The aim of the study is to validate at the biochemical level (presence of myocardial damage) the discharge diagnosis code ICD-9-CM 410.x1, and to compare the acute myocardial infarction
based on pure administrative data with the epidemiology
based on troponin
and clinical data.
The health-related administrative databases
of the Italian Region Friuli Venezia Giulia were used as the source of information. All the databases are anonymous and can be linked with each other at the individual patient level through a univocal stochastic key. Two methods were used to assess incidence in 2017: the first used the main hospital discharge diagnosis, validated by biochemical myocardial necrosis; the second identified from the cohort of all patients with any myocardial injury those with ischemic origin.
The positive-predictive value of the clinical diagnosis of AMI (410.x1), validated at the biochemical level, was 96.2%.
About 40% of patients with a not trivial biochemical myocardial injury and an ischemic heart disease diagnosis (e.g. 411) were discharged without either ST-elevation myocardial infarction (STEMI) or non-ST elevation myocardial infarction (NSTEMI) diagnosis, leading to a sensitivity of clinical discharge diagnosis of 47.6%.
Thirty-day and 90-day mortality at multivariate analysis resulted respectively, 1.8 and 4.0% in NSTEMI, 6.6 and 9.8% in STEMI, 8.8 and 12.2% in patients with biochemical AMI and discharge diagnosis other than 410.x1.
Pure administrative data (clinical discharge diagnosis) are today insufficient to catch the whole hospital epidemiology
of myocardial infarction missing an important proportion of AMI with an adverse prognosis comparable with STEMI.