Improving case ascertainment of a population-based birth defects registry in New York State using hospital discharge data

Abstract

The assessment of the data quality of population-based registration systems is essential to understanding the reliability and usefulness of disease surveillance and research findings resulting from the use of registry data. Since the New York State Congenital Malformations Registry (CMR) uses passive case ascertainment, the completeness of the registry data is an important aspect of the quality of information. This paper presents the results of hospital audits, which were conducted to capture the unreported cases using hospital discharge files, and evaluates the effectiveness of the audits. Children age 2 years or younger and diagnosed with reportable birth defects for the birth years 1998-2000 were selected from hospital discharge files of all reporting hospitals in the New York Statewide Planning and Research Cooperative System (SPARCS) and matched to the CMR database for the same birth year period. The unmatched reports from the SPARCS hospital discharge files that the CMR possibly missed were sent to hospitals, requesting submission of the missed reports. Two audits on all reporting hospitals in New York State were conducted: 1) 1998 and 1999 birth cohorts audited from June 2000 to March 2002, and 2) 2000 birth cohort audited from November 2001 to November 2002. Hospital audits using SPARCS hospital discharge data identified 5,460 reports that the CMR missed for the selected 66 hospitals analyzed. About 86% of these reports had reportable conditions and were added to the CMR, which comprised 21.4% of all reports from the 66 hospitals for the birth years 1998-2000. The number of reports that would have been missed without audits decreased from the 1998 and 1999 birth cohort (25.1%) to the 2000 birth cohort (13.9%). Low reporting rates and, thus, a high percent of added reports, were found for hospitals with a relatively small number of annual reports and for some specific birth defects such as chromosomal anomalies, anencephalus and congenital anomalies of the urinary system. The current study demonstrates that using hospital discharge data to improve case ascertainment is a valuable and effective method of enhancing birth defect surveillance, particularly for those hospitals with low reporting rates.