Background Uses of administrative healthcare data are increasing, particularly with respect to both clinical and health services research, where investigators have used these data to identify at-risk patients and to explore variation in healthcare services and outcomes across hospitals or regions.
Purpose To determine the accuracy of ICD-9-CM administrative codes for neonatal conditions associated with congenital malformations.
Methods The study population consisted of newborns admitted to Childrens Hospital Los Angeles in 2002. ICD-9-CM discharge codes reported to the California Office of Statewide Health Planning and Development were obtained. Data abstracted from clinical charts were merged with administrative data to categorize each patient by both clinical chart review (gold standard) and administrative data. The frequency, sensitivity, specificity, accuracy, predictive value positive, predictive value negative were calculated for each of 36 neonatal conditions. Data were analyzed using SAS (Statistical Software Package v.9.0, Cary, NC) and Stata/SE 8.0 (Stata Corporation College Station, TX).
Results Overall, the accuracy of administrative data coding was generally high with over 90% of the clinical diagnoses accurately coded within each organ system. The coding specificities for most congenital malformations were usually better than 90%. On the other hand, coding sensitivities varied widely, and for some conditions were lower than 20%.
Conclusions Although not uniformly sensitive, when compared to chart review, administrative data tended to provide accurate and specific identification of congenital malformations among newborns.
Statistics from Altmetric.com
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.