This review provides information on the overall incidence of D-DIs as a cause of adverse patient outcomes, although there is still uncertainty about the impact of D-DIs on adverse patient outcomes. Our results suggest that a limited number of drugs are involved in the majority of cases and that the number of reasons for admission as a consequence of D-DIs seems to be modest.

Although the number of clinically relevant drug-drug interactions (DDIs) is probably low, DDIs may be responsible for a substantial number of hospital admissions. In some countries, the pharmacist is responsible for preventing the use of unsafe or non-effective drug regimens. Specifically they should avoid the dispensing of combinations of drugs that may cause serious DDIs. In order to assess the determinants related to community pharmacies and associated with these dispensings, a systematic literature review was conducted. Medline and International Pharmaceutical Abstracts were searched for articles published in English between 1993 and 2003. Additional relevant articles were identified by screening the reference lists of relevant articles. Seven papers were located. The determinants described in the literature were divided into three groups. The first group focussed on the relationship between the pharmacist and the prescriber. The number of prescribers is of importance as well as the number of dispensing pharmacies. Both a high number of primary care physicians and multiple dispensing pharmacies increased the risk of DDIs. The availability, quality and sensitivity of the medication surveillance software appeared to be a second important determinant. Both too many and too few signals increased the risk of dispensing interacting drugs. The third group of determinants was related to the pharmacist and pharmacy organisation. Signals from the surveillance program are usually judged first by technicians and subsequently managed by the pharmacist. Consequently, knowledge, instructions and supervision are important determinants. A fourth group of determinants was identified in literature assessing interventions by pharmacists, including interventions for DDIs. A higher workload was associated with lower intervention rates, which indicated a higher risk of dispensing interacting drugs. The determinants identified in this review can be used to develop strategies to minimise patient harm resulting from DDIs. Further assessment of the relation between these determinants and the dispensing of DDIs and of the relation between DDI-associated dispensing and patient harm is recommended.

Objective: There are many drug-drug interactions (D-DI) of which some may cause severe adverse patient outcomes. Dispensing interacting drug combinations should be avoided when the risks are higher than the benefits. The objective of this study was to identify determinants of dispensing undesirable interacting drug combinations by community pharmacies in the Netherlands. Methods: A total of 256 Dutch community pharmacies were selected, based on the dispensing of 11 undesirable interacting drug combinations between January 1st, 2001 and October 31st, 2002. These pharmacies were sent a questionnaire by the Inspectorate for Health Care (IHC) concerning their process and structure characteristics. Main outcome measure: The number of times the 11 undesirable interacting drug combinations were dispensed.Results: Two hundred and forty-six questionnaires (response rate 96.1%) were completed. Dispensing determinants were only found for the D-DI between macrolide antibiotics and digoxin but not for the other 10 D-DIs. Pharmacies using different medication surveillance systems differed in the dispensing of this interacting drug combination, and pharmacies, which were part of a health care centre dispensed this interacting drug combination more often. Conclusion: Medication surveillance in Dutch community pharmacies seems to be effective. Although for most D-DIs no determinants were found, process and structure characteristics may have consequences for the dispensing of undesirable interacting drug combinations.

Ill-founded off-label prescribing in Dutch general practice is limited for 48 relevant combinations of drugs and off-label indications. In order to overcome limitations in registration databases, it is useful to look at as much information as possible--for example, co-medication and co-morbidity--when determining off-label prescribing. Studying ill-founded off-label prescribing should be performed on a day-to-day basis, especially for recently introduced drugs.