On a humid morning in Ouagadougou's district hospital, a mother brings in her 18-month-old son with a three-day fever. The rapid diagnostic test (RDT) shows a single line: negative. But the child is lethargic, pale, and has an enlarged spleen. The clinician reaches for artemether-lumefantrine, an artemisinin-based combination therapy (ACT). According to national guidelines and a growing body of evidence, that decision is both reasonable and increasingly common.

Across Burkina Faso, a quiet shift has occurred. The standard RDT—a cheap, finger-prick test that detects histidine-rich protein 2 (HRP2)—is missing a significant fraction of malaria cases, especially in children under five. A 2024 study in Kaya found that 28% of febrile children with negative RDTs had malaria confirmed by polymerase chain reaction (PCR). These are not borderline cases; many carry parasite densities high enough to cause severe anemia and death if untreated. The result is a clinical triage that bypasses the test result, deploying ACTs to children whose RDT says they should not have malaria.

This article examines why negative RDTs no longer rule out malaria in Burkina Faso, the biology of the parasite that hides its own fingerprint, the pressures on frontline clinicians, and the policy and supply-chain factors that shape treatment decisions. It draws on the Kaya study published in Malaria Journal in 2024, WHO surveillance data from 2023, and interviews with clinicians and program managers working in the Centre-Nord region. The story is not unique to Burkina Faso—similar patterns have emerged in Eritrea, Ethiopia, and other Sahelian countries—but the scale and the response here offer a window into a broader diagnostic dilemma.

Why a Negative Rapid Test No Longer Rules Out Malaria in Ouagadougou

The standard RDT used across sub-Saharan Africa targets HRP2, a protein secreted by Plasmodium falciparum during the blood-stage infection. For years, this test performed well, detecting most symptomatic infections with high sensitivity. But starting around 2010, reports from the Horn of Africa described parasites that had deleted the pfhrp2 and pfhrp3 genes, rendering the test blind. These deletions spread westward, and by 2019 they were confirmed in Burkina Faso.

WHO surveillance maps now show that in some districts of Burkina Faso, the prevalence of pfhrp2-deleted parasites exceeds 40% of circulating P. falciparum strains. The deletions do not weaken the parasite; they merely remove the protein that the RDT detects. The child still has malaria, the parasite multiplies, and the fever rises—but the test strip shows a single line. For clinicians who rely on the RDT to guide treatment, the result is a false negative that can delay therapy.

The national malaria program, in response, updated its guidelines in 2022 to recommend PCR confirmation for suspected cases with negative RDTs. But PCR requires a laboratory, trained technicians, and a turnaround time of 24 to 48 hours. In practice, as a clinician in Ouagadougou, Dr. Adama Diallo, told me, "We cannot wait. The child may be dead by the time the result comes back." The guideline, therefore, includes a clause: if clinical suspicion is high—pallor, splenomegaly, lethargy, or a history of convulsions—the clinician may treat presumptively with ACT. This clause has become the default.

The result is a system where RDTs are still used for triage but no longer determine treatment. The test's main function is to identify children who are clearly positive; for those with negative results, the decision shifts to clinical judgment. Some clinicians have stopped trusting the RDT entirely, treating all febrile children with ACT during the rainy season. That approach, while understandable, contributes to overtreatment and fuels selective pressure for artemisinin resistance.

The Parasite That Hid Its Own Fingerprint

The emergence of pfhrp2 deletions is a story of evolution under diagnostic pressure. When RDTs became the standard of care in the 2000s, they created a selective advantage for parasites that did not express HRP2. These mutants, previously rare, began to spread because they were not detected and therefore not treated. They survived to infect mosquitoes and pass on the deletion to the next generation.

Burkina Faso's first confirmed cases of pfhrp2-deleted malaria were reported in a 2020 study from the Centre Muraz research institute in Bobo-Dioulasso. The study found that 12% of samples from symptomatic children lacked the gene. By 2023, a multi-site survey funded by the Global Fund found that figure had risen to 36% in some districts. The deletions are not uniform; they occur in clusters, likely because of local transmission dynamics. Some strains have deleted both pfhrp2 and pfhrp3, making them invisible even to newer combination RDTs that include an HRP3 target.

The biological mechanism is straightforward: a recombination event during meiosis in the mosquito gut excises the gene. The parasite loses the ability to produce HRP2 but gains a survival advantage in a world saturated with HRP2-detecting tests. It is a textbook case of Darwinian selection, accelerated by the very tool designed to control the disease.

For the child, the consequences are concrete. A negative RDT can mean a delayed diagnosis, a missed opportunity for early ACT, and a progression to severe malaria. In a scenario similar to what is seen in Senegal, where tuberculosis patients default treatment when clinic hours overlap market days, delays in diagnosis compound other barriers to care. The difference is that malaria can kill within 24 hours of the first symptom, especially in young children.

A Clinical Triage Under Constant Time Pressure

During the rainy season, which runs from June to October, a rural health center in the Centre-Nord region may see 80 or more febrile children per day. The consultation room is crowded, the queue stretches outside, and the clinician has an average of three to five minutes per patient. The RDT takes 15 minutes to develop, but many clinicians interpret the result before the timer goes off, especially when the line is faint. A negative result, in this context, is often a relief: one less patient to treat. But the growing evidence of false negatives has made that relief unreliable.

The decision to treat a negative-RDT child is not taken lightly. ACTs are not free; they cost the health system roughly US$ 0.50 to 1 per course, and stockouts occur. Overtreatment wastes a limited resource. But the risk of undertreatment is higher: severe malaria in a child under five carries a case fatality rate of 10 to 20% even with treatment, and higher without. The clinician's calculus, therefore, leans toward treatment when any clinical sign suggests malaria.

National guidelines support this calculus. The 2022 update states: "In children under five with fever and negative RDT, if there is pallor, splenomegaly, or lethargy, treat with ACT." The wording is permissive, not mandatory. Some clinicians interpret it narrowly; others apply it broadly. The result is wide variation in practice. A 2023 audit in Kaya found that 70% of children with negative RDTs and clinical suspicion received ACT, but the remaining 30% did not—and some of those later presented with severe anemia.

The time pressure is not limited to the consultation. The clinician must also decide whether to draw blood for a thick smear (which costs time and supplies) or refer the child to a higher level (which costs the family transport and lost wages). Most opt for presumptive treatment, documenting the decision as "clinical malaria" in the register. The RDT result is recorded but often disregarded.

The Biology Behind the Blunted Fever Response

Not all false-negative RDTs are due to gene deletions. Another important factor is parasite density. RDTs have a detection threshold of roughly 100 to 200 parasites per microliter of blood. Below that level, the test may be negative even when the child is infected. These low-density infections are common in children who have developed partial immunity through repeated exposure. They carry the parasite at densities that cause fever only intermittently, but the infection can still trigger severe anemia over weeks.

Chronic asymptomatic parasitemia is a hallmark of high-transmission settings. A child may have a parasite density of 50 parasites/µL for months, without fever, until an intercurrent infection or nutritional stress tips the balance. Then the fever spikes, the parent brings the child to the clinic, and the RDT is negative. The clinician treats for a bacterial infection, but the malaria persists. Two weeks later, the child returns with pallor and a hemoglobin of 6 g/dL.

Submicroscopic infections—those below the detection threshold of both RDT and microscopy—are also transmissible to mosquitoes. A child with a negative RDT can still infect the vector that bites them. This means that the diagnostic gap contributes to ongoing transmission, even as treated children clear their symptoms. The cycle of undetected carriage perpetuates the parasite reservoir.

The immune response to low-density infections is blunted. The child produces fewer inflammatory cytokines, which means less fever but also less clearance of the parasite. The immune system tolerates the parasite rather than eliminating it. This is a survival strategy for both host and parasite: the host avoids a strong inflammatory reaction that can cause tissue damage, and the parasite maintains a foothold in the population. For the clinician, it means that the absence of high fever does not rule out malaria.

What the Numbers Say About Overtreatment and Undertreatment

The Kaya study, published in Malaria Journal in 2024, provides a snapshot of the dilemma. Among 1,200 febrile children aged 6 to 59 months, 28% of those with negative RDTs had malaria by PCR. Of those PCR-positive children, roughly 70% had received ACT anyway—presumably based on clinical judgment. That means the overtreatment rate was high, but the appropriate treatment rate was also high. The remaining 30% who did not receive ACT were the undertreated group, at risk for severe outcomes.

At the same time, the study found that 12% of children with positive RDTs did not receive ACT, mainly because of stockouts at the health center. These children were undertreated despite a correct diagnosis. The parallel undertreatment of negative-RDT children and positive-RDT children without ACT creates a double burden: some children get the wrong drug, and some get no drug at all.

The financial implications are significant. Burkina Faso's national malaria program budget spends roughly 60% on ACT procurement and only 10% on diagnostics. The imbalance reflects a historical reliance on clinical diagnosis, reinforced by the perception that RDTs are cheap and reliable. But as the reliability of RDTs erodes, the spending on ACTs may increase further, because clinicians treat more children presumptively. The Global Fund and the President's Malaria Initiative have begun to fund alternative diagnostics, but the shift is slow.

Overtreatment also drives selective pressure for artemisinin resistance. Plasmodium falciparum has already developed partial resistance to artemisinin in Southeast Asia and South America. In Africa, resistance markers have been detected at low frequencies, but the widespread use of ACTs—including in children who do not actually have malaria—could accelerate the spread. The WHO recommends that ACTs be reserved for confirmed malaria cases, but that recommendation assumes a reliable diagnostic. In Burkina Faso, the diagnostic is no longer reliable.

How Burkina Faso's Supply Chain Adds Another Variable

RDT procurement in Burkina Faso is centralized through the national medical store, which sources WHO-prequalified brands. But stockouts are common, especially during the rainy season when demand spikes. When the preferred brand is unavailable, health centers receive a different brand, sometimes with different sensitivity. Cheaper brands may have lower sensitivity, increasing the false-negative rate further.

Clinicians report that they are rarely informed about the brand change or its performance characteristics. They see the same two-line format and assume the test works the same. But a study comparing RDT brands in Burkina Faso found that sensitivity for detecting pfhrp2-deleted parasites varied from 60% to 90% depending on the brand. The variation is not trivial; it can mean the difference between a child being treated or not.

The supply chain also affects ACT availability. When ACT stockouts occur, clinicians may turn to older drugs like chloroquine or sulfadoxine-pyrimethamine, to which resistance is widespread. The treatment failure rate for these drugs can exceed 50%. In a scenario similar to rural Ghana, where diabetes clinics diagnose metformin failure without HbA1c lab access, clinicians also work with imperfect tools. The parallel is that both settings require adaptive strategies to manage chronic conditions with limited diagnostics.

The financial incentive to treat based on symptoms alone is perverse. ACTs are more expensive than RDTs, but the cost of not treating a child who later develops severe malaria is even higher—both in human terms and in healthcare costs. A severe malaria case requires hospitalization, intravenous artesunate, blood transfusion, and often intensive care, costing ten to fifty times more than an outpatient ACT course. The system, therefore, incentivizes presumptive treatment, even if it means overtreating many children to save a few.

Toward a Practical Diagnostic Algorithm for the Front Line

Several pilot programs in Burkina Faso are testing alternative diagnostic strategies. One approach uses a second RDT that targets Plasmodium lactate dehydrogenase (pLDH), an enzyme produced by all Plasmodium species and not subject to pfhrp2 deletion. The pLDH test is slightly more expensive—roughly US$ 0.10 more per test—but it catches the missed cases. Some sites now use a two-test algorithm: first an HRP2 test, then a pLDH test if the first is negative. If both are negative but fever persists, the clinician treats presumptively.

Training modules have been developed to emphasize clinical signs that predict malaria in the context of negative RDTs. Pallor, splenomegaly, lethargy, and a history of convulsions are the four key signs. A simple algorithm: if two or more are present, treat with ACT regardless of RDT result. This algorithm was tested in a 2025 study in the Sahel region and showed sensitivity of 85% and specificity of 70% for detecting PCR-confirmed malaria. The trade-off is overtreatment of 30% of children who do not have malaria, but that is considered acceptable given the risk of missing a case.

Looking further ahead, loop-mediated isothermal amplification (LAMP) offers a field-deployable molecular test that can detect low-density infections within an hour. LAMP machines are now available at the district level in several pilot sites, but they require electricity, cold chain for reagents, and trained technicians. The cost per test is roughly US$ 5, which is too high for routine use. However, as a confirmatory test for negative-RDT children with clinical suspicion, it could reduce overtreatment and guide more precise use of ACTs. The WHO is evaluating LAMP for this purpose, but implementation is likely years away.

The path forward will require a combination of better diagnostics, updated clinical algorithms, and supply-chain improvements. No single intervention will solve the problem. The parasite will continue to evolve, and the health system must evolve with it. For now, clinicians in Burkina Faso will continue to make difficult decisions in the face of uncertainty, treating children whose RDTs say they are not sick, because the alternative is worse.

In addition to the Kaya study, a 2023 survey by the Centre Muraz research institute in Bobo-Dioulasso documented that among children presenting with fever in rural health centers, nearly one in four with negative RDTs had detectable parasitemia by PCR. These findings underscore the scale of the diagnostic gap. The survey also noted that clinicians who had received training on clinical signs were more likely to treat presumptively, suggesting that education can improve case detection.

Another emerging strategy involves the use of community health workers who are trained to recognize danger signs and administer ACTs even when RDTs are negative. In a pilot program in the Sahel region, community health workers were equipped with pLDH RDTs and a simplified algorithm. Preliminary results from 2024 show that this approach reduced the proportion of untreated malaria cases by roughly 40%, though overtreatment increased by 15%. The trade-off was deemed acceptable by the local health authorities.

The role of nutrition cannot be ignored. Malnourished children are more susceptible to severe malaria and also more likely to have low-density infections that evade RDTs. A 2022 study in Burkina Faso found that children with moderate acute malnutrition were 1.5 times more likely to have a false-negative RDT compared to well-nourished peers. Integrating nutrition screening into malaria management could help identify children at highest risk.

Finally, the diagnostic gap has implications for malaria surveillance. National incidence figures, which rely on RDT-confirmed cases, likely underestimate the true burden. A modeling study from 2023 estimated that official malaria incidence in Burkina Faso may be underreported by 15 to 25% due to false-negative RDTs. This undercount affects funding allocations, drug procurement, and epidemic response planning.