The research team, led by Augustino Hellar of Prime Health Initiative Tanzania together with collaborators from the Tanzanian Ministry of Health, Jhpiego, and regional authorities, took advantage of an unusual opportunity: a complete census of antenatal care delivery rather than a sample. Between January 2023 and August 2024, all 5,936 pregnant women who enrolled in group antenatal care, often abbreviated G-ANC, across six public health facilities in Geita Region were included in the analysis. These facilities spanned the typical tiers of the Tanzanian health system, comprising two dispensaries, two health centers, and two district hospitals. The women were organized into 149 group cohorts, a model in which small groups of expectant mothers attend antenatal sessions together, combining clinical assessment with peer learning and discussion.

Group antenatal care has attracted growing interest across low- and middle-income countries because it promises to improve both the experience and the quality of care. Instead of the traditional one-on-one consultation, women in a G-ANC cohort meet repeatedly with the same facilitators and peers, which proponents argue increases satisfaction, knowledge retention, and retention in care. Tanzania’s Mlinde Mama Project embedded this model into routine government services rather than running it as a standalone research intervention, which is precisely what makes the new findings valuable. When a model moves from a carefully resourced pilot into the everyday machinery of a health system, its performance becomes a test of the system itself.

The headline numbers from the analysis were genuinely encouraging. Coverage of four or more antenatal care visits ranged from 90.2 to 95.0 percent across the six facilities, facility-based delivery ranged from 94.1 to 97.7 percent, and malaria testing reached 91.5 to 95.5 percent of enrolled women. Blood pressure monitoring was recorded for every single participant at every facility, a perfect 100 percent. These figures suggest that the group care model, when integrated into routine services, can achieve and sustain very high levels of service utilization, confirming that the attendance half of the maternal health equation is being solved at these sites.

But the picture changed sharply when the researchers looked beyond utilization to what clinicians actually delivered during those visits. Coverage of three or more doses of intermittent preventive treatment for malaria in pregnancy, the standard sulfadoxine-pyrimethamine regimen known as IPTp3+, ranged from 60.1 percent at the lowest-performing facility to 86.3 percent at the best, a gap of more than 26 percentage points for an intervention that protects both mother and fetus from a potentially devastating infection. Iron and folate supplementation, a cornerstone of prenatal nutrition, ranged from 75.0 to 92.7 percent. Most striking were the diagnostic services: HIV and syphilis testing coverage spanned from 59.9 to 95.0 percent, and urine testing, which screens for conditions such as pre-eclampsia risk and urinary tract infections, ranged from a troubling 30.1 percent to 62.7 percent. In other words, at one facility fewer than one in three women received a urine test that guidelines recommend for all.

To synthesize these differences, the team calculated a composite facility performance score as the mean coverage across nine selected indicators. The scores ranged from 80.1 to 91.9 percent, with Bwanga Health Center ranking highest and Butengorumasa Dispensary lowest. Health centers demonstrated the strongest overall performance across most indicators, outperforming both the smaller dispensaries and, in several respects, the district hospitals. Statistical testing using Pearson’s chi-square confirmed that the variation in coverage of the nine composite indicators across facilities was significant, with p-values below 0.05, meaning the differences were unlikely to be products of chance.

The pattern of variation carries important technical implications for how maternal health programs are monitored. Attendance-based metrics, such as the proportion of women completing four antenatal visits, are the traditional currency of maternal health surveillance because they are easy to measure and track. Yet the Geita data demonstrate that these metrics can mask profound differences in effective coverage, the proportion of the population actually receiving the interventions they need. A woman can attend every scheduled session and still leave without syphilis screening, without her full course of malaria prophylaxis, or without anemia assessment. Effective coverage, a concept increasingly emphasized by global health measurement initiatives, requires looking at what happens inside the consultation, not merely whether the consultation occurred.

Why would facilities within the same region, operating under the same national guidelines and implementing the same group care model, diverge so widely? The study did not formally test explanatory mechanisms, but the structure of the findings offers clues. Interventions requiring reliable supply chains, such as sulfadoxine-pyrimethamine for malaria prophylaxis and rapid diagnostic tests for HIV and syphilis, showed the largest gaps, while interventions depending only on clinical skill and a blood pressure cuff achieved universal coverage. This suggests that stock-outs, laboratory capacity, and equipment availability, rather than staff knowledge or patient demand, may drive much of the variation. Facility tier also mattered, with health centers occupying a sweet spot of sufficient resources without the patient volumes and staffing pressures that may strain larger district hospitals.

The authors argue that their findings support a shift in how antenatal programs are managed: facility-level performance monitoring should track the content of care alongside attendance, supported by supportive supervision and continuous quality improvement cycles. Because the analysis drew on routinely collected implementation data rather than a bespoke research dataset, the approach is inherently scalable; health managers could, in principle, reproduce these composite scorecards using existing health management information systems. The study also carries a caution about interpretation, since it reflects a single region and a program supported by external funding from the Gates Foundation, and the authors note that the published version is an early-release article subject to further editorial refinement.

For the global maternal health community, the message from Geita is both sobering and actionable. Tanzania has made substantial progress in getting pregnant women into contact with the health system, and the near-universal blood pressure monitoring and delivery coverage documented here show that the system can perform at scale. The unfinished agenda is consistency: ensuring that the woman at the lowest-performing dispensary receives the same malaria prophylaxis, the same syphilis test, and the same urine screening as the woman at the best-performing health center. Closing that gap does not require new medical discoveries, only the disciplined application of measurement, supervision, and supply chain management to interventions that are already proven, affordable, and sitting on national guidelines waiting to be delivered.

Subject of Research: Facility-level variation in the delivery of essential antenatal care interventions during routine group antenatal care implementation in Tanzania

Article Title: Beyond attendance: facility-level variations in the delivery of essential antenatal care interventions during routine group antenatal care implementation in Tanzania: a secondary analysis

Article References: Beyond attendance: facility-level variations in the delivery of essential antenatal care interventions during routine group antenatal care implementation in Tanzania: a secondary analysis. (n.d.). https://doi.org/10.1186/s12913-026-15769-4

Image Credits: AI Generated

DOI: 10.1186/s12913-026-15769-4

Keywords: antenatal care, group antenatal care, Tanzania, maternal health, effective coverage, health facility performance, quality of care, malaria in pregnancy, IPTp, health services research, implementation science, Geita Region

The retrospective study covered January 2023 through October 2025 and began with a hospital-wide sweep of emergency department and inpatient records carrying any ICD-10 malaria code. Of 519 laboratory-confirmed malaria diagnoses recorded during the period, 428 (82.5 percent) were P. vivax, 88 (17.0 percent) were P. falciparum, two were mixed infections, and one was P. malariae. From this pool, 127 patients met the analytic criteria: microscopy-confirmed malaria, at least one dose of intravenous artesunate given for severe disease or inability to tolerate oral therapy, survival of at least 24 hours after admission, and laboratory results available at two or more time points. Notably, 22.7 percent of vivax patients and 30.7 percent of falciparum patients entered the cohort, a proportion of hospitalization far higher than the 1.6 to 3.2 percent reported in earlier Peruvian and Brazilian Amazonian series.

The headline result is stark in its symmetry. Severe malaria, defined by at least one criterion from the 2022 WHO guidelines, occurred in 51 of 97 vivax patients (52.6 percent) and 14 of 27 falciparum patients (51.9 percent). That equivalence emerged despite a marked asymmetry in underlying vulnerability: recognized risk factors or comorbidities were recorded in 59.3 percent of falciparum patients but only 26.8 percent of vivax patients, a difference largely driven by pregnancy. In other words, vivax patients reached comparable severity with a lighter load of predisposing conditions, suggesting that the parasite itself, rather than patient frailty, was driving the clinical picture.

The spectrum of severity criteria also overlapped substantially between species. Prostration and altered consciousness were the most frequent clinical manifestations, while severe anaemia led the biochemical criteria. Sixty-six of the 67 patients who ultimately fulfilled a severity criterion already met one at admission, but nine patients developed new or recurrent WHO-recognized manifestations during hospitalization, including impaired consciousness, shock, metabolic acidosis, jaundice, severe anaemia, haemoglobinuria and renal impairment at various time points after admission. The authors argue that this longitudinal evolution supports continued clinical and laboratory reassessment throughout the hospital stay, even though formal repeated screening against all criteria is not mandated by current WHO guidance.

Laboratory data reinforced the impression that vivax malaria is haematologically aggressive. At baseline, patients with severe disease had significantly lower haemoglobin and haematocrit values than non-severe patients, and haemoglobin remained lower in severe cases after treatment. In non-severe malaria, falciparum infection was actually associated with a less favourable baseline profile, with lower haemoglobin and haematocrit than vivax. Strikingly, platelet counts were significantly higher in falciparum infections in both severity strata, echoing prior reports that thrombocytopenia in vivax malaria can be as marked as, or more marked than, that seen in falciparum disease, and often out of proportion to the peripheral parasite density visible on a blood smear.

That dissociation between circulating parasites and clinical damage has a plausible biological explanation. Human splenectomy studies have shown that P. vivax–infected erythrocytes accumulate alongside immature reticulocytes in the red pulp of the spleen, creating a hidden reservoir that peripheral microscopy systematically undercounts. A larger total parasite biomass has been linked to thrombocytopenia, inflammatory activation, endothelial dysfunction and disruption of the vascular glycocalyx. Endothelial activation in acute vivax malaria is pronounced and associated with impaired microvascular reactivity and reduced tissue perfusion, and unlike in severe falciparum malaria it does not appear to be driven by intravascular haemolysis, pointing instead to depleted arginine and nitric oxide bioavailability as a key pathway toward organ dysfunction.

The relapsing biology of P. vivax adds a cumulative dimension to its threat. Reactivation of dormant hepatic hypnozoites or release of parasites from extravascular reservoirs can produce recurrent episodes before haematological recovery is complete, progressively eroding haemoglobin reserves and nutritional status and priming patients for severe disease in later attacks. This mechanism may be especially relevant in the Peruvian Amazon, where vivax predominates and more than half of the study patients, 52.8 percent, reported previous malaria episodes. Evidence from Indonesia suggests that although early mortality after a first episode is higher with falciparum, the mortality gap narrows and may even reverse after repeated vivax infections.

Treatment outcomes in the cohort were generally favourable. The median time to parasite clearance after intravenous artesunate was 48 hours, with no substantial difference by severity, and 93.3 percent of the 89 patients evaluated at 72 hours had completely cleared parasites on smear. The median hospital stay was five days, and all surviving patients were switched to oral antimalarial therapy after parenteral treatment. Yet the study also exposed gaps in pre-hospital care: 60.6 percent of patients had received no antimalarial treatment before admission, and 14.2 percent had been given empirical therapy with no antimalarial activity, a pattern at odds with WHO recommendations for prompt effective treatment after diagnosis.

The single death in the cohort underscores how conventional severity criteria can fail pregnant patients. A pregnant woman with falciparum malaria, admitted eight days after symptom onset, met neither clinical nor biochemical severity criteria at admission and had a normal platelet count. Roughly 48 hours after starting artesunate, her platelets had collapsed from 253 to 36 thousand per microlitre, she remained parasitaemic, and moderate anaemia persisted. She suffered foetal loss on hospital day five, followed by massive obstetric haemorrhage, and died the next day despite timely obstetric intervention and transfusion. The authors note that while the recorded cause of death was uterine haemorrhage, falciparum infection in pregnancy is an established cause of stillbirth, and the case illustrates how difficult it can be to disentangle immediate from contributing causes of maternal death.

Coinfections added further complexity to the clinical landscape. Dengue was identified in 14 patients, eight with severe and six with non-severe malaria, alongside two cases of bacterial sepsis and one of leptospirosis, while intestinal parasitosis was detected in 21 patients. Previous work has suggested that dengue–malaria coinfection may worsen haematological complications, though the small number of coinfected cases here precluded any conclusion. The authors also acknowledge the limits of their retrospective, single-centre design: incomplete follow-up laboratories, cultures obtained only when infection was clinically suspected, reliance on microscopy with limited sensitivity for low-level or mixed infections, and the likelihood that the most severely ill patients died before ever reaching the hospital. Even so, the central message stands. In this low-transmission Amazonian setting, P. vivax carried a burden of WHO-defined severe malaria comparable to P. falciparum, with broadly similar clinical, laboratory and parasitological profiles, and the authors conclude that both locally prevalent species warrant the same degree of clinical vigilance.

Subject of Research: Comparative severity of Plasmodium vivax and Plasmodium falciparum malaria in hospitalized patients in the Peruvian Amazon

Article Title: Comparable severity of Plasmodium vivax and Plasmodium falciparum malaria in hospitalized patients in the Peruvian Amazon

Article References: Bresciani, L., Guardiani, M., Rojas-Blasquez, M., Ramirez, E., Iaquinandi, S., Martel, C., Velarde-Mera, M., Sifuentes-Vidigal, S., Carraro, A., Russo, G., Pasquazzi, C., Celis, J. C., Casapia Morales, M., Nicastri, E., Asayag, C. R., Lichtner, M., & Vita, S. (2026). Comparable severity of Plasmodium vivax and Plasmodium falciparum malaria in hospitalized patients in the Peruvian Amazon. New Microbes and New Infections, 73, Article 101849. https://doi.org/10.1016/j.nmni.2026.101849

Image Credits: AI Generated

DOI: 10.1016/j.nmni.2026.101849

Keywords: malaria, Plasmodium vivax, Plasmodium falciparum, severe malaria, Peruvian Amazon, WHO severity criteria, intravenous artesunate, thrombocytopenia, severe anaemia, malaria in pregnancy, parasite clearance, endothelial dysfunction