NEONATAL INTENSIVE CARE UNIT CENSUS INFLUENCES DISCHARGE OF MODERATELY PRETERM INFANTS
Profit J, McCormick MC, Escobar GJ, et al. Pediatrics. 2007;119:314-319.
A prospective multicenter cohort study was conducted with 850 infants at 10 neonatal intensive care units in Massachusetts and California to evaluate the impact of unit census on the decision to discharge moderately preterm infants.
Infants with gestational ages between 30 and 34 weeks plus 6 days were recruited from 3 level II and 7 level III neonatal intensive care units upon discharge. Data were abstracted from the patients' charts, hospitals' administrative records, and 3-month postdischarge follow-up telephone survey.
Researchers divided each unit's census into quintiles. It is expected that discharges should be evenly distributed among quintiles, with 20% occurrence in each. Across all quintiles, unit census and discharges were correlated (rs = 0.79, P = .007). In the lowest quintile of unit census, patients were less likely to be discharged (standard discharge rate = 0.8, P <= .01). In the highest quintile of unit census, patients were more likely to be discharged (standard discharge rate = 1.32, P < .001).
Using multivariate logistic regression, patient-nurse ratio was the only variable associated with discharge in the lowest quintile of unit census (odds ratio = 0.04, P < .001), with discharge more likely at a lower patient-nurse ratio. For patients discharged in the highest quintile of unit census, the odds of discharge were higher with a higher patient-nurse ratio (odds ratio = 14.2, P = .003). This opposing influence of patient-nurse ratio at low and high census is explained by the finding that as the census rises, the patient-nurse ratio also rises and vice versa.
The researchers note that discharges depend not only on clinical status but also on unit workload. They provide several recommendations for improved work flow and efficiency of patient discharge, including case management, primary nursing, and clear discharge guidelines.
VALIDATION OF OXYGEN SATURATION MONITORING IN NEONATES
Shiao SP, Ou C. Am J Crit Care. 2007;16(2):168-178.
A prospective clinical study of 78 neonates was conducted to validate oxygen saturation readings on clinical monitors against laboratory analyses and to present an oxyhemoglobin dissociation curve for neonates.
Blood samples were obtained via umbilical artery and vein catheters in the neonates. Differences between measurements were examined using paired t tests. A multivariate linear mixed model was used to examine differences between oxygen saturation, mixed effects, and repeated measures. Oxyhemoglobin dissociation was examined by the use of multiple regression curve-fitting analysis.
There was a 3% bias for oxygen saturation against oxyhemoglobin for arterial and venous samples as well as total arterial and venous samples (all P < .001). Regarding the effects of fetal hemoglobin, it was found that the greater the percentage of fetal hemoglobin, the smaller the difference between measurements of the blood oxygen saturation and blood oxyhemoglobin.
In summary, clinical monitors were less precise than blood analyses for neonates. Since safety ranges for oxygen saturation are narrower for neonates than for adults, clinical guidelines for neonates may require modification.