A Journey Undertaken by Families to Access General Surgical Care for their Children at Muhimbili National Hospital, Tanzania; Prospective Observational Cohort Study

Background A majority of the 2 billion children lacking access to safe, timely and affordable surgical care reside in low-and middle-income countries. A barrier to tackling this issue is the paucity of information regarding children’s journey to surgical care. We aimed to explore children’s journeys and its implications on accessing general paediatric surgical care at Muhimbili National Hospital (MNH), a tertiary centre in Tanzania. Methods A prospective observational cohort study was undertaken at MNH, recruiting patients undergoing elective and emergency surgeries. Data on socio-demographic, clinical, symptoms onset and 30-days post-operative were collected. Descriptive statistics and Mann–Whitney, Kruskal–Wallis and Fisher’s exact tests were used for data analysis. Result We recruited 154 children with a median age of 36 months. The majority were referred from regional hospitals due to a lack of paediatric surgery expertise. The time taken to seeking care was significantly greater in those who self-referred (p = 0.0186). Of these participants, 68.4 and 31.1% were able to reach a referring health facility and MNH, respectively, within 2 h of deciding to seek care. Overall insurance coverage was 75.32%. The median out of pocket expenditure for receiving care was $69.00. The incidence of surgical site infection was 10.2%, and only 2 patients died. Conclusion Although there have been significant efforts to improve access to safe, timely and affordable surgical care, there is still a need to strengthen children’s surgical care system. Investing in regional hospitals may be an effective approach to improve access to children surgical care.


Introduction
Access to health care is one of the most basic human rights supported by the Universal Declaration of Human Rights [1]. Indeed, the 2030 Agenda for Sustainable Development, approved by the United Nations in 2015, includes the key health-related target (Sustainable Development Goal 3.8) of universal health coverage (UHC) [2]. However, more than 94% of the global population lacking access to safe, timely and affordable surgical are from low-and middleincome countries (LMICs), and majority are children [3]. Approximately, 88 million individuals incurred catastrophic expenditures from seeking surgical and anaesthesia care [3]. Children represent the majority of the population in Tanzania and other LMICs [4,5]. Failure to meet surgical needs to children may be a barrier to achieving the UHC and advancing the human rights agenda [6].
There is paucity of data, especially in LMICs, on access to children's surgical care and related contextual challenges [7][8][9][10]. Over 70% of Tanzanians reside in rural setting and are served by district hospitals [11]. The ongoing developments to ensure access to essential surgical care even in district hospitals may not always address surgical needs for children. In Tanzania, Muhimbili National Hospital (MNH) is one of the centres in the country with developed capacity to provide children surgical care -in terms of available workforce and infrastructure -but is considerably far from some places where it receives referrals. Inadequate data on patients' journeys to access surgical care limit strategic design and implementation of policies for improvement [12]. Much of the data currently being used are from the Global North, where the situation is vastly different, hence may fail to reflect and realise paediatric surgical care needs in Tanzania [13]. This study, therefore, aimed to assess the journey patients make to receiving general paediatric surgical care at MNH and explore other Lancet Commission on Global Surgery (LCoGS) indicators related to safety and cost burden of healthcare.

Methodology
A prospective observational cohort study was undertaken from 2019 to 2020 at MNH in Dar Es Salaam (Coastal region of Tanzania). This is a tertiary national referral hospital capable of providing care for complex surgical conditions, receiving diversity of paediatric surgery patients from all over the country. The centre has two paediatric operating theatres rooms and a 60-bed paediatric surgery ward.
We randomly included patients undergoing elective or non-elective general surgery at MNH, aged 11-year-old or younger, and whose parent/caregiver consented on their behalf for participation and follow-up. We excluded those needing cardiac, trauma, neuro and plastic surgery as they are treated in separate respective institutes or units of MNH. Participation was voluntary and did not impact or change the care that they were receiving.
Collection of demographics, clinical and follow-up data was done by two study coordinators, both medical doctors and registrars at the department of paediatric surgery with a minimum of 2 years of experience. They had undergone the necessary research governance and ethics training for data collection. A Swahili structured questionnaire was used to interview and collect information from the parent/caregiver from the onset of the child's symptoms to 30-days post-operatively. This included participant and caregiver demographics, time to seeking, reaching and receiving surgical care, referral pattern, mode of transportation, insurance status and dates of admission, surgery and discharge/death. Distances travelled (km) from home to a referring health facility and/or to MNH were estimated by using Google Maps (https://www.google.com/maps): a free online tool which has been reported to be an accurate way of assessing distances [14,15]. We used Clavien-Dindo system for grading adverse events (i.e. complications) which occur because of surgical procedures [16]. Out-of-pocket (OOP) expenditure incurred by the patient's family on their journey to receiving care were collected in Tanzanian Shillings (TZS) and converted to US dollars (USD), a conversion rate of as of 12/07/2021 ($1 = 2319 TZS). A case report form (CRF) was used to collect data on patients' clinical information and outcomes, and patients were followed-up for 30 days while in the ward, or by phone and/or during clinic visits after discharge. Anonymous data were collected and stored in a secure REDCap database hosted by MNH that was accessible only to researchers.
Data were described in proportions for categorical variables, and medians and interquartile ranges (IQR) for quantitative variables. The Mann-Whitney Test and Kruskal-Wallis Test were used to determine differences Green Zone (green) = 16 participants. Zanzibar (purple) = 2 participants. Unknown = 1 participant. The black star represents the location of Muhimbili National Hospital between sub-groups where the explanatory variable was categorical, and the response variable was quantitative while Fisher's exact test for differences between subgroups where both explanatory and response variables were categorical. A multiple linear regression was calculated to predict time taken for a patient to present at MNH based on the distance of their home and referring centre from MNH. Data were analysed using Stata 15.1. We used the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement for observational

Demographics
A total of 154 children with a median age at admission of 36 (IQR: 18 -56) months participated in this study. The majority were from Coastal zone (n = 109/154, 71.2%) ( Fig. 1) and male (n = 99/154, 64.3%) ( Table 1). Appendix S1 details the regions-as per political divisions-the children resided in [18]. Most participants (n = 120/154, 77.9%) were referred from another health facility. Of these, 63.3% (n = 76/120) were from regional hospitals, with lack of paediatric surgery expertise being the main reason for referral (n = 106/120, 88.3%) (Fig. 1 ). Travel to a referring health facility could involve rough roads and/or tarmac roads. 13 patients had to travel on rough roads, where they spent a median time of 2 h (IQR: 1.5 -2). The medial time spent on tarmac roads was 1 h (IQR: 0.5 -3). A significant regression equation was found (F (2, 113) = 94.96, p \ 0.0001), with an R2 of 0.627. The predicted time (hours) taken is equal to 1.061 ? 0.003 (distance in km of their home from MNH) ? 0.018 (distance of their referring health facility from MNH). Time taken to present to MNH increased by 0.003 h for each km their home was from MNH (p = 0.070) and 0.018 h for each km their referring health facility was from MNH (p \ 0.001). The time taken from first symptom of current diagnosis to seeking care was significantly greater in those who selfreferred (median: 7 days; IQR: 2 -30) compared to those referred through a health facility (median: 2 days; IQR: 1 -14) (p = 0.0186). A significant difference was also noted in the time to seeking care between different caregivers' age groups (p = 0.0112) ( Fig. 4 and Appendix S4). The time taken to reaching care at a referring health facility was significantly greater in those who did not travel by a motorcycle (median: 1. Appendix S4). The time taken to reach care at MNH was significantly greater in those who used public transport (median: 5 h; IQR: 2 -11.5) compared to those who did a c d b Fig. 3 a Days taken from first symptom of current diagnosis to seeking care b Hours taken from deciding to seek care to reaching a referring health facility c Hours taken from deciding to seek care to reaching MNH among self-referred patients d Hours taken from deciding to seek care to reaching MNH among all patients. The results are segregated based on the participant's zone of residence. The height of the bars represents the median, and the error bars represent the interquartile range not (median: 3 h; IQR: 1.5 -5) (p = 0.0280) ( Fig. 5 and Appendix S4). Appendix S5 and S6 show the original location of patients who were able to present to MNH within 2 h of deciding to seek care and the factors affecting time taken to access care. The majority of participants were referred to MNH for elective surgery (n = 141/153, 92.2%) and had a pre-operative ASA score of 1 (n = 137/153, 89.5%). The incidence of post-operative SSI was 10.2% (15/147) ( Table 2). The time taken from first symptom of current diagnosis to seeking care from a healthcare provider was significantly shorter in those who had emergency surgery (p = 0.0198) (elective surgery -median: 3 (IQR: 1 -14) days; emergency surgery -median: 1 (IQR: 1 -4) day. There was no significant difference in the time taken to reach care at a referring health facility between those who had emergency surgery and elective surgery (p = 0.4361) (elective surgery -median: 1.5, IQR: 0.5-3 h; emergency surgery -median: 1.25, IQR: 0.5 -2 h). The time taken to present at MNH was significantly shorter in those who had emergency surgery (p = 0.0396) (elective surgery -median: 4.25, IQR: 2 -11 h; emergency surgery -median: 2.75, IQR: 1 -5 h). The median time from admission to receiving surgical care was 3 (IQR: 1 -14) days; all emergency surgery was conducted within a day.

Insurance Status
Most participants had insurance (n = 116/154, 75.32%). Children were significantly more likely to have insurance if they were undergoing elective surgery (n = 111/141, 78.7%) over emergency surgery (n = 4/12, 33.3%) a b Fig. 4 The days taken to seek care from first symptom of current diagnosis based on (a) the type of referral, and (b) the age of the caregiver. The height of the bars represents the median, and the error bars represent the interquartile range (p = 0.002). The median total out of pocket expenditure for receiving care at both referring health facility and MNH was $69.00 (Table 3). This was significantly greater among those referred through another health facility compared to those who self-referred to MNH (p = 0.0135). There was weak evidence that out-of-pocket expenditure was greater in those who did not have insurance (p = 0.0755) (insurance -median: $60.37, IQR: $17.90-$155.24; no insurance -median: $97.02, (IQR: $40.97-$232.86). Table 4 shows the relationship between self-reported socio-economic status and the various factors identified above to be significantly related to timely surgical access. Patients travel long distances, navigate a complicated referral system, and incur significant costs in seeking and receiving paediatric surgical care. More than two-thirds of children saw a healthcare provider at a referring health facility within 2 h, but approximately a third of all children reached a tertiary hospital (MNH) within 2 h of deciding to seek care. This is still in stark contrast to other low-resource settings, where approximately four-fifths of the population are unable to access surgical care within 2 h [19]. For those who required emergency procedures, 50% reached MNH within 2 h of deciding to seek care and were more likely to have post-operative complications.

Recommendations
Lack of paediatric surgery expertise was the main reason that 77.9% patients were referred from other healthcare facilities. This is a reversal of the 2008 findings that selfreferrals accounted for 72.5% of presentations at MNH for both surgical and non-surgical conditions [20]. Self-referral is thought to be associated with later presentation when the disease is more severe and worse prognosis. In our study, self-referring was associated with an increased time to seeking care. Similar findings have been reported in previous studies: a study in Uganda reported although 90% of participants were identified by family members to be suffering from an illness, only 14% sought medical attention immediately [21]. Existing benchmarks define paediatric surgical procedures that can be provided at various levels of healthcare based on resources, and these guide effective referral [22]. Based on these benchmarks, the majority of the referrals in our study needed tertiary level care. However, about 20.8% (32/154) of children had conditions (hernias, appendicitis and lipoma) which could have been treated at lower-level hospitals. Increased burden of managing these cases in tertiary hospitals may limit surgical care provision for both complex and common conditions [23]. If adequate resources are available, regional hospitals become the cornerstone of LMICs surgical care [23][24][25].
In addition to infrastructure developments, training by local and international providers need to be prioritized  [26]. This can be achieved by training multidisciplinary teams of children surgical providers [27] as well as including task-shifting and sharing [28,29]. Defining regional hospitals as centres for providing paediatric surgery and incorporating telemedicine may leapfrog physical barriers and surgical specialist shortages. This will ensure timely access to surgical care, reduce the number of preventable referrals and overcrowding at higher-level hospitals [30]. Unit costs and the relative shares of capital costs are generally lower at primary-level hospital [31]. Effective treatment depends on all steps of a healthcare system working harmoniously, from timely seeking and reaching healthcare, appropriate triage for surgery or referral, to proper transportation for care in an adequately resourced facility for better outcomes [32]. The modern concepts of improving value in healthcare emphasize the importance of considering value across the whole patient pathway from symptoms to care and rehabilitation [33]. It was shown that surgical outcomes will remain poor in Africa unless perioperative care is improved [24]. This include the pathway to care, which is a critical and the most challenging period that may determine treatment outcomes. Although there was a higher overall health insurance coverage (75%), those who were undergoing emergency surgery had 33.3%, which is comparable to findings of another study done on surgical patients in Northern Tanzania (45.5%) and to Tanzania's general population (32% in 2019) [34,35]. Patients are likely to have received their health insurance after being planned for elective surgery. This may explain the considerable out of pocket expenses among our study participants, related to both medical and non-medical expenses, and weak evidence that having health insurance protected patients from significant out of pocket expenditures. It was higher in those who were referred through another health facility. About 76.8% of Tanzanian are living below the poverty line ($3.20 per day)  [3,25]. It is important that paediatric surgical care is also financially accessible [37]. A median out-of-pocket expenditure of $69 for receiving surgical care in this study was considerably higher than the cost incurred for paediatric inpatient care in district hospitals in Kenya ($14.1) and Tanzania ($5.5) [37,38]. The consequences of out-ofpocket expenditure are pushing individuals and households into poverty, most of these are in rural settings of Tanzania and other LMICs. In these settings, $69 can equate to a month's salary for many, forming a barrier to individuals seeking care [39,40]. An argument for deductibles and co-  payments is to reduce the moral hazard of patients, but it is highly unlikely that children are considering the cost of their care when they are asking for it [41,42]. Therefore, one method of reducing the cost burden on caregivers would be a policy of free inpatient care for all children. However, a large proportion of costs in hospitals in Tanzania is related to food [38]. MNH has a policy of free provision of food for children coming from far. Activitybased costing may be adopted and utilised in Tanzania and similar settings to reduce costs of hospital food and other direct nonmedical costs [43].

Limitations
A limitation of this study is that it is based on patients who presented to MNH; we are unable to ascertain the treatment pathways for those who did not get to MNH. Future studies should consider understanding pathways to the regional hospitals. Furthermore, future studies should use qualitative methods to explore patient experiences in seeking and receiving surgical care.

Conclusion
This is the first report on whether paediatric patients in Tanzania have access to safe, timely and affordable surgical care. The majority of patients are able to access paediatric surgical care at referring health facilities within 2 h, especially those who need emergency surgery. There is a low rate of post-operative complications after paediatric surgery in Tanzania. However, paediatric surgery leads to considerable out of pocket expenses. Whilst great strides have been made by the Tanzanian government and various external partners to strengthen the surgical system in Tanzania, there now needs to be a greater focus on policies for paediatric patients. Indeed, efforts to scale up surgical care in in Tanzania and other LMICs should consider the needs of paediatric patients.

Appendix
See Tables 5, 6, 7,8    Hours taken from deciding to seek care to reaching MNH among self-referred patients

Declarations
Conflicts of interest This study received ethical approval from the Muhimbili National Hospital (MNH) IRB (MNH/IRB/2019/036) and required permissions from the respective department. An informed consent form was voluntarily provided to caregivers who signed on behalf of their children to participate in the study.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons. org/licenses/by/4.0/. For columns A), B), and C), the modes of transport being considered were those taken from the patient's home. For column D), the modes of transport being considered were those taken from the patient's home to MNH for self-referrals, and those taken from the referring health facility to MNH for referrals from health facilities