RESUMEN
BACKGROUND: The Medical Death Certificate and the Death Statistics Bulletin were unified and implemented in the year 2009 in Spain. National statistics detected an unusual increase for diabetes mellitus (DM) and hypertensive disease (HT) deaths in 2009, in relation to previous years trend.The objective is to study the documental causes of the increase, and describe the procedures and consequences in rates, after the revision and recodification of DM and HT. METHODS: All death certificates in 2009 for diabetes and hypertension in the Region of Murcia (cases=670) were revised, according to previous guidelines for direct recoding after consultation to the certifying physician. A telephone survey to certifying physician was designed to determine the accuracy of the pattern of recoding. Kappa index and 95% confidence intervals (95%CI) were performed between initial and recoded causes. Confirmation rate and 95%CI was estimated after phone inquiry to the certifying physician, and the annual age-adjusted and age-specific rates from 1999 to 2009 (uncorrected and corrected) for DM and HT were calculated. RESULTS: Simple agreement was 37% for DM and 30% for HT. The Kappa index between the initial and final causes was 49% (95%CI, 45 to 54%). Confirmation rates were 47% (95%CI, 43 to 52%) for DM and 38% (95%CI, 34 to 43%) for HT. The initial annual rates of 2009 for DM were corrected from 21.4 per 100,000 inhabitants to 17.1, and from 19.0 to 14.0 for hypertension. The respective specific age rates of 70 to 84 and older experienced similar reductions. CONCLUSIONS: The revision restored temporal trends in mortality of DM and HT in 2009, and identified no variations from previous years. It was detected that the erroneous fulfillment of DM AND HT came from the new death certificate.
