Malaria in supported health centres: data analysis, ACT quantification and free care

HC1: good utilisation, but 2,600 ACTs for 2,030 positives: 570 treatments without a positive result; 200 untested suspected cases explain only 200 treatments: about 370 treatments given to negative patients. A practice that does not comply with the recommendation to confirm every suspected case with a test before treatment; other causes of fever are not investigated.

HC2: compliant practice (almost all tested, ACTs ≈ positives). Lower utilisation (0.70): possible financial barrier (document 3).

HC3: the RDT stock-out explains the low testing rate (44%) and presumptive treatment: 790 positives + 1,400 untested suspected cases = 2,190, for 2,350 ACTs, i.e. still 160 treatments of negative patients. ACT consumption is inflated by the RDT stock-out: a supply failure, not only a practice issue.

HC4: highly suspicious profile: utilisation of 1.80 for a small catchment area, positivity of 95% (far above the other centres, even though some fevers have other causes), ACTs exactly equal to positives, 100% tested, "perfect" reports submitted before the end of the month. Hypotheses: fabricated or rounded data, diversion of ACTs and RDTs, or a population coming from other catchment areas. To be checked before drawing any conclusion.

Overall: 8,260 ACTs for 6,120 positives, i.e. about 2,140 more treatments than confirmed cases (1.35).

Limitations: (1) third-quarter consumption corresponds to the peak transmission of the rainy season; from October to December, demand generally falls: applying the same AMC risks overstock and expiries; (2) consumption is inflated by the treatment of negative patients (HC1, HC3) and possibly by false data or diversion (HC4): about 2,140 of the 8,260 treatments do not correspond to a confirmed case, i.e. 26%; (3) the AL 12 stock-out may have been offset by other presentations (splitting blister packs), which distorts the breakdown.

Correction: cross-check the consumption method with a morbidity-based estimate (expected confirmed cases per month and per weight band, data from previous years for seasonality); order in monthly tranches rather than all at once; secure RDTs as a priority, since their stock-out drives ACT overconsumption; revise after the verification of HC4's data.

Data quality: data quality audit in the 4 centres, starting with HC4: recount of consultation and laboratory registers against monthly reports, physical inventory of ACTs and RDTs compared with reported issues, verification of where patients come from. Responsible: health M&E manager, with the district management team. Indicator: discrepancy between registers and reports below 10%.

Supply: RDT safety stock, alert at 1 month of stock, weekly stock monitoring during the season. Responsible: pharmacist or health logistician. Indicator: number of stock-out days for RDTs and for each ACT presentation (target: 0).

Rational use: training and supportive supervision on the "test, treat, track" algorithm and the management of RDT-negative fevers (search for other causes, integrated management of childhood illness); monthly feedback to each centre. Responsible: clinical supervisor. Indicators: % of suspected cases tested (target ≥ 95%), ratio of ACTs to positives (target close to 1.0), % of negative fevers correctly managed (case file review).

Governance: monthly review meeting with the health delegation; if the HC4 audit confirms irregularities, reporting procedure in line with the anti-fraud policy and informing the donor.

Sample answer: "In the third quarter, the seasonal peak, the 4 supported centres carried out 15,510 new consultations and tested 7,890 of the 9,500 suspected malaria cases (83%); 6,120 cases were confirmed. A 5-week RDT stock-out at HC3 reduced the testing rate to 44% in that centre and led to presumptive treatment. In total, 8,260 ACT treatments were dispensed, i.e. 1.35 per confirmed case: some were given without confirmation. The data from one centre show inconsistencies and are currently being verified. Measures: RDT safety stock and weekly monitoring, supportive supervision on fever management, data quality audit. We will inform you of the conclusions of the verification. ACT quantification for the fourth quarter takes the seasonal decline into account."

Finding: 30% of the caregivers interviewed at HC2 (18 out of 60) paid for care that should be free; this is a breach of the agreement, a barrier to access for the poorest (possible link with HC2's lower utilisation, 0.70) and double funding, since the NGO already funds the medicines and the incentives. I cannot turn a blind eye.

But the problem is real: the guard and the cleaner are needed for the security and hygiene of the centre. I propose a sustainable solution: include these two posts in the project budget or have them covered by the centre's management committee under transparent rules, in discussion with the health delegation and the donor.

Measures: meeting with the head of HC2 and the health committee to stop the payments immediately; visible display of free care in the local language; accessible complaints mechanism; new exit interviews in one month; discussion with the district chief medical officer to recall the agreement, without confrontation. If the payments continue, application of the clauses of the agreement (suspension of the incentive), after escalation to the medical coordinator. I document the exchanges.