Intervention procedure: structured medication review

The medication review was performed by one of the pharmacists of the research group (ML, PhD student) who had completed advanced training in geriatric pharmacy and was experienced (2.5 years) in supplying LTC residents. The medication review procedure comprised the identification of DRPs including evaluation of the relevance, forwarding clinically relevant pharmacist recommendations to healthcare professionals and determining healthcare professionals’ feedback. Finally, pharmacist recommendations were determined in terms of actually implemented recommendations in order to assess resolved DRPs.

For the medication review, all available patient and medication-related information was accessed by the research pharmacist through review of residents’ medical records, including the medication charts in the LTC facilities, and considering the following items: prescribed medications, diagnosis, relevant laboratory data (e.g., renal status) and relevant physical parameters (e.g., weight, blood pressure). To enable a standardised procedure, a DRP checklist was used. Items related to the ‘prescribing problems’ were based on standard classification systems [6, 27, 28] that were adapted for this setting by including an additional category of potentially inappropriate medication (PIM) for older persons [29, 30]. These categories were completed in categories that covered ‘medication use problems’ and ‘documentation problems’. Table 1 shows all DRP categories and criteria for evaluation.

Explanation of DRP categories and evaluation criteria

Abbreviations: PIM Potentially inappropriate medication, PRN pro re nata, SmPC Summary of Product Characteristics

First, to identify DRPs, residents’ medication records were reviewed at the prescribing level according to the predefined DRP checklist. Drug information was obtained from the summary of product characteristics (SmPCs), a drug-drug interaction database used in public pharmacies (ABDA Database, ABDATA, Eschborn, Germany) and clinical therapy guidelines. Additionally, we considered DRPs at the medication use level by including further information about storage, documentation, dispensing as well as preparation and administration of medication. Therefore, nursing staff was observed in preparation (e.g., modification of medication, including crushing tablets or mixing medications) and administration of medication (e.g., verification of the right medication for the right resident and at the right time) during a morning drug round, once per resident. Drug-taking habits and health-related information, such as swallowing difficulties, were also collected from residents’ medication records and by a structured questioning of the nursing staff. Furthermore, therapy and documentation in the residents’ medication records were compared with stored and dispensed drugs. Documentation problems in the medication records were recorded in terms of correctness and completeness, such as missing information for PRN medication. All identified DRPs were evaluated in regard of the clinical relevance. Not relevant DRPs were excluded considering all available sources and clinical parameters. For example, the interaction of two antihypertensive drugs, which only needs to be considered at the beginning, could be neglected in long-term use where these additive effects may even be desired. Therefore, these DRPs were not classified as clinically relevant and not further forwarded to the physicians. Since PRNs could also potentially have been administered and thus pose a risk for the occurrence of DRPs, such as drug-drug interactions, we considered all prescribed PRN and chronic medications for the medication review.

Secondly, recommendations deemed relevant to resolving the identified DRPs were compiled. These included suggestions for changes such as discontinuing a medication without a clear indication or recommendations in the absence of documentation in medication records. To ensure consistency, accuracy, and relevance, the DRPs and recommendations were checked by a second experienced research pharmacist (SC). In cases of uncertainty, the DRPs were discussed until agreement was reached.

Third, the involved physicians (GPs and specialists), nursing staff and the community pharmacies were contacted, and the relevant pharmacist recommendations were forwarded to them and reconsidered in a multi-professional exchange. The healthcare professionals’ feedback was collected in regard of the acceptance of pharmacist recommendations. For the physician contact, a brief explanation of the scientific background and the sources used was added in written form. The nursing staff was usually contacted first and further relevant information to resolve the DRPs was collected, such as whether PRN medications were actually administered. Pharmacist recommendations concerning the nursing staff (e.g., storage) were discussed directly. Then, DRPs were forwarded to the physicians and, if necessary, also to the supplying pharmacies (e.g., supply issues). First choice of contact was a personal face-to-face meeting. Personal contact with nursing staff took place in the LTC facilities, with the physicians in their medical practices and with the community pharmacies in their pharmacies. If this was not possible, they were contacted by phone or in writing. Although contact with different healthcare professionals was possible for a DRP, the healthcare professional (physician, nursing staff, or supplying pharmacist) who was primarily involved in the resolution of the DRP was counted for evaluation.