RESUMEN
Preterm birth is the leading global cause of neonatal morbidity and mortality. Reliable gestational age estimates are useful for quantifying population burdens of preterm birth and informing allocation of resources to address the problem. However, evaluating gestational age in low-resource settings can be challenging, particularly in places where access to ultrasound is limited. Our group has developed an algorithm using newborn screening analyte values derived from dried blood spots from newborns born in Ontario, Canada for estimating gestational age within one to two weeks. The primary objective of this study is to validate a program that derives gestational age estimates from dried blood spot samples (heel-prick or cord blood) collected from health and demographic surveillance sites and population representative health facilities in low-resource settings in Zambia, Kenya, Bangladesh and Zimbabwe. We will also pilot the use of an algorithm to identify birth percentiles based on gestational age estimates and weight to identify small for gestational age infants. Once collected from local sites, samples will be tested by the Newborn Screening Ontario laboratory at the Children's Hospital of Eastern Ontario (CHEO) in Ottawa, Canada. Analyte values will be obtained through laboratory analysis for estimation of gestational age as well as screening for other diseases routinely conducted at Ontario's newborn screening program. For select conditions, abnormal screening results will be reported back to the sites in real time to facilitate counseling and future clinical management. We will determine the accuracy of our existing algorithm for estimation of gestational age in these newborn samples. Results from this research hold the potential to create a feasible method to assess gestational age at birth in low- and middle-income countries where reliable estimation may be otherwise unavailable.
RESUMEN
BACKGROUND: To avoid biased estimates of standard errors in regression models, statisticians commonly limit the analytical dataset to one observation per patient. OBJECTIVE: Measure and explain changes in model performance when a model predicting 30-day risk of death or urgent readmission (derived on a dataset having one hospitalization per patient) was applied to all hospitalizations for study patients. METHODS: Using administrative data from Ontario, we identified all hospitalizations of 499,996 patients between 2004 and 2009. We calculated the expected risk for 30-day death or urgent readmission using a validated model. The observed-to-expected ratio was determined after categorizing patients into quintiles of rates for hospitalization, emergent hospitalizations, hospital day and total diagnostic risk score. RESULTS: Study patients had a total of 858,410 hospitalizations. Compared with a dataset having one hospitalization per patient, model performance declined significantly when applied to all hospitalizations [c-statistic decreased from 0.768 to 0.730; the observed-to-expected ratio increased from 0.998 (95% confidence interval 0.977-0.999) to 1.305 (1.297-1.313)]. Model deterioration was most pronounced in patients with higher hospital utilization, with the observed-to-expected ratio increasing to 1.67 in the highest quintile of emergent hospitalization rates. CONCLUSIONS: The accuracy of predicting 30-day death or urgent readmission decreased significantly when the unit of analysis changed from the patient to the hospitalization. Patients with heavy hospital utilization likely have characteristics, not adequately captured in the model, that increase the risk of death or urgent readmission after discharge from hospital. Adequately capturing the characteristics of such high-end hospital users may improve readmission models.