Not Crying After Birth as a Predictor of Not Breathing

This was an observational nested study to evaluate the scale up of the HBB quality improvement (QI) package in hospitals of Nepal (Nepal Perinatal Quality Improvement Project).^27,28 

This study was conducted in 4 publicly funded hospitals providing comprehensive emergency obstetric and neonatal care and sick neonatal and pediatric referral services. These hospitals were Koshi Zonal Hospital, Bharatpur Hospital, Western Regional Hospital, and Lumbini Zonal Hospital.

Births in these hospitals occurred in 2 units: the labor and delivery unit for vaginal deliveries performed by nurse-midwives and obstetricians and the operation theater for cesarean deliveries performed by the obstetricians and anesthetists. As per the standard protocol, fetal heart rate is first assessed at the time of admission. Fetal heart rate monitoring is done by using the intermittent auscultation method during the first and second stages of labor. Women who had obstetric and/or neonatal complications were considered for cesarean delivery.

In both the units, nurse-midwives were the primary responders for neonatal resuscitation. The annual number of hospital births ranges from 9007 to 11 318 a year, with an intrapartum-related mortality (intrapartum stillbirth and first-day mortality) rate of 8.6 to 14.2 per 1000 total births.

After the introduction of the HBB QI package in the hospitals, health workers were trained to provide stimulation to “noncrying but breathing” infants.

The study was conducted between July 2017 and September 2018.

Eligible participants were consenting women in labor at ≥22 weeks’ gestation in admission to the labor and delivery room.

We excluded women whose fetus had died before admission to labor and delivery (no fetal heart sound), those referred to the operation theater for cesarean delivery, and women who were referred to other facilities before delivery.

This was a secondary data analysis of an estimated sample of 20 000 birth observations as part of the HBB scale-up evaluation to detect any change in the health workers’ practice after QI interventions.²⁷ 

We included the following variables:

• noncrying infants after birth: infants who did not cry immediately after birth;

• noncrying but breathing infants after birth: infants who did not cry but spontaneously breathed immediately after birth;

• noncrying and nonbreathing infants after birth: infants who did not cry and did not breathe spontaneously immediately after birth;

• nonbreathing at 1 minute: infants who were not breathing 1 minute after birth;

• nonbreathing at 5 minutes: infants who were not breathing at 5 minutes after birth;

• extremely preterm: infants between 22 and 28 weeks’ gestational age;

• very preterm: infants between 29 and 32 weeks’ gestational age;

• late preterm infants between- 33 to 36 weeks’ gestational age;

• term: infants ≥37 weeks’ gestation;

• mode of delivery: infants who were delivered either by spontaneous or assisted vaginal delivery (the assisted vaginal delivery included vacuum and forceps deliveries);

• induction of labor: labor that was started by artificial rupture of membranes, prostaglandins, and/or oxytocin;

• appropriate for gestational age: birth weight in the ≥10th percentile for gestational age based on the INTERGROWTH-21st study reference point for South Asia²⁹ ;

• small for gestational age (SGA): birth weight in the <10th percentile for gestational age based on the INTERGROWTH-21st study reference point for South Asia²⁹ ;

• predischarge neonatal death: infants who died before discharge from the hospital; and

• intrapartum stillbirth: infants born at or after 22 weeks’ gestation who had a fetal heart sound during the antepartum period and no fetal heart sound during labor, with no signs of life (breathing or heart rate) after neonatal resuscitation at birth.

The data collection team consisted of clinical nurse-midwives who have at least one year of experience in midwifery and research. The data collection team received a week of rigorous training before the start of training on the HBB program as well as the labor and delivery clinical observation tool. The data collection team was placed in admission, labor and delivery, postnatal, and sick newborn care units in each hospital. The team had 8 data collectors with a data coordinator to ensure data collection and management quality. Periodic data quality assessments were done by an independent team of clinicians. On a quarterly basis, an independent clinical team assessed the accuracy of the clinical observation recorded by the data collectors. Discrepancies on the accuracy of the clinical observation made by data collectors were reviewed, and feedback was provided by the independent clinical team.

Data collectors observed clinical events and clinical service provided to the women and infants in the labor and delivery room but not in the operation theater . Observers recorded the induction of labor, time and mode of delivery, and the status of infant crying and infant breathing after birth at 3 distinct times: immediately after birth and at 1 and 5 minutes. Immediate newborn care practices, including any additional stimulation, suctioning or clearing of airways, BMV, additional resuscitation care, and mortality outcomes of the infant, were assessed via observation. Recording of the observation was done in real time on paper-based data collection proforma. The pretesting of the observation checklist was done in another tertiary hospital of Nepal.

The collected information on the observed data were reviewed by the data coordinator to check for completeness. The completed data collection tools were transferred by data coordinators to the central research office in Kathmandu, where they were indexed and entered into the Census and Survey Processing System database. Data were backed up on an external hard drive every month.

To ensure data entry accuracy, 10% of the data were reentered by the data coordinator each month, with feedback provided to the data entry officers.

We categorized infants into 2 overlapping groups by their status at birth: (1) crying or noncrying and (2) breathing or nonbreathing were categorized into noncrying but breathing and noncrying and nonbreathing after birth. We analyzed the incidence of noncrying after birth and nonbreathing after birth and at 1 and 5 minutes. We analyzed the incidence of noncrying, noncrying and nonbreathing, and noncrying but breathing after birth by different determinants: gestational age, birth weight as per gestational age, induction of labor, and mode of delivery. The sensitivity, specificity, positive predictive value, and negative predictive value of noncrying infants after birth with nonbreathing infants were calculated.

A bivariate analysis was conducted to assess the level of association of noncrying but breathing and noncrying and nonbreathing infants after birth with predischarge neonatal mortality. A multiple regression analysis of noncrying but breathing and noncrying and nonbreathing after birth with predischarge neonatal mortality, after adjusting for gestational age, birth weight, induction of labor, mode of delivery, and sex of the infant, was performed. Missing variables and loss to follow-up were excluded during the data analysis.

Written consent from each participant included in the study was taken during admission to the hospital. Ethical approval was received from the Ethical Review Board of Nepal Health Research Council (17/2017). Registration in the ISRCTN registry was done for the large-scale evaluation of the scale up of the HBB QI package.

Of the total 33 430 pregnant women admitted into the hospitals during the 15-month study period, 26 345 were assessed for eligibility, of whom 22 340 (84.8%) women were recruited. Among the total women recruited, 183 (0.8%) had stillbirths, 169 (0.8%) had neonatal deaths, and 8% were lost to follow-up (Fig 1).

Of the 19 977 infants observed, 11.1% did not cry after birth. Of the infants observed, 5.2% were noncrying and nonbreathing and 5.9% were noncrying but breathing after birth. Of the infants observed, 1.9% (373) did not breathe spontaneously until 1 minute, and 0.8% (168) did not breathe spontaneously until 5 minutes (Fig 2).

Among noncrying and nonbreathing infants after birth, 71.0% received stimulation, 25.1% were nonbreathing at 1 minute, 23.8% received BMV, and 13.9% were nonbreathing at 5 minutes. Among noncrying but breathing infants after birth, 96.3% received stimulation, 9.5% were nonbreathing at 1 minute, 9.3% received BMV, and 2.0% were nonbreathing at 5 minutes (Fig 3).

Among the term noncrying and nonbreathing infants after birth, 73.1% received stimulation, 22.3% were nonbreathing at 1 minute, 22.9% received BMV, and 10.8% were nonbreathing at 5 minutes. Among the term noncrying but breathing infants after birth, 96.4% received stimulation, 9.2% were nonbreathing at 1 minute, 8.9% received BMV, and 1.8% were nonbreathing at 5 minutes (Fig 4). Among the term noncrying but breathing infants after birth, 9 of them died before discharge.

Among the extremely preterm infants (22–28 weeks), 53.5% were noncrying after birth, with 48.8% noncrying and nonbreathing after birth and 4.7% noncrying but breathing after birth. Among the infants born with assisted vaginal delivery, a higher proportion (23.5%) were noncrying after birth, with 11.4% noncrying but breathing and 12.0% noncrying and nonbreathing after birth. Among women with induced labor by using artificial rupture of membranes, 23.6% of infants were noncrying after birth, with 21.4% noncrying but breathing and 2.2% noncrying and nonbreathing after birth. Among SGA infants, 15.6% were noncrying after birth, with 5.7% noncrying but breathing and 8.1% noncrying and nonbreathing after birth (Table 1).

Compared with infants crying and breathing after birth, noncrying and nonbreathing infants had 46 times higher odds of predischarge neonatal mortality (adjusted odds ratio [aOR] 46.0; 95% confidence interval [CI], 25.8–81.9) after adjustment for preterm birth, assisted vaginal delivery, and SGA. Preterm birth, assisted vaginal delivery, and SGA revealed an association with predischarge neonatal mortality. Compared with infants crying and breathing after birth, noncrying but breathing infants had 12 times higher odds (aOR 12.3; 95% CI, 5.8–26.1) of predischarge mortality after adjustment for risk of preterm birth, assisted vaginal delivery, and SGA (Table 2).

This study revealed that all nonbreathing infants after birth were noncrying (sensitivity = 100%). The positive predictive value after birth of noncrying infants with nonbreathing was 46.0% (95% CI, 44.5%–48.7%). The negative predictive value after birth of noncrying infants with nonbreathing was 100% (95% CI, 100%–100%) (Tables 3 and 4).

This is the largest study to date in which we report on observation at birth of noncrying and nonbreathing infants. Among the total 19 977 observed births, approximately one-tenth of infants did not cry after birth. Almost half of the noncrying infants after birth were breathing. The proportion of noncrying infants increases with lower gestational age, with >50% of infants at 22 to 28 weeks’ gestation not crying at birth.

In the ILCOR guideline for neonatal resuscitation, the clinical algorithms use noncrying as decision points for resuscitation steps. In this cohort, all nonbreathing infants were noncrying after birth (100%). The positive predictive value of noncrying infants was 50% for nonbreathing infants.

In the cohort of noncrying and nonbreathing infants at birth, more than two-thirds received simple measures of stimulation, and almost one-fourth of them received BMV. In the cohort of noncrying but breathing infants at birth, almost all of them received simple measures of stimulation, and 1 in 10 infants received BMV. Infants who are noncrying but breathing may have primary or secondary apnea. These infants would require additional stimulation to assist in transition to spontaneous breathing, whereas some would require BMV.

Among the noncrying but breathing infants after birth, 9.5% were nonbreathing at 1 minute and 2.0% did not breathe at 5 minutes and had 12-fold odds for predischarge mortality. Crying assists in clearing lung fluid, and it may assist in establishment of functional residual capacity. Noncrying of the infant coupled with perinatal depression might increase the risk of mortality. Further mechanistic studies are required to understand the physiology.

So far, the incidence of infants who received resuscitation at birth is mostly derived from intervention-based studies in which the effects of neonatal resuscitation interventions on clinical performance and mortality are reported.^19,20,30  For example, in a study conducted in Tanzania, Ersdal et al¹⁹  proposed that 16% of the infants who did not breathe spontaneously received basic resuscitation. From our study, of the noncrying infants after birth, 16.8% did not breathe at 1 minute and required BMV.

Defining the population of infants who require resuscitation is crucial for measuring the coverage of neonatal resuscitation interventions. The Every Newborn Action Plan Measurement Improvement Roadmap (2016–2020) prioritized research to validate the coverage indicators for neonatal interventions (including neonatal resuscitation) to improve tracking of national and global progress to the Sustainable Development Goals for newborn survival.^31,32  The Every Newborn–Birth Indicators Research Tracking in Hospitals study in Nepal, Bangladesh, and Tanzania is designed to validate these indicators for use in maternal-report surveys and in routine facility data.³³  Because documentation of facility data is often done by the health worker providing the care, ensuring measurement (both numerator and denominator) that is meaningful to reinforce clinical practice is crucial.³³ 

This study has several strengths, including that observation was done by trained researchers in multiple hospitals in Nepal in a consistent way and with a large sample size. Using observational methods limited recall bias.

However, there are several limitations. First, this observation was only done for vaginal deliveries and not for cesarean deliveries. Second, there might have been observer reporting bias, especially when the case load for deliveries was high; however, periodic quality assessment was done to reduce the bias. Third, we did not assess whether the infant cried or breathed after stimulation. Fourth, the physiologic response to hypoxia is gasping (deep and irregular breathing), which might have been captured by the clinical observers as breathing; however, these infants who were gasping would have required BMV. Finally, we did not have cord base values to determine if infants who did not cry had perinatal depression.

With this study, we provide incidence of noncrying and nonbreathing infants after birth, prediction of infants receiving resuscitation, and risk of neonatal mortality in Nepal. Eleven percent of the infants did not cry after birth. One-fifth of infants born by assisted vaginal delivery and/or with induction of labor (artificial rupture of membranes) were noncrying at birth. Almost half of these infants were nonbreathing after birth. Noncrying but breathing infants after birth had 12-fold odds of predischarge mortality compared with crying infants. All noncrying infants require further assessment and initial steps of resuscitation. With this study, we provide evidence of the association between noncrying and nonbreathing. Our findings support noncrying as an indicator for initiating resuscitation and closer monitoring and potentially as a denominator for coverage measurement to maintain integrity between clinical and measurement practice.

We thank Omkar Basnet and data collectors for the contribution to the study. Most of all, we thank the women, their newborns, and their families.

aOR

adjusted odds ratio

BMV

bag-and-mask ventilation

CI

confidence interval

HBB

Helping Babies Breathe

ILCOR

International Liaison Committee on Resuscitation

QI

quality improvement

SGA

small for gestational age