Indian Pediatrics 2001; 38: 1289-1294  

Factors Contributing to Outcome in Newborns Delivered Out of Hospital and Referred to a Teaching Institution

Arvind Sehgal
M.S. Roy
N.K. Dubey
M.C. Jyothi*

From the Division of Emergency and Critical Care, Department of Pediatrics, Kalawati Saran Children’s Hospital and Lady Hardinge Medical College, New Delhi 110 001, India and *Department of Pediatrics, Sanjay Gandhi Memorial Hospital, Delhi, India.

Correspondence to: Dr. Arvind Sehgal, Flat No. 26, UCO Apartments, Sector-IX, Rohini, Delhi 110 085, India.

Manuscript received: November 8, 2000;
Initial review completed: December 5, 2000;
Revision accepted: May 9, 2001.

Non-institutional births constitute a significant proportion of total births with a high incidence of low birth weights, hypo-thermia and perinatal and neonatal mortality. In a developing country like India, where neonatal support systems are mainly concentrated in metropolitan cities, it is important to aunderstand factors that are contributory to neonatal mortality amogst newborns referred to tertiary centres. The present work was done to assess the associations between mortality and condition of newborns delivered outside (home, Government health centers and private nursing homes) at arrival to a tertiary care center. Assessment of factors which contri-bute to mortality followed by their rectifica-tion may improve the outcome of sick referred newborns.

The study was conducted at the Emergency Unit of Kalawati Saran Children’s Hospital, New Delhi, India during the period from August-November 1999. During this period data was collected from one hundred neonates born at home, government institutions (like Primary Health Centres, District Hospitals, etc.) and private hospitals and referred to our hospital for management. The subjects were assessed in terms of place of birth, persons conducting delivery, problems at birth, reasons for referral, condition of the babies at arrival, time taken during transport and need for any immediate resuscitation. Patient’s outcome was assessed in terms of death, discharge and duration of stay. Those stabilized in emergency were shifted to general wards or nursery depending on availability of cots in the latter.

Definitions: (a) Hypothermia was taken as rectal temperature less than 36.5ºC by low reading theromometer; (b) Cyanosis meant the presence of dusky soles with peri-oral cyanosis and not cyanosis of mucosa; (c) Hypoglycemia was taken as blood glucose less than 40 mg/dl with reagent strips(1); (d) Delayed capillary filling time (CFT) was taken as more than three seconds; (e) Respiratory distress was (RD) defined as respiratory rate more than 60/minute in a quiet baby associated with sub costal and/or inter-costal retractions, which may or may not be accompanied by expiratory grunt(2); (f) Mothers were designated as "booked" when they had three antenatal check ups.

Computerized analysis of data was done by the software "STATISTIX". Comparison of means for the continuously distributed variables like birth weight, gestational age was done by using Students ‘t’test. For dichotomous variables, the Chi-square, Chi-square with Yates correction and Fisher Exact test was used, wherever applicable. Multi-variate analysis of data was done by fitting logistic model for identifying independent factors affecting mortality.

Out of a total of 100 newborns, 95 were admitted while five were cases of neural tube defects and were referred to other centers for neuro-surgical interventions. Sixty eight (71.6%) babies arrived at less than 24 hours of age, 23 (24.2%) between 24 to 72 hours and the rest beyond 72 hours of age. Table I depicts a comparison of neonatal character-istics on arrival at the hospital between survivors and non-survivors. Survivors had a significantly higher weight, gestation and temperature on arrival than non-survivors. Survivors spend significantly less time on transport, were more often euglycemic and had better perfusion than no-survivors. Place of birth, prior treatment and type of morbidity was not associated with outcome.

Out of a total 95 babies who formed the study group, 35 (36%) died. About half the deliveries took place at home and were conducted by untrained personnel or relatives. Mortality was higher in this group (21/47) as compared to those who were born at government health institutions and private set ups by trained health personnel (14/48). About half the deliveries were to unbooked mothers. The reasons for referral in our study group included the following, with many babies having multiple reasons; birth asphyxia (32%), prematurity (30%), clinical sepsis (21%), hyaline membrane disease (HMD) (22%), meconium stained liquor/meconium aspiration syndrom (MSL/MAS) (21%), neonatal hyperbilirubinemia (NNH) (12%), birth injuries (3%).

None of the referrals had been pre informed to our institution. As far as mode of transport were concerned, babies were brought in various motor vehicles like cars, jeeps, auto-rickshaws and bicycle-rickshaws in which the chances of occurrence of hemodynamic derangements are more. Only 5 babies were transported in ambulances and the adequacy of equipment in these were not checked. None of the referrals were accompanied by trained health personnel, 3 were accompanied by ward boys and 2 by drivers. Children were brought covered mainly with cotton, in bed sheets and blankets while oxygen was provided only to those transported in ambulances. None of the babies were fed during transport.

Resuscitation was required in 34 patients with endotracheal intubation in the emergency, 13 immediately as they were brought in gasping state and in others within two hours of arrival. Out of a total of 35 deaths, 25 (70%) occurred within 6 hours of arrival. Out of those who were hemodynamically stabilized and transferred to wards, only one death was reported. Seven babies could be accommodated in nursery after initial management.

To adjust for confounders, factors found to be significant on univariate analysis were extended into a logistic regression model. The significant mortality contributory factors were lower birth weight, prematurity, hypothermia, hypoglycemia, poor perfusion and longer transport time (Table II). Cyanosis (proxy for hypoxia, hypothermia) and prolonged CFT (proxy for perfusion) had the highest odds for mortality.

Table II__Factors Independently Associated with Mortality by Logistic Regression Analysis

Deliveries taking place outside health institutions, those conducted by untrained birth attendants and inadequately qualified physicians constitute a significant proportion of total births. These births have been associated with low birth weight, perinatal asphyxia, hypothermia and elevated rates of neonatal and perinatal mortality(3,4). These observations corroborate the results of our study. Unbooked status of mothers also heightens the risks as the opportunity of health and nutritional education and early detection of complications is not available. Preterm infants have special problems that puts them at a disadvantage for temperature maintenance. These include high surface area to weight ratio, decreased stores of subcutaneous and brown fat, decreased caloric intake and muscular activity. Bhoopalam et al had also implicated hypothermia as the commonest morbidity in babies born before arrival at hospital(5). Capillary filling time and peripheral pulse volume are good indicators of hemodynamic status and delayed CFT with poor peripheral circulation compound the problems already existing in a hypothermic baby.

Hypoglycemia is quite common in babies born outside and referred. The reasons could be deliberate withholding of feeds thinking that the colostrum is too thick to be digested, replacement of breast-feeds by other pre-lacteal feeds, clinical reasons such as respira-tory distress, sepsis, asphyxia, etc. Batemann et al. have also documented that the incidence of hypoglycemia and the need for resuscita-tion are more in outside deliveries(1). Our findings are in agreement with earlier experience. Hypoglycemia not only affects immediate survival but also predicts abnormal outcome in terms of Neuro Biologic Risk Score (NBRS)(6). In babies who were given prior treatment, half were getting improper therapy which was improper drugs, faulty routes of administration, inadequate dosage, etc. These included administration of mannitol, steriods, oral anti-convulsants and establishment of early feeds in severely asphyxiated babies and lack of oxygen and vitamin K supplementation.

Hypothermia, hypoglycemia, delayed CFT and poor peripheral circulation are common in out of hospital deliveries. Maintenance of oxygenation and euglycemia on arrival at emergency goes a long way in reducing the mortality. The use of low cost transportable devices (Transparent insulating bags) may permit simultaneous application of heat and attention to infant’s respiratory status during transport(1). The use of plastic tenting (Saran wrap) has been shown to be effective in preventing both convection heat loss and insensible water loss(7). A small clear plastic heat shield around small infants also prevents convection heat loss(8). Those babies in stable condition can be dressed with woolen clothes and double layered caps. Continued supply of oxygen by hood, tents, or prongs in a transport incubator is important. Running intravenous lines with appropriate fluids would ensure maintenance of blood sugar during transport. In our study also in babies who were immediately stabilized (hemodynamically), the mortality was negligible (1/65), it is the derangement during transport and status at arrival which are of extreme importance. If prior information regarding arrival of a sick newborn and his hemodynamic status is given, the institution would be better equipped and ready for management in case the need for artificial ventilation arises. Instructing mothers as part of prenatal care on the handling of a possible out of hospital delivery may also help prevent these complication.

It is important to put a system of regionalized perinatal care in place by implementing out-reach education programs. A prospective review of medical care provided during all neonatal transport should be done. This will include four categories; first, assessing demographic characteristics such as birth weight, gestation age and reasons for referral. Second, stabilization measures performed during transport which include intravenous line placement and oxygen supplementation; third, clinical and laboratory assessment at arrival including detection of hypoxia, hypothermia and hypoglycemia and fourth, assessment of transport related mortality and final disposition. Following the implementation of out of reach education programs, Shenai et al. (9) demonstrated that the frequency of stabilization measures per-formed before and during transport increase significantly (intravenous line placement from 12% to 58% and oxygen supplementation from 10% to 33%). The incidence of complications during transport decreased significantly (cyanosis from 25% to 8%, hypothermia 30% to 33%, acidemia from 3% to 13%). Both transport related mortality and neonatal mortality also decreased significantly (2.9% to 8%, 17% to 7%, respectively).

Contributors: NKD and AS conceptualized and designed the study. AS, MSR and MCJ were involved in data collection. AS, MCJ and NKD were instrumental in data analysis, interpretation and result formulation. AS will be the guarantor of the study.

Funding: None.

Competing interests: None stated.

 

1. Batemann DA, Bryan L, Nicholas S, Heagarty M. Outcome of unattended out of hospital births in Harlem. Arch Pediatr Adol 1994; 148: 147-152.

2. Singh M. Respiratory disorders. In: Care of Newborn, 4th edn. New Delhi, Sagar Publications, 1993; pp 196-217.

3. Burnett CA, Jones JA, Rooks J, Chen CH, Tyler CW, Miller A. Home delivery and neonatal mortality in North Carolina. JAMA 1980; 244: 2741-2745.

4. Schramm WF, Barnes BF, Bakewell JM. Neonatal mortality in Missouri home births. Am J Pub Health 1987; 77: 930-935.

5. Bhoopalam PS, Watkinson M. Babies born before arrival at hospital. Br J Obstet Gynec 1991; 98: 57-64.

6. Contra ctor CP, Leslie GI, Brown JR, Arnold J. Neonatal Neuro biologic risk score. Indian Pediatr 1996; 33: 95-101.

7. Chatson K, Fant ME, Clotherty JP. Temperature control, In: Manual of Neonatal Care, 4th edn. Philadelphia, Lippin Cott Raven, 1998; pp 139-141.

8. LeBlanc MH. Thermoregulation, incubators, radiant warmers, artificial skins and body hoods. Clin Perinatol 1991; 37: 403.

9. Shenai JP, Major CW, Gaylord MS. A successful decade of regionalized perinatal care in Tenessee. The National Experience. J Perinatal 1991; 11: 137-142.