Evaluation of Paediatric Critical Care Needs and Practice in Nigeria: Paediatric Residents' Perspective

Background There is a dire need for paediatric critical care (PCC) services, but their availability in tertiary hospitals in Nigeria is not well defined. Objective We evaluated self-reported PCC practice, resources, and perceived challenges in various zones of the country, using paediatric residents' perspective. Methods This is a descriptive cross-sectional survey, carried out at an Intensive Course in Paediatrics at the University of Benin Teaching Hospital, Nigeria. Participants' PCC practice and perceived adequacy of PCC resources and services were assessed using a 100 mm uncalibrated visual analogue scale (VAS). A comparison between northern and southern zones was done. A 2-sided p value < 0.05 was considered significant. Results A total of 143 residents participated in the study, 37.1% of them were male, and 62.9% were female. Their mean age was 34.6 ± 3.2 years. They were mainly (86.7%) from federal institutions across the country. Less than a half (46.7%) of the trainees attended to critically ill children daily, but only 4 out of every 10 respondents stated that such severely ill children survived till hospital discharge; 12.1% of the trainees had PICUs in their institutions. Financial constraints hindered PICU admissions. PCC staff were relatively fewer in northern zones than southern zones (p < 0.05). Their perceived adequacy of PCC equipment and services were low (VAS scores 32.7 ± 2.6 and 30.9 ± 2.8, respectively) with a strong positive correlation between the two measurements (r = 0.839; p < 0.001). Conclusion There is an unmet need for PCC practice in Nigerian tertiary hospitals with a resultant low survival rate of critically ill children. PCC training curricula and improved critical care resources are desirable in the setting.


Introduction
Paediatric critical care (PCC) or intensive care is a relatively new medical specialty, especially in low-and middle-income countries [1]. PCC ensures optimal monitoring and organsystem support of critically ill children while taking cognizance of their unique anatomic and developmental stages. Recovery from childhood critical illnesses is dependent on clinical-laboratory monitoring, prompt escalation of clinical care, effective team dynamics, and multidisciplinary care as well as adequate communication with caregivers [1,2]. However, PCC is not readily available in a resource-limited setting, despite the dire need for such levels of care [2,3]. A multicountry survey by Muttalib et al. PALISI Global Health Subgroup reported major gaps in the availability of essential PCC resources in resource-limited settings [4]. A national audit of critical care resources in South Africa by Bhagwanjee and Scribante [5] found that, regarding the public sector ICU bed, the population ratio ranged from less than 1 : 20,000 to 1 : 80,000 but less than one-fifth of beds are dedicated to paediatric and neonatal patients. Also, Touray et al. [6] in Gambia reported in 2018 that only one hospital among seven surveyed public hospitals (treating more than 50 critically ill patients a month) had a dedicated intensive care unit, resulting in an estimated 0.4 ICU beds/100,000 population in the country, while Siaw-Frimpong et al. [7] in Ghana reported 0.5 ICU beds/100,000 people in 2020. Again, Atumanya et al. [8] in Uganda reported 1.3 ICU beds per million population in a recent survey, showing limited accessibility to critical care services in the country. Likewise, a recent single-centre study in southern Nigeria showed that 33.3% of children admitted into the Emergency Unit required critical care but only 2.8% of these critically ill children were transferred to the hospital's general ICU due to limited relevant resources and financial constraints [3].
Paucity of critical care services worsens the outcome of childhood illnesses in low-and middle-income countries. In a review of PCC in resource-limited settings, Turner et al. [1] affirmed that an upgrade of critical care resources and services is needed to reduce child mortality in the region. A comparative study of two PICUs by Abdelatif et al. [9] showed that the inadequate structure and higher burden of severely ill children at a resource-limited Egyptian PICU contributed to higher mortality in it than in a resource-rich Japanese PICU. Also, considering the multilevel, family-societal factors that promote late presentation of sick children to health facilities in medically underserved regions, prompt access to optimal critical care remains an important strategy to reduce childhood mortality in the setting [10,11]. In a 5-year review of hospitalized children in an ICU in Cameroon, Nguefack et al. [12] reported that deaths occurred mainly (90.0%) within 72 hours of admission and malnutrition was the underlying cause of death in nearly a quarter of the patients. Likewise, Akindolire and Tongo [3] in Ibadan southern Nigeria reported that almost a quarter of their patients died within 48 hours of presentation in the Emergency Unit, despite the majority of their illnesses being potentially reversible with prompt critical care [3].
is highlights the fact that children often present in extreme condition in the setting, requiring specialized multisystemic supports to survive the illness. Assessment of available critical care resources in the health system is therefore a major step towards improving paediatric health indices in the subregion [1,9].
Considering the foregoing, this national survey evaluated PCC practice and resources in Nigeria, using frontline clinician's viewpoints. We hypothesized that PCC services are inadequate in all geopolitical zones of the country. We aimed to determine the burden (frequency/types/outcome) of critical illnesses seen by paediatric residents in the preceding 3 weeks as well as PCC resources available for their management. Hence, this study quantified the existing gap in critical care in the Nigerian health system and harvested relevant recommendations.

Study Design and Sample Size.
is was a descriptive cross-sectional study. It was a total population study of all consenting resident doctors attending the NPMCN's Intensive Course in Paediatrics.

Data Collection.
e questionnaire was designed based on the principal investigator's conception of the study and relevant literature searches [3,13]. Each item on the questionnaire was reviewed independently by the researchers to ensure clarity. It comprised sections on (a) sociodemographic and practice characteristics, (b) burden of critical illnesses treated in the preceding 3 weeks, (c) availability of physical/ human resources for PCC at the training institutions, and (d) residents' perceived adequacy of critical care practice using 100 mm uncalibrated visual analogue scales (VAS). Critical illnesses were classified based on the Paediatric Assessment Triangle (PAT) categories [13]. Perceived challenges and recommendations to improve critical care practice in the setting were elicited using open-ended questions. e critical illness description subscale was a six-item Likert scale in which items (e.g., " ey are mainly infectious" and "Late presentation is common") were answered on a four-point scale, from "strongly agree" to "strongly disagree." Information about the study was provided to the paediatric trainees at the conference registration desk, emphasizing voluntary participation. e pretested self-administered questionnaire with a study information statement was included in their conference package. Also, they were centrally reminded during lunchtime daily in the first week of the conference, and a research assistant went around the hall to retrieve filled questionnaires. e response rate was 100%.

Pretesting.
e initial survey instrument was pretested on 15 paediatric residents at the Department of Child Health, UBTH, Benin City, to ensure clarity of its items and validity. On feedback, all the questionnaire items were readily understood by the trainees. e computed reliability rating (Cronbach's alpha) of the critical illness description subscale was adequate (0.68). No further changes were made to the questionnaire.

Statistical Analysis.
e data were analysed using the Software Package for Social Science (SPSS) version 20.0 (Windows Inc., Chicago, IL, USA). Categorized data were presented on frequency tables. Percentages are denominated on the total number of respondents per variable. A mean score of 2.5 on the Likert scales was taken as adequate. Residents' perception of the adequacy of PCC practice was reclassified as adequate or inadequate, using a 50 mm cut-off point on VAS. Pearson's Chi-square was used to compare federal and state residents' responses and to assess for any significance between northern and southern geopolitical zones. A 2-sided p value < 0.05 was considered significant. e intraclass correlation coefficient was calculated for equipment and service adequacy. Specified challenges and recommendations were subjected to thematic analysis and presented in text boxes.

Ethical Consideration. Approval was obtained from the Research Ethics Committee (REC) of College of Medical
Sciences, University of Benin (REC Approval no.: CMS/ REC/2019/050). Permission was sought from the Local Organizing Committee of the NPMCN Intensive Course. Informed consent was obtained from every participant. Specific names and addresses of respondents' institutions were not required to ensure confidentiality.

Baseline Characteristics of Participants and eir Institutions.
A total of 143 paediatric trainees took part in the study, 37.1% of them were male, and 62.9% were female. eir mean age was 34.6 + 3.2 years. One hundred and two (72.9%) were registrars, 35 (25.0%) were senior registrars, and 3 (2.1%) were other cadres of trainees. ey were mainly from federal tertiary institutions (86.7%). All parts of the country were represented with the south-south zone having the highest number of participants 44 (31.2%) and the northeastern zone the least 10 (7.1%) (supplementary file Figure 1). e commonest ICU facility available in the participants' institutions was general/mixed ICU (44.3%), followed by adult ICU (37.9%) and distantly by PICU (12.1%). e relative distribution of respondents with access to ICU facilities among the six geopolitical zones is shown in supplementary Table 1, while the distribution of those with access to PICU is shown in supplementary Figure 2. Less than half (45.5%) of the paediatric trainees had certification related to paediatric critical care and only 8 (11.9%) of them have done paediatric advanced life support (PALS) course. Further details of their baseline characteristics of the participants are as shown in Table 1.

Critical Care Needs in Participants' Self-Reported Clinical
Practice. Less than a half (46.7%) of the trainees attend to critically ill children in their clinical practice daily, while 17.0% of the participants treat such children twice weekly. e rest of them encountered critical illness less frequently in their practice. Figure 1 shows the frequencies of treatment of critical illness by the participants.
Also, most of them (83.8%) have recently managed children with severe respiratory distress; 29.8% of the respondents who managed such children reported that the children were admitted into ICU, while 74.6% of them said such patients eventually survived. Also, about a half of the participants have managed children in other Paediatric Assessment Triangle (PAT) categories with only 3 out of every 10 respondents stating that such children were able to assess ICU care. e commonest reasons for failed ICU admission were a "lack of appropriate equipment" and "financial constraint" (supplementary Table 2). Concerning the overall outcome, only 4 out of every 10 respondents stated that children with respiratory failure, decompensated shock, or cardiopulmonary failure survived till hospital discharge. e complete details on the types and outcome of the critical illnesses managed by the participants in their respective units in the preceding 3 weeks are shown in Table 2, using PAT classification. Table 3 shows participants' level of agreement with the description of the nature and severity of critical illnesses seen in their clinical practice in the preceding 3 weeks. One hundred and nineteen participants (83.2%) adequately agreed with the descriptive statements on the nature and prognosis of the critical illnesses seen in their practice. Although the illnesses were infection-related and potentially reversible with critical care, late presentation to health facilities was common among the patients (3.51 ± 0.85; Table 3). e least available resources were capnographs (4.3%), mechanical ventilators (11.3%), syringe drivers (20.6%), and multiparameter monitors (31.9%). Sphygmomanometers, CPAP devices, multiparameter monitors, and pulse oximeters were relatively less available in the northern zones compared to the southern zones (p < 0.05). e complete zonal distribution of resources is shown in Table 4. e availability and functionality of resources in the PICUs are shown in supplementary file Table 3.

Distribution of Paediatric
Altogether, human resources that were available in their training institutions include paediatric intensivists (23.6%) paediatricians (82.1%), ICU nurses (63.4%), ICU technicians (26.8%), and other cadres of ICU workers (4.9%). Forty-two trainees (29.4%) expressed a desire to specialize in paediatric intensive care. Paediatricians, ICU nurses, and ICU technicians were fewer in the northern zones than the southern zones, but there was no paediatric intensivist in the southeast zone (p < 0.05; Table 5).

Perceived Adequacy of PCC Resources, Challenges, and
Recommendations.
e mean VAS scores of the participants on 100 mm uncalibrated scales on the perceived adequacy of PCC equipment and PCC services in Nigeria were low, 32.7 ± 2.6 and 30.9 ± 2.8, respectively.
ere is a strong positive correlation between their perceived adequacy of equipment and services (r � 0.839; p < 0.001; Figure 2). Also, there is an excellent intraclass correlation coefficient (ICC) between the two measurements, ICC � 0.91 (95% CI: 0.87-0.94). Perceived challenges to PCC and recommendations by participants are listed as follows.

Discussion
is study shows the unmet need for paediatric critical care in Nigeria with PICUs being only available to 12.1% of the respondents, similar to prior reports of limited critical care resources in subregions by other researchers [14][15][16]. ere is a high burden of treatable severe childhood illnesses in the setting, encountered daily by nearly a half of the participants, but the majority of such children die due to the lack of optimal critical care, consistent with the single-centre report by Akindolire and Tongo [3] which reported that about a quarter of patients in their series die within 48 hrs of presentation in the emergency room (ER). Likewise, Nguefack et al. [12] in Cameroon found that death occurred early on admission especially among malnourished children. e aforementioned studies reported that most of the illnesses were infection-related and potentially reversible but the patients presented late to the referral centre, requiring intensive care services for survival [3,12]. Factors promoting the delayed presentation of acutely ill children to health facilities in the setting are protean [11,17]. Although they are being addressed at various levels, late presentation to the emergency room is still widely prevalent [10,11]. Hence, there is a need to boost PCC service delivery in the country.
Moreover, basic monitoring equipment like sphygmomanometers and pulse oximeters are inadequate in over        one-fifth of the training institutions, reflecting suboptimal care of acutely ill children and subsequently high paediatric mortality rates in the country [18,19]. In this survey, ultramodern equipment like ventilators, capnographs, and multiparameter monitors were rarely available at the tertiary centres, consistent with earlier reports of major disparity between resource-limited and resource-rich centres by Muttalib et al. [4] and Abdelatif et al. [9] Similarly, Touray et al. [6] in Gambia and Siaw-Frimpong et al. [7] in Ghana reported less than 1 ICU bed/100,000 population in their countries. is highlights the need to improve health care financing and prioritize investment in critical care services in the region [20,21]. Also, restructuring of the health system to include designated centres of excellence in PCC in various zones of the country may improve the availability of PCC resources nationwide in the public sector. is study found an intraclass correlation coefficient (ICC) of 0.91 between PCC equipment and services which is indicative of excellent reliability based on Koo and Li's guideline [22]. is confirms that the provision of the equipment is pertinent to enhancing PCC services in the setting.
Nonetheless, PCC human resources were not sufficiently available at the training institutions, perhaps partly due to the limited training opportunities in PCC in the setting as well as suboptimal healthcare financing. Brotherton et al. [23] in Kenya reported the creation of an Emergency and Critical Care Clinical Officer (ECCCO) higher diploma program since 2015 but there are currently no accredited degree or fellowship programs in PCC in Nigeria. Also, critical care-related courses like PALS are often expensive; only 11.9% of our participants had current certification in PALS. is highlights the need for collaboration with other postgraduate colleges for PCC training [23][24][25]. Several international training programs such as the Paediatric Emergency and Critical Care (PECC) Kenya fellowship training program, the African Paediatric Fellowships Program (APFP) in South Africa, and the Commonwealth Medical Fellowships program in the United Kingdom are available to build critical care capacity in resource-limited settings [26][27][28].
ere should also be regular in-house training for all cadres of health workers as well as sponsored capacity-building workshops to enhance their PCC proficiency.
e main themes of perceived challenges to PCC in this national survey focus on poor staffing, lack of equipment, late presentation, and financial constraints, comparable to prior reports from resource-limited settings [9,14,16].
ere is a need to upgrade service delivery at an affordable rate to the populace, considering the potentially catastrophic effects of out-of-pocket expenditures on families [29]. Recommendations stated in this survey to strengthen PCC comprise subsidizing healthcare services, possibly through the National Health Insurance Scheme (NHIS) and provision of more PICU bed spaces and relevant paediatric equipment in the country. ese can be attained through viable NHIS and public-private partnership funding [21,30].
is national survey has several strengths; its 100% response rate captures the views of all paediatric residents, attending the update course from all geopolitical zones of the country. e 3-week recall interval used in this study was short, ensuring the accurate provision of information. Also, many of the respondents were keenly aware of the dire need for PCC and the dearth of relevant resources in their institutions because they had spent 2 or more years in residency programs at their centres. Nonetheless, using the CROSS checklist [31], the limitations of this survey include the possibility of a recall bias and imprecision as well as a small sample size that can limit the generalizability of the findings. is sample size was estimated as 17.7% of paediatric trainees nationwide [32]. Future studies may focus on the on-site survey and key informant interviews at the training institutions to ascertain trainers' viewpoints. Also, a detailed assessment of critical care nursing manpower in the country is a desirable line of future study to strengthen PCC in the setting.

Conclusions.
ere is an unmet need for PCC practice in Nigerian tertiary hospitals with a resultant low survival rate of critically ill children in the setting. e scarcity of human resources can be ameliorated by incorporating relevant aspects of the PCC curriculum into the clinical training of health professionals as well as investing substantially into PCC physical resources.
Data Availability e completed questionnaires and excel spreadsheet of the research reported in this article are available from the authors on request.

Conflicts of Interest
e authors declare that they have no conflicts of interest.

Authors' Contributions
MTA designed the study and wrote the initial draft of the paper. FEK carried out data collection. Both authors participated in data analysis and critical review of the final manuscripts.
institutions not included) (powered by Bing© GeoNames, Microsoft, Tom Tom). Table 1: distribution of respondents with access to intensive care facilities by geopolitical zones. Table 2: reasons for not admitting critically ill children into ICU indicated by the respondents (N � 136). Table 3: availability and functionality of PICU resources indicated by the respondents (N � 17). . (Supplementary Materials)