Effectiveness of serial C-reactive protein point-of-care testing in optimizing antibiotic treatment in neonates and children in Zanzibar: an open-label, individual randomized controlled clinical trial protocol Authors: Daniel Joshua, Jeremiah Seni, Hanna Rahimi, Mayassa Salum Ally, Muhiddin Omar, Farid Said Ali, Jyoti Joshi, Said Mohammed Ali, Anja Poulsen, Stine Lund, Khadija Omar, Khadija Omar, Ahlam Amour, Mwanakhamis Seif, Thuwaiba Talib, Nadhra Khamis, Salama Bakari, Hafsa Omar, Hafidh Hassan, Simon Kühnert Published: 2026 ABSTRACT IntroductionAntimicrobial resistance is increasing globally. Children are disproportionately affected due to the overuse and misuse of antibiotics which is linked to lack of prompt diagnostics for suspected bacterial infections, limited access to health care facilities and financial constraints. The C-reactive protein (CRP) point-of-care test (POCT) presents an opportunity for quick and cost-effective diagnosis to establish the likelihood of a bacterial infection. However, information on the effectiveness of point-of-care CRP testing in hospitalized children in Zanzibar, Tanzania is limited.Methods and analysisThis multicenter, open-label, individual randomized controlled trial with 28 days follow-up is conducted at four hospitals in Zanzibar. Well-appearing neonates at risk of early onset sepsis, neonates (0 – 28 days) with clinical signs of infection on admission and children (6 months – 12 years) with febrile illness and/or diarrhea on admission are eligible for participation if their caretakers provide informed consent. The hospitals are supplied with CRP POCT equipment as well as a training session on CRP interpretation to aid in the clinical evaluation and antibiotic management of neonates and children. The primary outcome is the duration of antibiotic treatment within 14 days of the admittance of included neonates and children in each study arm and days to relapse of infection (defined as reinstitution of antibiotics within 72 h after completion of treatment).Ethics and disseminationThe study is approved by the Zanzibar Health Research Ethics Committee Ref. No. ZAHREC/04/AMEND/DEC/2023/11. The findings from this study will be used to generate policy briefs and technical reports to inform rational antimicrobial prescriptions’ clinical practices and guide future research. Findings will be presented at global conferences and published in peer-reviewed medical journals to foster access to wider communities within Tanzania and beyond.Trial registrationThis trial was registered with ISRCTN registry (ISRCTN25937092) on 22/01/2024.Supplementary InformationThe online version contains supplementary material available at 10.1186/s12887-026-07096-8. Background Antimicrobial resistance (AMR) is estimated to be associated with nearly 5 million deaths, with 1 in 5 deaths occurring in children [1]. Globally, bacterial pathogens are increasingly becoming resistant to antibiotics, constituting a significant threat to global health. AMR is driven by multiple factors, including antibiotic use. Pediatric populations experience high rates of antibiotic exposure, which contribute to this growing burden [2]. The indiscriminate use of antibiotics in children, both in the health care settings and through over the counter sales in community pharmacies and other medicine outlets, has also contributed the development of resistant strains [3], especially in resource limited settings like Tanzania [4]. Febrile illness and diarrhea are among the common reasons for health care visits in children and are major drivers of antibiotic prescribing, often in the absence of confirmed bacterial infection [5, 6]. Combating AMR in children requires comprehensive strategies. These include improving health care infrastructures, such as diagnostic tests to guide specific antimicrobial therapies [7], regulating antibiotics use [8] and implementing proper infection prevention and control measures [9]. Current practice for diagnosing suspected bacterial infections in Zanzibar, Tanzania is primarily clinical assessment, supported by basic laboratory investigations such as full blood picture [10]. Blood culture with antimicrobial susceptibility testing (AST) remains the gold standard for identifying causative pathogens and their resistance profiles, but is not routinely available [11]. Inflammatory markers such as interleukin, procalcitonin and CRP are potential alternatives for diagnosis of suspected bacterial infections [12]. Testing interleukin and/or procalcitonin is relatively expensive to carry out on a routine basis and requires sophisticated infrastructures [13]. C-reactive protein is suitable for use in routine clinical practices, as it is relatively cheaper and easy to sustain. It can be used as a point-of-care (POC) test [14, 15]. CRP testing has already been optimized in Tanzania, though it has not yet been fully integrated as a routine diagnostic tool across health care settings and especially in Zanzibar [16–18]. However, evidence on the use of serial CRP measurements to guide antibiotic treatment decisions in this setting remains limited [19]. C-reactive protein is an inflammatory biomarker in blood which can serve as a proper diagnostic test to quickly ascertain and establish the likelihood of a bacterial infection. Through POC testing of CRP, health care providers can make more accurate and targeted decisions regarding antibiotic prescription as they wait for a 3-day final culture and AST results [20]. However, CRP alone lacks absolute specificity in distinguishing bacterial from non-bacterial infections, as elevated levels can also be observed in non-bacterial conditions such as viral infections, autoimmune diseases and other inflammatory states. To improve diagnostic accuracy and antibiotic stewardship, an integrated approach combining serial CRP measurements with microbiological confirmation has increasingly been advocated [21]. Evidence supports the use of CRP POCT to guide antibiotic prescribing in children. However, most studies have been conducted in outpatient or high-resource settings. Evidence on the effectiveness of serial CRP measurements, particularly in hospitalized neonates and children in low-resource settings, remains limited. Therefore, knowledge gaps exist regarding the clinical utility and stewardship impact of serial CRP POCT in settings such as Zanzibar. With escalating AMR in neonates and children in Tanzania [22], this randomized controlled trial aims to determine the feasibility and effectiveness of using serial CRP POC tests for clinical decision-making on antibiotic treatment in neonates and children presenting with febrile illness and diarrhea in selected hospitals in Zanzibar. Methods Study design The study is a multicenter, open-label, individual randomized controlled clinical trial with 28 days follow-up, which is conducted in selected hospitals in Zanzibar. Neonates at risk of early onset sepsis, admitted neonates (0–28 days) with clinical signs of infection and admitted children (6 months to 12 years) with febrile illness or diarrhea that are eligible for participation and parents are asked for informed consent. Participating children are randomized to either intervention group (CRP-POC testing) or control group (usual care). Participants in the control group will receive usual care, which includes standard clinical assessment and routine laboratory investigations, with antibiotic initiation and duration determined according to Zanzibar guidelines [23] and clinician judgment, without the use of CRP point-of-care testing. The study flow is shown in Fig. 1.Fig. 1The schematic diagram of trial flow The schematic diagram of trial flow Study setting The trial is conducted in Zanzibar, part of the United Republic of Tanzania which has a total population of 1,889,773, with females accounting for 974,281 (52%) and children under 18 years of age accounting for 826,848 (48%) of the population. Zanzibar is composed of two main islands, Unguja and Pemba. The two islands are separated by a channel approximately 50 km (about 31 miles) wide [24]. The health care infrastructure comprises over 300 health facilities, including 189 public health facilities, which serve as the backbone of accessible health care and 135 private health facilities [25]. The study is conducted across four hospitals, with two located on each island, Mnazi Mmoja Hospital (MMH) and Kivunge Hospital (in Unguja), Mkoani Hospital and Vitongoji Hospital (in Pemba). The usual care in these facilities is primarily based on clinical diagnosis with limited laboratory diagnostics access and requested tests often take longer to process. Laboratory infrastructures for the study Public Health Laboratory-Ivo de Carneri (PHL-IdC) is a WHO Collaborating Centre for Neglected Tropical Diseases located in Pemba. It assists in implementing disease control strategies, conducting operational research and strengthening the capacities of health staff. PHL-IdC also provides microbiological diagnostic services, including blood culture using the BD BACTEC FX40 semi-automated system and antimicrobial susceptibility testing (AST) performed manually using the disk diffusion method in accordance with EUCAST guidelines. Routine laboratory investigations, including full blood picture and urinalysis, are conducted at referring hospitals such as Vitongoji and Mkoani. The microbiology laboratory at Makunduchi Health Centre, established in partnership with the Bernhard-Nocht Institute for Tropical Medicine (BNITM), provides diagnostic infrastructure for the study in Unguja island. The laboratory performs blood culture using the BD BACTEC FX40 system and AST using disk diffusion following EUCAST guidelines, in addition to routine tests such as full blood picture and urinalysis. These laboratories contribute to microbiological diagnosis and antimicrobial resistance surveillance in Zanzibar. Patients’ involvement There was no involvement of parents and caregivers in the development of the research question, design, or technique of the study. Nevertheless, their involvement in the feasibility study played a role in improving the study processes and protocols. Engagement of parents and caregivers is important, as their perspectives can help shape the relevance, acceptability and potential uptake of the study and its interventions. Although they were not involved in the early design phase, their insights are valuable for informing communication strategies and ensuring effective dissemination of results to the wider community. Study participants The study has three study population groups 1) Neonates with risk of early onset sepsis according to the WHO criteria [26], 2) neonates between 0 and 28 days admitted with suspected infection and 3) children between 6 months and 12 years admitted with either febrile illness or/and diarrheal disease. Neonates and children up to 12 years of age were included to reflect real-world clinical practice, where empiric antibiotic strategies and clinical decision-making are frequently applied across pediatric age groups, particularly in acute care settings. Additionally, including a broad age spectrum enhances the generalizability of the findings and allows evaluation of the intervention across clinically relevant sub population. Only neonates and children admitted with suspected infection are enrolled whereas those who develop infections after admission for other conditions than possible infections are excluded. Neonates and children attending the selected hospitals at any time, 24 h a day, meeting the inclusion criteria stipulated in Table 1 are eligible for participation.Table 1Inclusion and exclusion criteriaInclusion criteriaNeonates 0–28 days, born at termwith birth weight > 2000 g admitted with1) Well-appearing newborns with a risk factor of early onset sepsis according to the WHO criteriai. Membranes ruptured over 18 h before deliveryii. Maternal fever (over 38 °C) during laboriii. Amniotic fluid is foul smelling or purulent2) Suspected infection (clinical signs subjectively judged by the clinician)Children 6 months and 12 years admitted with3) Febrile illness with temperature above 38 °C or below 36 °C at admission, or a history of febrile illness within the last 72 hORDiarrhea disease (defined as the passage of 3 or more loose or liquid stools per day) with or without feverParents/caregivers of admitted children are able and willing to comply with all study requirementsParents/caregivers of admitted children are able and willing to give voluntary Informed consentExclusion criteriaSeverely ill where measurement of CRP POCT would delay the treatment processNeonates who need surgery within the first week of life or have major congenital malformationsPatients with known immunosuppression or severe chronic disease (HIV, hepatic disease, history of neoplastic disease, long-term systemic steroid use, Severe Acute Malnutrition (SAM) or similar conditions as assessed by the health care worker)Positive rapid diagnostic test for malariaHave taken antibiotics within 24 h before admittance Table 1Inclusion and exclusion criteriaInclusion criteriaNeonates 0–28 days, born at termwith birth weight > 2000 g admitted with1) Well-appearing newborns with a risk factor of early onset sepsis according to the WHO criteriai. Membranes ruptured over 18 h before deliveryii. Maternal fever (over 38 °C) during laboriii. Amniotic fluid is foul smelling or purulent2) Suspected infection (clinical signs subjectively judged by the clinician)Children 6 months and 12 years admitted with3) Febrile illness with temperature above 38 °C or below 36 °C at admission, or a history of febrile illness within the last 72 hORDiarrhea disease (defined as the passage of 3 or more loose or liquid stools per day) with or without feverParents/caregivers of admitted children are able and willing to comply with all study requirementsParents/caregivers of admitted children are able and willing to give voluntary Informed consentExclusion criteriaSeverely ill where measurement of CRP POCT would delay the treatment processNeonates who need surgery within the first week of life or have major congenital malformationsPatients with known immunosuppression or severe chronic disease (HIV, hepatic disease, history of neoplastic disease, long-term systemic steroid use, Severe Acute Malnutrition (SAM) or similar conditions as assessed by the health care worker)Positive rapid diagnostic test for malariaHave taken antibiotics within 24 h before admittance Inclusion and exclusion criteria Neonates 0–28 days, born at term with birth weight > 2000 g admitted with 1) Well-appearing newborns with a risk factor of early onset sepsis according to the WHO criteria i. Membranes ruptured over 18 h before delivery ii. Maternal fever (over 38 °C) during labor iii. Amniotic fluid is foul smelling or purulent 3) Febrile illness with temperature above 38 °C or below 36 °C at admission, or a history of febrile illness within the last 72 h OR Diarrhea disease (defined as the passage of 3 or more loose or liquid stools per day) with or without fever Intervention CRP POCT equipment, Aidian QuickRead go CRP POCT (Aidian, Espoo, Finland) [27] is supplied at the selected hospitals, along with a training package for HCWs and research assistants (RAs) on the use and interpretation of results to support clinical evaluation and management of the neonate and child. Also, the training includes understanding basic information of infectious diseases, AMR and inflammatory marker CRP kinetics as well as cases where a low CRP might need cautious interpretation, such as a fever lasting less than 24 h. Clinical decision flow charts are provided and included in the HCWs training together with CRP performing standard operating procedures, these will help to guide antibiotic prescribing based on CRP results. While final management decisions remain at the clinician’s discretion, RAs will encourage HCWs to follow the CRP-guided algorithm for intervention randomized children. The CRP-guided algorithm was developed based on published evidence and adapted to reflect subgroup-specific clinical risk profiles and feasibility in Zanzibar. Given that CRP responses may vary across age groups [28], the selected thresholds were chosen to balance sensitivity and specificity and to support safe and practical implementation. Serial measurements were incorporated to support interpretation beyond single CRP values. Neonates delivered at the participating hospitals who are well-appearing upon clinical assessment, without danger signs but have one of the WHO-defined risk factors for early-onset sepsis, are admitted for observation and have a CRP POCT measured at 18–24 h of life. A measurement below 30 mg/L in neonates who are still well-appearing upon clinical assessment indicates that the neonate most likely does not need antibiotic treatment [29] and the parents are informed of seeking care if the neonate show any danger signs. CRP levels between 30 mg/L and 50 mg/L warrant caution and the neonate should remain admitted and be clinically assessed with a repeat CRP measurement after 6 h. CRP levels above 50 mg/L indicate that neonate requires antibiotic treatment for early onset sepsis. If the neonate at any time shows signs of neonatal infection, antibiotics will be initiated according to the Zanzibar Standard Treatment Guidelines for Children and Adolescents [23]. For neonates admitted with clinical signs of infection and suspected infection, antibiotic treatment is initiated regardless of randomization. In the intervention group CRP levels, including blood culture, are assessed at admission and reassessed after 18–24 h of treatment. If both CRP levels are below 10 mg/L, antibiotics are most likely not needed [30–32]. Treatment continues until the blood culture result is available. If negative, the clinician may consider a differential diagnosis and discontinuing antibiotic treatment if the neonate is well. CRP levels above 100 mg/L suggest severe disease and adverse outcomes following potential rapid disease progression. For children aged 6 months to 12 years admitted with fever, a history of fever and/or diarrhoea and assigned to the intervention group, CRP is measured at admission and again at 18–24 h. If both CRP levels are less than 10 mg/L, the disease is likely not severe and antibiotics are most likely not needed and can either not be started or discontinued. If the second CRP is between 10 mg/L and 50 mg/L, antibiotics might be needed depending on the trend of the CRP levels. If antibiotics have been initiated, the clinician may consider stopping them earlier. CRP levels above 50 mg/L suggest that antibiotics are likely needed and treatment should continue. HCWs are instructed to consider the clinical picture alongside the CRP value. With obvious symptoms of bacterial infection and affected general condition, antibiotic treatment may be initiated and continued based on clinical indication. HCWs uses CRP POCT for patients in the intervention group to guide diagnosis and treatment decisions (Figs. 2, 3 and 4).Fig. 2CRP intervention flowchart for neonates with risk factors of early-onset sepsis (EOS). Adapted from [19]. CRP values are expressed in mg/LFig. 3CRP intervention flowchart for neonates with signs of infection. Adapted from [19]. CRP values are expressed in mg/LFig. 4CRP intervention flowchart for children with febrile illness or diarrhea. Adapted from [19]. CRP values are expressed in mg/L CRP intervention flowchart for neonates with risk factors of early-onset sepsis (EOS). Adapted from [19]. CRP values are expressed in mg/L CRP intervention flowchart for neonates with signs of infection. Adapted from [19]. CRP values are expressed in mg/L CRP intervention flowchart for children with febrile illness or diarrhea. Adapted from [19]. CRP values are expressed in mg/L Study procedures Participants are screened for eligibility at the participating hospitals by a research assistant (RA) prior to routine examination by a clinician. Screening and recruitment will be conducted 24 h a day by trained RAs with clinical background working in shifts (morning, afternoon and overnight), ensuring that all eligible patients presenting at any time can be approached for participation. Parents/caregivers of eligible children are requested for informed consent, after having received oral and written information regarding the trial. Participants are then randomized to the intervention group (CRP POCT) or control group (usual care) by a responsible RA through the REDCap Randomization Module [33]. Case report forms (CRFs) are filled out by RA for all participants at the first consultation and every subsequent day the participant is admitted. Participants assigned to the intervention group are provided with CRP POCT during consultation with their clinician at admittance and after 18–24 h of treatment. The CRP POCT is carried out as a prick test using capillary blood. Children in the control group do not have the CRP POC test taken. Decision on the treatment of participants is exclusively the responsibility of the clinicians of the included hospitals. To assess adherence, all treatment decisions including deviations from CRP recommendations are recorded in CRFs. Reasons for deviation are documented for transparency and allow for subgroup analysis. During admission RA follow the enrolled children daily and register.Demographic details including sex, age, co-morbidities, admission detailsObstetric and neonatal background information including birthweight, relevant maternal or perinatal historyClinical history, including duration of clinical symptoms prior to consultingClinical symptoms, recorded using a structured checklist and including general symptoms (e.g., poor feeding, lethargy), respiratory symptoms (e.g., cough, nasal flaring, apnea), neurological signs (e.g., seizures, irritability), gastrointestinal symptoms (e.g., vomiting, diarrhea) and other system-specific featuresClinical assessments and vital signs including oxygen saturation, respiratory rate, blood pressure, heart rate, body temperature and other examination findingsLaboratory and diagnostic investigations, including blood tests, cultures (e.g., blood, urine, cerebrospinal fluid) and other relevant investigations performed during admissionClinical management, including antibiotic prescriptions (type, dose, route, frequency and duration), as well as other medications and supportive therapies (e.g., intravenous fluids)Additional advice: preventive measures for themselves and their family members, time-off, home isolation, hygienic measuresReferral to other hospitals Demographic details including sex, age, co-morbidities, admission details Demographic details including sex, age, co-morbidities, admission details Obstetric and neonatal background information including birthweight, relevant maternal or perinatal history Obstetric and neonatal background information including birthweight, relevant maternal or perinatal history Clinical history, including duration of clinical symptoms prior to consulting Clinical history, including duration of clinical symptoms prior to consulting Clinical symptoms, recorded using a structured checklist and including general symptoms (e.g., poor feeding, lethargy), respiratory symptoms (e.g., cough, nasal flaring, apnea), neurological signs (e.g., seizures, irritability), gastrointestinal symptoms (e.g., vomiting, diarrhea) and other system-specific features Clinical symptoms, recorded using a structured checklist and including general symptoms (e.g., poor feeding, lethargy), respiratory symptoms (e.g., cough, nasal flaring, apnea), neurological signs (e.g., seizures, irritability), gastrointestinal symptoms (e.g., vomiting, diarrhea) and other system-specific features Clinical assessments and vital signs including oxygen saturation, respiratory rate, blood pressure, heart rate, body temperature and other examination findings Clinical assessments and vital signs including oxygen saturation, respiratory rate, blood pressure, heart rate, body temperature and other examination findings Laboratory and diagnostic investigations, including blood tests, cultures (e.g., blood, urine, cerebrospinal fluid) and other relevant investigations performed during admission Laboratory and diagnostic investigations, including blood tests, cultures (e.g., blood, urine, cerebrospinal fluid) and other relevant investigations performed during admission Clinical management, including antibiotic prescriptions (type, dose, route, frequency and duration), as well as other medications and supportive therapies (e.g., intravenous fluids) Clinical management, including antibiotic prescriptions (type, dose, route, frequency and duration), as well as other medications and supportive therapies (e.g., intravenous fluids) Additional advice: preventive measures for themselves and their family members, time-off, home isolation, hygienic measures Additional advice: preventive measures for themselves and their family members, time-off, home isolation, hygienic measures Referral to other hospitals Referral to other hospitals Detailed variables and response options are provided in the Case Report Form (Supplementary File 1). Furthermore, HCWs in participating hospitals answer a questionnaire on professional education as well as acceptability of CRP POCT. The questionnaire was developed by the study team based on objective and existing literature and has been previously described in our feasibility study [19]. Follow-up by phone After discharge from the hospital, all participants are contacted on the 7th and 28th day by phone calls from a RA who is blinded to the group allocation. Information gathered on the follow-up call includes general medical history of the child post discharge, additional diagnostic testing (with pictures of medical records or clinical follow-up if possible), complications (such as hospitalization, days in hospital, ICU stay, medication/oxygen/ventilation received, pneumonia, or death) and any medicine bought over-the-counter or prescribed by other health care workers (including the name of the drug, route of administration, frequency of administration and dose). Primary study outcome The primary study outcome is the duration of antibiotic treatment (number in 24-h periods between start and end of antibiotic treatment) within 14 days of the admittance of included neonates and children in each study arm (superiority analysis). Antibiotic use in this study includes administration via any systemic route (e.g., intravenous, intramuscular, or oral), as recorded in the case report form, excluding topical treatments. Antibiotic use during admittance and a composite measure of antibiotic use during admittance and discharge are reported. For every participant, the types and quantity of antibiotics given during the study period will be assessed based on individual drug administration records. To ensure patient safety in case of reduction in antibiotic use a second primary outcome relapse of infection is measured as reinstitution of antibiotics within 72 h after completion of the initial course of antibiotic treatment. The secondary study outcomes are:Antibiotic use (prescribed daily doses)Choice of antibiotic drug and dose compared to local standard treatment guidelinesNumber of days admitted in hospitalNumber of days until resolution of symptomsDevelopment of clinical signs of sepsisProportion of cases with blood cultures collected before antibiotic treatmentBacterial pathogens and AST in blood culturesHospital re-admittance within 14 daysMortality at follow up day 7 and 28 Antibiotic use (prescribed daily doses) Antibiotic use (prescribed daily doses) Choice of antibiotic drug and dose compared to local standard treatment guidelines Choice of antibiotic drug and dose compared to local standard treatment guidelines Number of days admitted in hospital Number of days admitted in hospital Number of days until resolution of symptoms Number of days until resolution of symptoms Development of clinical signs of sepsis Development of clinical signs of sepsis Proportion of cases with blood cultures collected before antibiotic treatment Proportion of cases with blood cultures collected before antibiotic treatment Bacterial pathogens and AST in blood cultures Bacterial pathogens and AST in blood cultures Hospital re-admittance within 14 days Hospital re-admittance within 14 days Mortality at follow up day 7 and 28 Mortality at follow up day 7 and 28 Withdraw of individual participants Participants are entitled to withdraw from the study at their discretion without any negative repercussions. The investigator may remove a participant from the study in urgent medical situations. If any participant chooses to leave voluntarily or is removed, the reasons are recorded in the clinical registry form and all data collected up to the point of withdrawal are included in the study analysis. Safety reporting The trial registers serious adverse events (SAEs) including death and hospitalization after discharge until the trial’s conclusion. SAEs are captured in the patients’ files and in the follow up forms on days 7 and 28 after discharge. In the event of any SAEs, the investigator assesses the event and determines its causality (certainly, probably, possibly, improbable, unrelated) and reports it to the Data Safety Monitoring Board (DSMB) within one week. The DSMB provides independent oversight and reviews safety data throughout the study. Data Safety Monitoring Board (DSMB) A Data Safety Monitoring Board (DSMB) has been established to conduct interim analyses, including the evaluation of SAEs, which will be presented and discussed. The primary duties of the DSMB include regularly reviewing and assessing the accumulated study data for participant safety, study progress and, when applicable, efficacy and providing recommendations to the Principal Investigator regarding the continuation, modification, or termination of the trial. The DSMB considers both study-specific data and relevant background information about the disease, test agent or patient population under investigation. Members of the DSMB are independent clinicians and researchers from Denmark and Tanzania not directly involved in the current study. Data analysis plan Power calculations For the superiority primary outcome, the sample size calculation is based on an estimated 20% reduction in duration of antibiotic treatment. Assuming the mean duration of antibiotic treatment of 7 days (168 h) and an SD of 5 days (120 h), with an α of 0.05 and a β of 0.1. The trial needs 262 patients in each group (524 patients in total). However, some patients may be discharged before reaching the stopping rules and these patients would not be stopped according to the CRP guidelines. We assume that 20% of patients are going to be discharged before the stopping rule was enacted or lost to follow-up. To ensure enough power for independent analysis for the three study population groups we apply this power calculation for each group. Furthermore, for children 6 months to 12 years further inflation of sample size by 50% is done to ensure exploration of sub-groups with different symptoms of infection (e.g. pneumonia, diarrhoea etc.). To establish non-inferiority of the second primary outcome of relapse of infection, assuming event rate of 5% of relapse, non-inferiority margin of 6%, one-sided alpha of 2.5%, power of 90% the sample size is 278 per group i.e. total 556 children plus 20% loss to follow-up equaling 646 children. Hence, the study aims to include 646 neonates at risk of infection, 646 neonates with suspected infection and 969 children aged 6 months to 12 years with infection—an overall total of 2261 children included. Power calculations were performed using sample size calculators provided by Sealed Envelope [34]. Analysis of outcomes Data analysis will include comparing baseline variables between the intervention and control groups using X2 tests for categorical variables and t-tests for continuous variables, without adjustment. Baseline characteristics such as age, sex, urban/rural residence, number of children and study site will be compared. Separate analyses will be conducted for all three sub-groups neonates at risk of infection, neonates with signs of infection and children aged 6 months to 12 years. The primary outcome duration of antibiotic treatment (days), which is a continuous variable, will be compared between the intervention and control groups using a multivariable linear regression model. Neonates at risk of early onset sepsis are recommended for two days of antibiotic treatment by the WHO. Therefore, in this subgroup, an additional non-inferiority analysis will be conducted using a multivariable logistic regression model, comparing CRP POCT with prophylactic two-day antibiotic treatment on the binary outcome of antibiotic prescribed yes/no. All comparisons will be conducted using both unadjusted and adjusted models for sex, age, and clinical status, following an intention-to-treat (ITT) approach. To account for deviations from CRP-based recommendations, a secondary Per-Protocol (PP) analysis will be conducted, excluding participants who did not adhere to their assigned intervention. For neonates at risk of early-onset sepsis, a non-inferiority comparison will be performed per protocol, assessing whether CRP-guided management is not inferior to standard clinical decision-making. Adherence to CRP-based recommendations will be systematically recorded and analyzed to evaluate its impact on treatment decisions by calculating proportion of adherent decisions. Statistical analyses will be performed after entering data from the final follow-up and data will be stored in a REDCap database for analysis using STATA. Trial status The study commenced on 4th March 2024 and aims to achieve the desired sample size within a 13-month period, having begun patient inclusion on this date. Discussion The CRP-POCT randomized controlled trial could bring positive impact to health care in a low- and middle-income country, where such testing is not commonly practiced. The test has the potential to enhance clinical decision-making by offering real-time inflammatory marker values, thereby improving the accuracy of diagnoses and reducing unnecessary antibiotic prescriptions, which in turn minimizes risks such as AMR. Petel et al. [35] reported that CRP-guided strategies can reduce antibiotic use and shorten treatment duration without compromising safety. This more targeted approach can lead to better patient outcomes, including shorter hospital stays and fewer antibiotic-related complications. However, CRP-POCT has limitations that should be considered. CRP is a non-specific inflammatory marker and may be elevated in non-bacterial conditions, potentially reducing diagnostic specificity [36]. To mitigate this, CRP-POCT was performed serially and complemented by microbiological culture and AST to support pathogen identification and guide antibiotic selection. CRP levels may also be influenced by underlying comorbidities; this was partly addressed through the study’s exclusion criteria. Additionally, variability in CRP cut-off thresholds may affect clinical interpretation. In this study, interpretation was guided by cut-offs derived from the existing literature [29–32]. The reduced misuse of antibiotics could contribute to slowing the spread of AMR in the community, addressing a pressing public health issue, particularly in low- and middle-income countries, as highlighted in the WHO GLASS report [37], while also potentially promoting greater trust in evidence-based pediatric care. Goodbout et al. [38] have shown that antimicrobial stewardship programs improve prescribing practices and reduce inappropriate antibiotic use. C-reactive protein testing at the point of care could reduce the strain on laboratory services, promote resource optimization on appropriate selection of children who would benefit from culture and antimicrobial susceptibility testing and may help reduce overall health care costs by limiting unnecessary treatments and reducing hospital readmissions. Keitel et al. [39] reported that point-of-care testing supports quick clinical decision-making in pediatric infectious diseases. Moreover, the trial’s findings could inform national antibiotic stewardship guidelines, potentially leading to widespread adoption of CRP-POC testing in routine care, thereby aligning with global efforts to control AMR and improve pediatric health care practices, particularly in resource-limited settings, where Cals and Ebell [40] demonstrated that CRP-guided prescribing can safely reduce unnecessary antibiotic use. C-Reactive Protein is already recognized in the Zanzibar Standard Treatment Guidelines for Children and Adolescents [23], in alignment with the WHO Essential Diagnostics List [41], but is not yet widely used as a point-of-care test. Scaling up CRP-POCT will require integration into routine clinical workflows, targeted training of health care workers and the establishment of reliable supply chains for CRP test kits and related consumables. Additionally, clear implementation strategies and supportive supervision will be important to ensure appropriate and sustained use in clinical practice. Supplementary Information Supplementary Material 1: Case Report Form for daily clinical data collection. Supplementary Material 2. Supplementary Material 1: Case Report Form for daily clinical data collection. Supplementary Material 2.