[Acute myocardial infarction incidence estimated using a standard algorithm based on electronic health data in different areas of Italy].

Aim: to define and implement an algorithm, based on current databases, in order to estimate acute myocardial infarction (AMI) incidence in six Italian areas.

Setting: Local Health Units of Firenze and Venezia, and the municipalities of Pisa, Roma, Taranto, and Torino.

Participants: residents in the above mentioned six areas in the period 2002-2004, for a total of about 4,447,000 subjects (30th June 2003).

Methods: acute myocardial infarction incident cases were identified through hospitalization databases and causes of death. Hospital discharges (excluding outpatient discharges) with ICD9-CM code 410* as primary discharge diagnosis, or as secondary diagnosis when associated with selected codes suggestive of ischemic symptoms in primary diagnosis, and deaths with the ICD9-CM code 410* as underlying cause were selected. Patients without a previous hospitalization for ICD9-CM codes 410* and 412* during the previous 60 months were considered as incident cases. Crude, age-specific and age-standardized incidence rates (standard: total Italian population at the 2001 census) were calculated. A number of data quality indicators were also evaluated.

Results: age-standardized incidence rates show different levels of incidence in the areas included in the study. Both for males and females, higher incidence is observed in Rome and Turin (males: respectively 260.5 and 260.2 cases/100,000; females: 105.6 cases/100,000 in both areas). The lowest incidence is observed in Taranto (males: 219.5 cases/100,000; females: 87.0 cases/100,000). Quality indicators suggest a good comparability of incidence estimates among the studied areas. In particular, in both genders, the differences observed in the incidence rates are consistent with the differences of current AMI mortality rates.

Conclusions: although limitations in data comparability among the studied areas and in the quality of disease coding cannot be completely excluded, results suggest that the algorithm we used provides estimates of AMI incidence rates comparable among the studied areas. Only a validation study with standardized criteria will make it possible to more closely evaluate the diagnostic quality and comparability of AMI cases identified through this algorithm.