Failure to screen for sickle cell disease at the University Hospital of the West Indies

ABSTRACT

OBJECTIVE: Universal neonatal screening for sickle cell disease using cord blood has taken place in the University Hospital of the West Indies (UHWI) since 1997. Comprehensive quality control of the screening programme has not been implemented. Our aim was to determine how many deliveries were left unscreened, and to ascertain factors contributing to failure to screen for sickle cell disease. METHODS: All live births in 1999 entered in the delivery book of the Labour Ward, UHWI, were scrutinized against the sample book of the Neonatal Screening Laboratory, Sickle Cell Unit. Identified unscreened infants were matched with a randomly chosen control, born on the same day. Time and mode of delivery, persons assisting the delivery, birth weight, level of urgency, location of delivery and Apgar scores at 1 and 5 minutes were collected for unscreened infants and screened controls. Secular variation within the year was assessed by comparing the frequency of unscreened births in each month to the previous month. Frequencies of potential determinants of failure to screen were tabulated. The independent predictive power of these potential determinants was assessed using conditional logistic regression, to account for correlations between predictors, and for the case-control design of the study. RESULTS: Of 2,763 live births in 1999, 139 (5 percent) infants were not screened for sickle cell disease. There was no statistically significant secular variation within the year (relative risk 1.02, 95 percent confidence interval 0.97, 1.07, x2 = 0.77, p = 0.38). Independently significant predictors of failure to screen were emergency delivery (odds ratio 4.8, 95 percent confidence interval 2.3, 10.0), nigttime delivery (OR2.5, 0.9) CONCLUSIONS: This study identified that 5 percent of infants born in UHWI in 1999 were discharged from hospital without having been screened for sickle cell disease. Statistically significant associations with failure to screen were found to be emergency delivery, night-time delivery, and low 5 minute Apgar score. The risk of being left unscreened increased 5-fold for emergency deliveries. Increased awareness of these risk factors may reduce failure to screen. Of course, such effort does not obviate the need for regular quality control, including timely identification and tracing of unscreened infants. (AU)