The study, led by Faiza Mohamed Juma of the Catholic University of Health and Allied Sciences in Mwanza and Mnazi Mmoja Hospital, was nested within the ZanTOTO project, an initiative supported by the International Centre for Antimicrobial Resistance Solutions in Copenhagen. Researchers deployed two WHO instruments in parallel: the Point Prevalence Survey tool, applied at two time points between February 2024 and May 2025 to capture which hospitalized children were on antibiotics and why, and the Health Care Facility Core Element Indicators Tool, administered once in April 2024 to score how well each hospital’s antimicrobial stewardship program was actually functioning. The survey covered the full spectrum of Zanzibar’s hospital tiers: one tertiary referral center, one regional hospital, and three district hospitals spread across the islands of Unguja and Pemba.

Of the 943 eligible pediatric patients, all aged thirteen or younger, 826 were on antibiotics on the survey days. The gradient across hospital tiers was striking and ran in an unexpected direction. At the regional Abdulla Mzee Hospital, 99.4 percent of children received antibiotics; district hospitals ranged from 92.0 to 96.9 percent. Yet at Mnazi Mmoja, the tertiary facility best equipped with diagnostics and specialist expertise, the figure fell to 71.8 percent. Statistical modeling confirmed the pattern: after adjustment, antibiotic use was significantly more prevalent in regional hospitals (adjusted prevalence ratio 1.39) and district hospitals (1.35) than at the tertiary center. Severity mattered too, with children classified by the McCabe score as having an ultimately fatal underlying condition significantly more likely to be treated.

The reverse gradient, with the weakest-resourced facilities prescribing most intensively, is the study’s most technically telling signal. The authors interpret it as a reflection of limited diagnostic capacity and fewer clinical decision-support resources at lower-tier hospitals, where clinicians confronting febrile infants have little choice but to treat empirically. The demographic profile of the wards reinforces the point: nearly two-thirds of admissions were infants under twelve months, a population in which bacterial sepsis can kill within hours and in which clinical signs alone cannot reliably distinguish bacterial from viral illness. Without laboratory confirmation, broad-spectrum coverage becomes the defensive default.

That defensive default is visible in the drug list. Gentamicin accounted for 35.7 percent of the 1,385 recorded prescriptions, ampicillin for 31.8 percent, and ceftriaxone for 16.1 percent; together these three agents made up 83.6 percent of everything prescribed. Categorizing the prescriptions with the WHO’s AWaRe framework reveals the deeper concern. The framework divides antibiotics into Access drugs, first- and second-choice agents with lower resistance potential; Watch drugs, higher-resistance-potential antibiotics that demand prioritized stewardship; and Reserve drugs, last-resort options. WHO targets call for more than 60 percent of use to come from the Access category and no more than 28 percent from Watch. Zanzibar’s pediatric wards inverted those targets: Watch drugs made up 53.8 percent of prescriptions while Access drugs accounted for only 45.8 percent. Reserve use, at 0.4 percent, was appropriately minimal.

Why children were being treated is equally revealing. Community-acquired infections drove 60.4 percent of antibiotic indications, with pneumonia at 27.3 percent and clinical sepsis at 18.0 percent the leading diagnoses. Hospital-acquired infections accounted for another 20.6 percent, while prophylaxis was comparatively rare at about 12 percent combined. But the diagnostic pipeline feeding these clinical labels was nearly empty: only 10.9 percent of antibiotic courses were supported by bacteriological culture results, and among the cultures that were done, blood cultures overwhelmingly dominated. Antimicrobial susceptibility testing was performed for just 30.4 percent of patients at Vitongoji District Hospital, the best performer, and fell to between 0.6 and 4.6 percent elsewhere, with none at all at Kinyasini. Only two of the five hospitals had functional microbiology laboratory capacity during the study period.

The stewardship infrastructure meant to restrain and rationalize this prescribing was, by WHO standards, barely functioning. Hospitals scored from 7 to 41 percent on the Core Element assessment, and none approached the 50 percent threshold that marks even a partially operational program. Mnazi Mmoja, the tertiary hospital, scored highest at 41 percent; the regional and district hospitals scored below 25 percent, with Kivunge at just 7 percent. The component scores map the anatomy of the failure: leadership commitment was recorded at zero in two hospitals, accountability and responsibility at zero in two, drug and therapeutic committee function at zero in one, and monitoring, surveillance, reporting and feedback were weak nearly everywhere. The assessment also showed that most prescribing was done by general practitioners, 77.7 percent of prescriptions, with specialists contributing only 15 percent, a workforce structure very different from the multidisciplinary, specialist-led stewardship teams of high-income systems.

There were brighter spots. Adherence to the Zanzibar Standard Treatment Guidelines stood at 77.8 percent, suggesting that when clinicians prescribe, they are largely following the national rulebook; the problem is less disobedience than the structure around the rulebook. The near-total reliance on the parenteral route, 98.6 percent of doses, is consistent with the severity of the hospitalized population, though the authors note that missing data on intravenous-to-oral switching made it impossible to assess de-escalation practices, a core stewardship lever. The researchers also point to a promising local precedent: a feasibility study within the same ZanTOTO collaboration found that serial C-reactive protein point-of-care testing could help optimize antibiotic treatment in hospitalized Zanzibari children, offering a low-cost biomarker pathway where culture capacity is absent.

Set against international benchmarks, the numbers are sobering. Antibiotic use in the Zanzibar pediatric wards exceeds estimates from mainland Tanzania, where pediatric prevalence has been reported around 84 percent and all-ages prevalence near 62 percent, and from other sub-Saharan African settings including Kenya at 67.7 percent, Botswana at 70.6 percent, and Nigeria at 69.7 percent. It even edges past Bangladesh’s 91.2 percent at the district level and stands far above the sub-45 percent figures typical of European acute care hospitals, where robust stewardship systems and accessible diagnostics constrain empirical treatment. Comparable WHO point prevalence surveys in Uganda, Ghana, and across the Global-PPS network have consistently shown community-acquired infection as the leading indication, confirming that Zanzibar’s pattern is an intensified version of a regional norm rather than an anomaly.

The study’s authors are careful about its limits: incomplete medical documentation prevented assessment of dosing accuracy, comorbidities, and temporal trends, and the cross-sectional design captures snapshots rather than trajectories. But the direction of travel is clear. For Zanzibar, which adopted its own Action Plan on antimicrobial resistance in 2019 aligned with the WHO Global Action Plan, the survey provides the first comparable local evidence that stewardship governance, laboratory capacity, and prescribing culture all need reinforcement simultaneously. The authors call for facility-specific interventions tailored to each hospital tier, expanded diagnostic capacity, and future research into whether the prescriptions being written are actually appropriate, and into the barriers preventing stewardship programs from taking root. In a decade when antimicrobial resistance threatens to undo a century of pediatric medicine, the wards of Zanzibar offer a precise, data-rich map of where the system leaks, and a reminder that saving the antibiotics may depend less on new drugs than on the unglamorous machinery of committees, laboratories, and feedback loops that decide when to reach for them at all.

Subject of Research: Antibiotic use and antimicrobial stewardship performance in pediatric wards of Zanzibar public hospitals

Article Title: Drivers influencing antibiotics use in pediatric wards across five public hospitals in Zanzibar: a WHO multiple point prevalence survey

Article References: Drivers influencing antibiotics use in pediatric wards across five public hospitals in Zanzibar: a WHO multiple point prevalence survey. (n.d.). https://doi.org/10.3389/frabi.2026.1874316

Image Credits: AI Generated

DOI: 10.3389/frabi.2026.1874316

Keywords: antibiotic use, antimicrobial stewardship, antimicrobial resistance, pediatrics, Zanzibar, WHO point prevalence survey, AWaRe classification, low- and middle-income countries, diagnostic capacity, hospital prescribing, Tanzania, public health

The study team, led by Chukwuma J. Okafor of the State University of Zanzibar together with colleagues Islam H. Mwita, Rodolfo Isidro Bosch Bayard, and Aziza Japhary, designed their investigation as a cross-sectional survey, a snapshot approach that captures patient attitudes at a single point in time. They adapted the Patient Satisfaction Questionnaire, a widely used instrument in health services research, into a structured 28-item tool spanning nine distinct domains: the physical environment, courtesy and interpersonal manner, communication, accessibility, time spent with the doctor, convenience, technical quality, expenses, and general satisfaction. This multidimensional framing matters because patient satisfaction is not a single feeling but a composite judgment, assembled from everything from the cleanliness of the waiting area to whether a clinician explained a diagnosis in words the patient could understand.

The granularity of the questionnaire allowed the researchers to see precisely where the service was succeeding and where it was collapsing. Administrative interactions fared comparatively well: 83.2 percent of respondents expressed satisfaction with registration staff, suggesting that the clerical backbone of the outpatient department functions acceptably. But the picture deteriorated sharply when the researchers turned to the human and environmental dimensions of care. Communication and patient-centred care items registered very low satisfaction, and a striking 92.9 percent of respondents reported dissatisfaction with cleanliness. Long waiting times emerged as another prominent source of frustration, compounding the sense that patients were spending hours in an unhygienic environment only to receive hurried, impersonal consultations.

Methodologically, the study makes several choices that reflect a growing sophistication in how health services researchers handle common outcomes and imperfect data. Because dissatisfaction affected well over 10 percent of the sample, the authors used modified Poisson regression with robust variance rather than ordinary logistic regression. The technical rationale is important: when an outcome is common, odds ratios derived from logistic models systematically overstate the underlying prevalence ratios, inflating the apparent strength of associations. Poisson regression with robust standard errors directly estimates prevalence ratios, giving a more honest picture of how much more likely one group is to be dissatisfied than another. Reporting adjusted prevalence ratios with 95 percent confidence intervals, the team identified lower satisfaction linked to perceived affordability, accessibility, and perceived socioeconomic barriers.

Equally notable is what the authors chose to disclose about the limits of their own statistics. Two adjusted prevalence ratios — one for perceived socioeconomic influence and one for female sex — came out with magnitudes exceeding 20, values so extreme that no prevalence ratio can plausibly represent them as true effects. Rather than presenting these numbers as headline findings, the researchers flagged them as almost certainly reflecting instability in the underlying data, reporting them for completeness but explicitly declining to interpret them as precise or generalisable effect sizes. They also reported that a logistic-regression specification of the same model failed to converge, a phenomenon known as quasi-complete separation, and chose to state this openly rather than dress up unstable odds-ratio estimates as reliable results. In a field where inflated effect estimates are too often laundered into policy claims, this kind of transparency is rare and welcome.

The reliability of the survey instrument itself received similarly candid treatment. The overall 28-item scale achieved a Cronbach’s alpha of 0.685, a figure the authors describe as acceptable-to-borderline, sitting just below the conventional 0.70 threshold that many researchers treat as the floor for internal consistency. Domain-level reliability varied widely, ranging from 0.42 to 0.82 for domains containing three or more items, and several domains did not reach conventional reliability thresholds at all. Instead of burying this variation behind a single summary statistic, the team reported it domain by domain, acknowledging that some subscales of the adapted questionnaire measured their intended constructs less consistently than others. For readers, this means the domain-level satisfaction percentages should be read as descriptive indicators of where problems lie rather than as precision instruments.

The authors are equally careful about the study’s sampling design. The 309 participants were recruited through a non-probability approach that combined purposive selection of the study hospital with convenience sampling of outpatients during the data-collection period. Convenience sampling, while practical in a busy clinical setting, cannot guarantee that the surveyed patients represent the full population of outpatients, and the researchers accordingly frame their multivariable associations as hypothesis-generating for this particular facility rather than as confirmed, generalisable predictors of patient satisfaction. This is a methodologically honest posture, and it does not diminish the study’s practical value; it simply locates its conclusions at the level of Jitimai District Hospital, where they are most actionable.

Ethical oversight for the research was provided by the Zanzibar Health Research Ethics Committee, which granted approval under number ZAHRE/02/ST/FEB/2025/51, with additional administrative permission from the State University of Zanzibar and the hospital administration. The study was conducted in accordance with the Declaration of Helsinki, and every participant provided written informed consent after being told the study’s purpose, their right to decline or withdraw at any stage without any effect on their care, and the confidentiality protections in place. Only adults aged 18 and above were eligible, so no parental or guardian consent procedures were required. The authors received no external funding for the work and declare no competing interests, and the article is published open access under a Creative Commons licence.

What elevates the study beyond a single hospital’s report card is its identification of modifiable targets. The domains that performed worst — communication, interpersonal care, cleanliness, and waiting time — are precisely the ones that facility-level quality improvement programmes can address without waiting for new buildings, new equipment, or new budgets of the magnitude required for tertiary care upgrades. Training clinicians in patient-centred communication, reforming provider attitudes, tightening cleaning protocols, and reorganising patient flow to cut waiting times are all interventions with documented track records in low- and middle-income country settings. The authors conclude that interventions targeting patient-centred communication, provider attitudes, and financial and access barriers are essential to improve outpatient service quality, and their data give hospital managers a concrete priority list.

The broader significance lies in what the findings suggest about health systems across similar settings. Outpatient departments in low-resource contexts are the primary interface between patients and healthcare systems, and service quality at that interface critically shapes whether people seek care early, adhere to treatment, or trust public medicine at all. Evidence from Zanzibar has historically been limited, making this study a useful data point for regional health planners. If nearly 59 percent dissatisfaction can be documented at one district hospital with transparent methods and honest uncertainty, the question becomes how many neighbouring facilities would produce similar numbers under the same scrutiny — and whether the modest, well-defined fixes this study points toward, from cleaner corridors to kinder consultations, might be among the highest-yield investments available to health systems operating on thin margins.

Subject of Research: Patient satisfaction and its associated factors among outpatients at a district hospital in Zanzibar

Article Title: Patient satisfaction and associated factors among outpatients at a district hospital in Zanzibar: a facility-based cross-sectional study

Article References: Okafor, C. J., Mwita, I. H., Bosch Bayard, R. I., & Japhary, A. (2026). Patient satisfaction and associated factors among outpatients at a district hospital in Zanzibar: a facility-based cross-sectional study. BMC Health Services Research. https://doi.org/10.1186/s12913-026-15685-7

Image Credits: AI Generated

DOI: 10.1186/s12913-026-15685-7

Keywords: patient satisfaction, outpatient care, Zanzibar, district hospital, health services research, PSQ-18, modified Poisson regression, low- and middle-income countries, healthcare quality, communication, waiting times, cleanliness