Institute of General Practice and Family Medicine
  1. Homepage LMU Hospital
  2. Institute of General Practice and Family Medicine
  3. Research
  4. Studies
  5. VALUE Study

VALUE Study

Validation of algorithms for the detection of adverse events in routine inpatient data

  • The detection of adverse events (AEs) and adverse drug reactions (ADRs) in routine data is of particular importance for various reasons. On the one hand, the identification of AEs is necessary in order to be able to investigate new relationships between drugs and health damage in the context of pharmacoepidemiological research or to quantify known relationships.

    ADRs, on the other hand, are needed, among other things, to monitor the status quo of drug therapy safety, derive potential for improvement and evaluate the effectiveness of measures to improve drug therapy safety, but also to point out possible ADRs in clinical practice.

    As part of the cross-consortia use case POLAR of the Medical Informatics Initiative, algorithms were therefore developed that can be implemented in routine data from inpatient care in order to detect ADRs and potential ADRs. The focus was on AEs that were prioritized by an expert consensus procedure (RAND/UCLA Appropriateness Method) due to their high clinical relevance. Each algorithm consists of a logical combination of variables that can be measured in structured data sources (e.g. ICD codes for diagnoses, OPS codes for procedures, laboratory values, ATC codes for drugs). In order to take account of the different data availability in different settings, a selection of different algorithms was developed for some UEs, which differ in terms of the type and number of data sources required.

    Both for the use of the algorithms in pharmacoepidemiological research and in the context of clinical practice, the sensitivity and specificity with which the algorithms can detect actual AEs/UAWs are crucial. Empirical validation is therefore necessary in order to be able to assess the benefits of the algorithms developed for the detection of AEs and ADRs in science and practice.