Correspondence (letter to the editor): Explanation of Methods Is Needed

Some explanations regarding the methods are needed for the study reported by Buth et al. (1). The defined daily dose (DDD) is not—as misleadingly presented—a recommended daily dose but an “assumed mean daily dose for the main indication.” The DDD is a merely technical unit of measurement.

Table 1 in the article shows for recipients of the individual substance group the proportion with a prescription of the selected relevant active substance from this group. In my opinion it is misleading to term this a prescription prevalence. Also, the proportion of patients taking lorazepam has not more than doubled between 2006 and 2016 but has increased 1.4 times. What has changed is the rank of lorazepam in all patients treated with benzodiazepines. The definition of long term use in Figure 3 is also misleading. According to the eMethods, long-term use was assumed if at least one relevant prescription was issued in each quarter of the patient year under observation.

The study, which is based on data from four regions, shows the pressing importance of routine monitoring for topics relevant for public health on the basis of a nationwide dataset that includes all national health insurance funds, since biases arise for all data sources, for example as a result of regional particularities or the type of data collection (in the case of Buth, incomplete data collection from pharmacies). In theory, such a dataset is available in the shape of DIMDI healthcare data (3), so that reliable data are feasible for members of statutory health insurance funds (with underestimates because of private prescriptions). But to be able to use these data, organizational hurdles will need to be cleared in order to facilitate more current data and timely analyses, which are urgently required for managing healthcare.