May 2009 This publication was produced for review by the United States Agency for International Development. It was prepared by the University of Zambia and submitted to and reviewed by Chemonics International Inc. PREVENTING THE MEDICAL TRANSMISSION OF HIV IN ZAMBIA FINAL EVALUATION REPORT The author’s views expressed in this publication do not necessarily reflect the views of the United States Agency for International Development or the United States Government. PREVENTING THE MEDICAL TRANSMISSION OF HIV IN ZAMBIA FINAL EVALUATION REPORT Contract No. GHS-I-00-03-00025-00, Task Order 01 Acknowledgements The University of Zambia is greatly indebted to the Chemonics-led Zambia Medical Injection Safety Project (MISP) for funding the study and for the useful comments. Special acknowledgements are given to the District Directors of Health in Chingola, Monze, and Solwezi for the logistical support given to facilitate the smooth implementation of the study. Special tribute also goes to the various focal point persons and other local leaders in the three districts for providing information that was very relevant for the study. The University of Zambia would also like to express thanks to the research assistants and all the people in the districts who gave their precious time to go through the long list of questions during interviews and focus group discussions that were conducted. Their views and experiences are very insightful and have been invaluable in shaping our understanding of the issues relating to injection safety and infection prevention. Lastly we would like to thank our reviewers, particularly Dr. C. Mazimba and Mr. S. Chanda, for the helpful comments. iv CONTENTS List of Tables i Acronyms iii SECTION I Background 1 A. Statement of Problem 1 B. Purpose 1 C. Objective 2 D. Methodology 2 SECTION II Findings 5 A. Overall Performance Indicators 5 A1. Average Number of Medical Injections per Person per Year 5 A2. Proportion of Persons Age 15-49 Reporting Last Injection Was 6 Given With a Syringe/Needle from a New, Unopened Package A3. Post-Exposure Prophylaxis Provision and Immunization against 6 Hepatitis B B. Commodity Management and Procurement 8 B1. Stock-Outs of IP/IS Equipment 8 B2. Availability of Sufficient Injection Equipment 10 B3. Availability of Personal Protective Equipment 11 B4. Procurement Systems 12 C. Capacity Building and Training 12 C1. Health Care Providers’ Behavior 12 C2. Stick-Needle Injuries 14 C3. Use of Personal Protective Equipment 14 D. Behavior Change Communication 15 D1. Health Care Providers’ Perception of Patient Treatment Preference 15 D2. Patient Treatment Preference 16 E. Medical Waste Management 19 F. Policy Environment 21 G. Monitoring and Evaluation 21 SECTION III Discussion and Conclusions 23 A. Overall Performance Indicators 23 B. Commodity Management and Procurement 24 C. Capacity Building 24 D. Behavior Change Communication 24 E. Conclusions and Recommendations 25 v i LIST OF TABLES Table 1: Number of Observations Collected per Collection Tool and Districts Visited 3 Table 2: Average Number of Medical Injections per Person per Year by District, Facility, Type and Age Group 5 Table 3: Proportion of Patients Age 15-49 Reporting Last Injection Was Given With a Syringe/Needle from a New, Unopened Package 6 Table 4: Proportion of Facilities Reusing Needles and Syringes for Injection 6 Table 5: Facilities/Departments Providing Post-Exposure Prophylaxis to Staff after Injury 7 Table 6: Supervisors Reporting Hepatitis B Vaccine to Facility Staff 7 Table 7: Proportion of Health Workers Immunized Against Hepatitis B 8 Table 8: Percent of Facility Supervisors Reporting Stock-outs of IP/IS Equipment 9 Table 9: Percent of Facility Supervisors Reporting Stock-outs of IP/IS Equipment during Final Evaluation 9 Table 10: Providers Reporting Sufficient Quantities of Injection Equipment Available 10 Table 11: Quantities of Puncture-Proof Safety Containers Observed In Stock 10 Table 12: Availability of Personal Protective Equipment to Health care Workers 11 Table 13: Procurement Systems Used By Health Facilities As Reported By Facility 12 Table 14: Observed Behaviors of Health care Providers 13 Table 15: Proportion of Health care Workers Interviewed Reporting Needle-Stick Injuries 14 Table 16: Use Of Personal Protective Equipment in Sharps Disposal by Auxiliary Staff As Observed During Facility Inspection 15 Table 17: Percent of Providers Reporting Patient Preference for Injections When Presenting With Febrile Illness 16 Table 18: Percent of Patients Reporting Preference for Injections over Oral Medications 16 Table 19: Treatment Prescribed to Patients Interviewed 17 Table 20: Patients Who Received an Injection Reporting That Provider Suggested the Treatment 18 Table 21: Percent of Facilities in Which BCC Materials Were Appropriately Displayed 18 Table 22: Percent of Patients Reporting They Had Heard or Seen BCC Messages About Injection Safety 18 Table 23: Percent of Health Facilities with Satisfactory Disposal of Used Injection Equipment 19 Table 24: Primary Methods Used for Sharps Waste Disposal 20 ii Table 25: National IP/IS Guidelines Implemented at the Facility Level 21 Table 26: Supervisors Reporting That Facility Has an M&E Plan to Monitor IP/IS and Other Practices 21 Table 27: Supervisors Reporting IP/IS Activities in Current Facility Action Plans 22 iii ACRONYMS AIDS Acquired Immune Deficiency Syndrome AD Auto-disable BCC Behavior Change Communications DHMT District Health Management Team FGDs Focus Group Discussions GRZ Government of the Republic of Zambia HP Health Post HIV Human Immunodeficiency Virus IP Infection Prevention IS Injection Safety M&E monitoring and evaluation MISP Zambia Medical Injection Safety Project MOH Ministry Of Health PEPFAR President’s Emergency Plan for AIDS Relief PEP Post-exposure prophylaxis PPE Personal Protective Equipment RHC Rural Health Center SIGN Safe Injection Global Network UHC Urban Health Centre WHO World Health Organization PREVENTING THE MEDICAL TRANSMISSION OF HIV IN ZAMBIA 1 SECTION ONE Background The World Health Organization (WHO) estimates unsafe administration of injections in health care settings is responsible for 8 to 16 million cases of hepatitis B infections, 2.3 to 4.7 million cases of hepatitis C, and 80,000 to 160,000 cases of HIV infections annually.1 In response to this crisis the WHO, in collaboration with partners from the Safe Injection Global Network (SIGN), developed an intervention strategy aimed at reducing the incidence of unsafe and unnecessary injections. The main areas of focus of the SIGN are: 1. Behavior change of health care workers and patients to ensure safe injection practices and reduce unnecessary injections. 2. Ensuring availability of equipment and supplies. 3. Managing medical waste safely and appropriately. Zambia was one of 15 priority countries identified in the President’s Emergency Plan for AIDS Relief (PEPFAR) for preventing the spread of HIV infections. In 2004, in accordance with PEPFAR activities aimed at eliminating the preventable medical transmission of HIV, Chemonics International — in partnership with Jhpiego and the Manoff Group — began implementing the Zambia Medical Injection Safety Project (MISP). The project focuses on improving systems and practices related to infection prevention and injection safety (IP/IS). Since 2004 Chemonics has expanded the project to a national scale, with the MISP team providing training, procurement, follow-up and supportive supervision, and behavior change communication (BCC) programs in health care facilities in all 72 districts of the nine provinces of Zambia. A. Statement of the Problem The Global Burden of Health study conducted by the World Health Organization in collaboration with the Ministry of Health in 2000 (unpublished) showed that despite improvements in injection safety practices in Zambia, there were still gaps in knowledge, attitudes and practices among health care providers and the community at large that increase the risk of transmission of HIV and other blood-borne diseases. A baseline assessment of injection safety practices conducted by the Zambia Medical Injections Safety Project and the Ministry of Health (MOH) in May 2006 revealed that, notwithstanding improvements in specific areas, injections were frequently administered improperly and that the quality of the sharps boxes was often inadequate. A midterm evaluation conducted by the project in collaboration with the MOH in June 2008 also revealed areas requiring further attention. B. Purpose To assess the extent to which the project met its long-term objectives and to identify areas in which attention still needs to be paid. 1 World Health Organization. WHO/EHT/04.04 Safety of Injections. Global Facts and Figures. PREVENTING THE MEDICAL TRANSMISSION OF HIV IN ZAMBIA 2 C. Objective The main objective of the final evaluation was to assess progress over time compared to results from the baseline evaluation conducted in May 2006. To achieve this objective the evaluation specifically looked at: 1. Availability of IP/IS equipment and methods of managing stock-outs 2. IP/IS practices of health care providers 3. Availability of equipment/materials for the collection, transportation, and elimination of waste 4. Procedures and practices related to medical waste injection management activities 5. Availability and accessibility of reference documents (national policy, standards, guidelines) in health facilities 6. Experiences related to IP/IS in health facilities and the community of those patients (parents/families of those patients) who will have received injections on the day of the survey D. Methodology Before commencing the research exercise, an orientation meeting was convened between the Client (Chemonics International/MISP) and the Consultants (University of Zambia) at which the Client elaborated on the vision for the assignment. Issues pertaining to sample selection, tools for data collection, duration of the assignment, and logistics were discussed. The Consultants then carried out a review of the baseline and midterm reports as well as the end-of-evaluation research protocol. The documents, together with the monitoring and evaluation plan developed by the MISP team, gave the Consultants the necessary background information on MISP. Sixteen research assistants (with a combination of social science and medical backgrounds) were recruited and trained to assist the Consultants in data collection. The Consultants and MISP staff facilitated a one-day training workshop on March 28, 2009 at the University of Zambia. They introduced IP/IS concepts and trained the research assistants to accurately observe IP/IS practices. The research assistants were divided into three data collection groups by district (Chingola, Monze, and Solwezi). During the baseline evaluation the MISP team picked Chingola, Monze, and Solwezi to survey because of their representation of urban versus rural settings and the diversity of health facilities accessible in each district. These same districts were surveyed to ensure comparability of results. D1. Data Collection Data collection activities were conducted between April 1 and April 15, 2009 and involved an inventory of district facilities, provider observations, and interviews with in￾charges, prescribers, providers, and waste handlers. PREVENTING THE MEDICAL TRANSMISSION OF HIV IN ZAMBIA 3 The survey team collected data in 63 health facilities in Chingola, Solwezi, and Monze through purposive selection (meaning a non-representative non-random convenience sampling method): 13 health facilities (1 district hospital and 12 health centers/health posts) in Chingola, 20 health facilities (1 district hospital and 19 heath centers/health posts) in Monze, and 30 health facilities (1 district hospital and 29 health centers/health posts) in Solwezi. Before data collection, the teams presented themselves at the District Health Management Team (DHMT) offices in each district and held meetings with the district directors of health. At these meetings, maps of all the health facilities in each district were given to the teams and plans on how to execute the assignment were discussed. Immediately after the meetings, the teams went into the field to commence data collection. Each team was assigned a district focal point person who was familiar with the location of the facilities as well as the terrain in the study sites. Both quantitative and qualitative data were collected to inform the evaluation, using the same questionnaires from the baseline and midterm evaluations, to ensure comparability across studies. A number of data collection methods including direct observation, face-to face interviews, and focus group discussions were used. The main sources of information/data included: procurement/stores managers in charge of IP/IS equipment, medications, and vaccines; injection prescribers; injection providers; facility managers (in-charges); waste handlers; and patients receiving services at the facility at the time of the survey. The breakdown of sources of information and tools in the three sites is presented below. Table 1. Number of Observations Collected Per Collection Tool and Districts Visited Tool Total at Baseline Total at Midterm Total at Final Disaggregated by District Total at Final Chingola Monze Solwezi Facility inventory observations 59 69 13 20 30 63 Injection provider interviews 162 83 22 22 33 77 Injection provider observations 146 208 65 79 35 179 Injection prescriber interviews 97 83 19 26 30 75 Supervisor interviews 73 77 15 20 30 65 Interviews with waste handlers 77 66 13 20 30 63 Client/patient exit interviews 282 247 60 71 76 207 Focus group discussions 9 9 3 3 3 9 PREVENTING THE MEDICAL TRANSMISSION OF HIV IN ZAMBIA 4 A total of 738 activities (487 interviews, 179 observations, 63 inventories, and 9 focus group discussions) were conducted to provide data for the final evaluation. D2. Data Management To ensure data quality, all questionnaires were checked for errors and consistency on the day of administration before being entered into the database. Each team had one person designated for data entry so that data entry started in the field to ensure completion of the evaluation within the stated time. Each data-collection activity was led by a team leader with a medical background. Their role was to participate in and closely supervise the data-collection exercise. There were daily meetings to discuss the events of the day and plans for the next day. The Lead Consultant travelled to Chingola on April 3, 2009 to ensure quality data collection and check for progress in the field. He also participated in some of the daily preview and review meetings. The team used thematic analysis for qualitative data and the Statistical Package for Social Sciences (SPSS) software for quantitative data analysis. D3. Limitations The baseline survey was conducted in May 2006. When the baseline survey was designed, budgetary and time constraints limited surveying to three districts. The three districts allowed comparison between urban and rural settings in districts that had yet to receive MISP activities but would be eventual recipients and which possessed all types and levels of health facilities. However, the locations chosen and the sample sizes surveyed were not statistically representative samples of the entire population and were never intended as such. The team was attempting to gain some understanding of IP/IS activities in Zambia and gaps in knowledge, practices, and behaviors. In much the same way, the data gathered during the final evaluation is an attempt to identify trends in infection prevention and injection safety and highlight areas still requiring further attention, but should not be interpreted as representative of the situation in Zambia as a whole. Interventions in the three districts occurred after the baseline evaluation, between 2006 and 2007. The midterm evaluation was not conducted until spring 2008 and the final evaluation in spring 2009. It is difficult to make meaningful comparisons or analyze trends with such a short time lapse in between the midterm and final evaluations, and for that reason the midterm evaluation has not been included in this report. Similar constraints of budget and time also resulted in the MISP team having to utilize the Consultants services, introducing a possible element of bias to the data collection and interpretation process, as the Consultants were not as familiar with the data collection instruments and definitions as the original surveyors. However, the MISP team, having worked closely with the Consultants, can with a degree of confidence say that the information provided is in line with current literature on the subject and illustrates that while progress is being made, injection safety continues to be an important issue and requires further attention in Zambia to meet international standards. PREVENTING THE MEDICAL TRANSMISSION OF HIV IN ZAMBIA 5 SECTION II FINDINGS A. Overall Performance Indicators As a PEPFAR recipient, the MISP project regularly reports on the following medical injections indicators: • Average number of injections per person per year for persons age 15-49 years • Proportion of individuals age15-49 reporting that their last health care injection was given from a syringe set from a new, unopened package • Number of health care providers trained in IP/IS The number of health care providers trained was recorded through project training activities and therefore not directly measured through evaluation studies. The remaining PEPFAR indicators were explored through exit interviews of patients at health care facilities. A1. Average Number of Medical Injections per Person per Year The average number of injections per person per year was found to be 2.96 (as shown below in Table 2). The average number of injections received by any person in the three age group categories per year has increased for both the 0-14 and the 15-49 age groups from the baseline survey, although a decrease was noticed in the 50+ age group. Table 2. Average Number of Medical Injections per Person per Year by District, Facility Type, Age Group and Gender Stud y Age Group District Facility Gender Overal l Chingol a Monz e Solwez i RH C UH C Hosp . Missio n Hosp. Mal e Femal e Base￾line 0-14 (n=100) 0.40 1.68 2.32 1.89 0.50 2.00 3.38 2.28 1.48 1.91 15-49 (n=118) 0.91 1.05 3.20 2.06 0.60 3.67 1.64 1.49 2.67 2.18 50+ (n=27) 0.00 1.07 4.90 0.82 0.00 13.33 3.33 3.05 1.50 2.46 Final 0-14 (n=38) 3 2 3 3.40 1.71 6 4 2 3.57 2.66 15-49 (n=151) 2.4 3.1 3.86 3.38 2.58 3.4 3.38 3.78 2.68 3.01 50+ (n=18) 1 0 3 3 1 0 0 3 2.33 1.33 Increases within age groups were witnessed in all districts, though decreases were most noticeable amongst the 50+ age group within hospital settings. PREVENTING THE MEDICAL TRANSMISSION OF HIV IN ZAMBIA 6 A2. Proportion of Persons Age 15-49 Reporting Last Injection Was Given With a Syringe/Needle from a New, Unopened Package Like many other countries in the sub-Saharan region, Zambia has also adopted PEPFAR and WHO protocols where use of single-use needles and syringes is encouraged. Table 3 provides information on the proportion of patients receiving injections from new, unopened packages, and Table 4 shows the proportion of facilities reusing needles and syringes. Table 3. Proportion of Patients 15-49 Years of Age Reporting Last Injection Given With a Syringe/Needle from a New, Unopened Package Study District (%) Facility (%) Overall (%) Chingola Monze Solwezi RHC UHC Hosp. Mission Hosp. Baseline (n=59) 100 100 97.2 100 100 92.9 100 98.3 Final (n=207) 100 100 100 100 100 100 100 100 Table 4. Proportion of Facilities Reusing Needles and Syringes for Injection Health care worker Study District (%) Facility (%) Overall (%) Chingola Monze Solwezi RHC UHC Hosp. Mission Hosp. Reuse of syringes or needles for immunization injection Baseline (n=59) 0 0 0 0 0 0 0 0 Final (n=63) 7.7 0 0 2.2 0 0 0 2.5 Reuse of syringes or needles for curative injections Baseline (n=59) 0 0 0 0 0 0 0 0 Final (n=63) 7 0 6.7 6.7 0 0 0 4.4 A significant success in injection safety promotion is evidenced by the 100 percent of patients 15-49 reporting that they received their last injection from a new, unopened package. There was a slight increase in the number of facilities witnessed to be still reusing syringes and needles for immunization and curative injections (2.5 percent and 4.4 percent respectively). It is the hypothesis of the researchers that this is a sampling error. It is also important to note that the trend predominates in rural health centers, suggesting a need for further interventions in harder-to-reach areas. A3. Post-Exposure Prophylaxis Provision and Immunization against Hepatitis B Administration of injections by health personnel exposes them to blood-borne diseases such as HIV/AIDS and hepatitis B. This evaluation prioritized the need to obtain information on the number of health facilities providing post-exposure prophylaxis (PEP) and the percentage of health care workers immunized against Hepatitis B in response to the emphasis placed on the importance of this indicator by the project. Close to half (42.6 percent) of health facilities in this survey as reported by supervisors and more than half (62.4 percent) the surveyed facilities reported by providers themselves state that PEP is provided, well above the target of 40 percent of providers providing PEP. PREVENTING THE MEDICAL TRANSMISSION OF HIV IN ZAMBIA 7 Table 5. Facilities/Departments Providing Post-Exposure Prophylaxis to Staff after Injury Health care worker Study District (%) Facility (%) Overall Ching (%) ola Mon ze Solw ezi RH C UH C Hos p. Mission Hosp. Supervisors reporting facilities provide PEP to staff Baseline (n=67) 53.9 34.6 32.1 12. 2 33. 3 92. 3 85.7 37.3 Final (n=64) 71.4 40 16.7 20. 9 61. 5 100 50 42.6 Providers reporting facilities provide PEP to staff Baseline (n=143) 43.2 57.5 51.5 38. 2 19. 1 84. 2 100 51.5 Final (n=69) 81.8 77.8 27.6 39. 4 32. 5 100 80 62.4 While the overall trend shows improvements in the percentage of providers and supervisors reporting PEP from baseline to final, a decrease in the percentage of providers and supervisors reporting PEP provided to staff was noticeable in Solwezi. This is especially relevant because of the high number of rural health centers in Solwezi and because the complexity of traveling and accessing these centers makes routine provision for these facilities difficult. Approximately 13 percent of health facilities reported offering hepatitis B immunization at final evaluation, up from 2.7 percent at baseline, which is a noticeable increase. However, up-take of hepatitis B immunization decreased from 10.7 percent to 3.3 percent among health care providers, with a moderate increase witnessed among waste handlers. Due to time constraints, survey teams were not able to confirm results from supervisors against a registry of immunizations or health care workers medical records, so reporting is based on perception, which may not be accurate. Table 6. Supervisors Reporting Hepatitis B Vaccine Provided to Facility Staff Health care worker Study District (%) Facility (%) Overall (%) Chingola Monze Solwezi RHC UHC Hosp. Mission Hosp. Supervisors reporting Hepatitis B vaccine provided to staff Baseline (n=71) 15.4 0 0 2.3 16.7 0 0 2.7 Final (n=63) 26.7 11 3.3 9.3 14.3 20 0 13.3 PREVENTING THE MEDICAL TRANSMISSION OF HIV IN ZAMBIA 8 Table 7. Proportion of Health Workers Immunized Against Hepatitis B Health care worker Study District (%) Facility (%) Overall (%) Chingola Monze Solwezi RHC UHC Hosp. Mission Hosp. Health care provider Baseline (n=159) 25.6 9.3 2.7 8.2 28 7.3 0 10.7 Final (n=65) 6.6 0 3.3 2.3 0 20 100 3.3 Waste handler Baseline (n=77) 0 3.7 0 0 0 0 12.5 1.3 Final (n=62) 7.7 0 3.3 4.8 0 0 0 3.7 In the case of health care providers immunized against hepatitis B, the sample size surveyed at final was approximately 40 percent smaller, making comparison of these results difficult. One hypothesis for the lower number of health care providers immunized against hepatitis B at the time of the final survey is that staff attrition may have resulted in those initially immunized moving on, and a lack of sensitization of new people to the importance of immunization upon arrival at the new facility. B. Commodity Management and Procurement MISP worked closely with the MOH to build capacity in the financing, procurement, and distribution of appropriate levels of injection equipment. The result has been systematic data collection on the stock of injection equipment available at health facilities. Information collected includes stock-outs of IP/IS equipment, quantities of injection equipment available, quantities of puncture-proof safety containers in stock, availability of personal protective equipment, and standard procurement systems used by health facilities. B1. Stock-Outs of IP/IS Equipment The project monitored stock levels of essential IP/IS equipment — including disposable syringes, disposable needles, auto-disabled (AD) syringes, and sharps boxes — that should be continuously available at all facilities. The final evaluation indicated a sharp decrease in stock-outs of IP/IS equipment compared to the baseline (disposable syringes stock-outs dropped from about 45.2 percent to 9 percent). PREVENTING THE MEDICAL TRANSMISSION OF HIV IN ZAMBIA 9 Table 8. Percent of Facility Supervisors Reporting Stock-Outs of IP/IS Equipment Equipment Study District (%) Facility (%) Overall Chingol (%) a Monze Solwezi RHC UHC Hosp. Mission Hosp. Disposable Syringe Baseline (n=73) 30.8 36.7 60 46.7 16.7 46.2 55.6 45.2 Final (n=68) 13.3 0 13.8 11.7 14.2 0 0 9 Disposable needles Baseline (n=73) 15.4 41.4 55.2 45.5 16.7 46.2 37.5 42.3 Final NA NA NA NA NA NA NA NA AD Syringes Baseline (n=61) 20 36.4 44.8 32.6 50 33.3 100 37.7 Final (n=68) 13.4 0 6.7 4.7 14.2 0 0 6.7 Sharps Boxes Baseline (n=65) 50 44 70 47.6 50 90.9 66.7 56.9 Final (n=68) 20 5 6.8 7 14.2 25 0 10.6 Table 9. Percent of Facility Supervisors Reporting Stock-Outs of IP/IS Equipment during Final Evaluation Equipment Length District (%) Facility (%) Overall (%) Chingola Monze Solwezi RHC UHC Hosp. Missio n Hosp. Disposable Syringes and needles Never 86.7 100 82.8 100 100 100 100 89.8 Less than 1 month 13.3 0 6.9 0 0 0 0 6.7 More than 3 months 0 0 1 0 0 0 0 0.3 AD Syringes Never 86.6 100 92.9 95.3 84.6 100 100 93.2 Less than 1 month 6.7 0 7.1 4.7 7.7 0 0 4.6 More than 3 months 6.7 0 0 0 7.7 0 0 2.2 Sharps Boxes Never 80 95 93.2 93 85.7 75 100 89.4 Less than 1 month 13.3 5 3.4 4.7 7.1 0 0 7.2 More than 3 months 6 0 3.4 2.3 7.1 0 0 3.1 Whereas information in the baseline was collected separately for needles and syringes, the final evaluation instrument combined the two instruments making comparison among the survey points difficult. What was captured, however, were instances of stock-outs of IP/IS equipments for cases categorized as “never,” “less than one month,” and “more than three months.” The majority of health institutions indicated that they never run out of syringes and needles (Table 8). From Table 9, it is possible to see that with the PREVENTING THE MEDICAL TRANSMISSION OF HIV IN ZAMBIA 10 exception of Solwezi, most health facilities surveyed did not experience stock-outs of any essential equipment of more than three months. B2. Availability of Sufficient Injection Equipment Tables 10 and 11 present information on health providers reporting availability of sufficient quantities of injection equipment and puncture-proof safety containers at their health facilities in study districts. Table 10. Percent of Providers Reporting Sufficient Quantities of Injection Equipment Available Equipment Study District (%) Facility (%) Overall (%) Chingola Monze Solwezi RHC UHC Hosp. Mission Hosp. New, single￾use syringes and needles available Baseline (n=161) 90.1 95.5 93.2 95.3 96.2 85.4 100 93.2 Final (n=77) 100 100 97 100 97.5 100 50 99 Sharp boxes available to safety dispose of sharps Baseline (n=161) 61.4 63.6 24.7 50.6 80.8 19.5 11.1 45.3 Final (n=77) 100 100 96.2 97.1 100 100 100 98.7 Almost all health facilities (99 percent) reported having sufficient quantities of new, single-use syringes and needles available with the exception of a noticeable decrease in mission hospitals (from 100 percent to 50 percent). Table 11. Quantities of Puncture-Proof Safety Containers Observed In Stock Number of puncture￾proof safety containers Study District (%) Facility (%) Overall Chingola Monze Solwezi RHC UHC Hosp (%) . Mission Hosp. 0 Baseline 77.8 50 42.3 45.5 55.6 100 75 50.9 Final 7.7 10.5 3.3 6.8 8.3 0 0 7.2 1-4 Baseline 10 17.4 46.2 30.4 33.3 0 0 28.8 Final 23.1 26.3 23.3 20.5 33.3 0 0 24.2 5-9 Baseline 0 26.1 3.9 15.2 0 0 0 11.9 Final 0 10.5 26.7 22.7 0 0 0 12.4 10-20 Baseline 11.1 8.3 7.7 9.1 11.1 0 0 8.5 Final 30.8 26.3 26.7 29.5 0 0 50 27.9 20+ Baseline 0 0 0 0 0 0 0 0 Final 50 50 44 44 58.3 100 100 47.1 Similarly, 98 percent of health facilities have sharp boxes available, with only 7.2 percent of facilities having no puncture-proof safety containers available, down from 50.9 percent at baseline. There was a noticeable percentage decrease in the stocking of 1-4 puncture￾proof safety containers from baseline, however in all other denominations there were PREVENTING THE MEDICAL TRANSMISSION OF HIV IN ZAMBIA 11 noticeable increases, with 47.1 percent of health facilities reporting having more than 20 puncture-proof safety containers. It is important to point out that at baseline no hospitals had zero puncture-proof containers. B3. Availability of Personal Protective Equipment (PPE) The availability of uniforms, surgical gloves, gumboots, face masks, and plastic aprons improved significantly between baseline and final evaluation (as seen in Table 12), indicating an increased demand for PPE among health care workers and the recognition of its importance by facility managers and those procuring the equipment. There was a decrease in the availability of some personal protective equipment such as heavy duty gloves, utility gloves, and examination gloves. The availability of heavy duty gloves decreased from 49.2 percent to 27.3 percent, and from 88.1 percent to 35.8 percent in the case of examination gloves, between baseline and final evaluation. Table 12. Availability of Personal Protective Equipment to Health care Workers Personal Protective Equipment Study District (%) Facility (%) Overall C (%) hingola Monze Solwezi RHC UHC Hosp. Mission Hosp. Uniform Baseline (n=59) 100 12.5 19.2 81.8 77.8 50 25 28.8 Final (n=63) 53.8 60 30 40 58.3 50 100 47.9 Surgical gloves Baseline (n=59) 77.8 37.5 46.2 43.2 66.7 50 50 47.5 Final (n=63) 46.2 30 76.7 60 41.7 100 50 51 Gumboots Baseline (n=59) 0 20.8 7.7 13.6 0 0 25 11.9 Final (n=63) 38.5 80 36.7 46.7 50 100 100 25.1 Facemasks/ Goggles Baseline (n=59) 0 8.3 7.7 6.8 0 0 25 6.8 Final (n=63) 38.5 15 30 24.4 16.7 100 100 22.8 Heavy duty gloves Baseline (n=59) 88.9 41.7 42.3 40.9 77.8 50 75 49.2 Final (n=63) 38.5 85 43.3 55.6 41.7 100 100 27.3 Utility gloves Baseline (n=59) 77.8 41.2 42.3 40.9 66.7 50 75 47.5 Final (n=63) 30.8 30 30 26.7 33.3 100 50 20.3 Examination gloves Baseline (n=59) 88.9 95.8 80.8 86.4 88.9 100 100 88.1 Final (n=63) 30.8 50 76.7 62.2 33.3 100 100 35.8 Plastic aprons Baseline (n=59) 11.1 12.5 11.5 9.1 11.1 50 75 11.9 Final (n=63) 30.8 85 60 62.2 50 100 100 30.3 As a result of interventions made by the MISP project, demand for much of the equipment above increased as health care workers became sensitized to the importance of PPE in protecting themselves and their patients. This is witnessed in the decrease in PREVENTING THE MEDICAL TRANSMISSION OF HIV IN ZAMBIA 12 availability of all types of gloves (with the exception of surgical gloves), which is not surprising considering the frequency that gloves are used in a health care setting. B4. Procurement Systems There are two procurement systems often used by health service providers — the push system and the demand-based system. (In some instances both systems are used). Because of its responsiveness to the needs of the individual institution, the MISP project advocated use of the demand-based system during trainings and follow-up supervision visits to health facilities. The push system relies on pre-determined assessments of what is needed by a facility usually during planning stages, but is not necessarily as responsive to changing needs and demands. Table 13. Procurement Systems Used by Health Facilities as Reported by Facility Managers Procurement system Study District (%) Facility (%) Overall (%) Chingola Monze Solwezi RHC UHC Hosp. Mission Hosp. Push System Baseline (n=72) 7.7 0 3.33 2.2 0 7.7 0 2.7 Final (n=65) 0 35 3.3 14 7.1 0 0 12.8 Demand￾Based Baseline (n=72) 92.3 100 90 95.6 100 84.6 100 93.1 Final (n=65) 100 45 83.3 69.8 85.7 100 100 76.1 Both Baseline (n=72) 0 0 6.7 2.2 0 7.7 0 2.7 Final (n=65) 0 20 13.3 16.3 7.1 0 0 11.1 Despite efforts of the MISP project, the MOH continues to rely on centralized procurement systems that may not be addressing the needs of health facilities adequately. Data presented in Table 13 shows an increase in the percentage of health facilities using the push system, from about 2.7 percent at baseline to 12.8 percent at final evaluation. Given the decrease in the use of demand-driven systems between the baseline and final evaluation (93 percent to 76 percent), more needs to be done to support systems that are responsive to individual facilities’ needs. C. Capacity Building and Training The project aimed to improve health workers’ practices through capacity building and by reinforcing best practices. C1. Health Care Providers’ Behavior Noticeable improvements were seen in important hygiene behaviors, with a marked increase in the number of health care providers observed washing hands before injection (up from 26.4 percent at baseline to 73.6 percent at final evaluation) and after injection PREVENTING THE MEDICAL TRANSMISSION OF HIV IN ZAMBIA 13 (up from 35.5 percent to 71 percent). In addition, safe disposal of sharps in boxes was observed 99.6 percent of the time at final evaluation. Also, the proportion of health care providers observed leaving a needle inserted in the vial for multiple doses dropped from 42.9 percent at baseline to 29 percent at final evaluation. The proportion of health care providers observed recapping increased from 12.6 percent at baseline to about 17 percent at final evaluation. It is possible that evaluators did not understand that recapping is often necessary in surgical and dental procedures and that the project promoted one-hand recapping. This is a serious data limitation. Table 14. Observed Behaviors by Health Care Providers Behavior/ Practice Study District (%) Facility (%) Overall Chingola Monze (%) Solwezi RHC UHC Hosp. Mission Hosp. Patient shown new needle/syringe Baseline (n=146) 43.2 56.4 50 54.2 50 39.4 57.1 46.6 Final (n=179) 40 40 57.1 37.5 46.1 9.1 65.2 45.7 Recapped after injection Baseline (n=143) 23.7 11.1 6.7 6.4 8 27.3 28.6 12.6 Final (n=179) 38.5 6.3 5.7 9.4 28.9 36.4 0 16.8 Disposal in sharps box Baseline (n=130) 60.6 84.1 79.2 87.8 83.3 37.5 62.5 76.2 Final (n=179) 100 98.7 100 100 100 100 96.3 99.6 Hands washed before injection Baseline (n=144) 23.7 32.6 23.3 28.2 8 33.3 37.5 26.4 Final (n=179) 86.2 71.8 62.9 70.3 78.9 63.6 80.8 73.6 Hands washed after injection Baseline (n=138) 39.5 41.3 25.9 36.7 24 36.7 37.5 35.5 Final (n=179) 90.6 45.3 77.1 63.5 77.3 72.7 54.2 71 Needle left inserted in vial Baseline (n=120) 61.8 26.8 42.2 37.3 47.6 52 42.9 42.5 Final (n=179) 27.7 22.2 37.1 61.9 59.2 36.4 52.4 29 Comparisons by facility type indicate that the proportion of injection providers observed showing a patient a new needle or syringe dropped significantly in all types of health facilities, except mission hospitals, between 2006 and 2009. The proportion of health care providers observed recapping increased from 6.4 percent to 9.4 percent in rural health centers and from 8 percent to 28.9 percent in urban health centers, although the proportion declined in hospitals from 28.6 percent to 0 percent during the reference period. Furthermore, the proportion of injection providers observed leaving a needle inserted in a vial for the purpose of drawing several doses decreased overall. However, in rural health centers, urban health centers, and mission hospitals, increases were noted between 2006 and 2009. PREVENTING THE MEDICAL TRANSMISSION OF HIV IN ZAMBIA 14 Encouragingly, the proportion of health care providers disposing needles/syringes in sharps boxes increased significantly from 87.8 percent, 83.3 percent and 37.5 percent (in rural health centers, urban health centers, and hospitals, respectively) to 100 percent. At mission hospitals, the increase was from 62.5 percent to 96.3 percent between baseline and final evaluation. Also encouraging is the finding that the proportion of health care workers observed washing hands before and after giving an injection increased in all facility types. C2. Needle-Stick Injuries The project aimed to reduce the incidence of needle-stick injuries among health care workers, while simultaneously creating an environment in which workers readily report needle-stick injuries and receive appropriate care to mitigate disease transmission. The proportion of health care providers reporting needle-stick injuries dropped from an average of 17.6 percent at baseline to 6.6 percent at final evaluation. The drop in the proportion of health care providers reporting needle-stick injuries was particularly remarkable in Monze, where it dropped by more than 23 percentage points between baseline and final evaluation, compared to 4.5 and 7.8 percent drops for Chingola and Solwezi, respectively. Table 15. Proportion of Health Care Workers Interviewed Reporting Needle-Stick Injuries Position Study District (%) Facility (%) Overall (%) Chingola Monze Solwezi RHC UHC Hosp. Mission Hosp. Health care provider Baseline (n=159) 13.6 27.9 13.9 18.1 11.5 14.6 44.4 17.6 Final (n=77) 9.1 4.5 6.1 2.5 8 16.7 20 6.6 Waste handler Baseline (n=76) 0 7.7 3.3 4.4 0 0 14.3 3.9 Final (n=63) 7.7 10 3.3 7.3 7.1 0 0 7 As shown in Table 15 the incidence of needle-stick injuries increased among waste handlers from 3.9 to 7 percent between baseline and final evaluation. Traditionally support staff are forgotten in the compliance of safety measurements and are not always supported properly. They often lack sufficient provision of PPE and do not feel empowered to advocate for themselves. PREVENTING THE MEDICAL TRANSMISSION OF HIV IN ZAMBIA 15 C3. Use of Personal Protective Equipment The use of PPE by auxiliary staff (waste handlers) — including uniforms, gumboots, facemasks, heavy duty gloves, and plastic aprons — all increased during facility inspections, which again suggests an increase in demand. However, as noted previously, the ability of facilities to meet demand through their procurement systems still needs to be improved. Again availability of examination gloves, utility gloves, and overalls declined by 46.1, 17.1, and 75.7 percentage points respectively, suggesting a continued need to address procurement issues for some of the more in-demand equipment. Table 16. Use of Personal Protective Equipment in Sharps Disposal by Auxiliary Staff as Observed During Facility Inspection Personal protective equipment Study District (%) Facility (%) Overall Chingola Monze Solwezi RHC UHC Hosp. (%) Mission Hosp. Uniform Baseline (n=55) 100 12.5 30.8 20.5 77.8 100 50 33.9 Final (n=63) 61.5 30 43.3 42.2 41.7 100 50 44.9 Overalls Baseline (n=55) 100 83.3 100 93 100 100 75 93.2 Final (n=63) 15.4 20 17.2 13.3 16.7 100 50 17.5 Gumboots Baseline (n=55) 11.1 12.5 15.4 13.6 11.1 0 25 13.6 Final (n=63) 53.8 75 56.7 66.7 41.7 100 50 61.8 Facemasks Baseline (n=55) 0 0 7.7 4.6 0 0 0 3.4 Final (n=63) 23.1 5 23.3 15.6 16.7 50 50 17.1 Heavy duty gloves Baseline (n=55) 66.7 20.8 42.3 31.8 66.7 50 25 37.3 Final (n=63) 46.2 80 46.7 60 41.7 50 100 57.6 Utility gloves Baseline (n=55) 44.4 16.7 34.6 27.3 44.4 50 0 28.8 Final (n=63) 0 15 20 20 0 0 0 11.7 Exam Gloves Baseline (n=55) 44.4 70.8 73.1 75 55.6 50 25 67.8 Final (n=63) 7.7 20 36.7 28.9 16.7 0 0 21.7 Plastic Aprons Baseline (n=55) 0 4.2 3.9 4.6 0 0 0 3.4 Final (n=63) 15.4 80 56.7 60 33.3 50 100 50.7 D. Behavior Change Communication D1. Health Care Providers’ Perception of Patient Treatment Preference One of the objectives of the project has been to reduce demand for and provision of unnecessary injections by changing the beliefs of providers and community members that injections are more effective than oral medications. Injection prescribers and providers perceived that their patients’ preference for injections decreased (see Table 17). It is difficult to determine what percentage of this decrease is directly attributable to the behavior change communication initiatives undertaken by the project and is a measure of PREVENTING THE MEDICAL TRANSMISSION OF HIV IN ZAMBIA 16 providers’ perception of patient preferences, and not actual patient preferences, as shown in Table 18. Table 17. Percent of Providers Reporting Patient Preference for Injections When Presenting With Febrile Illness Provider Study District (%) Facility (%) Overall Chingola Monze (%) Solwezi RHC UHC Hosp. Mission Hosp. Injection providers reporting patient preference Baseline (n=148) 62.5 38.5 58 52.6 70.8 54.1 22.2 54.1 Final (n=77) 45.5 33.3 56.3 47.4 44 83.3 20 45 Injection prescribers reporting patient preference Baseline (n=96) 75 39 66.7 59.7 75 41.7 30 56.3 Final (n=75) 36.8 38.5 63.3 50 47.4 40 50 45.9 D2. Patient Treatment Preference Table 18 presents findings on patients’ preference for injections versus an equivalent oral medication. The findings indicate no significant difference in patient preference for injection between the time of the baseline and final evaluations. Table 18. Percent of Patients Reporting Preference for Injections over Oral Medications Preference Study District (%) Facility (%) Overall (%) Chingola Monze Solwezi RHC UHC Hosp. Mission Hosp. Preferred Injection Baseline (n=170) 64.1 53.9 44.6 48.3 73.3 47.1 37.5 51.2 Final (n=207) 51.7 54.9 4.61 45.2 52.4 63.6 53.8 50.9 Examination of focus group discussions supports the evidence that a significant proportion of patients still prefer injections. As detailed below, patients’ responses to the question of preference show a multitude of reasons for why there may not have been significant change in this indicator. PREVENTING THE MEDICAL TRANSMISSION OF HIV IN ZAMBIA 17 Figure 1. Results of Focus Group Discussion Suggesting Patient Preference for Injection • Feel that the injections are more powerful and better. • “We want injections to heal quickly.” • “We prefer injections because injections are stronger, work faster with almost 100% of curing common diseases.” • Injections are better especially on children because they tend to vomit when they take drugs orally. • Injections are stronger and always work better than oral drugs. • All injections are better and heal faster, because the medicine goes directly to the blood. • When patients are given injections, they quickly recover, even when people are treated for malaria or TB. • Oral drugs are not as effective as injections. • Injections and IV-fluids are better for patients who cannot swallow. • Injections and IV-fluids go directly to the blood and therefore work faster. • “We prefer injections because they are very effective.” Diseases for which injections were believed to be particularly effective include TB, malaria, diarrhea, sharp pains, syphilis, rash, sores, snake bites, dog bites, crocodile bites, toothache, pain in the waist area, bilharzias, abscesses, and blood in the stool. Injections are also believed to be more effective for patients who are weak or very ill, have difficulties in swallowing oral medicines, such as those who vomit after taking oral medications, and children. The conclusion that patients prefer injections is further supported by an increase in the proportion of injection prescriptions from 39.7 percent in 2006 to more than half (50.9 percent) in 2009. On the other hand, oral medication prescriptions dropped from 50.5 percent to 33.6 percent between 2006 and 2009. The rest of the results are presented in Table 19. Table 19. Treatment Prescribed to Patients Interviewed Treatment Study District (%) Facility (%) Overall (%) Chingola Monze Solwezi RHC UHC Hosp. Mission Hosp. Injection Baseline (n=277) 43.8 32.7 44 21.7 33.3 59 21.7 39.7 Final (n=207) 51.7 54.9 46.1 45.2 52.4 63.6 53.8 50.9 Oral Baseline (n=277) 37.5 57.7 49.6 52.8 50 30.8 65.2 50.5 Final (n=207) 40 42.3 18.4 25 40.5 27.3 38.5 33.6 Both Baseline (n=277) 16.7 9.6 6.4 7.9 16.7 7.7 13 9.4 Final (n=207) 8.3 2.8 35.5 29.8 7.1 9.1 7.7 15.5 None Baseline (n=277) 2.1 0 0 0 0 2.6 0 0.4 Final (n=207) 0 0 0 0 0 0 0 0 PREVENTING THE MEDICAL TRANSMISSION OF HIV IN ZAMBIA 18 Patient perception is difficult to change in as short a timeframe as the implementation period. The literature supports the fact that globally, the use of injections has completely overtaken the real need, reaching proportions no longer based on rational medical practice.2 Patients often feel that if they have traveled long distances or have waited a long time to meet with a doctor they should receive an intervention, and they associate receiving an injection as proper treatment despite what might be medically advised (see Table 20). This ultimately puts pressure on providers. Table 20. Patients Who Received an Injection Reporting That Provider Suggested the Treatment Provider Study District (%) Facility (%) Overall (%) Chingola Monze Solwezi RHC UHC Hosp. Mission Hosp. Health care provider suggested injection Baseline (n=133) 100 95.2 100 97.6 100 100 100 98.5 Final (n=207) 31.7 15.5 38.3 32.1 25 54.5 19.2 28.5 At baseline, the proportion of health facilities in which project-developed BCC materials were appropriately displayed was reported at 46.6 percent. Appropriate display means the messages were not displayed in combination with different health messages, enough space was provided between messages, and there was a balance between text and graphic materials for non-literate patients. By 2009, at the time of final evaluation, this proportion had increased to about 86 percent (Table 21). Table 21. Percent of Facilities in Which BCC Materials Were Appropriately Displayed Study District (%) Facility (%) Overall Chingola Monze Solwezi RHC UHC Hospital Mission (%) Hosp. Appropriate display of BCC material Baseline (n=70) 61.5 33.3 53.3 37.8 33.3 76.9 55.6 46.6 Final (n=63) 73.3 100 84.2 90.6 76.9 80 50 85.8 The percentage of patients who reported they had seen or heard about BCC messages related to injection safety increased by almost 5 percentage points from 54.1 percent in 2004 to 59 percent in 2009 as seen below. Table 22. Percent of Patients Reporting That They Had Heard or Seen BCC Messages about Injection Safety Study District (%) Facility (%) Overall (%) Chingola Monze Solwezi RHC UHC Hosp. Mission Hosp. Patients exposed to IP/BCC message Baseline (n=27) 41.3 78.4 38.5 60.1 38.9 31.6 68.2 54.1 Final 65 47.9 64.5 67.5 59.5 27.3 46.2 59 2 Gisselquist, D; Rothenberg, R; Potterat, J; Drucker, E. “HIV Infections in Sub-Saharan Africa Not Explained by Sexual or Vertical Transmission,” International Journal of STD & AIDS, 2002; 13: 657-666. PREVENTING THE MEDICAL TRANSMISSION OF HIV IN ZAMBIA 19 (n=207) Despite a dramatic decrease in the percentage of providers suggesting injections (see Table 20 above) and increases in the numbers of patients reporting that they have been exposed to BCC messages, it is clear that more interventions are needed to curb the demand for unnecessary injections. E. Medical Waste Management The study looked at medical waste disposal in the 63 health facilities and noted that there was a reduction in the incidence of overflowing sharps containers (from 10 percent at baseline to 5.9 percent at final). It was interesting to observe that there was a reduction in the presence of sharps in facilities’ immediate surroundings, from 22 percent at baseline to 0 percent at final evaluation. There was also an improvement in the securing of disposal sites from 22 percent to 35.8 percent at end of evaluation. There was also a marked improvement in the labeling of disposal sites from 1.7 percent at baseline to 34.8 percent at end of evaluation. Table 23. Percent of Health Facilities with Satisfactory Disposal of Used Injection Equipment In terms of sharps waste disposal methods, it was observed that open burning in a hole or enclosure was the primary method used (51.7 percent), followed by incineration (24.7 percent). The collection of waste by a hired contractor was also becoming a common method of waste disposal, increasing from zero to 10.3 percent. Observation Study District (%) Facility (%) Overall (%) Chingola Monze Solwezi RHC UHC Hosp. Missi on Hosp . Sharps in open containers Baseline (n=59) 0 33.3 15.4 20.5 0 0 75 20.3 Final (n=63) 0 0 10 4.4 8.3 0 0 3.3 Overflowing sharps containers Baseline (n=59) 11.1 4.2 15.4 9.1 11.1 50 0 10.2 Final (n=63) 7.7 10 0 2.2 16.7 0 0 5.9 Presence of used sharps in immediate surroundings Baseline (n=58) 44.4 4.2 30.8 15.9 55.6 50 0 22 Final (n=63) 0 0 0 0 0 0 0 0 Secured disposal site Baseline (n=59) 22.2 20.8 26.9 20.5 11.1 100 50 23.7 Final (n=63) 23.1 80 43.3 54.5 25 100 100 35.8 Properly labeled disposal site Baseline (n=59) 11.1 0 0 0 0 50 0 1.7 Final (n=63) 20 61.1 23.3 40.5 16.7 50 100 34.8 PREVENTING THE MEDICAL TRANSMISSION OF HIV IN ZAMBIA 20 Table 24. Primary Methods Used for Sharps Waste Disposal Disposal practice Study District (%) Facility (%) Overall (%) Chingola Monze Solwezi RHC UHC Hosp. Mission Hosp. Open burning on the ground Baseline (n=59) 44.4 45.8 15.4 25.0 55.6 0.0 75.0 32.2 Final (n=63) 0 25 0 12.2 0 0 0 8.3 Open burning in a hole or enclosure Baseline (n=59) 22.2 25.0 42.3 36.4 33.3 0.0 0.0 32.2 Final (n=63) 38.5 50 66.7 58.5 50 0 33.3 51.7 Incineration Baseline (n=59) 55.6 20.8 11.5 13.6 33.3 100.0 50.0 22.0 Final (n=63) 30.8 20 23.3 22 21.4 100 33.3 24.7 Burial Baseline (n=59) 0.0 0.0 7.7 2.3 11.1 0.0 0.0 3.4 Final (n=63) 0 0 0 0 0 0 0 0 Dumping in a pit latrine or other secure pit Baseline (n=59) 11.1 50.0 7.7 25.0 11.1 0.0 75.0 25.4 Final (n=63) 0 5 6.6 4.8 0 0 33.3 3.9 Dumping in an unsupervised area Baseline (n=59) 0.0 8.3 23.1 18.2 0.0 0.0 0.0 13.6 Final (n=63) 0 0 3.3 2.4 0 0 0 1.1 Collected by contractor Baseline (n=59) 0 0 0 0 0 0 0 0 Final (n=63) 30.8 0 0 0 28.6 0 0 10.3 The study has established that sharps disposal is becoming a priority at most health facilities surveyed. However, it was also noted that in some facilities the different types of waste were not segregated, and once outside the facility the waste was thrown into the same shallow pit. Sometimes it would be partially burned or (partially) buried. Information/education is only one factor that can address this problem. Lack of clear norms, lack of implementation follow-up and supervision, lack of recognizable, adequate waste bins for waste segregation, and a lack of space around the facility for proper elimination are but a few barriers that can be ascribed to this situation. Technologies are available for the treatment and disposal of wastes. Their use can be controversial, however, particularly when the type chosen depends more on the economics of the system than on environmental performance. A lack of adequate funding remains a serious issue, particularly for publicly operated hospitals. Frequently the budgets are stretched. As a result, managers naturally look at what offers the best performance for the money available. PREVENTING THE MEDICAL TRANSMISSION OF HIV IN ZAMBIA 21 F. Policy Environment To establish a policy environment that ensured the availability of relevant guidelines and adequate resources for safe injection practice, MISP worked with the Ministry of Health to make IP/IS guidelines available at all facilities and to integrate the policies into their standard operating procedures. Table 25. National IP/IS Guidelines Implemented at the Facility Level Action Plan Study District (%) Facility (%) Overall (%) Chingola Monze Solwezi RHC UHC Hosp. Mission Hosp. IP/IS guidelines available Baseline (n=71) 42.2 50 47.7 46.7 66.7 46.2 44.4 48 Final (n=75) 40 90 94.7 88.2 61.5 40 100 74.9 Health care waste management guidelines available Baseline (n=66) 38.5 26.7 6.7 11.1 16.7 38.5 44.4 20.6 Final (n=75) 60 68.4 33.3 50 57.1 40 0 53.9 Of the supervisors interviewed, 74.9 percent indicted the availability of IP/IS guidelines at their facilities, an increase of 26.9 percent since the baseline study. The availability of health care waste management guidelines has also showed a significant increase, from 20.6 percent at baseline to 53.9 percent at end-line. One of the project’s major challenges, however, is to get IP/IS in the top priorities of the MOH so that it is also budgeted for and activities related to IS are included in the routine work plans and supervisions. G. Monitoring and Evaluation MISP worked with the Ministry of Health at the district and facility levels to establish a system for continuous monitoring and evaluation of injection safety. The study has established that 74.4 percent of the supervisors interviewed indicated that their facilities had an M&E plan in place to monitor the IP/IS activities, compared to 31.5 percent at baseline. Table 26. Supervisors Reporting That Facility Has an M&E Plan to Monitor IP/IS and Other Practices Action Study District (%) Facility (%) Overall (%) Chingola Monze Solwezi RHC UHC Hosp. Mission Hosp. M&E in place to monitor activities Baseline (n=73) 46.2 36.7 20 26.7 66.7 23.1 44.4 31.5 Final (n=65) 80 83.3 60 70.7 78.6 80 50 74.4 PREVENTING THE MEDICAL TRANSMISSION OF HIV IN ZAMBIA 22 A significant number (83 percent) of the supervisors reported that IP/IS activities had been incorporated in their action plans. Table 27. Supervisors Reporting IP/IS Activities in Current Facility Action Plans Action plan Study District (%) Facility (%) Overall Chingol (%) a Monze Solwezi RHC UHC Hosp. Mission Hosp. IP/IS activities in action plan Baseline (n=62) 38.5 70 43.3 53.3 66.7 23.1 88.9 53.4 Final (n=65) 100 90 60 72.1 92.9 100 50 83.3 PREVENTING THE MEDICAL TRANSMISSION OF HIV IN ZAMBIA 23 SECTION III DISCUSSION AND CONCLUSIONS A. Overall Performance Indications There have been considerable successes and challenges for the project as witnessed in the overall performance indicators. The number of injections has increased by about 1 injection per person per year for both males and females, resulting in an average of 2.96 injections per person per year. The averages obtained in the survey were slightly above the national average of less than 1 medical injection per person per year, as suggested in the 2007 Zambia Demographic Health Survey. Differences in sample population and sample size can account for much of the difference in estimates. Other factors include the inability of the project, due to limited resources and time, to provide continuous supportive supervision to each district, resulting in a lack of reinforcement of positive IP/IS messages. In addition, private sector promotion of injections and injection equipment and the proliferation of immunization and vaccination campaigns can send mixed messages about the necessity and effectiveness of injections. Ultimately patient preference plays a large part in the demand for injections, but changing perception and preference will also be one of the hardest and slowest changes to effect. It was encouraging to note that there was remarkable improvement in the percentage of people between 15-49 years of age reporting having been given an injection with a syringe/needle from a new, unopened package, and that the results were consistent across all districts. There was a slight trend, particularly in Solwezi district, of needle reuse, suggesting that rural facilities continue to be a cause for concern and require further attention. The literature suggests that up to 70 percent of rural facilities in some provinces in Zambia are managed by Classified Daily Employees who lack the education and training to implement hygiene best practices and encourage behavior change among patients.3 The project was unable to provide training in all rural health centers and hoped that districts could scale up efforts and that information would trickle down, but many of the facilities remain difficult to access or visit routinely both for the project and for DHMTs. In the area of PEP and hepatitis B vaccinations, there is a general record of improvement between the baseline and the final evaluation. More supervisors reported offering PEP to health personnel at the final evaluation compared to the baseline. Similarly there are more health care providers reporting that the facility offers PEP now compared to the baseline (51.5 percent at baseline compared to 62.4 percent at final evaluation). There seems to be a gradual decrease in hepatitis B vaccinations for health workers but a slight increase for 3 Ministry of Health HRIS Data Base 2004–2005 and National Health Strategic Plan 2006. PREVENTING THE MEDICAL TRANSMISSION OF HIV IN ZAMBIA 24 workers who handle health waste. These data point to mixed achievements. These areas will be improved when the MOH develops and actively promotes policy recognizing PEP and hepatitis B as integral components of the IP/IS policy and provides adequate resources and budget to support implementation. B. Commodity Management and Procurement The case of commodity management procurement has also posted both successes and challenges. Few health facilities are currently experiencing stock-outs of essential IP/IS commodities, which is a significant improvement: Close to half (45 percent) of health facilities reported stock-out of disposable syringes and needles at baseline, the current overall stock-outs of the same equipment is now 9 percent. Furthermore, health facilities have recorded an increase in the quantities of new single-use syringes and needles and the availability of sharp boxes and puncture-proof containers for the period of the project. This is an example of the successes made during the entire project. Another critical aspect of achievement in terms of commodity management procurement is the availability of personal protective equipment for health workers. It is worth noting that there has been a general improvement of availability of personal protective equipment except for examinations gloves, which have seen a decrease from close to 90 percent at baseline to just about 35 percent at final evaluation. C. Capacity Building There have been a number of successes during the project’s lifespan from 2006 to 2009 in the area of capacity building. Notable among these have been increases in the proportion of: health care providers observed washing hands before and after injection (from 26.4 percent to 73.6 percent and from 35.5 percent to 71 percent, respectively); safe disposal of needles (from 76 percent to 99.6 percent); and health care providers reporting needle-stick injuries (from 17.6 percent to 6.6 percent at final evaluation). Also the proportion of health care providers observed leaving needles inserted in the vial for multiple doses dropped from 42.5 percent to 29 percent during the reference period. D. Behavior Change Communication In the area of BCC the project showed improvements in: the proportion of health facilities in which BCC materials were appropriately displayed (from 46.6 percent to 86 percent); and (b) the percentage of patients who reported they had seen or heard about BCC messages related to infection prevention (from 54.1 percent to 59 percent) between 2006 and 2009. There still remain some challenges, including the fact that (a) the proportion of health care providers observed recapping increased from 12.6 percent to about 17 percent at final evaluation, (b) a significant proportion of injection providers (45 percent) and prescribers (45.9 percent) still believe that patients prefer injections when they present with febrile illness, and (c) there was no significant difference in patient preference for injections between 2006 (51.2 percent) and 2009 (50.9 percent). Probably as a consequence of the above, there was an increase in the proportion of injection PREVENTING THE MEDICAL TRANSMISSION OF HIV IN ZAMBIA 25 prescriptions from 39.7 percent in 2004 to more than half (50.5 percent) between 2006 and 2009. On the other hand, oral medication prescriptions dropped by almost 17 percentage points (16.9 percent) between 2006 and 2009. Another worrisome finding is that the incidence of needle-stick injuries increased among waste handlers from 3.9 to 6.6 percent between baseline and final evaluation, although the proportion reporting needle-stick injury at midterm was 1.6 percent among this group. Also, although needle-stick injuries among health care providers reportedly dropped from 17.6 percent to 6.6 percent between 2006 and 2009, this drop is far below desirable levels. The increase in needle-stick injuries and modest decline in the same among health care providers could probably be attributed to the observed decline in the use of personal protective equipment, particularly examination gloves and utility gloves as a result of increased demand and decreased availability. The proportion of health care providers observed using these personal protective equipment declined quite significantly by about 46.1 and 17.1 percentage points, respectively. In focus group discussions, providers and patients alike said it was more risky to receive unneeded shots than to accept oral prescriptions. However, in individual interviews providers and prescribers acknowledged feeling a lot of pressure from clients to receive injections, particularly when a patient had traveled long distances to receive treatment or had waited a long time to be seen by a doctor. E. Conclusions and Recommendations After five years of implementation of the Medical Injection Safety Project (MISP), the evaluators have acknowledged that there are some important conclusions that can be drawn from the data collected. Hand-washing practices before and after injections improved considerably. This single practice is still a big challenge in most important hospitals in the world, including the United States and Europe. The Ministry of Health should continue targeting this simple approach to guarantee continued prevention of infections. Provision of hepatitis B vaccine for health care workers and waste handlers also requires increased attention and commitment from the MOH. As provision has not become standardized policy, individual facilities must prioritize and budget for these vaccines in their annual budgets. National policy would help facilities in making funds available to procure prophylaxis for their staff. Increases in the provision of PPE for health care workers and waste handlers were witnessed across facilities and show an improvement in the commitment of facility supervisors to the protection of their staff. However, it is clear that availability still does not match demand. The MOH should re-invigorate the demand-based procurement system — giving provincial and district offices autonomy in their procurement plans — and push for budgets that allow facilities to provide PEP to their staff. These plans should be prepared using data produced by their health information systems. This practice could introduce important efficiencies in their systems and allow managerial teams to use PREVENTING THE MEDICAL TRANSMISSION OF HIV IN ZAMBIA 26 resources in a more efficient way. In addition, it is important for continue advocating for PPE for waste handlers, as they are often unable to advocate for themselves but are continuously at risk as a result of the nature of their job. It is important to recognize that behavior change is a process that requires long-term investment. Important setbacks occurred between the baseline and the final evaluation as a result of a lack of repeated and continuous supportive supervision to reinforce behaviors. The MOH should discuss the results with the provincial and district offices to identify strategies that can allow DHMTs to address preferences, perceptions, and practices between providers and patients in terms of unneeded injections versus oral treatments. In the planning process of these campaigns, providers from the private sector should be included to analyze different messages used by them in the promotion of some devices and treatments. It would be interesting to do policy research on the effects of immunization and vaccination campaigns on patient perception of efficacy of injections and uptake.4 The area of incentives and recognition should be addressed with providers and policy makers. For those facilities and providers that have reduced the number of unneeded injections, special recognition campaigns could be designed and promoted to highlight best practices. It is clear that the MISP project has contributed significantly in the training and follow-up of different cadres of human resources. But it was also evident that important shortfalls still exist. Many of these are produced because of the high mobility of already-trained staff to more attractive posts. To mitigate this effect, the MOH should use the inventory of trained human resources that MISP shared with provincial and district managers to design systems and put them in place before these movements occurred. It is also recommended that the information collected during the life of the project be shared by the MOH with other donors and agencies, especially those that will be working with retention schemes. That way, cadres already trained in IP/IS could remain working, especially in the hard-hit rural areas. 4 World Health Organization. WHO/EHT/04.04 Safety of Injections. Global Facts and Figures.