Completeness of state administrative databases for surveillance of congenital heart disease

Area of Science:

• Public Health Surveillance
• Cardiology
• Data Quality Assessment

Background:

• Accurate tracking of congenital heart defects (CHD) birth prevalence is crucial for identifying clusters and risk factors.
• Resource constraints often necessitate using existing state administrative databases for defect ascertainment.
• This study assesses the data quality of state administrative databases for identifying CHD and specific diagnoses.

Purpose of the Study:

• To evaluate the accuracy of state administrative databases in identifying congenital heart defects (CHD).
• To compare the effectiveness of different state data sources for CHD ascertainment.
• To determine the performance of these databases in matching specific CHD diagnoses.

Main Methods:

• Infant medical records with CHD (n=373) born 1997-1999 were abstracted for diagnoses from definitive reports.
• Abstracted CHD data were linked to state birth/death records and hospital discharge summaries.
• State database codes (generic CHD) and exact diagnosis matches were evaluated against abstracted data.

Main Results:

• State databases identified 58% of generic CHD cases.
• Postnatal hospital discharge summaries identified 48% of cases, followed by birth summaries (27%), birth certificates (9%), and death records (4%).
• Exact matches were found for 52% of 633 specific CHD diagnoses, with postnatal discharge summaries yielding the most matches.

Conclusions:

• State databases captured 60% of generic CHD and approximately half of specific CHD diagnoses.
• Postnatal hospital discharge summaries demonstrated the highest performance for both generic CHD identification and specific diagnosis matching.
• Vital records showed limited utility for ascertaining CHD data.