Single-step gastric aspirate shake test: bedside predictor of neonatal pulmonary morbidity.

Abstract

A simplified, single-step shake test has been performed on gastric aspirate samples from 85 preterm infants and a control group of 214 term infants. Respiratory symptoms were seen in 25 of 30 preterm infants with a negative or intermediate test, but in only 2 of 55 infants with a positive test (P < 0 001). No infant with a positive test developed respiratory distress syndrome (RDS) while RDS developed in 66% of those preterm infants with a negative test. False-negative results were not seen in the preterm group. The gastric aspirate shake test had better sensitivity and selectivity in prediction of clinical outcome than did the amniotic fluid lecithin/sphingomyelin ratio in 37 preterm infants in whom both results were available. Serial shake tests were performed on samples from a number of infants with RDS and the results were seen to change in parallel with clinical recovery. This and other observations suggest that the result of the gastric aspirate shake test depends more upon direct swallowing of fetal lung liquid than on swallowing of amniotic fluid. Estimation of the lecithin/sphingomyelin (L/S) ratio in amniotic fluid (Gluck et al., 1971) had made prediction of fetal lung maturity a reality. However, the equipment required to estimate the L/S ratio is not universally available and this test is not immediately available to all patients. The development of the foam stability test, or 'shake test', for use with amniotic fluid (Clements et al., 1972) provided expediency, simplicity, and economy yet preserved sensitivity and selectivity, but could only be performed in centres where amniocentesis was available. Sefton et al. (1972) assumed that the fluid in the newborn infant's stomach was amniotic fluid swallowed immediately before delivery and showed that the L/S ratio could be performed on this fluidwith results comparable to analysis of maternal amniotic fluid. Evans (1975) and Cowett et al. (1975) showed that this was also true of the shake test used with gastric aspirate. Why then has the shake test not become widely used when it has been shown to be a satisfactory clinical tool for prediction of neonatal pulmonary morbidity? Firstly, previous studies have used a seemingly complex series of dilutions (Cowett et al., 1975) or interpretations (Evans, 1975) in describing the test, which may have detracted from its acceptance. Secondly, reports of false-negative results in Received 30 November 1976 term babies as a source of error (Gebhart and Ruys, 1976) may have cast doubt on the specificity of the test. We have therefore prospectively assessed the gastric aspirate shake test performed at a single dilution in 85 preterm infants. In addition, the results have been compared to amniotic fluid L/S ratios obtained within 24 hours of delivery in 37 pre-term infants in whom both indices were assessed. Materials and methods Gastric aspirate samples were obtained within one hour of birth from 92 consecutive infants, of 37 weeks' gestation or less, admitted to the neonatal intensive care unit, by passage of a no. 10 Fr polyethylene suction catheter with a mucus trap.* Seven samples were discarded either because of inadequacy of sample (<0 71 ml) or because of contamination with blood or meconium, as recommended by Clements et al. (1972). The 85 infants included in this study had a mean (±SD) weight of 2348±640 g and a mean (±SD) gestation of 35 *3 ±1 * 8 weeks. The male : female ratio was 1 45 :1 and 16% were delivered by caesarean section. As a control, shake tests were *Argyle HR1 8888-257527 DeLee Suction Catheter with mucus trap.