Discussion

AMR is a critical global health challenge driven mainly by inappropriate antibiotic use in clinical settings. Understanding antibiotic prescribing patterns through hospital-based surveillance is essential for informing stewardship strategies.

In this study, among inpatients at Tanga RRH in Tanzania, the prevalence of antibiotic use was higher (77.6%) than that reported in previous studies from Tanzanian hospitals, with a range of 44% to 62.3% across the six referral hospitals. ^{Reference Seni, Mapunjo and Wittenauer11} Similar high rates have been observed in Uganda (73.7%) ^{Reference Kiggundu, Wittenauer and Waswa14} and Kenya (62%). ^{Reference Okoth, Opanga, Okalebo, Oluka, Baker Kurdi and Godman15} However, the rate at Tanga RRH is lower than that reported in some West African countries. ^{Reference Okedo-Alex, Akamike, Iyamu and Umeokonkwo16} Compared to middle-income countries such as Thailand ^{Reference Anugulruengkitt, Charoenpong and Kulthanmanusorn17} and HICs such as Canada and Europe, ^{Reference Browne, Chipeta and Haines-Woodhouse18} the prevalence observed in Tanga RRH was markedly higher. These findings underscore the persistent global disparities in antibiotic use. ^{Reference German, Frenette and Caissy5–Reference Abubakar and Salman9} Such regional differences likely reflect variations in AMS capacity, diagnostic infrastructure, regulation of prescribing, and the overall infectious disease burden. Stronger oversight and access to diagnostics in HICs help contain use, while limited resources and greater reliance on empirical therapy contribute to higher use in LMICs.

Older patients and those in surgical wards showed greater reliance on antibiotics, consistent with the clinical expectations and demands of these settings, where surgical procedures and empirical treatments are frequently employed. Similar ward-specific trends have been reported in Tanzanian hospitals, where pediatric and surgical patients are often prescribed antibiotics. ^{Reference Seni, Mapunjo and Wittenauer11} Likewise, in Kenya, most pediatric inpatients received antibiotics, ^{Reference Hooft, Ndenga and Mutuku19} reinforcing the notion that empiric antibiotic use in high-risk groups is a common regional practice. Beyond these groups, cancer patients and other immunocompromised individuals are also recognized as high-risk populations, in whom frequent antibiotic use is often justified by increased susceptibility to infections.

Broad-spectrum antibiotics and pharmacological classes such as imidazoles, cephalosporins, and penicillins collectively dominate prescribing patterns. Prescribing was skewed toward Watch antibiotics, with Access agents falling short of the WHO target of 60% use. ^{Reference Zanichelli, Sharland and Cappello20,Reference Sharland, Cappello and Ombajo21} This overreliance on broad-spectrum agents such as ceftriaxone, a Watch antibiotic listed in both WHO AWaRe and Tanzania’s NEMLIT, raises concern, particularly in the absence of routine AST to guide use.

These findings align with those of previous studies in Tanzania and other African countries that have documented similar trends in prescription patterns. ^{Reference Kizito, Lalitha, Kajumbula, Ssenyonga, Muyanja and Byakika-Kibwika22–Reference Yeika, Ingelbeen, Kemah, Wirsiy, Fomengia and van der Sande24} However, antibiotic use patterns vary globally. For example, cefotaxime and amoxicillin are predominant in Botswana, ampicillin and gentamicin in the Republic of Congo, ^{Reference da Costa, Serrano, Chambel and Oliveira25} piperacillin-tazobactam and cefazolin in Canada, ^{Reference Lima, Silva da, Barbosa and Barreiro26} and penicillin–beta-lactamase inhibitor combinations in Europe. ^{Reference Roy, Calon, Simonyan and Bergeron27} These variations may reflect differences in disease burden and drug availability, as demonstrated by multi-country analyses of antimicrobial prescription patterns. ^{Reference Klein, Impalli and Poleon28,Reference Hsia, Lee and Versporten29} In many African settings, the wide availability and low cost of drugs such as metronidazole and ceftriaxone contribute to their frequent use. ^{Reference D’Arcy, Ashiru-Oredope and Olaoye30} Without AST data to support these prescriptions, the continued use of Watch Group antibiotics increases the risk of accelerating AMR. ^{Reference Zanichelli, Sharland and Cappello20,Reference Hsia, Lee and Versporten29}

Despite the high prevalence of antibiotic use, microbiological testing to guide therapy is rarely performed. Prescribing seldom relies on laboratory results, reflecting broader patterns of dependence on empirical treatment in Tanzania due to diagnostic constraints. ^{Reference Seni, Mapunjo and Wittenauer11} Similar trends of limited microbiological support for prescribing have been documented in Botswana, Ghana, and Sierra Leone. ^{Reference Kamara, Kanu and Maruta31,Reference Kwame Yevutsey, Ohene Buabeng and Aikins32} Several barriers may contribute to the underutilization of microbiology services, including clinicians not requesting tests, high costs, stockouts of reagents, incomplete patient records, and a lack of skilled personnel. As a result, empirical antibiotic prescribing remains the norm at Tanga RRH, increasing the risk of inappropriate use and driving AMR. ^{Reference Kiggundu, Wittenauer and Waswa14} Particularly in surgical wards, the routine practice of administering antibiotics before and after procedures—often in the absence of clear prophylactic guidelines—reflects a reliance on empirical treatment to mitigate infection risk, a phenomenon also reported in other African settings. ^{Reference Hassan, Chan, Stevens and Stupans33,Reference Nnadozie, Umeokonkwo and Maduba34} Similar practices have been observed in the management of low-birthweight neonates and postpartum women. ^{Reference Letouzey, Foix-L’Hélias and Torchin35} Strategic interventions are needed to address the underutilization of microbiology services. Strengthening the laboratory infrastructure, ensuring a regular supply of reagents, and building personnel capacity are critical steps. Integrating AST into clinical workflows through antibiotic stewardship programs can help clinicians move from empirical to evidence-based prescribing.

Our finding that only peripheral vascular catheter use remained independently associated suggests the importance of vascular access in driving prescription practices at Tanga RRH. This finding aligns with previous reports that patients with indwelling vascular devices are perceived to be at a higher risk of infection, leading clinicians to initiate antibiotics empirically as a preventive measure. ^{Reference Serane and Kothendaraman36–Reference Kolasiński38} Although admission to surgical and mixed wards, as well as surgery since admission, showed strong associations in the univariate analysis, these factors did not retain significance after adjustment. This suggests that their effect may be mediated through device-related exposures or the clinical context, rather than being independent predictors of outcomes. Similar patterns have been observed in other Tanzanian settings, where surgical patients frequently receive antibiotics; however, the influence diminishes once device use and ward-level factors are taken into account. ^{Reference Kilipamwambu, Bwire, Myemba, Njiro and Majigo39} Our findings suggest that surgery per se may not be an independent predictor of antibiotic use once device-related exposure and ward-level factors are considered. This highlights the importance of AMS interventions in prioritizing device management and ward-level prescribing practices, rather than focusing solely on surgical status.

Despite apparent alignment with national treatment guidelines, prescribing diverged from WHO stewardship priorities by leaning heavily toward Watch antibiotics rather than Access drugs. Similar gaps in adherence have been documented in Tanzanian primary healthcare settings and during earlier phases of stewardship implementation, underscoring the ongoing challenge of aligning practice with policy. ^{Reference Sangeda, Kibona and Munishi10,Reference Massele, Rogers and Gabriel40} These findings suggest an insufficient alignment with stewardship principles. Prioritizing narrow-spectrum Access antibiotics where clinically appropriate is a core stewardship strategy, as these agents carry lower resistance potential compared with broad-spectrum Watch and Reserve agents. Therefore, strengthening the diagnostic capacity to support de-escalation and targeted prescribing is essential to curb AMR in this setting. ^{Reference Zanichelli, Sharland and Cappello20,Reference Sharland, Cappello and Ombajo21}

Taken together, the high prevalence of antibiotic use, low reliance on AST, and disproportionate use of Watch antibiotics underscore the urgent need to strengthen AMS at Tanga RRH. Interventions should include enforcing adherence to national guidelines, improving prescriber training, and addressing structural barriers, such as the lack of laboratory capacity and timely AST results. Integrating routine AST into clinical decision-making, particularly in surgical and high-risk wards, could curb empirical overuse. Furthermore, periodic PPSs and audits can help monitor antibiotic use patterns and guide policy adjustments aligned with WHO stewardship targets.

The limitations of this PPS include its single-site scope, which restricts generalizability to other hospitals in Tanzania, and the relatively small sample size. Because data were collected at a single point in time, the study could not capture seasonal variations in infection patterns or changes in prescription over time. The observed association between peripheral catheter use and antibiotic prescription, while consistent with the global literature, should be interpreted cautiously, as it reflects empiric practice in the absence of microbiological diagnostics rather than a novel finding. Despite these limitations, this study offers valuable baseline insights into antibiotic use patterns at a RRH in Tanzania, utilizing the standardized WHO PPS methodology. This underscores the need for future multi-center or national PPSs, as well as repeated surveys linked to stewardship interventions.

This study noted the underutilization of microbiology laboratory services, suggesting a heavy dependence on empirical prescribing. These findings underscore the need to strengthen AMS by improving diagnostic capacity, promoting targeted therapy, and enhancing adherence to treatment guidelines. Collaboration with regional and national health authorities is essential to align hospital practices with national policies and address the systemic drivers of inappropriate antibiotic use. Further qualitative research, including interviews with prescribers, and expanded PPS in other facilities, is recommended to understand prescribing patterns better and scale up stewardship interventions.