Project Overview

The Kenya Strategic Information (Kenya SI) project is a CDC-funded mechanism primarily focusing on supporting the Ministry of Health by providing technical assistance in HIV/AIDS strategic information through surveillance in Kenya. Earlier iterations of the project also included monitoring and evaluation (M&E) activities. Informatics activities under this project are in support of the overall project goal.

Areas of Support

Through Kenya SI, our informatics team works and has worked across the following activities:

• Case-Based Surveillance — prioritizing longitudinal follow-up of individual HIV cases 
• Recency Surveillance — focusing on detection of recent HIV cases among newly diagnosed HIV-positive persons aged >=15 years 
• Mortuary and Hospital-Based Surveillance of HIV-associated Mortality — centered on HIV-associated mortality in key hospitals
• General surveillance activities

Past activities

• DREAMS Monitoring and Evaluation — prioritizing prevention programming for adolescent girls and young women 
• Partner Progress & Performance Monitoring (3PM) — tracking progress against targets for CDC-funded partners
• M&E Tools Tracking Platform — tracking distribution and utilization of government health reporting registers and tools
• e-Learning Platform — facilitating online continuous learning for health practitioners on HIV surveillance
• Early Infant Diagnosis Drug Resistance (EID DR) — HIV drug resistance for newly HIV diagnosed infants
• Integrated Key Populations Surveillance (IKePS) — HIV monitoring among key populations served at Drop in Centers
• Key Population Size Estimate (KPSE) — estimating population size and mapping hotspots for Key populations
• DREAMS Recency — survey for recent HIV cases among females aged 15 years or more and less than 25 years
• Lost to Follow-up — survey on magnitude of and reasons for LTFU among patients in HIV care

Accomplishments

Case-based Surveillance Database, Dashboards and Reports

• Developed the case-based surveillance data warehouse to clean, merge and deduplicate individual HIV-patient level data from the national data warehouse, national viral load database, paper sites and recency data to structure data along selected sentinel events and enable facilities, counties and national levels to focus on the treatment journey of individual HIV patients
• Developed an ETL process with advanced SQL-based scripts producing dimensions and fact tables to compute the status of each individual in the database, with one final fact table summarizing all sentinel events of a patient
• Reports on key sentinel events, which include HIV diagnosis, start on treatment, any regimen changes, viral load status, death and lost to follow up, as well as some MER indicators, including TX_CURR
• Compute and derive the individual outcome of over 1.1m people in Kenya — suppressed, not suppressed, lost, not initiated on ART, or died
• Developed comprehensive interactive dashboards and reports to disclose the information in the CBS database, allowing end users to look at aspects of the epidemic through different angles and perspectives
• Developed monthly Public Health Response reports based on the data in the CBS database, enabling counties and health facilities to follow up on pre-selected critical treatment events, for example, individuals with prolonged non-suppression of HIV

Data Insight Reports

Developed “Data Insight Reports”, which are based on additional advanced analysis of CBS data, providing insights in the epidemic and the data of the CBS database:

• Cause of Death from Case-based Surveillance Data: Analyzed quality of data and documented causes of death for persons in the HIV care and treatment program
• Quantification of Incomplete Linkage Between HIV Testing, Care and Treatment: Illustrate the disparity between individuals testing positive for HIV and those enrolled in HIV care and treatment services
• Incomplete HIV Case Reporting: Compared CBS with Kenya Health Information System (KHIS) to analyze new HIV diagnoses reported in facilities with electronic medical record systems (EMRs) through both case-based and aggregate reporting systems
• Missing Viral Load Data: Assessed the impact of merging viral load results from the lab database to the data collected in electronic medical records (EMRs) for Case-based Surveillance, focusing on data completeness and accuracy

DREAMS Service Delivery Monitoring and Progress Tracking

• Developed a web-based and mobile system for managing intervention services provided to young women enrolled in the DREAMS program
• Web-based system built on Django-Python for web access to facilitate registration of beneficiaries and capture of services/interventions delivered to beneficiaries
• Offline system built using ODK for mobile Android devices for facilities with less reliable Internet connectivity with periodic syncing of data to the web system
• Developed complex analytics to process data and develop key performance metrics, including progress towards targets, and delivery of the right package of intervention based on beneficiaries’ baseline characteristics and eligibility criteria
• Comprehensive PowerBI dashboard for reporting and analytics capabilities, including monitoring of service delivery, monitoring of progress towards targets, and determining whether the right beneficiaries were getting the right services
• Manages data for over 350,000 active beneficiaries and monitored the delivery of over 2 million health interventions provided to beneficiaries

3PM Reporting Platform

• Developed a system that allows partners to submit monthly reports with key PEPFAR and CDC indicators built on DHIS2
• Development of advanced analytic scripts using SQL for computing metrics not supported by the DHIS2 platform, such as performance over time and comparison with previous reporting periods
• Developed PowerBI dashboards to support additional analysis and reporting of key metrics

Recency Surveillance System

• Developed a system to support Recency Surveillance which aims to establish whether individuals recently diagnosed with HIV have a recent or long-term infection
• Collect demographic information, blood sample information and lab results from clients who newly tested HIV positive and voluntarily participated in the Recency study
• Built to incorporate five interoperable systems: an HTS system, a lab sample collection and tracking system, a lab results system and a analytics and visualization system to seamlessly exchange data
• Leveraging ODK for patient demographics and HIV testing status, OpenHIE for data exchange between the lab and Recency database, SQL server for storage and data analysis, and PowerBI for dashboard development, supporting seamless interoperability of the five systems 
• Advanced analytics to support correct implementation of the recency testing algorithm, which includes the ASANTE test, and the viral load test to confirm suspect recent individuals

Additional Accomplishments 

• Development of 3PM software system used to monitor progress of CDC partners performance for the various CDC funded initiatives
• Development of DREAMS software system for managing and reporting on the DREAMS program
• Development of a Recency data collection tool and dashboards for collection and reporting of information for clients who participate in the Recency study
• Development of HIV drug resistance web based system to automate extraction and submission of HIV DNA sequences from DBS samples taken from HIV infected infants from the lab systems to the Stanford HIV drug resistance database in order to find drug resistances of the HIV DNA sequence

Key Leadership

UCSF

• Principal Investigator: George Rutherford, MD, MA
• Leadership Team: Dr. Mary Mwangome, Wanjiru Waruiru, Dr. Peninah Masibo, Fitti Weissglas, Vincent Yahuma, Kevin King'utu