gms | German Medical Science

17th Annual Meeting of the German Drug Utilisation Research Group (GAA)

Gesellschaft für Arzneimittelforschung und Arzneimittelepidemiologie

25.11. - 26.11.2010, Osnabrück

Identification of subjects at risk using a medication-based algorithm: first results of the medication module of a case-management program (Casaplus®)

Meeting Abstract

Search Medline for

  • corresponding author Sebastian Harder - Institut für Klinische Pharmakologie Universitätsklinik, Frankfurt am Main, Germany
  • Christian Schug - Institut für Klinische Pharmakologie Universitätsklinik, Frankfurt am Main, Germany
  • Julia Fleckenstein - Institut für Klinische Pharmakologie Universitätsklinik, Frankfurt am Main, Germany

Gesellschaft für Arzneimittelanwendungsforschung und Arzneimittelepidemiologie e.V. (GAA). 17. Jahrestagung der Gesellschaft für Arzneimittelanwendungsforschung und Arzneimittelepidemiologie. Osnabrück, 25.-26.11.2010. Düsseldorf: German Medical Science GMS Publishing House; 2010. Doc10gaa26

doi: 10.3205/10gaa26, urn:nbn:de:0183-10gaa263

Published: November 22, 2010

© 2010 Harder et al.
This is an Open Access article distributed under the terms of the Creative Commons Attribution License ( You are free: to Share – to copy, distribute and transmit the work, provided the original author and source are credited.



Context: Recently, an algorithm was developed which based on multivariate predictors for an early hospitalisation and comprises variables like sex, age, previous hospitalisations, kind of diagnosis, and number of prescriptions. These high-risk multi-morbid patients aged 55 or more were voluntarily enrolled in the case management Casaplus® addressing geriatric related health risks like falls, malnutrition, cardiovascular and gastrointestinal diseases. A further algorithm should identify Casaplus® members at risk for medication induced hospitalisations. Those candidates then were scheduled for a “medication-tailored” case management within Casaplus®.

Material and Method: A sub-algorithm which consisted of 10 combinations of chronic or intermittent prescriptions and 3 drug-disease combinations should identify patients with medication related problems (see Harder et al. Candidate patients were identified by their prescription history (ACT-coded, referring to the previous year) taken from the medical insurance records. Incident patients were contacted and underwent a problem-oriented case management by experienced pharmaceutical health care advisers which was aimed to advice patients regarding appropriate medication use. The curriculum covered relevant information from 7 categories of general problems with medications. The counselling was complemented by a written prescription history for the responsible physician in order to enhance transparency. We now report on the first data on the performance of this part of the Casaplus® program.

Results: The algorithm was applied to a cohort of 7424 patients who participate in the Casaplus® program (see above). Overall, 579 patients (=candidates; 7.8% of all pts.) have been detected by the sub-algorithm as having a risk-prone drug-drug- or drug-disease- combination. 211 patients (= participants, 2.8% of all pts.) actually had been subject to counselling, in case of the remaining patients, the incident drug-drug- or drug-disease- combination was not longer present at the time of the contact interview or a counselling intervention was declined. In the group of candidates, 461 pts. had 1, 85 pts. had 2 and 20 pts. had >2 problematic drug-drug combinations, these numbers were 184, 11 and 0 for the group of participants. Most often, candidates and participants had the risk-prone drug-drug combination “ACE-inhibitors or ARB + potassium-sparing diuretic” (C09 + [C03AB or C03D or C03E]; N=171 candidates and 91 participants), “platelet inhibitor + NSAR” (B01AC + [M01A excl. M01AB55]; N=116 candidates and 13 participants), and “cardiac glycoside + diuretic w/o potassium-sparing effects or potassium suppl.” (C01AA + [C03AA or C03BA or C03CA or C07B]; N=106 candidates and 43 participants). Symptomatic heart rhythm disturbances together with the use of tricyclic antidepressants were the most often detected drug-disease combination (N=61 candidates and 15 participants). In the group of participants, a total of 214 interaction-related and 935 general medication-related problems like e.g. keeping track of medications or application/handling problems, have been addressed during the counselling.

Conclusions: It is concluded that a considerable proportion of patients with a high risk profile for hospitalisation had additional risk-prone pharmacotherapy, albeit the total number of candidates was lower than expected, possibly due to a selection bias (voluntary participation and informed consent). The largest cluster of potentially harmful combinations involves NSAR and drugs which might interfere with potassium levels. It needs to be shown whether counselling and case management will prevent drug induced hospitalisations.