Predictors of Outcome in Neonates with Respiratory Distress

Address for correspondence: Dr. B M John, Associate Professor (Paediatrics), Command Hospital (Air Force), Bangalore-560007, India E-mail: drbmj1972@yahoo.com Tel No: 09632483900 1Dr. BM John, Associate Professor (Paediatrics), Command Hospital (Air Force), Bangalore, India, 2Dr. V Venkateshwar, Professor (Paediatrics), Command Hospital (Air Force), Bangalore, India, 3Dr. Vinod Dagar, Paediatrician, Military Hospital, Danapur, Bihar, India. Abstract


Introduction
R espiratory distress is one of the commonest causes of admission of a neonate to the neonatal intensive care unit (NICU).It is a challenging problem and accounts for signifi cant morbidity and mortality.In various Indian studies, it occurred in up to four to seven percent of the neonates 1,2 .There are various factors which determine the progress and outcome in neonatal respiratory distress.The birth weight, gesta onal age and the degree of respiratory compromise are the key factors which decide the level of care the neonate would require.Clinical monitoring is most important as sophis cated equipments may not always be available in resource limited se ngs.Clinical scores such as Downe's score 3 , Silverman score 4 , APGAR score 5 and ACoRN respiratory score 6 are being used for assessing the severity of respiratory distress while CRIB (Clinical risk index for babies) 7 and SNAPPE (Score for neonatal acute physiology-perinatal extension) 8 are being used for determining illness severity.Calcula on of some of the above men oned scores need es ma on of frac onal inspired oxygen, arterial blood gas analysis and monitoring of vitals including blood pressure.However, invasive monitoring and ven lator facili es are not available in all the neonatal care units.There is a dearth of studies on role of simple clinical parameters like Downe's score and pulse oximetry early on during the course of respiratory distress in predic ng which neonates may have a higher mortality, need mechanical ven la on and need higher dura on of respiratory support.Simple clinical scores if me culously documented could be useful to determine the progression of the respiratory distress.This would then enable mely transfer of these neonates to higher centres from the primary neonatal care facility available in most of the developing countries.
This study was therefore conducted to assess the suitability of simple parameters like birth weight, gesta on age, baseline oxygen satura on, APGAR score at fi ve minutes and Downe's score as predictors of certain short term outcomes like requirement of respiratory support and mortality.

Material and Methods
Over a period of 18 months, all consecu vely admi ed neonates developing respiratory distress were studied.Inclusion Criteria: Neonates delivered in the hospital, which developed respiratory distress within 6 hours of birth, irrespec ve of gesta onal age and ae ology.Exclusion Criteria: All neonates admi ed for indica on other than respiratory distress and neonates with congenital anomalies.
Any newborn showing one or more of the following signs within 06 hours of birth was considered to have respiratory distress: (i) Respiratory rate of sixty per minute or more (ii) grun ng (iii) intercostal or subcostal retrac on (iv) cyanosis Following data was recorded: Birth weight (babies were weighed as soon as possible a er birth, nude, using standardized digital weighing scale), Gesta onal age (as per expected date of delivery (EDD) and confi rmed by USG report men oned in the antenatal follow up card), APGAR score at fi ve minutes a er birth, Downe's score (at admission and a er 2 hours, 6 hours, 12 hours and 24 hours), SpO 2 at admission ( in room air) and a er 2 hours, 6 hours, 12 hours and 24 hours.Neonates in the study group were treated as per unit protocol which had specifi ed indica ons for oxygen therapy, con nuous posi ve airway pressure (CPAP), mechanical ven la on, surfactant therapy, an bio cs, fl uid therapy, thermoregula on and suppor ve care.The neonates were followed up for the following outcomes: Need for mechanical ven la on at any stage during admission, Need for any respiratory support in the form of oxygen therapy, CPAP or mechanical ven la on at 72 hours a er admission.Mortality and survival data was recorded at me of discharge.
For this study, neonates were divided into four groups based on their birth weight and gesta onal age for diff eren al analysis as depicted in Table 1.
The diagnos c criteria for various causes of respiratory distress were adopted from the recommenda ons made by Na onal Neonatology Forum and also published in Na onal Neonatology Perinatology database report 2002-03 1 .
Data analysis was done by using SPSS (Sta s cal package for social sciences) version 17.0.Fisher's exact test, Chi-square test, Odds ra os were used to fi nd out the rela onship between various parameters.Binary logis c regression analysis was used to fi nd out the rela onship between defi ned predictors and outcomes.Receivers Opera ng Characteris c curves were used to fi nd the cut off values of the predictors for the specifi ed outcomes.All the sta s cal tests were used at 95% confi dence interval (C.I.) (5 % level of signifi cance) and p-value of < 0.05 was considered as a signifi cant rela on between studied parameters.

Results
During the study period, a total of 165 neonates having respiratory distress fulfi lled the inclusion criteria of the study.They were followed up for studying the rela onship between the specifi ed predictors and the outcomes.The demographic profi le of the studied popula on is depicted in Table 1.
In univariate analysis along with ROC curves : Birth weight < 1620 gram, gesta onal age of < 31 weeks, APGAR score of < 6, Downe's score of >3 and baseline oxygen satura on of < 86 % were found to be signifi cantly associated with mortality ( baseline oxygen satura on of < 87 % were signifi cantly associated with requirement of mechanical ven la on (Table 3) .Birth weight < 1894 gram, gesta onal age of < 37 weeks, APGAR score of < 7, Downe's score of >3 and baseline oxygen satura on of < 89 % were signifi cantly associated with requirement of respiratory support (O 2 , CPAP, mechanical ven la on) at 72 hrs of life (Table 4).

Discussion
In India, and many developing na ons, only few 'Neonatal intensive care units' have level III facili es for newborn care.Most of the centres, especially in the rural districts, have only level II care facili es.Respiratory distress is among the most common symptom complexes seen in the newborn infant.It may result from pulmonary and non pulmonary causes 1,2,9,10,11,12,13 .Prognosis depends not only on the birth weight and gesta onal age, but also on other perinatal factors and physiological condi ons of the individual neonate, in par cular the disease severity in the fi rst hours of life 14,15 .The predic on of requirement of higher level of care like mechanical ven la on and prolonged requirement of oxygen support is of paramount importance at the peripheral level where sophis cated medical equipments are not available.This study was therefore conducted to assess the u lity of birth weight, gesta onal age, baseline oxygen satura on, Downe's score and APGAR score at 5 min as predictors of mortality and requirement of respiratory support/ mechanical ven la on at 72 hours in neonates with respiratory distress.In our study, the main causes of respiratory distress requiring mechanical ven la on were Respiratory distress syndrome (61%) and meconium aspira on syndrome (14%).The study results are similar to other Indian studies 16,17 .The overall mortality was 14.5 % which was similar to another Indian study by Bhat et al (18%) 2 .

Predictors versus defi ned outcomes:
Mortality-In our study birth weight of ≤ 1620 grams was associated with higher mortality (sensi vity 75 %, posi ve predic ve value (PPV) 94.64 %, Odds ra o (OR)-9, confi dence interval (95% CI): 3.3-24.6).Our birth weight associated with higher mortality was compara vely lower in value compared to earlier studies done by Mathur et al 18 and Malhotra et al 16 showing birth weight of <2000 gram associated with higher chance of mortality.This shows the trend that with be er neonatal care facili es younger neonates are surviving well.Gesta on age of≤31 weeks was predic ve of mortality (sensi vity-88%, PPV-94%, OR-14.5, 95% CI: 5.42-39.2),which is comparable to studies done by Gera et al (30.4 weeks) 19 .Our gesta on age cut off value was lesser than as observed by Monir et al (34 weeks) 20 .APGAR score of <6 was associated with higher odds of mortality compared to higher scores (sensi vity-92%, PPV-92%, OR-12.7,95% CI: 4.6-34.4).Our score predic ve of mortality was comparable to study by Onama et al (APGAR score ≤6) 21 but our value of APGAR score was higher than many previous studies 22,23,24,25 .Downe's score of >3 had a sensi vity of 52.48 % with posi ve predic ve value of 97.37 % for mortality.And there was 12.1 mes odds (95% CI: 2.7-53.6) of dying with Downe's score of > 3. Neonates having oxygen satura on of ≤ 86 % at admission had higher mortality (sensi vity-76%, PPV-93%, OR-6.29, 95% CI: 2.47-15.9).In our study mortality was 49 % in the ven lated neonates which is higher than study by MA Xio et al (15.4 %) 6 .
Mechanical ven la on-Our study has showed that birth weight of ≤2000 grams is associated with higher requirement of mechanical ven la on (sensi vity-67 %, PPV-48%).The odds of requirement of mechanical ven la on with birth weight of < 2000 grams was 4.58 mes (95% CI: 2.24-9.36).This was comparable to the study by Mathur et al 18 .Gesta on age of < 32 weeks was predic ve of higher requirement of mechanical ven la on (sensi vity-55%, PPV-59%, OR-6.27, 95% CI: 2.96-13.23).APGAR score of ≤ 7 had a sensi vity of 83.67 % with posi ve predic ve value of 48.8 % and a high nega ve predic ve value of 90.12 meaning thereby that APGAR score of >7 signifi cantly excludes requirement of mechanical ven la on.Eugene et al 26 and MA Xiao et al 6 also had similar fi ndings with 5 minute APGAR score of < 7 predic ng requirement of mechanical ven la on.In our study Downe's score of >4 had sensi vity of 59% and PPV of 50 % towards predic ng ven lator support with an OR of 4.94 (95% CI: 2.35-10.39).Baseline oxygen satura on of ≤ 87 % had a sensi vity and PPV of 57.14% and an OR of 6.03(95% CI: 2.88-12.60)predic ng the likely requirement of mechanical ven la on.
However on mul variate analysis, only gesta onal age and birth weight were independent predictors of mortality (p-value <0.05).For requirement of mechanical ven la on, gesta on age and APGAR score were independent predictors of ven la on (p-value <0.05).No studied parameter signifi cantly predicted requirement of respiratory support (O2, CPAP, ven la on) at 72 hours on mul variate analysis.This may have been because of the limited sample size or type of sample popula on which was a limita on of the study.

Conclusion
Neonatal respiratory distress is one of the common causes of NICU admission in India and other developing countries.Illness severity assessment is important for the management.The APGAR score, Downe's score and SpO2 monitoring can be done non invasively with rela ve ease.The suggested cut off s for the observed (Birth weight and gesta on age) and monitored (Downe's score, APGAR score and oxygen satura on) parameters may be together used to predict the requirement of respiratory support, mechanical ven la on and mortality in neonates with respiratory distress thereby guiding the decisions for treatment of these babies in a given neonatal care facility as against referral to a higher centre.

Table 1 :
Demographic characteris cs of the studied popula on

Table 2 :
Predictors and Mortality

Table 3 :
Predictors and mechanical ven la on

Table 4 :
Predictors and respiratory support at 72 hours of life