“Mid-Term Performance Evaluation of 
USAID’s Bilateral VIH Program in 
Nicaragua”
Solicitation Number: (SOL-524-13-000006)
Component: Assessing PrevenSida Coverage and Quality 
(Q7, Q8)
This document was produced for review by the United States Agency for International Development. It was 
presented by consultant Carlos Hernandez on behalf of the team that also included consultants Yelba Jarquín and 
Francisco Ochoa and Maricela Larios under contract No. AID-524-O-13-00004.
Assessing PrevenSida Project Coverage - “2014 Mid-Term Performance Evaluation of HIV/USAID Nicaragua Program”
1
CONTENTS
GLOSSARY....................................................................................................................................................... 2 
EXECUTIVE SUMMARY..................................................................................................................................... 4 
I. BACKGROUND-RATIONALE....................................................................................................................... 6 
II. METHODOLOGY ....................................................................................................................................... 9 
2.1 Study Population: Universe and Sample ...................................................................................................... 9 
2.2 Pertinent and Relevant Information ........................................................................................................... 11 
2.3 Fieldwork .................................................................................................................................................... 13 
2.4 Quality Control…………………………………………………….…………………………………………………… ............................ 15 
2.5 Ethical Framework………………………………………………………………………………………………................................... 15 
2.6. Processing and Analysis.............................................................................................................................. 16 
III. RELEVANT COVERAGE FINDINGS............................................................................................................. 17 
3.1. Characterization of Studied MARPs ............................................................................................................ 17 
3.2. Total MARP Coverage................................................................................................................................. 22 
3.3 PrevenSida Global Coverage....................................................................................................................... 23 
3.4 PrevenSida Exclusive Coverage .................................................................................................................. 31 
IV. APPROXIMATIONS TO PREVENSIDA COVERAGE EFFECTIVENESS .............................................................. 38
4.1 PrevenSida/HIV Test Positivity ................................................................................................................... 38
4.2 New Infection Risk Index: PLACE ................................................................................................................ 40 
4.3 Consistent Condom Use ............................................................................................................................. 43 
4.4 History of Sexually Transmitted Infections................................................................................................. 46 
4.5 Effectiveness Measured by Number of Promotional Contacts.................................................................. 48
V. CONCLUSIONS........................................................................................................................................ 50 
VI. RECOMMENDATIONS ............................................................................................................................. 56 
VII. BIBLIOGRAPHY....................................................................................................................................... 58
VIII. ANNEXES 
Annex 1. MARP Exposure to Six Preventive Activities. PrevenSida Coverage
Annex 2. Basic Tabulation Tables. Descriptive Level
Annex 3. PrevenSida Municipal Prioritization – HIV Test Positivity
Annex 4. List of Venues by Region
Annex 5. Total Expected Samples by Venue
Assessing PrevenSida Project Coverage - “2014 Mid-Term Performance Evaluation of HIV/USAID Nicaragua Program”
2
GLOSSARY
ADESENI Asociación por los Derechos de la Diversidad Sexual Nicaragüense (Association forthe Rights of 
Sexual Diversity) 
AHCV Asociación de Hombres contra la Violencia (Association of Men against Violence) 
AIDS Acquired Immunodeficiency Syndrome
AMODISEC Asociación de Movimiento de la Diversidad Sexual Costeña (Association of the Coastal Sexual 
Diversity Movement) 
ANICP + VIDA Asociación Nicaragüense de Personas Positivas por la Vida (Nicaraguan Association of Positive 
Persons Fighting for Life) 
ASONVIHSIDA Asociación Nicaragüense de VIH/SIDA (Nicaraguan Association of People Living with HIV/AIDS) 
CEPRESI Centro para la Educación y Prevención del SIDA (Center for AIDS Education and Prevention) 
CI Confidence Index
CONISIDA Comisión Nicaragüense del SIDA (Nicaraguan AIDS Commission)
CP Combination Prevention
GAO Asociación Grupo de Autoayuda de Occidente (Self-Help Group Association in the West Region) 
GF Global Fund
GSB Gender-Sensitive Budgeting
HIV Human Immunodeficiency Virus
M&E Monitoring and Evaluation
MARP More-at-Risk Population
MOH Ministry of Health
MOSAFC Modelo de Salud Familiar y Comunitario (Family and Community Health Model) 
MOVITEP Movimiento de Teatro Popular sin Fronteras (Movement of Popular Theater without Borders) 
MSD Movement for Sexual Diversity
MSM Men who have sex with men
Assessing PrevenSida Project Coverage - “2014 Mid-Term Performance Evaluation of HIV/USAID Nicaragua Program”
3
NGO Nongovernmental organization
OR The ratio of odds of an event occurring
OVI Asociación Organización Vida Integral (Integral Life Organization) 
PLHIV Persons living with HIV
PLACE Priorities for Local AIDS Control Efforts
PPA Priority Prevention Area
PrevenSida Program for Preventing Transmission of HIV/AIDS among High Risk Populations in 
Nicaragua
RAAN North Atlantic Autonomous Region
RAAS South Atlantic Autonomous Region
STI Sexually Transmitted Infection
SW Sex Workers
TOR Terms of Reference
TRANS Transgender
UR Unique Register
USAID U.S. Agency for International Development
VCT Voluntary Counseling and Testing for HIV
WHO World Health Organization
Assessing PrevenSida Project Coverage - “2014 Mid-Term Performance Evaluation of HIV/USAID Nicaragua Program”
4
EXECUTIVE SUMMARY
The findings of the “Assessing PrevenSida Coverage” study, a component of the “Mid-Term Performance Evaluation 
of USAID’s HIV Program in Nicaragua,” are presented here. The study was conducted to determine project 
coverage and effectiveness, with a view at identifying factors that enable or hinder MARP access to preventive 
activities and their impact on HIV protection capacities. Additionally, the study was seen as an opportunity to apply 
the main elements of the PLACE approach and methodology proposed by MEASURE/USAID for local epidemic 
studies. The most relevant evidence and PLACE methodological validation will strengthen the project at its second 
implementation phase. 
The PrevenSida management team selected 29 Priority Prevention Areas (PPAs/PLACE) in five regions based on 
HIV positive rapid tests performed in the last three years, and other risk and accessibility criteria. A total of 1,393 
MARP interviews were conducted, of which 48% were MSM, 32% SW, and 20% TRANS. Fieldwork was organized 
in such a way that 46% of MARPs were convened by PrevenSida implementing organizations and 54% were 
contacted at their living and working places to control selection error bias.
Only 11% of MARPs did not receive any preventive activity in the past year; 89% received one or more preventive 
activities, broken down into 44% of MARPs covered by “other provider agents” and 45% of MARPs reached by 
some activity from the PrevenSida project (60% TRANS, 46% MSM, and 34% SW). This coverage includes 30% of 
MARPs under mixed coverage (PrevenSida and other provider agents) and the remaining 15% receives preventive 
services exclusively from PrevenSida (17% MSM, 16% TRANS, and 11% SW). Four PrevenSida preventive activities 
have been included in this coverage, namely: promotional contacts covering 26% of MARPs (65% of total coverage), 
training events covering 25% of MARPs (49% of total coverage), peer socialization groups accounting for 15% of 
MARPs (41% of total coverage), and VCT accounting for 16% of MARPs (66% of total coverage).
Evidence shows PrevenSida global and exclusive coverage stems from combination care by implementing 
organizations to the social fabric of MARPs in their immediate environment and the effects of a “pickup” venue 
mapping developed by PrevenSida since 2013. This translates into a higher probability of access to PrevenSida 
services by MARPs claiming to participate in some organization and by those meeting “pickups” at bars and 
discotheques. The prevalence of this logic determines an inverse relationship between expected coverage and 
municipal HIV test positivity rates, since the largest coverage is found in municipalities with the lowest positivity 
rates, and vice versa. This means that the territorial risk criterion (PPA) has not been built in PrevenSida planning. 
While this condition has ensured better access to MARPs with higher HIV risk occupations, these modes of 
encounter are strongly conditioned by a wide range of factors facilitating or hindering access, i.e., higher schooling 
facilitates access, whereas access is hindered by having a school education below 6th grade; not having children 
facilitates access, whereas having them is a limitation; belonging to the unemployed non-student MARP segment is a 
facilitating factor, whereas working in places other than bars, discotheques, and motels is a strong coverage 
limitation; living in the capital city or in a departmental capital facilitates access, as opposed to living in peripheral 
municipalities; looking for “pickups” at bars and discotheques increases accessibility, whereas seeking them at 
commercial sites significantly reduces accessibility. These and other coverage-conditioning factors evidenced in this 
study make it possible to build collective reflection and dialogue to reach a consensus on actions for change.
The study provides evidence of PrevenSida coverage effects on result indicators, particularly those dealing with
behavioral changes for protection, such as consistently using condoms and reducing the number of sexual partners 
and the number of new sexual partners in the last month. In this behavioral modification, varying degrees of 
association are identified with PrevenSida global and exclusive coverage, and with specific activities, particularly
Assessing PrevenSida Project Coverage - “2014 Mid-Term Performance Evaluation of HIV/USAID Nicaragua Program”
5
those fostering more interaction and dialogue on account of their nature, such as training events, peer group 
socialization, and VCT with its counseling component. In these specific activities significant differences were found 
in terms of a higher PrevenSida impact, as compared to other preventive service providers.
It must be mentioned, however, that the high risk indicator used in the PLACE approach for persons with two or 
more sexual partners and at least one new partner in the last month is not only too broad (78% of total MARPs) 
but also requires longer and deeper exposure for modification. Nonetheless, exposure to training and peer groups 
shows a reduction, regardless of the preventive service provider. 
The study validates the main PLACE approach methodological components, inasmuch as they are crucial 
epidemiological elements to address the epidemic at the local level and provide key inputs to project and program 
management decisions. PrevenSida has already started applying some of these elements, particularly for identifying
local scenarios where “new pickups” can be met, and found that in addition to their contribution to MARP 
coverage and accessibility, there were also some limitations in the narrow “mapping” focus on commercial and 
population concentration areas where 36% of MARPs seek “pickups”. The importance of such elements as a PPA 
definition based on easy-to-process information like HIV test positivity, rating MARP individual risks, including the 
“new partner” concept, and improving such valuable tools like “mapping”, could strengthen the capacities of the 
PrevenSida network and the project as a whole.
The study validated the inclusion of local and national scenario structural elements, such as the incidence of 
tourism, population mobility, and high concentration of workers in enclave economies, identifying them at both the 
individual level (each MARP’s relationship by virtue of its work) and the aggregate level of influence on the 
economy and social life of PPAs targeted by the study. Evidence was found of their influence on both PrevenSida 
coverage and modification of their impacts on outcome indicators. The study identified their importance to the HIV 
epidemic and the need for incorporating these elements in local intervention analyses and strategies, including 
targeting actions and partnerships.
The number of interpersonal contacts per MARP was found to have a limited impact on outcome indicators and a
strong impact on increased coverage of other preventive activities, evidencing its promotional and motivating role. 
While this ratifies its importance and role in the project’s planning model, it also points to the need for its insertion 
in the combination prevention model.
The research team highlights the usefulness of this report as educational material to encourage collective reflection 
and dialogue at different project and program levels, in order to strengthen them. 
Assessing PrevenSida Project Coverage - “2014 Mid-Term Performance Evaluation of HIV/USAID Nicaragua Program”
6
I. BACKGROUND-RATIONALE
The first phase of the evaluation process of the USAID/HIV program in Nicaragua has been developed, as stated in 
Solicitation SOL-524-13-000006, covering the performance of the program as a whole. TORs included a second 
phase aimed at answering two additional questions as the basis for recommending adjustments to PEPFAR1 Central 
America Partnership Framework in the remaining years (2010-2015): 
Q7: What is the coverage of HIV prevention actions implemented by the PrevenSida project in each key 
population? 
Q8: What activities are deemed to be successful or not successful in these populations, and why? How can 
these outcomes be used in planning future programs? 
PrevenSida is the specific project that develops national policies and standards in the proposed combination 
prevention model2 targeted on MARPs and other vulnerable populations. It operates through a civil society 
organization network that has to address the wide accessibility gaps previously found in these populations by 
ensuring delivery of a minimum package of (behavioral, biomedical, and structural) care for each person, in order to 
increase their capacity to control HIV infection risk. 
The USAID/HIV program has strengthened the capacities of the health system to comprehensively address the 
epidemic by developing novel proposals that have materialized into a body of regulatory, methodological, 
pedagogical, and monitoring tools, as well as sustainable processes seen from the perspective of the different 
prevention levels and coordination of the different subsystems and social actors in the comprehensive epidemic 
approach. 
This study is assumed as the second phase of the same process, which seeks to delve deeper into some analytical 
elements, focusing on the PrevenSida Project as its main strategic axis. To the extent that it closes existing 
accessibility gaps for key populations in the prevention area, it continuously shapes and validates in practice a 
preventive care model that requires support from all program components and provides essential inputs to the 
components for achieving their own goals.
The program performance evaluation3focused on the compliance of a set of indicators established for the four 
programmatic components, which in turn have translated into significant contributions to the overall HIV epidemic
approach in Nicaragua. The wide diversity of achievements, particularly those directly related to the preventive 
care model of key populations, which has been evolving from a clear exclusion of MARPs (MSM, TRANS, sexual 
workers) in the context of national policies promoting a Family and Community Health Model4 (MOSAFC): 
• MARP coverage has been attained through interpersonal communication activities on prevention, as a 
complement to the other initiatives undertaken in the national response (GF). Based on the MARP population 
estimated and approved by CONISIDA for 2013, PrevenSida has reached an estimated 42% of sex workers, 40% 
of MSM, and 43% of the national TRANS population. In turn, this coverage shows significant progress has been 
made in terms of reducing stigma and distrust as accessibility barriers, while remarkable increases have been 
achieved in the three MARP groups that have had two or more contacts, revealing the development of 
communication methodologies and other central elements of the preventive care model, such as mapping pickup 
venues and building local networks promoted and led by the groups themselves.
Assessing PrevenSida Project Coverage - “2014 Mid-Term Performance Evaluation of HIV/USAID Nicaragua Program”
7
• Currently, PrevenSida has a network of civil society organizations (MARP and HIV associations, local and 
national NGOs, businesses, universities) that grows every year and engages with other national response 
initiative networks. This interaction is meant to share a unified institutional strengthening strategy to constantly 
build management, administrative, and monitoring and evaluation capacities based on a Unified Register5 of 
people served and services rendered.
• The program evaluation also showed that combination prevention action coverage makes a difference in HIV 
seroprevalence probabilities by reducing reaction to rapid tests, in comparison with populations that had not 
been exposed to preventive actions (OR 0.55, CI 95% 0.3 – 0.9). This relationship, however, was only found in 
NGOs that had been working in the PrevenSida network for over a year. 
Evidence gathered during the program performance evaluation has contributed to highlight the meaning of the two 
questions to be answered in its entire dimension by seeking to explain whether the effectiveness criterion of the 
actions could be associated to a stronger link with MARP grassroots in line with the implementation of strategies 
more adjusted to the specificity of each site. 
In this context, the reference framework becomes a suitable scenario to implement the PLACE method developed 
by the USAID-sponsored MEASURE evaluation6 through a methodological systematization of the approximation 
process steps for identifying local networks used by MARPs to find new sexual partners, which are the main 
scenario for the transmission of new infections according to available scientific evidence. 
The PLACE method focuses on venues where MARPs usually gather in search of new partners, either “commercial 
sexual clients” or just new “pickups”, and emphasizes the “network” nature of these sites inasmuch as they 
encompass the entire process of making contact, encouraging a mutual interest and consolidating the new liaison
until finally a sexual encounter is agreed. These can include outdoor spaces (parks, streets, docks, etc.), restaurants, 
bars, discotheques, and motels. This “new sexual encounter network” notion implicitly entails a territorial concept 
because these sites are usually located in close proximity to each other and are related by communication and 
transportation mechanisms. This territorial notion is called “VENUES” (translation by PLACE).
It should be noted that the configuration of these venues with an appropriate set of establishments makes them 
predominantly urban, although they are created and consolidated away from residential areas on account of their 
“pickup venue” nature. The factors that determine where a “pickup venue” is created and developed in a city or 
geographic location in Nicaragua are highly varied, but they always follow high concentrations of people, whether 
or not permanent, such as:
• Recreational areas where people are present with no specific origin and which are largely located according to 
urban development and commercial criteria.
• Industrial zones or production enclaves (off-shore assembly, mines, seasonal harvests, etc.) with a high 
concentration of permanent or seasonal workers. 
• Permanent or seasonal tourist destinations with a strong presence of national or foreign tourists. 
• Cargo or passenger transport network nodes, such as bus terminals, borders, ports, and rest areas for 
international cargo transporters, which create concentrations of transport workers or users.
• Defined commercial sex areas with their own clients (bars, brothels, movie theaters, etc.)
While general elements with similar characteristics are found in the various types of “pickup venues”, it is also clear 
that each venue exhibits unique features derived from the ecological, climate, demographic, social and cultural 
conditions of each location. This leads to one fundamental characteristic of the PLACE method, which consists of 
Assessing PrevenSida Project Coverage - “2014 Mid-Term Performance Evaluation of HIV/USAID Nicaragua Program”
8
methodologically recognizing and addressing the fact that national or regional epidemics are actually a combination 
of local epidemics, each having a unique nature contingent on how the HIV epidemic transmission chains are built at
each particular location. Thus, the method is used in studying transmission networks to characterize local 
epidemics, in order to design or tailor locally-defined intervention strategies, although it also allows to make more 
general inferences to be used as inputs for broader policies and strategies.
By and large, the epidemiological analysis being carried out enables a broad diversification of the benefits of the 
PLACE method, even more so when it goes beyond sporadic studies to become an ongoing local capacity-building 
process to follow and use it. At national level, it also builds capacities for consolidating baseline studies to produce 
monitoring and dialogue inputs for public policy. 
The potential benefits of the “PLACE” method could be summarized as follows: 
Table 1: PLACE Methodology Summary
RELEVANT PROCESSES “PLACE” METHOD OUTPUTS
Micro localized incidence surveillance —Periodic indicators
—Network maps
Behavior monitoring and coverage of 
the program 
—Indicators, maps
—MARP and network behavior
—Coverage of actions in MARPs and networks
Differentiated intervention design —Contents and methodologies
—Service delivery modes (interpersonal 
communication, peers, exchanges, etc.)
—Building support networks
Program scaling —Model and methodology validation
—Process systematization implementation
—Lessons learned
Community mobilization —Motivates and stimulates encounters. 
CS/Inst
—New project inputs and support
Evidence as a powerful educational
tool
—Encourages reflection, dialogue, and 
agreement
—Exchange for building broader networks
—Public policy inputs
Assessing PrevenSida Project Coverage - “2014 Mid-Term Performance Evaluation of HIV/USAID Nicaragua Program”
9
II. METHODOLOGY 
In order to answer the two questions set forth in the TORs within the 
context of the PLACE method application, the general objective of this 
study was defined as:
“Assessing PrevenSida/USAID project coverage and effectiveness 
in the different contexts where it operates, in order to provide 
inputs that contribute to tailor local HIV preventive intervention 
strategies in MARPs.”
To this end, it became necessary to address complexity in different 
levels of analysis, which was translated into specific objectives, 
• Characterizing MARPs according to types of venues identified as 
HIV transmission network scenarios. 
• Characterizing MARPs covered by PrevenSida actions.
• Identifying a series of factors associated to populations and social 
scenarios influencing effectiveness of PrevenSida preventive action 
coverage.
To achieve these objectives, the PLACE method will require a single 
database containing data on venue characterization and MARP 
population in each venue. 
These outlined objectives made it easier to highlight the main findings 
that were deemed significant in the analysis for promoting a collective 
dialogue among different social actors involved in the national 
response, which in addition to making progress in the interpretative 
models of the findings also incorporates consultation and consensus￾building processes for adjusting HIV prevention strategies in Nicaragua.
2.1 Study Population: Universe and Sample
In line with national policies, PrevenSida prioritizes preventive care for 
men who have sex with men (MSM), transgenders (TRANS) and sex 
workers (SW), and has been expanding its coverage in 11 departments 
in the country8, covering 20,869 MSM, 7,269 TRANS and 2,344 SW in 
2013. For 2014, an expansion of the NGO network, as well as a wider 
population and territory coverage, has been planned and implemented.
In this framework, the study adopted a selection process based on a 
stratification by clusters according to the PLACE method, resulting in 
the selection of Priority Prevention Areas (PPAs), ranked as having the 
highest likelihood of HIV transmission on account of their 
PrevenSida Goal
Increasing healthy behaviors in higher risk people to 
decrease HIV/AIDS transmission through condom use, 
reduced number of sexual partners, and increased 
access to HIV counseling and promotion of testing.
PrevenSida Impact Indicators
• A 50% increase (from baseline) in consistent 
condom use in all sexual contacts, including stable 
partners. 
• A 30% decrease (from baseline) in the number of 
multiple partners among high risk population.
• A 60% increase (from baseline) in the use of 
counseling and test promotion. 
PrevenSida is an innovative project of the United States 
Agency for International Development (USAID) for 
HIV/AIDS transmission prevention in high risk 
populations. It will be developed in five years (2010-
2015) through a 5 million dollar investment. 
The program is managed by University Research Co., 
LLC (URC)7 under cooperative agreement number AID￾524-A-10-00003.
Table 2
Municipality Prioritization. 
PrevenSida Proposal
Region
Incidence (%) in 2011-2013 
No 
Cases
0.01% 
- 1 
Case
0.02% 
to 
0.03%
0.04% 
or 
more
Managua 1 3 4 1
Western 8 6 8 2
SouthernPacific 14 9 4 4
Central 35 15 12 5
Caribbean 7 7 5 7
Total 
Municipalities
65 40 33 19
PLACE Sample 4 4 9 8
Assessing PrevenSida Project Coverage - “2014 Mid-Term Performance Evaluation of HIV/USAID Nicaragua Program”
10
characteristic typologies and being representative of those specific 
territories covered by this study. 
The PPA selection process was based on two major stratification levels
(region and municipality) and the following main criteria:
 HIV infection prevalence and incidence, applying the methodology 
principle of the PLACE approach for tracing infection transmission 
to define PPAs. This criterion was contributed by the HIV/USAID 
Program and PrevenSida based on the review of the reactivity of all 
rapid tests performed in the last three years, which led to four 
levels of positivity (reactive tests).
 The incidence of major underlying factors (tourism/recreation, 
production enclaves, transport network),to the extent that they 
determine MARP concentration venues and new sexual encounter 
networks. 
 PrevenSida coverage. To the extent that its evaluation is the 
purpose of this study and determines the level of access to MARP 
clusters in each selected PPA. In this regard, the application of the 
efficiency principle was anticipated because the selected PPA must 
ensure a minimum MARP to justify mobilizing resources thereto. 
To establish the starting point, the study has used information 
provided by national response governing institutions, defining five 
representative regions with different characteristics making up the 
main risk areas at the national level, in terms of detected case 
prevalence. 
A national risk mapping clearly identifies three regions with two risk 
levels: the Pacific and Caribbean regions with a high risk, and the 
Central region with a predominant medium and low risk.9
The Department of Managua is different due to its political, 
administrative and demographic characteristics, which condition the 
highest concentrationof MARPs and potential “VENUES” to be 
selected as PPAs in the study. 
Consequently, five regions were defined as a first stratification level, 
which delimits sample representativeness and therefore should 
concentrate sufficient population to support a stratified analysis.
The selection process was complemented with the identification of 
venues at the municipal level. To this end, the PrevenSida network was 
used as a starting point because it has a presence in over 60 
municipalities, in the majority of which (44) HIV rapid tests are 
performed. The selection process was carried out by sequentially
Table 3
Estimated Population 2013 – CONISIDA Data
Region MSM TRANS SW Total
Managua 16,127 6,121 1,296 23,544
West 9,224 2,281 1,043 12,548
South Pacific 8,804 2,507 1,369 12,680
Central 10,257 3,181 1,217 14,655
Caribbean 7,677 3,149 485 11,311
Nicaragua 52,090 17,239 5,410 74,739
Table 4
Served by PrevenSida 2013 – MARP UR*
Region MSM SW TRANS Total
Managua 8,954 3,736 752 13,442
Western 5,100 1,610 990 7.700
South 
Pacific
2,650 2,763 363 3,490
Central 1,085 296 196 1,577
Caribbean 3,080 1,150 42 4,272
Nicaragua 20,869 7,269 2,343 30,481
*Unique Register
Assessing PrevenSida Project Coverage - “2014 Mid-Term Performance Evaluation of HIV/USAID Nicaragua Program”
11
applying the three criteria agreed between the research team and 
PrevenSida coordinating team. 
The first selection criterion was the incidence detected by PrevenSida 
(reactive rapid tests) in four years of work. Thus, 15 municipalities 
emerged where at least one HIV reactive test had been detected, all of 
which were included in the sample, regardless of how the other 
criteria will be combined. 
The second criterion applied was the actual coverage of PrevenSida 
based on specific knowledge on “VENUES” in terms of MARP density
and accessibility during fieldwork. The main information was provided 
by the PrevenSida team, which has a mapping10 of the most frequently 
used venues in each municipality for “new pickups”, as well as the 
knowledge and linkages of each organization. Based on the jointly 
formulated proposal, NGO managers and officials in the preselected 
municipalities were consulted, whose cumulative knowledge and 
experience quality them to provide an “expert opinion” in relation to
the preselected PPAs. Minor adjustments were made to the initial list.
The last criterion was based on the incidence of structural (underlying) 
factors, also applied in consultation with “expert” organization staff, 
whose knowledge enabled to rank incidence and infection probabilities 
in tourism, transport network, and labor enclaves, which served as the 
basis for final adjustments to the share of each MARP type. 
Based on available information, such as the official MARP population 
estimates at the departmental level and PrevenSida coverage, 25 PPAs 
were selected (one per municipality, 4–5 in the Managua municipality),
accounting for an estimated sample of 1,600 MARPs to be interviewed.
2.2 Pertinent and Relevant Information 
In accordance with the guideline questions and objectives of the study, 
instruments were developed to gather information relevant to the 
characterization of the coverage status of each MARP, as well as for 
the characterization of each site to be measured. 
In the characterization of the SITE, a guide was applied by field team 
coordinators, together with the staff of NGOs that belong to the 
PrevenSida NETWORK and have a presence at the site.
Significant progress has been made to the extent that most PrevenSida 
network organizations have mapped new pickup “nodes” with very 
relevant basic information. It can be affirmed that the “PLACE” 
Table 5
PrevenSida Coverage 2013
MARP 
Type
Official 
CONISIDA 
Data
MARP 
UR 
2013
PrevenSida 
Coverage
MSM 52,090 20,869 40%
TRANS 17,239 7,269 42%
SW 5,410 2,343 43%
Total MARP 74,739 30,481 41%
Communication 
Actions
PrevenSida / “Others” 
Yes/No
Contact
Peer Group
Training
VCT
MARP
% Exposure • Coverage
• MSM, SW, TRANS
• Regions
• Occupation
Modify Access
Determinant 3
i.e. social networks, 
support networks
Determinant 1
i.e. typology of 
households, size, 
household head
Determinant 2
i.e. schooling, gender 
roles, occupation
Determinant 4
i.e. production model, 
macro tourism, 
mobility, enclaves
Determinant 5
i.e. dynamic of 
networks for new 
encounters
Determinant 7
i.e. State tutelage and 
protection
Determinant 6
i.e. meso variables, 
other agents, E&D 
setting
MARP 
Protection
Yes/No • First sexual 
relationship
• Number of partners
• New partners
• Condom use
• HIV test
Communication 
Actions
PrevenSida / “Others” 
Yes/No
Contact
Peer Group
Training
VCT
Modify Effect
Determinant 3
i.e. social and 
support networks
Determinant 1
i.e. typology of 
households, size, 
household head
Determinant 2
i.e. schooling, gender 
roles, occupation
Determinant 4
i.e. production model, 
macro tourism 
incidence, mobility, 
enclaves
Determinant 5
i.e. dynamic of 
networks for new 
encounters
Determinant 7
i.e. State tutelage and 
protection
Determinant 6
i.e. meso variables, 
other agents, E&D 
setting
Assessing PrevenSida Project Coverage - “2014 Mid-Term Performance Evaluation of HIV/USAID Nicaragua Program”
12
methodology, focused on characterizing “pickup” establishments and 
venues has been highly applied in its more specific and local dimension.
A questionnaire for individual dialogue interviews targeted to MARPs 
was developed as the main methodological tool of this study since the 
most trustworthy source in both coverage and its effects on 
protection capacities is the own “VOICE” of MARPs that could be 
potentially exposed to the actions of the PrevenSida network. 
Given the evaluative bias of the study raised in the guideline questions 
and objectives, the methodological tool is mostly targeted on: 
• Assessing coverage achieved through interpersonal 
communication actions in the MARP universe.
• Assessing coverage effectiveness of actions by their impact on 
MARP protective capacities.
In compliance with these two objectives, the analysis identified factors 
conditioning access and thus modifying coverage. Likewise, the 
incidence of these factors on the achievement of expected results was 
analyzed, that is to say, which modify effectiveness of exposure to the 
actions through their association with changes in MARP protective 
capacities, i.e., their potential impact.
This methodological guideline led to the design of a dialogue interview 
questionnaire, including questions to enable the construction of three 
types of indicators:
1. Exposure to preventive services. Coverage
These indicators show the share of MARPs that have received 
some of the studied preventive services, including 6 activities. 
Changes or differences in MARP stratification variables are 
analyzed for each activity. 
Based on the nominal identification of the organization providing 
or promoting each service provided by the interviewed MARPs, 
we were able to differentiate the coverage provided by PrevenSida 
from that offered by “Other Providers”, so as to enable a contrast 
analysis between exposed and non-exposed MARPs, and between 
MARPs served by PrevenSida and those served by “Others”. A 
confluence indicator was developed for the 4 activities more 
directly related to the combination prevention model, which was 
used to develop the global coverage indicator analyzed using the 
same contrasts. 
Table 6
PrevenSida Network Organizations. 
UR-MARP 2014
ACCCS CREPRESI IDSDH
ADESENI CEPS Ingenio San 
Antonio
ADISNIC COSEP IXCHEN
AGENTES DE 
CAMBIO PADCANIC MDS RAAS
AMODISECRAAN PSL MOJUDS
ANCP¿VIDA GAO NAVINIC
ASONVIHSIDA GAVIOTA OVI
CASUR HIJAS DE LA 
LUNA PrevenSida
CBGODEM ICAS RDS
Result Indicators:
Protective Behaviors
 Age at first sexual relation
 Number of sex partners
 Number of new sex partners
 New infection risk index. “Place”
 STI Background
 Use of condom
 Current HIV test
Exposure to Interventions
MARPs WITH/WITHOUT COVERAGE
PrevenSida/Other Providers
 Prevention Messages
 Personal Contact 
(promotion/motivation)
 Training events
 Reflection and dialogue in peer 
groups
 Voluntary counseling and 
testing (VCT)
Assessing PrevenSida Project Coverage - “2014 Mid-Term Performance Evaluation of HIV/USAID Nicaragua Program”
13
2. Protection against HIV. Outcomes.
Indicators targeted to measure whether or not protective 
practices are used to prevent HIV infection, in order to detect 
changes or differences that could be attributable to exposure to 
preventive actions provided by the PrevenSida network or “other 
provider agents”. 
The inclusion of the specific contribution of the PLACE approach 
regarding new sexual partners is worth mentioning, as well as a 
new infection risk index that combines the number of new sexual 
partners and the number of total sexual partners each person has 
had in the last month (high risk) and year (moderate risk). A low 
risk refers to stable and monogamous couples based on fidelity, or 
persons with no sexual partners in the last year.
3. Conditioning factors. Effect modifiers
These are particular characteristics of MARPs, their families, or 
place of residence or location that could have a bearing on the 
probabilities of accessing preventive services, or affect protection
capacities.
Age, schooling, type of work, participation in organizations, 
habitual “pickup” venues, sexual initiation conditions, having 
children, and living with their partners, among other things, are 
MARP characteristics that could facilitate or hinder access to 
prevention services and thus modify coverage, as well as household 
characteristics, i.e., head of household, economic contributions, or 
characteristics of the site, such as the incidence of a production 
model, tourism, transport or type of urban center where site is 
located.
Whenever a significant basic association was found in evaluating 
coverage or action effectiveness, a review was carried out to 
determine whether any of the characteristics of the person, family, 
or site were acting as an effect modifier, in order to provide 
evidence about the conditions in which PrevenSida is linked to 
MARPs or the effects generated by such link. 
2.3 Fieldwork
Based on a consensus reached with the PrevenSida team in relation to 
the selection of venues to be measured and the MARP questionnaire, 
three field teams of four people were formed, namely, three MARP 
peer interviewers(MSM, TRANS, SW) and one coordinator/supervisor.
Field staff was trained in a two-and-a-half-day workshop (July 10-12, 
2014) covering an 8-item agenda (see attached table).
PLACE Risk Index
Focusing on people with 
High rates of new partners
HIGH
MODERATE
LOW
Effect Modifiers
Personal, Family and Site
Cluster variables at the site
Tourism or mobile population incidence; 
recreation or commerce places
Person: Age, sex, sexual identity, schooling, 
occupation, has a partner, lives with partner
Household type: Own/other head of household, 
children, economic contribution
Trust in services/social support networks
Recognizes unmet rights 
 
Last Month
 Some new relation
 Multiple partners
Last Year
 Some new relation
 Multiple partners
Last Year
 Have not had sex
 An “old” partner
Assessing PrevenSida Project Coverage - “2014 Mid-Term Performance Evaluation of HIV/USAID Nicaragua Program”
14
The training workshop ended with the validation of the MARP 
questionnaire through interviews with members of the three 
population groups invited or contacted in habitual concentration 
venues. A collective evaluation session of the validation was held on 
the basis of which final adjustments were made to the questionnaire. 
Fieldwork began when the three supervisors got in touch with, or paid 
an on-site visit to, all PrevenSida network organizations serving the 
selected municipalities, for the purpose of getting their “expert 
opinion” on MARP accessibility conditions in each municipality in order 
to adjust the sample and define MARP contact strategies (call, 
staggered group concentration, contact in venue network nodes, etc.). 
Strategies were balanced to reduce and prevent as much as possible 
any “selection bias” that could arise if only one approach was used, 
which would result in an overrepresentation of MARPs close to 
PrevenSida.
Three routes were established in accordance with proximity and 
channels of communication. The planned route made it possible to
develop, in consensus with local NGOs, an arrival schedule at each 
site, and work agendas were determined by contact strategies. This 
preparatory work was critical to ensure fieldwork efficiency in terms 
of population goals and the highest potential efficiency since at least 
one work day (8–12 hours) was scheduled for each site, which was 
extended in some places according to conditions of MARP share in 
each municipality. 
Fieldwork started on July 14th and ended on August 6th. On the first 
day, the three teams worked at a single venue in the city of Managua, 
in order to encourage an exchange of experiences and consolidate the 
introduction and greeting, build trust with the interviewed MARP, 
manage the questionnaire, dialogue interview techniques and recording
quality. At the end of the first day, quality control guidelines and timing 
were strengthened by the coordinating team.
On the second day of fieldwork, teams began their own route plan, 
getting in touch with relevant NGO contacts at each site two days 
prior to the scheduled date, in order to confirm agendas and MARP 
shares, as previously agreed. 
In most measurement sites, an adequate call and collaboration was 
found in relation to concentration and pickup venues, in order to find 
and interview MARPs, albeit the scheduled share could not always be 
completed. We highlight the participation of many PrevenSida network 
NGOs that made valuable contributions through their connections, 
knowledge, and venue mapping. Major difficulties were found in some 
Training Workshop Agenda
Three field teams - 2½ days
a. Contextualization and study objectives
b. “Place” approach and method
c. Study framework and ethical principles
d.Questionnaire review
e.Dialogue interview methods and 
techniques
f. Registration quality
g.Quality control levels and processes
h.Questionnaire validation
Contact Modalities
Balance avoids “selection bias”
 Call
(NGOs – Lead associations of the groups)
 Snowball...
First ones invite others
 Search in concentration sites
(Town squares, shopping malls and 
recreational areas)
 Search in “pickup” venues
(Bars, discotheques, motels)
Assessing PrevenSida Project Coverage - “2014 Mid-Term Performance Evaluation of HIV/USAID Nicaragua Program”
15
venues because the agreed call had not been issued despite prior 
scheduling and confirmation before the team’s arrival and, in some 
cases,it was not possible to contact the reference persons. When the 
scheduled MARP share could not be reached, the team scheduled 
another visit for the next following days, according to the route plan.
A sample of 1,393 MARP members was gathered, accounting for 
approximately 87% of the scheduled sample. The sample maintains a
desirable internal balance between MARP members who were called 
by the PrevenSida network (47%) and those contacted on the streets 
and shopping centers or “pickup” venues (54%), which remains the 
same for the three populations. As seen in the analysis, it validates 
intrinsic coverage rates.
2.4 Quality Control
Quality control occurred at four points. Firstly, it took place at the time 
of the interviews through the team coordinator’s direct observation 
and periodic and crossed-checked interview reviews. Secondly, the 
supervisor reviewed the set of interviews conducted by each 
interviewer before certifying, signing and delivering them to the data 
entry process.The third quality control takes place when the data of 
each interview is entered, based on the experience of the data entry 
clerk in applying the “right” parameters, which are then fed back to 
supervisors to correct duplication errors along the way. Finally, each 
interview is typed by two different data entry clerks, and the two 
databases are compared automatically to identify disparities and 
correct errors. The combination of four control points ensures that 
the database accurately contains the VOICE of MARPs.
A team made up of six data entry clerks and one coordinator was 
responsible for designing a capture screen, including controls to 
minimize errors and speed up data entry. Data entry clerks were 
organized in three pairs to perform the aforementioned dual data 
entry. As a result, 2,786 interviews were digitized to build a debugged 
2,393-record database. The data entry process included mesovariable 
data from venue level forms, in order to enrich the analysis. 
2.5 Ethical Framework
Since the study addresses issues usually highly sensitive to people’s 
privacy, in addition to the marked stigma and discrimination suffered by 
most key groups, special attention was given to ethical aspects in the 
entire research in the design of the protocol, questionnaires and 
fieldwork organization. 
Quality Control
Entire process. Four points
Around the time of MARP interview
Certification by field supervisor
Evaluation of “right” parameters by 
entry data clerk
Dual data entry
Principle of Parsimony
“No more than the minimum needed”
The questionnaire is very simple and the 
number of questions is limited. It consists of 
one page, on both sides, and did not take 
more than 15 minutes of the informant´s 
time. It is a norm that also assumes the 
privacy of the interview out of respect for 
people because of the intimate topic being 
explored and the time of the people 
interviewed because they must consent to 
interrupt the activity in which they are 
engaged.
Assessing PrevenSida Project Coverage - “2014 Mid-Term Performance Evaluation of HIV/USAID Nicaragua Program”
16
Given the type of study and the characteristics of the population to be 
interviewed, several methodological principles, including respect, 
voluntariness, parsimony and immediacy, were applied in the design of 
the questionnaire, to ensure instrument suitability, as well as trust, 
freedom and voluntariness amongst interviewed MARPs. 
As mentioned already, the ethical framework was significantly relevant 
in the interviewer and field supervisor training process, emphasizing 
ethical principles and respect for the rights of people in general, 
particularly those with HIV/AIDS. A reflection on how dialogue
interview methods and techniques embody key ethical principles, i.e., 
confidentiality, respect of people’s dignity, and the right of sexual 
diversity and to express an opinion, aroused special interest. Training 
included examples of attitudes, visual or body behaviors, ways of 
talking or asking questions, and other details that might be construed 
as disrespect.
Emphasis was placed on principles of confidentiality and voluntariness 
to participate in the study. Free and informed participation was 
ensured by greeting the interviewed persons and making a brief and 
clear presentation of the purposes of the study, and explicitly making it 
clear that they had the right to refuse to answer any question or stop 
the interview. 
Respondent names were not recorded to protect confidentiality, and 
an independent code was developed for the identity of the respondent. 
Proper safeguard of paper interviews and database was ensured for the 
information provided. 
2.6 Processing and Analysis
Based on primary data, went on to build indicators, which were used to conduct three analysis levels: 
• Descriptive, using contingency tables for the three indicator groups (exposure, results, conditioning factors) by 
MARP group and territory, expressed as simple frequencies (%, rates) or averages or mode values. Tables were 
prepared, which serve as a preliminary report to be used as a basis for planning risk analysis and stratification.
• Main association, comparing coverage indicators to their conditioning factors, or contrasting results between
people covered and not covered by PrevenSida project actions. In conducting this analysis, the Mantel-Haenszel 
procedure, in combination with Cornfield& Miettinen11 (Cornfield 1956) confidence intervals, were used. 
Significant associations are reported as raw or weighted odds ratio (OR), with 95% confidence level significance, 
confirmed by the Chi square (X2) test. 
• Stratification. Major and significant associations will be subject to the sequential significance and stratification 
test (Mantel, Haenszel 1958 and Mantel 1963), and to the Woolf heterogeneity test across strata (Woolf 1955), 
in order to detect effect modifiers or co-variants with a significant incidence on study assessment and 
management goals. Significant heterogeneities are found when tests show values of X2het. 
This report is prepared for the purpose of presenting the main findings, without much interpretation. 
Principle of Immediacy
Interview as a “colloquial dialogue”
The semantics of the questions sought to build 
trust and achieve “immediacy”, i.e. easily 
understood questions and immediate answers. It 
is a moment for measuring aspects that are not 
part of a usual conversation and compromise 
many beliefs and taboos, so when the answers are 
more spontaneous (less reasoned), they are more 
respectful and reliable.
Principle of Respect and Voluntariness
No pressure exerted or induced answers
The formulation of the questions respects the 
intimacy and dignity of the informants. Most of 
the questions seek a YES/NO answer, which 
sometimes can be maybe or I don´t know.
Open questions are preferred, so as not to induce 
preconceived answers.
Assessing PrevenSida Project Coverage - “2014 Mid-Term Performance Evaluation of HIV/USAID Nicaragua Program”
17
3. MAJOR COVERAGE FINDINGS
Major findings are presented in relation to set objectives, focusing attention on an analysis of preventive action 
coverage and PrevenSida project’s contribution to said coverage. Subsequently, the evidence found in reference to 
the effectiveness attributable to MARP exposure to general preventive actions is shown, specifically those derived 
from the dynamics of the PrevenSida network. 
3.1 MARP Characterization
3.1.1 Age
More than two thirds of the population sample (68%) is made up of 
people under 30 years of age, mostly MSM (76%) and TRANS (78%). In 
the case of SW, the sample is equally distributed across this age 
bracket (50%) at the expense of having the smallest share in the less￾than-20-year-old group (7%), whereas this age bracket accounts for 
one fifth and one fourth in the other two groups (20% MSM, 25% 
TRANS). The older-than-40-year-old age bracket is larger in SW (15%) 
and less significant in the other two groups (8% MSM and 4% TRANS).
[Table 1] 
3.1.2 Schooling and Study
Over half of MARPs (52%) have a high school level education, making it 
the majority segment in the three groups; about half of SW, 
however,completed6th grade or lesswith43%in the 1st–6th grade bracket 
and 8% with no schooling at all. More than one third (36%) of MSM 
have received technical or higher education, whereas TRANS only 
reached 17% in this segment, and only 5% of SW reached this level.
No significant differences are evident in regional schooling distribution 
amongst MSMs, even in the higher schooling segment, which accounts 
for 49% and 38%in the Central and Caribbean region, respectively. In 
the case of TRANS, Managua and the South Pacific region show the 
lowest technical and higher education rates (12% and 6%), in contrast 
with26% and 33% in the Central and Caribbean region, respectively. As
regards SWs, the Caribbean region stands out. Despite exhibiting a 
large segment without schooling (10%), it also shows the largest 
segment with higher schooling (11%). [Tables 7,8,9] 
One third of MARPs (32%) are studying. The highest rates are found in
the Central (40%), Caribbean (39%), and West region (34%), whereas 
the lowest rates are found in the South Pacific region (29%) and 
Managua (23%). MSM is the group with the highest education rate 
(43%), of which 63% and 75% are studying in the Central and 
Caribbean regions. One third of TRANS (32%) are studying. The 
Caribbean and Central regions stand out with 91% and 52% rates, 
respectively. Only 15% of SW are studying, but Managua and the 
63%
27% 24%
10%
42%
17%
Under 25 years Over 25 years
MSM SW TRANS
Graph 4 
MARPs who Study, by Age
Assessing PrevenSida Project Coverage - “2014 Mid-Term Performance Evaluation of HIV/USAID Nicaragua Program”
18
Caribbean region stand out with 34% and 25% rates. As expected, the 
highest schooling rates are found in younger groups. [Tables 3 through
6] 
3.1.3 Household
One third of MARPs (33%) live in their own home as head of 
household (27%) or as spouse of the head of household (6%). This 
condition is found in the majority of SW (55%) and only in a minority
of TRANS (23%) and MSM (22%). As a result, two thirds of MARPs 
(67%) live with their parents, grandparents or in-laws [Tables 17, 18].
About 35% of MARPs have children. Similar rates prevail across 
regions, except for Managua with 42%. Rates increase with age, widely 
ranging from 8% among those under 20 years to 61% among those 
more than 40 years old. The group with the highest rates is SW with 
83%, showing 41% among those under 20 years of age and 90% among 
those above 25 years of age. The Caribbean region has the lowest rate 
(68%) and Managua has the highest rate (91%). Sixteen percent of 
MSMs have children. The west region stands out with 7%. Rates range 
from 4% among those under 20 years of age to 39% among those 
above 40 years of age. Only 4% of the TRANS population has children.
[Tables 10 through 13] 
Twenty-two percent of MARPs had their first child when they were 
under 20 years of age (58% of those with children), and 6% had their 
first child when they were under 15 years of age (16% of those with 
children). The highest rate (27%) is reported in Managua and the 
lowest rate (16%) in the south Pacific region. Sixty percent of SWs had 
a child when they were under 20 years of age, with the highest rate
(65%) reported in Managua and the lowest (44%) in the Caribbean 
region. Only 5% of MSMs and 1% of TRANS had a child when they 
were under 20 years of age [Tables 14 through 16] 
3.1.4 Work
Seventy-two percent of MARPs work. The highest rate is found in 
Managua (79%) and the lowest rates in the south Pacific and Caribbean 
regions (67%). Ninety-two percent of SW claim to be working, as well 
as 68% of TRANS, and 62% of MSMs. 
Sixty percent of working MARPs claim to be self-employed, as well as 
78% of SW, 60% of TRANS, and 42% of MSM. Forty percent of MARPs 
work for someone and earn a salary, as well as 58% of MSMs, 40% of
TRANS, and 22% of SWs [Tables 19 through 22]. Seventy percent of 
SWs and eleven percent of TRANS claim to be sexual workers. 
Fourteen percent of MARPs work at bars, hotels and restaurants, a 
sector that employs 18% SW, 14% TRANS, and 11% MSM. Most 
44%
58%
62%
62%
65%
60%
Caribbean
South Pacific
West
Central
Managua
Total SW
Graph 6
SWs with 1Child < 20 years, by region 
7%
25%
65%
90%
2% 6%
Under 25 years 25 and older
HSH TS TRANS
Graph 5
MARPs with Children, by Age
Assessing PrevenSida Project Coverage - “2014 Mid-Term Performance Evaluation of HIV/USAID Nicaragua Program”
19
MARPs (53%) claim other occupations, namely, 87% of MSM, 67% of
TRANS, and 11% of SW. Seven percent of TRANS and two percent of 
SWs claim to perform housework [Table 23]. 
Thirty-nine percent of MARPs work in activities related to tourism. 
The highest rates are found in the Caribbean (54%) and South Pacific 
region (43%), accounting for 50% of TRANS, 40% of SW, and 33% of
MSM. Forty-six percent of MARPs work in activities related to mobile 
populations (passenger or cargo transport). The highest rates are 
found in the West (54%) and Caribbean (51%) regions, and the lowest 
rates are found in the Central (46%), South Pacific (43%), and Managua
region (41%). More than half (53%) of working SWs and half (50%) of
working TRANS are related to mobile populations. This rate is 
somewhat lower among MSM (37%) [Tables 24, 25]. 
Forty-four percent of MARPs search for work. Sixty-four percent are 
unemployed and 36% are working. but want to change jobs. The 
highest rate of MARPs who want to change jobs is found among SWs 
and TRANS, accounting for 40% of those who work in each group. 
Thirty percent of working MSMs are also searching for another job. 
Consequently, 46% of MARPs work and do not search jobs, 26% work 
but keep searching for another job, 18% do not work and are 
searching for jobs, and 10% are unemployed and are not searching for 
jobs. 
This situation is expressed differently in each MARP group. As can be 
seen, although SWs make up the largest segment of workers not 
looking for a job (54%), they are also the largest segment (37%) of 
workers looking for another job to change working conditions, and 
certainly some want to change activities. This segment is also large for 
TRANS (27%) and MSM (19%), which is significant in relation to the 
other groups on account of their associations with self-esteem and 
risky practices [Table 26]. 
3.1.5 Participation in Organizations 
Forty percent of MARPs claim to be involved in some organization, 
namely, 56% of TRANS, 41% of MSM, and 28% of SW. The highest 
participation rate is found in the West region at 54%, and the lowest 
rate is found in the South Pacific region at 25%, while the other regions 
range from 37% to 43%. Twenty-nine percent identify their 
participation with an NGO or an association working on HIV, 
particularly TRANS (46%). Three percent name some government 
program (GSP), while 8% mention a political organization (mostly MSM
at 13%) [Tables 27, 28]. 
Assessing PrevenSida Project Coverage - “2014 Mid-Term Performance Evaluation of HIV/USAID Nicaragua Program”
20
3.1.6 Awareness of Rights
To the question, “Could you tell me one right that has not been 
realized for … SW, MSM, TRANS?,” nine percent answered, “None is 
realized”, and 5% said “All are realized.” Fifteen percent of MARPs, 
including 21% of SW, 15% of MSM, and 7% of TRANS, replied “I don’t 
know.” Fifty-four percent of MARPs mentioned some rights related to 
the group’s stigma and discrimination status (respect, non-violence, 
freedom of expression, equality), accounting for 59% of MSM, 40% of
SW, and 64% of TRANS. Seventeen percent of MARPs, including 26% 
of TRANS, 18% of SW and 13% of MSM, pointed out some basic rights 
that are not realized (health, education, housing, employment) [Tables 
29 through 33]. 
3.1.7 Trust in HIV Care Network
Seventy-four percent of MARPs are aware of some organization that 
can provide HIV support services, accounting for 79% of TRANS, 75% 
of SW and 72% of MSM [Table 34], with regional differences ranging 
from the highest in the Caribbean (85%) to the lowest in the South 
Pacific region (66%).
Thirty-six percent of MARPs mention an organization in the 
PrevenSida network, 16% mention other NGOs, and 19% mention
MOH services and other public services, while 26% did not recognize 
any support organization. The proportional distribution of the three 
types of services stand out among SW, ranging from 23% to 25%, while 
the percentage of MSMs and TRANS that mention the PrevenSida 
network is by far the majority (40% and 48%, respectively). The 
Caribbean region and Managua have the highest PrevenSida network 
awareness rates, while the Central region shows the lowest rate
[Tables 35, 36]. 
In addition to awareness, interviewed MARPs were asked if they had 
used those services in the last year, and the response was that 59% of 
MARPs had been exposed to the services, including 65% TRANS, 63% 
SW, and 53% MSM. The highest coverage was reported in Managua 
(67%) and the lowest in the Central region (50%) [Table 37]. The 
PrevenSida network contributes to this coverage by serving 31% of all 
MARPs (46% in the Caribbean, 17% in the Central region). PrevenSida 
support services cover 32% of MSM (50% in the Caribbean, 22% in the 
Central region), 21% of SW (40% in the Caribbean, 2% in the Central 
region), and 43% of TRANS (55% in Managua and the West region, 
24% in the Central and South Pacific regions) [Table 38]. 
15% 21%
7%
59% 40%
64%
13% 18% 26%
MSM SW TRANS
Social human 
rights
E&D related 
human rights
Do not know
Graph 10 
MARPs by Awareness of Unrealized
Rights
Assessing PrevenSida Project Coverage - “2014 Mid-Term Performance Evaluation of HIV/USAID Nicaragua Program”
21
In addition to the region and MARP group, PrevenSida network 
coverage with HIV support services is associated with other factors: 
• Coverage increases with schooling level, accounting for 17% (X2
39.4) at the elementary school level, 35% (X2 6.5) at the secondary 
school level, and up to 37% (X2 29.7 2Gl) at the technical and higher 
education level. 
• Not having children (35%), as compared to MARPs that have 
children (23%) (OR 1.9, CI95% 1.4-2.4, X2 23.0).
• Having a tourism-related job (34%) as opposed to other jobs (26%) 
(OR 1.4, CI95% 1.1-1.9, X2 5.6). 
• Recognizing any right that has not been realized for them as a group 
(34%), in contrast with MARPs that are not aware of unrealized 
rights (19%) (OR 2.1, CI95% 1.5-3.0, X2 21.0).
3.1.8 Usual Venues for Meeting New Sexual Partners
In applying the central elements of the PLACE approach, inquiries were 
made about the places most often used by MARPs to meet “new 
pickups”. Forty-eight percent responded that they looked for new 
partners at bars, discotheques and hotels, usually located in 
recreational areas or near population concentration sites on account 
of their mobility (terminals, ports, etc.). Thirty-eight percent indicated 
that they seek new sexual partners in commercial areas, including 
shopping centers, streets, or markets. Finally, 14% pointed out that 
they met new pickups in communication networks based on mobile 
phones or Internet. 
By and large, the three MARP groups exhibit a similar structure in the 
use of the three types of venues to find “new pickups”, and the 
broader use of “networks” by MSM and commercial areas by TRANS 
stand out.
3.1.9 Level of Influence of Local Structural Factors 
At each venue to be mapped, fieldwork team coordinators interviewed 
key informants, including PrevenSida network NGO staff, MARP group 
leaders, or local institution officials with broad knowledge of, and 
experience with, local reality. This was done to gather qualitative 
information about structural factors that might be affecting MARP living 
or working dynamics (for instance, commercial sex, meeting venues, 
etc.), including periodicity/seasonality of these dynamics. From the 
perspective of the PrevenSida coverage analysis, the level of incidence 
of tourism, mobile population flows and cargo transport rest areas 
stands out. 
23%
35% 34% 34%
23% 19%
Have children Tourism work Recognize
unrealized right
Yes No
Graph 13 
MARPs who have used PrevenSida HIV 
Support Services, by factors of significant 
impact
48% 51% 43%
36% 36% 47%
16% 13% 10%
MSM SW TRANS
Communication 
networks
Commercial 
zones
Bars / 
Discotheques
Graph 14 
MARPs by Usual “New Pickup” 
Venues
Graph 15 
MARPs in municipalities with high tourism 
and mobility rates by region
Tourism Population Mobility
Managua West South Pacific Central Caribbean
Assessing PrevenSida Project Coverage - “2014 Mid-Term Performance Evaluation of HIV/USAID Nicaragua Program”
22
3.2 Total MARP Coverage
This section deals with the exposure of interviewed MARPs to the combined actions of different preventive service 
providers. To this end, four iconic activities were included as the main transformation axis in HIV epidemic 
programs targeted on high risk population groups (MARPs), based on interpersonal preventive action 
communication, which as a whole makes up the combination prevention model, to wit: 
1. Exposure to promotional contacts in MARP living spaces and recently focused on concentration venues where 
new sexual partners are sought (bars, discotheques, shopping centers, streets, tourist sites, etc.).
2. Training or exchange events dealing with general or HIV-specific issues that address behavioral change, gender￾based violence, stigma and discrimination, etc. 
3. Peer groups, where 6–9 individuals from the same MARP group share experiences in relation to HIV 
protection. 
4. HIV voluntary counseling and testing in accordance with national protocols and under such conditions that it 
can be offered by health institutions or NGOs that have taken on this responsibility.
Two indicators have been built through a simple summation of these activities in each interviewed MARP: 
• Number of activities that each MARP has been exposed to in the past year, regardless of the service providing 
source, referred to as MARP “global coverage.” The variable resulting from the summation has 0–4 values, 
with 0 identifying persons who have not received any activity (no coverage) and the remaining values (1–4) 
indicating the number of activities to which they have been exposed. A second variable is a dichotomous 
variable, which shows if each individual MARP member has been covered (by one or more activities) or not 
covered at all (no activity).
• As shown above, a simple summation of activities provided by some PrevenSida network organization was 
done, obtaining two similar variables. In other words, the number of activities provided by PrevenSida to which 
each MARP member has been exposed, and the dichotomous variable in which each of the two values 
expresses either “With PrevenSida network coverage” (one or more activities provided by these 
organizations) or “No PrevenSida network coverage” (no network organization was named as provider). 
In terms of global coverage, 11% of the interviewed MARPs did not 
receive any of the four activities. The highest rate of uncovered MARPs 
was in the South Pacific region (14%) and the lowest in Managua (7%); 
21% have received one activity, 28% have received two, 24% have 
received three activities, and 16% have received all four activities.
In terms of PrevenSida network coverage, 55% of the interviewed 
MARPs have not received any of the four activities provided by the 
PrevenSida network, 45% have received one or more activities from 
the PrevenSida network, broken down as follows: 21% have received 
one activity, 14% have received two activities, 8% have received three 
activities, and 2% have received the four activities from PrevenSida. 
If the global coverage (one or more of the four activities mentioned) of
the interviewed MARPs is 89% and the coverage of PrevenSida is 45%, 
Assessing PrevenSida Project Coverage - “2014 Mid-Term Performance Evaluation of HIV/USAID Nicaragua Program”
23
it could be inferred that coverage by “other” providers (NGOs, MOH, 
private organizations) is 44%, but such statement would not be very 
accurate considering that all preventive service providers deal with the 
same MARP universe in their day-to-day activities of searching, 
detecting, contacting, convening and implementing events or group 
sessions, which are usually accessible in limited living spaces. 
In this scenario, delimiting the contribution made by each organization 
or program to global coverage is a complex proposition, even more so 
where different programs subsidize the same organization. In the 
search for greater accuracy, a process was undertaken to differentiate 
MARPs that had been exposed exclusively to the PrevenSida network 
from those exclusively exposed to “other providers”, in order to also 
differentiate those MARPs that had received activities from different 
sources. The coverage was classified into four groups: 
1. MARPs with none of the 4 activities. No coverage, 11%.
2. MARPs served exclusively by “Other providers,” 44%.
3. MARPs receiving combined care (PrevenSida and others), 30%.
4. MARPs served exclusively by the PrevenSida network, 15%.
We can therefore say that the actual coverage of “other providers” is 
74% because this is the share of MARPs that have received one or 
more activities from this source, whereas the actual coverage of the 
PrevenSida network is 45% of the total MARPs interviewed. Clearly, 
there is a coverage overlapping of the two global preventive service 
agencies analyzed. In effect, 30% of the MARP segment is shared 
(mixed coverage) as it has received activities from both sources. Thus, 
the PrevenSida coverage analysis, which is the main purpose of this 
study, would have to provide evidence of the characteristics and 
conditioning factors of the coverage of this segment that has global 
access to PrevenSida (45%), as well as the coverage of the segment 
that has exclusive access to PrevenSida (15%).
In fact, the distribution of the four coverage groups is very 
heterogeneous within each MARP group, standing out the majority 
coverage of “other providers” among SWs, as well as the scope of the 
mixed coverage among TRANS. 
3.3 PrevenSida Global Coverage
This analysis deals with the MARP segment exposed to one or more 
activities provided by the PrevenSida network, and therefore includes 
both mixed and exclusive coverage. 
The highest global coverage rate of the PrevenSida network is 
reported in the Caribbean (61%) and the lowest in the Central region 
Graph 17 
Preventive Activity Coverage by service 
provider
Exclusive
PrevenSida 
15%
Mixed 
“PrevenSida
+ Others”
30%
Exclusive 
“Others” 
44%
No coverage 
11%
Assessing PrevenSida Project Coverage - “2014 Mid-Term Performance Evaluation of HIV/USAID Nicaragua Program”
24
(28%). The three remaining regions maintain an average level (Managua, 
49%, West region, 48%, and South Pacific region,43%). The highest 
PrevenSida coverage is found in the TRANS population (60%) and the 
lowest in SWs (34%), whereas coverage in MSMs is 47%. The most 
relevant factors influencing this coverage are: 
• PrevenSida global coverage is higher in MARPs older than 24 years 
than in younger working MARPs (48% versus 42%, X2 5.1), which 
means that older MARPs have a 28% higher probability of access
(OR 1.3, CI 95% 1.0-1.6). This association is stronger when: 
 MARPs are unemployed. The PrevenSida coverage rate is 66% 
among MARPs more than 24 years old, and remains at 41% 
among younger people. So the probability of having access to 
PrevenSida is almost triple in older MARPs, compared to 
younger MARPs (OR 2.8, CI 95% 1.7-4.6, X2 19.5). A significant
contrast (X2het 11.8, p=0.0006) is that there is no difference in 
probabilities of accessing PrevenSida amongst employed MARPs
(42% versus 45%, X2 0.7).
 This effect is even higher when MARPs are studying, but 
unemployed. In this segment, coverage reaches 70% among
older MARPs and remains at 40% in younger MARPs, which 
means older MARPs have 3.5 times more probability of 
accessing PrevenSida than younger unemployed MARPs that 
study (OR 3.5, CI 95% 1.5-7.6,X2 10.2).
• At 34%, PrevenSida global coverage is lower in MARPs that have 
not completed primary school, as compared to 49% in MARPs that 
have completed 6th grade and beyond. Consequently, MARPs with 
less schooling have half the probability of accessing PrevenSida, as 
compared to those that have a higher schooling level (OR 0.55, CI 
95% 0.4-0.7,X2 21.3).The ratio between PrevenSida coverage and 
schooling is modified when: 
 MARPs with primary school education are employed, since their 
PrevenSida coverage is further reduced (32%), whereas those 
that are working and have a schooling level beyond 6th grade 
remain at 49%. This widens the probability gap in accessing 
PrevenSida coverage to slightly less than half to the detriment of 
MARPs with less schooling (OR 0.49, CI95% 0.4-0.7,X2 24.1).In a 
significant contrast (X2het 5.1, p=0.02), schooling does not make 
any difference in the probabilities of unemployed MARPSs 
accessing PrevenSida (50% versus 48%, X2
0.06).
 The above condition is only found among people employed in 
low-risk occupations, while MARPs with some primary schooling
have a 27% coverage and those completing 6th grade and beyond 
Graph 20 
Global PrevenSida Coverage by age and 
employment status
Under 24 years 24 years and older
All MARPs Work Do not work
34% 32%
49% 49% 50% 48%
All MARPs Work Do not work
Up to 6th grade Beyond 6th grade
Graph 21 
PrevenSida Global Coverage
By schooling and employment status
Assessing PrevenSida Project Coverage - “2014 Mid-Term Performance Evaluation of HIV/USAID Nicaragua Program”
25
have a 46% coverage (OR 0.44, CI 95% 0.3-0.7,X2 16.4). On the 
other hand (X2het 3.6, p=0.05), no significant difference was found 
in PrevenSida coverage between schooling levels within the high￾risk occupation MARP stratum (45% versus 50%, X2 0.9). 
 MARPs are not aware of unrealized E&D-related rights, which 
reinforces the restrictive effect of low schooling on PrevenSida 
coverage, accounting for just 25%, whereas those with higher 
schooling remain at 47% (OR 0.38, CI 95% 0.2-0.6, X2 24.2). A 
significantly different situation (X2het 7.0, p=0.008) was found in 
MARPs aware of unrealized E&D-related rights because the 
difference in PrevenSida coverage between both schooling levels 
is not significant (43% versus 50%, X2 2.6). 
• MARPs that study have a higher PrevenSida coverage that non￾students (49% versus 43%), which translates into a 30% higher 
probability of accessing PrevenSida services (OR 1.3,CI 95% 1.0-1.6, X2
4.9). This ratio is modified by various factors, the most important 
including: 
 Employment status because the PrevenSida coverage gap 
between students (54%) and non-students (41%) is widened, as 
well as the probability of accessing PrevenSida services, which 
becomes 72% higher among employed students versus employed 
non-students (OR 1.7,CI 95% 1.3-2.3, X2 12.8). In a significant 
contrast (X2het 12.7, p=0.0004), unemployed student MARPs do 
not make any significant difference in the likelihood of accessing 
PrevenSida coverage (44% versus 53%, X2 3.0).
 In MARPs that have recently been victims of abuse or violence, 
the coverage of those that study increases to 58%, whereas the 
coverage of non-students remains at 42% (OR 1.9, CI 95% 1.2-2.9, 
X2
 8.4).In a significantly different situation (X2het 4.0, p=0.04), 
studying does not make any difference in PrevenSida coverage 
for MARPs that have not recently been victims of violence (46% 
versus 44%, X2 0.7). 
 In MARPs unaware of unrealized E&D rights, PrevenSida 
coverage is 50% for students and 38% for non-students (OR 1.66, 
CI 95%1.2-2.4, X2 8.1). On the other hand (X2het4.1, p=0.04), 
PrevenSida coverage between students and non-students is not 
significantly different (49% versus 48%, X2 2.6) in MARPs aware of 
unrealized E&D rights. 
• MARPs with tourism-related jobs have higher PrevenSida coverage 
(50%) than MARPs whose work is not related to tourism (39%), 
creating a 54% probability gap of accessing PrevenSida services in 
MARPs linked to tourism, as compared to those whose work is not 
Graph 24 
PrevenSida Global Coverage
By tourism-related work and occupational 
risk
Tourism work Non-tourism work
 All MARPs High-Risk Low-Risk
 Occupation Occupation
Graph 22 
PrevenSida Global Coverage
By schooling and unrealized E&D-related 
rights
Up to 6th grade Beyond 6th grade
 All MARPs Mention Do not mention 
E&D rights E&D rights
Graph 23 
PrevenSida Global Coverage
By study and employment status
Study Do not study
All MARPs Work Do not work
Assessing PrevenSida Project Coverage - “2014 Mid-Term Performance Evaluation of HIV/USAID Nicaragua Program”
26
related to tourism (OR 1.5, CI 95%1.2-2.0, X2 10.8). This basic ratio is 
contingent on different situations, including: 
 When MARPs work in higher risk occupations, their linkage to 
tourism does not make a significant difference in terms of 
PrevenSida coverage (52% versus 47%, X2 1.3). In a different 
situation(X2het 5.1, p=0.02), we found that MARPs in lower risk 
occupations have a reduced coverage, and the gap increases 
between MARPs with tourism-related jobs (47%) and those with 
non-tourism-related jobs (28%). This entails more than two 
times the probability of coverage for MARPs with tourism￾related jobs than those without that linkage (OR 2.3, CI 95% 1.5-
3.4, X2 16.4).
 When MARPs are employed in tourism-related jobs, schooling 
level has a noticeable impact on PrevenSida coverage. Coverage 
38% for MARPs with a primary school level, 51% for those at a 
high school level, and 65% for technical or higher education level 
(X2 11.7, 2 GI).
 Coverage is markedly increased in MARPs living with a stable 
partner when they have tourism-related jobs (64%) and it is 
reduced when their job is unrelated to tourism (34%). This 
significantly widens (from 3.4 to 1) the gap between both 
segments in terms of the likelihood of accessing PrevenSida 
services (OR 3.4, CI 95% 1.7-6.9, X2 11.5). In a significant 
heterogeneity condition (X2het 5.6, p=0.01), the coverage gap 
resulting from tourism-related jobs and the probabilities of 
accessing PrevenSida services by MARPs not living with their 
partner is smaller (47% versus 40%, OR 1.4, CI 95% 1.0-1.8, X2 4.7).
 As expected, MARPs working in tourism-related jobs in 
municipalities ranked as highly influenced by tourism have a 
much higher PrevenSida coverage (57%) than MARPs not related 
to the sector (38%), which means that MARPs working in 
tourism-related jobs have twice the probability of accessing 
PrevenSida services (OR 2.1, CI 95% 1.4-3.2, X2 13.4). In a 
significantly heterogeneous situation (X2het 4.1, p=0.04), tourism￾related jobs in municipalities with low or no tourism incidence 
make no difference in PrevenSida coverage (44% versus 39%, X2
1.2). 
• PrevenSida coverage changes according to municipality size and 
political-administrative location. Coverage is 49%in the capital, 46% 
in departmental capitals, and 40% in peripheral municipalities(X2 7.0, 
p=0.008, 2 GI).The main difference is found in MARPs living in smaller 
municipalities, which have a 26% lower probability of accessing 
PrevenSida services, as compared to those living in departmental 
capitals or in the capital city (40% versus 47%, OR 0.74, CI 95% 0.6-0.9, 
Graph 25 
PrevenSida Global Coverage
By tourism-related employment and schooling
Tourism work Non-tourism work
Primary Secondary Higher
Graph 26 
PrevenSida Global Coverage
By tourism-related job and cohabitation with 
partner
Tourism work Non-tourism work
 All MARPs Cohabit Do not cohabit
 with partner with partner
Graph 27 
PrevenSida Global Coverage
By tourism-related employment and 
municipal HIV test positivity rate
Tourism work Non-tourism work
All MARPs Municipality Municipality
 with high + with low +
 test rate test rate
Assessing PrevenSida Project Coverage - “2014 Mid-Term Performance Evaluation of HIV/USAID Nicaragua Program”
27
X2
6.4). This association is clearly evident in SW and MSM, but there 
is no difference for TRANS at the three municipal levels. 
The strongest trend is found in SW, which in addition to starting 
with the lowest overall coverage (34%) declines from 37% in the 
capital city and departmental capitals to 17% in peripheral 
municipalities (X2 6.2, p=0.02, 2Gl). This situation is similar for MSM, 
although they start with a better overall coverage (46%), but their 
PrevenSida coverage is 56% in the capital, 47% in departmental 
capitals, and 40% in peripheral municipalities (X2 9.4, P=0.002, 2Gl).
 MSM in peripheral municipalities have a 34% probability of 
accessing PrevenSida services as compared to MSM living in the 
capital and departmental capitals (40% versus 50%, OR 0.66, CI 95%
0.5-0.9, X2 6.5). Unlike SW in municipalities with a high incidence 
of mobile population, there is no difference in PrevenSida 
coverage between peripheral and other municipalities (52% versus 
52%, X2 0.1.). In a significant contrast (X2het 18.1, p=0.000), 
PrevenSida coverage for MSM living in peripheral municipalities 
with low mobile population incidence drops to 15%, whereas 
MSM living in departmental capitals or capital city areas without 
mobile population incidence have a 48% PrevenSida coverage
(OR 0.19, CI 95% 0.1-0.4, X2
26.7). As a result, MSM in peripheral 
municipalities with low mobile population incidence have 5 times 
less probability of accessing PrevenSida services than MSM living 
in the capital or departmental capitals.
 Coverage differences in SW living in municipalities have one 
third (1 out of 3 times) probability of accessing PrevenSida 
services, as compared to SW living in the capital and 
departmental capitals (17% versus 37%, OR 0.34, CI 95% 0.1-0.6, X2
11.7). This association is stronger in municipalities with a higher 
mobile population incidence since SW living in peripheral 
municipalities reduce their PrevenSida coverage down to 5%, 
whereas they have a PrevenSida coverage of 43% in 
departmental capitals and capital city areas with mobile 
population incidence (OR 0.1, CI 95% 0.02-0.2, X2 19.9). In a 
significant contrast (X2het 10.6, p=0.001), there are no coverage 
differences for SW in municipalities with low mobile population 
incidence between peripheral municipalities and other 
municipalities (29% versus 29%, X2 1.0).
• PrevenSida coverage is 65% for MARPs participating in some 
organization and 35% for those that do not participate in any 
organization. Therefore, participating MARPs have almost three 
times as much probability of accessing PrevenSida services, as 
opposed to those not participating in any organization (OR 2.9, CI 
Graph 28 
PrevenSida Global Coverage
By MARP group and type of location
MSM SW
 Capital Departmental Peripheral
 Capitals Municipalities
50% 52%
15%
40%
52% 48%
All MSM Municipality with
high mobile
population
Municipality with low
mobile population
Capital and Deptal capital Peripheral Municipalities
Graph 29 
MSM with PrevenSida Global Coverage
By type of municipality and mobile population 
influence on municipality
50%
43%
29%
39%
5%
29%
All SW Municipality with
high mobile
population
Municipality with
low mobile
population
Capital and Deptal capital Peripheral Municipalities
Graph 30
SW with PrevenSida Global Coverage
By type of municipality and mobile population 
influence on municipality
Assessing PrevenSida Project Coverage - “2014 Mid-Term Performance Evaluation of HIV/USAID Nicaragua Program”
28
95% 2.3-3.7, X2 92.3). PrevenSida coverage for MSM with a 45% 
organization rate is similar to the overall rate (61% versus 36%, OR 
2.8, CI 95% 2.0-3.9, X2 39.8). This association is even more marked in 
TRANS, which have a 78% PrevenSida coverage if they participate in 
organizations, as compared to 38% for those that do not participate 
in any organization. Hence, organized TRANS have six times more 
probability of accessing PrevenSida services than unorganized 
TRANS (OR 5.9, CI 95% 3.6-9.9, X2 47.6). No significant difference was 
found in PrevenSida coverage for SW with a 28% participation rate 
in some organization (39% versus 32%, X2 2.2). The effect of 
organizational participation on PrevenSida coverage is modified 
when: 
 MARPs have children, in which case PrevenSida coverage is 
reduced. Organized MARPs have a coverage of 45%, and 
unorganized MARPs have a coverage of 31%. The gap between 
organized MARPs with children and unorganized MARPs with 
children in terms of probabilities of accessing PrevenSida 
services is reduced (OR 1.8, CI 95% 1.2-2.8, X2 8.9). In a significant 
contrast (X2het 5.3, p=0.02), the same effect described for MARPs 
in general is maintained for MARPs with no children (66% versus 
38%, OR 3.2, CI 95% 2.4-4.3, X2 69.6).
 PrevenSida coverage increases markedly in MARPs working in 
higher risk occupations and participating in organizations (71%) 
and remains the same (34%) in unorganized MARPs.
Consequently, the probability gap of accessing PrevenSida 
services increases significantly to a ratio of almost 5 to 1 in favor 
of organized MARPs (OR 4.7, CI 95% 3.3-7.0, X2 74.2). In a 
significantly heterogeneous condition (X2het 14.4, p=0.0002), in 
MARPs working in lower risk occupations, the probability gap of 
PrevenSida coverage and care is reduced between organized and 
unorganized MARPs (45% versus 33%, OR 1.7, CI 95% 1.1-2.6, X2 
6.1).
• MARPs that have had their first sexual relation before 13 years of 
age have a PrevenSida coverage of 51%, and those that had their 
first sexual relation at 13 years or older have a PrevenSida coverage 
of 44%. MARPs that had an earlier sexual initiation have a 30% 
higher probability of accessing PrevenSida services in comparison 
with those that had it after 12 years of age (OR 1.3, CI 95% 1.0-1.7, X2
4.0). This association is modified when: 
 MARPs do not study or work. In this case, coverage for those 
who had an early sexual life initiation increases noticeably (71%), 
and coverage for those who had it after 12 years of age remains 
the same (48%). The probability gap of accessing PrevenSida 
services is 2.6 times larger for MARPs that had an earlier sexual 
Graph 31 
PrevenSida Global Coverage
By MARP group and participation in 
organizations
Organized  Not organized
MSM SW TRANS
Graph 32 
PrevenSida Global Coverage
By participation in organizations and 
parenting status
Organized Not organized
All MARPs Have children Do not have children
Graph 33 
PrevenSida Global Coverage
By participation in organizations and 
occupational risk level
Organized Not organized
All MARPs High Risk Job Low Risk Job
Assessing PrevenSida Project Coverage - “2014 Mid-Term Performance Evaluation of HIV/USAID Nicaragua Program”
29
initiation than for those who had it after 12 years of age (OR 2.6, 
CI 95% 1.2-5.7, X2 6.2). In a significantly different situation (X2het 3.8, 
p=0.05), sexual initiation age does not make a difference in 
PrevenSida coverage for MARPs that study and/or work (47% 
versus 43%, X2 1.3).
 Where MARPs are not aware of unrealized E&D-related rights, 
PrevenSida coverage increases for those initiating sexual life 
before 13 years of age (55%) and is reduced for those that 
initiated it at 13 years of age or older (38%), thus doubling their
probability of accessing PrevenSida services (OR 2.0, CI 95% 1.3-
3.1, X2 10.8). In a significantly heterogeneous condition (X2het 7.7, 
p=0.005), sexual initiation age does not make any significant 
difference in PrevenSida coverage for MARPs that are aware of
E&D-related rights (48% versus 49%, X2 0.1).
• MARPs that had their first sexual relation under pressure (did not 
want to) have a lower PrevenSida coverage (44%) than MARPs
whose first sexual relation was voluntary (57%). The probability of 
accessing PrevenSida is reduced to half for MARPs that had a sexual 
initiation under pressure, as compared to those who had it of their 
own free will (OR 0.57, CI 95% 0.4-0.8, X2 10.1). This association is 
modified under some conditions: 
 MARPs without a stable partner who experienced their first 
sexual relation under pressure have an even lower coverage
(37%), while those who had a voluntary sexual initiation increase 
their PrevenSida coverage (59%). As a result, the gap in the 
probability of accessing PrevenSida services between both 
segments is increased from 1 to 2.5 times (OR 0.4, CI 95% 0.2-0.7, 
X2 13.5). In a significantly different situation (X2het 4.1, p=0.04), 
voluntary sexual initiation in MARPs with a stable partner does 
not make a significant difference in PrevenSida coverage (51% 
versus 56%, X2 0.5). 
 MARPs with secondary education or less (11th grade or less)
have a 36% PrevenSida coverage if they had their first sexual 
relation under pressure, whereas coverage increases to 57% for 
those who had their first sexual relation voluntarily, so the gap 
in the odds of accessing PrevenSida services increases to a ratio 
of 1 to 2.3 times (OR 0.43, CI 95% 0.3-0.6, X2 17.5). In contrast 
(X2het9.1, p=0.002), MARPs with technical or college education 
have an increased coverage, although voluntary sexual initiation 
does not make a significant difference in the odds of accessing 
PrevenSida (70% versus 60%, X2 1.3). 
• MARPs looking for a job (60% are employed, 40% are not) have a 
higher PrevenSida coverage (50%) than those not looking for a job 
Graph 34 
PrevenSida Global Coverage
By age of 1st sexual relation and 
employment/study status
1
st SR <13 years 1st SR > 12 years
 All MARPs Do not study or work Study and/or
 work
Graph 35
PrevenSida Global Coverage
By assertiveness in 1st sexual relation and 
stable partner status
1st SR under pressure 1st SR voluntary
 All MARPs With stable partner No stable partner
Graph 36 
PrevenSida Global Coverage
By assertiveness in 1st sexual relation and 
schooling level 
1st SR under pressure 1st SR voluntary
 All MARPs Up to 11th grade Technical or 
 college education
Assessing PrevenSida Project Coverage - “2014 Mid-Term Performance Evaluation of HIV/USAID Nicaragua Program”
30
(41%). Hence, MARPs that want to change jobs or look for a job 
have a 40% higher probability of accessing PrevenSida services than 
those not looking for a job (OR 1.4, CI 95% 1.1-1.7, X2 9.9).This 
association is modified by various influencing factors, such as: 
 In MARPs looking for new sexual partners in commercial venues 
(shopping centers, streets, markets), PrevenSida coverage 
increases to 53% in MARPs looking for a job, whereas 
PrevenSida coverage is reduced to 36% in MARPs not looking 
for a job, increasing two times the probability of accessing 
PrevenSida services (OR 2.0, CI 95% 1.4-3.0, X2 14.5). In contrast 
(X2het 4.6, p=0.03), looking for a job does not make any difference in 
PrevenSida coverage for MARPs seeking new partners in other 
venues (49% versus 44%, X2 2.2).
 In peripheral municipalities, MARPs looking for a job maintain a 
50% PrevenSida coverage, but those not looking for a job 
noticeably reduce their coverage to 31%. Hence, the probability 
of accessing PrevenSida is two times greater for MARPs that 
want to change jobs or look for a job, as compared to those not 
looking for work (OR 2.1, CI 95% 1.4-3.3, X2 13.1). In contrast (X2het
5.2, p=0.02), looking for work in departmental capitals or in the 
capital city does not make any difference in PrevenSida coverage 
(50% versus 45%, X2 2.3). 
 In municipalities with a high concentration of workers (off-shore 
assembly factories, large work centers, ports, mines), MARPs
looking for work have a 54% coverage, and those not looking for 
work have a 38% coverage. Hence, MARPs that want to change 
jobs or look for work have a 90% higher probability of accessing 
PrevenSida services than MARPs remaining in the same job (OR 
1.9, CI 95% 1.3-2.8, X2 12.9). In contrast (X2het 4.5, p=0.03), looking for 
work in municipalities with a low presence of workers does not 
make a significant difference in PrevenSida coverage (47% versus 
43%, X2 1.4). 
 In municipalities with a high influence of cargo transport (ports, 
rest and food areas, borders), MARPs looking for a job have a 
55% coverage, while those not looking for a job have a 37% 
coverage. Hence, MARPs looking for a job have twice as much 
probability of accessing PrevenSida services than MARPs that 
remain in the same job(OR 2.1, CI 95% 1.4-3.2, X2 12.8).In contrast 
(X2het 5.1, p=0.02), looking for a job in municipalities without any 
cargo transport incidence does not make a significant difference 
in PrevenSida coverage (48% versus 43%, X2
2.1).
It should be noted that PrevenSida global coverage was not directly 
associated with municipal risk levels linked to HIV positivity tests 
carried out by the PrevenSida network (45% versus 46%, X2 0.24), and no 
Graph 37 
PrevenSida Global Coverage
By job search and types of venue for new
“pickups”
Search for work Do not search for work
 All MAPRs New “pickups” New “pickups”
 Commercial site Bar-Disco/Networks
Graph 38 
PrevenSida Global Coverage
By job search and type of municipality
Search for work Do not search for work
All MARPs Peripheral Departmental 
 Municipality Capital/Capital
Graph 39 
PrevenSida Global Coverage
By job search and level of concentration of 
workers in the municipality
All MARPs Municipality Municipality
 with high with low
 concentration concentration
Search for work Do not search for work
Assessing PrevenSida Project Coverage - “2014 Mid-Term Performance Evaluation of HIV/USAID Nicaragua Program”
31
significant differences were found in coverage on account of pickup 
venues preferred by MARPs (bars/discotheques/motels account for 
47%, shopping centers/streets account for 44%, and social networks 
account for 43%) (X2 1.4, 2 GI). 
3.4 PrevenSida Exclusive Coverage
Exposure to PrevenSida actions without receiving any actions from other HIV preventive service providers in a 
multi-presence and sometimes competitive scenario of confluence, is in itself a unique differentiating condition of 
the coverage offered by the PrevenSida network. Since the main objective of this study is to assess PrevenSida 
coverage, different situations that might characterize it have been explored to contribute evidence that can be used 
in analysis and management decisions. 
Clearly, exclusivity does not stem from an intentional action to establish follow-up standards for PrevenSida 
network field teams, or from commitments with associations and NGOs to guide their affiliates or usersto 
“loyalty” to PrevenSida network services or any specific organization providing such services. Although 
PrevenSida´s overall planning is based on agreements reached with national response authorities to provide 
coverage to 15% of MARPs, this is just an indicative goal to establish and agree on PrevenSida subsidy goals. 
Nevertheless, exclusivity is not a guiding criterion in the implementation of preventive actions, i.e., seeking, 
contacting, convening and holding sessions or events with MARPs. 
As already mentioned, PrevenSida exclusive coverage is 15% of all 
interviewed MARPs. The highest rate (30%) is found in the Caribbean
and the lowest rate (9%) is reported in the Central region. The 
coverage in the other three regions remains at an average level 
(Managua 13%, West region 14%, South Pacific region 14%). In the 
three MARP groups, the lowest rate is 12%in SWs, whereas exclusive 
coverage is 16% in TRANS and 17% in MSM. Various factors are 
related to PrevenSida exclusive coverage variations, including, among 
others: 
• MARPs that participate in some organization have 18% PrevenSida 
exclusive coverage, while those not participating in any organization 
have 13% exclusive coverage. Therefore, organized MARPs have a
45% higher probability of getting exclusive care from PrevenSida, in 
comparison with unorganized MARPs (OR 1.45, CI 95% 1.1-2.0, X2 6.1).
This association is greater when: 
 MARPs are less than 25 years old. If they are not organized, 
their exclusive coverage is reduced to 9%, while organized 
MARPs maintain an 18% exclusive coverage. So the gap in 
probabilities of receiving exclusive care from PrevenSida 2.2 
times greater for organized MARPs (OR 2.2, CI 95% 1.3-3.8, X2
10.4). In contrast (X2het 4.2, p=0.04), participation in organizations 
does not make any significant difference in PrevenSida exclusive 
Graph 40 
PrevenSida Exclusive Coverage
Global and PrevenSida Coverage by Region
Total Managua West South Central Caribbean
MARPs Pacific
Graph 41 
PrevenSida Exclusive Coverage
By participation in organizations and age 
Organized Unorganized
 All MARPs < 25 years > 24 years
Assessing PrevenSida Project Coverage - “2014 Mid-Term Performance Evaluation of HIV/USAID Nicaragua Program”
32
coverage for MARPs more than 24 years old (18% versus 16%, X2
1.2).
 MARPs that had their first sexual relation after 16 years of age 
and are organized increase their PrevenSida exclusive coverage 
to 25%, while exclusive coverage in unorganized MARPs is only 
7%. Hence, organized MARPs have four times more probability 
of receiving exclusive care from PrevenSida (OR 4.1, CI 95%2.0-8.2, 
X2
 15.8).In a significantly different situation (X2het 10.5, p=0.001), 
participating in an organization does not make any significant 
difference in PrevenSida exclusive coverage in MARPs that had 
their first sexual relation before 17 years of age (16% versus 14%, 
X2 1.2).
 MARPs that are NOT aware of unrealized E&D rights and are 
organized have a 21% PrevenSida exclusive coverage, while 
unorganized MARPs have a mere 10% coverage, which means 
that not being aware of E&D rights and being organized 
increases by 2.4 times the probability of receiving exclusive care 
from PrevenSida, as compared to unorganized MARPs (OR 2.4, CI 
95% 1.6-3.8, X2 15.6). In contrast (X2het 9.6, p=0.001), being 
organized does not make any difference in the probabilities of 
receiving exclusive care from PrevenSida in MARPs that are
aware of unrealized E&D rights (15% versus 16%, X2 0.04). 
 In MARPs looking for new sexual partners through 
communication networks (telephones/Internet), being organized 
accounts for a noticeable increase in PrevenSida exclusive 
coverage (24%), while PrevenSida exclusive coverage is 
significantly reduced (7%) in MARPs using networks to find new 
“pickups”, but not participating in any organization. Hence, 
MARPs looking for “pickups” in networks have four times more 
probability of receiving exclusive care from PrevenSida, as 
opposed to unorganized MARPs (OR 4.0, CI 95% 1.7-9.6, X2 9.6). In 
a significantly heterogeneous situation (X2het 5.7, p=0.01), 
participating in any organization does not make any difference in 
the probabilities of receiving PrevenSida exclusive care for 
MARPs looking for new “pickups” through other means (17% 
versus 14%, X2 2.0). 
 In peripheral municipalities, MARPs participating in organizations 
have 22% PrevenSida exclusive coverage, whereas those not 
participating in any organization reduce their coverage to 10%, 
which means that organized MARPs have 2.5 times more 
probability of receiving PrevenSida exclusive care, as opposed to 
unorganized MARPs (OR 2.5, CI 95% 1.4-4.9, X2 10.8). In contrast
(X2het 5.5, p=0.02), participating in any organization in the 
departmental capitals and capital city does not make any 
All MARPs 1st SR > 16 years 1st SR <17 years
Graph 43 
PrevenSida Exclusive Coverage 
By participation in organizations and unrealized 
E&D rights 
Organized Unorganized
 All MARPS Mention E&D Do not mention
 E&D rights E&D rights
Graph 42 
PrevenSida Exclusive Coverage
By participation in organizations and age at first 
sexual relation 
Organized Unorganized
All MARPs 1st SR <16 years 1st SR <17 years
Graph 44 
PrevenSida Exclusive Coverage
By participation in organizations and usual 
places for new “pickups”
Organized Unorganized
 All MARPs New “pickups” New “pickups”
 Through networks bars/commerce 
Assessing PrevenSida Project Coverage - “2014 Mid-Term Performance Evaluation of HIV/USAID Nicaragua Program”
33
difference in the probabilities of receiving exclusive care from 
PrevenSida (16% versus 14%, X2 0.7). 
 In municipalities with a high influence of cargo transport, 
MARPs participating in organizations have a 23% PrevenSida 
exclusive coverage, whereas those not participating in any 
organization reduce their coverage to 8%, which means that 
organized MARPs have 3.3 times more probability of receiving 
exclusive care from PrevenSida than unorganized MARPs (OR 3.2, 
CI 95% 1.8-6.5, X2 16.0).In a significant contrast(X2het 10.4, p=0.001), 
participation in any organization does not make any difference in 
probabilities of receiving exclusive care from PrevenSida in 
municipalities with a low cargo transport incidence (16% versus
15%, X2 0.14). 
• MARPs with children have 11% PrevenSida exclusive coverage and 
those without children have 17%, which means that MARPs with 
children have 41% less probability of receiving exclusive care from 
PrevenSida than those without children (OR 0.59, CI 95% 0.4-0.8, X2
9.7). Two conditions modify this association: 
 In peripheral municipalities, MARPs with children significantly
reduce their exclusive coverage (4%), while those without 
children increase it (19%), in such a way the gap in the 
probability of receiving exclusive care increases in a 1 to 6.2 
ratio (OR 0.16, CI 95% 0.1-0.4, X2 14.7). In a significant contrast (X2het
7.7, p=0.005), the fact that MARPs have children does not make 
any difference in PrevenSida exclusive coverage (13% versus 16%, 
X2 2.0) in larger municipalities (departmental capitals and capital 
city). 
 In municipalities with a high concentration of workers due to 
enclave economies (off-shore assembly factories, mines, ports), 
PrevenSida exclusive coverage is reduced to 7% in MARPs with 
children and reaches 19% in MARPs without children. Hence, 
the probability of receiving exclusive care from PrevenSida is 
three times less for MARPs with children, as compared to 
MARPs without children(OR 0.3, CI 95% 0.2-0.6, X2 14.6).In a 
significant contrast (X2het 6.8, p=0.009), in municipalities with low 
or no concentration of workers, the fact that MARPs have 
children does not make any difference in PrevenSida exclusive 
coverage (13% versus 16%, X2 0.9).
• In MARPs that think there could be situations where beating their 
partner is justified, PrevenSida exclusive coverage is 10%, while it 
reaches 16% in those MARPs that think there is no situation that 
justifies beating their partner, which means 42% less probability of 
accessing PrevenSida exclusive care in MARPs that could justify 
Graph 47 
PrevenSida Exclusive Coverage
By custody of children and level of labor 
concentration in the municipality
Have children Do not have children
All MARPs Municipality Municipality
 with high with low
 Concentration Concentration
Graph 46 
PrevenSida Exclusive Coverage
By MARPs with children and type of municipality
Have children Do not have children
All MARPs “Peripheral” Capital / Dept.
 Municipality Capitals
Graph 45 
PrevenSida Exclusive Coverage
By participation in organizations and type of 
municipality
Organized Unorganized
All MARPs “Peripheral” Capital / Dept.
 Municipality Capitals
Assessing PrevenSida Project Coverage - “2014 Mid-Term Performance Evaluation of HIV/USAID Nicaragua Program”
34
partner violence (PV), as compared to MARPs that do not accept 
any situation that justifies partner violence (OR 0.58, CI 95% 0.3-0.9, 
X2
5.2). Three situations modify this association: 
 PrevenSida exclusive coverage is 7% in MARPs under 30 years of 
age that think there could be situations that justify beating their 
partner, compared to 16% when they think the opposite. The 
gap between these two groups in the probabilities of accessing 
PrevenSida exclusive care is 1 to 2.5 times(OR 0.4, CI 95% 0.1-0.7, 
X2 8.3). In contrast (X2het 4.5, p=0.03), MARPs over 30 years of age
that justify partner violence does not make any difference in 
PrevenSida exclusive coverage (17% versus 16%, X2 0.1). 
 PrevenSida exclusive coverage is reduced to barely 4% in MARPs
that do not participate in any organization and think there could 
be situations that justify beating their partner, whereas the 
coverage is reduced to 14% in those MARPS that think 
otherwise. Hence, the probabilities of having exclusive access to 
PrevenSida are almost four times less in MARPs that justify 
violence, as compared to those that do not justify it (OR 0.28, CI 
95% 0.1-0.6, X2 8.4). In contrast (X2het 4.1, p=0.04), justification of 
partner violence does not make any difference in PrevenSida 
exclusive coverage in MARPs participating in some organization
(16% versus 18%, X2 0.4). 
 In municipalities with a high concentration of workers in 
factories (off-shore assembly), ports, or mines, there is no 
difference in PrevenSida exclusive coverage associated to the 
opinion regarding partner violence (15% versus 15%, X2 0.0).In 
contrast (X2het 5.9, p=0.01), in municipalities with low labor 
concentration, MARPs that justify beating their partner 
significantly have their PrevenSida exclusive coverage reduced to 
just 6%, whereas it increases to 18% in those MARPs that think 
no situation justifies beating their partner. Therefore, the
probabilities of having exclusive access to PrevenSida are three 
times less for MARPs that justify violence than for those that do 
not justify it (OR 0.31, CI 95% 0.1-0.6, X2
9.5).
• The type of venues where MARPs more often seek new sexual 
partners makes a significant difference in PrevenSida exclusive 
coverage, which is 11% if they seek new sexual partners in 
commercial venues (shopping centers, streets),15% through 
networks, and 18% at bars/discotheques (X2 8.8, p=0.003, 2GL). 
Consequently, MARPs looking for new “pickups” in commercial 
venues have 36% less probability of receiving PrevenSida exclusive 
care, as compared to other MARPs (11% versus 17%, OR 0.64, CI 95%
0.4-0.9, X2 7.1). On the other hand, MARPs meeting new “pickups” at 
bars and discotheques have a 57% higher probability of receiving 
PrevenSida exclusive care, as compared to other MARPs (18% versus
Graph 48 
PrevenSida Exclusive Coverage
by justification of partner violence and age of 
MARPs
Justify PV Do not justify PV
All MARPs Under 30 years 30 years and more
Graph 49 
PrevenSida Exclusive Coverage
Depending on whether or not MARPs justify partner 
violence and participate in any organization 
Justify PV Do not justify PV
 All MARPS Unorganized Organized
Graph 50 
PrevenSida Exclusive Coverage
by justification of partner violence and 
municipal labor concentration
Justify PV Do not justify PV
 All MARPs Municipality Municipality
 with high labor with low labor
 concentration concentration
Assessing PrevenSida Project Coverage - “2014 Mid-Term Performance Evaluation of HIV/USAID Nicaragua Program”
35
12%, OR 1.57, CI 95% 1.2-2.2, X2 8.7). This association is modified by 
two conditions: 
 When MARPs work in some activity related to mobile 
populations, PrevenSida exclusive coverage increases for those 
looking for new pickups at bars/discotheques, and decreases for 
MARPs seeking new pickups in commercial venues. For MARPs 
working in activities related to mobile populations, “picking up” 
new partners at bars/discotheques increases their PrevenSida 
exclusive coverage to 22%, whereas those looking for new 
“pickups” in commercial venues reduce it to 8%.This means that 
MARPs looking for pickups at bars/discotheques have 3.5 times 
more probability of accessing PrevenSida exclusive care than 
those looking for pickups in commercial venues (OR 3.5, CI 95%
1.8-8.8, X2 14.5).In a significantly different situation (X2het 7.0, 
p=0.008), a “new pickup” venue does not make any difference in 
PrevenSida exclusive coverage for MARPs working in activities 
unrelated to mobile populations (14% versus 13%, X2 0.15). 
 In municipalities with high tourism, MARPs looking for “pickups” 
at bars/discotheques increase their PrevenSida exclusive 
coverage to 25%, whereas those meeting them in commercial 
venues have a 12% coverage, which shows that MARPs picking 
up at bars/discotheques have a 2.6 higher probability of accessing 
PrevenSida exclusive care than those picking up at commercial 
venues (OR 2.6, CI 95% 1.5-5.1, X2 9.5).In a significantly different 
situation (CI 90%, X2het 3.2, p=0.07), a “new pickup” venue does not 
make any difference in PrevenSida exclusive coverage in 
municipalities with low tourism incidence (15% versus 11%, X2
 1.8).
• MARPs working in higher risk occupations (bars, discotheques, 
commercial sex) have a 43% higher probability of PrevenSida 
exclusive care than those in lower risk occupations (17% versus 12%, 
OR1.43, CI 95% 1.1-2.0, X2 3.8). This association is modified when: 
 MARPs participate in some organization, in which case the 
probabilities of receiving PrevenSida exclusive care are 2.8 times 
higher for MARPs with higher risk occupations (23% versus 10%, 
OR 2.8, CI 95% 1.7-5.8, X2 10.6). In contrast (X2het 9.4, P=0.002), when 
they do not participate in any organization, occupational risk 
does not make any difference in the probabilities of receiving 
PrevenSida exclusive care (12% versus 13%, X2 0.4).
 MARPs in lower risk occupations that live in municipalities with 
a high incidence of PrevenSida HIV positive tests reduce their 
PrevenSida exclusive coverage by half, and those in higher risk 
occupations double their probabilities of receiving PrevenSida 
exclusive care as opposed to the other group(16% versus 8%, OR 
2.1, CI 95% 1.4-3.5, X2 8.6).In a significantly different situation (X2het
Graph 51 
PrevenSida Exclusive Coverage
by habitual venues in search of new sexual 
partners (“pickups”) 
 Commercial Networks Bars, discos,
 venues (Cel/Internet) motels
All MARPs Work with mobile Do not work with 
 population mobile population
Graph 52
PrevenSida Exclusive Coverage
By habitual venues for seeking new 
“pickups” and work with mobile population
Bars/Discos Commerce/Networks
Graph 53 
PrevenSida Exclusive Coverage
by habitual venues in search of new “pickups” 
and municipal tourism rate
Bars/discos Malls, networks
All MARPs Muncipality Municipality
 with high with low 
 tourism rate tourism rate
Assessing PrevenSida Project Coverage - “2014 Mid-Term Performance Evaluation of HIV/USAID Nicaragua Program”
36
6.5, p=0.01),occupational risk in municipalities with a low 
incidence of PrevenSida HIV positive tests does not make any 
difference in PrevenSida exclusive coverage (17% versus 20%, X2
0.5). 
 In municipalities with high tourism incidence, occupational risk 
does not make any difference in PrevenSida exclusive coverage
(16% versus 20%, X2 0.8). In a significantly different situation(X2het
6.2, p=0.01), MARPs in lower risk occupations in municipalities with 
low tourism incidence have reduced their PrevenSida exclusive 
coverage to 9%, while those in higher risk occupations have two 
times more probability of receiving PrevenSida exclusive care in
comparison to those with lower risk occupations (17% versus 9%, 
OR 2.0, CI 95% 1.3-3.1, X2 8.7).
 In municipalities with a high incidence of cargo transport, the 
PrevenSida exclusive coverage gap is wider and MARPs in higher 
risk occupations have three times more probabilities of receiving 
PrevenSida exclusive care (22% versus 8%, OR 3.2, CI 95% 1.8-7.3, X2
10.5) .In a significantly different situation (X2het 7.2, p=0.007), 
occupational risk does not make any difference in PrevenSida 
exclusive coverage in municipalities with low incidence of cargo 
transport (15% versus 14%, X2 0.1).
• MARPs working in tourism-related activities have a 50% higher 
probability of receiving exclusive care, as compared to MARPs not 
working in tourism-related activities (17% versus 12%, OR 1.5, CI 95%
1.1-2.0, X2 4.4). This association is modified by the incidence of 
different conditional factors: 
 I nMARPs with primary schooling, those working in tourism￾related activities increase their PrevenSida exclusive coverage 
(19%), whereas MARPs not working in tourism-related activities 
reduce their PrevenSida exclusive coverage to half (6%). Hence, 
tourism-related MARPs have 3.7 times more probability of 
receiving PrevenSida exclusive care than those unrelated to 
tourism (OR 3.7, CI 95% 1.9-8.6, X2 12.1). In a significantly different 
condition (X2het8.4, p=0.003), tourism-related work does not make 
any difference in PrevenSida exclusive coverage for MARPs with 
a schooling level greater than 6th grade (16% versus 15%, X2 0.1). 
 For MARPs working in higher risk occupations, tourism-related 
work does not make any difference in the probability of 
receiving PrevenSida exclusive coverage (16% versus 15%, X2 0.1). 
However, in a significant contrast (X2het 4.7, p=0.03), MARPs in 
lower risk occupations unrelated to tourism reduce their 
coverage, so the probabilities of PrevenSida exclusive care for 
MARPs working in tourism-related activities increases 2.5 times 
(17% versus 8%, OR 2.5, CI 95% 1.5-4.5, X2 9.4).
Graph 54 
PrevenSida Exclusive Coverage
By occupational risk level and participation in 
organizations 
Higher risk Lower risk
All MARPs Organized Unorganized
Graph 55 
PrevenSida Exclusive Coverage
By level of occupational risk and municipal 
incidence of HIV positive tests
Higher risk Lower risk
All MARPs Municipality with Municipality with
 High HIV + tests Low HIV + tests
Graph 56 
PrevenSida Exclusive Coverage
By level of occupational risk and municipal 
tourism incidence
Higher risk Lower risk
All MARPs Municipalities Municipalities 
 with high with low
 incidence incidence 
 of tourism of tourism 
 
Assessing PrevenSida Project Coverage - “2014 Mid-Term Performance Evaluation of HIV/USAID Nicaragua Program”
37
 In the capital city, tourism-related work does not make any 
significant difference in PrevenSida exclusive coverage (10% versus
13%, X2 0.6). However, in a significant contrast(X2het 4.4, 
p=0.03),the probability of receiving exclusive care from 
PrevenSida is 85% higher for MARPs working in tourism-related 
activities, as compared to those not working in tourism-related 
activities in departmental capitals and peripheral municipalities
(20% versus 12%, OR 1.9, CI 95% 1.3-2.8, X2 8.3).
 In MARPs living in municipalities with a high incidence of 
PrevenSida HIV positive tests, tourism-related work does not 
make any difference in the probability of receiving PrevenSida 
exclusive care (11% versus 12%, X2 0.3). In a significantly different 
times their probabilities of receiving exclusive care from 
condition (X2het 9.0, p=0.02), MARPs working in tourism-related 
activities increase their coverage significantly and multiply by 2.6 
PrevenSida in municipalities with a low PrevenSida HIV test 
positivity (26% versus12%, OR 2.6, CI 95% 1.7-4.5, X2
12.2).
Tourism related Not related to tourism
Graph 59 
PrevenSida Exclusive Coverage
By tourism-related work and municipal HIV test 
positivity rate
Tourism related Unrelated to tourism
All MARPs Municipality with Municipality with
 High HIV + tests Low HIV + tests
Graph 57 
PrevenSida Exclusive Coverage
By tourism-related work and schooling level
Tourism related Unrelated to tourism
All MARPs Up to Beyond 
 6
th grade 6
th grade 
Graph 58 
PrevenSida Exclusive Coverage
By tourism-related work and occupational risk 
level 
Tourism related Unrelated to tourism
All MARPs High Risk Low Risk
 Occupation Occupation
Assessing PrevenSida Project Coverage - “2014 Mid-Term Performance Evaluation of HIV/USAID Nicaragua Program”
38
IV. APPROXIMATIONS TO PREVENSIDA COVERAGE EFFECTIVENESS
Major associations between PrevenSida preventive action coverage and various approximation indicators for 
program deliverables are shown in the context of interviewed MARP HIV protective capacities. 
It is not an impact assessment, inasmuch as that was not the purpose of the study. For this reason, the design of the 
sample and questionnaires does not strictly match the methodological demands of such an assessment. 
Notwithstanding the above, the structure of the coverage actions permits to differentiate the preventive services 
provided by the program from those provided by other agencies, facilitating a comparison of exposures in terms of 
their association with proxy indicators regarding the protection status of MARPs, which are seen as program 
results. 
In this regard, it should be viewed as an evidence-based exercise that seeks to respond to the question included in 
the TORs on whether PrevenSida actions are “successful”, and whether they could be replicated or scaled up in 
the context of the USAID program or elsewhere. To this end, an analysis was carried out in two methodological 
phases: 
• Searching for associations between PrevenSida coverage and results indicators that would allow to determine 
whether it is “successful”, to the extent that MARP exposure to PrevenSida care creates some difference 
(modification) in the outcome indicators. 
• Searching for conditional factors that modify the significant associations found (CoverageResults), in order 
to explore conditions that facilitate or hinder its replication or scaling. 
4.1 PrevenSida/HIV Test Positivity
The starting point is a municipal-level risk rating carried out by the 
PrevenSida team, based on a review of rapid tests performed in the last 
three years, according to which the positivity rate was estimated. The 
numerator were the reactive tests in each municipality and the 
denominator was the total number of tests performed in that same 
municipality. Municipalities with 0.04% or more reactive tests were 
ranked as “high positivity”, and the remainder were rated as “low 
positivity.” This dichotomous variable (high=1, low=2) was assumed to 
be a mesovariable, namely, the entire MARP cluster was assigned a 
value corresponding to the relevant municipality. For this reason, it 
cannot actually be assumed as an outcome indicator because that 
would require an individual value resulting from a specific exposure, 
but it is included in this chapter to visualize the first effectiveness 
factor, which is targeting, i.e., matching risk distribution and coverage 
distribution.
Fifty-nine percent of MARPs live in high positivity municipalities, 
broken down as follows: 55% of MSM, 65% of SW, and 61% of TRANS. 
This is a predictable situation given that it was a sample selection 
criterion. No significant differences were found in PrevenSida global 
coverage between high and low HIV test positivity municipalities, or in 
total MARP sample or in any of the specific groups. That is to say, no 
Assessing PrevenSida Project Coverage - “2014 Mid-Term Performance Evaluation of HIV/USAID Nicaragua Program”
39
territorial prioritization related to test positivity has been in place in 
PrevenSida overall coverage. 
In turn, PrevenSida exclusive coverage is lower in municipalities with a 
high HIV test positivity than in municipalities with low positivity, which 
means that MARPs living in higher positivity municipalities have a 35% 
lower probability of accessing PrevenSida exclusive care(13% versus18%, 
OR 0.65, CI 95% 0.5-0.8, X2 8.0).This association was even stronger in the 
case of SWs (OR 0.37, CI 95% 0.3-0.7, X2 11.6)and TRANS (OR 0.4, CI 95%
0.2-0.8, X2 8.1). No difference was found in PrevenSida exclusive 
coverage between high and low HIV test positivity amongst MSM (X2
0.2). 
In MARPs as a whole, the reverse association between municipal HIV 
test positivity and PrevenSida exclusive coverage is modified: 
• MARPs under 20years of age further reduce their exclusive 
coverage (8%) when they live in higher positivity municipalities, 
whereas those living in lower positivity municipalities have an 
increased coverage(23%). Hence, younger MARPs living in high 
positivity municipalities reduce their probability of receiving 
exclusive care from PrevenSida to one third as compared to 
younger MARPs living in lower positivity municipalities (OR 0.3, CI 
95% 0.1-0.6, X2 9.7). In a significant contrast(X2het 4.5, p=0.03),the 
municipal HIV test positivity rate does not make a significant 
difference in PrevenSida exclusive coverage (14% versus 17%, X2 2.8)
for MARPs over 19 years old.
• In MARPs working in higher risk occupations, the level of municipal 
HIV test positivity does not make any difference in PrevenSida 
exclusive coverage (16% versus 17%, X2 0.1). In a significant contrast
(X2het 6.5, p=0.01), in MARPs working in lower risk occupations and
living in higher HIV positivity municipalities have reduced their 
coverage to 8%. Hence, the probabilities of receiving exclusive care 
from PrevenSida are 2.7 times lower for MARPs living in high HIV 
positivity municipalities, as compared to those in low HIV positivity 
municipalities (8% versus 20%, OR 0.37, CI 95% 0,2-0.6, X2 12.1).
• In MARPs working in tourism-related activities, their PrevenSida 
exclusive coverage is reduced in high HIV positivity municipalities, 
and increases in those municipalities with low positivity. Hence, the 
probabilities of receiving exclusive care from PrevenSida are three 
times lower in high positivity municipalities than in municipalities 
with low positivity (11% versus 26%, OR 0.36, CI 95% 0,2-0.6, X2 14.2). In 
contrast (X2het 9.0, p=0.002), the level of municipal HIV test positivity 
does not make any difference in the odds of receiving PrevenSida 
exclusive care in MARPs working in activities unrelated to tourism 
(13% versus 12%, X2 0.13).
Graph 63
PrevenSida Exclusive Coverage 
by Municipal HIV test positivity rate and risk 
occupational level 
High + Rate Low+ Rate
All MARPs High Risk Low Risk
 Occupation Occupation
Graph 64
PrevenSida Exclusive Coverage
by municipal HIV test positivity rate and 
tourism related work
All MARPs Tourism Not related 
 Related to Tourism
High + Rate Low+ Rate
Graph 62 
PrevenSida Exclusive Coverage 
by Age and Municipal HIV test positivity rate
High + Rate High + Rate
 All MARPs <20 years 20 years 
 or more
Assessing PrevenSida Project Coverage - “2014 Mid-Term Performance Evaluation of HIV/USAID Nicaragua Program”
40
• A similar situation is found in: 
 MARPs working in activities related to mobile populations (11%
versus 22%, OR 0.4, CI 95% 0,3-0.7, X2 10.8), as opposed (X2het 4.5, 
p=0.03) to MARPs working in activities not related to mobile 
populations (14% versus 15%, X2 0.15). 
 MARPs living in municipalities with a high incidence of mobile 
populations (14% versus 21%, OR 0.6, CI 95% 0,4-0.8, X2 6.5), as 
opposed to (X2het 4.7, p=0.03) to MARPs living in municipalities 
with a low incidence of mobile populations (12% versus 9%, X2 1.0). 
 MARPs living in municipalities with a high incidence of cargo 
transport (11% versus 21%, OR 0.5, CI 95% 0,3-0.7, X2 12.1), as 
opposed to (X2het 5.1, p=0.02) MARPs living in municipalities with a 
low incidence of cargo transport (14% versus 14%, X2 0.02). 
No association was found between the municipal HIV test positivity 
rate and coverage of the different communication activities offered by 
PrevenSida. PrevenSida promotional contacts have the same coverage, 
regardless of the municipal positivity rate (27% versus 26%, X2 0.03), 
which is similar to the coverage of PrevenSida training events (24% 
versus 27%, X2 1.6) and peer group participation (14% versus 15%,X2 0.12). 
The only exception is VCT coverage provided by PrevenSida, which 
was lower in high positivity municipalities, as compared to those with 
low positivity (13% versus 20%, OR 0.58, CI 95% 0.4-0.8, X2 13.4). Only two 
factors modifying this association were found: 
 MARPs living with any of their parents in the same home. 
Those living only with their mothers (33%) increase the 
PrevenSida VCT coverage gap between municipalities with high 
and low HIV test positivity (10% versus 24%, OR 0.37, CI 95% 0,2-
0.6, X2
15.5).In contrast (X2het 6.6, p=0.01), there is no difference 
in PrevenSida VCT coverage (16% versus 16%, X2 0.01) in 
MARPs with a paternal presence at home (father and mother 
21%, father only 3%). 
 Municipal mobile population incidence. Differences in PrevenSida 
VCT coverage between high and low HIV test positivity 
municipalities is similar to the overall situation in high mobility 
municipalities (16% versus 26%, OR 0.55, CI 95% 0,4-0.8, X2 10.8).In 
contrast (X2het 17.8, p=0.00), the VCT coverage gap between high 
and low HIV test positivity municipalities is inverted in 
municipalities with low mobile population incidence (10%versus
1%, OR 7.7, CI 95% 2.7-14.2, X2 10.8).
Graph 65 
PrevenSida Activity Coverage 
by municipal HIV test positivity rate
High + Rate Low+ Rate
Contacts Training Peer Groups VCT
Graph 66 
PrevenSida VCT Coverage
by municipal HIV test positivity rate, cohabitation 
with parents and municipal incidence of mobile 
population 
Low+ Rate High + Rate
Only Paternal High Low 
Maternal (F&M/F) Mobility Mobility
Assessing PrevenSida Project Coverage - “2014 Mid-Term Performance Evaluation of HIV/USAID Nicaragua Program”
41
4.2 New Infection Risk Index: PLACE
As indicated in the methodology chapter, PLACE focuses attention on 
HIV epidemic control in the transmission chain driven by new sexual 
partners that a person has in a particular period of time, and 
complemented with the number of sexual partners this same person 
has during the same period. A high risk level is assigned to those who 
have had a new sexual partner and/or more than one sexual partner in 
the past month; a moderate risk level is assigned to those who have 
had a new sexual partner and/or more than one sexual partner in the 
past year; and a low risk level is assigned to those who have not had 
sex or have had only one stable sexual partner in the past year.
Since this study is focused on MARPs, 78% of them have been ranked 
as “high risk”, 18% as “moderate risk”, and only 4% qualify as “low 
risk”. At the territorial level, the highest rate of “high risk” is found in 
Managua (84%) and the lowest in the West region (72%); the other 
three regions revolve around the average (South Pacific region 76%, 
Central region 77%, Caribbean 78%). In population terms, the highest 
rate of “high risk” is found in SW (93%), the lowest is found in MSM
(67%) and a mean rate is reported for TRANS (80%). 
No association was found between the PLACE risk level with 
PrevenSida exclusive coverage or with communication activities, as 
compared to those provided by other agents (Exclusive coverage X2 0.02, 
Contacts X2 2.1, VCT X2 0.0, Events X2 0.6, Peer groups X2 0.7). 
Nevertheless, we did find that “PLACE high risk” rate can be modified 
through participation in training or exchange events, and through peer 
group socialization offered by all providers, as compared to MARPs 
that were not covered by these two actions. MARPs involved in some 
peer group session have 37% less probability of being rated as “PLACE 
high risk”, as compared to MARPs that have not participated in any 
peer group session (73% versus 81%, OR 0.63, CI 95% 0,5-0.8, X2
12.4).
The association between a reduced “PLACE high risk” rate and peer 
group participation was only explained by: 
o Having had the first sexual relation before age 16. In this segment, 
participation in peer groups reduces their probability of being 
included in “PLACE” high risk” almost to half (48%), as compared 
to MARPs that have not participated in peer groups (74% versus 85%, 
OR 0.52, CI 95% 0.4-0.7, X2 20.2).In contrast, (X2het 8.3, p=0.004), in
MARPs that initiated their sexual life at 16 years or older, peer 
group participation does not make any difference in the “PLACE 
high risk” rate (70% versus 67%, X2 0.5). 
Graph 68 
“PLACE” High Risk
by interpersonal communication activities and 
exclusive covered provided by PrevenSida
“Other Agents”
Exclusive Contact VCT Training Peer
Coverage Events/ Groups
 Internet
Graph 69 
“PLACE” High Risk
By exposure to four interpersonal 
communication activities
Exposed Not exposed
Contact VCT Training Peer Groups
 Events/Inter 
Assessing PrevenSida Project Coverage - “2014 Mid-Term Performance Evaluation of HIV/USAID Nicaragua Program”
42
o Living with parents. The odds of being included in “PLACE high 
risk” is reduced by 61% (62% versus 80%, OR 0.39, CI 95% 0.2-0.7, X2
14.5) in MARPs living with their mother and father and participating
in peer groups. In contrast (X2het 5.8, p=0.01), peer group 
participation does not modify the “PLACE high risk” rate (75% versus 
78%, X2 0.5) in MARPs living only with their mother.
The above situation results from the likelihood that preventive services 
are able to reduce the number of sexual partners or the number of 
new sexual partners that MARPs have had in the past month. This is 
why the potential association between the coverage of each of the four 
activities included in the global coverage index and PrevenSida 
exclusive coverage was reviewed, finding the following evidence:
• MARPs exposed to some promotional contact regarding HIV do 
not significantly modify their average number of sexual partners 
(14.1 versus 14.3, X2K-W 0.02), nor the average of new sexual partners 
(5.1 versus 8.5, X2K-W 1.4), as compared to MARPs unexposed to 
promotional contacts. However, MARPs exposed to PrevenSida 
network contacts significantly reduce their average number of 
sexual partners, as compared to MARPs exposed to contacts from 
“other providers” (12.2 versus 17.6, X2K-W 5.7, p=0.02). This association 
was not found in the case of new sexual partners (7.0 versus 8.5, X2K-W 
0.56). Nevertheless, no association was found between the number 
of contacts and the number of partners (X2K-W 3.4, p=0.48) or number 
of new partners (X2K-W 2.4, p=0.72).
• MARPs participating in peer socialization groups significantly 
reduced both their average number of sexual partners (8.4 versus
17.6, X2K-W 40.4, p=0.000) and their average number of new sexual 
partners (6.2 versus 8.0, X2K-W 19.2, p=0.000) in the past month, in 
contrast with MARPs that have not participated in any peer group 
session. No difference was found in the average number of sexual 
partners (6.5 versus 9.7, X2K-W 0.6) or number of new partners (5.3 
versus 6.4, X2K-W 0.1) in relation to exposure originating from 
PrevenSida or “other agents”.
• MARPs that have participated in training events significantly reduced 
both their average number of sexual partners (10.0 versus 18.0, X2K￾W 34.8, p=0.000) and their average number of new sexual partners
(5.4 versus 9.0, X2K-W 26.6, p=0.000) in the past month, in 
comparison with MARPs that have not participated in any training 
event. No difference was found in the average number of sexual 
partners (8.8 versus 12.0, X2K-W 0.27) or number of new partners
(4.6 versus 6.1, X2K-W 0.13) in relation to exposure originating from 
PrevenSida or “other agents”.
Total New Total New
Partners Partners Partners Partners
Graph 71 
Average sexual partners, past month, 
by exposure to promotional contacts
 With/Without PrevenSida/“Others” 
……….Coverage
Total New Total New
Partners Partners Partners Partners
With/Without Coverage PrevenSida/“Others”
Graph 72 
Average sexual partners, past month,
by participation in peer groups
With/Without Coverage PrevenSida/“Others”
Total New Total New
Partners Partners Partners Partners
 Total New Total New
 Partners Partners Partners Partners 
Graph 73 
Average sexual partners, past month, by 
participation in training events
Assessing PrevenSida Project Coverage - “2014 Mid-Term Performance Evaluation of HIV/USAID Nicaragua Program”
43
• MARPs that have received VCT in the past year significantly reduce 
their average number of sexual partners (10.0 versus 16.0, X2K-W 
12.0, p=0.005), although they do not significantly modify the average 
number of new sexual partners (6.5 versus 7.7, X2K-W 0.22) in the 
past month, as compared to MARPs that have not received VCT in 
the past year. No difference was found in the average number of 
sexual partners (14.0 versus 14.3 X2K-W 0.4) or number of new 
sexual partners (4.5 versus 7.8, X2K-W 0.22) in relation to exposure 
originating from PrevenSida or “other agents”.
• Significant differences were found between the average number of 
sexual partners (12.5 versus 14.0 versus 15.0, X2K-W 6.5, p=0.03, 2 GL) and 
the average number of new sexual partners (5.0 versus 7.0 versus 8.4, 
X2K-W 5.8, p=0.05, 2 GL) in the past month among MARPs, depending 
on whether they received PrevenSida exclusive coverage, mixed 
coverage, or exclusive coverage from “other agents.” 
4.3 Consistent Condom Use 
A “consistent condom use index” was developed based on the answers 
to two interview questions inquiring whether: 
• A condom was always used with a stable partner, occasional 
partners and new partners.
• A condom had been used in all sexual acts performed in the last 
sexual relation. 
Condom use was rated as “consistent” when it was always used with 
the three types of partners and in all sexual acts during the last 
relationship. 
Thirty-nine percent of interviewed MARPs consistently use a condom, 
a percentage obtained by combining 49% of MARPs that always use a 
condom with all partners and 68% of MARPs that used a condom in all 
sexual acts during the last relationship. The resulting global reduction 
mainly results from 29% of MARPs that adequately used a condom in 
their last relationship, but claim that they do not always use it with all 
partners. A major contribution is that half of SW do not use a condom 
with all partners (particularly with their stable partner). This same 
condition is found in 19% of MSM and 17% of TRANS. Ten percent of
MARPs use a condom with all their partners, although they did not 
adequately use it in their last sexual relationship (17% in TRANS, 11% 
in MSM, and 5% in SW). Lastly, 22% of MARPs do not use a condom 
with all their partners and did not adequately use it in their last sexual 
relationship (25% of MSM, 21% of TRANS, and 18% of SW).
Graph 76 Condom Use
In all sexual acts in last relationship
Always with all partners (stable, occasional, new)
Consistent use
Total MSM SW TRANS
Total Partners New Partners
Graph 74 
Average number of sexual partners in the past 
month, by VCT carried out in the past year
With/Without Coverage PrevenSida/“Others”
Total New Total New
Partners Partners Partners Partners
12.5
14.0 15.0
5.0
7.0
8.4
PrevenSida Mixed "Others Agents"
Total partners New partners
Graph 75 
Average number of sexual partners in the past 
month by exclusive coverage provided by 
PrevenSida and “Other Agents” 
Assessing PrevenSida Project Coverage - “2014 Mid-Term Performance Evaluation of HIV/USAID Nicaragua Program”
44
Consistent condom use rates are higher in the South Pacific and 
Caribbean regions (46% and 45%), and lower in Managua (30%). The 
rate in the other two regions is close to average (West region at 42% 
and Central region at 37%). In MSM and TRANS, the rate is 45%, 
although significantly reduced in SW, reaching just 26%.
The type of global preventive service coverage of each MARP clearly 
makes a difference in the rates of consistent condom use (X2 22.1, 
p=0.0000, 3 GL). In MARPs that have not received any of the 4 activities
included in the coverage, the rate is 34%, while the rate in MARPs with 
exclusive coverage from “other agents” is 33%. The rate is 45% among 
MARPs with mixed coverage, and 46% with PrevenSida exclusive 
coverage. In this same context, the significance of potential contrasts 
between the four global coverage groups is worth noting: 
• No significance was found in consistent condom use rate 
differences between MARPs exposed to some of the 4 activities and 
those exposed to none (39% versus 34%, X2 1.5).
• MARPs exposed to PrevenSida exclusive coverage have 42% more
probability of achieving consistency in condom use, as compared to 
MARPs in the other three coverage groups (46% versus 37%, OR 1.4, 
CI 95%1.1-1.9, X2 5.4). 
• MARPs exposed to PrevenSida exclusive coverage have 73% more
probability of achieving consistency in condom use, as compared to 
MARPs exclusively served by “other agencies” providing preventive 
services (46% versus 33%, OR 1.7, CI 95% 1.2-2.4, X2 11.4). This 
association was stronger where: 
 MARPs have an education lower than 6th grade, the odds of 
consistent condom use is four times higher when they are 
exclusively served by PrevenSida than those served by “others”
(54% versus 22%, OR 4.2, CI 95% 1.9-9.2, X2 13.2). In contrast (X2het
6.0, p=0.01), the odds of consistent condom use in MARPs with a 
schooling level of 6th grade or higher is just 40% higher in those 
served by PrevenSida (44% versus 35%, OR 1.4, CI 95% 1.0-2.1, X2 
4.3).
 MARPs are heads of household, in which case PrevenSida 
exclusive coverage increases almost three times the odds of 
consistent condom use, as compared to MARPs covered by 
“other” providers (46% versus 23%, OR 2.8, CI 95% 1.5-5.3, X2 13.2).
In contrast (X2het 4.4, p=0.03), PrevenSida coverage does not make 
any significant difference in respect of coverage provided by 
“others” in MARPs that are not heads of household (46% versus
38%, X2 2.5).
 MARPs are rated as “high risk of new transmissions” (PLACE 
index), in which case PrevenSida exclusive coverage doubles the
Graph 77 
Consistent Condom Use by Region
Managua West South Central Caribbean
 Pacific
Graph 78 
Consistent Condom Use 
by Preventive Coverage Provided by Agents
No coverage PrevenSida Mixed Other agents
“Other Providers”
Graph 80
Consistent Condom Use 
by Exclusive Coverage from PrevenSida and 
Other Agents, and head of household
“Other Providers”
 All MARPs PLACE High Risk PLACE Low/
 Medium Risk
Graph 79
Consistent Condom Use by Exclusive 
Coverage of Preventive Service Providers and 
Schooling
All MARPs Below 6th grade 6th grade and beyond
Assessing PrevenSida Project Coverage - “2014 Mid-Term Performance Evaluation of HIV/USAID Nicaragua Program”
45
odds of consistent condom use, as compared to MARPs with the 
same characteristics covered by “other” preventive service 
providers (49% versus 31%, OR 2.2, CI 95% 1.5-3.2, X2 18.3). In 
contrast (X2het 7.5, p=0.006), no significant difference was found in 
consistent condom use between PrevenSida coverage and 
“other” preventive service providers in MARPs rated as 
moderate/low risk in the PLACE index (33% versus 41%, X2 0.8).
 MARPs that do not participate in any organization because their 
preventive coverage is provided exclusively by the PrevenSida 
network increase 2.4 times their odds of consistent condom use, 
as compared to unorganized MARPs served by “other agents” 
(52% versus 31%, OR 2.4, CI 95% 1.5-3.8, X2 16.6). In contrast (X2het
5.6, p=0.02), no significant difference was found in consistent 
condom use rates between MARPs served exclusively by 
PrevenSida and those served by “other” preventive service 
providers in the MARP segment that does not participate in any 
organization (39% versus 37%, X2 0.2). 
 In MARPs that identify any organization that could provide HIV￾related support because PrevenSida exclusive coverage doubles 
their odds of consistent condom use, as compared to those 
served by “other” providers (51% versus 32%, OR 2.2, CI 95% 1.4-
3.3, X2 14.6). In contrast (X2het 5.0, p=0.02), no difference was found 
in consistent condom use rates between MARPs served by 
PrevenSida and those served by “other” preventive service 
providers in MARPs that are not aware of any HIV-related 
support organization (32% versus 33%, X2 0.04). 
• Some preventive activities provided by PrevenSida affect consistent 
condom use: 
 Coverage of PrevenSida promotional contacts increases by 60% 
the odds of consistent condom use (47% versus 38%, OR 1.4, CI 
95% 1.1-1.9, X2 7.0), as compared to “other providers”. No 
difference was found, however, in consistent condom use rate in 
relation to the number of promotional contacts (X2 0.5, p=0.47).
 Participation in training events provided by PrevenSida increases 
by 64% the odds of consistent condom use (47% versus 35%, OR 
1.6, CI 95% 1.2-2.3, X2
9.8), as compared to “other providers”. 
 Participation in peer group socialization sessions provided by 
PrevenSida does not make any difference in odds of consistent 
condom use (43% versus 39%, X2 0.9) in comparison to “other 
providers”.
 VCT care provided by PrevenSida increases by 67% the 
probabilities of consistent condom use (49% versus 37%, OR 1.7, CI 
95% 1.2-2.3, X2 10.8)in comparison to “other providers”.
Graph 81
Consistent Condom Use by PrevenSida or 
“other” exclusive coverage and “PLACE” risk 
level
Other Providers
All MARPs “PLACE” High “PLACE” Low
 Risk Risk
Graph 82
Consistent Condom Use
by PrevenSida or “Other” exclusive coverage 
and participation in organization
Other Providers
 All MARPS Participate in Do not participate
 organization in organization
Graph 83 
Consistent Condom Use
by Action Coverage by Agent
Other Providers
Promotional Training Peer Group VCT
Contact 
Assessing PrevenSida Project Coverage - “2014 Mid-Term Performance Evaluation of HIV/USAID Nicaragua Program”
46
 MARPs that often get their condom supply from PrevenSida 
increase by 38% their odds of consistent condom use (46% versus
38%, OR 1.4, CI 95% 1.0-1.8, X2
4.9) in comparison to “other 
providers”.
4.4 History of Sexually Transmitted Infections
Twenty percent of MARPs claim a history of STI. The highest rates are 
found in Managua (25%) and the Caribbean (21%), and the lowest in 
the South Pacific region (16%). Rates in the West and Central regions 
remain average (17% and 18%). By type of MARP, SW have the highest 
rate at 29%, and MSM the lowest at 14%, while TRANS have a rate of 
20%. This means that SW have twice the odds of an STI history than 
the other two groups (29% versus 16%, OR 2.1, CI 95% 1.7-2.9, X2 32.3).
STI history rate is significantly higher in municipalities with higher 
PrevenSida HIV test positivity (23% versus 15%, OR 1.7, CI 95% 1.3-2.3, X2
12.8). This association is even stronger in municipalities with a high 
incidence of mobile populations (24% versus 13%, OR 2.1, CI 95% 1.4-3.1, 
X2 14.5). In a significant contrast (X2het 5.5, p=0.02), test positivity rates do 
not make any difference in STI history in municipalities with a low 
incidence of mobile populations (23% versus 22%, X2 1.0).
There are no significant differences in STI history rates in the four 
preventive action segments of global coverage. Rates are 20% in 
MARPs without any coverage in the four activities, 18% in MARPs
under mixed coverage, 19% in MARPs exposed exclusively to 
PrevenSida, and 21% in MARPs covered only by “other providers” (X2
0.7).
No significant differences were found between the potential coverage 
in the three MARP types, although SW always have the highest STI 
history rates in all coverage segments. The TRANS group has the 
lowest rate under PrevenSida exclusive coverage. 
Except for the VCT case, no significant differences were found in STI 
history rates in a comparison between MARPs exposed to each 
prevention action and MARPs not exposed to the same actions: 
• Having been exposed to some HIV promotional contact does not 
make any difference in STI history rate (19% versus 22%, X2 2.5). 
• There was no difference in STI history rate as a result of 
participating in any training event (19% versus 21%, X2 1.1).
Graph 84 
STI History by regions
Managua West S. Pacific Central Caribbean
Graph 85
STI History by Municipal HIV test positivity 
rate and mobile population incidence rate
High + Rate Low + Rate
 All MARPs High Mobile Low Mobile
 Population Rate Population Rate
Graph 86
STI History by type of MARP and coverage 
segment by agent
 SW TRANS  MSM
 No PrevenSida Mixed “Others” 
coverage
Graph 87 
STI history by exposure to preventive 
activities
 
Exposed Unexposed
Promotional Training Peer Group VCT
Contact Event
Assessing PrevenSida Project Coverage - “2014 Mid-Term Performance Evaluation of HIV/USAID Nicaragua Program”
47
• Having participated in peer socialization groups in relation to HIV is 
not associated with an STI history rate (19% versus 20%, X2 0.2).
• Having carried out VCT is associated with a 2.4 times higher 
probability of having an STI history, as compared to MARPs that 
have not carried out the test (21% versus 10%, OR 2.4, CI95% 1.6-4.2, 
X2 16.1). 
No differences in STI history rates were found between MARPs
covered by each preventive action provided by PrevenSida and those 
covered by “other providers”.
• In the case of contacts, STI history rates were 16% with PrevenSida 
and 20% with other providers (X2 2.4). 
• Similarly, when workshops/events have been offered by PrevenSida 
or “other providers” (17% versus 21%, X2 2.1).
• The same situation arises when peer groups have been promoted 
by PrevenSida or “other providers” (19% versus19%, X2 0.0). 
• The STI history rate was similar with VCT provided by PrevenSida 
or by “other providers” (19% versus 23%, X2 1.5). 
In MARPs that have recognized any organization of the PrevenSida 
network as an HIV support organization, the STI history rate is 
significantly lower than in MARPs that have recognized “other 
organizations.” Thus, MARPs that recognize PrevenSida have a 38% 
lower probability of having an STI history, as compared to MARPs that 
recognize “other organizations” (17% versus 25%, OR 0.62, CI 95% 0.4-0.8, 
X2 9.2).
STI history rate is significantly lower in MARPs using a condom 
consistently than in those not doing so. Consistent condom use 
reduces by 32% the odds of having an STI history (16% versus 22%, OR 
0.68, CI 95% 0.5-0.9, X2 7.5). The protective effect of consistent condom 
use is even greater when: 
 MARPs are less than 24 years old because they reduce by 60% 
the odds of having an STI history when they use a condom 
consistently, as compared to those not using a condom 
consistently (7%versus 17%, OR 0.4, CI 95% 0.2-0.7, X2 11.1). In 
contrast (X2het 5.5, p=0.02), consistent condom use by MARPs 
older than 24 years did not make any difference in STI history 
rates (24% versus 26%, X2 0.8).
 MARPs do not have children, in which case consistent condom 
use reduces by 45% the odds of having an STI history, as 
compared to those not using a condom consistently (12% versus
19%, OR 0.55, CI 95% 0.4-0.8, X2 9.3). In contrast (X2het 6.5, p=0.01),
consistent condom use by MARPs with children did not make 
any difference in STI history rates (29% versus 26%, X2
0.5). 
Graph 88 
STI History by coverage of actions by agent
Others PrevenSida
Promotional Training Peer Group VCT
Contact Event
Graph 88 
STI History by coverage of actions by agent
Others PrevenSida
 Promotional Training Peer Group VCT
 Contact Event
Graph 89
STI history 
by consistent condom use and age
Consistent Use Inadequate Use
All MARPs Under 24 years 24 years and more
Assessing PrevenSida Project Coverage - “2014 Mid-Term Performance Evaluation of HIV/USAID Nicaragua Program”
48
 The probability of having an STI history is reduced by 41% in 
MARPs with a schooling level of 6th grade or more and 
consistent condom use, as compared to those not using a 
condom consistently (13% versus 20%, OR 0.59, CI 95% 0.4-0.8, X2
10.1). In contrast (X2het 6.2, p=0.01), consistent condom use does 
not make any difference in STI history rates in MARPs with a 
schooling level below 6th grade (37% versus 30%, X2 1.1). 
 MARPs have carried out VCT offered by a PrevenSida network 
organization and use a condom consistently. In this case, they 
reduce to less than one third the odds of having an STI history, 
as compared to those not using a condom consistently (9% versus
28%, OR 0.26, CI 95% 0.1-0.6, X2 12.2). In contrast (X2het 7.5, p=0.006),
consistent condom use by MARPs that have carried out VCT 
with “other agents” does not make any difference in STI history 
rates (21% versus 24%, X2 1.0).
 MARPs identify a PrevenSida network organization that provides 
HIV support. In this segment, the odds of an STI history is 
slightly reduced by more than half through consistent condom 
use, as compared to those not using it adequately (11% versus
22%, OR 0.45, CI 95% 0.3-0.7, X2 9.7). In contrast (X2het 4.0, p=0.04),
consistent condom use by MARPs that recognize any HIV 
support organization does not make any difference in STI history 
rates (19% versus 22%, X2 1.0).
4.5 Effectiveness Measured by Number of Promotional Contacts
Given the relevance of the number of contacts in PrevenSida planning and management models, this section has 
been reserved to present the findings related to the effects associated with an increased number of contacts 
received by each MARP served by PrevenSida, some of which have already been specifically set forth in previous 
chapters.
Based on the assumption that a relationship exists between the number of contacts and the probabilities of 
modifying risk levels or increasing protective behaviors, this association was explored to determine whether the 
number of contacts could have the effect of reducing the likelihood of HIV infection: 
• No significant difference was found in the number of contacts provided by PrevenSida, as per municipal HIV 
test positivity rate, or in the one-to-one increase trend (X2 0.95, 3Gl), or in the contrast of 1-2 contacts versus 3 
and more contacts (X2 1.2). In other words, the number of contacts to the same MARP is not modified by the 
territorial risk level (high municipal rate of positive tests). 
• No association was found between the number of contacts provided by PrevenSida (one-to-one increase 
contrast, X2 1.3, 3GL) and the percentage of MARPs ranked as “PLACE high risk”. Having three or more 
PrevenSida contacts is associated with a higher share of MARPs ranked as “PLACE high risk”, as compared to 
MARPs with one or two PrevenSida contacts (77% versus 67%, OR 1.7, CI 90% 1.0-2.7, X2 3.0). This means a 
contact increase is associated with targeting the search for MARPs on concentration venues with a higher risk 
level. Delving deeper into this indicator, no significant differences were found in the average number of sexual 
Graph 90
STI history by consistent condom use and 
schooling 
Consistent Use Inadequate Use
All MARPs 6th grade or more Less than 6th grade
Graph 91
STI history by consistent condom use and
VCT PrevenSida
Consistent Use Inadequate Use
All MARPs VCT PrevenSida VCT Others
Assessing PrevenSida Project Coverage - “2014 Mid-Term Performance Evaluation of HIV/USAID Nicaragua Program”
49
partners during the past month per the number of PrevenSida contacts received, or in the one-to-one increase 
contrast (X2kw 0.7, p=0 .88, 3Gl), or in the contrast of 1-2 contacts versus 3 or more (X2kw 0.6, p=0.42). The 
same lack of association is reported in the average number of new sexual partners, in both the one-to-one 
increase (X2kw 0.65, p=0.88) and the contrast of 1-2 contacts versus 3 or more (X2kw 0.5, p=0.46).
• No association was found between the number of contacts provided by PrevenSida and the consistent condom 
use rate. This applies to both the one-to-one increase contrast (X2kw 1.2, p=0.25, 3Gl) and to the contrast 
between one contact and two or more contacts (X2 0.9), and also to the contrast of 1-2 contacts versus 3 or
more (X2 0.6). 
• No association was found between the number of contacts provided by PrevenSida and an STI history rate, or 
in the one-to-one increase contrast (X2kw 0.3, p=0.53, 3Gl), or in the contrast of 1-2 contacts versus 3 or more
contacts (X2 0.5).
The possible association between the number of contacts provided by PrevenSida and the probabilities of MARPs 
accessing PrevenSida network global preventive care, or the other specific activities covered, was explored: 
• The increase in the number of contacts provided by PrevenSida significantly increases the odds of participating 
in PrevenSida training events, with a trend that ranges from 29% with 1 or 2 contacts, to 43% with three 
contacts, and 55% with four or more contacts (X2 12.0, 3Gl). This means that a MARP that has received three or 
more contacts from PrevenSida has an almost triple probability of attending some training event offered by 
PrevenSida (52% versus 29%, OR 2.7, CI 95% 1.5-5.4, X2 11.1). 
• No association was found between the number of contacts provided by PrevenSida and the share of MARPs 
participating in peer socialization groups. This applies to the one-to-one increase contrast (X2kw 2.4, p=0.12,
3Gl), to the contrast between one contact and two or more contacts (X2 2.2), and to the contrast of 1-2 
contacts versus 3 or more contacts (X2 0.96). 
• The increase in the number of contacts provided by PrevenSida significantly increases the probability that 
MARPs carrying out VCT in PrevenSida network organizations, with a trend of 14% with 1 or 2 contacts, 27% 
with three contacts, and 36% with four or more contacts (X2 10.9, 3Gl), which means that MARPs that have 
received three or more contacts from PrevenSida have three times more probability of carrying out VCT 
offered by PrevenSida (34% versus 14%, OR 3.2, CI 95% 1.6-6.6, X2 10.1). 
As stated above, the number of contacts have a remarkable effect in PrevenSida in the coverage of the other 
prevention activities, although it also has some effect on outcome indicators. That is why this study has 
characterized promotional contacts, which duration and implementation do not permit an interactive dialogue, but 
they do have a clear motivating function in relation to HIV, as well as a function of calling to participate in other 
activities. 
Assessing PrevenSida Project Coverage - “2014 Mid-Term Performance Evaluation of HIV/USAID Nicaragua Program”
50
V. CONCLUSIONS 
The study provides valuable evidence for different types of analysis and could also be used as an input to planning 
processes at different management levels to enhance action targeting, either at the level of the program, project or 
implementing organization, and even in specific thematic areas where performance of some indicators suggests a 
match or mismatch with “expected” parameters. 
Since the basic review is a contrast analysis of indicators at different sample strata, those differences that constitute 
statistically significant evidence are presented as relevant findings (with a 95% confidence interval) and reflect the 
incidence of some factor that conditions PrevenSida coverage distribution or effectiveness. In the absence of major 
assessable or interpretative elements on the part of the research team by reason that it does not know all 
“expected” parameters at an operational level, and recognizing that the social actors involved in the project have 
the interpretative elements derived from the day-to-day dynamics of the project and operating management 
decisions that guide it, we highlight the value of the evidence to encourage collective reflection and dialogue within 
field teams of PrevenSida network implementing organizations, as well as between this level, project management 
and the HIV/USAID Program. 
As already pointed out, while a specific analysis of the different preventive actions was conducted, it has been
included as an annex to this report, in order to provide evidence derived from the operational dynamics, each of 
which is influenced by various factors. 
Although PrevenSida reaches 45% of MARPs through one or more of its main four preventive activities, this global 
coverage includes 30% of total MARPs receiving preventive actions from PrevenSida and “other providers.” Hence, 
PrevenSida provides exclusive coverage to 15% of MARPs. The main characteristics of PrevenSida coverage 
targeting are: 
1. The level of approximation between PrevenSida global coverage data in this study (45%) and the level estimated 
for 2013 (41%), based on MARP Unified Register (UR) and population estimates by CONISIDA, stands out. 
While SW show the lowest coverage in both estimates, the coverage estimated for each group is different. 
More differentiated rates are reported in this study, with the highest coverage found in TRANS (60%) and the 
lowest in SW (34%), while MSM have an average rate (46%). Coverage similarity in the estimates (43%, 42%, and
40%) seems to reflect the effect of using MARP-UR data as the basis for the population estimate. 
2. Coverage reflects access modalities for reaching MARPs, which have been used by PrevenSida since the 
beginning, and particularly mapping methods of concentration and pickup venues since 2013. Major factors 
associated to an increased coverage are: 
 
 MARPs participation in some organization, which largely influences PrevenSida global coverage (OR 2.9), 
as well as its exclusive coverage (OR 1.4). In turn, this condition is related to the fact that in MARPs 
claiming to be organized (40% of total), the vast majority (29% of total) claim to be linked to some NGO 
or association engaged in HIV prevention, where both the global and exclusive coverage exhibit the 
highest rates (69% in global coverage and 20% in exclusive coverage).
 MARPs searching for “pickups” in bars or discotheques increase PrevenSida exclusive coverage (OR 1.5) 
without affecting global coverage (X2 2.4).
Assessing PrevenSida Project Coverage - “2014 Mid-Term Performance Evaluation of HIV/USAID Nicaragua Program”
51
 Working in higher risk occupations (commercial sex, employed in bars, discotheques, motels) (OR 1.7 in
global coverage, OR 1.4 in exclusive coverage). 
 Studying increases PrevenSida global coverage (OR 1.3), but does not modify exclusive coverage (X2 0.8). 
PrevenSida has built in some epidemiological planning elements derived, in principle, from the PEPFAR 
Central America Partnership Framework, with a view to assume the HIV situation as a concentrated 
epidemic in order to prioritize MARP groups. Therefore, the first approximation consisted of establishing
alliances with organizations (associations and NGOs) representing or linked to key populations, which have 
developed a wide network. Each organization, in turn, has a social fabric that keeps a population segment 
very close to the organization, a situation expressed as a sense of ownership, although many organizations 
are NGOs without representativeness. This population segment accounts for 29% of total MARPs studied 
(72% of 40% claim to participate in organizations), including persons who are activists and develop leadership
with convening power that facilitates access to preventive services. Obviously, this 29% of MARPs is the first 
universe of access. As of 2013, PrevenSida added the search of the three MARP groups at “pickup” venues, 
including mapping as an operational planning tool.
Undoubtedly, the association of these factors with an increased PrevenSida coverage shows that strategies 
are sought to reach population groups with complex accessibility, given the combined effect of stigma and 
discrimination, and the difficulties of security, working hours and willingness of the establishments and venues 
to locate MARPs. Hence, coverage manages to reach MARPs under an individual risk condition. However, it 
should be noted that 54% of PrevenSida global coverage and 48% of exclusive coverage originate from the 
organized MARP segment, whereas the unorganized MARP universe is a majority (60% of total MARPs), of 
which about half (30% of total) remains as a segment with few probabilities of being covered by PrevenSida, 
including half (15% of total) of MARPs working in high risk activities.
3. In addition to the aforesaid factors directly associated with PrevenSida project coverage, various factors 
related to MARP personal or family characteristics, sexual life, schooling, and social insertion (study, work, 
awareness of rights, stance concerning violence, etc.) act as effect modifiers to increase or reduce coverage, 
that is to say, they act as facilitators or barriers to MARP accessibility to PrevenSida preventive services. We 
mainly highlight those factors that characterize coverage or make a difference due to their reiterative
presence, among others:
 Factors that increase coverage or facilitate access:
 Being over 24 years of age, with a higher schooling level (above 6th or 11th grade), not heads of 
household, students or members of an unemployed non-student group. 
 Having the first sexual relation after 16 years of age, looking for new “pickups” at bars or 
discotheques, being aware of unrealized rights and linking them to stigma and discrimination. 
 Working in activities related to tourism and mobile populations, having higher risk occupations (at
bars, discotheques, motels), searching for work. 
 Factors that reduce coverage or hinder access:
 Being under 24 years of age, living with any parent, having children, having a stable partner, living in the 
capital city, living in a municipality with a high HIV test positivity rate, and belonging to an employed 
non-student group. 
 Having the first sexual relation before age 15 or under pressure, looking for new pickups in 
commercial venues, being unaware of unrealized rights, and justifying violence against partner.
Assessing PrevenSida Project Coverage - “2014 Mid-Term Performance Evaluation of HIV/USAID Nicaragua Program”
52
 Working in activities unrelated to tourism, having occupations other than sex work, or working at 
bars or discotheques.
All of the above factors operate in combination in relation to both project coverage and each preventive 
activity, which specificity in terms of MARP recruitment, relaxed or hasty conditions, and dialogue 
possibilities, among other things, also determine the different influences of the studied factors (a particular 
analysis of each activity is attached as an annex). Some of them could be linked to strategic management 
decisions (targeting, action redirection) and others to operating management decisions in terms of search 
modes, recruitment, and even dialogue and message contents. 
4. In continuation of the foregoing, and considering that this study applies in its methodological design the basic 
elements of the PLACE approach proposed by MEASURE/USAID, evidence shows both progress and 
limitations in PrevenSida project coverage concerning strategic and operating management, which validate the 
epidemiological approach in HIV program management. 
 The first PLACE methodological element is in line with the basic epidemiology principle of identifying 
territories where an incidence or prevalence of damage is concentrated, in order to target preventive 
intervention on them. As already mentioned, priority prevention areas (PPAs) were identified in this 
study, according to the PLACE methodology, based on municipal positivity rate of HIV rapid tests 
conducted by PrevenSida in the past three years. Evidence shows that the PrevenSida project has not 
based its territorial targeting decisions on HIV test positivity rates, as explained below:
 Global coverage in PPAs (municipalities with a high positivity rate) is similar to that in municipalities 
with low positivity (44% versus 46%, X2 0.2), i.e., there is no coverage prioritization by territorial risk 
level. 
 The situation is counterintuitive in exclusive coverage, since PrevenSida care favors the low positivity 
segment, which has a 53% higher probability of accessing PrevenSida services, as compared to MARPs 
in PPAs that have a 35% lower probability of accessing PrevenSida prevention services. This means 
exclusive coverage does not prioritize PPAs, but rather focuses subsidy allocations on low positivity 
municipalities.
 No association was found between municipal HIV test positivity rates and coverage of the different 
PrevenSida communication activities. PrevenSida promotional contacts have the same coverage, 
regardless of municipality positivity (27% versus 26%, X2 0.03). The same happens with PrevenSida 
training event coverage (24% versus 27%, X2 1.6) and peer group participation (14% versus 15%, X2
0.12). The only exception is the case of VCT coverage provided by PrevenSida, which was lower in 
municipalities with the highest positivity (PPAs), as compared to municipalities with low positivity (13% 
versus 20%, OR 0.58, CI 95% 0.4-0.8, X2 13.4). 
Evidently, prioritization by territorially defined risk levels has not been included in PrevenSida strategic 
management and planning. It seems that the implementing organization selection process has guided 
decisions that have not been associated to an understanding of the epidemic situation in territorial terms. 
 A second methodological element in the PLACE approach consists of identifying specific venues within 
PPAs where MARPs seek new relationships (“pickups”), in the understanding that these relations 
contribute more to the risk of new HIV infections. It should be pointed out that PrevenSida had already 
incorporated since 2013 a “mapping of pickup venues” as a method for planning and organizing its 
preventive service offer, which has had clear effects on coverage:
Assessing PrevenSida Project Coverage - “2014 Mid-Term Performance Evaluation of HIV/USAID Nicaragua Program”
53
 PrevenSida global coverage does not show significant differences among the three types of pickup 
venues identified in this study (bars/discotheques 47%, streets and shopping centers 44%, and 
communication networks 43%).
 PrevenSida exclusive coverage focused on MARPs looking for new “pickups” at bars and discotheques
(OR 1.6, X2 8.7), but it is not very accessible to MARPs (36% of total) seeking new relationships on 
streets, shopping centers, or markets (OR 0.6, X2 7.1).
 An association was found only between the new pickup venue and VCT coverage provided by the 
PrevenSida network, identifying greater coverage in bars/discotheques (OR 1.5, X2 7.0) and less access 
for MARPs searching for pickups on streets and commercial venues (OR 0.7, X2 4.9). A similar 
coverage was found in other preventive activities offered by PrevenSida, regardless of the venue where 
MARPs meet new “pickups.”
It is quite clear that the correct inclusion of “mapping” as an operating planning and organization method 
has improved PrevenSida coverage of MARPs, but it has been biased towards venues assumed to be of
higher risk, resulting in the creation of access barriers to 34% of MSM and SW and 45% of TRANS, who 
habitually look for “pickup” on streets, shopping centers, or markets. 
 The third element in the PLACE approach is the individual rating of risk based on the total number of 
sexual partners and the number of new sexual partners in the past month and past year, in the 
understanding that each person with whom an individual has sex increases the risk of new infections. This 
risk is higher with “new partners” whose background is unknown and the possibility of negotiating 
condom use cannot be foreseen. It should be noted that the number of partners is used by PrevenSida as 
one of its outcome indicators without differentiating them from “new” partners. No parameters were 
found to rate the individual risk level of served MARPs. This condition determines whether this element is 
assumed to characterize coverage without evaluation purposes. 
 No significant differences were found in PrevenSida global coverage (44% versus 49%, X2 2.3) or 
PrevenSida exclusive coverage (15% versus 15%, X2 0.0), according to “PLACE individual risk” rating. 
 A lower “PLACE high risk” rate was found where MARPs have been exposed to training events (OR 
0.69, X2 8.1) and peer socialization groups (OR 0.63, X2 12.4), regardless of the agent that provided 
the service. No differences were found in the “PLACE risk” level between PrevenSida preventive 
activity coverage and preventive activities offered by “other agents.” 
 PrevenSida global coverage is associated with a smaller average total number of sexual partners (X2kw
3.9, p=0.04) and the average number of new sexual partners (X2kw 4.2, p=0.04) in the past month, but 
PrevenSida exclusive coverage is not associated with significant differences in average total number of 
sexual partners (X2kw 0.04) or new sexual partners (X2kw 0.08) in the past month. 
 Of the four preventive actions provided by PrevenSida only exposure to promotional contacts is 
associated with significantly smaller averages in the number of sexual partners (X2kw 5.7, p=0.02). None 
of PrevenSida preventive actions are associated with smaller averages in the number of new sexual 
partners in the past month.
 Except for promotional contacts, exposure to the other three preventive activities (regardless of the 
providing agent) is associated with a reduction in the average total number of sexual partners in the 
past month (Training X2kw 34.8, Peer group X2kw 40.4, VCT X2kw 12.0). The average number of new 
sexual partners was significantly smaller in MARPs exposed to training events (X2kw 26.6) and peer 
socialization groups (X2kw19.2), regardless of the agent providing the service. 
Assessing PrevenSida Project Coverage - “2014 Mid-Term Performance Evaluation of HIV/USAID Nicaragua Program”
54
We have highlighted the effects of exposing MARPs to behavior-changing activities that may lead to 
dialogue and interaction, such as training events, peer groups and counseling as part of VCT, where the 
number of sexual partners is part of the educational content. In turn, promotional contacts exhibit very 
limited effects in modifying this indicator.
5. The study provides clear evidence validating a review of structural factors in the socioeconomic and 
institutional environment in the methodological design of the study. In this regard, the dialogue interview 
questionnaire explored the relationship between MARP jobs and specific production models that might lead
to potential higher HIV infections, such as tourism and mobile population services. Additionally, based on 
interviews with key informants, the level of collective incidence of tourism and population mobility (borders, 
passenger flow, transport terminals) on each PPA selected in the study was reviewed. In effect, tourism and 
population mobility have an impact on both coverage and outcomes at the individual MARP level and at the 
aggregate level (mesovariable). 
 Forty percent of working MARPs claim their job is related to tourism and have higher probabilities of 
accessing PrevenSida global coverage (OR 1.54) and exclusive coverage (OR 1.48). While no significant 
differences were found in the PLACE risk level, they do have much higher averages of new sexual 
partners than MARPs working in activities unrelated to tourism (11.5 versus 8.1, X2kw 10.1, p=0.001), 
although they report a higher consistent condom use rate (OR 1.3, X2 3.9). In turn, as a mesovariable, 
municipalities with a high incidence of tourism do not show differences in accessing PrevenSida global 
coverage, but they do show a difference in accessing exclusive coverage (OR 1.5) and have a higher 
consistent condom use rate (OR1.3, X2 6.9), which seems to match the lower probability of municipalities 
with a high HIV test positivity (OR 0.48, X2 43.3). 
 Forty-six percent of working MARPs who claim to work in mobile population related services, do not 
have differences in their probabilities of accessing PrevenSida global or exclusive coverage, have a higher 
“PLACE high risk” rate (OR 1.4, X2 4.5) related to the fact that they report higher averages in the 
number of sexual partners (19.4 versus 16.9, X2kw 12.7, p=0.000) and new sexual partners in the past 
month (13.2 versus 6.6, X2kw 48.5, p=0.000), and exhibit a higher STI history rate (OR 1.44, X2 5.7). 
Viewed as a mesovariable, municipalities with a strong mobile population incidence have a higher 
probability of higher HIV test positivity rates (OR 1.4, X2 8.0), but show no differences in the probabilities 
of accessing PrevenSida coverage.
Evidence shows that while visualization of MARP concentration venues is very relevant because they either 
work or find new “pickups” there, it is also necessary to visualize risks related to environmental structural 
elements conditioning personal interrelationships at “pickup” venues, either due to the incidence of tourism 
or high population mobility at borders or venues where there is a passenger or cargo flow. The relevance of 
this condition increases even more if we consider the changes Nicaragua is currently going through and the 
changes it will face in the near future with tourism development and a constant increase in people and cargo 
flow within Central America. Both situations will be compounded in the interoceanic canal construction 
scenario.
6. PrevenSida coverage has clear effects on consistent condom use. We consider that consistent condom use 
rate (39%) is relatively low, and critically lower in SW (26%), despite it is the group with the highest 
frequency use (77%) and availability (76%) rates, but less use condoms with their stable partners (31%). In the 
other two groups, the rate is almost half (44% and 45%). However: 
Assessing PrevenSida Project Coverage - “2014 Mid-Term Performance Evaluation of HIV/USAID Nicaragua Program”
55
 Both PrevenSida global coverage (OR 1.7) and exclusive coverage (OR 1.7) are associated with higher 
rates of consistent condom use, as compared to other MARPs. 
 An association was also found between a higher rate of consistent condom use and PrevenSida coverage
through specific preventive actions, as compared to the same activities offered by “other providers”, such 
as promotional contacts (OR 1.6), training participation (OR 1.6), and VCT (OR 1.7). This association was 
not found in the case of peer socialization groups (X2 2.3), nor with an increase in the number of 
interpersonal contacts.
 The effect of PrevenSida exclusive coverage on consistent condom use is even higher in MARPs with a 
low schooling level, those who are heads of household, which are rated as “PLACE high risk,” and those 
that participate in some organization. 
7. MARP claim of an STI history is validated in this study as an adequate “proxy” outcome indicator, to the 
extent it is associated with higher municipal HIV test positivity rates (OR 1.7) and lower rates when 
condoms are consistently used (OR 0.68). In this context, it is remarkable that no association was found with 
PrevenSida project coverage (global: X2 1.2, exclusive: X2 0.3). Similarly, no association was found with 
specific preventive action coverage. 
8. The number of contacts provided by PrevenSida in the same MARP does not have any effect on outcome 
indicators, i.e., increasing MARPs exposure to a larger number of per-capita contacts under current 
implementation conditions does not have any effect on protective behaviors or risk levels. Nevertheless, 
passing from one to two, and from two to three and more contacts does have a clear consequence of 
increasing coverage of other preventive activities offered by PrevenSida. In other words, contacts as an 
activity have a clear promotional and motivational purpose that translates into MARPs transitioning to new 
forms of relationship and exposure to preventive activities (training, VCT, peer groups), which are more 
likely to modify behavior and risk indicators by enabling dialogue and interaction conditions, as shown above.
9. As expected, exposure to PrevenSida coverage increases the rate of MARPs who are aware of unrealized 
rights related to stigma and discrimination (58% versus 50%, OR 1.4, CI 95% 1.1-1.7, X2 8.1), as compared to 
MARPs unexposed to PrevenSida activities. The relevance of this association is that awareness of E&D￾related HR is, in turn, associated with a significant increase in consistent condom use (45% versus 32%, OR 
1.7, CI 95% 1.4-2.2, X2 24.4) and with a significant reduction in the number of sexual partners during the past 
month (10.2 versus 19.0, X2kw 21.5, p=0.0000). The same effect was not found in the case of new sexual 
partners (X2kw 1.9, p=0.165).
VI. RECOMMENDATIONS
The team considers that the main recommendation consists of using the evidence in this study as a tool for 
motivating reflection and dialogue at all levels in the USAID/HIV Program, and particularly in the PrevenSida 
Project. The team is convinced this exchange will ultimately achieve the best interpretation of said evidence and will 
translate it into decisions for change, in order to strengthen and expand obvious achievements, as well as to adjust 
elements that have been shown to weaken or hinder the achievement of PrevenSida goals.
We draw attention to the use of the annex that contains the specific analysis of each preventive activity. In addition 
to the four activities analyzed in the entire coverage (contacts, training, peer groups, and VCT), we also added an 
analysis of exposure to messages and condom accessibility. We believe this annex can be very useful in dialogues 
within implementing organization work teams and MARPs themselves.
Assessing PrevenSida Project Coverage - “2014 Mid-Term Performance Evaluation of HIV/USAID Nicaragua Program”
56
In order to target elements where robust findings were made, we would like to suggest some broad courses of 
action: 
1. A closer look should be taken at the possibility of including key PLACE methodological elements as axis that 
can articulate an overall adjustment to the project approach that can translate into planning and organization 
processes at the different levels: 
 Prioritization by territorial risk level should be included that translate not only into targeting decisions, 
but also in designing specific strategies for each location, taking into account the PLACE definition which 
states that a national epidemic is nothing more than an aggregation of various local epidemics.
Monitoring and evaluation should include follow-up of the coverage shift towards the highest risk and an 
adjustment of work strategies to local characteristics.
 “Mapping” development should continue in more depth as a local planning and organizational
instrument, defining effective strategies to reach MARPs looking for “new pickups” in venues other than 
bars and discotheques, in order to transfer relevance to MARPs exposed to new infection risk on 
streets, commercial areas, markets, transport terminals, etc.
 Consideration should be given to the possibility of including MARP rating criteria to individually identify
riskier or less risky behaviors, and thus monitor changes and define adequate follow-up strategies. The 
combined prevention model cannot be viewed as an across-the-board quantitative standard for all 
MARPs. We believe that the number of partners and new partners, as well as consistent condom use, 
could be monitored more closely.
2. The two well-differentiated coverage segments in terms of access and outcomes, namely, MARP segments 
with PrevenSida exclusive coverage and mixed coverage, are large enough (30% of total MARPs) to warrant 
a more in-depth analysis and determine whether targeting strategies need to be further adjusted with other 
provider agents, in order to make them truly complementary and, if possible, synergistic in seeking the same 
results. Likewise, consideration should be given as to whether monitoring systems should include elements 
identified by covered MARPs, in order to establish evidence-based dialogues with other major players at the 
local level in seeking better results and optimizing scarce resources. 
3. The series of coverage and outcome conditioning factors should be considered, particularly those that could 
become: 
 Collective, group, or individual risk trackers to develop guiding policies for targeting actions. All MARPs 
have a risk level, and while the highest risk should be prioritized, it cannot be done at the expense of a 
limited coverage of LOWER risk, especially in a population universe with a 78% HIGH RISK according to 
the PLACE indicator. As an example, PrevenSida HIV rapid test positivity rates or STI history are 
indicators relatively easy to gather and process, and can be monitored and inserted in periodic planning 
exercises. 
 Limited accessibility trackers are understood as access to the entire combination prevention model and 
not only to contacts. As noted in the analysis, priority targeting of risk is not enough, and therefore 
targeting should be accompanied with search, contact, and follow-up strategies. Such tools as “mapping,” 
adjusting working hours to MARP life dynamics, and partnerships with relevant players (municipal tourism 
boards, transport companies, transport terminal authorities, etc.) can greatly contribute to a more 
effective coverage. 
 Special attention should be given to MARPs who claim that personal or family conditions limit their 
participation, such as being young, low schooling level, having children, not having autonomy at home, 
Assessing PrevenSida Project Coverage - “2014 Mid-Term Performance Evaluation of HIV/USAID Nicaragua Program”
57
history of violence and abuse (1st relationship, partner relationships), and self-awareness as a subject of 
rights. These factors result not only in attraction strategies, but also in dialogue modes and contents. 
4. Findings that confirm, quantify, and highlight the incidence of structural factors in the economic 
environment, such as HIV risk related to tourism, mobile populations, or workers’ concentration in 
enclaves, point to the need for analyzing the epidemic in the light of public policies and processes that are 
changing these scenarios, in order to identify territories and population groups currently exposed, or more 
exposed in the near future, and develop risk control strategies linked to the development of tourism 
projects, infrastructure (revamping of ports, roads, the Grand Canal), or others. Similarly, this visualization 
enables to more clearly identify public or private responsibilities in controlling the epidemic, which guides 
dialogue and joint actions agreed in local partnerships between public institutions (mayor’s offices, INTUR, 
municipal tourism boards) and private companies (large, medium, and small) that can and should assume 
leadership roles in prevention.
5. The project should examine in more depth the role and purpose assigned to interpersonal contacts carried 
out in MARP life scenarios. Their clear effects on coverage of other activities strongly justify their 
implementation and the fact that they are the focus of operating planning. However, their limited effects on 
outcome indicators warrant a clearer inclusion of training events, peer groups and VCT in programming 
approaches to complete the combination prevention model.
6. The evidence contained in this report can be of great educational value in working with MARPs and 
implementing organization staff in the field if it is used in dialogues and educational materials. The idea is to 
promote evidence-based communication leading to immediate individual and collective self-reference 
concerning their own experiences, interpretations, and behaviors because it is immediately seen that it is 
not just about data, but data expressing the voice of peers who live in the same or similar places and who 
have equal or similar life dynamics. As a result, data has the power to encourage reflection and places the 
specific and local situation as a benchmark for a dialogue between various explanatory models seeking 
consensus on actions for change, generating collective dynamics that communications based on theoretical 
concepts fail to encourage.
7. Coverage and outcome achievements warrant further systematization in order to be improved and 
replicated, which should not only highlight work methodologies or modes more related to the actions 
carried out, but also be a clear reference to the various scenarios where actions take place, in order to 
adjust capacity-building training guides, methodologies and processes for implementing NGOs to recognize 
the uniqueness of each local epidemic.
Assessing PrevenSida Project Coverage - “2014 Mid-Term Performance Evaluation of HIV/USAID Nicaragua Program”
58
VII. BIBLIOGRAPHY
1. PEPFAR. Partnership Framework to Support Implementation of the Central American Regional HIV
Response (March 2010). www.pepfar.gov/documents/organization/139196.
2. PrevenSida. (March 2013). Estrategia de prevención combinada de VIH aplicada a poblaciones en mayor 
riesgo (PEMAR), Managua, Nicaragua.
3. USAID. Consultants Hernandez C, Jarquin Y, Ochoa F (2007-2013). “Evaluación intermedia del desempeño 
del programa bilateral de VIH en Nicaragua” Nicaragua.
4. MOH. (February 2007). Marco conceptual del Modelo de salud familiar y comunitario. MOSAF. Nicaragua.
5. PrevenSida. PPT. Registro único de identificación de PEMAR, PVIH, prueba de VIH y capacitaciones. 2013. 
6. Carolina Population Center. Priorities for Local AIDS Control Efforts Overview of the PLACE Method. 
www.cpc.unc.edu/measure/publications/tr-02-10/at
7. USAID. PrevenSida. Aplicación Técnica año 2010. Nicaragua. 
8. CONISIDA. (2013). Estimación de Poblaciones en Mayor Riesgo. Nicaragua.
9. MOH. (2013) Registros Componente ITS-VIH y SIDA. Nicaragua.
10. PrevenSida. USAID (2013). Partner Organization Mapping.