EVALUATION OF PREVENTIVE CARE PACKAGE PROJECT FOR
PEOPLE LIVING WITH HIV/AIDS IN ETHIOPIA
Woldemariam Girma 
May 2011 
With support from the American people, the Preventive Care Package (PCP) project funded by USAID 
was implemented by a team of three international health organizations: Population Services International 
(PSI), Intrahealth, and Population Council. 
The PCP project was supported by the U.S. President’s Emergency Plan for AIDS Relief (PEPFAR), the 
largest commitment ever by any nation for an international health initiative dedicated to addressing a 
single disease. PEPFAR is a multi-faceted initiative for combating HIV/AIDS around the world through 
programs to prevent HIV infection and provide comprehensive care and treatment to those affected and 
infected by HIV/AIDS. 
Contract/Project No.: GHH-I-00-07-000-62-00 
Submitted to: United States Agency for International Development/Ethiopia 
 
DISCLAIMER 
The authors’ views expressed in this publication do not necessarily reflect the views of the United 
States Agency for International Development (USAID) or the United States Government. 
TABLE OF CONTENTS 
ACKNOWLEDGEMENTS ............................................................................................................................... I
EXECUTIVE SUMMARY ............................................................................................................................... II
I. BACKGROUND ........................................................................................................................................... 1
II. PREVENTIVE CARE PACKAGE PROJECT FOR PLWHA IN ETHIOPIA............................................ 2
III. RESEARCH DESIGN & METHODOLOGY ............................................................................................ 3
TARGET GROUPS ............................................................................................................................................................4
SAMPLE SIZE AND SELECTION ............................................................................................................................................4
INDICATORS AS MEASURE OF CHANGE.................................................................................................................................4
DATA ANALYSIS ..............................................................................................................................................................5
LIMITATIONS OF THE STUDY ..............................................................................................................................................6
IV. RESULTS................................................................................................................................................... 7
A. SAMPLE CHARACTERISTICS......................................................................................................................................7
B. SEX DISAGGREGATED CHARACTERISTICS OF RESPONDENTS.............................................................................................8
C. EXPOSURE TO AND UTILIZATION OF PCP PROJECT ITEMS.............................................................................................10
D. AVAILABILITY, UTILIZATION AND EFFECT OF PROJECT ON CONSISTENT CONDOM USE .........................................................12
Availability and utilization of condoms...........................................................................................12
Effect of project on consistent condom use ....................................................................................13
E. AVAILABILITY, UTILIZATION AND EFFECT OF PROJECT ON CONSISTENT WUHAAGAR USE.....................................................14
Availability and use of WuhaAgar...................................................................................................14
Effect of project on consistent use of drinking WuhaAgar treated water.......................................15
Hand washing using soap at critical times......................................................................................16
F. DE‐WORMING DRUG UTILIZATION..........................................................................................................................17
G. AVAILABILITY, USE OF ITN AND EFFECT OF PROJECT ON YEAR ROUND ITN UTILIZATION.....................................................18
Effect of the project on year round ITN utilization..........................................................................18
H. OPPORTUNITY, ABILITY AND MOTIVATION TO PRACTICE PCP COMPONENTS ...................................................................19
Condom use.....................................................................................................................................19
WuhaAgar/PUR use ........................................................................................................................22
ITN use.............................................................................................................................................22
De‐worming agent use....................................................................................................................23
I. OUTCOME/IMPACT OF PREVENTIVE CARE PACKAGE PROJECT .......................................................................................25
Morbidity history and hospitalization .............................................................................................25
Diarrhea ..........................................................................................................................................25
Respiratory infections .....................................................................................................................25
Malaria............................................................................................................................................25
Febrile illnesses other than malaria ................................................................................................25
Skin diseases....................................................................................................................................26
Visits to health facilities..................................................................................................................26
V. CONCLUSION AND RECOMMENDATIONS..........................................................................................28
VI. REFERENCES..........................................................................................................................................30
VII. APPENDIX ..............................................................................................................................................31
LIST OF TABLES 
Table 1: Objectives, indicators and variables for evaluating the PCP project ............................................................... 5
Table 2: Sample characteristics in control and intervention areas, 2009-2011 .............................................................. 8
Table 3: Characteristics of respondents by sex, 2009-2011 ........................................................................................... 9
Table 4: Exposure to PCP items and source of information, 2009-2011 ..................................................................... 11
Table 5: Information on PCP kits received by respondents in intervention sites, 2011 ............................................... 12
Table 6: Sexual practice and condom use by sex, 2009-2011 ..................................................................................... 13
Table 7: Predictors of consistent condom use by sex and time of survey among respondents who had sex in one year 
before the survey ........................................................................................................................................... 14
Table 8: Availability of WuhaAgar and consistent use of WuhaAgar-treated water for drinking all the days in the 
week prior to the survey, 2009-2011 ............................................................................................................. 15
Table 9: Predictors of consistently drinking WuhaAgar-treated water by sex and time of survey .............................. 16
Table 10: Hand washing using soap at critical times, 2009-2011 ................................................................................ 17
Table 11: De-worming agent use at baseline (2009) and endline (2011) by survey site ............................................. 17
Table 12: ITN use in malaria endemic areas (altitude 2000 meters below sea level), 2011 ........................................ 18
Table 13: Predictors of year round ITN utilization in malaria endemic areas by sex and time of survey ................... 19
Table 14: Mean score on opportunity, ability and motivation of condom use ............................................................. 21
Table 15: Mean scores on opportunity, ability and motivation of WuhaAgar/PUR use ............................................. 22
Table 16: Mean score on opportunity, ability and motivation of ITN use, 2011 ......................................................... 23
Table 17: Mean score on opportunity, ability and motivation of de-worming agent use ............................................ 24
Table 18 Morbidity history of respondents at baseline and endline surveys, 2009-2011 ............................................ 26
Table 19: Morbidity and hospitalization status of respondents at baseline and endline surveys, 2009-2011 .............. 27
Table A1: Reliability analysis, condom use ................................................................................................................ 31
Table A2: Reliability analysis, WuhaAgar use ............................................................................................................ 31
Table A3: Reliability analysis, ITN use in malaria endemic areas .............................................................................. 31
Table A4: Reliability analysis, ORS use ..................................................................................................................... 32
Table A5: Reliability analysis, de-worming use .......................................................................................................... 32
ACKNOWLEDGEMENTS 
The authors wish to acknowledge the contributions of various institutions and individuals in undertaking 
this survey. We are grateful to the regional health bureaus and the HIV/AIDS Prevention and Control 
Offices of Amhara, Oromiya, Southern Nations Nationalities and Peoples, Tigray, and the Addis Ababa 
Administrative Region. We are also grateful to all participating health facilities in these regions for their 
collaboration and facilitation of the survey. 
We are grateful to Mekuria Yilma for his assistance in reviewing the questionnaire, training of 
interviewers and coordination of the baseline data collection. Belayneh Anteneh and Meskerem Abera of 
PSI-Ethiopia made considerable contributions to the protocol and questionnaire development and we 
acknowledge the coordinators, supervisors, and data collectors for their hard work during data collection. 
Finally, we express our gratitude to all respondents, who generously gave their time and shared 
experiences for the study. 
i
ACRONYMS 
AIDS Acquired Immune Deficiency Syndrome 
ART Antiretroviral Therapy 
BCC Behavior Change Communication 
EDHS Ethiopia Demographic and Health Survey 
FMoH Federal Ministry of Health 
HAPCO HIV/AIDS Prevention and Control Office 
HIV Human Immunodeficiency Virus 
ITN Insecticide Treated Net 
PCP Preventive Care Package 
PEPFAR President’s Emergency Plan for AIDS Relief 
PLWHA People Living with HIV/AIDS 
PSI/E Population Services International/Ethiopia 
ORS Oral Rehydration Salt 
STI Sexually Transmitted Infection 
SWS Safe Water System 
USAID United States Agency for International Development 
SNNPR Southern Nations, Nationalities, and Peoples Region 
ii
EXECUTIVE SUMMARY 
Preventive Care Package (PCP) for People Living with HIV/AIDS (PLWHA) in Ethiopia was a three￾year project, funded by the United States Agency for International Development (USAID)/President’s 
Emergency Plan for AIDS Relief (PEPFAR). The goal of the project was to reduce morbidity and 
mortality due to opportunistic infections among PLWHA by promoting the use of basic hygiene practices 
and disease prevention items provided in a pre-packaged kit. The PCP, pre-packaged kits that include 
soap, condoms, insecticide treated bednets (ITNs), WuhaAgar/PUR, de-worming agents, and narrow 
mouthed water vessels, were distributed free-of-charge to PLWHA at government health facilities. The 
project was implemented in four major regions of Ethiopia (Amhara; Oromiya; Southern Nations, 
Nationalities, and Peoples Region (SNNPR); and Tigray) and the Addis Ababa City Administration. 
Population Services International (PSI) led the project, with IntraHealth as its partner. Population Council 
oversaw the project evaluation. 
In order to measure changes in knowledge, attitudes, practices, utilization of the intervention products, 
and morbidity associated with the PCP project, the Population Council conducted evaluative surveys 
among PLWHA in project sites. This evaluation report is based on a baseline survey, conducted in 2009 
and an endline survey undertaken in 2011, among female and male PLWHA clients in selected health 
facilities. The list of facilities where the project was to be implemented in the first year of the project 
served as a sampling frame for the intervention sites, and the list of facilities where the project was 
planned to be implemented in the third year of the project served as a sampling frame for the control sites. 
Respondents were selected in two stages. First, 33 health facilities from the intervention sites and 12 
health facilities from the control sites were selected for inclusion in the study based on probability 
proportional to size of ever-enrolled HIV-positive clients and equal allocation of respondents. Second, 
clients within the catchment area of the facilities were selected systematically until the predetermined 
sample of 1,588 respondents was reached (34 respondents from each of 38 selected intervention and 
control health facilities in the four regions and 44 respondents from selected health facilities in Addis 
Ababa). Selection was conducted without replacement. Respondents were asked questions related to 
exposure to PCP products, utilization of the product items, morbidity history, utilization of health 
services, sexual behavior, and opportunity, ability, and motivation related to health behavior concerning 
the promoted PCP products. In total, 1,487 and 1,585 PLWHA were interviewed for the baseline and 
endline surveys, respectively. 
More than two-thirds of the respondents were females and more than 40 percent were aged 30-39 years. 
The background characteristics of respondents at baseline and endline surveys were different for marital 
status, level of education, and socio economic status at control and intervention sites. 
Data shows that the PCP project has reached the majority of beneficiaries in the intervention sites. About 
78 percent received the PCP kit; and among those more than 95 percent received the kit with soap, 
WuhaAgar, condoms, water containers and de-worming agents. Insecticide treated nets (ITNs) were not 
distributed among 87 percent of the respondents at the intervention sites. The PCP project has 
significantly increased condom availability and consistent condom use among respondents in the 
iii
intervention sites. Female respondents were significantly more likely to report consistent condom use as a 
result of the program intervention than male respondents. The proportion of respondents at endline who 
ever treated their drinking water using WuhaAgar and who drank WuhaAgar treated water during all the 
days of the week prior to the endline survey was significantly higher in the intervention sites than the 
control sites. Male and female respondents in the intervention site were 4.9 times more likely to drink 
WuhAgar treated water consistently compared to male and female respondents in the control sites, 
showing the effectiveness of the program towards consistent use of the product. Regular use of soap for 
hand washing was significantly higher in the intervention sites compared to control sites at baseline and 
endline indicating the insignificance of the PCP program towards improving regular use of soap for hand 
washing. A significantly higher proportion of respondents at endline were ever told by health providers to 
use a de-worming agent every six months compared to respondents in the control sites. Increased 
utilization of a de-worming agent every six months was also observed related to the program. The 
program did not show significant increase in utilization of ITN or hand washing with soap. 
There is a need to scale up the project interventions specifically the promotion of WuhaAgar, water vessel 
and De-worming agent to address the health status of people living with HIV/AIDS living in all regions. 
There is also a need to promote behavioral change for effective utilization of program interventions in 
future project initiatives. Reviewing the project implementation and designing appropriate strategy is 
recommended to achieve the desired outcome for hand washing with soap and ITN use. 
1
I. BACKGROUND 
Sub-Saharan Africa accounts for only 11 to 12 percent of the world’s population (UNAIDS, 2008; PRB, 
2007). However this area contains the largest number of people living with HIV/AIDS (PLWHA) (67 
percent), the majority of all new HIV infections (70 percent), and the greatest number of deaths related to 
AIDS (75 percent). In Ethiopia alone, it is estimated that more than one million people are living with 
HIV. In 2009, there were an estimated 44,750 deaths related to AIDS, 84,190 pregnancies among HIV￾positive women, 14,140 HIV-positive births to women living with the virus, and 336,160 PLWHA 
requiring antiretroviral therapy (ART) (FMoH, 2007). 
In developing countries such as Ethiopia, opportunistic infections are the leading causes of morbidity and 
mortality among PLWHA—ART alone is insufficient for a healthy life. Diarrhea, one of the most 
common opportunistic infections, affects 90 percent of PLWHA and results in significant morbidity and 
mortality, particularly among children with HIV and AIDS (Katabira, 1999). The World Health 
Organization estimates that 85 to 90 percent of diarrheal illnesses in developing countries can be 
attributed to unsafe water and inadequate sanitation and hygiene practices. Improvement of hygiene is a 
key approach to reducing diarrheal disease by promoting practices such as hand-washing with soap, water 
treatment and safe storage of drinking water. Introducing household-based safe water systems (SWS), 
such as treating the water with sodium hypochlorite solution and safe water storage, can also reduce 
diarrheal disease including among PLWHA (Quick et al., 2002, Lule et al., 2005). 
Other basic preventive care interventions can also reduce illness, prolong life, and prevent HIV 
transmission for PLWHA in Africa. Utilization of de-worming agents on a daily basis was found to be 
associated with reduced morbidity and mortality, and had beneficial effects on CD4 cell count and viral 
load among PLWHA in rural Uganda (Mermin et al., 2004). Insecticide treated nets reduce the risk of 
malaria infection, which is associated with reductions in CD4 cell counts (Mermin et al., 2006b). Oral 
rehydration salts (ORS) combat dehydration among the general population including PLWHA. Finally, 
the provision of condoms to HIV-discordant couples has been associated with a more than 80 percent 
reduction in HIV transmission (Allen et al., 2003) and is highly likely to prevent HIV re-infection among 
PLWHA. 
While there is some evidence from Uganda that some preventive care package items do yield beneficial 
health outcomes for PLWHA (Colindress et al., 2007), the outcome of using the preventive care items 
among PLWHA in the Ethiopian context is unknown. This report evaluates the outcome of a three-year 
preventive care package project for PLWHA in Ethiopia funded by the United States Agency for 
International Development (USAID)/President’s Emergency Plan for AIDS Relief (PEPFAR). The 
evaluation was based on findings from baseline and endline surveys conducted by the Population Council.
2
II. PREVENTIVE CARE PACKAGE PROJECT FOR PLWHA IN ETHIOPIA 
USAID/Ethiopia awarded the Preventive Care Package project to Population Services 
International/Ethiopia (PSI/E) in 2008, which was implemented in partnership with IntraHealth 
International and Population Council. The goal of the project was to reduce morbidity and mortality due 
to opportunistic infections among adults and children living with HIV/AIDS in Ethiopia. The project was 
multi-faceted and included the delivery of products and services, facilitating trainings and behavior 
change communication (BCC), all of which target the most common opportunistic infections among 
PLWHA. The project has four major components: 1) refining and developing a PCP; 2) bundling and 
distributing PCP elements for adult and pediatric cases 3) supporting utilization and adherence of 
preventive PCP elements through behavior change interventions at facilities and in the community; and 4) 
evaluating the outcome of the project. 
The PCP kit was developed through a coordinated development process led by PSI/E, program partners, 
and stakeholders (Federal and Regional HIV/AIDS Prevention and Control Office and Regional health 
bureaus). The PCP kit was further modified and finalized by technical working group from PSI/E, 
program partners and stakeholders. 
The project targeted the distribution of 215,000 PCP kits to clients receiving ART, pre-ART and PMTCT 
services in Tigray, Amhara SNNPR, Oromia regional states and Addis Ababa City Administration. The 
PCP kit elements include: 
 Six bottles of Water Guard (Wuha Agar)and 24 sachets of PUR 
 One 20 liter safe water storage vessel, 
 Two long lasting insecticide treated bed nets (only in specified malaria areas), 
 Four bars of soap, 
 Four dose of de-worming tablets, 
 100 condoms 
 An educational booklet that references correct procedures for utilizing each commodity. 
 About 211,000 PCP kits of the 215,000 PCP kits produced by PSI/E were distributed during the project 
period in Tigray, Amhara, Oromiya, SNNP regions and Addis Ababa Administrative City through 197 
selected health facilities. The target health facilities were selected in consultation with respective regional 
health bureaus and selection was primarily based on the size of registered PLWHA in the facility. In 
addition, 505 health service providers and 745 community agents selected from the 197 health facilities 
were trained on preventive PCP and customer service. Different BCC activities were conducted through 
public mass media, printed media, and interpersonal communications to increase PCP uptake and 
utilization. 
3
III. RESEARCH DESIGN & METHODOLOGY
Data collection and design 
The objective of this study is to examine changes associated with the PCP project for PLWHA. The study 
is quasi experimental by design. Two cross-sectional surveys were undertaken – the first at baseline just 
before project implementation and the second at endline four months before project completion. The 
baseline survey was conducted between May and August 2009 from respondents in 38 health facilities in 
four regions (Amhara, Oromiya, the Southern Nations Nationalities and Peoples Region (SNNPR); and 
Tigray) and 7 facilities in Addis Ababa City Administration. The endline survey was undertaken from 
January to mid-March 2011 from respondents in the same facilities. 
The list of facilities where the project was implemented in the first year of the project served as a 
sampling frame for the intervention sites. The list of facilities where the project was planned to be 
implemented in the third year of the project served as a sampling frame for the control sites. Respondents 
were selected in two stages. First, 33 health facilities from the intervention sites where the project was 
implemented in the first year and 12 health facilities from the control sites where the project was to be 
implemented in the third year were selected for inclusion in the study based on probability proportional to 
size of ever-enrolled HIV-positive clients. Facilities with at least 100 ever-enrolled HIV-positive clients 
were eligible for selection at baseline and the same health facilities selected at baseline were used for the 
endline survey. Second, clients within the catchment area of the facilities were selected systematically 
with equal allocation of respondents and sampling without replacement (34 respondents from each of 38 
intervention and control sites in the four regions and 44 from Addis Ababa sites). This selection was 
based on the consecutive appearance of clients at the health facility, and the sampling interval varied 
according to the average daily client flow in each facility. 
The survey questionnaire included questions on demographic characteristics; utilization of project items 
such as condoms, WuhaAgar, ITNs, and safe water systems, as well as hygiene practices. The instrument 
also included questions related to perception, behavior, opportunity, ability and motivational factors to 
use PCP items in the kit. Opportunity, ability and motivation constructs were measured through likert 
scale type four point ranging from ‘1’ for “strongly disagree” to ‘4’ for “strongly agree.” The 
questionnaire was drafted in English, translated into Amharic, and pretested in Addis Ababa. In cases 
where the respondent did not speak Amharic, interviewers used other regional languages; Tigrigna and 
Oromiffa. Interviewers had a minimum 12th grade education and previous experience with data collection. 
Interviewers and supervisors took part in a four-day training provided by Population Council staff just 
prior to the baseline and endline surveys. 
All respondents at baseline and endline provided their informed consent to participate in the study. The 
study received ethical approval from the Ethiopian Public Health Association ethical review board in 
Ethiopia. 
4
Target groups 
The target groups for the survey were female PLWHA aged 15-49 and male PLWHA aged 15-59 who 
were enrolled as clients in selected health facilities at the time of survey. The eligibility of females and 
males in the specified age range was defined to maintain compatibility and comparison with other major 
surveys like the Ethiopian Demographic and Health Survey (EDHS). 
Sample size and selection 
The sample size was determined using the following formula and assumptions: 
    
2
21
2
1 11z12z 22111 n
PP
α β PPPPPPD

     ; 
Where n is the minimum sample size per group, = 0.05 (one-tailed test of significance), = 0.20 (or 1-
power = 0.80), D= design effect=1.5, P1 is the assumed proportion for baseline group, P2 is the assumed 
proportion for endline group, and P is the average of the two proportions. The sample is calculated for an 
aggregate of the four regions and for Addis Ababa separately assuming differences in the behaviors 
promoted by the project. It is also assumed that the intervention will result in an 8 percent increase in 
WuhaAgar utilization. P1 is assumed 0.5 for respondents from the four regions in order to achieve optimal 
sample size and 7 percent for Addis Ababa based on a survey in Uganda (Colindress et al., 2007). The 
sample size was calculated to be 884 for the four regions and 301 for Addis Ababa, and 408 from control 
sites which include a 20 percent adjustment for non-response. The sample size of respondents from each 
health facility was determined to be 34 for the regional sites and 43 for Addis Ababa. 
Indicators as measure of change 
The most important variables to measure changes associated with the project were identified during the 
preparation of the questionnaire (Table 1). 
 
5
Table 1: Objectives, indicators and variables for evaluating the PCP project 
PCP product Objective Indicator Variable 
Condom To promote consistent use 
of condom 
Increased availability of 
condom 
 Condom available at the 
household currently 
 Availability of condom 
within walking distance 
 Increased consistent condom 
use 
 Ever use of condom 
 Consistent condom use in 
the last month/last year 
WuhaAgar To increase consistent use 
of WuhaAgar treated 
drinking water and safe 
storage system & reduce 
incidence of diarrhea 
Increased availability 
utilization of piped safe water 
vessel 
 Availability of piped safe 
water vessel at household 
 Utilization of safe water 
vessel 
 Increased availability and 
utilization of WuhaAgar 
 Currently have WuhaAgar 
at household 
 Ever used WuhaAgar 
 Drank WuhaAgar treated 
water every day in the last 
week 
 Ever bought WuhaAgar 
 Agree that WuhaAgar 
makes water free of germs 
and safe to drink 
Soap To increase soap use to 
wash hands at critical times 
and reduce other 
opportunistic infections 
Increased availability and 
utilization of soap 
 Availability of soap at 
household currently 
 Regular use of soap before 
eating and after eating food 
 Regular use of soap after 
using toilet 
 Regular use of soap before 
preparing food 
 Regular use of soap after 
changing baby diaper 
De-worming drug To increase use of de￾worming drug to every six 
months and prevent 
parasitic infections 
Increased availability and 
utilization of de-worming 
drug 
 Received information by 
health care provider to use 
de-worming drug every six 
months 
 Use de-worming drug 
every six months 
 ITN To increase consistent use 
of ITNs and reduce malaria 
infection in malaria prone 
areas of the project 
Increased availability and 
utilization of ITN 
 Availability of ITN at 
household currently 
 Use of ITN during all 
nights in last year 
Data analysis 
The data was entered in Microsoft Access and converted to SPSS for analysis. Separate results are 
presented for Addis Ababa and the four regions. Data were weighted by the number of registered 
PLWHA in the selected health facilities to adjust for unequal probability of selection, and results are 
presented based on the weighted data except for the sample size. Respondents from Addis Ababa were 
analyzed separately. based on the assumption that the preventive care practice of respondents from 
regions varies from urban Addis Ababa. In order to control the effect of the project, comparison was 
made between intervention and control sites for selected variables to evaluate the project. For Addis 
Ababa comparison was made between baseline and endline results. 
6
Respondents were read a series of assets that households may own and asked if their household possessed 
that item. Principle components analysis was used to determine the socioeconomic status of the 
respondent. 
Factor analysis was used for each of the opportunity, ability and motivation measures using scale items, 
to determine which items contributed to a single factor. Rotated loading was examined to determine if any 
sub-factors existed within the proposed item for each construct. Items loading a single factor were 
retained and reliability analysis was conducted for these items. Scales were considered reliable if they had 
a Chronbach’s alpha of 0.70 or higher and items were dropped to improve the alpha (PSI, 2005). A 
minimum of three items were required to create a multi-item scale composite variable. The composite 
variable is the sum of each of the individual variable item scores by the total number of variable items for 
that scale. 
Means were used to find levels of indicators and behavioral determinants for monitoring the project, and 
logistic regression modeling was used to identify which behavioral determinants and population 
characteristics are predictors of a promoted behavior. 
Limitations of the study 
Our sample was limited to PLWHA who are government health facility beneficiaries and did not include 
PLWHAs served by private health facilities and those who unaware of their status. This may result in a 
bias towards respondents who attend government health facilities hence may not be representative of all 
PLWHA in Ethiopia. 
7
IV. RESULTS 
A. Sample characteristics 
Data was collected from a total of 3,072 survey participants. Response rates were high, 98 percent for 
Addis Ababa, and 96 percent for the four major regions (Amhara, Oromiya, SNNP and Tigray). 
A total of 1,487 respondents were interviewed at baseline survey, of which 19 percent were from Addis 
Ababa city Administration, 37 percent from Amhara, 29 percent from Oromiya, 8 percent from SNNP, 
and 7 percent from Tigray regions. Similarly, endline survey captured 1,585 respondents with similar 
proportions as the baseline survey in the five regions. 
The majority of the respondents were female (68 percent in the four regions; 71 percent in Addis Ababa). 
Respondents who started ART increased from 83 percent at baseline to 88 percent at endline in the four 
regions, while it remained stagnant in Addis Ababa (88 percent). The majority of respondents were aged 
25- 39 years (69 percent from four regions; 62 percent from Addis Ababa). Respondents in the 
intervention sites were more likely to be never married, divorced or widowed compared to the control 
sites (p<0.05). Respondents in the control sites were less educated and more impoverished compared to 
respondents in the intervention sites at baseline and endline. 
Since there was no control site in Addis Ababa, characteristics of baseline and endline respondents’ were 
compared (Table 2). There were no significant differences among baseline and endline respondents for 
most of the socio-demographic variables. Respondents’ marital status, educational status, and 
socioeconomic status in the intervention sites appeared to be significantly different from the control sites 
at both baseline and endline survey. A greater proportion of respondents in the intervention sites had 
never been married (18 percent) compared to respondents in the control sites (13 percent). Respondents at 
the endline survey were more likely to be illiterate (18 percent) compared to respondents at baseline (11 
percent), while a high proportion respondents at baseline were poor (38 percent) compared to respondents 
at endline (24 percent). 
 
8
Table 2: Sample characteristics in control and intervention areas, 2009-2011 
The Four Regions Addis Ababa 
Variables Categories Baseline Study (2009) Endline Study (2011) Baseline 
(2009) 
Endline 
(2011) Intervention 
(n=817) 
Control 
(n=381) 
Intervention 
(n=883) 
Control 
(n=402) (n=289) (n=300) 
Treatment Status Pre-ART 16.8 16.6 11.6 11.9 12.2 12.1 
On-ART 83.2 83.4 88.4 88.3 87.8 87.9 
Sex of Respondent Male 34.3 27.4 * 31.9 33.6 29.5 28.7 
Female 65.7 72.6 68.1 66.4 70.5 71.3 
Age group 15-24 8.4 10.1 9.8 9.3 10.1 5.7 
25-29 28.8 23.6 25.1 23.8 19.4 14.3 
30-39 40.4 42.7 46.1 46.6 45.4 44.8 
40-44 11.1 13.6 10.5 11.5 12.2 15.2 
45+ 11.4 10.0 8.5 8.8 12.9 20.0 
Mean age 33.1 33.1 33.0 33.8 34.4** 36.6 
Marital status Never married 15.4 5.1 *** 11.2 5.7 *** 17.7 12.8*** 
Married 32.2 34.1 41.4 44.6 29.5 39.9 
Widowed 18.2 24.1 17.4 19.3 24.7 18.8 
Divorced 23.9 27.1 25.6 29.6 16.7 26.2 
Cohabiting 8.8 7.6 3.7 0.0 9.0 1.6 
Separated 1.6 1.9 0.8 0.8 2.4 0.6 
Highest schooling 
completed
Illiterate 21.6 34.5 ** 19.5 42.4 ** 10.7 17.8** 
Grade 1-6 31.6 32.9 25.6 30.2 36.3 21.7 
Grade 7-8 19.9 13.0 21.6 11.9 19.7 24.5 
Grade 9-10 13.7 12.8 17.8 9.0 10.4 13.1 
Grade 11-12 9.1 4.6 8.3 2.8 17.3 17.2 
Grade 12+ 4.1 2.2 7.2 3.7 5.6 5.7 
Socio economic 
status 
Poor 43.5 54.2*** 34.0 47.7*** 37.8 24.2*** 
Medium 28.0 14.5 26.1 12.7 54.4 56.5 
Rich 28.5 31.3 39.9 39.6 7.7 19.4 
Migration Resident 37.3 32.2 38.8 45.4 38.9 33.7 
Migrant 62.7 67.8 61.2 54.6 61.1 66.3 
* Significant at 0.05; ** significant at 0.01; significant at 0.001 
B. Sex disaggregated characteristics of respondents 
ART initiation was gender biased in the four regions, favoring males (92 percent) compared to females 
(86 percent) at endline (Table 3). The mean age for male respondents was 37 years and for females it was 
31 years (t-test, p<0.05). Female were less likely to be single, divorced or widowed compared to male 
respondents. Females were also more likely to be illiterate and less likely to have more than 9 years of 
education compared to males.
9
Table 3: Characteristics of respondents by sex, 2009-2011 
The Four Regions Addis Ababa 
Variables Categories 
Baseline Endline Baseline Endline 
Male Female Male Female Male Female Male Female 
(n=391) (n=807) (n=418) (n=867) (n=73) (n=216) (n=88) (n=212) 
Treatment Status 
Pre-ART 13.6 18.2 7.7 13.6** 10.6 13.3 13.3 11.6 
On-ART 86.4 81.8 92.3 86.4 89.4 86.7 86.7 88.4 
Age group 
15-24 2.1 12.0** 3.8 11.8** 6.0 11.8** 5.6 5.4 
25-29 18.6 31.3 10.9 29.9 2.4 26.5 2.2 19.2 
30-39 42.8 40.2 46.3 46.8 66.7 36.3 32.2 50.0 
40-44 18.9 8.6 20.3 7.3 7.1 14.1 22.2 12.5 
45+ 17.6 7.9 18.8 4.2 17.8 11.3 37.8 12.9 
Marital status 
Single 14.4 11.2** 13.5 7.6*** 16.7 18.2** 18.7 10.3** 
Married 47.9 25.6 66.0 31.1 64.3 15.3 60.4 31.3 
Widowed 8.6 25.4 7.0 23.2 3.6 33.0 7.7 23.7 
Divorced 15.4 29.5 13.0 33.4 14.2 17.7 11.0 32.6 
Cohabiting 12.8 6.2 0.0 3.9 1.2 12.4 1.1 1.8 
Separated 0.9 2.1 0.5 0.8 0.0 3.4 1.1 0.3 
Education 
Illiterate 14.1 31.1** 17.6 30.7** 3.5 13.8** 12.2 20.1** 
Grade 1-6 32.5 31.8 25.8 27.6 28.2 39.9 18.9 22.8 
Grade 7-8 17.5 17.9 18.1 18.9 15.3 21.7 20.0 26.3 
Grade 9-10 20.2 10.1 21.2 12.2 11.8 9.4 12.2 13.4 
Grade 11-12 15.7 8.8 17.3 10.6 41.2 15.2 36.7 17.4 
Socioeconomic status 
Poor 45.8 47.3 37.1 38.7 36.5 38.3 29.1 22.3 
Medium 27.7 21.9 21.7 22.2 56.4 53.7 48.8 59.4 
Rich 26.5 30.8 41.2 39.1 7.1 8.0 22.1 18.3 
Migration 
Resident 28.3 39.2** 34.9 43.6** 64.7 59.6 63.7 67.4 
Migrant 71.7 60.8 65.1 56.4 35.3 40.4 36.3 32.6 
* Significant at 0.05; ** significant at 0.01; significant at 0.001 
10
C. Exposure to and utilization of PCP project items 
One of the most important goals of the PCP project was to increase exposure and utilization of items in 
the PCP, through provision of information/education and the care items to PLWHA. 
Exposure to PCP project items at baseline and endline was assessed among respondents in the 
intervention and control sites (Table 4). A significantly higher proportion of respondents in the four 
regions received information/education at endline (87 percent in intervention sites, 58 percent in the 
control sites) compared to respondents who received information/education at baseline (36 percent in 
intervention sites, 26 percent in the control sites). The main sources of information/education for the four 
regions were government health facilities (75 percent at baseline, 87 percent at endline) followed by radio 
(18 percent at baseline, 20 percent at endline) and television (12 percent at baseline, 14 percent at 
endline). At endline, a significantly higher proportion of respondents received information from 
government health facilities on the importance of PCP use in the intervention sites (93 percent) compared 
to respondents in the control sites (77 percent). A significantly higher proportion of respondents in Addis 
Ababa received information from the radio on the importance of PCP items at endline (35 percent) 
compared to respondents who received the information at baseline (19 percent). 
A very small proportion (2 percent) of the respondents in the intervention sites and the control sites 
received the PCP items at baseline; this is probably from sources outside of the project (Table 4). At 
endline, 78 percent of respondents in the intervention sites received the PCP, while only one percent 
received PCP in the control sites. Similarly, 64 percent of the respondents in Addis Ababa (intervention 
site) received the PCP at endline compared to less than 1 percent who received the PCP at baseline. This 
shows that there was no significant contamination in the control sites or at baseline, and that the project 
package was well implemented in the intervention sites. 
The survey showed that almost all of the respondents who received the PCP intervention in the four 
regions and Addis Ababa also received WuhaAgar, 98 percent received water container/vessel, and 97 
percent received condoms or soap. 
11
Table 4: Exposure to PCP items and source of information, 2009-2011 
The Four Regions Addis Ababa 
Baseline Endline 
Intervention Control Intervention Control Baseline End 
line 
(n=817) (n=381) (n=883) (n=402) (n=381) (n=402) 
Received Information/Education 
on PCP 
35.5 26.1 86.9*** 58.4 45.5 90.8*** 
Source of Information 
Government health facility 74.8 65.3 93.0** 77.0 79.4 77.9 
Radio 14.1 24.2* 12.6 23.5* 19.1 35.2** 
Television 8.6 3.2 8.2 9.3 24.4 33.0 
PLWHA Association 11.0 5.3 4.8 19.0 16.8 12.7 
Private health facility 8.3 8.4 2.6 6.2 12.2 6.0* 
NGO 7.6 7.4 2.2 5.3 13.7 4.9** 
Pamphlets/Leaflets 2.1 6.3 1.5 1.3 3.8 10.9* 
News letter 2.1 2.1 1.0 0.0 8.4 10.6 
Billboard 0.0 0.0 1.6 0.0 0.0 4.2** 
Received PCP Kit 2.0 1.9 78.0*** 1.3 0.4 64.3*** 
Contents of the PCP Kit 
WuhaAgar 93.8 0.0 99.0 60.0 0.0 99.5 
Water Container 25.0 0.0 98.6 0.0 0.0 98.0 
Soap 25.0 0.0 96.8 80.0 0.0 98.0 
Condom 25.0 0.0 96.7 80.0 100.0 97.0 
De-worming medication 0.0 0.0 94.3 0.0 0.0 88.1 
PUR 0.0 0.0 91.0 20.0 0.0 89.1 
IEC Material 18.8 0.0 88.1 0.0 0.0 81.1 
ITN (among residents of 2000 
meter below sea level) 
0.0 0.0 13.0 0.0 - - 
* Significant at 0.05; ** significant at 0.01; ***significant at 0.001 
In more than 93 percent of the cases in the four regions and more than 87 percent of the cases in Addis 
Ababa, the providers (government health facility workers) of the PCP explained why the PCP items are 
given and how/when to use each of the items. Information, education, communication (IEC) materials 
were given along with the PCP components to more than 91 percent of the respondents (Table 5). 
12
Table 5: Information on PCP kits received by respondents in intervention sites at endline, 2011 
 Explained 
why given 
Explained 
How/when to use 
Received 
IEC material 
Four 
Regions 
Addis 
Ababa 
Four 
Regions 
Addis 
Ababa 
Four Regions Addis 
Ababa 
Type PCP Kit 
Soap 92.8 89.2 92.3 89.0 87.7 70.7 
WuhaAgar 97.9 93.0 97.7 93.0 92.8 85.9 
PUR 95.3 87.1 95.0 87.1 90.4 74.0 
Condom 96.5 92.8 96.1 92.8 89.7 84.2 
Water Container 96.9 89.3 96.8 89.2 92.9 84.0 
De-worming agent 97.1 89.2 96.8 89.3 92.8 84.0 
IEC Material 96.1 90.8 96.4 90.8 91.8 90.5 
D. Availability, utilization and effect of project on consistent condom use 
Availability and utilization of condoms 
There was no significant difference in the proportion of respondents who ever had sex, number of sexual 
partners, ever use of condoms, and consistent condom use between intervention and control sites at 
baseline survey (Table 6). However, at endline the proportion of respondents who ever use condoms, 
who have condoms at their household, and who consistently use condoms in the four regions significantly 
increased in the intervention sites compared to respondents in the control sites. The increase was observed 
for both male and female respondents, and thus shows that the PCP project has significantly increased 
condom availability and consistent condom use among respondents in the intervention sites in the four 
regions. The endline survey in the four regions showed that more female respondents in the control sites 
(11 percent) had more than one sexual partner compared to females in the intervention sites (7 percent); 
this requires further investigation. 
In Addis Ababa, a significantly higher proportion of respondents were found to have condoms in their 
households at endline compared to respondents at baseline. However consistent condom use at the endline 
among male respondents was significantly lower than consistent condom use among males at baseline 
survey. This shows that condom availability in the household is not always associated with consistent 
condom use, especially among male respondents. On the other hand, consistent condom use at the endline 
survey was not significantly different from baseline survey among females and overall respondents of the 
survey. In the intervention sites, consistent condom use increased from 68 percent at baseline to 73 
percent at endline, while it deceased from 59 percent at baseline to 48 percent at endline in the control 
site. This indicates that the program has prevented consistent condom use from declining. 
According to the data (not shown), the most preferred source of condoms at endline survey was health 
facilities (42 percent), followed by groceries/shops (33 percent) and pharmacies (17 percent). The 
condom types most widely used were Sensation (32 percent), Hiwot Trust (30 percent) and unbranded (9 
percent). Preferred source of condom at endline was not different from baseline for respondents in the 
four region and Addis Ababa.. 
 
13
Table 6: Sexual practice and condom use by sex, 2009-2011 
 Baseline 
(Four Regions) 
Endline 
(Four Regions) 
Addis Ababa 
 Intervention 
(n=817) 
Control 
(n=381) 
Intervention 
(n=883) 
Control 
(n=402) 
Baseline 
(n=289) 
Endline 
(n=300) 
Had sex in the last one year 46.9 49.0 54.3 48.3 48.0 46.7 
Have more than 1 regular/non regular 
partner in the past one year 
 
-Male 14.0 11.6 16.9 10.0 10.9 15.8** 
-Female 5.7 7.8 6.9 10.6* 2.5 9.2 
-All 9.4 9.4 5.9 10.4 5.9 12.4 
Ever used condom 50.5 44.4 53.3** 38.4 53.0 53.7 
Currently have condoms at household 
-Male 48.2 47.0 68.1*** 38.5 45.8 52.9 
-Female 27.5 25.4 49.2*** 18.6 23.1 47.4*** 
-All 37.5 29.0 58.4*** 24.9 28.9 49.0*** 
Consistent condom use last month/last year 
-Male 71.7 62.0 71.1*** 52.4 87.3*** 62.1 
-Female 65.2 57.3 74.5*** 43.7 64.2 73.0 
-All 68.1 59.1 73.0*** 47.6 74.1 68.7 
* Significant at 0.05; ** significant at 0.01; *** significant at 0.001
Effect of project on consistent condom use 
Multivariate logistic regression analysis was used to identify the effect of the project interventions on the 
likelihood of consistent condom use by adjusting for socio demographic variables. Age, marital status, 
educational status, socio economic status and site were entered into the logistic regression model as 
predictors of consistent condom use (Table 7). 
In the male model, the intervention and educational status of respondents were significant predictors of 
consistent condom use at baseline. Male respondents in the intervention site were 2.1 times more likely to 
use condom consistently at baseline compared to male respondents in the control site. However, at 
endline consistent condom use between male respondents in intervention and control sites was not 
significant. This suggests that the intervention was not effective in promoting consistent condom use 
among male respondents, rather implies that male respondents in the intervention sites were less likely to 
use condom at the end of the project period. 
On the other hand, female respondent in the intervention site were 3.2 time more likely to use condom 
consistently compared to female respondents in the control site, suggesting that consistent condom use 
was effective as a result of the intervention among female respondents 
At endline , respondents in the older age group were more likely to use condoms consistently compared to 
the reference and the youngest age group 15-24 in both male and female models. Female respondents who 
have completed some level of education were also more likely to use condoms consistently compared to 
illiterate female respondents at endline. 
14
Table 7: Predictors of consistent condom use by sex and time of survey among respondents who had sex in 
one year before the survey 
Variables Categories Male Female 
Baseline Endline Baseline Endline 
OR p-value OR p-value OR p-value OR p-value 
Site Intervention 2.10 0.029 1.46 0.211 1.39 0.229 3.16 0.000 
Control (Ref) 1.00 - 1.00 - 1.00 - 1.00 - 
Age 
group 
15-24 (Ref) 1.00 - 1.00 - 1.00 - 1.00 - 
25-29 1.94 0.557 4.62 0.074 2.36 0.008 2.62 0.007 
30-39 1.00 0.997 4.20 0.079 1.88 0.069 2.03 0.061 
40-44 1.40 0.759 7.63 0.011 2.36 0.019 2.10 0.050 
45+ 1.69 0.643 6.19 0.025 0.06 0.023 2.31 0.345 
Marital 
status 
Never married (Ref) 1.00 - 1.00 - 1.00 - 1.00 - 
Married 0.77 0.677 0.34 0.040 1.50 0.329 1.67 0.329 
Widowed/separated/
divorced 
0.82 0.792 0.50 0.345 1.23 0.629 1.26 0.672 
Cohabiting 0.63 0.523 0.22 0.566 1.57 0.335 1.11 0.873 
Education
al status 
Illiterate (Ref) 1.00 - 1.00 - 1.00 - 1.00 - 
Grade 1-6 5.91 0.000 1.44 0.355 1.34 0.355 2.29 0.003 
Grade 7-8 3.42 0.016 1.62 0.294 0.84 0.620 2.61 0.004 
Grade 9-10 6.74 0.000 5.63 0.000 0.71 0.390 6.01 0.000 
Grade 11-12 3.99 0.020 1.61 0.314 1.10 0.844 7.39 0.006 
Above Grade 12 4.73 0.016 1.23 0.671 0.37 0.140 1.06 0.917 
Socio 
economic 
status 
Poor 1.33 0.460 0.51 0.022 0.98 0.935 0.48 0.006 
Medium 0.64 0.232 0.81 0.517 1.29 0.403 0.97 0.905 
Rich (Ref) 1.00 - 1.00 - 1.00 - 1.00 - 
E. Availability, utilization and effect of project on consistent WuhaAgar use 
Availability and use of WuhaAgar 
Table 8 shows availability and usage of safe water systems among PLWHA. At baseline, none of the 
respondents in the intervention and control sites had the promoted water vessels in their households, 
which is evidence of the absence of contamination in the project. Availability of safe storage systems in 
the household and use of the storage for treating water was significantly higher in the intervention sites 
(76 percent) compared to the control sites (none of the respondents). More than 61 percent of respondents 
in Addis Ababa used the safe water storage vessel at endline compared to none at baseline. Behavioral 
change, measured by treating water using WuhaAgar, showed a difference from baseline to endline. In the 
absence of significant difference between intervention and control sites in treating drinking water using 
WuhaAgar at baseline; the proportion of respondents who ever treated water using WuhaAgar (80 percent 
in intervention sites versus 61 percent in control sites) and who drank WuhaAgar treated water during all 
the days of the last week at endline (63 percent in intervention sites versus 23 percent at control sites)was 
significantly higher in the intervention sites compared to the control sites. This behavior holds for both 
males and females. This finding shows that the project has increased ever and current use of WuhaAgar 
among both male and female respondents. A significantly higher proportion of respondents in Addis 
Ababa had heard of WuhaAgar, ever used it, and were currently using the product at endline (96 percent 
15
heard, 70 percent ever used, and 60 percent currently using) compared to baseline (82 percent heard, 8 
percent ever used, and 3 percent currently using). 
Table 8: Availability of WuhaAgar and consistent use of WuhaAgar-treated water for drinking all the days in 
the week prior to the survey, 2009-2011 
 Baseline (Four Regions) Endline (Four Regions) Addis Ababa 
 Intervention 
(n=817) 
Control 
(n=381) 
Intervention 
(n=883) 
Control 
(n=402) 
Baseline 
(n=289) 
Endline 
(n=300) 
Currently have piped safe water Vessel 0.0 0.0 79.0*** 0.0 0.0 68.6 
Currently using the safe water vessel 0.0 0.0 76.0*** 0.0 0.0 61.4 
Heard about WuhaAgar/PUR 
-Male 85.3 80.0 95.0 87.8 84.7 90.6 
-Female 81.2 74.2 99.2 89.9 80.3 98.7** 
-All 82.6 75.7 97.7 89.2 81.6 96.4** 
Ever used WuhaAgar/PUR 
-Male 41.8 33.7 77.0*** 43.1 7.1 47.1*** 
-Female 38.3 18.9 79.1*** 44.7 8.1 78.1*** 
-All 47.8 30.3 80.2*** 49.6 7.8 69.6*** 
Currently have WuhaAgar at household 
-Male 31.1 24.0 60.6*** 22.3 3.5 47.8 
-Female 29.6** 12.7 65.7*** 23.4 3.0 65.2 
-All 30.0** 16.0 61.1*** 23.1 3.2 60.2*** 
Ever bought WuhaAgar 
-Male 13.0 13.0 18.8 11.5 4.7 25.0*** 
-Female 9.7 8.6 18.2 15.6 4.5 46.0*** 
-All 10.9 9.8 18.4 14.2 4.6 40.4*** 
Drank WuhaAgar treated water every 
day in last week 
 
-Male 23.0 19.2 59.2*** 19.1 3.5 37.6 
-Female 25.1 12.1 62.5*** 22.6 2.0 54.3 
-All 24.3 14.4 60.8*** 20.6 2.1 49.5** 
Agree that WuhaAgar makes water free 
of germs and safer to drink 
70.5 97.0** 89.5 92.6 55.0 10.1*** 
* Significant at 0.05; ** significant at 0.01; *** significant at 0.001
Effect of project on consistent use of drinking WuhaAgar treated water 
Table 9 shows gender disaggregated predictors of consistent use of WuhaAgar treated drinking water 
(defined by use of WuhaAgar-treated drinking water during all the days of the week prior to the survey). 
Age, marital status, educational status, socio economic status and program intervention were entered into 
the logistic regression model as predictors of use of WuhaAgar treated water. 
Both male and female logistic regression models showed a significant difference in consistent use of 
WuhaAgar treated water as the result of the program intervention. In the male model, there was no 
significant difference in consistent use of WuhaAgar treated water between intervention and control sites 
at baseline. However, respondents in the intervention site were 4.9 times more likely to use WuhaAgar 
treated water compared to respondents in the control sites, implying the effectiveness of the intervention. 
In the female logistic regression model, a significant difference in consistent use of WuhaAgar treated 
water was observed both at baseline and endline. At baseline, respondents in the intervention sites were 
16
1.7 times more likely to use WuhaAgar treated water compared respondents in the control sites, while the 
likelihood of using WuhaAgar treated water increased to 4.9 times after the program intervention, 
showing the contribution of the program for increased use of WuhaAgar. 
 Table 9: Predictors of consistently drinking WuhaAgar-treated water by sex and time of survey 
Variables Categories Male Female 
Baseline Endline Baseline Endline 
OR p-value OR p-value OR p-value OR p-value 
Site Intervention 1.34 0.355 4.91 0.000 1.71 0.016 4.87 0.000 
Control (Ref) 1.00 - 1.00 - 1.00 - 1.00 - 
Age group 15-24 (Ref) 1.00 - 1.00 - 1.00 - 1.00 - 
25-29 7.29 0.246 0.92 0.898 1.30 0.374 0.65 1.047 
30-39 3.85 0.431 0.82 0.729 1.15 0.651 0.66 1.066 
40-44 4.76 0.358 0.57 0.310 0.97 0.910 0.48 0.776 
45+ 5.10 0.343 0.64 0.430 0.84 0.671 0.16 0.322 
Marital status Never married (Ref) 1.00 - 1.00 - 1.00 - 1.00 - 
Married 2.73 0.048 1.50 0.185 0.83 0.511 1.16 1.939 
Widowed/separated/di
vorced 
2.69 0.074 0.75 0.415 0.64 0.079 0.71 1.156 
Cohabiting 6.31 0.002 0.00 1.000 1.03 0.941 0.31 0.691 
Educational 
status 
Illiterate (Ref) 1.00 - 1.00 - 1.00 - 1.00 - 
Grade 1-6 0.41 0.022 0.74 0.351 0.82 0.395 1.13 1.609 
Grade 7-8 0.64 0.294 0.58 0.125 0.89 0.658 1.04 1.531 
Grade 9-10 0.81 0.622 1.03 0.923 0.86 0.635 2.90 4.767 
Grade 11-12 0.20 0.007 1.21 0.645 1.24 0.524 0.89 1.510 
Above Grade 12 0.25 0.036 0.95 0.886 0.42 0.272 0.73 1.467 
Socio economic 
status 
Poor 0.70 0.287 0.81 0.394 0.94 0.774 0.78 1.076 
Medium 0.59 0.123 1.55 0.085 0.68 0.104 0.81 1.120 
Rich (Ref) 1.00 - 1.00 - 1.00 - 1.00 - 
Hand washing using soap at critical times 
The vast majority of respondents at baseline and endline had soap in their households (Table 10). At 
baseline, there was no significant difference in the availability of soap at respondents’ households 
between the intervention sites (90 percent) and control sites (89 percent). However, a significantly higher 
proportion of respondents in the intervention sites (97 percent) had soap in their household compared to 
those in the control sites (91 percent) at endline. The increased availability of soap in respondents’ 
household in the intervention sites at endline was not related to regular use of soap for hand washing. This 
is evidenced by the absence of significant difference in regular use of soap for hand washing in the 
intervention sites at baseline and endline. 
Washing hands using soap before and after eating food, after using the toilet, and before preparing food 
were found to be significantly higher in the intervention sites compared to the control sites at both 
baseline and endline surveys. Hand washing using soap at critical times was not significantly improved 
by the project except before feeding children. A significant increase in using soap to wash hands after 
17
using the toilet was observed at endline (80 percent) compared to baseline (64 percent) among the Addis 
Ababa respondents. 
Table 10: Hand washing using soap at critical times, 2009-2011 
Baseline Endline Addis Ababa 
Intervention 
(n=817) 
Control 
(n=381) 
Intervention 
(n=883) 
Control 
(n=402) 
Baseline 
(n=289) 
Endline 
(n=300) 
Currently have soap at household 90.1 88.6 96.8*** 91.4 92.6 97.6** 
Regular users of soap to wash hands 74.3*** 54.5 76.1*** 61.2 74.3 88.7*** 
Users of soap to wash hands 
Before eating food 58.9* 51.4 62.0** 45.6 71.0 71.8 
After eating food 53.2*** 38.4 55.6*** 43.9 49.8 57.8 
After using toilet 57.3* 47.3 79.9*** 63.1 64.2 79.9*** 
Before preparing food 38.9*** 24.3 43.2*** 28.9 49.1 52.4 
Before feeding children 3.9 3.8 7.2* 3.6 2.9 3.6 
After changing baby diaper 1.7 1.6 3.0 2.8 2.2 1.3 
* Significant at 0.05; ** significant at 0.01;*** significant at 0.001
F. De-worming drug utilization 
A higher proportion of respondents in the intervention sites (37 percent at baseline, 89 percent at endline) 
received a de-worming drug from government health facilities compared to respondents in control sites (9 
percent at baseline, 33 percent at endline) (Table 11). However, the high proportion of respondents 
receiving a de-worming agent in the intervention sites was not necessarily due to the project intervention. 
Use of a de-worming agent every six months was significantly higher at endline in the intervention sites 
(98 percent) compared to the control site (89 percent), in the absence of significant difference between 
respondents in the intervention and control sites at baseline. A significantly higher proportion of 
respondents at endline were ever told by health providers to use a de-worming agent every six months 
compared to respondents in the control sites. This demonstrates the project’s contribution to the increased 
utilization of a de-worming agent every six months and the role of health care providers to communicate 
the information. In Addis Ababa, a significantly higher proportion of respondents at endline had received 
a de-worming agent from government health facilities (97 percent versus 82 percent), used de-worming 
agent every six months (70 percent versus 25 percent), and were ever told by health providers to use a de￾worming agent every six months (50 percent versus 36 months) compared to baseline respondents. 
Table 11: De-worming agent use at baseline (2009) and endline (2011) by survey site 
Baseline Endline Addis Ababa 
Intervention 
(n=817) 
Control 
(n=381) 
Intervention 
(n=883) 
Control 
(n=402) 
Baseline 
(n=289) 
Endline 
(n=300) 
Got de-worming from government HF 36.9*** 9.0 88.5** 32.9 82.3 97.4*** 
Use de-worming agent every 6 month 90.8 89.7 98.3*** 89.2 25.1 69.5** 
Ever told by health provider to use de￾Worming agent every six months 
15.0* 10.1 66.7*** 9.1 36.2 49.5** 
* Significant at 0.05; ** significant at 0.01; *** significant at 0.001
18
G. Availability, use of ITN and effect of project on year round ITN utilization 
Table 12 shows respondents’ knowledge about ITNs, their availability in the household and consistent use 
of ITN in the last year in the four regions. Only respondents who lived in areas 2000 meters below sea 
level were included for this analysis. Ninety-eight percent of respondents at baseline and 99 percent of 
respondents at endline were aware of ITNs, and there were no significant differences in ITN knowledge 
between intervention and control sites. ITN availability was significantly higher in the intervention sites 
(60 percent at baseline, 70 percent at endline) compared to control sites (43 percent at baseline, 66 percent 
at endline) showing that increased availability of ITNs in the intervention sites was accompanied by a 
similar increase in availability of ITNs in the control sites. The increased availability of ITNs in the 
control sites may be attributed to other interventions by the Ethiopian government. At baseline, year 
round use of ITNs was significantly higher in the control sites (70 percent) compared to intervention sites 
(62 percent); however there was no significant difference at endline. This finding suggests that the project 
made some contributions in the intervention sites for ITN use in the last year. 
Table 12: ITN use in malaria endemic areas (altitude 2000 meters below sea level), 2011 
Baseline Endline 
Intervention 
(n=279) 
Control 
(n=160) 
Intervention 
(n=306) 
Control 
(n=164) 
Know ITNs 98.2 97.9 98.9 99.3 
ITN available at home 60.4** 43.4 70.0** 66.0 
Used ITN all the nights in the last year 62.1 69.8** 74.3 74.2 
Get ITNs from Health Facility 10.5 10.2 31.5** 13.8 
* Significant at 0.05; ** significant at 0.01
Effect of the project on year round ITN utilization 
Table 13 shows multivariate results for predictors of year round utilization of ITN. Age group, marital 
status, educational status, and socio economic status were entered in to the logistic regression model as 
predictors of consistent use of ITNs. 
The multivariate logistic regression model for male showed that respondents in the intervention sites were 
about 60 percent less likely use ITN as compared to respondents in the control site both before and after 
the program intervention. This implies that the program intervention did not bring about improvement in 
ITN utilization among male respondents. This could be due to other strong interventions in the control 
sites that might canceled out the visibility of the program intervention. 
Similarly, ITN utilization was not improved due to the program intervention among female respondents. 
Other important predictors of year round ITN utilization were marital status and educational status. 
Married male and female respondents were more likely to use ITN compared to never-married male and 
female respondents at endline. Further female respondents with some level of education were less likely 
to use ITN compared to illiterate female respondents. 
19
Table 13: Predictors of year round ITN utilization in malaria endemic areas by sex and time of survey 
Variables Categories Male Female 
Baseline Endline Baseline Endline 
OR p-value OR p-value OR p-value OR p-value 
Site Intervention 0.42 0.005 0.43 0.001 0.86 0.443 0.82 0.256 
Control (Ref) 1.00 - 1.00 - 1.00 - 1.00 - 
Age group 15-24 (Ref) 1.00 - 1.00 - 1.00 - 1.00 - 
25-29 3.50 0.447 0.65 0.549 2.12 0.021 0.60 0.051 
30-39 2.42 0.591 0.19 0.019 2.39 0.009 0.83 0.463 
40-44 2.80 0.527 0.20 0.019 1.88 0.059 0.59 0.046 
45+ 2.31 0.613 0.24 0.041 0.72 0.486 0.53 0.103 
Marital 
status 
Never married (Ref) 1.00 - 1.00 - 1.00 - 1.00 - 
Married 6.05 0.006 8.29 0.000 1.77 0.064 2.46 0.003 
Widowed/separated/
divorced 
2.26 0.256 5.39 0.009 1.00 0.996 1.18 0.588 
Cohabiting 3.61 0.086 0.00 1.000 1.22 0.627 0.29 0.085 
Educational 
status 
Illiterate (Ref) 1.00 - 1.00 - 1.00 - 1.00 - 
Grade 1-6 0.40 0.027 0.54 0.083 0.72 0.108 0.54 0.002 
Grade 7-8 0.95 0.905 1.37 0.401 0.53 0.018 0.54 0.005 
Grade 9-10 0.45 0.084 0.61 0.188 0.26 0.001 0.66 0.092 
Grade 11-12 0.40 0.102 1.07 0.886 0.44 0.027 0.29 0.001 
Above Grade 12 0.30 0.060 1.02 0.967 0.24 0.090 0.76 0.489 
Socio 
economic 
status 
Poor 0.42 0.013 0.72 0.237 0.77 0.211 0.86 0.392 
Medium 0.69 0.287 0.82 0.499 0.70 0.123 0.62 0.013 
Rich (Ref) 1.00 - 1.00 - 1.00 - 1.00 - 
H. Opportunity, ability and motivation to practice PCP components 
Factors affecting behavioral change are modeled into three components for the purpose of this study. 
These components are opportunity to practice intended behaviors (includes availability and social norm), 
ability to perform intended behavior (includes knowledge, self-efficacy and social support), and 
motivational factors (includes attitude, belief, and outcome expectation). In addition to comparing change 
in behavior related to PCP implementation, we also looked at changes in these factors that are highly 
likely to affect behavior in the long term. Reliability and principal component analyses were used to 
identify variables to reliably measure these factors. A Cronbach’s alpha of 0.70 or above is accepted and 
the analysis is based on this criteria. All factors are measured using a Likert scale ranging from 1 to 4; 1 
for strongly disagree, 2 for disagree, 3 for agree and 4 for strongly disagree. 
Condom use 
Ability of respondents to use condom significantly increased among respondents in the intervention sites 
compared to respondents in the control sites at endline. At endline respondents in the intervention sites 
acquired significantly higher levels of knowledge about condoms’ ability to prevent HIV re-infection, 
STIs, and their use as a family planning method compared to respondents in the control site. Respondents 
at intervention sites also developed significantly higher levels of self efficacy to correctly and consistently 
use condoms during the intervention period. Significantly greater levels of social support from friends, 
sexual partner/spouse was observed at both intervention and control sites at endline. Similarly, analyses of 
motivational factors showed a significant increase in outcome expectation among respondents in the 
20
intervention sites. They expect condoms to be an effective tool to prevent HIV re-infection and STIs. 
Further, erroneous beliefs that free condoms break easily/are poor in quality and that condoms decrease 
sexual pleasure, intimacy or emotional closeness decreased in the intervention sites towards the end of the 
project period, while this was not the case in the control sites. The change in the knowledge, self efficacy 
and outcome expectation could be attributed to the PCP project. On the other hand the contribution of the 
PCP project towards improving negative attitudes about condoms and receiving social support to use 
condoms was not noticeable. Detailed analysis is shown in Table 14 below. 
21
Table 14: Mean score on opportunity, ability and motivation of condom use 
 
Four Regions Addis Ababa 
Intervention sites Control sites 
Baseline End line Baseline End line Baseline End line 
(mean) (mean) (mean) (mean) (mean) (mean) 
Ability 
Knowledge on condom use 3.37 3.45*** 3.31 3.23 3.56 3.52 
Condoms can be used to prevent HIV re-infection 3.43 3.51** 3.27 3.32 3.69 3.56 * 
Condoms can be used to prevent STI 3.36 3.50*** 3.28 3.26 3.64 3.57 
Condoms are not effective in preventing HIV re-infection (R) 3.03 3.21*** 3.23 2.97*** 3.29 3.25 
Condoms should be used consistently and correctly to prevent 
STIs 3.42 3.43 3.29 3.20 3.50 3.57 
Condoms can be used as a family planning method 3.47 3.59*** 3.45 3.41 3.72 3.62 * 
Social support 2.58 3.00*** 2.29 2.62** 2.79 2.86 
My friend and I encourage each other to use condoms correctly 
and consistently 2.49 2.86*** 2.26 2.45* 2.98 3.04 
Most of my friends approve the use of condoms 2.59 2.90*** 2.41 2.69*** 2.89 2.92 
Most of my friends encourage consistent and correct use of 
condoms 2.54 2.88*** 2.35 2.69*** 2.74 2.90 
There is someone that I can talk when I have questions or 
concerns about condoms 2.70 3.39*** 2.82 3.04** 2.58 3.22 *** 
My partners/spouse supported me to use condoms 2.51 2.91*** 2.15 2.35* 2.86 2.72 
Most of my community support for making decisions about my 
sex life 2.51 2.75*** 1.89 2.38*** 2.62 2.29 ** 
Self efficacy 3.21 3.41** 3.01 3.02 3.52 3.40 
I am confident that I can correctly use condoms 3.09 3.42** 2.97 2.92 3.43 3.39 
I am confident that I can consistently use a condom with my 
regular partner and causal partner 3.14 3.39** 2.98 2.97 3.50 3.36 
I can refuse to have sex with a new partner if we don`t have a 
condom even if I want to have sex 3.28 3.40** 3.09 3.15 3.51 3.42 
Motivation 
Outcome expectation 3.17 3.25* 3.10 3.15 3.15 3.40 ** 
Condoms are effective in preventing HIV infection 3.34 3.42* 3.25 3.24 3.49 3.59 
Condoms are effective in preventing STIs 3.34 3.45** 3.27 3.24 3.40 3.51 ** 
If one uses condoms correctly and consistently they don’t have 
to worry about contracting HIV 2.79 2.86 2.72 3.00*** 2.50 3.11 ** 
Attitude 3.05 3.15** 2.99 2.83** 3.39 3.18 
Using a condom all the time is difficult for me (R) 2.98 3.12** 2.82 2.63** 1.76 1.85 
The idea of using condoms does not appeal to me (R) 3.00 3.08 3.05 2.72*** 1.63 1.79 
Condoms are uncomfortable (R) 3.12 3.20 2.97 2.86 1.59 1.72 
Condoms break easily (R) 2.96 3.10** 3.04 2.98 1.62 2.04 ** 
Belief, condom strength 2.13 1.96** 2.11 2.09 1.85 1.82 
Free condoms mostly break easily 1.86 1.67** 1.76 1.93** 1.74 1.64 
Condoms given for free are of poor quality 1.86 1.65* 1.78 1.88 1.72 1.59 * 
Condoms in general are not safe 2.09 1.84** 2.12 2.04 1.98 1.87 
Using two condoms at the same time is better than using just one 2.44 2.42 2.54 2.43 2.23 2.50 * 
Most people who use condoms are promiscuous 2.08 1.73** 1.97 2.06 1.70 1.64 
Condoms decrease sexual pleasure 2.29 2.22 2.30 2.14* 1.75 1.99 * 
Condoms decreases intimacy or emotional closeness 2.26 2.14* 2.20 2.21 1.76 1.73 
 * Significant at 0.05; ** significant at 0.01; *** significant at 0.001; (R) indicates reversed items 
22
WuhaAgar/PUR use 
Opportunities related to WuhaAgar use behaviors expressed in terms of availability and social norms have 
brought about a significant positive change both in intervention and control sites. Ability factors 
expressed in terms of self efficacy to treat drinking water using WuhaAgar and social 
support/encouragement received from friends and neighborhood showed positive significant change at 
both intervention and control sites. However, the rate of increase for all these positive changes was higher 
in the intervention sites compared to the control sites (Table 15). 
Table 15: Mean scores on opportunity, ability and motivation of WuhaAgar/PUR use 
 The Four Regions Addis Ababa 
Intervention sites Control sites 
Baseline End line Baseline End line Baseline End line 
(mean) (mean) (mean) (mean) (mean) (mean) 
Opportunity 
Availability of WuhaAgar/PUR 2.09 2.68** 1.66 2.13** 1.67 2.85 ** 
It is difficult to find WuhaAgar/ PUR around here.(R) 2.49 2.10** 2.92 2.38** 2.92 1.92 ** 
I have easy access to organization when I want to get WuhaAgar/ 
PUR 2.21 2.58** 1.65 2.26** 1.43 2.65 ** 
I know where to obtain WuhaAgar/PUR 1.58 2.55** 1.33 1.53** 1.55 2.70 ** 
Social norms 2.17 2.83** 1.65 2.21** 1.58 2.76 ** 
Majority of PLWHA always use WuhaAgar to treat their 
drinking water 2.26 3.00** 1.71 2.26** 1.58 3.02 ** 
It is customary to treat drinking water among PLWHA 1.83 2.66** 1.43 1.96** 1.43 2.58 ** 
Most people in the community often take precaution to drink safe 
water 2.37 2.81** 1.89 2.31** 1.96 2.76 ** 
Most of my friends take action to treat their drinking water to 
make it safer 2.09 2.78** 1.68 2.28** 1.63 2.68 ** 
Ability 
Self efficiency 2.98 3.30*** 2.69 2.95** 2.73 3.26 ** 
I can treat the water that I drink whenever I need to 2.87 3.35** 2.37 2.91** 2.13 3.43 ** 
I am easily able to treat my water to make it safer to drink 3.02 3.42** 2.66 2.98** 2.76 3.50 ** 
Treating my water with WuhaAgar/PUR is not easy to do 2.34 2.03** 2.29 2.17** 2.26 2.24 
I know how to make water safe for drinking 3.09 3.41** 2.64 3.00 3.03 3.31 ** 
Social support 1.94 2.49** 1.67 2.21** 1.35 2.71 ** 
I can get WuhaAgar/PUR from my PLHWA association 1.90 2.48** 1.56 1.93** 1.33 2.72 ** 
People in my neighborhood have taught me how to treat my 
water 1.90 2.42** 1.68 2.19** 1.27 2.45 ** 
My friends encourage me to use WuhaAgar/PUR 1.93 2.51** 1.75 2.28** 1.40 2.66 ** 
Fellow PLWHA have shown me how to use WuhaAgar/PUR 1.98 2.56** 1.69 2.39** 1.37 2.74 ** 
Motivation 
Outcome expectation 3.45 3.44 3.29 3.25 3.22 3.40 ** 
Treating water with WuhaAgar/PUR is the safest way to protect 
my family from water-borne diseases. 3.36 3.40 3.22 3.15 2.80 3.30 ** 
WuhaAgar/PUR is effective in making water safe for drinking 3.47 3.49 3.30 3.31 3.34 3.47 * 
Treating water with WuhaAgar/PUR gets rid of contaminants 
that can harm my health 3.49 3.44 3.33 3.30 3.39 3.45 
* Significant at 0.05; ** significant at 0.01; *** significant at 0.001; (R) indicates reversed items 
ITN use 
ITN use was an expected and normal standard behavior/social norm among respondents in the 
intervention sites and control sites, and this behavior significantly increased at endline in both 
intervention and control sites. On the other hand, self efficacy to use ITNs did not significantly change in 
the intervention sites and the rate of increase was less than the rate of increase in control sites. Contrary to 
the intervention sites, self efficacy in the control sites significantly increased at endline compared to 
baseline. The outcome expectation that ITNs protect against malaria were significantly reduced in the 
23
intervention sites during the intervention period. This requires further investigation as to why PLWHA 
lost positive expectation. .Generally, there is no intervention related change in behavior to use ITN (Table 
16). 
Table 16: Mean score on opportunity, ability and motivation of ITN use, 2011 
 Intervention sites Control sites 
Baseline End line Baseline End line 
(mean) (mean) (mean) (mean) 
Opportunity 
Social norm 2.83 3.20** 2.21 3.04** 
Using ITN daily is normal among PLWHA 2.77 3.13** 2.08 2.90** 
Majority of PLWHA know how to take precautions to avoid malaria 2.86 3.24** 2.20 3.06** 
Most of my friends and neighbors use mosquito nets 2.85 3.24** 2.36 3.14** 
Ability 
Self efficacy 3.50 3.49 3.37 3.69** 
I know how to use a bed net. 3.43 3.52 3.28 3.66** 
Hanging a bed net is easy for me to do 3.46 3.44 3.30 3.66** 
It is my responsibility to make sure I sleep under a bed net 3.60 3.51 3.54 3.72** 
Social support 2.37 2.66** 2.00 2.62** 
People in my neighborhood have told me how to use bed nets 2.33 2.62** 2.04 2.57** 
People in my neighborhood have encouraged me to use bed nets 2.35 2.66** 1.97 2.62** 
My friends have counseled me on how to use bed nets 2.41 2.71** 2.00 2.68** 
Motivation 
Outcome expectation 3.50 3.35** 3.46 3.54 
The best protection against malaria is the use of ITN 3.56 3.43** 3.58 3.56 
Using a bed net provides restful sleep for PLWHA 3.52 3.42** 3.42 3.65** 
When I sleep under ITNs, I do not to have worry about getting malaria 3.45 3.16** 3.45 3.48 
ITNs will kill mosquitoes, when they touch them 3.56 3.42** 3.46 3.58 
ITNs will not protect me from malaria even if I use them ( R ) 1.60 1.68 1.65 1.56 
* Significant at 0.05; ** significant at 0.01; *** significant at 0.001; (R) indicates reversed items 
De-worming agent use 
Opportunity of receiving de-worming agent significantly increased from a mean of 2.77 to 2.91 among 
intervention sites, while it significantly decreased in control sites. Ability and motivational factors to use 
de-worming agents also significantly increased in the intervention sites, while significantly decreasing in 
the control sites towards the end of the project period. This shows the project contribution towards using 
de-worming agents (Table 17). 
 
24
Table 17: Mean score on opportunity, ability and motivation of de-worming agent use 
 The Four Regions Addis Ababa 
Intervention sites Control sites 
Baseline End line Baseline End line Baseline End line 
(mean) (mean) (mean) (mean) (mean) (mean) 
Opportunity 
Availability 2.88 2.72** 3.09 2.62** 2.97 2.84 * 
I know a place where I can obtain de-worming agents 3.18 2.98** 3.37 2.71** 3.48 3.06 ** 
It is hard to find de-worming agents when I need them ( R ) 2.21 2.32* 2.10 2.31** 2.00 2.21 * 
De-worming agents are always available in an outlet within 
walking distance of my home 2.59 2.47* 2.99 2.43** 2.32 2.66 ** 
Social norms 2.67 3.04** 2.40 2.51* 2.94 3.03 
Taking de-worming agent every six months is normal among 
PLWHA 2.33 2.87** 1.83 1.95* 2.55 2.93 ** 
It is not common for PLWHA to take de-worming agents every 
six months ( R ) 2.72 2.24** 2.94 2.71** 2.33 2.22 
It is acceptable for PLWHA to take more precaution to avoid 
parasites 3.46 3.47 3.34 3.27 3.65 3.41 ** 
Ability 
Knowledge 3.19 3.32** 3.20 2.78** 3.22 3.32 
De-worming can help prevent parasites 3.43 3.43 3.46 3.11** 3.55 3.48 
Everybody should take a de-worming drug every six months 
for life 2.79 3.01** 2.72 2.20** 2.74 3.05 ** 
Preventing parasites will help me to live a longer and healthier 
life 3.41 3.49** 3.40 3.08** 3.42 3.42 
Motivation 
Outcome expectation 3.16 3.38** 3.17 3.01** 3.38 3.42 
Taking a de-worming agent every six months is effective in 
preventing parasites among PLWHA 3.27 3.47** 2.83 2.57** 3.29 3.47 * 
Taking a de-worming agent every six months cannot protect 
people from parasites ( R ) 1.96 1.68** 1.72 1.76 1.69 1.65 
De-worming has nothing to do with improving my health status 1.75 1.64** 1.68 1.78* 1.57 1.59 
Intention 2.96 3.32** 3.30 2.77** 2.86 3.31 ** 
I intend to use a de-worming agent every six months for the 
rest of my life 2.99 3.46** 3.26 2.83** 3.02 3.48 ** 
I intend to start using de-worming agents (if not yet started 
using) 2.99 3.39** 3.30 2.99** 2.82 3.29 ** 
I intend to convince my friends who are HIV-positive to use 
de-worming agents every six months 2.83 3.21** 3.17 2.53** 2.73 3.25 ** 
 * Significant at 0.05; ** significant at 0.01; *** significant at 0.001; (R) indicates reversed items 
25
I. Outcome/impact of preventive care package project 
Morbidity history and hospitalization 
The expected outcome of the PCP project is to decrease incidence of HIV-related opportunistic infections 
among the target groups. The overall analysis shows a significance reduction in prevalence of morbidity 
in the three months before the endline survey as reported by respondents. Incidence of diarrhea, acute 
febrile illness, and respiratory infection significantly decreased in the intervention sites suggesting the 
contribution of the project intervention to the reduction of morbidity levels (Figure 1, Table 18). Visits to 
health facilities were significantly reduced in the intervention sites, while there was no significant 
difference in the level of hospitalization among respondents in the intervention sites and control sites. 
Diarrhea 
There was a significant reduction in diarrhea episodes, from 18 percent at baseline to 9 percent at endline 
in the intervention sites (χ2 test, p<0.01, odds ratio (OR) = 0.47). This suggests that 53 percent of the 
respondents in the intervention sites were protected from diarrhea. Although there was a significant 
decrease in episodes of diarrhea in the control sites, (χ2 test, p<0.01, OR 0.64), the 36 percent protection 
may be due to other national interventions. The 17 percent difference in protection is likely to be directly 
attributed to the PCP project. The frequency and mean incidence of diarrhea episodes among those who 
acquired diarrhea was also significantly reduced. 
Respiratory infections 
Respiratory infections significantly reduced in the intervention sites from a prevalence of 17 percent at 
baseline to 6 percent at endline, (χ2 test, p<0.01, OR=0.31). There was no significant change between 
endline and baseline respiratory infection in the control sites. The reduction of respiratory infection 
among respondents in the control sites was achieved in the absence of significant difference between 
respondents in the intervention sites and control sites at baseline, which implies a 69 percent protection 
against respiratory infection is attributable to the PCP project. Similarly, the frequency of respiratory 
infection episodes significantly decreased in the intervention sites over the course of the project. This 
reduction of respiratory infection could be as a result of increased immunity due to reduction of diarrhea 
incidence. 
Malaria 
Malarial infection among respondents residing in malaria prone areas during the three months prior to the 
survey did not show significant reduction in the intervention sites compared to control sites. The mean 
number of malaria incidence also remained insignificant between the intervention and control sites at both 
endline and baseline surveys. In 
general, the PCP project 
contribution to reduction of 
malaria-related morbidity was not 
noticeable. 
Febrile illnesses other than 
malaria 
There was a significant reduction 
in febrile illness from 30 percent 
at baseline to 20 percent at 
30.1 26.5 17.8 16.8 13.1 19.8 15.4 9.2 6.0 10.3
0
10
20
30
40
50
Fever other
than malaria
Malaria Diarrhea Resparatory
infection
Percentage
Skin infection
Type of illness
Figure 1. Morbidity status of respondents in intervention sites of
four regions at endline and basline
Baseline Endline
26
endline in the intervention sites (χ2 test, p<0.01, OR=0.57). This implies that the project produced a 
protection rate of more than 43 percent against febrile illnesses, as there was a significant increase in the 
incidence of febrile illness among respondents in the control sites from 21percent at baseline to 28 percent 
at endline (χ2 test, p<0.05, OR=.45). 
Skin diseases 
There was a comparable 3 percentage point reduction in skin disease among respondents in intervention 
and control sites at endline compared to baseline. This reduction is statistically insignificant between 
baseline and endline surveys among respondents in the intervention and control sites (χ2 test, p > 0.05). 
This might be a result of absence of specific intervention that is strong enough to reduce skin diseases as 
part of the PCP intervention. 
Table 18 Morbidity history of respondents at baseline and endline surveys, 2009-2011 
The Four Regions Addis Ababa 
Baseline Endline 
 Intervention Control Intervention Control Baseline End line 
Diarrhea 17.8 19.6 9.2 13.4* 17.7* 11.1 
Have diarrhea more than once last 
three months 53.1** 38.4 42.5 48.1 62.2** 28.6 
Mean frequency of diarrhea 2.11** 1.63 1.93 1.97 3.8 1.7 
Respiratory Infection 16.8 15.9 6.0 12.5*** 20.5* 13.3 
Mean number of respiratory infections 1.7 1.6 1.4 1.5 1.8*** 1.2 
Fever other than malaria 30.1*** 21.3 19.8 28.2** 34.6 32.5 
Mean number of fever condition 2.9* 2.7 3.0 2.7 
Skin Disease 13.1 13.4 10.3 10.9 21.9 18.8 
Malaria 21.1 19.4 14.4 17.7 - - 
Mean number of malaria incidence 1.8 1.8 1.9 1.8 - - 
* Significant at 0.05; ** significant at 0.01; *** significant at 0.001
Visits to health facilities 
Visits to health facilities have significantly decreased from 34 percent at baseline to 17 percent at endline 
among respondents in the intervention sites (χ2 test, p<0.01, OR=0.40), implying a protection rate of 60 
percent (Table 19). There was also a reduction in visits to the health facility among respondents in the 
control sites, from 35 percent to 27 percent (χ2 test, p<0.05, OR=0.68) during the implementation period 
and this has produced a protection rate of 32 percent. The difference in protection rate (28 percent) 
between respondents in the intervention sites and the control sites can be attributed to the PCP project 
intervention. Rate of hospitalization also decreased from 12 percent to 9 percent among respondents in the 
intervention sites; however this reduction is not statistically significant. There was an increase in mean 
hospital stay at time of hospital admission from 13.8 days to 22.2 days among those in the intervention 
sites (t-test, p<0.05). 
 
27
Table 19: Morbidity and hospitalization status of respondents at baseline and endline surveys, 2009-2011 
The Four Regions Addis Ababa 
Intervention sites Control sites 
Baseline End line Baseline End line Baseline End line 
(n=817) (n=883) (n=381) (n=402) (n=389) (n=300) 
Diarrhea 17.8 9.2** 19.6 13.4* 17.7 11.1* 
No of Episodes 
One 46.2 56.6 62.5 51.9 37.3 73.5** 
Two 24.8 22.4 20.8 32.7 15.2 2.9 
Three or more 29.0 21.1 16.7 15.4 47.1 23.5 
Respiratory Infection 16.8 6.0** 13.7 12.5 20.5 13.3* 
No of Episodes 
One 50.7 65.4 56.9 60.4 40.4 85.4** 
Two 23.5 23.1 25.9 18.8 36.8 7.3 
Three or more 25.7 11.5 17.2 20.8 22.8 7.3 
Malaria 26.5 15.4 ** 18.8 18.3 - - 
No of Episodes 
One 54.1 74.5** 56.2 46.7 - - 
Two 23.0 6.4 28.1 25.8 - - 
Three or more 23.0 19.1 15.6 22.6 - - 
Fever other than malaria 30.1 19.8** 21.3 28.2* 34.6 32.5 
No of Episodes 
One 22.4 20.2 39.7 22.9 16.3 46.0** 
Two 23.2 23.6 15.4 25.7 9.2 4.0 
Three or more 54.5 56.2 44.9 51.4 74.5 50.0 
Skin Disease 13.1 10.3 13.4 10.9 21.9 18.8 
No of Episodes 
One 54.0 65.9 64.7 57.1 38.7 69.5 
Two 21.5 14.3 21.6 14.3 27.4 18.6 
Three or more 21.5 19.8 13.7 28.6 33.9 11.9 
Visit to health facility in last 3 Months 33.8 16.9** 35.0 26.8* 35.1 24.5** 
Hospitalization in last 3 months 11.8 9.3 6.2 7.8 7.0 7.8 
* Significant at 0.05; ** significant at 0.01; *** significant at 0.001
28
V. CONCLUSION AND RECOMMENDATIONS 
A significantly higher proportion of respondents in the four regions received information/education about 
PCP at endline (87 percent in intervention sites, 58 percent in the control sites) compared to respondents 
who received information/education at baseline (36 percent in intervention sites, 26 percent in the control 
sites). The vast majority of respondents in the intervention sites were exposed to PCP kits containing 
soap, condoms, WuhaAgar/PUR, water container, de-worming medication and IEC material over the 
course of the project period. More than 78 percent of the respondents received the PCP kits through the 
project in the intervention sites, while only one percent received the kits in the control sites. 
A significantly higher proportion of beneficiaries in the intervention sites were having the PCP items 
(condoms, WuhaAgar, soap and de-worming agent) in their household compared to respondents in the 
control sites. Availability and utilization of condoms, WuhaAgar, and de-worming agents were 
significantly higher among respondents in the intervention sites compared to respondents in the control 
site and this was as a result of the project intervention. Consistent condom use and regular use of 
WuhaAgar treated water for drinking were significantly higher for both male and female respondents as a 
result of the project intervention. However, utilization of soap for washing hands at critical times and of 
ITNs to prevent malaria did not show significant increases as a result of the PCP project intervention. 
Multivariate logistic regression analysis demonstrated that the intervention was not effective in promoting 
consistent condom use among male respondents. On the other hand, female respondent in the intervention 
site were 3.2 time more likely to use condom consistently compared to female respondents in the control 
site, suggesting that consistent condom use was effective as a result of the intervention among female 
respondents. At endline, male and female respondents in the intervention site were 4.9 times more likely 
to use WuhaAgar treated water compared to respondents in the control sites, implying the effectiveness of 
WuhaAgar intervention in increasing utilization. ITN utilization was not improved due to the program 
intervention among male and female respondents. 
The PCP project has resulted in decreasing incidence of opportunistic infections; thus, decreasing 
morbidity from these opportunistic infections. The overall analysis shows a significant decrease in 
prevalence of morbidity as reported by study participants. Incidence of diarrheal diseases, respiratory 
infection, and febrile illnesses other than malaria were significantly reduced. However, there was no 
significant reduction in morbidity related to skin and malaria infection due to the PCP project. Frequency 
(episodes) of diarrhea, febrile illness other than malaria and respiratory infection was also significantly 
decreased. 
Based on the findings of the evaluation, the following recommendations are suggested for further 
consideration. 
1. The project has resulted in significant reduction in morbidity from common opportunistic infections, 
mainly diarrhea, respiratory infection and febrile illnesses other than malaria. Therefore there is a 
need to scale up the project interventions to reach PLWHA living in all regions. 
29
2. A considerable behavioral change to use the PCP items was observed as assessed by opportunity, 
ability and motivational factors. This indicates the need to strongly promote behavioral change for 
effective utilization of program interventions in future project initiatives. 
3. Using soap for washing hands and year round ITN utilization in malaria prone areas were not well 
addressed by the project. There is a need to review the current implementation strategy or change the 
strategy to promote use of these items. 
30
 VI. REFERENCES
Allen, S., Meinzen-Derr, J., Kautzman, M., et al. 2003. “Sexual behavior of HIV discordant couples after 
HIV counseling and testing,” AIDS, Vol 17, No 5 (2003): 733-40. 
Colindress R, Mermin J, Ezati E, et al. “Utilization of basic care and prevention package by HIV-infected 
persons in Uganda,” AIDS Care, 24 January, 2007. 
Federal Ministry of Health (FMoH). Single Point HIV Prevalence Estimate, Addis Ababa, Ethiopia, 2007. 
Lule, J.R., Mermin, J., Ekwaru, J.P., et al. “Effect of home based water chlorination and safe storage on 
diarrhea among persons with human immunodeficiency virus in Uganda,” American Journal of Tropical 
Medicine and Hygiene, Vol 73, No 5 (2005): 926-933. 
Katabira, E.T. “Epidemiology and management of diarrheal disease in HIV-infected patients,” 
International Journal of Infectious Disease Vol 3, No 3 (1999):164-7. 
Mermin, J., Lule, J., Ekwaru, J.P., et al. “Effect of cotrimoxazole prophylaxis on morbidity, mortality, 
CD4 cell count, and HIV viral load among persons with HIV in rural Uganda,” Lancet, Vol 364 
(2004):142-34. 
Mermin, J., Lule, J.R., & Ekwaru, J.P. “Association between malaria and CD4 cell count decline among 
persons with HIV,” Journal of Acquired Immunodeficiency Syndrome, Vol 41 (2006b),129-30. 
Population Reference Bureau (PRB). World Population Data Sheet, 2007. 
PSI Research Division. The dashboard: A tool for social marketing, 2005. 
Quick, R.E., Kimura, A., Thevos, A., et al. “Diarrhea Prevention Through household-Level Water 
disinfection and Safe Storage in Zambia,” American Journal of Tropical Medicine and Hygiene, Vol 66, 
No 5 (2002), 584-89. 
UNAIDS. Report on the Global AIDS Epidemic, 2008. 
www.hip.watsan.net\content\download\2032\10826\\file\HIP_AEDcx.pdf 
31
VII. APPENDIX 
Table A1: Reliability analysis, condom use 
Items Cronbach’s 
alpha 
No of 
items 
Questionnaire number of selected items 
Opportunity 
Social norms 0.56 3 v1022, v1023, v1025a 
Ability 
Knowledge on condom use 0.78 5 v1026, v1027, v1028, v1029, v1030 
Social support 0.84 6 v1031, v1032, v1033, v1034, v1035, v1036 
Self efficacy 0.86 3 v1037, v1038, v1039 
Motivation 
Outcome expectation 0.74 3 v1040, v1041, v1042 
Attitude 0.73 4 v1043, v1044, v1047, v1048 
Belief, condom strength 0.73 7 v1049, v1050, v1051, v1052, v1053, v1054, v1055 
Belief, decision making 0.53 3 v1056, v1057, v1058 
Fear 0.54 3 v1059, v1060, v1061 
Table A2: Reliability analysis, WuhaAgar use 
Items Cronbach’s 
alpha 
No of 
items 
Questionnaire number of selected items 
Opportunity 
Availability 0.84 6 v1139, v1140, v1141, v1141a, v1141b, v1141c 
Social norm 0.87 5 v1142, v1143, v1144, v1145, v1145a 
Ability 
Knowledge 0.64 5 v1146, v1147, v1148, v1149, v1149a 
Self efficiency 0.74 5 v1150, v1151, v1152, v1153, v1153a 
Social support 0.89 4 v1154, v1155, v1156, v1157 
Motivation 
Attitude 0.54 3 v1158, v1160, v1162 
Intention 0.67 4 v1163, v1164, v1165, v1165b 
Threat 0.56 3 v1166, v1168, v1169a 
Outcome expectation 0.83 5 v1170, v1171, v1172, v1172a, v1172b 
Table A3: Reliability analysis, ITN use in malaria endemic areas 
Items Cronbach’s 
alpha 
No of 
items 
Questionnaire number of selected items 
Opportunity 
Social norm 0.78 6 v1216, v1217, v1218, v1218a, v1218b, v1218c 
Ability 
Knowledge 0.49 8 v1219, v1221, v1222, v1223, v1224, v1224a, v1224b, v1224c 
Self efficiency 0.76 5 v1225, v1226, v1227, v1227a, v1227b 
Social support 0.92 4 v1228, v1229, v1230, v1230a 
Motivation 
Threat 0.64 6 v1231, v1232, v1233a, v1233b, v1233d, v1233e 
Outcome expectation 0.78 7 v1234, v1235, v1236, v1237, v1238, v1238a, v1238b 
Belief 0.64 4 v1241, v1241a, v1241b, v1241c 
32
Table A4: Reliability analysis, ORS use 
Items Cronbach’s 
alpha 
No of 
items 
Questionnaire number of selected items 
Opportunity 
Availability 0.80 4 v1411, v1412, v1413, v1413a 
Ability 
Knowledge 0.72 4 v1414, v1416, v1416a, v1416b 
Social support 0.86 4 v1417, v1418, v1419, v1419a 
Self efficiency 0.74 4 v1420, v1421, v1422, v1422a 
Motivation 
Outcome expectation 0.69 3 v1423, v1425, v1425a 
Attitude 0.62 4 v1426, v1427, v1428, v1428a 
Intention 0.67 4 v1429, v1430, v1431, v1431a 
Table A5: Reliability analysis, de-worming use 
Items Cronbach’s 
alpha 
No of 
items 
Questionnaire number of selected items 
Opportunity 
Availability 0.82 5 v1510, v1511, v1512, v1512a, v1512b 
Social norm 0.78 5 v1513, v1514, v1515, v1515a, v1515b 
Ability 
Knowledge 0.72 5 v1516, v1517, v1518, v1518a, v1518b 
Motivation 
Outcome expectation 0.77 5 v1519, v1520, v1521, v1521a, v1521b 
Intention 0.85 4 v1522, v1523, v1524, v1524a