Predicting preterm infants’ length of hospital stay (LOS) using gestational age, birth weight, and early breastmilk feeding

Abstract

Aims

The study aims to find a mathematical formula for predicting the length of hospital stay (LOS) of preterm infants with a case study from a hospital in Banyumas Regency, Central Java, Indonesia.

Method

The study uses a retrospective approach with a convenience sample of 96 preterm infants. The outcome variable was the preterm infants’ LOS. The predictor variables were the infants’ gestation, birth weight, and early oral feeding. Data were collected from a medical record at a hospital in Banyumas Regency, Central Java Province, Indonesia. The respondents’ inclusion criteria were preterm infants born before 37-week gestation at the hospital and treated at levels 1, 2, or 3 neonatal care in 2019. The respondents’ exclusion criteria were preterm infants born in another hospital and referred to the hospital because their data were often incomplete. Three experts designed the instrument for data collection with a validation agreement of 100%. The data collection was conducted between March and August 2021. The statistic analysed used univariate analysis, Kolmogorov–Smirnov test, Spearman Rank, linear regression, and ANOVA test.

Results

The mathematical formula for predicting preterm infants’ LOS was y = 64.939 − 1.306 (infant's gestational age) − 0.007 (infant's birth weight) + 1.294 (early oral feeding). The ANOVA test shows that the mathematical formula was feasible to predict preterm infants’ LOS in Indonesia (p = 0.000). The R² was 0.490, which indicates that the mathematical formula could predict the Indonesian preterm infant's length of hospital stay at 49%.

Conclusions

The predictors of preterm infants’ LOS are gestational age, birth weight, and early breast milk feeding. Therefore, nurses at levels 1, 2, and 3 neonatal care should give attention to infants’ gestation and birth weight. Nurses should facilitate early feeding as soon as possible to increase daily weight and shorten the LOS.