As in other Mediterranean countries, human cystic echinococcosis (HCE) is considered an endemic disease in Spain.1 The implementation of prevention and control programs for hydatidosis/echinococcosis in the 1980s led to a significant decrease in human incidence rates (fig. 1); nonetheless, HCE remains one of the most important anthropozoonoses in Spain.2 The national epidemiological surveillance network is based on 3 interdependent systems: Compulsory Notifiable Diseases (CND), Outbreak Alerts (OA) and Microbiological Information (MI). All practicing doctors from both the public and private healthcare sectors are obliged to notify incident and suspected HCE cases to the CND, although since 1996, reporting is only required in those autonomous regions where the infection is considered endemic.3 Hence, the CND is regarded as the universal notification system in Spain, providing the official figures published in the Spanish Report on Trends and Sources of Zoonoses by the European Commission. The MI is based on voluntary weekly reporting of confirmed microbiological diagnoses of individual cases provided by a network of parasitology laboratories, mainly in hospitals. This system covers approximately 25% of the Spanish population, since only 5 of 19 autonomous regions in Spain currently report their cases. Because they have a lower coverage range and limited usefulness, the OA and MI systems generally play complementary and/or confirmatory roles. However, the effectiveness of the CND in HCE reporting may be seriously compromised by a number of factors derived from its operating design and the particular biological features of the disease: 1) errors and omissions made by overworked or poorly motivated professionals can lead to underreporting, and 2) the system is inadequately adapted to reporting diseases characterized by slow progression rates, long asymptomatic periods, and late diagnosis, as is the case of HCE. Indeed, the sensitivity of the CND in estimating the incidence of HCE was found to be 47% to 57% in the autonomous regions of Aragon4 and the province of Salamanca5 by the capture-recapture method and active search of hospital records, respectively. These findings clearly illustrate the need for new epidemiological data sources to help improve the low sensitivity of the CND.
Computerized hospital discharge records, particularly the minimum basic data set (MBDS), have been increasingly used to retrieve scientific evidence over the last 20 years in Spain. The MBDS is a nationwide medical-administrative database, approved in 1987 by the National Health System Interterritorial Council and based on recommendations of the European Union. The system uses clinical codes (CIE-10-MC) according to the International Classification of Diseases, tenth revision, Clinical Modification, and in 1998 it covered an estimated 97% of public hospitals in Spain.6 The quality and consistency of the MBDS relies on meticulous recording of data in hospital discharge reports and coding of the variables considered. Errors and omissions in these 2 key procedures have been reported frequently in the literature7–9 and indicate considerable differences in the quality of data between hospitals and regions. Because of its wide coverage, convenience and relatively low cost, the MBDS is a potentially powerful resource for epidemiological and research purposes, although it is important to bear in mind that the system was not set up as an instrument to promote diagnostic accuracy. Considered together, these observations indicate that the MBDS, like any other methodological tool, must be adequately validated as a research resource. In the last few years the MBDS has also been used to estimate the incidence of HCE in the province of Salamanca10 and the autonomous region of Madrid.11 Both surveys found relevant epidemiological data showing higher incidence rates than those expected or provided by the CND, confirming the lack of sensitivity of the CND reported previously. However, the reliability of the MBDSs investigated was not determined in either of the surveys (for example validating the data obtained against individual patient case reports), a fact that sheds doubts on the degree of accuracy of the figures presented.
In conclusion, it is now clear that the CND alone is insufficient for estimating the incidence of HCE in Spain. This task requires complementary epidemiological data sources, including the MI and the MBDS. However, the MBDS must be evaluated prior to its use in order to guarantee the quality and accuracy of the data provided.