Objective: The National Mortality Register (NMR) is used to monitor and prioritize health problems, but it may underestimate deaths from external causes, among which are deaths from acute reaction to psychoactive substances (ARPS). To evaluate the discrepancies between death from ARPS collected in the NMR and those collected in the specific register of the State Information System for Drug Addiction (SEIT).
Methods: We evaluated the discrepancy between ARPS deaths in 15-49 years olds registered in the NMR (code E850-858) and deaths from acute reaction to opium and cocaine in the SEIT, in Barcelona, Bilbao, Madrid, Sevilla, Valencia and Zaragoza, between 1984-93.
Results: 3,491 ARPS deaths were registered in SEIT, 1,285 in NMR. Varying degrees of discrepancies appeared between the two registers in all six cities. The NMR in Madrid showed 92% fewer deaths than the SEIT, while in Barcelona the NMR figure was 23% lower. Between 1984 and 1993 the differences between registers in all cities, except Madrid and Seville, narrowed (p < 0.01).
Conclusions: Even though detection of ARPS deaths in the NMR has been improving, it still underestimates the real situation, and is therefore not a useful tool in the evaluation of temporal-spatial variations. To improve detection of ARPS deaths (and those from other external causes) a systematic linkage with data from the coroners' registers must be done.