Setting:
MAUL was awarded a five-year United States (U.S.) President’s Emergency Plan for AIDS Relief (PEPFAR) grant in 2011, through the U.S. Department of Health & Human Services (HHS) under the U.S. Centers for Disease Control & Prevention (CDC) to implement the Procurement & Supply Chain Strengthening Project (PSSP). The main goal was to support procurement and logistics management services for HIV/AIDS‐related commodities – including ARVs, medicines for opportunistic infections (OIs), laboratory reagents, equipment, and consumables to private, not‐for‐profit (PNFP) HFs and strengthen institutional capacity for management of HIV/AIDS logistics services. Project funding increased from U.S. dollars (USD) 7.6 million in 2011-12, to USD 46 million in 2012-13 and USD 60 million in 2015-16. PSSP provides ARVs to over 250,000 patients in 216 participating HFs across 62 districts in Uganda in 2015-16.
Overtime, we implemented four major interventions to address these constraints at the HF level, including Mentorship and Technical Health Systems Strengthening (MaTHSS) supportive supervision model in 2011-12, the Field Report on Stock Tracker (FROST), Geographic Information Systems (GIS), and WhatsApp® Messenger in 2014 in an attempt to improve SCM indicators.
Interventions to address the gaps:
The Mentorship and Technical Health Systems Strengthening (MaTHSS) Supportive Supervision Model:
Starting in 2012 and implemented in 198 HFs, we combined on-site mentorship, training on Logistic Management Information Systems (LMIS) and SCM and monthly supportive supervision to achieve defined goals. Regional Field Support Officers (RFSOs) performed on-site mentorship of HF staff during monthly supportive supervision visits. Trainings were performed by a team comprised of supply chain technical officers, monitoring and evaluation, LMIS, and field operations teams.
HFs were assessed using the Supportive Supervision Monitoring Tool (SSMT) that scored them in six broad categories: 1) stock management, 2) product organization, 3) dispensary, lab, and stores management, 4) dispensing aids and tools, 5) ordering and reporting, and 6) expiry tracking. Each category had a maximum score possible and the categories were added up to give a maximum total score. Facilities were graded in each category and their scores added up to give a facility total score. Facilities were then categorized – based on their total score as a percentage of the maximum total score possible – as unranked (<50%), bronze ranking (50-69%), silver ranking (70-90%), and gold ranking (>90%). Direct feedback was provided to HF staff, and senior management to ensure sustainable improvement in overall performance. Facilities could graduate to a new category if they implemented recommended actions and scored in a higher category on two-three consecutive visits.
Facilities were also assessed using three internal performance monitoring indicators that recorded the number of HF personnel trained in LMIS and SCM, and the number of supportive supervision visits monthly, quarterly, and annually.
Field Report on Stock Tracker (FROST):
In 2014, we developed a field-based tool, FROST, to monitor stock levels at 191 HFs receiving ARVs and laboratory supplies. This field-based tool enabled RFSOs and logisticians to manage commodities by visualizing ARV stock levels at all HFs.
HF consumption rates and physical counts for HIV commodities were updated into FROST on a monthly basis. The tool generated available months of stock, thereby facilitating commodity decision-making for borrowing or lending of HIV commodities to and from nearby HFs.
Geographic Information Systems (GIS):
In 2014, we utilized a four-stage approach of linking LMIS to GIS. Logistics staff members were trained in GIS spatial and temporal analyses. GIS coordinates were then collected from 216 HFs. Confirmation of coordinates was done using mobile-phone reconnaissance and merged into a central-level LMIS database, to form cross-walk tables. GIS data were used together with FROST data to spatially visualize stock-on-hand, stock outs, and other important stock indicators.
WhatsApp®:
For each HF with communication problems, we identified a health worker with a smart phone. WhatsApp® Messenger was installed on the smart phone and health workers trained on how to take pictures of order reports to be forwarded to the warehouse. The image was transcribed into an electronic version and processed for resupply. Acknowledgement of receipt for all orders received at the warehouse was done immediately by warehouse staff.
Performance assessment:
All the interventions were assessed following a pre-specified Performance Monitoring Plan, a set of performance indicators developed in-house to evaluate our achievement of specified goals. These indicators are broken down into seven major categories: 1) product selection, 2) forecasting and quantification, 3) procurement, 4) storage and warehousing, 5) order processing, 6) inventory management and facility reporting, and 7) supervision and training.
Variables and definitions:
For purposes of assessing performance, the following six indicators were used;
A - Percentage of facilities with adequate stock levels of indicator commodities to ensure near-term continuous product availability (total number of HFs reporting stock levels of indicator commodities within minimum/maximum range divided by total number of HFs).
B - Inventory accuracy for on-hand inventory at the end of the reporting period (number of HFs whose physical count tallies the stock card record divided by total number of HFs visited).
C - Average percentage stock in levels at the facilities (total number of HFs that did not report a stock out for any indicator commodity divided by the total number of HFs that reported in a cycle).
D - Order fill rate in terms of customer receipts (total product ordered minus total product received by HFs divided by total number of product ordered by HFs). We also assessed timeliness, completeness, and efficiency of the distribution system.
E - Facility reporting rates (number of HFs that submitted complete LMIS reports according to the defined reporting schedule divided by the total number of HFs reporting).
F - Number of HF personnel trained in LMIS or SCM and number of supportive supervision visits conducted in HFs.
Data management and statistical analysis:
For this analysis, data were aggregated from several data sources including Excel spreadsheets and electronic databases developed in-house. SSMT data entry was performed using Epidata 3.1[14]. Data cleaning and consistency checks were performed on aggregated data to ensure correctness.
For HF characteristics, categorical variables were described using frequency statistics and proportions and for continuous variables we reported median and interquartile range (IQR). For changes in performance monitoring indicators we tested for the significance in the change in scores using the test for difference in proportions. Kaplan Meier estimates were computed for time to silver and gold ranking and univariate and multivariate Cox proportional hazards models were computed for time to gold ranking. All analyses were done using Stata 12.1[15].
Ethics approval and consent to participate:
All participating HFs, the Ministry of Health, and CDC Uganda co-authors provided their concurrences to publish this information. Ethical approval was obtained from Mildmay Uganda Research and Ethics Committee (#REC REF 0501-2017). This activity was reviewed in accordance with CDC human research protection procedures and determined to be research, but CDC investigators did not interact with human subjects or have access to identifiable data or specimens for research purposes.