In this study, we present the first dataset on pediatric emergency visits to an OPD of a district-level hospital in rural Tanzania.

Overall, we observed a hospitalization rate of approximately 12% among the OPD, which was lower than the rates previously reported in other studies [16,17,18,19]. Instead, the mortality rate (0.4%) was higher than reported in high-income countries (1.5/100,000 visits) [20], but it was comparable to the rate found by Enyuma et al. [16], where in their study (0.5%), and even lower compared to other sub-Saharan countries like Nigeria where the rates was ranged 2-17.5% [21,22,23,24,25] or Cameroon where Chiabi et al. found a mortality rate of 1,6 − 1,9% [19] and Chelo et al. found a rate of 5,76% [26]. Our findings revealed a higher risk of hospitalization for the admissions to OPD among patients without insurance, those from more distant districts, those attended by a non-paediatric trained staff and those accessing the OPD for burn/wound, cardiovascular, constitutional/malnutrition, fever, gastrointestinal, respiratory, ingestion/inhalation, injury. Conversely, we observed a lower risk of hospitalization when the age of children increased.

The findings suggest that living in suburban or otherwise remote areas, far from health centers, is a risk factor for hospitalization. Juran et al. found that 92.5% of the sub-Saharan African population lived within 2 h of a major hospital for surgical procedures [27]. The increased risk is likely due to the greater distance to the hospital, which may lead to a delayed access to OPD for the children. Manongi and colleagues in Tanzania found that hospitalised children who came from areas less than 3 h away from the hospital had a mortality rate of 3.4% compared to 8.0% for children who came from areas more than 3 h away [28]. These delays can be attributed to socioeconomic factors related to high costs of transportation, inadequate transportation infrastructures and vehicles, which in turn may result in the exacerbation of the disease not treated in a timely manner. This observation may also reflect the large number of challenges in traveling to the hospital that children face even when they are seriously ill [29]. A WHO report on quality of care recommends timely referral for every child with conditions that cannot be managed effectively at first-level facilities [30].

Previous studies highlighted that the reasons for late presentations at pediatric emergency units includes poor identification of early sign of disease severity by care givers of low level health facilities, high costs of hospital treatments as patient pay out of pocket, poor health seeking behaviors and beliefs in remedies with unproven efficacy [31, 32].

Although challenging and costly, the establishment of an efficient emergency call and ambulance transport service would be crucial to ensure that critically ill children from both suburban and urban areas have timely access to OPD. Safe transport is required by many sick children seen in primary care facilities to referral hospitals. However, implementing such a service may not always be feasible, and this need to be balanced against the potential risk of transport, distance to referral hospital, costs and the needs of other patients. In Tanzania, transportation from primary care health posts to hospitals is even more difficult, and other modes of emergency transport are employed; these included bicycles with trailers, tricycles with platforms, motorboats and ox carts [33].

Another relevant observation of the present study is that owning health insurance is a protective factor for hospitalization risk. The existing health insurance schemes in Tanzania only cover medical costs at healthcare facilities, but do not compensate patients for travel and time costs incurred in accessing care, resulting in financial burden on households and delayed access to care [34]. The health financing system in Tanzania is highly fragmented involving different resource providers including general taxation (34%), private donors, non-governmental organizations (NGOs), foreign states etc. (36%), direct payments (22%) and health insurance contributions (8%) [35]. According to a 2018 analysis of the health sector only 33% of Tanzanians are covered by health insurance, leaving the remaining two-thirds 2/3 of population exposed to financial arising from direct health care payments [35]. This finding could be partly correlated with the above, assuming that the family with health insurance is in better economic condition and thus can more easily afford transportation to the hospital. In addition, by not having to pay for the health service, the parent would tend to bring the child in early for a medical examination as shown by Huang at al. in Taiwan [36], with opportunities then for health personnel to intervene early on the ongoing condition. Moreover, we can hypothesize that the higher level of education of families with health insurance may also play a role, being more aware of the warning signs/symptoms of the child’s pathology. As demonstrated by Agelebe at al [37]., children who were socially disadvantaged presented significantly later to the hospital than their non-socially disadvantaged counterparts following onset of illnesses. Low utilization of healthcare services due to delay in making decision and delay in assessing medical services is because of the ripple effects of unemployment and poverty [37].

Our study showed that children were more likely to be hospitalized if they had cardiovascular, constitutional, neurological, gastrointestinal, respiratory symptoms, or had burns. In the literature other studies in LMICs report found similar rates. Specifically, in LMICs the prevalence of preventable communicable diseases (such as malaria, pneumonia and diarrheal diseases) and acute and chronic malnutrition is high in sub-Saharan African Countries [19, 21, 26, 38,39,40,41].

In addition to the high burden of pediatric infectious disease, there is at present an increasing incidence of non-communicable and hereditary diseases and their complications, that require special care by specialized personnel. Such symptomatology in the pediatric patient takes on different peculiarities and complexity than in the adult one, thus emphasizing that specific expertise is needed [16, 23, 42,43,44].

Our study supports this hypothesis, as it indicates that visiting an outpatient clinic not specifically dedicated to pediatric care is predictive of an increased risk of hospitalization. Indeed, our findings may suggest that better management of the pediatric patient even at the OPD level could reduce avoidable hospitalizations, which can cause stress and economic burden to families, as well as increase the risk of hospital-acquired infections. Several studies have shown that children that arrived at emergency departments during off hours tend to experience longer stays [21, 45–46]. This is attributed to the unavailability of highly skilled personnel and certain investigations during such periods of the day [31, 32]. Also, healthcare personnel trained in pediatric emergency medicine principles are shown to reduce childhood mortality in LMICs likely through dissemination of education, practice patterns, and advocacy measures [47].

Training health personnel in early identification of critical illness is a crucial step in improving disease prognosis, as shown by several studies [48, 49]. Priority should be given to training in the early recognition and management of pediatric conditions that most commonly lead to death in the local area. This issue holds particular significance in LMICs [50], such as Tanzania, where, like many sub-Saharan Africa countries, there is a shortage of trained healthcare professionals with an estimated 3 doctors and 39 nurses per 100 000 inhabitants in the country [51, 52].

Early triage assessment and prompt identification of signs of critical illness, and rapid initiation of appropriate treatment should be top priorities for all hospitals providing emergency care for children [53]. However, according to the study of Ardsby et al. [54], only 9.1% of the hospitals reported specific triage protocols for children < 5 years of age. Inadequate triage is a widespread problem in LMICs and represents an important challenge in addressing emergency conditions in hospitals [55, 56].

Finally, our study revealed that the children admitted to hospital were in average younger than not admitted (about 10 months older), a result in agreement with what other studies have observed [18]. However, in our study are included children who came for follow up for post-therapy, post hospitalization or anthropometric and neurodevelopmental assessment (12.3% of the total population examined).

This study provides information about pediatric emergency visits to an OPD of a district-level hospital in Tanzania on a large sample of children, a still under-researched field of research that needs as much work as possible for reduce preventable deaths.

Moreover, the availability of data collected through dedicated software is another strength of our study. Optimization of data recording is an important area on which to focus resources to obtain increasingly accurate data. The data collection was done through the new HMIS introduced in early 2022 at Tosamaganga Hospital. The availability of this software allowed to collect good quality information in terms of accuracy, reliability, consistency. In fact, the current data recording system that is used in the hospital allows for good data recording with minimal gaps that are likely to be reduced over time. The HMIS proved to be effective and easily usable in the data collection process, confirming that having better data quality simplifies the monitoring of health care delivery and evaluation of the impact of health interventions [57].

This is a single-center study of a district-level hospital; it may not represent the rest of Tanzania or other LMICs, but it is reasonable to assume that our data provide some useful information about pediatric emergency care at the local and national levels that can help optimize the distribution and use of resources, as well as plan more appropriate feasible and effective interventions to improve pediatric emergency care within an integrated system of care.

For some variables we could not collect information for all subjects because they had not been registered in the electronic database, which became operational just in January 2022, when the study started. Furthermore, we could not collect information on the last three months of 2022, a limitation that did not allow to assess whether there are seasonal differences on the outcome. However, we performed analyses by month of admission to OPD.

Moreover, data on comorbidities different from those collected are not reported in the OPD database, thus in the inpatient database they are mentioned inconsistently and, therefore, could not be analyzed.

This would be extremely valuable information to include in multivariate analysis, as underlying conditions such as malnutrition, HIV, and other infectious or chronic diseases have been shown to be associated with need for hospitalization and mortality [58, 59]. Systematic collection of major comorbidities should be pursued to be able to properly interpret the data to improve care and optimize organization and resource utilization.

Finally, it was not possible to define the aetiological diagnosis of the diseases due to instrumental and laboratory diagnostic limitations.