Nationwide data base on medically diagnosed spontaneous abortions in Finland

The content and technical quality of the Finnish data base on medically diagnosed spontaneous abortions, retrieved from hospital discharges and polyclinic records, are described. The validity and suitability of the data for epidemiological studies is also evaluated. The rate was 8.9 in 1973-83 and it increased from 7.8 to 10.2 during the study period. The rates are at the same level as the ones obtained in other register-based studies, and slightly lower than those of the interview studies. A comparison of self-reported spontaneous abortions to those in our data base showed that 20% of all the abortions reported by the nurses, and 17% of those by the solvent-exposed women, were not found in the hospital register. About half of these are likely to be due to erroneous personal identification codes in the records of the patients. The absence of a spontaneous abortion was related to the length of gestation, with early abortions more often missing. Women in different occupational groups confirmed spontaneous abortions retrieved from our data base to a different degree: nurses in 91% of the cases, whereas industrial workers in only about 79% of the cases. Confirmation was also related to time elapsed since abortion. As the nationwide data base on medically diagnosed spontaneous abortions provides data, independent of an individual's own definition, recognition and reporting, it is a powerful tool in retrospective reproductive studies. Its limitations regarding early abortions need to be recognized, however.