Attachment_J.8_-_SOP_for_Reporting_Data_from_DHIS2.pdf

Standard Operating Procedure # 1:

Generating PEPFAR Reports from the National eHMIS/DHIS2 System

Monitoring and Evaluating Emergency Plan Progress (MEEPP) Project

Version 1.0 Updated: June 9, 2015

The Monitoring and Evaluation Emergency Plan Progress (MEEPP) II project and this standard operating procedure are made possible by the generous support of the American people through the United States Agency for International Development (USAID) under the terms of

USAID/Social & Scientific Systems (SSS) Contract No. 697-C-10-00008. The contents do not necessarily reflect the views of USAID.

Contents Acronyms

Introduction

National Data Systems

MER Indicators for PEPFAR Reporting

Key Terms

Scope of SOP

Use of the SOP

Target Audience of SOP

I. Responsibilities for collection, cleaning, and reporting DHIS2 Data

1. Facility Staff

2. District-level M&E Staff

3. PEPFAR –supported Implementing Partners (IPs)

4. National-level MoH Staff

5. PEPFAR M&E Contractor

6. PEPFAR Coordinating Office (PCO) Staff

7. PEPFAR Technical Working Groups (TWG)

II. Procedures for Generating PEPFAR Indicators from DHIS2 Data

1. Data Collection, Aggregation, and Reporting at Health Facility

2. Data Review and Cleaning at District/Health Sub-District Levels

2.1. Review of HMIS Data Forms Submitted to the District

2.2. Running Completeness Reports and Validation Checks within DHIS2

2.3. Data Extraction, Applying Data Quality Checks and Providing Feedback on Data Quality 22

3. Data Review and Cleaning at National Level

3.1. Obtaining Access to the DHIS2 Database

3.2. Checking for Completeness of Reporting within DHIS2

3.3. Data Extraction from DHIS2 using the Pivot Table Tool

3.4. Data Extraction from DHIS2 using Web API

3.5. Linking DHIS2 to PEPFAR IP Program Data

3.6. Running Validation Checks on the Downloaded Data

3.7. Final Data Cleaning Process

4. Aggregation and Mapping of Data Elements to MER Indicators

4.1. Aggregating facility data across the reporting period

4.2. Other Considerations

4.3. Mapping PEPFAR indicators from DHIS2 data variables

5. Submitting Reports into DATIM

5.1. Update the Integrated PEPFAR-Supported Site List (iPSL) in DATIM:

5.2. Mapping sites to their iPSL IDs:

5.3. Code and Format the Data for DATIM Submission

5.4. Check for Duplicates

5.5. Size and Format Datasets for Submission

5.6. Access the DATIM Test Exchange Environment

5.7. Import Test Datasets Using a “Dry Run” on Test Server

5.8. Data Verification by OGAC

5.9. Final Import to Production Server

6. Archiving Cleaned Datasets for PEPFAR Reporting

III. Attachments

Acronyms

AIDS Acquired Immune Deficiency Syndrome

APR Annual Program Results

CAO Chief Administrative Officer

COP Country Operational Plan

CPHL Central Public Health Laboratory

DAPTS Data Accountability Platform for Type of Support

DATIM Data for Accountability Transparency Impact Monitoring

DFID Department for International Development

DHIS2 District Health Information System, Version 2

DHO District Health Officer

DoD Department of Defense eHMIS Electronic Health Management Information System

EID Early Infant Diagnosis

HIBRID HIV Based Real-time Integrated Database

HISSIP Health Sector Strategic Investment Plan

HIV Human Immunodeficiency Virus

HMIS Health Management Information System

HSD Health Sub-District

HSS Health Sector Strengthening

HTC HIV Testing and Counseling

IM Implementing Mechanism

IP Implementing Partner iPSL Integrated PEPFAR-supported Site List

JAF Joint Action Framework

M&E Monitoring & Evaluation

MCH Maternal and Child Health

MDGs Millennium Development Goals

MER Monitoring, Evaluation, and Reporting

MoGLSD Ministry of Gender, Labour, and Social Development

MoH ACP Ministry of Health AIDS Control Program

MoH RC Ministry of Health Resource Center

NDP National Development Plan

NGI Next Generation Indicators

NOP National OVC Policy

NSPPI National Strategic Program Plan for OVC interventions

OGAC Office of the Global AIDS Coordinator

OVC Orphans and Vulnerable Children

OVC MIS Orphan and Vulnerable Children Management Information System

PCO PEPFAR Coordinating Officer

PCR Polymerase Chain Reaction

PEPFAR President's Emergency Plan for AIDS Relief

PMTCT Prevention of Mother to Child Transmission

PRS Partner Reporting System

SAPR Semi-Annual Program Results

SI TWG Strategic Information Technical Working Group

SMC Safe Medical Circumcision

SOP Standard Operating Procedure

TB Tuberculosis

TWG Technical Working Group

UAC Uganda AIDS Commission

UBTS Uganda Blood Transfusion Service

UPDF Uganda People's Defense Forces

USG United States Government

VMMC Voluntary Male Medical Circumcision

Introduction

The Monitoring and Evaluation of Emergency Plan Progress – Phase II (MEEPP II) project is a 6-year contract awarded by the U.S. Agency for International Development (USAID) to Social &

Scientific Systems, Inc., in September 2010. MEEPP II is designed primarily to support program performance management for the United States (US) President’s Emergency Plan for AIDS

Relief (PEPFAR), including data use for program improvement, target setting, and results reporting. The secondary objective is to work with multiple stakeholders to provide structured support to the national HIV/AIDS monitoring and evaluation (M&E) reporting system with the expectation that after the successful completion of the MEEPP II project, PEPFAR will rely on a centralized/consolidated national system for most of its data needs.

Through its work over the past 4 years, MEEPP II has demonstrated that PEPFAR can meet most of its needs for data reporting using Uganda’s national M&E systems. Under the technical direction and oversight of the PEPFAR/Uganda program, MEEPP has worked with the Ministry of Health Resource Center (MoH/RC), National AIDS Control Program (ACP), and U.S.

Government (USG) implementing partners (IPs) to provide support for the customization, deployment, and use of DHIS 2 by all 112 districts in Uganda. As importantly, MEEPP II has worked collaboratively with these entities to support data cleaning and quality improvement processes for HIV/AIDS programming. MEEPP has also worked with the Ministry of Gender, Labour and Social Development (MoGLSD), providing technical assistance in the improvement of the orphans and vulnerable children management information system (OVC MIS).

This standard operating procedure (SOP) is the first of a series that describe processes used for generating PEPFAR annual and semi-annual program results for a subset of the current (US)

President’s Emergency Plan for AIDS Relief (PEPFAR) Monitoring, Evaluation and Reporting (MER) indicators1 from data that are collected, reported and stored in national monitoring and evaluation (M&E) systems. This is in line with the UNAIDS “Three Ones” principle that underscores the need for stakeholders in countries’ HIV/AIDS programs to support and utilize

“one M&E Framework” that utilizes “one M&E system.2”

1 PEPFAR. March 2015. Monitoring, Evaluation, and Reporting Indicator Reference Guide Version 2.1.

2 UNAIDS. (2004) “Three Ones” key principles: “Coordination of National Responses to HIV/AIDS” Guiding principles for national authorities and their partners. http://data.unaids.org/UNA-docs/Three- Ones_KeyPrinciples_en.pdf http://data.unaids.org/UNA-docs/Three-Ones_KeyPrinciples_en.pdf

National Data Systems

Routine information related to HIV/AIDS and orphans and vulnerable children’s (OVC) programs is mainly gathered from two major national monitoring and evaluation systems in Uganda: the national health management information system (HMIS) and the OVC management information system (MIS).

The essential function of the HIMIS is to equip all heath stakeholders with information for planning, managing, monitoring and evaluating the health care delivery system. HMIS was developed to address the information needs of the Health Sector Strategic Investment Plan

(HSSIP)3 for monitoring the health sector towards the Joint Action Framework, Uganda’s

National Development Plan (NDP)4 and the Millennium Development Goals (MDGs)5 and it can provide all national information requirements and some international reporting indicators, including a subset of the PEPFAR MER indicators. HMIS includes both paper based reporting tools and an electronic database the District Health Information System version 2 (DHIS2). In addition to the HMIS, a national reporting system managed by the Central Public Health

Laboratory (CPHL) collects data on laboratories (lab) and lab tests performed, while that of the

Uganda Blood Transfusion Centre (UBTS) collects data on blood transfusion services and

Uganda Virus Research Institute collects data on lab proficiency testing.

The National Orphans and Other Vulnerable Children (OVC) Policy and National Strategic

Programme Plan of Interventions for OVC (NSPPI)6 provide the framework for strategic interventions, implementation, coordination, monitoring and evaluation (M&E) of OVC programs in the country. The OVC MIS was conceptualized in 2008 as a web-based management information system to support the monitoring of implementation of OVC interventions, as well as to measure progress and evaluate performance in the effective implementation of NSPPI.

A large amount of information can be gathered from these national data systems to guide

HIV/AIDS related program management. However different stakeholders including PEPFAR may require specific information that is not available in these national systems. In this case, the information will be collected from alternative sources, foremost of which is the Partner

Reporting System (PRS) that collects data from the US Government (USG) implementing

3 Ministry of Health. 2010. Health Sector Strategic Investment Plan 2010/11-2014/15.

4 National Planning Authority (2010). Uganda National Development Plan 2010/11-2014/15.

5 United Nations (2011): The Millennium Development Goals Report 2011. New York.

6 Ministry of Gender, Labour and Social Development, May 2011. National Strategic Programme Plan of Interventions for Orphans And Other Vulnerable Children 2011/12—2015/16 partners (IPs) and stores it in the PRS database – the HIV Based Real-time Integrated Database

(HIBRID) - hosted by the PEPFAR M&E contractor.

MER Indicators for PEPFAR Reporting

This document refers to the set of 44 level one & level two indicators detailed in the PEPFAR

Monitoring, Evaluation, and Reporting (MER) Indicator Reference Guide, Version 2.1.1 In

FY2015, MEEPP supported PEPFAR/Uganda to transition to reporting the MER indicators from the previous Next Generation Indicators (NGI), though not all disaggregations of the MER indicator set are reported yet.

At present, 22 PEPFAR MER indicators can be derived from different tools within the

HMIS/DHIS2, CPHL, UBTS, or OVC MIS national reporting systems, though some disaggregations are not available. It will be possible to aggregate an additional four (4) MER indicators from the revised HMIS tools, once these have been rolled out in the country, and one (1) from the OVC

MIS. This gives the potential for 27 MER indicators to be reported using national systems.

Another 17 level one and level two MER indicators cannot be derived at all from national reporting systems. Table I details the MER indicators and their availability from national M&E systems.

Table 1: MER Indicators that can be derived from national data systems7

Indicator Level

Indicator Label Data Source Comments

1 OVC_SERV OVC MIS Not yet in effect, PRS still used

1 TB_IPT HMIS, Revised 2015 Not yet in effect, TB IPT not yet rolled out in Uganda

2 TB_OUTCOME HMIS, Revised 2015 Not yet in effect, PRS still used

1 FN_THER HMIS, Revised 2015 Not yet in effect, PRS still used

2 FN_ASSESS HMIS, Revised 2015 Not yet in effect, PRS still used

1 HTC_TST HMIS

1 PMTCT_STAT HMIS

1 PMTCT_ARV HMIS

1 PMTCT_EID HMIS

1 PMTCT_FO HMIS

Sub-disaggregation not available, reported through

PRS

2 PMTCT_CTX HMIS

1 VMMC_CIRC HMIS PEPFAR and HMIS age disaggregations differ

7 A detailed mapping of MER indicators to the DHIS2 and OVC MIS tools can be found in the annex document titled “4_Mapping MER Indicators to National Data Systems.xlsx”

Indicator Level

Indicator Label Data Source Comments

1 FPINT_SITE HMIS

Disaggregation by support type reported through

PRS

1 TB_STAT HMIS PEPFAR age disaggregation not available

1 TB_ART HMIS

PEPFAR age, positivity status, & ART initiation timeline disaggregations not available

2 TB_SCREEN HMIS

1 TX_NEW HMIS PEPFAR and HMIS age disaggregations differ

1 TX_CURR HMIS PEPFAR and HMIS age disaggregations differ

1 TX_RET HMIS PEPFAR Age-Sex disaggegations not available

1 LAB_CAP HMIS

Disaggregation by testing facility type and site support type reported through PRS

2 HRH_DENS MoH HRH Database

1 CARE_NEW HMIS

1 CARE_CURR HMIS

1 SC_STOCK HMIS

1 LAB_ACC CPHL Data provided by CPHL

1 LAB_PT CPHL Data provided by CPHL/Specialized Labs

1 BS_COLL UBTS Data provided by UBTS

The PEPFAR M&E contractor also aggregates seven8 National Level 1 MER Indicators annually.

Four of the seven national Level 1 indicators are derived from HMIS/DHIS2 sources. The computation of these indicators is similar to the process outlined in this SOP, with the exception that the indicators are aggregated for the nation across all reporting sites, without restriction to PEPFAR-supported facilities. Table 2 contains a list of National Level 1 Indicators for Uganda.

Table 2: National Level 1 Indicators for Uganda derived from national data systems

Program Area

Indicator Label Data Source Comments

Prevention

PMTCT_STAT_NAT HMIS

PMTCT_ARV_NAT HMIS

VMMC_CIRC_NAT HMIS PEPFAR and HMIS age disaggregations differ

Treatment TX_CURR_NAT HMIS PEPFAR and HMIS age disaggregations differ

Additionally, the PEPFAR M&E contractor will design a survey to gather the data for the

Essential Survey Indicators and supervises consultants who field the survey once every two

8 The number of people who inject drugs (PWID) on medication assisted therapy (KP_MAT_NAT) is not a relevant indicator in Uganda, as there are few PWID.

years. For a complete flowchart of the PEPFAR reporting process as of May 2015, see Figure 1 below and the annex document “6_Flowchart_PEPFAR Reporting from National

Systems_SAPR.pdf”.

Figure 1: Data Flow for PEPFAR Results Reporting

Key Terms

CPHL: The Uganda National Central Public Health Laboratories, responsible for the development of the national laboratory policy, guidelines and standards; monitoring and evaluating national health laboratory services; mobilizing resources for laboratory services;

supervising laboratory quality assurance and control countrywide and participating in operational research. They provide reference and specialized testing services to support both clinical and public health programs.

DAPTS: A module within the HIBRID database that enables PEPFAR implementing partners to specify reliable and accurate information on the type of support provided to all PEPFAR-supported sites.

Data Quality Assessment and Improvement (DQAI): a critical component of the PEPFAR data management process that includes specific procedures for verifying effectiveness of data management systems for producing reliable and accurate data.

DHIS2: DHIS2 is open-source software for web-based health information system databases.

This software was developed under the supervision of the Health Information Systems Programme (HISP) at the University of Oslo with support from Norad, the Research Council of Norway, PEPFAR, and the Global Fund. However, throughout this SOP, DHIS2 refers to the Ugandan adaptation of the DHIS2 tool that is hosted by the MoH RC.

District: Uganda is divided into 112 district-level administrative units (111 districts and the city of Kampala). Under the decentralized health sector management, the districts play a major role in the oversight of the health sector at the lower levels.

HIBRID: The HIV Based Real-time Integrated Database is a web-based data reporting and warehousing tool hosted by the PEPFAR M&E contractor for the purpose of collecting data from PEPFAR IPs operating in Uganda. HIBRID uses the same design as the DHIS2 database hosted by the MoH RC for collection of Uganda HMIS data.

HMIS: The Health Management Information System (HMIS) was a key component to fill information needs related to the HSSIP. It has been designed by the MoH using paper based data collection tools, paper-based summary tools, and an electronic web-based database

(DHIS2).

MOH/Resource Centre: The national MoH is divided into disease and management specific programs, but also has a national body that coordinates the overall HIV/AIDS M&E strategy across program areas – the Uganda AIDS Control Commission (UAC). The MoH RC is responsible for the HMIS data collection tools and is the custodian of the DHIS2 database.

MOH/NACP: The National AIDS Control Programme of Ugandan MoH is the primary organization responsible for HIV/AIDS data collection tools and validation of national-level data.

The PEPFAR M&E contractor works closely with the MoH NACP (aka ACP) to establish consensus on data cleaning guidelines and to share cleaned data.

OVC MIS: An M&E system that aims to measure progress towards achieving the goal of the National Strategic Program Plan of OVC Interventions (NSPPI). It consists of community, district and national level tools and processes that ensure data that informs the national OVC Program.

PEPFAR Implementing Partner: Implementing partners are organizations that support the joint USG-Government of Uganda (GOU) effort to address the HIV/AIDS epidemic with PEPFAR funding support. They provide support through either direct service delivery (DSD) or technical assistance for service delivery improvement (TA-SDI). Implementing partners are aligned with implementing mechanisms.

PRS: The Partner Reporting System refers to the system used by the PEPFAR M&E contractor to gather information from IPs from the service delivery points they support. IPs collect this data using existing HMIS data collection tools, where possible, or through partner-specific data collection and reporting tools. Partners gather this data only for indicators and disaggregates that cannot be obtained from national reporting tools. The data collected is reported to the HIBRID database hosted by the PEPFAR M&E contractor.

“The Three-Ones Framework”: A UNAIDS Global declaration that signatory countries and development partners adopted for improving coordination of HIV/AIDS interventions. The framework calls for: “one agreed-upon HIV/AIDS Action Framework that provides the basis for coordinating the work of all partners; one National AIDS Coordinating Authority, with a broad-based multi-sectoral mandate; and one agreed-upon country-level Monitoring and Evaluation System.”9

9 UNAIDS. (2004). “Three Ones” key principles: “Coordination of National Responses to HIV/AIDS” Guiding principles for national authorities and their partners. http://data.unaids.org/UNA-docs/Three- Ones_KeyPrinciples_en.pdf

Scope of SOP This standard operating procedure describes the processes used for generating PEPFAR annual and semi-annual program results for a subset of the current MER indicators from data that are collected, reported and stored in the national eHMIS/DHIS2 system. This SOP will be updated to include protocols for collecting, validating, and reporting on MER indicators from the OVC MIS, CPHL, and UBTS sources once procedures for these national data systems have been finalized.

Use of the SOP This SOP can be used as a training manual or as a stand-alone reference and user’s guide for those responsible for aggregating PEPFAR reports in Uganda.

Target Audience of SOP

Staff from USG PEPFAR Coordinating Office (PCO), USG Strategic Information (SI) team members, activity managers, and other technical officers; and implementing partners.

MoH/MoGLSD M&E staff.

I. Responsibilities for collection, cleaning, and reporting DHIS2 Data

The MoH has elaborated procedure manuals for data collection and reporting at community, health facility, health sub-district and district levels. This section outlines the responsibilities of the current actors in the HMIS system towards generating PEPFAR reports. The MoH/RC has detailed the district-level processes and responsible entities in the HMIS procedure manuals including the schedule of reporting – see annex document

8_TABLE N1a_Health Facility Reporting schedule. Table 3 also present the HMIS health facility reports and their due dates.

1. Facility Staff

Health facility staff members are required to maintain standardized paper-based registers that contain patient-level information on diagnoses and treatment. These registers are the primary data collection tools of the Health Management Information

System (HMIS). It is pertinent that all services provided are documented in the registers, including those provided across multiple departments and during outreach.

For reporting, health facility staff responsible for data management should aggregate and transcribe data from the patient registers into the HMIS summary reporting tools. It is the responsibility of the health facility staff to ensure that all relevant HIV/AIDS data are entered into the HMIS summary forms and that these forms are delivered to the district on-time following procedures outlined in the HMIS Standards Manual for health units10. The report should be verified and counter signed by the In-Charge of the facility.

Table 3: Health Facility Reports Submitted to the District for HMIS/DHIS2

Report Title Name Period Date Due HMIS 033 Surveillance Report Weekly Every Monday of the following week

HMIS 105 Outpatient Report Monthly 7th of following month

HMIS 009* Outpatient Addendum Report Monthly 7th of following month

HMIS 108 Inpatient Report Monthly 7th of following month

HMIS 106a Quarterly Report 1st Q (July−Sep) 2nd Q (Oct−Dec) 3rd Q (Jan−Mar) 4th Q (Apr−June)

7th of month following end of quarter

HMIS 101 Physical Inventory

Annual (July−June)

7th August

HMIS 102 Equipment Inventory 7th August

HMIS 103 Staff Listing 7th August

- HSSIP Indicators 7th August

HMIS 107 Annual Report 7th August

HMIS 109 Population Report 7th August

* Revised HMIS tools will incorporate the data elements from HMIS 009 Monthly Outpatient Addendum Report into HMIS 105.

2. District-level M&E Staff

The district biostatistician and other district level and health sub-district (HSD) M&E staff members should review the HMIS reports submitted by health facilities within the district and evaluate them in accordance with the standards manual. It is the responsibility of the district-level/ HSD M&E staff (district biostatistician and/or district

HMIS focal person, in collaboration with the district program technical staff, to support the facility staff in ensuring that the data submitted by the health facility is accurate as per the HMIS procedure manuals for health facilities. However, depending on the capacity of the district M&E staff, the data cleaning processes may not be sufficient and additional validation and data cleaning is required at the national level. The district-level

M&E staff are required to make the actual submission of data on behalf of the health facilities into the web-based DHIS2 database, as per the guidance in the HMIS Standards

Manual for Districts/Health Sub-districts11. Cleaned data sets should be finalized and in the DHIS2 by the 28th of the month following the end of the reporting period.

10 Ministry of Health Resource Center. “The Health Management Information System (Volume 1): Health Unit and Community Procedure Manual.” October 2014.

11 Ministry of Health Resource Center. “The Health Management Information System (Volume 3): District/HSD Procedure Manual.” Oct 2014.

3. PEPFAR –supported Implementing Partners (IPs)

It is the responsibility of all PEPFAR IPs working at the district level to support the district-level/HSD M&E and facility M&E staff in ensuring that quality data is submitted into the DHIS2 as per the HMIS procedure manuals. Implementing partners will be evaluated based on the data that facilities submit, and it is in their interest to make sure that information is of quality. IPs are the primary channel through which USG agencies and the M&E contractor will address data quality issues arising from facilities and districts supported by the IPs.

4. National-level MoH Staff

The national-level Ministry of Health Resource Center (MoH RC) is responsible for carrying out activities that support the production, management and access of timely and quality information of the Health Sector. The MoH RC staff are responsible for coordinating the creation, maintenance, and distribution of HMIS system reporting tools. This includes creation of HMIS procedure manuals and the training of national, district, and facility M&E staff on the use of reporting tools. In December of 2014, the

MoH RC released a new set of Health Unit Reporting Forms (HMIS forms) that are to be used for reports generated after April 1, 2015.12

The MoH RC is also the custodian of the Ugandan DHIS2 database. It is responsible for storing the data on its server and for maintaining a web platform for data submission and access. The MoH RC also controls user access to the DHIS2 database and provides credentials to users.

While the Resource Center is the custodian of the reporting tools and hosts the DHIS2 data server, it is the responsibility of the individual technical programs within the MoH to produce the relevant patient registers and to validate the data reported through

DHIS2. The primary technical program responsible for HIV/AIDS control and management in the country is the AIDS Control Program (ACP) of the MoH. Working with USG and their partners, the ACP has established standard data quality checks and cleaning procedures that other stakeholders, including the PEPFAR M&E contractor, should use when validating DHIS2 data. These procedures should be reviewed periodically and updated if necessary.

12 At the time of writing (May 2015), the transition to the new HMIS tools has not been completed and facilities continue to use the old reporting forms. As the new forms contain additional fields that are needed for certain MER indicators, the delayed transition impedes the use of DHIS2 for PEPFAR reporting in FY2015. The remainder of this SOP refers to the old tools.

The PEPFAR M&E contractor shares cleaned HIV/AIDS data from DHIS2 with the ACP and the MoH RC for strategic information purposes. These departments archive the cleaned

HIV/AIDS datasets.

MoH RC also validates the lists of health facilities operating in each district. In this effort, MoH RC should receive support from the PEPFAR M&E contractor, which obtains information on facilities through IPs. When changes occur in health facilities that need to be reflected in DHIS2, the MoH first confirms the changes with the district health officers (DHOs). Then working together, the PEPFAR M&E contractor should support the MoH RC to maintain the list of facilities included in the DHIS2 database on a continuous basis.

5. PEPFAR M&E Contractor

The PEPFAR M&E contractor currently has the responsibility for a) extracting HIV/AIDS data from DHIS2, b) for undertaking data cleaning and validation using standardized tools, c) following up with implementing partners and other stakeholders to address data quality issues, d) aggregating the MER indicators derived from DHIS2 (and other sources), and e) submission of the data to USG’s DATIM system. In addition, the PEPFAR

M&E contractor collects data that cannot be obtained from the national reporting systems, through the Partner Reporting System/HIBRID database.

6. PEPFAR Coordinating Office (PCO) Staff

The PCO provides critical coordination across the six USG agencies involved in HIV/AIDS programming in Uganda. It is responsible for developing the PEPFAR Country Operation

Plan and for submission of PEPFAR reports to the Office of the (US) Global AIDS

Coordinator and Health Diplomacy (OGAC). It is responsible for facilitating access to the

OGAC hosted DATIM database that aggregates PEPFAR program data and gives final approval over the content of the COP, annual and semi-annual PEPFAR reports that are submitted to OGAC.

7. PEPFAR Technical Working Groups (TWG)

PEPFAR Technical Working Groups (TWGs) are clusters of specialized staff drawn from the six USG agencies supporting HIV/AIDS programs in Uganda. The TWGs devoted to

HIV/AIDS technical areas (e.g. PMTCT) provide quality assurance oversight for PEPFAR reports and provide guidance to the IPs on data quality issues.

An additional SI TWG is composed of technical and M&E staff across USG agencies. It works with the PCO and the USG M&E contractor to determine the set of MER indicators that apply in Uganda and the data collection tools that will generate them.

They also set standards and troubleshoot issues related to PEPFAR reporting. During

FY2015, the SI TWG had the responsibility for de-duplicating facilities that are supported by more than one IP in the same technical area. This is necessary for proper attribution of sites (e.g. health facilities) to IPs and programming information.13

13 It is expected that by FY2016 the process of ‘rationalization’ will prevent multiple IPs from providing the same support to the same sites, simplifying the process of attribution.

II. Procedures for Generating PEPFAR Indicators from DHIS2 Data

1. Data Collection, Aggregation, and Reporting at Health Facility

Health facilities should collect patient information and maintain service delivery records using standardized, paper-based registers developed by the MoH RC. The following section outlines how the facility-level register data is reported to the district level.

1. At the end of each reporting period, facility staff should transcribe data from the patient registers to the reporting forms. Depending on the resources of the facility, this task may be performed by frontline care staff, a dedicated M&E officer, or by the in-charge of the facility. Regardless, the in-charge of the facility is required to certify and approve the reports.

2. Facility reports should then be submitted to the district and the health sub-district by the 7th of the month following the end of the reporting period. It is the responsibility of health facility and district M&E staff to ensure that all relevant HIV/AIDS data are submitted to the DHIS2 on time and in the format outlined in the HMIS procedure manual for health facilities.

The HMIS reporting tools from which HIV/AIDS data is obtained include14. Samples of these reports are included in the annex folder “9_HMIS Reporting Tools”:

HMIS 105 Outpatient Monthly Report

HMIS 009 Outpatient Monthly Addendum Report

HMIS 106a Outpatient Quarterly Report

2. Data Review and Cleaning at District/Health Sub-District Levels

Between the 7th and 15th of the month following the end of the reporting period, the district level M&E staff (district biostatistician) should review the reports from health facilities and then enter the information into DHIS2. The DHIS2 database serves as the repository for all

HMIS data reported from the community and health facilities in the country.

It is the responsibility of all other partners and stakeholders operating at district level to support the district and facility M&E staff in ensuring that quality data is submitted by the district into the DHIS2 as per the procedure manuals. The procedures and timelines for data

14 This refers to the 2010-2015 Reporting Tools that were intended to be replaced April 2015, but which continue to be in use as of May 2015.

collection and reporting at district level are outlined in the MoH’s HMIS Procedure Manual for Districts/Health Sub-Districts, last revised in October 2014.

2.1. Review of HMIS Data Forms Submitted to the District

District/HSD M&E staff are responsible for receiving reports from health facilities. As they do so, they should assess the timeliness of report submission (ideal is on or before the 7th of the month following the end of reporting period for monthly and quarterly reports), as well as the completeness and validity of reported data. To assist in monitoring the timeliness of report submission, the district/HSD is provided with Table N1a “Health Facility Record of

Reporting” to track reporting progress in the HMIS Manual – see annex 8.

In addition, the District/HSD should review the reports submitted by health facilities prior to entering the data into DHIS2. This is commonly performed by the district biostatistician and other support staff from the district and HSD. The district/HSD staff are trained to look for missing data elements, disaggregation counts that exceed aggregated counts, anomalously large or small numbers, data inconsistent with previous reports, data entered into incorrect cells, and transcription errors. When data quality or completeness issues are identified, the district level staff should get in touch with the in-charge of health facility for clarification and/or correction. Being closer to the health facilities, the HSD frequently assist the district level M&E staff in communicating with the health facilities.

As the data is entered, hard copies are printed and filed for backup record-keeping. Also of relevance, as the reports are entered for health facilities, the districts/HSD must also aggregate the data from facilities into district/HSD level reports submitted to the MoH RC.

Between the 7th and 15th of the month, all stakeholders with an interest in HMIS data especially the USG IPs, should work with and support the district biostatistician, District

Health Officer (DHO), and other M&E staff to ensure submission of relevant data from facilities into the DHIS2. They should also support the DHO to extract the data from the

DHIS2 for purposes of data cleaning and for providing feedback to health facilities in order to improve data quality, so that by the 28th of the month following the reporting period, all data within the DHIS2 is of good quality and ready to be extracted for PEPFAR reporting.

2.2. Running Completeness Reports and Validation Checks within DHIS2

DHIS2 facilitates the process of data validation by offering tools and support mechanisms for data validation at the time of entry.15 First, DHIS2 will not accept invalid data (e.g.

15 DHIS 2 End-user Manual. V2.19. https://www.dhis2.org/doc/snapshot/en/end-user/dhis2_end_user_manual.pdf https://www.dhis2.org/doc/snapshot/en/end-user/dhis2_end_user_manual.pdf characters in place of a number, negative numbers). Secondly, data validation checks can be run as the report data is being entered (see Figure 2).

Figure 2: Running Validation checks within the DHIS2

Thirdly, DHIS2 allows comparison of the entered value for the current period with values entered for previous periods for the same facility. These tools assist the district M&E staff to flag transcription errors made either at the time of reporting or when entering the reported values into DHIS2. However, data validation flags do not prevent submission of reports.16

Despite controls, reports can be entered with invalid data.

Within DHIS2, districts can track reporting completeness from facilities as the reports are being entered. Reporting completeness is based on the number of health facilities within the district and the number of reports they are expected to submit during the current reporting period. The completeness checks are incumbent on the district and the MoH RC, with support of the PEPFAR M&E contractor, keeping the list of health facilities in DHIS2 current. See Section 3.2 for steps in checking data completeness within the DHIS2.

2.3. Data Extraction, Applying Data Quality Checks and Providing Feedback on Data

Quality

As DHIS2 is the repository for health facility data, districts/HSDs may return to DHIS2 to extract data in an electronic format for data validation and analysis. To extract data from

DHIS2, the user should refer to section 3.3 of this SOP for guidance on the pivot table mode

16 In some cases, data producing a validation flag, such as the number of women receiving 4th ANC exceeding the number receiving 1st ANC, could be valid data. The district M&E staff are expected to vefify these cases with the in-charge of the health facility.

of extraction. Once extracted, the same data validation checks established by the MoH program working groups for countrywide HIV/AIDS data cleaning should be applied.

Districts/HSD are expected to provide feedback to health facilities on a regular basis on both data quality and health care performance. In relation to data quality, districts routinely conduct data quality assessments with health facilities, wherein they compare register data with that submitted in HMIS reports and assess the processes used for data collection and reporting.

In addition, districts should prepare quarterly reports and disseminate the results of data to district-level stakeholders. These may include the members of Local Council V, the district health office (DHO), the Chief Administrative Officer (CAO), NGOs supporting the district and HSD staff. Under Uganda’s decentralized health model, the reports can be used for assessing the performance of health personnel hired by the district and to improve the allocation of supplies among facilities within the district. The meeting can also be used to validate reported data.

3. Data Review and Cleaning at National Level

Between the 15th and the 28th of the month following the end of a reporting period, national-level stakeholders (MoH, PEPFAR M&E contractor, Implementing Partners, and other development partners) should extract all relevant HIV/AIDS datasets from DHIS 2 and check for data quality and completeness. The following steps are intended to allow the user of this SOP to access, extract, and validate DHIS2 data for this purpose.

3.1. Obtaining Access to the DHIS2 Database

To extract data from the DHIS2 database, one will need to have user credentials that provide “Data User” access rights. The DHIS2 database is hosted by the MoH RC. Those seeking to access and extract data from DHIS2 must first go through the MoH RC to obtain login credentials. Implementing partners should coordinate their access requests with their funding agencies. The form used for requesting access to DHIS2 is included in the annex document titled “5_Ministry of Health Data Access Guideline.pdf”.

3.2. Checking for Completeness of Reporting within DHIS2

The DHIS2 system allows users to see the number of reports submitted from health facilities at the district level compared to the number of expected reports. However, this completeness check in the DHIS2 system does not ensure that the expected data elements within the reports are complete; it only indicates that the district has affirmed submission of a specific report. Nonetheless, checking completeness through the DHIS 2 can be a start to identifying districts that may require additional support for timely reporting and to know when is best to pull data for cleaning and reporting.

Follow these steps to run a data completeness report in DHIS2:

1. Log into DHIS2 http://hmis2.health.go.ug/ using your assigned credentials.

Figure 3: Uganda eHMIS/DHIS2 Login Screen:

2. Navigate to “Apps” and select “Reports” from the drop-down menu.

Figure 4: Selecting Reports from the Applications Menu http://hmis2.health.go.ug/

3. From the navigation tab on the left, select “Reporting Rate Summary.”

4. In the menu that appears, ensure that the radio button for “Based on complete data set registrations” is selected. The alternate option, “Based on compulsory data elements,” is not set up for use. (Figure 5)

5. In the open menu for “Organizational Unit,” select a facility, a sub-county, a district, a region, or the whole of Uganda for which a data completeness report is required.

6. Select a dataset and an appropriate data period. If the data period selected is shorter than the reporting period of the report (e.g. weekly for the monthly reported HMIS 105), no information will appear.

7. Select “get report” to review and/or download the report-level completeness data.

Figure 5: Reporting Rate Summary Tool in DHIS2

The user has selected completeness of reporting for the monthly HMIS 105 (Section 1 – OPD) for Mar 2015 for facilities within Masaka district.

The resulting reporting rate summary for Masaka District for March 2015 HMIS 105:1 is shown in Figure 6.

Figure 6: Sample Reporting Rate Summary in DHIS2:

Once the reporting rate summary is generated, it can be downloaded as a PDF, CSV, or Excel file. Under “Show more options”, the user can disaggregate report completeness based on a number of facility characteristics (functionality, ownership, authority, facility type, accreditation, and ARV supply chain).

However, to check the completeness of data elements within reports, the data must be extracted from DHIS2 and data completeness checks applied externally. Section 3.3 discusses data extraction and Section 3.5 addresses completeness and validation checks.

3.3. Data Extraction from DHIS2 using the Pivot Table Tool

Extracting data from DHIS2 is done on a routine basis, to provide status updates on reporting by districts and sites as well as for routine data cleaning. There are two primary methods of extracting data from DHIS2: the “Pivot Table” tool, which is discussed below, and “Web API” described in section 3.4.

The process of generating data from the DHIS2 system is aided by use of the Pivot Table tool. Pre-set table formats for extraction of data for PEPFAR reporting have been created and saved to “Favorites.” Using these pre-set formats, data extraction only requires selecting the appropriate format from favorites, then changing the reporting period, and, finally, the organizational unit(s) as required. Users can also specify the data elements they wish or create and save their own “favorites.” Below are the steps for data extraction:

1. Log into DHIS2 http://hmis2.health.go.ug/ using your assigned credentials

2. Under the “Apps” tab click on “Pivot Table” to access the pivot table module http://hmis2.health.go.ug/

Figure 7: Selecting Pivot Table from the Applications Menu:

3. In the Pivot table window at the right, click on the “Favorites” button. (Figure 8)

Type “HIV Data Cleaning” in the search window to find a list of pre-generated results table formats. The five (5) most useful tables for PEPFAR reporting are:

a. HIV Data Cleaning: HMIS 105 Facility HCT SMC Stock outs – Monthly

b. HIV Data Cleaning: HMIS 105 Facility PMTCT EID – Monthly

c. HIV Data Cleaning: HMIS 105:6 Facility Laboratory Tests – Monthly

d. HIV Data Cleaning: HMIS 106a Facility ART – Quarterly

e. HIV Data Cleaning: HMIS 106a Facility HIV-TB – Quarterly

Figure 8: ‘Favorites’ within the Pivot Table

4. Select the data period and organizational unit desired for the report

Note: When using “Favorites,” a default period and organizational unit will appear.

To change from the defaults, click on “Periods” and add the period you want and click on “Organizational Units” and change to the desired district, sub-county-level unit, or facility. CTRL+Click can be used for multiple selections.

Figure 9: Selecting the Period for a Pivot Table

Figure 10: Selecting Organizational Units for a Pivot Table

5. After selecting the required reporting period and organization unit(s) click on

Update to refresh the results table. Results should display with the reporting period identified. If this is not the case, use the Layout tab to ensure that “Periods” is included as a row variable in the table layout and click on Update. (Figure 11)

Figure 11: Table Layout

6. To improve the presentation, click on “Options” and unselect all of the subtotals.

Also, deselect “hide empty rows” and ensure that “show hierarchy” is selected and that the “digit group separator” is set to “none”. (Figure 12)

Figure 12: Table Options Selections

7. Once you have the desired data for the district and program area in the results panel, you will need to copy them into a new Excel worksheet to run standard data cleaning validation rules. To copy and paste the results:

a) Select all results by, clicking anywhere in the table and Pressing CTRL + A

b) Copy the selected results: Press CTRL + C

c) Open a new the Excel workbook

d) Click in Cell A2 and click CTRL + V to paste the copied results into the Excel workbook

Alternatively, pivot table data can be downloaded directly to CSV, Excel, and other table data formats using the “Download” tab as shown in Figure 13.

Figure 13: Download Options

Steps 3-7 are repeated for each dataset, period, and organization unit selected. The Pivot table provides data dynamically by allowing dimensional data extraction and analysis.

However, this process has one major limitation: Extraction of large datasets is restricted.

The default number of allowed records/cells per dataset extracted is ~50,000 cells. This constrains the number of indicators and sites that one can download/extract per dataset at once. If one is frequently going to extract large datasets, it is possible to do so quickly through Web API (Section 3.4).

For most datasets, one can extract data for multiple districts (an average of six), however for some program areas that have many fields, like PMTCT, the dataset limits may be reached and extraction will need to take place one district at a time. For some districts like

Kampala, datasets have to be extracted by divisions and for Makindye and Nakawa

Divisions, one has to extract them by selecting a few health facilities at a time. It is advised to restrict the number of organizational units (districts, counties, facilities) rather than to limit the time period. This is done to ensure that the output remains in the form desired with multiple records for the facility representing different periods.

As data extracts are being transferred to Excel files, the Excel files should be made to contain all the data of a given type for a given district and then saved into a file directory organized by the region of Uganda. Such a file system will allow for district-level reporting to be monitored while also facilitating the aggregation of data from districts.

3.4. Data Extraction from DHIS2 using Web API

Extraction of large datasets is faciliated with Web API, however the process is more technologically complex than pivot table extracts. To use the Web API to extract data, please refer to the attachment “7_Guidelines for extracting DHIS2 datasets using Web

API.docx”.

3.5. Linking DHIS2 to PEPFAR IP Program Data

National reporting tools do not identify PEPFAR programming information. Therefore, before DHIS2 data can be used for aggregating PEPFAR indicators, it must be linked using the VLookup function to a dataset that identifies each site with a) the implementing partner providing support, b) the implementing mechanism, c) the USG funding agency, and d) the technical area supported by a PEPFAR-funded mechanism and e) the support type (DSD/TA-

SDI) provided by the IP for the site. IPs submit information on the sites they serve through the Data Accountability Platform for Type of Support (DAPTS) module of the HIBRID database, which is hosted by the PEPFAR M&E contractor. The information is validated by the PCO and the PEPFAR TWGs. Maintaining this database is a necessary component of the

PEPFAR reporting process.

As some facilities receive support in the same program areas from more than one IP in

FY2015, it has been necessary to rationalize suppport and assign only one IP/IM per service area.17 In addition, USG/Uganda has directed that the definition of direct service delivery

(DSD) and Technical Assistance for Service Delivery Improvement (TA-SDI) in the country should be based on the ownership status (public/private) of the site. The PCO, SI TWG and program area TWGs will guide and clarify which classification system for DSD/TA-SDI is to be applied and which facilities are exceptions to the general rule. Figure 14 depicts how the

17 It is expected that in FY2016, duplication will no longer be an issue due to the USG-led process of rationalization of services at the district level.

different identifiers from DHIS2 and PEPFAR DAPTS datasets are linked. It also makes reference to the DATIM and iPSL IDs that will be described in Section 5.

The PEPFAR M&E contractor should assist the MoH to update the national list of health facilities at least before each reporting period, using information provided by the implementing partners and confirmed at the district level. This list is updated into DHIS2 by the MoH RC, wherein the DHIS2 system assigns the sites a database identifier that is unique within the DHIS2 system only.

While assisting the MoH to update DHIS2, the PEPFAR M&E contractor also works with USG

OGAC to update the DATIM database for Uganda. The DATIM system assigns another database identifer code to the facility. Lastly, the HIBRID system, which contains the data that cannot be derived from national systems, also has a system identifier for each facility/service outlet. These different identifiers for the facilities have to be harmonized.

As of yet, there is no unique ID code for all health facilities in Uganda. Although PEPFAR has created unique iPSL IDs for Uganda, the methodology has not been adopted by the MoH for use. The identifying information in DHIS2 (district, HSD/sub-county, facility name, facility level) is intended to be specific. However, due to formatting and spelling variations and errors, this information does not yet match facilities perfectly between data sets. There are also a few private facilities with identical names and levels within the same sub-county-level jurisdiction. To link facilities/service outlets across these three databases, a matching key has been created by the PEPFAR M&E contractor. Users of this SOP seeking to link DHIS2 data to other databases will need to either obtain this key from the PEPFAR M&E partner or they will need to generate one through a thorough matching process.

Figure 14: Linking PEPFAR Programming Information to DHIS2 Datasets and DATIM IDs

Once the matching key has been created, it is possible to link the data extracts from DHIS2 to PEPFAR program data as well as to DATIM and to their PEPFAR-issued (integrated

PEPFAR-supported site list (iPSL) IDs. It is recommended that this is done using Vlookup…