|Year : 2022 | Volume
| Issue : 3 | Page : 257-263
Risk factors predicting early in-hospital mortality among underfive children and need for decentralization of pediatric emergency care services
Aditya Soni1, Sumaira Khalil2, RM Pandey3, Harish Chellani4
1 Senior Resident, Department of Pediatrics, Vardhman Mahavir Medical College and Safdarjung Hospital, New Delhi, India
2 Assistant Professor, Department of Pediatrics, University College of Medical Science & GTB Hospital, New Delhi, India
3 Professor and Head, Department of Biostatistics, All India Institute of Medical Sciences, New Delhi, India
4 Professor, Department of Pediatrics, Vardhman Mahavir Medical College and Safdarjung Hospital, New Delhi, India
|Date of Submission||05-Apr-2022|
|Date of Decision||16-Jul-2022|
|Date of Acceptance||25-Jul-2022|
|Date of Web Publication||22-Sep-2022|
Department of Pediatrics, Vardhaman Mahavir Medical College and Safdarjung Hospital, New Delhi
Source of Support: None, Conflict of Interest: None
| Abstract|| |
Background: Lack of pediatric triage and emergency care system in peripheral healthcare centers leads to unnecessary referral of low- and medium-risk patients. This study was conducted to study the risk factors predicting mortality within 48 h of admission in neonates and under-five children referred to the pediatric emergency of a tertiary care hospital in India. Methods: This prospective study was conducted on children (0–5 years) referred to the pediatric emergency who were enrolled and followed up. The outcome was defined as “survival” or “death” at 48 hours. Logistic regression analysis was conducted to assess the predictors of early in-hospital mortality. Results: A total of 246 consecutive pediatric (62 neonates, 52 young infants, and 132 children aged 1–5 years) referral cases were enrolled; mortality within 48 hours was 20%. Lack of pediatric intensive care (odds ratio [OR] 4.07, 95% confidence interval [CI] 2.0, 8.32, P = 0.02), lack of neonatal intensive care (OR 2.10, 95% CI 1.01,4.28, P ≤ 0.001), distance from referral center >20 km (OR 4.61, 95% CI 2.01, 10.58, P = 0.0003), >1 h taken during transport (OR 7.75, 95% CI 2.93, 20.46, P < 0.001), lack of ambulance facility (OR 0.04, 95% CI 0.009, 0.143, P < 0.0001), very sick condition on arrival (OR 210.1, 95% CI 12.1, 3643.41, P = 0.0002), and unstable temperature-oxygenation-perfusion-sugar on arrival were the independent risk factors predicting in early in-hospital mortality. Conclusion: Developing a pediatric triage and monitoring system, tele-pediatric intensive care unit, regionalizing referral-back-referral services with robust interhospital communication, and strengthening pediatric emergency services are the need of the hour to reduce early in-hospital mortality.
Keywords: Death, emergency, medical audit
|How to cite this article:|
Soni A, Khalil S, Pandey R M, Chellani H. Risk factors predicting early in-hospital mortality among underfive children and need for decentralization of pediatric emergency care services. Indian J Public Health 2022;66:257-63
|How to cite this URL:|
Soni A, Khalil S, Pandey R M, Chellani H. Risk factors predicting early in-hospital mortality among underfive children and need for decentralization of pediatric emergency care services. Indian J Public Health [serial online] 2022 [cited 2022 Sep 30];66:257-63. Available from: https://www.ijph.in/text.asp?2022/66/3/257/356613
| Introduction|| |
Even though there has been a substantial reduction in under-five deaths since 1990, about 1.2 million under-five children continue to die annually in India, which accounts for 20% of the world. Given the current scenario, India is unlikely to achieve the National Health Policy targets of 2025 or Sustainable Development Goals 2030 targets. A multi-tiered healthcare system is existing in our country, and patients are referred from peripheral to a higher level of care depending on the availability of resources and condition of the patient. However, with nonuniform distribution of resources and concentration of trained workforce only in urban areas, patients hailing from rural and distant areas are at a significant disadvantage.
Reports published in the past have found that lack of prereferral communication, poor documentation, and lack of emergency care practices,,, directly affect the outcome in low- and middle-income countries (LMICs). The World Health Organization (WHO) in the 72nd World Health Assembly adopted a resolution supporting countries in the development of services delivering timely care to critically ill children. We are a tertiary care referral hospital aiming to reduce early in-hospital neonatal and child mortality as we cater to numerous pediatric emergencies from the city and surrounding states. Sylverken et al. reported that redirecting limited resources to pediatric emergency care is associated with the reduction in early in-hospital mortality, especially in LMICs. Therefore, this death audit of pediatric referrals was conducted to ascertain the risk factors predicting early in-hospital mortality within 48 hours of admission and to determine the demographic profile, disease-related factors, and system-based factors related to referral of under-five children to the pediatric emergency of a tertiary care hospital.
| Materials and Methods|| |
Study design and setting
This prospective, cross-sectional study was conducted in the pediatric emergency department of a tertiary care hospital in North India from April 2019 to March 2020. All consecutive children (0–5 years) referred and admitted to the pediatric emergency were enrolled in the study. The study was approved by the institutional ethics committee (IEC/VMMC/SJH/Thesis/October/2018/4 dated October 29, 2018), and informed written consent was taken from caregivers of eligible participants.
Inclusion and exclusion criteria
All consecutive children (0–5 years) referred and admitted to the pediatric emergency were included in the study. Referred patients were defined as those patients who had visited some other healthcare center and were refused admission. Children referred for head injury or surgical indication were excluded as they were referred to other departments after admission and brought in dead (BID) were excluded. Children who visited emergency without a referral were also excluded from the study.
All children aged 0–5 years were triaged according to the Emergency Triage Assessment and Treatment WHO Protocol. At admission, details regarding temperature, oxygenation, perfusion, and sugar (TOPS) were recorded. At recruitment, children were divided into three groups, 0–1 month, 1 month to 1 year, and 1–5 years. Temperature was measured with a digital thermometer; cold stress was defined as an axillary temperature between 36°C and 36.4°C, moderate hypothermia was defined as an axillary temperature between 32°C and 35.9°C and severe hypothermia was defined as an axillary temperature <32°C. Children with SpO2 <94% were defined as hypoxic. Blood sugar was measured from a drop of blood, taken from the heel of a young infant, or by finger prick from an older infant or child using a glucometer; hypoglycemia was defined as measured blood glucose level is low I <2.5 mmol/l (45 mg/dl) in neonates or <3 mmol/l (54 mg/dl) in a severely malnourished child. Perfusion was assessed by measuring the capillary refill time (CRT). Pressure was applied for 5 seconds on the sternum, and the time taken to regain its original color was counted. A CRT of more than 3 seconds was taken as delayed perfusion. All these were managed as per the department's protocol.
Baseline clinical data collection
After initial stabilization in the emergency, a detailed clinical history with clinical examination was conducted on every child. Any treatment received before admission was recorded; resuscitations received at arrival in the form of oxygen, fluid boluses, intubation, use of emergency drugs such as antiepileptic and antihypertensive, endotracheal intubation, and chest compressions were taken note of. In case of neonates, a detailed antenatal history with birth details was recorded. Those with emergency signs who required immediate emergency treatment were labeled as very sick. Emergency signs included compromised airway, severe breathing difficulty, circulatory shock, coma, convulsions, and severe dehydration.
Details regarding indications for referral, such as worsening of clinical status, medical disorders with lack of availability of treatment, facilities, absence of neonatal and pediatric intensive care units (NICU and PICU), nonavailability of blood banks, and surgical expertise were recorded in a predesigned pro forma. Information regarding the referral details such as facility from where child was referred; government or private hospital, number of institutes child was seen before being referred, distance covered (kilometers), and time taken (hours) during transport was also recorded. Transportation details of referral such as mode of transport, whether the vehicle for referral of the patient was provided by the referring facility, whether patient was accompanied by healthcare professional, and interval between decision to transfer and availability of transport were also recorded. Details regarding communication with the referring facility were also recorded such as any telephonic communication between the referring facility and referral facility before transfer of the patient, any pretransfer counseling, and need for referral was explained to the attendant. A referral slip given or not, any resuscitation done (oxygen, intravenous fluids, emergency drugs such as vasopressors and antiepileptics) at referring facility was also noted.
Children were investigated, managed, and monitored as per the standard treatment protocols. The outcome was defined as children who “survived” and “died” at 48 hours of hospitalization. Comparison of all risk factors with outcomes was done at 48 h of hospitalization.
Sachan et al. in their study reported the adjusted odds ratio (OR) of mortality for duration of transport <1 hour and 1-2 hours as 0.01 (95% confidence interval (CI) 0.001-0.05) and 0.42 (95% CI 0.07-2.40) respectively in neonates. The minimum required sample size calculated with 80% power of study and 5% level of significance was 221. Taking lost to follow up as 10%, total sample size calculated was 246.
All categorical variables were presented as proportions, and all continuous variables were documented as mean ± standard deviation (for normally distributed data) and as median and interquartile range (non-parametric data). Patient demographics, hospital course, and referral details were summarized and compared among those who survived and expired. Fischer's exact test was used for categorical variables, while Student's t-test was used for continuous variables. Chi-square test or t-test was used to study the distribution of mortality among the different predictor variables. For multivariate analysis, logistic regression analysis was used, mortality was taken as a dependent variable, and clinically important factors or factors with P < 0.05 were taken as independent variable. The data were entered into MS EXCEL spreadsheet, and analysis was done using Statistical Package for the Social Sciences (SPSS IBM, Chicago, USA) version 21.0.
| Results|| |
A total of 300 children were enrolled in this study. Out of these, 246 referred children were included for analysis and 54 were excluded; 8 were BID, 25 were head injury patients, and 22 were referred for surgical indications. Of the ones included, a quarter were neonates (n = 62, 25%), 21% (n = 52) were between 1 month and 1 year, and 54% (n = 132) were 1–5 years of age. The demographic and disease-related profile of the study population is depicted in [Table 1]. Sixty percent were referred from one healthcare contact, whereas 40% had contact with two or more than two healthcare facilities before being referred to us. Sixty percent covered more than 20 km during transport and 20% patients took more than 2 hour to reach the hospital. In majority, indication for referral to our center was lack of intensive care services.
|Table 1: Demographic and disease-related profile of study population (n=246)|
Click here to view
The overall early in-house mortality in our study was 20% (n = 48/246). Neonates contributed 40% (n = 19/48) to the early mortality. Eighty-four percent of the neonates who expired were preterm and 32% were extremely low birth weight (ELBW). On univariate analysis, neonates (OR = 0.31, 95% CI 0.11, 0.83, P = 0.02), Extremely low birth weight (ELBW) (OR = 65, 95% CI 2.41, 1967.3, P = 0.01), prematurity (OR = 8.69, 95% CI 2.09, 35.97, P = 0.003), lack of NICU (OR 2.10, 95% CI 1.01,4.28, P = <0.001), lack of PICU (OR 4.07, 95% CI 2.0, 8.32, P = 0.02), contact with >1 health facility (OR = 2.95, 95% CI 1.47, 5.87, P = 0.002), distance traveled > 20 km (OR = 4.61, 95% CI 2.01, 10.58, P = 0.0003), >1 hour taken during transport (OR = 7.75, 95% CI 2.93,20.46, P < 0.0001), lack of ambulance (OR = 0.04, 95% CI 0.009,0.14, P < 0.0001), hypothermia (OR = 10.6, 95% CI 5.11, 21.8, P < 0.0001), hypoxia (OR = 13, 95% CI 6.2, 27.22, P < 0.0001), and shock (OR = 32.7, 95% CI 14.23, 75.08, P < 0.0001) at arrival were significant risk factors associated with early in-house mortality [Table 2].
[Table 3] depicts predictors of mortality within 48 h of admission by logistic regression analysis. Lack of PICU (OR 101.2, 95% CI 5.97, 1715.61, P = 0.02) and NICU (OR 76.77, 95% CI 4.46, 1319.40, P < 0.001), time taken >30 min between transfer decision and availability of transport (OR 2.1, 95% CI 1.09, 4.02, P = 0.02), >20 km traveled (OR 4.61, 95% CI 2.01, 10.58, P = 0.0003), >1 hour taken during transport (OR 7.75, 95% CI 2.93, 20.46, P < 0.001), lack of ambulance facility (OR 0.04, 95% CI 0.009, 0.143, P < 0.0001), shock on arrival (OR 4.19, 95% CI 1.08, 16.14, P = 0.03), requirement of fluid bolus/inotropes on arrival (OR 12.60, 95% CI 2.54, 62.49, P = 0.002), hypothermia (OR 7.97, 95% CI 2.42, 26.15, P = 0.001), and hypoxia (OR 3.40, 95% CI 1.03, 11.06, P = 0.04) on arrival were independent factors predicting early in-house mortality within 48 hours of admission.
|Table 3: Logistic regression analysis of significant factors predicting mortality among referred under-5 children|
Click here to view
| Discussion|| |
In this prospective study, the overall early in-house mortality reported was 20%, neonates contributed 40% to the early mortality, and prematurity was the most common cause. Pneumonia was the most common cause of under-5 mortality. Seventy-six percent were critical at arrival.
Our center is a tertiary care center that acts as a major referral center receiving patients from the public and private sectors equally. Majority of children referred were between 1 and 5 years of age. Pneumonia (17%) was the most common diagnosis at admission in neonates, one-fourth (24%) were admitted with prematurity. Among neonates, majority who expired were preterm and 30% were ELBW. On univariate analysis, prematurity and ELBW were the independent risk factors significantly associated with mortality. This is in synchronous with the National Representative Mortality Survey,, and the national and global estimates of child mortality.,,
Financial constraints were a leading cause of referrals from private hospitals, but a large percentage was also referred from public sector hospitals. Majority (60%) referrals were due to unavailability of intensive care services, while 2/3rd were critical at arrival requiring immediate resuscitation and management. Therefore, availability of resources, presence of trained staff, and prereferral management were important determinants seen in our study for timely referral. Lack of skilled professionals, limited intensive care expertise, lack of organized triage system, disorganized emergency management, and lack of protocol-based management are some major reasons reported to refer patients from public sector hospitals, in LMICs. Healthcare professionals skilled in triage can provide prereferral stabilization to these critical patients, and noncritical patients can be managed at the district hospitals/level II care facilities, thereby reducing the burden on tertiary care hospitals.
In our study, contact with >1 health facility before referral, distance traveled >20 km during transport, time taken >1 hour during transport, and presence of hypothermia, hypoxia, and shock at arrival were significantly associated with early in-hospital mortality. This was similar to Nicholl et al. These factors continue to remain significantly relevant nearly two decades later also as it highlights the lack of sustainable interventions and re-emphasizes a need for a regionalized referral system, improved health infrastructure, and better level II pediatric emergency services including transport services which are underdeveloped even in metropolitan cities.
Pediatric emergency care services and referral system is underdeveloped and nonexistent in most of the secondary level/district hospitals, taluks, and primary health centers (PHCs) that cater to socioeconomically deprived children of the community and refer children in hemodynamically unstable conditions to our hospital. Therefore, empowering these healthcare facilities by training junior doctors, nurses, technicians, and paramedical staff in handling pediatric emergencies and pediatric resuscitation can help bridge this gap. To strengthen facility-based pediatric care at district hospitals, the National Health Mission under the Ministry of Health and Family Welfare formed operational guidelines for planning and implementation in district hospitals. This includes standard operating procedures (SOPs) for setting up pediatric triage and emergency areas in district hospitals and provides specific guidance for strengthening pediatric care facilities compared to secondary level of care in public health systems within the framework of Indian Public Health Standards Guidelines for District Hospitals, Revised 2012. Following these SOPs strictly will complement the community-based interventions that focus on timely identification and prompt referral of sick children to health facilities. Giri et al. concluded that developing a pediatric triage and monitoring system can be implemented in the general emergency department in a low-resource setting, thereby reducing mortality. Learning through simulation is another intervention recommended by the WHO as an educational tool in low-resource settings to bridge this gap. Navjat Shishu Suraksha Karyakram, Facility-based Newborn Care, and Facility-based Integrated Management of Neonatal and Childhood Illness programs need to be implemented to cover at least 80% healthcare facility providers. The training should focus more on the skill component including communication skills. The health facility services need to be linked with community services with strong referral linkages.
On logistic regression analysis, lack of intensive care services, long distance traveled, longer duration taken, lack of ambulance facility, unstable condition, and unstable TOPS on arrival were independent risk factors predicting early in-house mortality. This is similar to Sheth et al. Therefore, each parameter of TOPS carries an independent risk associated with mortality. Ghbeis et al. reported a positive impact on survival after implementing a “Tele-Pediatric Intensive Care” program (tele-PICU) to critically ill pediatric patients in Syria. Most of our peripheral health centers are equally compromised in terms of pediatric intensive care services, lack of specialized trained staff, equipment, communication, and education. If similar models can be implemented where telePICU of a tertiary care facility can be linked with nearby primary or secondary level of care facility to receive timely consultation with pediatric intensivists, provide specialized patient care and monitoring, provide standardized treatment protocols, and centralize technical assistance directly reducing mortality. Future research on this model is recommended as majority were referred due to lack of intensive care expertise.
There is a need for regionalization of referral-back-referral services and a robust interhospital communication and referral system along with an outreach education program to improve the outcome. Kapoor et al. concluded that an effective referral/counter-referral system is essential in resource-limited healthcare systems like ours, to maximize efficiency, decrease unnecessary resource utilization, and provide safe and timely patient care within a tiered system. Similar study from India analyzed the quality of referrals and their association with outcome and reported that their model was successful in sensitizing healthcare providers in PHC regarding timely resuscitation and stabilization, thereby reducing the proportion of patients received in hemodynamically unstable conditions and strengthening the referral-back-referral process. [Table 4] depicts a summary of supporting evidence.
The strength of the present study is that this is a prospective study closely analyzing the disease- and nondisease-related risk factors, leading to pediatric referrals and the association of these risk factors on their early outcome. It also included children from neonates up to 5 years of age increasing its generalizability. The study also had some limitations. We did not analyze the quality of care at the facility and long-term outcomes of these referrals.
| Conclusion|| |
Developing a pediatric triage and monitoring system, tele-PICU, simulation, regionalizing referral-back-referral services with robust interhospital communication, and strengthening pediatric emergency services are the need of the hour to reduce early in-hospital mortality.
Financial support and sponsorship
Conflicts of interest
There are no conflicts of interest.
| References|| |
India State-Level Disease Burden Initiative Child Mortality Collaborators. Subnational mapping of under-5 and neonatal mortality trends in India: The global burden of disease study 2000-17. Lancet 2020;395:1640-58.
Subhan I, Jain A. Emergency care in India: The building blocks. Int J Emerg Med 2010;3:207-11.
Sachan R, Singh A, Kumar D, Yadav R, Singh DK, Shukla KM. Predictors of neonatal mortality referred to a tertiary care teaching institute: A descriptive study. Indian J Child Health 2016;3:154-8.
Roy M, Gupta R, Bhatt M, Meena SK, Aggarwal KC. Epidemiology of child mortality in a tertiary hospital, New Delhi. Saudi J Health Sci 2017;6:83. [Full text]
Hsiao M, Morris SK, Malhotra A, Suraweera W, Jha P. Time-critical mortality conditions in low-income and middle-income countries. Lancet 2013;381:993-4.
Reynolds TA, Sawe H, Rubiano AM, Shin SD, Wallis L, Mock CN. Strengthening health systems to provide emergency care. In: Jamison DT, Gelband H, Horton S, editors. Disease Control Priorities: Improving Health and Reducing Poverty. 3rd
ed., Ch. 13. Washington D.C: The International Bank for Reconstruction and Development/The World Bank; 2017. p. 247-66.
Sylverken J, Robison JA, Osei-Akoto A, Nguah SB, Addo-Yobo E, Balch A, et al.
Decreased mortality after establishing a pediatric emergency unit at an urban referral hospital in Ghana. Pediatr Emerg Care 2021;37:e391-5.
World Health Organization. Emergency Triage, Assessment and Treatment Manual for Participants. Geneva. World Health Organization; 2005. p. 3-9.
Narang M, Kaushik JS, Sharma AK, Faridi MM. Predictors of mortality among the neonates transported to referral centre in Delhi, India. Indian J Public Health 2013;57:100-4. [Full text]
Million Death Study Collaborators; Bassani DG, Kumar R, Awasthi S, Morris SK, Paul VK, et al.
Causes of neonatal and child mortality in India: A nationally representative mortality survey. Lancet 2010;376:1853-60.
Roy MP, Gupta R, Sehgal R. Neonatal transport in India: From public health perspective. Med J DY Patil Univ 2016;9:566-9. [Full text]
Inter-Agency Group on Child Mortality Estimation (IGME): Levels and Trends in Child Mortality, Report 2012 (13 September 2012). Available from: http://www.childinfo.org/mortality.html
. [Last accessed on 2021 Jan 14].
Morris SK, Bassani DG, Awasthi S, Kumar R, Shet A, Suraweera W, et al.
Diarrhea, pneumonia, and infectious disease mortality in children aged 5 to 14 years in India. PLoS One 2011;6:e20119.
Nolan T, Angos P, Cunha AJ, Muhe L, Qazi S, Simoes EA, et al.
Quality of hospital care for seriously ill children in less-developed countries. Lancet 2001;357:106-10.
Sharma S, Pandey S. Utilization and Impact of Referral Transport System on Institutional Deliveries in Delhi. IEG Working Paper; 2016. p. 367.
Nicholl J, West J, Goodacre S, Turner J. The relationship between distance to hospital and patient mortality in emergencies: An observational study. Emerg Med J 2007;24:665-8.
Giri S, Halvas-Svendsen T, Rogne T, Shrestha SK, Døllner H, Solligård E, et al
. Pediatric Patients in a Local Nepali Emergency Department: Presenting Complaints, Triage and Post-Discharge Mortality. Glob Pediatr Health. 2020.18;7:2333794X20947926. doi: 10.1177/2333794X20947926.
Rule AR, Tabangin M, Cheruiyot D, Mueri P, Kamath-Rayne BD. The call and the challenge of pediatric resuscitation and simulation research in low-resource settings. Simul Healthc 2017;12:402-6.
Sheth NM, Pandya N. Evaluation of TOPS score as predictor for outcome in sick newborns. Int J Contemp Pediatr 2020;7:1249-55.
Kumar PP, Kumar CD, Shaik F, Yadav S, Dusa S, Venkatlakshmi A. Transported neonates by a specialist team – How STABLE are they. Indian J Pediatr 2011;78:860-2.
Ghbeis MB, Steffen KM, Braunlin EA, Beilman GJ, Dahman J, Ostwani W, et al.
Tele-pediatric intensive care for critically Ill children in syria. Telemed J E Health 2018;24:621-3.
Kapoor R, Avendaño L, Sandoval MA, Cruz AT, Sampayo EM, Soto MA, et al
. Initiating a Standardized Regional Referral and Counter-Referral System in Guatemala: A Mixed-Methods Study. Glob Pediatr Health 2017;13;4:2333794X17719205.
Ezhumalai G, Jayashree M, Nallasamy K, Bansal A, Bharti B. Referrals to a pediatric emergency department of a tertiary care teaching hospital before and after introduction of a referral education module – A quality improvement study. BMC Health Serv Res 2020;20:761.
[Table 1], [Table 2], [Table 3], [Table 4]