NAMPROPA: the Namibia Project for Retention of Patients on ART

Home , Eenhana, Khomasdal, Ohangwena Region, Okuryangava, Otjomuise, Zambezi Region

NAMPROPA: The Namibia Project for Retention of Patients on ART

Results of a Quality Improvement Collaborative in 24 Health Facilities in Namibia

July 2018

Republic of Namibia Ministry of Health and Social Services Table of Contents

EXECUTIVE SUMMARY 1 Background 1 Key Findings 1 Key Recommendations 1

INTRODUCTION 2

BACKGROUND 2 Quality Improvement Collaborative Approach 2 Indicators 3 Site Selection 4 QI Coaching 4

KEY ACTIVITIES 4 Design Meeting 4 Pre-Work 5 Learning Sessions 5 Actions Periods 5

RESULTS 6 Data Collection and Presentation 6 Data Interpretation 6 Loss to Follow-Up 7 Viral Load Monitoring 13 Viral Suppression 19 Hypertension Screening 25 Hypertension Treatment 31 HIV Quality Management Capacity 33 Change Package 34 Knowledge Management 35 Plans for Sustainability 35 Implementation Challenges, Solutions, and Recommendations 36 Plans for Scale-Up 37

CONCLUSION 37

ACKNOWLEDGEMENTS 37

APPENDIX 38 Figure 9. NAMPROPA Driver Diagram 38 Figure 10. NAMPROPA Data Collection Tool 40 Executive Summary

Background

This report summarizes the results of the Namibia Project for Retention of Patients on ART (NAMPROPA), a quality improvement collaborative with the aim of improving retention, viral load monitoring and suppression, and hypertension screening and treatment among people living with HIV in Khomas, Ohangwena, and Zambezi Regions, Namibia. NAMPROPA was led by the Republic of Namibia’s Ministry of Health and Social Services (MoHSS) with technical support from CDC-Namibia and HEALTHQUAL at the University of California, San Francisco (UCSF) and funding through the U.S. President’s Emergency Plan for AIDS Relief (PEPFAR) as part of the Health Resources and Services Administration’s (HRSA) Quality Improvement Capacity for Impact Project (QICIP) award #U1NHA08599. The contents are the responsibility of HEALTHQUAL and do not necessarily reflect the views of the U.S. Government. Key Findings . Average monthly rates of loss to follow-up decreased by 5% from 17% in NAMPROPA’s first quarter (March 2017-May 2017) to 12% in its last quarter (December 2017-February 2018).

. Between the first and last quarter of NAMPROPA, the viral load monitoring rate across participating sites increased by 11% (84% vs. 95%).

. Viral suppression (<1,000 copies/mL) across participating sites increased from 80% in NAMPROPA’s first quarter to 90% in its last quarter.

. Throughout NAMPROPA implementation, an average of 14,044 ART patients per month were screened for hypertension.

. In the final six months of NAMPROPA, 1,508 ART patients were newly diagnosed with hypertension, of which 854 (57%) were initiated on antihypertensive treatment.

. HIV quality management capacity among NAMPROPA sites, as measured by organizational assessment scores, increased dramatically across all domains, with the most significant improvements seen in outcomes monitoring.

Conclusions

Results of NAMPROPA activities indicate improvements in loss to follow-up, viral load monitoring, viral suppression, and hypertension screening and treatment initiation. Moreover, NAMPROPA activities were associated with site-level improvements in HIV quality management capacity, particularly in the areas of data use and outcomes monitoring. As NAMPROPA activities are spread to other sites in CY 2018 and beyond, focused attention will be required to ensure that efficacious interventions are implemented with high fidelity at new sites, and that improvements made in existing sites are sustained.

Page 1 Introduction

Of the estimated 237,126 people living with HIV (PLHIV) in Namibia, approximately 169,568 are active on antiretroviral therapy (ART). Although retention among PLHIV on ART in Namibia is approximately 80%, retention rates according to geography, age and sex are nevertheless highly variable. Moreover, rates of viral suppression—the primary objective of efforts to retain PLHIIV on ART—are likewise highly variable in Namibia, reflecting geographic-, age-, and sex-specific gaps along the cascade of HIV care and treatment. For example, while an estimated 87% of all PLHIV on ART in Namibia were virally suppressed in 2016, this figure was only 71% among PLHIV aged 15-24, suggesting that an expansion in the coverage of HIV services alone is necessary, though not sufficient, to achieve epidemic control.

Ensuring that all PLHIV on ART in Namibia, regardless of geography, age, and sex, are retained in HIV care that results in the achievement of durable viral suppression is critical to achieving the UNAIDS’ 90-90-90 targets and improving the quality of life of PLHIV in Namibia. With funding through the U.S. President’s Emergency Plan for AIDS Relief (PEPFAR) as part of the Health Services and Resources Administration’s (HRSA) Quality Improvement Capacity for Impact Project (QICIP), HEALTHQUAL at the University of California, San Francisco (UCSF) has partnered with the Namibia Ministry of Health and Social Services (MoHSS) to implement the Namibia Project for Retention of Patients on ART (NAMPROPA), a quality improvement collaborative with the aim of improving retention, viral load monitoring, viral suppression, and hypertension screening and treatment among PLHIV on ART in Namibia. Launched in November 2016 and formally concluded in February 2018, NAMPROPA spanned 24 health facilities in PEPFAR-designated scale- up aggressive districts in the regions of Khomas, Ohangwena, and Zambezi. In total, it is estimated that NAMPROPA sites provide HIV care to nearly one-third of all PLHIV on ART in Namibia’s public sector.

This report summarizes the design, implementation, and results of NAMPROPA from its inception in November 2016 to its formal conclusion in February 2018. In particular, it highlights the adaptation of the quality improvement collaborative model to the Namibian context, the incorporation of NAMPROPA activities into existing national quality management infrastructures and frameworks, and the utility of the collaborative methodology in integrating HIV and non-communicable disease (NCD) services. The report concludes with a summary of plans for national scale-up of NAMPROPA activities, and an analysis of resources required for sustainability.

Background Quality Improvement Collaborative Approach The design of the NAMPROPA Quality Improvement Collaborative (QIC) was adapted from the Institute for Healthcare Improvement’s Breakthrough Series (BTS) Model (Figure 1), an improvement science methodology in which participating sites are convened to apply proven quality improvement (QI) methods (e.g., Plan-Do- Study-Act [PDSA] cycles, cascade analysis, process mapping, root cause analysis) to the identification and improvement of gaps in a specific area of healthcare service delivery. Originally piloted in the United States, QICs have been adapted with significant success in low- and middle-income country contexts and clinical settings as varied as HIV/AIDS, family planning and maternal and child health. Unlike traditional QI approaches in which individual projects are implemented and evaluated at individual sites over a period of months, QICs features more frequent performance measurement and smaller-scale tests of interventions, enabling sites to test multiple interventions and share experiences with peers over a compressed period of time. Through the QIC approach, accumulation of evidence-based interventions and performance improvements are therefore significantly accelerated, leading to the compilation of a “package” of proven interventions and a sustainable network of peer learning in which “all [participants] teach, and all learn.” Figure 1 outlines the structure of the NAMPROPA QIC, including key outputs associated with each step of implementation.

Page 2 Figure 1. NAMPROPA QIC Structure

Indicators Indicators were selected by MoHSS and key stakeholders, and reflect definitions outlined in the National Guidelines for Antiretroviral Therapy (2016) and the Namibia Standard Treatment Guidelines (STG) (2011) (Table 1). All Indicators were collected by sites from new or existing data sources and reported on a monthly basis using a pre-programmed Excel spreadsheet (Appendix), and disaggregated by age and sex beginning in September 2017. In response to the high burden of non-communicable diseases (NCD) in Namibia and national efforts to integrate NCD and HIV care, hypertension screening and treatment were included as NAMPROPA indicators.

Table 1. NAMPROPA Indicators Indicator Definition Data Source 1. Loss to follow-up (LTFU) Proportion of patients on ART without a clinic visit ePMS or medication pickup in the last 90 days. 2. Viral load (VL) monitoring Proportion of eligible patients on ART who ePMS, paper registers received a viral load test. 3. Viral suppression Proportion of patients who received a viral load ePMS, paper registers test result indicating a suppressed viral load (<1,000 RNA copies/mL). 4. Hypertension screening Proportion of adult patients (>15 years) on ART Paper registers who were screened for hypertension. 5. Hypertension treatment Proportion of adult patients on ART newly Paper registers diagnosed with hypertension who received antihypertensive treatment.

Page 3 Site Selection Twenty four (24) facilities in Khomas, Ohangwena, and Zambezi Regions participated in NAMPROPA (Table 2). Of these, 6 were tertiary healthcare facilities, 6 were secondary healthcare facilities, and 12 were primary healthcare facilities. All 24 sites had at least one registered nurse in residence, yet only 8 had an assigned physician. All participating facilities were located in PEPFAR-designated scale-up aggressive regions, defined as geographic areas with high burdens of HIV infection and significant gaps in ART coverage. In total, it is estimated that NAMPROPA facilities provide ART services to approximately 55,000 PLHIV.

Table 2. NAMPROPA Participating Sites Region Facility Name and Patient Volume Khomas Katutura Health Center*** Otjomuise Clinic* Katutura Intermediate Hospital*** Robert Mugabe Clinic** Khomasdal Health Center** Windhoek Central Hospital** Okuryangava Clinic** Ohangwena Eenhana Clinic** Ongenga Clinic* Ekoka Clinic* Ongha Health Center** Engela Hospital*** Oshaango Clinic* Odibo Health Center** Oshandi Clinic* Okongo Clinic* Oshikunde Clinic* Zambezi Bukalo Health Center* Ngweze Clinic* Katima Mulilo Clinic* Sesheke Clinic* Katima Mulilo Hospital** Sibbinda Health Center* Mavuluma Clinic* Legend: *0-999 active ART patients | **1,000-4,999 active ART patients | ***>5,000 active ART patients

QI Coaching Participating sites received QI coaching and support from regional clinical and district nurse mentors, an existing cadre of physicians and nurses employed by the MoHSS Directorate of Special Programs (DSP) with special expertise in HIV clinical care. As part of Collaborative activities, regional mentors visited sites at a minimum of a monthly basis to review sites’ performance, assist with the design, execution, and evaluation of PDSA cycles, and communicate Collaborative updates. Technical support for regional mentors in QI methods and coaching was provided by the MoHSS QM Team, IHI, and HEALTHQUAL, and monthly check-in calls between HEALTHQUAL and MoHSS were convened to monitor implementation and discuss challenges.

Key Activities Design Meeting

The Collaborative’s Design Meeting was convened in Windhoek, Namibia, from November 21-25, 2016, by MoHSS and key stakeholders from CDC-Namibia, HEALTHQUAL, IntraHealth International Namibia, I-TECH Namibia, Project Hope, and Development Aid from People to People (DAPP)/Total Control of the Epidemic (TCE). During the Meeting, attendees discussed the proposed scope, geographic focus, and duration of the Collaborative, and drafted an aim statement to guide Collaborative implementation: “To improve retention and attain viral suppression in Khomas, Zambezi, and Ohangwena Regions to reach the 90-90-90 goals for Namibia by 31 August 2018.” In addition, Design Meeting attendees drafted the Collaborative’s terms of reference and driver diagram, and selected five indicators to track Collaborative progress on a monthly basis. Furthermore, Lastly, Design Meeting attendees reviewed existing QI coaching capacity in Khomas, Ohangwena, and Zambezi Regions and discussed an approach to forming QI coaching teams.

Page 4 Pre-Work Period Figure 2. NAMPROPA storyboards The pre-work period began following the conclusion of the Collaborative’s Design Meeting in November 2016. During this period, regional coaching teams were identified, the Collaborative terms of reference were drafted, and initial site visits were completed. A total of 138 healthcare workers (36 in Khomas, 64 in Ohangwena, and 38 in Zambezi) comprised of physicians, nurses, pharmacists, data clerks, health assistants and field officers were reached during site visits. In addition, participating sites identified Collaborative focal persons, collected baseline performance data, and began construction of storyboards (Figure 2) in preparation for the first learning session. Learning Session 1 The first learning session (LS) was held February 27-March 1, 2017, in Ondangwa, Namibia, bringing together 107 participants from Collaborative sites, and stakeholders from MoHSS, CDC-Namibia, IntraHealth, I-TECH, Project HOPE, DAPP, IHI, and HEALTHQUAL. During the learning session, attendees received training on Collaborative methodology and QI methods such as the Model for Improvement, run chart analysis, data collection, Collaborative indicators, process mapping, change ideas, and Plan-Do-Study-Act (PDSA) cycle planning and evaluation. In addition, teams from participating sites presented storyboards (Figure 2) that summarized their baseline performance data and described current QI activities. Follow-up discussions and peer exchange were facilitated by MoHSS, HEALTHQUAL and IHI advisors. Sites were encouraged to share successful interventions and approaches. The meeting concluded with revision of the Collaborative’s driver diagram (see Appendix), dissemination of data collection tools and teams’ development of implementation plans for the first action period.

Action Period 1 Activities for Action Period 1 began at the conclusion of the first LS. During this period, sites conducted rapid tests of change to improve their existing systems and processes. Data were collected, analyzed and reported on a monthly basis to track the results of their interventions. Clinical and nurse mentors conducted monthly site visits, and provided ongoing support to sites in their planning and execution of PDSA cycles. In addition, the MoHSS QM team visited sites on a quarterly basis to provide overall technical support and the Collaborative management team conducted weekly virtual meetings to monitor implementation progress. Finally, sites prepared presentations of their activities and results for LS. Learning Session 2 The second LS was held July 4-6, 2017, in Otjiwarongo, Namibia. A total of 102 participants from Collaborative sites and MoHSS stakeholders (i.e., CDC-Namibia, HEALTHQUAL, IHI, IntraHealth International, I-TECH Namibia, Project HOPE, and DAPP/TCE) were in attendance. The aim of the second LS was to create a common learning platform for successful implementation of NAMPROPA, while specific objectives were to review progress of NAMPROPA through storyboard presentations, assess and benchmark results, and develop work plans for the second action period. During the LS, attendees received refresher training on Collaborative measures, the Model for Improvement, and PDSA implementation. In addition, Dr. Michael Mugavero, Professor of Medicine and Co-Director of the Center for AIDS Research, University of Alabama, Birmingham, delivered a virtual presentation on retention measurement and reengagement strategies, underscoring the importance of missed visits as an actionable item associated with mortality and evidence-based interventions involving peer support. Collaborative-wide data were presented by HEALTHQUAL, with visual display of regional trends for each indicator. Finally, teams from participating sites delivered presentations on their QI activities and results from the first action period, with follow-up discussions facilitated by regional nurse and clinical mentors. The meeting concluded with teams’ development of implementation plans for the second action period. Page 5 Action Period 2 Activities for Action Period 2 began at the conclusion of the second LS. During this period, sites conducted rapid tests of change to improve their existing systems and processes. Data were collected, analyzed and reported on a monthly basis to track the results of their interventions. QI coaching of participating sites continued with ongoing support from the MoHSS QM team. In addition to Collaborative-wide LS, and specifically in preparation for 3rd LS, regional teams used innovative ideas such as soliciting sponsorships for food and transportation to convene facility teams to share best practices and discuss common implementation challenges. Between July 2017 and February 2018, seven regional exchange meetings (two in Khomas, three in Ohangwena, and two in Zambezi) were held with facilitation by regional clinical and nurse mentors. Finally, sites prepared presentations of their activities and results for LS. Learning Session 3 The third LS was held February 6-8, 2018, in Otjiwarongo, with 96 participants from Collaborative sites and stakeholders from MoHSS and HEALTHQUAL. Attendees presented interventions that were tested and subsequently adopted, and discussed their relative importance, ease of implementation, and potential scalability. Small groups were convened to prioritize and rank changes in each region. The exercise involved detailed discussion and comparison of changes, leading to agreement about the best PDSA cycles which were implemented and yielded significant improvement, as well as the scale of regional implementation. These discussions culminated in the construction of a draft change package of interventions for use in scale up and spread of NAMPROPA activities to non-participating sites. To develop sustainability plans, participants were divided into groups by region to discuss proposed activities, with facility teams, while QI coaches, district and regional managers convened separately. The LS concluded with presentations of regions’ strategies for sustaining NAMPROPA activities following its formal conclusion in February 2018. These strategies are summarized in table 5. Results Data Collection and Presentation This section summarizes performance measurement reported by NAMPROPA sites between March 1, 2017, and February 28, 2018. It also reports one quarter of data following the formal conclusion of NAMPROPA activities. Performance measurement data for the five NAMPROPA indicators were submitted by sites to regional clinical and nurse mentors on a monthly basis. After submission, these data were reviewed by mentors and submitted to MoHSS QM team for further analysis and aggregation. Following analyses of trends, MoHSS staff liaised with regional mentors to identify focus areas for subsequent PDSAs, and target sites with low performance and/or data quality issues for immediate follow-up. In this report, NAMPROPA- and region-wide performance rates for all measures are calculated according to definitions detailed in Table 1, and are visualized month-to-month alongside their corresponding denominator, and site-level performance data are visualized as small multiples. Unless otherwise specified, all analyses reflect comparisons of distinct, cross- sectional panels of patients, and therefore do not intend to imply longitudinal improvements among a single cohort. Moreover, it should be noted that all data are self-reported by participating sites and have not undergone external evaluation. Finally, unless otherwise specified, all comparative analyses are reported for descriptive purposes only, and are not intended to imply statistical significance. Data Interpretation Small multiples of run charts are presented to display sites’ month-to-month performance. However, these run charts do not feature annotations of contingent factors that may have affected monthly performance, such as medication stock-outs, significant staff-turnover, and malfunction of blood pressure monitors. These factors are documented in site-level Collaborative databases and are available for review upon request. As part of pre- work activities prior to the launch of Collaborative activities in March 2017, sites collected baseline data on the five Collaborative indicators through queries of ePMS and EDT. Owing to significant data quality issues associated with these registers, however, these data are not reported. Since no Collaborative sites were routinely collecting data to measure rates of hypertension screening and treatment at the commencement of NAMPROPA activities, baseline performance on these indicators is assumed to be zero.

Page 6 Loss to Follow Up In the first quarter of NAMPROPA implementation, participating sites reported a monthly average of 46,785 active patients on ART (Figure 3a). Of these patients, a monthly average of 7,833 (17%) were reported by sites as lost to follow-up (LTFU). By the final quarter of implementation, this figure had decreased to 6,596 (12%), representing a 5% decrease in Collaborative-wide LTFU rate over 12 months. In the three months following the formal conclusion of NAMPROPA activities in February 2018, improvements were maintained. Rates of LTFU were highly variable by site and region (Figure 3b-g) In particular, Engela and Eenhana—two high-volume sites in Ohangwena Region—reported high rates of LTFU throughout NAMPROPA implementation. As these figures are the likely artifacts of an incomplete transfer/de-duplication of patient records in ePMS at the time of HIV treatment decentralization, they should be interpreted with caution.

Figure 3a. Number of Active Patients on ART and Loss to Follow Up Rate—NAMPROPA, March 2017-May 2018

60000 56167 56409 56716 57051 56976 25% 54581 54990 55522 56032 55248 51972 52580 48666 50000 20% 46368 20% 17% 39718 40000 14% 14% 15% 12% 13% 11% 12% 12% 12% 30000 11% 11% 11% 11% 10% 10% 20000 Loss to Follow Up Rate Up Followto Loss 5% 10000 Number of Active Patients on ART on Patients Active of Number

0 0% Jul-17 Oct-17 Apr-18 Apr-17 Jun-17 Jan-18 Mar-18 Mar-17 Feb-18 Nov-17 Dec-17 Aug-17 Sep-17 May-18 May-17

Figure 3b. Number of Active Patients on ART and Loss to Follow Up Rate—Khomas Region, March 2017-May 2018

30000 10%

9% 24393 24541 24406 24429 24640 24758 24781 24951 25000 24011 23789 22618 22798 8% 21872 7% 7% 7% 20000 16184 6% 15000 5% 12194 4% 10000 3% 3% 3% 3% 2% 2% 2%

2% Rate Up Followto Loss 2% 2% 1% 1% 2% 5000 1% 1% Number of Active Patients on ART on Patients Active of Number 1%

0 0% Jul-17 Oct-17 Apr-18 Apr-17 Jun-17 Jan-18 Mar-18 Mar-17 Feb-18 Nov-17 Dec-17 Aug-17 Sep-17 May-18 May-17

Page 7 Loss to Follow Up (Continued)

Figure 3c. Number of Active Patients on ART and Loss to Follow Up Rate—Ohangwena Region, March 2017- May 2018

25000 40%

19734 20951 21040 20953 21131 20975 21267 21072 21278 35% 20346 20403 20790 19799 19888 20065 20000 34% 28% 29% 31% 31% 27% 30% 26% 26% 26% 26% 25% 25% 24% 24% 24% 25% 15000

20%

10000 15% Loss to Follow Up Rate Up Followto Loss 10%

Number of Active Patients on ART on Patients Active of Number 5000 5%

0 0% Jul-17 Oct-17 Apr-18 Apr-17 Jun-17 Jan-18 Mar-18 Mar-17 Feb-18 Nov-17 Dec-17 Aug-17 Sep-17 May-18 May-17

Figure 3d. Number of Active Patients on ART and Loss to Follow Up Rate—Zambezi Region, March 2017-May 2018

12000 10% 9% 11101 11123 10808 10849 11026 10451 10167 10250 10339 9% 9717 9838 10000 9466 9019 8% 7%

7725 7% 8000 6% 7060 6% 6% 5% 5% 5% 6% 5% 5% 5% 6000 4% 4% 5% 5% 4% 4% 4000

3% Rate Up Followto Loss

Number of Active Patients on ART on Patients Active of Number 2% 2000 1%

0 0% Jul-17 Oct-17 Apr-18 Apr-17 Jun-17 Jan-18 Mar-18 Mar-17 Feb-18 Nov-17 Dec-17 Aug-17 Sep-17 May-18 May-17

Page 8 Loss to Follow Up (Continued) Figure 3e. Loss to Follow Up Rate—By Site, Khomas Region, March 2017-May 2018

Katutura Health Center Katutura Intermediate Hospital

60% 60% 50% 50% 40% 40% 30% 30% 20% 20% 10% 10% 0% 0% Loss to Follow Up Rate Up Followto Loss Loss to Follow Up Rate Up Followto Loss Jul-17 Jul-17 Apr-18 Oct-17 Apr-17 Apr-18 Oct-17 Apr-17 Jun-17 Jan-18 Jun-17 Jan-18 Mar-18 Mar-17 Feb-18 Mar-18 Mar-17 Feb-18 Nov-17 Dec-17 Aug-17 Sep-17 Nov-17 Dec-17 Aug-17 Sep-17 May-18 May-17 May-18 May-17

Khomasdal Clinic Okuryangava Clinic

60% 60% 50% 50% 40% 40% 30% 30% 20% 20% 10% 10% 0% 0% Loss to Follow Up Rate Up Followto Loss Rate Up Followto Loss Jul-17 Jul-17 Oct-17 Apr-18 Oct-17 Apr-18 Apr-17 Apr-17 Jan-18 Jan-18 Jun-17 Jun-17 Mar-18 Mar-18 Mar-17 Mar-17 Feb-18 Feb-18 Nov-17 Dec-17 Nov-17 Dec-17 Aug-17 Sep-17 Aug-17 Sep-17 May-18 May-18 May-17 May-17

Otjomuise Clinic Robert Mugabe Clinic

60% 60% 50% 50% 40% 40% 30% 30% 20% 20% 10% 10% 0% 0% Loss to Follow Up Rate Up Followto Loss Loss to Follow Up Rate Up Followto Loss Jul-17 Oct-17 Apr-18 Apr-17 Jul-17 Jun-17 Jan-18 Mar-18 Mar-17 Feb-18 Nov-17 Dec-17 Aug-17 Sep-17 Oct-17 Apr-18 Apr-17 May-18 May-17 Jun-17 Jan-18 Mar-18 Mar-17 Feb-18 Nov-17 Dec-17 Aug-17 Sep-17 May-18 May-17

Windhoek Central Hospital

60% 50% 40% 30% 20% 10% 0% Loss to Follow Up Rate Up Followto Loss Jul-17 Oct-17 Apr-18 Apr-17 Jun-17 Jan-18 Mar-18 Mar-17 Feb-18 Nov-17 Dec-17 Aug-17 Sep-17 May-18 May-17 Page 9 Loss to Follow Up (Continued) Figure 3f. Loss to Follow Up Rate—By Site, Ohangwena Region, March 2017-May 2018

Eenhana Clinic Ekoka Clinic 65% 60% 55% 50% 45% 40% 35% 30% 25% 20% 15% 10% 5% 0% Loss to Follow Up Rate Up Followto Loss Loss to Follow Up Rate Up Followto Loss -5% Jul-17 Jul-17 Apr-18 Oct-17 Apr-17 Jun-17 Jan-18 Mar-18 Mar-17 Feb-18 Nov-17 Dec-17 Aug-17 Sep-17 Apr-18 Oct-17 Apr-17 Jun-17 Jan-18 May-18 May-17 Mar-18 Mar-17 Feb-18 Nov-17 Dec-17 Aug-17 Sep-17 May-18 May-17

Engela Hospital Odibo Health Center

60% 60% 50% 50% 40% 40% 30% 30% 20% 20% 10% 10% 0% 0% Loss to Follow Up Rate Up Followto Loss Loss to Follow Up Rate Up Followto Loss Jul-17 Oct-17 Apr-18 Apr-17 Jan-18 Jun-17 Jul-17 Mar-18 Mar-17 Feb-18 Nov-17 Dec-17 Aug-17 Sep-17 May-18 May-17 Oct-17 Apr-18 Apr-17 Jun-17 Jan-18 Mar-18 Mar-17 Feb-18 Nov-17 Dec-17 Aug-17 Sep-17 May-18 May-17

Okongo Clinic Ongenga Clinic

60% 60% 50% 50% 40% 40% 30% 30% 20% 20% 10% 10% 0% 0% Loss to Follow Up Rate Up Followto Loss Loss to Follow Up Rate Up Followto Loss Jul-17 Jul-17 Oct-17 Apr-18 Apr-17 Jun-17 Jan-18 Mar-18 Mar-17 Feb-18 Oct-17 Apr-18 Apr-17 Nov-17 Dec-17 Aug-17 Sep-17 Jun-17 Jan-18 May-18 May-17 Mar-18 Mar-17 Feb-18 Nov-17 Dec-17 Aug-17 Sep-17 May-18 May-17

Ongha Clinic Oshaango Clinic

Page 10 Loss to Follow Up (Continued) Figure 3f. Loss to Follow Up Rate—By Site, Ohangwena Region, March 2017-May 2018 (Continued)

Oshaandi Clinic Oshikunde Clinic

60% 60% 50% 50% 40% 40% 30% 30% 20% 20% 10% 10% 0% 0% Loss to Follow Up Rate Up Followto Loss Loss to Follow Up Rate Up Followto Loss Jul-17 Jul-17 Apr-18 Oct-17 Apr-17 Jun-17 Jan-18 Apr-18 Oct-17 Apr-17 Jun-17 Jan-18 Mar-18 Mar-17 Feb-18 Nov-17 Dec-17 Aug-17 Sep-17 Mar-18 Mar-17 Feb-18 Nov-17 Dec-17 Aug-17 Sep-17 May-18 May-17 May-18 May-17

Page 11 Loss to Follow Up (Continued) Figure 3g. Loss to Follow Up Rate—By Site, Zambezi Region, March 2017-May 2018

Bukalo Health Center Katima Mulilo Clinic

Katima Mulilo Hospital Mavuluma Clinic

60% 60% 50% 50% 40% 40% 30% 30% 20% 20% 10% 10% 0% 0% Loss to Follow Up Rate Up Followto Loss Rate Up Followto Loss Jul-17 Jul-17 Oct-17 Apr-18 Oct-17 Apr-18 Apr-17 Apr-17 Jun-17 Jan-18 Jun-17 Jan-18 Mar-18 Mar-18 Mar-17 Mar-17 Feb-18 Feb-18 Nov-17 Dec-17 Nov-17 Dec-17 Aug-17 Sep-17 Aug-17 Sep-17 May-18 May-18 May-17 May-17

Ngweze Clinic Sesheke Clinic

Sibbinda Clinic

60% 50% 40% 30% 20% 10% 0% Loss to Follow Up Rate Up Followto Loss Jul-17 Apr-18 Oct-17 Apr-17 Jun-17 Jan-18 Mar-18 Mar-17 Feb-18 Nov-17 Dec-17 Aug-17 Sep-17 May-18 May-17 Page 12 Viral Load Monitoring Between March 2017 and May 2017, 12,678 ART patients were eligible for viral load monitoring (Figure 4a). Of these, 10,642 received a viral load (VL) test, corresponding to a VL monitoring rate of 84% in NAMPROPA’s first quarter. In the final quarter of NAMPROPA, 12,564 (95%) of all eligible patients were successfully monitored—an 11% increase compared to the first quarter. Data collected following the formal end of NAMPROPA activities indicated that improvements were sustained. Site-level and regional variability was generally modest (Figure 4b-g), with the majority of sites consistently reporting VL monitoring rates above 85%.

Figure 4a. Number of Active Patients on ART Eligible for a VL Test and VL Monitoring Rate—NAMPROPA, March 2017-May 2018

7000 95% 97% 95% 97% 95% 98% 97% 97% 96% 98% 100% 92% 90% 86% 84% 90% 6000 5765 76% 5407 5491 5428 5312 5219 80% 5016 5105 4839 4891 5000 4490 70% 4263 4089 3925 60% 4000 3270 50% 3000 40%

2000 30% Viral Load Monitoring Rate Monitoring Load Viral Number Number of Patients Eligible 20% 1000 10%

0 0% Jul-17 Oct-17 Apr-18 Apr-17 Jun-17 Jan-18 Mar-18 Mar-17 Feb-18 Nov-17 Dec-17 Aug-17 Sep-17 May-18 May-17

Figure 4b. Number of Active Patients on ART Eligible for a VL Test and VL Monitoring Rate—Khomas Region, March 2017-May 2018

97% 96% 97% 96% 3500 93% 94% 93% 100% 90% 92% 90% 85% 83% 2957 90% 3000 2769 2724 2693 72% 2639 73% 80% 2500 70% 2199 2253 2112 2097 68% 1962 60% 2000 1669 1534 1546 1576 50% 1500 1178 40%

1000 30% Number Number of Patients Eligible

20% Viral Load Rate Monitoring 500 10%

0 0%

Page 13 Viral Load Monitoring (Continued)

Figure 4c. Number of Active Patients on ART Eligible for a VL Test and VL Monitoring Rate—Ohangwena Region, March 2017-May 2018

98% 99% 98% 99% 100% 100% 100% 100% 100% 100% 100% 100% 100% 2500 94% 100% 2305 90% 2026 2055 1975 1956 2000 80% 1820 1852 1857 1836 1863 1841 1786 1763 66% 1607 70%

1500 60%

50% 1075 1000 40%

30% Viral Load Monitoring Rate Monitoring Load Viral Number Number of Patients Eligible 500 20%

10%

0 0% Jul-17 Oct-17 Apr-18 Apr-17 Jun-17 Jan-18 Mar-18 Mar-17 Feb-18 Nov-17 Dec-17 Aug-17 Sep-17 May-18 May-17

Figure 4d. Number of Active Patients on ART Eligible for a VL Test and VL Monitoring Rate—Zambezi Region, March 2017-May 2018

98% 97% 98% 97% 97% 98% 98% 98% 97% 98% 1400 94% 95% 94% 95% 100% 90% 1197 90% 1200 80% 1004 1017 967 1000 70% 906 894 911 852 852 829 772 774 60% 800 698 651 661 50% 600 40%

400 30% Viral Load Monitoring Rate Monitoring Load Viral Number Number of Patients Eligible 20% 200 10%

0 0% Jul-17 Oct-17 Apr-18 Apr-17 Jun-17 Jan-18 Mar-18 Mar-17 Feb-18 Nov-17 Dec-17 Aug-17 Sep-17 May-18 May-17

Page 14 Viral Load Monitoring (Continued) Figure 4e. VL Monitoring Rate—By Site, Khomas Region, March 2017-May 2018

Katutura Health Center Katutura Intermediate Hospital 100% 100%

80% 80%

60% 60%

40% 40%

20% 20%

0% 0% Viral Load Monitoring Rate Monitoring Load Viral Viral Load Monitoring Rate Monitoring Load Viral Jul-17 Jul-17 Apr-18 Oct-17 Apr-17 Apr-18 Oct-17 Apr-17 Jun-17 Jan-18 Jun-17 Jan-18 Mar-18 Mar-17 Feb-18 Mar-18 Mar-17 Feb-18 Nov-17 Dec-17 Aug-17 Sep-17 Nov-17 Dec-17 Aug-17 Sep-17 May-18 May-17 May-18 May-17

Khomasdal Clinic Okuryangava Clinic 100% 100%

80% 80%

60% 60%

40% 40%

20% 20%

0% 0% Viral Load Monitoring Rate Monitoring Load Viral Rate Monitoring Load Viral Jul-17 Jul-17 Oct-17 Apr-18 Oct-17 Apr-18 Apr-17 Apr-17 Jan-18 Jan-18 Jun-17 Jun-17 Mar-18 Mar-18 Mar-17 Mar-17 Feb-18 Feb-18 Nov-17 Dec-17 Nov-17 Dec-17 Aug-17 Sep-17 Aug-17 Sep-17 May-18 May-18 May-17 May-17

Otjomuise Clinic Robert Mugabe Clinic 100% 100%

80% 80%

60% 60%

40% 40%

20% 20%

0% 0% Viral Load Monitoring Rate Monitoring Load Viral Viral Load Monitoring Rate Monitoring Load Viral Jul-17 Oct-17 Apr-18 Apr-17 Jul-17 Jun-17 Jan-18 Mar-18 Mar-17 Feb-18 Nov-17 Dec-17 Aug-17 Sep-17 Oct-17 Apr-18 Apr-17 May-18 May-17 Jun-17 Jan-18 Mar-18 Mar-17 Feb-18 Nov-17 Dec-17 Aug-17 Sep-17 May-18 May-17

Windhoek Central Hospital 100%

80%

60%

40%

20%

0% Viral Load Monitoring Rate Monitoring Load Viral Jul-17 Oct-17 Apr-18 Apr-17 Jun-17 Jan-18 Mar-18 Mar-17 Feb-18 Nov-17 Dec-17 Aug-17 Sep-17 May-18 May-17 Page 15 Viral Load Monitoring (Continued) Figure 4f. Viral Load Monitoring Rate—By Site, Ohangwena Region, March 2017-May 2018 Eenhana Clinic Ekoka Clinic 100% 100%

80% 80%

60% 60%

40% 40%

20% 20%

Engela Hospital Odibo Health Center 100% 100%

80% 80%

60% 60%

40% 40%

20% 20%

0% 0% Viral Load Monitoring Rate Monitoring Load Viral Rate Monitoring Load Viral Jul-17 Jul-17 Oct-17 Apr-18 Oct-17 Apr-18 Apr-17 Apr-17 Jun-17 Jan-18 Jun-17 Jan-18 Mar-18 Mar-18 Mar-17 Mar-17 Feb-18 Feb-18 Nov-17 Dec-17 Nov-17 Dec-17 Aug-17 Sep-17 Aug-17 Sep-17 May-18 May-18 May-17 May-17

Okongo Clinic Ongenga Clinic 100% 100%

80% 80%

60% 60%

40% 40%

20% 20%

0% 0% Viral Load Monitoring Rate Monitoring Load Viral Viral Load Monitoring Rate Monitoring Load Viral Jul-17 Oct-17 Apr-18 Apr-17 Jun-17 Jan-18 Jul-17 Mar-18 Mar-17 Feb-18 Nov-17 Dec-17 Aug-17 Sep-17 May-18 May-17 Oct-17 Apr-18 Apr-17 Jun-17 Jan-18 Mar-18 Mar-17 Feb-18 Nov-17 Dec-17 Aug-17 Sep-17 May-18 May-17

Ongha Clinic Oshaango Clinic 100% 100%

80% 80%

60% 60%

40% 40%

20% 20%

0% 0% Viral Load Monitoring Rate Monitoring Load Viral Viral Load Monitoring Rate Monitoring Load Viral Jul-17 Jul-17 Oct-17 Apr-18 Apr-17 Oct-17 Apr-18 Apr-17 Jun-17 Jan-18 Jun-17 Jan-18 Mar-18 Mar-17 Feb-18 Mar-18 Mar-17 Feb-18 Nov-17 Dec-17 Aug-17 Sep-17 Nov-17 Dec-17 Aug-17 Sep-17 May-18 May-17 May-18 May-17

Page 16 Viral Load Monitoring (Continued) Figure 4f. Viral Load Monitoring Rate—By Site, Ohangwena Region, March 2017-May 2018 (Continued) Oshikunde Clinic Oshaandi Clinic 100% 100%

80% 80%

60% 60%

40% 40%

20% 20%

0% 0% Viral Load Monitoring Rate Monitoring Load Viral Viral Load Monitoring Rate Monitoring Load Viral Jul-17 Jul-17 Apr-18 Oct-17 Apr-17 Jun-17 Jan-18 Mar-18 Mar-17 Feb-18 Nov-17 Dec-17 Aug-17 Sep-17 Apr-18 Oct-17 Apr-17 Jun-17 Jan-18 May-18 May-17 Mar-18 Mar-17 Feb-18 Nov-17 Dec-17 Aug-17 Sep-17 May-18 May-17

Page 17 Viral Load Monitoring (Continued) Figure 4g. Viral Load Monitoring Rate—By Site, Zambezi Region, March 2017-May 2018

Bukalo Health Center Katima Mulilo Clinic 100% 100%

80% 80%

60% 60%

40% 40%

20% 20%

0% 0% Viral Load Monitoring Rate Monitoring Load Viral Rate Monitoring Load Viral Jul-17 Jul-17 Apr-18 Apr-18 Oct-17 Oct-17 Apr-17 Apr-17 Jun-17 Jan-18 Jun-17 Jan-18 Mar-18 Mar-18 Mar-17 Mar-17 Feb-18 Feb-18 Nov-17 Dec-17 Nov-17 Dec-17 Aug-17 Sep-17 Aug-17 Sep-17 May-18 May-18 May-17 May-17

Katima Mulilo Hospital Mavuluma Clinic 100% 100%

80% 80%

60% 60%

40% 40%

20% 20%

Ngweze Clinic Sesheke Clinic 100% 100%

80% 80%

60% 60%

40% 40%

20% 20%

Sibbinda Clinic 100%

80%

60%

40%

20%

0% Viral Load Monitoring Rate Monitoring Load Viral Jul-17 Apr-18 Oct-17 Apr-17 Jun-17 Jan-18 Mar-18 Mar-17 Feb-18 Nov-17 Dec-17 Aug-17 Sep-17 May-18 May-17

Page 18 Viral Suppression Of 12,626 viral load test results returned to NAMPROPA sites between March 2017 and May 2017, 10,114 (80%) indicated a VL <1,000 copies/mL (Figure 5a). In the final quarter of NAMPROPA implementation, this figure had increased to 90%, corresponding to a 10% improvement in viral suppression rate across all participating sites. Rates of viral suppression were highly variable across regions and sites, particularly among those with high volumes of adolescent patients (Figure 5b-g). Rates remained high following the formal conclusion of activities in February 2018.

Figure 5a. Number of Active Patients on ART Who Received a VL Test Result and Viral Suppression Rate— NAMPROPA, March 2017-May 2018

6000 92% 100% 90% 90% 91% 90% 88% 87% 87% 90% 87% 84% 90% 81% 82% 79% 80% 5079 4805 5000 4652 4539 4597 4534 80% 4514 4352 4374 4160 4106 3952 4040 70% 4000 3810 60% 2867 3000 50%

40% 2000 30% Viral Suppression Rate Suppression Viral 20% 1000 Number of Results Test of Load Number Viral 10%

0 0% Jul-17 Oct-17 Apr-18 Apr-17 Jun-17 Jan-18 Mar-18 Mar-17 Feb-18 Nov-17 Dec-17 Aug-17 Sep-17 May-18 May-17

Figure 5b. Number of Active Patients on ART Who Received a VL Test Result and Viral Suppression Rate— Khomas Region, March 2017-May 2018

2500 94% 96% 95% 96% 95% 95% 97% 96% 97% 95% 100% 91% 86% 85% 84% 2179 90% 2034 79% 2003 1936 2000 1873 1925 1913 1903 1924 80% 1747 1613 1587 70% 1466 1500 60% 1210 50%

1000 40% 809 30% Viral Suppression Rate Suppression Viral 500 20%

Number of Results Test of Load Number Viral 10%

0 0% Jul-17 Oct-17 Apr-18 Apr-17 Jun-17 Jan-18 Mar-18 Mar-17 Feb-18 Nov-17 Dec-17 Aug-17 Sep-17 May-18 May-17

Page 19 Viral Suppression (Continued)

Figure 5c. Number of Active Patients on ART Who Received a VL Test Result and Viral Suppression Rate— Ohangwena Region, March 2017-May 2018

2500 100% 90% 90% 91% 87% 87% 86% 87% 88% 88% 83% 84% 85% 85% 90% 77% 2000 1898 1890 1936 80% 1987 1820 1762 1735 68% 1742 1679 1682 1621 1638 1659 70% 1555 1500 60%

50% 1050 1000 40%

30% Viral Suppression Rate Suppression Viral

Number of Results Test of Load Number Viral 500 20%

10%

0 0% Jul-17 Oct-17 Apr-18 Apr-17 Jun-17 Jan-18 Mar-18 Mar-17 Feb-18 Nov-17 Dec-17 Aug-17 Sep-17 May-18 May-17

Figure 5d. Number of Active Patients on ART Who Received a VL Test Result and Viral Suppression Rate— Zambezi Region, March 2017-May 2018

100% 89% 1200 113881% 90% 83% 83% 83% 83% 80% 79% 80% 76% 1025 1008 74% 75% 74% 80% 1000 71% 933 68% 896 875 70% 789 804 800 755 829 768 738 716 60% 666 665 50% 600 40%

400 30% Rate Suppression Viral

20% Number of Viral Results Test Load Viral of Number 200 10%

0 0% Jul-17 Oct-17 Apr-18 Apr-17 Jun-17 Jan-18 Mar-18 Mar-17 Feb-18 Nov-17 Dec-17 Aug-17 Sep-17 May-18 May-17

Page 20 Viral Suppression (Continued) Figure 5e. Viral Suppression Rate—By Site, Khomas Region, March 2017-May 2018 Katutura Intermediate Hospital Katutura Health Center 100% 100% 80% 80% 60% 60% 40% 40% 20% 20% 0%

0% Rate Suppression Viral Viral Suppression Rate Suppression Viral Jul-17 Apr-18 Oct-17 Apr-17 Jun-17 Jan-18 Mar-18 Mar-17 Feb-18 Nov-17 Dec-17 Aug-17 Sep-17 Jul-17 May-18 May-17 Apr-18 Oct-17 Apr-17 Jun-17 Jan-18 Mar-18 Mar-17 Feb-18 Nov-17 Dec-17 Aug-17 Sep-17 May-18 May-17

Khomasdal Clinic Okuryangava Clinic 100% 100%

80% 80%

60% 60%

40% 40%

20% 20%

0% 0% Viral Suppression Rate Suppression Viral Rate Suppression Viral Jul-17 Jul-17 Oct-17 Apr-18 Oct-17 Apr-18 Apr-17 Apr-17 Jun-17 Jan-18 Jun-17 Jan-18 Mar-18 Mar-18 Mar-17 Mar-17 Feb-18 Feb-18 Nov-17 Dec-17 Nov-17 Dec-17 Aug-17 Sep-17 Aug-17 Sep-17 May-18 May-18 May-17 May-17

Otjomuise Clinic Robert Mugabe Clinic 100% 100%

80% 80%

60% 60%

40% 40%

20% 20% 0%

Viral Suppression Rate Suppression Viral 0% Viral Suppression Rate Suppression Viral Jul-17 Oct-17 Apr-18 Apr-17 Jul-17 Jun-17 Jan-18 Mar-18 Mar-17 Feb-18 Nov-17 Dec-17 Aug-17 Sep-17 Oct-17 Apr-18 Apr-17 May-18 May-17 Jun-17 Jan-18 Mar-18 Mar-17 Feb-18 Nov-17 Dec-17 Aug-17 Sep-17 May-18 May-17

Windhoek Central Hospital* 100%

80%

60%

40%

20%

0% Viral Suppression Rate Suppression Viral Jul-17 Oct-17 Apr-18 Apr-17 Jun-17 Jan-18 Mar-18 Mar-17 Feb-18 Nov-17 Dec-17 Aug-17 Sep-17 May-18 May-17 *May not reflect full clinical population due to Page 21 suboptimal screening rates Viral Suppression (Continued) Figure 5f. Viral Suppression Rate—By Site, Ohangwena Region, March 2017-May 2018 Ekoka Clinic Eenhana Clinic 100% 100% 80% 80% 60% 60% 40% 40% 20% 20% 0%

Engela Hospital Odibo Health Center 100% 100%

80% 80%

60% 60%

40% 40%

20% 20%

Okongo Clinic Ongenga Clinic 100% 100%

80% 80%

60% 60%

40% 40%

20% 20% 0%

Ongha Clinic Oshaango Clinic 100% 100%

80% 80%

60% 60%

40% 40%

20% 20% 0%

Page 22 Viral Suppression (Continued) Figure 5f. Viral Suppression Rate—By Site, Ohangwena Region, March 2017-May 2018 (Continued)

Oshaandi Clinic Oshikunde Clinic 100% 100%

80% 80%

60% 60%

40% 40%

20% 20%

0% 0% Viral Suppression Rate Suppression Viral Rate Suppression Viral Jul-17 Jul-17 Apr-18 Apr-18 Oct-17 Oct-17 Apr-17 Apr-17 Jun-17 Jan-18 Jun-17 Jan-18 Mar-18 Mar-18 Mar-17 Mar-17 Feb-18 Feb-18 Nov-17 Dec-17 Nov-17 Dec-17 Aug-17 Sep-17 Aug-17 Sep-17 May-18 May-18 May-17 May-17

Page 23 Viral Suppression (Continued) Figure 5g. Viral Suppression Rate—By Site, Zambezi Region, March 2017-May 2018

Bukalo Health Center Katima Mulilo Clinic 100% 100%

80% 80%

60% 60%

40% 40%

20% 20%

0% 0% Viral Suppression Rate Suppression Viral Viral Suppression Rate Suppression Viral Jul-17 Jul-17 Apr-18 Oct-17 Apr-17 Jun-17 Jan-18 Mar-18 Mar-17 Feb-18 Nov-17 Dec-17 Aug-17 Sep-17 Apr-18 Oct-17 Apr-17 Jun-17 Jan-18 May-18 May-17 Mar-18 Mar-17 Feb-18 Nov-17 Dec-17 Aug-17 Sep-17 May-18 May-17

Katima Mulilo Hospital Mavuluma Clinic 100% 100%

80% 80%

60% 60%

40% 40%

20% 20%

Ngweze Clinic Sesheke Clinic 100% 100%

80% 80%

60% 60%

40% 40%

20% 20% 0%

Sibbinda Clinic 100%

80%

60%

40%

20%

0% Viral Suppression Rate Suppression Viral Jul-17 Oct-17 Apr-18 Apr-17 Jun-17 Jan-18 Mar-18 Mar-17 Feb-18 Nov-17 Dec-17 Aug-17 Sep-17 May-18 May-17

Page 24 Hypertension Screening Prior to NAMPROPA implementation in March 2017, 0% of participating sites were routinely screening and documenting rates of hypertension screening among ART patients. As a result of NAMPROPA activities commencing in March 2017, an average of 14,044 patients were screened for hypertension each month, representing a dramatic increase in coverage relative to baseline (Figure 5a). The performance of some sites was adversely affected by a lack of functional blood pressure monitors (Figure 5b-g); however, with support from HEALTHQUAL, digital blood pressure monitors were procured and distributed to sites. As noted above, baseline in February 2017 for all sites was 0% since blood pressure measurement was not performed or tracked routinely at HIV clinics.

Figure 6a. Number of Adult Patients on ART and Hypertension Screening Rate—NAMPROPA, March 2017- May 2018

96% 96% 20000 19006 94% 92% 95% 100% 89% 89% 17885 86% 18000 83% 85% 17316 83% 90% 81% 16321 15975 78% 15966 18338 18620 16119 15931 16000 74% 80% 16353 16137 14512 14537 80% 14000 70% 12116 12000 60%

10000 50%

8000 40%

6000 30%

4000 20% Rate Screening Hypertension Number of Adult ART on Patients of Number Adult 2000 10%

0 0% Jul-17 Oct-17 Apr-18 Apr-17 Jun-17 Jan-18 Mar-18 Mar-17 Feb-18 Nov-17 Dec-17 Aug-17 Sep-17 May-18 May-17

Figure 6b. Number of Adult Patients on ART and Hypertension Screening Rate—Khomas Region, March 2017-May 2018

9000 95% 95% 93% 95% 100% 91% 92% 87% 86% 85% 86% 85% 86% 7875 90% 8000 82% 7446 82% 6964 6781 7000 6691 6713 6760 80% 6288 6408 6266 6228 69% 6280 6431 70% 6000 5140 60% 5000 4674 50% 4000 40% 3000 30% 2000

20% Rate Screening Hypertension Number of Adult ART on Patients of Number Adult 1000 10%

0 0% Jul-17 Oct-17 Apr-18 Apr-17 Jun-17 Jan-18 Mar-18 Mar-17 Feb-18 Nov-17 Dec-17 Aug-17 Sep-17 May-18 May-17

Page 25 Hypertension Screening (Continued)

Figure 6c. Number of Adult Patients on ART and Hypertension Screening Rate—Ohangwena Region, March 2017-May 2018

9000 96% 95% 100% 8447 93% 90% 91% 92% 88% 86% 8000 90% 82% 7381 80% 7349 81% 7285 80% 7000 6679 6784 6538 6408 6576 6394 6275 68% 6231 6230 76% 70% 6000 63% 58% 60% 5000 4580 50% 3933 4000 40% 3000 30%

2000 Rate Screening Hypertension Number of Adult ART on Patients of Number Adult 20%

1000 10%

0 0% Jul-17 Oct-17 Apr-18 Apr-17 Jun-17 Jan-18 Mar-18 Mar-17 Feb-18 Nov-17 Dec-17 Aug-17 Sep-17 May-18 May-17

Figure 6d. Number of Adult Patients on ART and Hypertension Screening Rate—Zambezi Region, March 2017-May 2018

100% 99% 98% 100% 99% 99% 98% 99% 99% 4500 94% 100% 90% 90% 91% 3961 4000 90% 82% 3701 3778 3540 3607 3608 3454 3430 80% 3331 3500 3268 80% 3244 3149 3059 3043 2924 70% 3000 60% 2500 50% 2000 40% 1500 30%

1000 Rate Screening Hypertension Number of Adult ART on Patients of Number Adult 20%

500 10%

0 0% Jul-17 Oct-17 Apr-18 Apr-17 Jun-17 Jan-18 Mar-18 Mar-17 Feb-18 Nov-17 Dec-17 Aug-17 Sep-17 May-18 May-17

Page 26 Hypertension Screening (Continued) Figure 6e. Hypertension Screening Rate—By Site, Khomas Region, March 2017-May 2018

Katutura Health Center Katutura Intermediate Hospital 100% 100%

80% 80%

60% 60%

40% 40%

20% 20%

0% 0% Jul-17 Apr-18 Oct-17 Apr-17 Jul-17 Jun-17 Jan-18 Mar-18 Mar-17 Feb-18 Nov-17 Dec-17 Aug-17 Sep-17 Apr-18 Oct-17 Apr-17 Hypertension Rate Screening Hypertension May-18 May-17 Jun-17 Jan-18 Mar-18 Mar-17 Feb-18 Nov-17 Dec-17 Aug-17 Sep-17 Hypertension Rate Screening Hypertension May-18 May-17

Khomasdal Clinic Okuryangava Clinic 100% 100%

80% 80%

60% 60%

40% 40%

20% 20%

0% 0% Jul-17 Jul-17 Oct-17 Apr-18 Oct-17 Apr-18 Apr-17 Apr-17 Jan-18 Jan-18 Jun-17 Jun-17 Mar-18 Mar-18 Mar-17 Mar-17 Feb-18 Feb-18 Nov-17 Dec-17 Nov-17 Dec-17 Aug-17 Sep-17 Aug-17 Sep-17 Hypertension Rate Screening Hypertension Rate Screening Hypertension May-18 May-18 May-17 May-17

Otjomuise Clinic Robert Mugabe Clinic 100% 100%

80% 80%

60% 60%

40% 40%

20% 20%

0% 0% Jul-17 Oct-17 Apr-18 Apr-17 Jul-17 Jun-17 Jan-18 Mar-18 Mar-17 Feb-18 Nov-17 Dec-17 Aug-17 Sep-17 Oct-17 Apr-18 Apr-17 Hypertension Rate Screening Hypertension May-18 May-17 Jun-17 Jan-18 Mar-18 Mar-17 Feb-18 Nov-17 Dec-17 Aug-17 Sep-17 Hypertension Rate Screening Hypertension May-18 May-17

Windhoek Central Hospital 100%

80%

60%

40%

20%

0% Viral Load Monitoring Rate Monitoring Load Viral Jul-17 Oct-17 Apr-18 Apr-17 Jun-17 Jan-18 Mar-18 Mar-17 Feb-18 Nov-17 Dec-17 Aug-17 Sep-17 May-18 May-17 Page 27 Hypertension Screening (Continued) Figure 6f. Hypertension Screening Rate—By Site, Ohangwena Region, March 2017-May 2018

Eenhana Clinic Ekoka Clinic 100% 100%

80% 80%

60% 60%

40% 40%

20% 20%

0% 0% Jul-17 Jul-17 Apr-18 Oct-17 Apr-17 Jun-17 Jan-18 Apr-18 Oct-17 Apr-17 Mar-18 Mar-17 Feb-18 Jun-17 Jan-18 Nov-17 Dec-17 Aug-17 Sep-17 Mar-18 Mar-17 Feb-18 Hypertension Rate Screening Hypertension May-18 May-17 Nov-17 Dec-17 Aug-17 Sep-17 Hypertension Rate Screening Hypertension May-18 May-17

Engela Hospital Odibo Health Center 100% 100%

80% 80%

60% 60%

40% 40%

20% 20%

Okongo Clinic Ongenga Clinic 100% 100%

80% 80%

60% 60%

40% 40%

20% 20%

0% 0% Jul-17 Jul-17 Oct-17 Apr-18 Apr-17 Oct-17 Apr-18 Apr-17 Jun-17 Jan-18 Jun-17 Jan-18 Mar-18 Mar-17 Feb-18 Mar-18 Mar-17 Feb-18 Nov-17 Dec-17 Aug-17 Sep-17 Nov-17 Dec-17 Aug-17 Sep-17 Hypertension Rate Screening Hypertension May-18 May-17 Hypertension Rate Screening Hypertension May-18 May-17

Ongha Clinic Oshaango Clinic 100% 100%

80% 80%

60% 60%

40% 40%

20% 20%

0% 0% Loss to Follow Up Rate Up Followto Loss Loss to Follow Up Rate Up Followto Loss Jul-17 Jul-17 Oct-17 Apr-18 Apr-17 Jun-17 Jan-18 Oct-17 Apr-18 Apr-17 Jun-17 Jan-18 Mar-18 Mar-17 Feb-18 Nov-17 Dec-17 Aug-17 Sep-17 Mar-18 Mar-17 Feb-18 Nov-17 Dec-17 Aug-17 Sep-17 May-18 May-17 May-18 May-17

Page 28 Hypertension Screening (Continued) Figure 6f. Hypertension Screening Rate—By Site, Ohagnwena Region, March 2017-May 2018 (Continued) Oshikunde Clinic Oshaandi Clinic 100% 100% 80% 80%

60% 60%

40% 40%

20% 20%

0% 0% Hypertension Rate Screening Hypertension Jul-17 Jul-17 Apr-18 Apr-18 Oct-17 Oct-17 Apr-17 Apr-17 Jun-17 Jan-18 Jun-17 Jan-18 Mar-18 Mar-18 Mar-17 Mar-17 Feb-18 Feb-18 Nov-17 Dec-17 Nov-17 Dec-17 Aug-17 Sep-17 Aug-17 Sep-17 Hypertension Rate Screening Hypertension May-18 May-18 May-17 May-17

Page 29 Hypertension Screening (Continued) Figure 6g. Hypertension Screening Rate—By Site, Zambezi Region, March 2017-May 2018 (Continued)

Katima Mulilo Clinic Bukalo Health Center 100% 100% 80% 80% 60% 60% 40% 40% 20% 20% 0% 0% Jul-17 Apr-18 Oct-17 Apr-17 Jun-17 Jan-18 Mar-18 Mar-17 Feb-18 Nov-17 Dec-17 Aug-17 Sep-17 Jul-17 Hypertension Rate Screening Hypertension May-18 May-17 Apr-18 Oct-17 Apr-17 Jun-17 Jan-18 Mar-18 Mar-17 Feb-18 Nov-17 Dec-17 Aug-17 Sep-17 Hypertension Rate Screening Hypertension May-18 May-17

Katima Mulilo Hospital Mavuluma Clinic 100% 100%

80% 80%

60% 60%

40% 40%

20% 20%

Ngweze Clinic Sesheke Clinic 100% 100%

80% 80%

60% 60%

40% 40%

20% 20%

Sibbinda Clinic 100%

80%

60%

40%

20%

0% Viral Load Monitoring Rate Monitoring Load Viral Jul-17 Oct-17 Apr-18 Apr-17 Jun-17 Jan-18 Mar-18 Mar-17 Feb-18 Nov-17 Dec-17 Aug-17 Sep-17 May-18 May-17 Page 30 Hypertension Treatment After the first 6 months of NAMPROPA implementation, the indicator for hypertension treatment was modified to reflect treatment rates among patients newly diagnosed with hypertension as opposed to those who were previously diagnosed and active on antihypertensive treatment. Between September 2017 and February 2018, 1,508 patients on ART were newly diagnosed with hypertension, of which 854 (57%) were successfully initiated on treatment (Figure 7a). Hypertension treatment performance rates were negatively impacted by a number of key system-level barriers, including stock-outs of hypertensive medications, lack of physicians to initiate treatment, and a lack of guidelines for referral of newly diagnosed patient to tertiary sites for treatment initiation.

Figure 7a. Number of Adult Patients on ART with Newly Diagnosed Hypertension and Hypertension Treatment Rate—NAMPROPA, September 2017-May 2018

450 100% 398 400 90% 80% 350 70% 70% 300 59% 60% 59% 267 56% 54% 57% 51% 60% 250 234 272 45% 251 205 50% 200 236 199 40% 150 30% 101 100 Hypertension Rate Treatment Hypertension

Diagnosed with Hypertension with Diagnosed Hypertension 20%

50 10% Number of Adult Patients on ART Newly ART on Newly Patients of Number Adult 0 0% Oct-17 Apr-18 Jan-18 Mar-18 Feb-18 Nov-17 Dec-17 Sep-17 May-18

Figure 7b. Number of Adult Patients on ART with Newly Diagnosed Hypertension and Hypertension Treatment Rate—Khomas Region, September 2017-May 2018

300 96% 100% 264 90% 250 80% 70% 70% 200 60% 60% 152 48% 48% 145 49% 49% 150 45% 50% 39% 151 136 107 40% 100 86 75 30% Hypertension Rate Treatment Hypertension

Diagnosed with Hypertension with Diagnosed Hypertension 20% 50 40 10% Number of Adult Patients on ART Newly ART on Newly Patients of Number Adult 0 0% Oct-17 Apr-18 Jan-18 Mar-18 Feb-18 Nov-17 Dec-17 Sep-17 May-18

Page 31 Hypertension Treatment (Continued)

Figure 7c. Number of Adult Patients on ART with Newly Diagnosed Hypertension and Hypertension Treatment Rate—Ohangwena Region, September 2017-May 2018

90 100% 100% 83 86% 90% 80 75 76 73 77% 75% 80% 70 63 70% 60 56 62% 53% 60% 50 63% 47% 50% 40 41% 40%

with Hypertension withHypertension 30 28 28 30%

20 Rate Treatment Hypertension 14 20%

10 10% Number of Adult Patients on ART Newly ART on Newly Diagnosed Patients of Number Adult 0 0% Oct-17 Apr-18 Jan-18 Mar-18 Feb-18 Nov-17 Dec-17 Sep-17 May-18

Figure 7d. Number of Adult Patients on ART with Newly Diagnosed Hypertension and Hypertension Treatment Rate—Zambezi Region, September 2017-May 2018

160 97% 100% 100% 145 86% 90% 140 83% 79% 86% 80% 120 70% 100 92 60%

80 50% 39% 35% 35% 59 59 40% 60 with Hypertension withHypertension 30% 37

40 33 Rate Treatment Hypertension 29 32 25 20% 20 10% Number of Adult Patients on ART Newly ART on Newly Diagnosed Patients of Number Adult 0 0% Oct-17 Apr-18 Jan-18 Mar-18 Feb-18 Nov-17 Dec-17 Sep-17 May-18

Page 32 HIV Quality Management Capacity Participating sites were administered pre- and post-implementation organizational assessments (OA), which are part of the MoHSS Quality Management (QM) capacity building framework. The OA is a validated tool developed by HEALTHQUAL to capture site-level HIV QM capacity according to nine domains: leadership support, quality committee, quality plan, staff engagement, performance measurement and data use, QI initiatives, patient engagement, program evaluation, and outcomes monitoring. Each domain is scored on a scale of 0-5, with 5 indicating the existence of a “full systematic approach to quality management” and 0 Indicating “no QM structures.” To assess the effect of Collaborative activities on sites’ QM capacity, sites were assessed at baseline during pre-work activities in January 2017, and re-assessed during Learning Session 3 in February 2018. Improvements in average OA score were noted across all domains, with the most significant improvement seen in outcomes monitoring (Figure 8). Although slight improvements were seen in patient engagement, post-implementation performance remained low, underscoring a need to integrate patients more fully into sites’ existing QM activities. Of note, the 24 participating sites convened a total of 204 quality management meetings during NAMPROPA implementation—a significant increase compared to baseline.

Figure 8. Average Pre- And Post-Implementation OA Scores---All Sites, by Domain

5.0 Pre-implementation (2017) Post-implementation (2018) 4.5

4.0 3.5 3.1 3.0 3.0 3.1 3.0 2.6 2.6

2.0

1.0 0.9 1.0 0.5 0.5 0.4 0.5 0.2 0.3 0.0 0.0 0.0 Leadership Quality Quality Plan Staff PM & Data QI Initiatives Patient Program Outcomes Support Committee Engagement Use Engagement Evaluation Monitoring

Page 33 Change Package As part of LS3, teams from participating sites presented adopted interventions (“change ideas”) that were associated with improvements in indicator performance. Of 63 distinct interventions tested during NAMPROPA implementation, 38 were subsequently adopted (Table 3). As part of scale-up activities in CY 2018, these interventions will be further characterized and then disseminated as an evidence-based “change package.”

Table 3. Adopted Interventions by Indicator

Indicator Adopted Intervention Loss to follow-up 1. Printing and analyzing list of LTFU before tracing 2. Telephonic tracing 3. Physical tracing of LTFU through home visits 4. Cross-checking LTFU lists with nearby facilities 5. Establish community-based ART (CBART) sites 6. Telephonic reminder of patients who missed follow-up appointments 7. Inspect health passport before ANC and escort PMTCT patients to ART clinic 8. Updating contact details for those who have changed contacts 9. Validating list of LTFU from ePMS with EDT

Viral load 1. Coordination of VL testing date with pharmacy pickup and follow-up appointments monitoring 2. Reminder put in patient health passport on when to come for VL monitoring 3. Development of VL monitoring tracking register 4. Placement of viral load monitoring reminders in patient care booklets 5. Telephonic tracing of patients who missed VL monitoring 6. 1-week supply of medication for patients eligible for VL monitoring who sent someone for medication pick-up 7. Placement of viral load monitoring register in phlebotomy room

Viral load 1. Provide directly observed therapy to selected patients with high VL suppression 2. Introduction of high VL register and close monitoring 3. Development of adherence support group TRIOs for patients with high VL 4. Intensified use of high VL tracking register 5. Perform pill counts to assess adherence 6. Involvement of treatment supporters for adolescents with high VL 7. Issuing of pill boxes to patients with high VL 8. Health education and intensive adherence counseling 9. Involvement of school matrons to monitor adherence

Hypertension 1. Redesigning patient flow in the clinic screening 2. Development of blood pressure (BP) monitoring register 3. Task shifting BP monitoring to health assistants 4. Introduction of digital BP machines 5. Recording of BP readings in health passport and patient care booklets 6. Provision of start dose in consulting room for patients with hypertension 7. Implementing BP monitoring at CBART sites 8. Queue jumping to facilitate follow-up BP screening for patients with high BP

Hypertension 1. Development of registers to capture patients with high BP treatment 2. On-site initiation of hypertension treatment when medical officers are available 3. Stocking hypertension medication at the ART clinic 4. Maintaining adequate stock of hypertension medications 5. Referral of patients with hypertension for initiation of treatment 6. Patients with elevated BP getting initial dose in consulting room

Page 34 Knowledge Management Between November 2016 and February 2018, results from NAMPROPA were presented at numerous national and international conferences and forums. Of particular note was the representation of NAMPROPA at IHI’s 1st Annual Africa Forum on Quality and Safety in Healthcare, February 19-21, 2018, in Durban, South Africa, where eight NAMPROPA-related abstracts were presented (Table 4).

Table 4. NAMPROPA Presentations and Abstracts Shared at IHI’s 1st Annual Africa Forum Title First Author “Setting Up a Quality Improvement Collaborative on Retention in HIV Care Julie Neidel Across 24 HIV Care Sites, The Namibian Experience” “Successful Integration of Hypertension Screening and Treatment in an Adult Munduu Tjondu Urban HIV Care Clinic” “Improving Viral Load Suppression Among Patients on Antiretroviral Therapy Dr. Mireille A Mpalang Kakubu (ART) at Khomasdal Health Center in Namibia” “Improving Retention in Care Among Patients on Antiretroviral Therapy at Dr. Mireille A Mpalang Kakubu Windhoek Central Hospital in Namibia” “Project to Improve HIV Viral Load Monitoring and Suppression at Otjomuise Ndahafa Ndeikemona Clinic in Namibia” “Improving Retention in Care for Patients on Antiretroviral Therapy in A Rural Dr. Simbarashe Mpariwa HIV Clinic in Namibia” “Improving Viral Load Monitoring in Patients on Antiretroviral Therapy at Dr. Simbarashe Mpariwa Mavuluma Clinic in the Zambezi Region of Namibia” “Using a Quality Improvement Collaborative Approach to Improve Routine HIV Dr. Simbarashe Mpariwa Viral Load Monitoring at Sesheke Clinic “Improving Retention in Care Among Patients on Antiretroviral Therapy at Dr. Mireille A Mpalang Kakubu Windhoek Central Hospital in Namibia” “Overview of NAMPROPA QI Collaborative” Dr. Apollo Basenero

Plans for Sustainability In an effort to maintain, and further build upon, gains made through NAMPROPA, participants developed facility, district/regional, and national sustainability plans during the third learning session. These strategies are summarized in Table 5.

Table 5. Facility, District/Regional, National Strategies for Sustainability Level Strategy Facility • Continue to convene monthly QI meetings • Generate and analyze monthly performance reports • Visually display performance data • Meaningfully engage consumers in QI activities • Continuous QI in-service training • Continuous QI coaching activities District/Regional • Incorporating QI support into existing structures • Clarifying roles and responsibilities • Ensuring active involvement of district and regional management teams • Data management and reporting • Using DOH vehicles/providing refreshments to support participation in regional meetings • Tracking implementation at each site • Continuing coaching and mentoring • Supporting in-service trainings • Convening quarterly regional QI meetings. National • Provision of continuous technical support and data analysis

Page 35 Implementation Challenges, Solutions, and Recommendations Like many large-scale quality improvement initiatives, NAMPROPA faced a number of system-level implementation challenges which spanned three broad domains: commodities/supply chain, human resources, and data systems.

Table 6. Implementation Challenges, Solutions, and Recommendations Domain Challenge Solutions/Recommendations Commodities/Supply Chain • Poor network coverage/ • Mentors collect facility dashboards internet connectivity from sites who cannot submit reports via email/get monthly reports telephonically

• Inadequate or no blood • 64 BP monitors were procured pressure monitors at launch of and distributed to all sites with NAMPROPA funds solicited through HEALTHQUAL • Transport of viral load • Dedicated days for VL specimen specimens from remote sites collection and increased pick-up days for VL samples • Improved communication whenever there is transport available to assist with sample transportation • Occasional stock-outs of • Training on stock management at antihypertensive medications all levels Human Resources • No resident physician to • Patients referred to nearest facility initiate antihypertensive with resident physician/ periodic treatment on site rotation of physicians to sites not covered • MoHSS to consider task-shifting of hypertension treatment initiation to registered nurses • Staff Turnover • Continuous in-service training and coaching provided to new staff

Data Systems • Poor initial documentation of • Completed example of PDSA change ideas at some sites template shared with sites for guidance • Incomplete and delayed • Data tracking tool developed at monthly facility reports the national level and feedback provided to regions accordingly • Sporadic documentation of • Template for capturing minutes facility QI meetings developed and shared with sites • Inaccuracy of ePMS- • Paper registers developed at generated performance facility level to track indicators indicators • Feedback provided to M & E unit • Initial lack of age- and sex- • Excel database modified to disaggregated data include age and sex variables • Ongoing challenges tracking • Excel database modified to track cross-border patients identified cross-border patients

Page 36 Plans for Scale-Up Beginning April 2018, MoHSS plans to scale-up NAMPROPA activities to other high-volume ART sites in two waves (Figure 9). In the first wave, NAMPROPA activities will be spread to 19 additional sites in the existing three regions: 4 in Khomas, 8 in Ohangwena, and 7 in Zambezi. In the second wave, activities will be introduced to an additional 29 ART sites in Namibia’s remaining 11 regions.

Figure 9. Scale-Up Strategy for NAMPROPA Activities

NAMPROPA Scale-Up Wave 1 Scale-Up Wave 2 3 Regions, 24 Facilities 3 Regions, 43 Facilities 14 Regions, 72 Facilities

Conclusion Improving rates of loss to follow-up, viral load monitoring, and viral suppression is crucial to reducing HIV- related morbidity and mortality and achieving UNAIDS’ 90-90-90 targets for Namibia. Implementation of NAMPROPA, a quality improvement collaborative, led to improvements in loss to follow-up, viral load monitoring, viral load suppression, and hypertension screening and treatment initiation across 24 sites in Khomas, Ohangwena, and Zambezi Regions.

Acknowledgements UCSF-HEALTHQUAL wishes to acknowledge Dr. Mireille Kakubu, Dr. Jacques Kamangu, Dr. Simbarashe Mpariwa, Dr. Apollo Basenero, and Ms. Hilaria Ashivudhi for their assistance with the creation of this report. This report was authored and reviewed by staff of UCSF-HEALTHQUAL, including Dr. Bruce Agins, Mr. Richard Birchard, Mr. Joshua Bardfield, Mr. Dan Ikeda, and Ms. Julie Neidel. This work was funded by the U.S. President’s Emergency Plan for AIDS Relief (PEPFAR) as part of the Health Resources and Services Administration’s (HRSA) Quality Improvement Capacity for Impact Project (QICIP) award #U1NHA08599. Any errors or omissions are the sole responsibility of HEALTHQUAL.

Page 37 Appendix

Figure 9. NAMPROPA Driver Diagram

Page 38 Figure 9. NAMPROPA Driver Diagram (Continued)

Page 39 Figure 9. NAMPROPA Driver Diagram (Continued)

Figure 10. NAMPROPA Data Collection Tool

Page 40