The lack of access to surgical care in low-and middle-income countries (LMIC) requires urgent attention. Currently, around 1.7 billion children and adolescents do not have access to potentially life-saving surgeries (1). In sub-Saharan Africa (SSA), paediatric disease is rife, with under-five mortality rates 15 times higher than in high-income countries (HIC) (2). Where paediatric surgery is available, outcomes in SSA are considerably worse than in HIC for common conditions in infants, such as repair of gastroschisis (75.5% mortality in SSA vs. 2.0% in HIC) and repair of anorectal malformation (11.2% mortality in SSA vs. 2.9% in HIC). Common surgical procedures such as paediatric hernia repair in SSA also result in higher and avoidable mortality (3).

In SSA, shortages have been reported for all members of the specialist surgical team (4). For example, in Tanzania, the current ratio of surgeons, anaesthesiologists and obstetricians (SAO providers) is 0.46 compared to the recommended ratio of 20 SAOs per 100,000 population (5). In SSA, there is an estimated one paediatric surgeon for every six million children aged 0 to 14 years (6). Outcomes for specific paediatric surgical conditions in LMICs correlate with paediatric surgical workforce density, as demonstrated in a recent systematic review, where a paediatric surgical workforce density greater than 0.4 per 100,000 children below 5yrs was significantly correlated with increased odds of survival in surgery undertaken for gastroschisis, oesophageal atresia, intestinal atresia and typhoid perforation (7). In many countries in SSA, front-line paediatric healthcare is provided by medical staff with no specialist paediatric training (8).

Malawi, Tanzania and Zambia (MTZ) are three countries in SSA that have a combined population of 92 million, of which at least 50% are under the age of 18 (9). A recent study in Malawi reported that 26% of children (approximately 2 million) live with a condition treatable with surgery (10). Similar data for Zambia and Tanzania has not yet been collected.

In order to inform the scale-up of paediatric surgical services, available data on existing services and resources at district hospitals in SSA need to be collected and analysed to enable local authorities to make informed decisions on how best to maximise the utilisation of scarce resources to meet population needs (11). However, little systematic evidence has been collected at the facility level (12), and no study has been done comparing the capacity of district facilities across countries. This study aimed to examine and compare the capacity of district hospitals in three SSA countries (MTZ) to provide paediatric surgical care, using tools specifically designed to measure the capacity components (13) so as to establish a baseline and inform policy and planning at country level.

We conducted a cross-sectional study using the PediPIPES surgical capacity assessment tool (13) which is an extension of the PIPES tool (14). Several studies have endorsed PIPES as a valid and reliable instrument to measure surgical capacity in resource-constrained settings (15). The PediPIPES tool measures a health facility's capacity to deliver emergency and essential surgical care (EESC) to infants and children under 18 years (13, 16) by examining the availability of procedures, personnel, infrastructure, equipment and supplies. The PediPIPES survey components include the following numbers of variables: procedures 46, personnel 7, infrastructure 17, equipment 22, and supplies 26 variables. Numbers are recorded of available staff, incubators, paediatric ventilators, paediatric beds and operating rooms. For other variables, a binary value of 1 is given where the item is reported adequate or always available, and where a surgical procedure is actually conducted at the hospital and a score of 0 if the item is unavailable/inadequate or where a surgical procedure is not conducted. The tool has no maximum score for the number of available personnel, operating rooms, incubators or ventilators. An overall hospital surgical capacity index is computed based on individual variable scores, which is then used to compare facilities and monitor trends over time.

The overall PediPIPES Index scores were similar across the three countries. Zambia (Mean [M] = 5.84, range: 4.92–7.29, standard deviation [SD] = 0.67) and Malawi (M = 5.85, range = 3.73–8.14, SD = 1.02) scored the lowest on average. Tanzania scored the highest (M = 6.24, range: 4.07–7.63, SD = 0.95). Differences between countries were not statistically significant (p = 0.231). The mean index score for faith-based hospitals in Tanzania and Zambia combined was 6.25 (range = 5.00–7.63, SD = 0.76), compared with 5.91 (range = 3.73–8.14, SD = 0.94) for government owned hospitals. This difference was not statistically significant (p = 0.203).

Despite the attention drawn to global surgery by the 2015 Lancet Commission on Global Surgery (LCoGS) report, few initiatives have specifically focused on paediatric surgery (19). Also, the LCoGS report did not include any paediatric-specific indicators to monitor progress in global surgery, although half of the population in SSA are children (20). This has contributed to the apparent neglect of the burden of surgical conditions among children, especially in LMICs (21). Yet, in our 2017 study of cases that presented to eight district hospitals in Malawi and nine in Zambia, over a 24-month period, children 0 to 15 years accounted for two-thirds of trauma cases and 40% of hernias in girls that were managed in operating theatres (8). Hence, the scale of paediatric surgical care undertaken at district hospitals in some SSA countries is being ignored as much as neglected.

The first step in addressing this priority is to assess district hospital capacity to deliver paediatric surgical care safely. Our study provides evidence not previously available in SSA, based on a sample of 67 district-level hospitals in three African countries. It systematically collected data using a standardised tool employed by other studies (13), which can be replicated over time to monitor trends. Despite limitations of the tool—see below, it provides evidence on some of the strengths and weaknesses of country-level responses to this burden by triangulating the findings with other published evidence (18). Our study reported an overall higher PediPIPES index score than in a similar study in 39 district level facilities in Pakistan (22). The paediatric surgical capacity of district hospitals in our study, for the respective PIPES components, was similar to a study in West Africa, where 32 of 37 surveyed hospitals were tertiary, two were secondary, and district, private and mission hospitals were one each (13).

While the overall PediPIPES index scores were similar—5.8 in Zambia and Malawi and 6.2 in Tanzania, these scores masked some notable country-specific patterns and differences in the specific components. Infrastructure, equipment and supplies, including those specific to the surgical needs of children, were more widely available in district hospitals in Tanzania. Whereas, hospitals in Malawi, which scored lowest in these components, reported more widespread capacity to conduct a range of paediatric procedures. This could reflect the efforts to improve district-level surgical care in Malawi through innovative surgical training and supervision programmes designed specifically to address the population's surgical needs in rural areas (23).

The overall pattern of paediatric interventions and procedures reported for district hospitals across the three countries is similar and plausible, considering no paediatric surgeons were reported to practise at district level. Most district hospitals undertake minor emergency interventions and elective cases, especially covering orthopaedic procedures for trauma and abdominal surgery, consistent with our 2017 study (8). Surgical interventions are not reported for more complex congenital cases, which usually are referred for specialist care at higher level hospitals (24). However, the surgical referral systems in LMIC also face significant challenges (25). Future interventions tackling shortages of paediatric surgical capacity at district level must also address deficiencies along the patient referral journey.

Lack of trained staff has been considered one of the most important barriers to deliver safe surgery to children (26). In our study, there were medical doctors performing surgery on children. Although they were very limited in numbers in Malawi and Tanzania, they were present in 85.7% of Zambian hospitals. Only 5 of 22 sampled hospitals in Malawi reported a cadre (general medical officers) who undertook surgery on children and none reported anaesthesiologists or nurse anaesthetists that provided anaesthesia. Other studies done in the region showed that the bulk of surgical care at district level is delivered by medical doctors working alongside non-physician clinicians (27). Indeed, most of the district-level surgical and anaesthesia providers in Malawi, and many in Zambia are non-physician clinicians (NPCs) and some do manage paediatric surgical cases (8, 28). However, the PediPIPES tool in its current form is not designed to capture these cadres. Hence, the first lesson from this study is the need to modify facility assessment tools to ensure that paediatric surgical capacity surveys take account of the realities of who is delivering surgical and anaesthesia care to children at district hospitals in SSA.

Essential supplies such as paediatric size nasogastric tubes, urinary catheters and chest tubes were particularly lacking in Malawi, where more hospitals reported undertaking a wider variety of cases. Such shortages have a major negative effect on the readiness to perform surgery. A study in Malawi demonstrated that surgical supply shortages contributed to 51% of cancelled elective procedures (29). When essential infrastructure is unavailable, surgical outcomes are compromised (30). The evidence from our earlier study (18) suggests that district hospitals in Malawi, Zambia and Tanzania experience shortages of infrastructure, equipment and supplies, needed to safely deliver elective and emergency surgical care and administer anaesthesia to adults. The evidence from this study demonstrated that shortages in equipment and supplies specific to paediatric surgery are greater. This is worrying because children in need of surgical care undoubtedly present to these settings (8).

Since the inclusion of essential and emergency surgery as a key component of Universal Health Coverage (31, 32), and the work of the Disease Control Priorities 3 (DCP3) Country Translation project (32), several countries have included surgery in their essential packages of health services. In addition, countries with relatively high enrolment rates into health insurance schemes have included surgery in the health benefit packages of these schemes [e.g., Ghana, Kenya, Rwanda (33–36]). However, these packages do not specifically mention paediatric surgery, and paediatric surgery continues to be absent from most National Health Strategic Plans in SSA, despite the majority of the population being young (36).

Also, a list of cost-effective, essential surgical procedures for children is yet to be defined and included in essential health packages and national surgical and health plans. Of equal importance is the need to define the health systems level (national, referral and district hospital) at which each procedure should take place and the capacity conditions that need to be in place to do so safely (37). National surgical plans need to define the paediatric surgical responses and capacities that are required at all levels of the health system and, based on these, agree on lists of paediatric surgical procedures that are allowed and should be undertaken at national, referral and district hospital levels, the latter being the first level of surgical care in SSA (38).

The Global Initiative for Children's Surgery (39), an international interest group, has been advocating for inclusion of children's surgical care in national health and surgical plans. They have developed practical recommendations on how to do so, including general guidance on optimal resources for children surgery (40), to be adapted to each country's context. We support their recommendations, which are reinforced by the results in this paper, which have highlighted critical surgical capacity gaps at district level hospitals in MTZ, particularly around the paediatric surgical workforce. Countries such as Zambia and Tanzania, where the current NSOAP has (Zambia) or is about to expire (Tanzania), are well placed to integrate a district hospital response to children's essential and emergency surgical needs in the next NSOAP planning cycle.

Our study has several limitations. Firstly, there is potential selection bias and limited generalisability. In Tanzania, only hospitals in the Northern Zone (one of seven zones, countrywide) were surveyed, for feasibility reasons (17). Sampling in Malawi and Zambia was country-level representative. Secondly, as with previous studies that used the PediPIPES tool, facility staff-reported data were collected however, recall or ascertainment bias was reduced by having a minimum of two respondents for each facility. Thirdly, the facility respondents, at least two of whom worked in the facility's operating theatre, generally lacked specific paediatric surgical training, which could have impacted on the accuracy of their responses. The research team, however, had 5 + years prior experience of researching district hospital surgical facilities in two of the three studied countries, enabling the team to detect and resolve obvious inaccuracies during data cleaning, analysis and manuscript writing. Finally, the PediPIPES tool has deficiencies when assessing surgical capacity at district hospitals in SSA in particular, it does not capture non-specialist and non-physician cadres who undertake surgical interventions in children in these settings. One of the recommendations in this paper points to the need to adapt and utilise an improved tool for future use in SSA countries.