Background Transitions of care are risky periods for development of medication-related problems. The authors aimed to identify any problems experienced by patients following an admission to the hospital's Acute Medical Unit and to pilot discharge telephony follow-up. Pharmacist from NHS Direct (our partner for the project) conducted follow-up interviews with selected patients after discharge using their inhouse systems which are set up nationally to handle calls 24 h a day about any health-related matters.

Purpose To describe and quantify medication-related problems in a sample of patients discharged from hospital.

Materials and methods Eligible patients were short-stay admissions (<3 days) to the Acute Medical Unit of the Chelsea and Westminster Hospital. Consented patients had their discharge summary relayed to NHS Direct, who then administered a structured telephone survey 3 weeks after discharge. The pharmacist categorised and attempted to remedy any problems identified. The categories were possible side effects; concordance/compliance; difficulties with packaging; misunderstanding/misinterpretation of directions; other. Responses were fed back to the project team and assessment was made of the potential for harm from their medicines.

Results 54 patients were initially consented; 34 were contacted and 7 were removed from analysis. 20 medication-related problems were identified in 12 patients (44.4%): five potential side effects; five problems with taking medication and four felt that their medication did not suit them. NHS Direct identified one other medication-related problem and three patients received counselling for other medication issues. Only one problem was considered potentially harmful. 19 (70.4%) found the call to be helpful and 25 (88.9%) would like to have a similar follow-up call if admitted to hospital again.

Conclusions Nearly half our cohort was reported to be experiencing medication-related problems, though a low level of potential harm was found. Many patients initially recruited were not able to be contacted by phone. This suggests that although acceptable to those patients who were contacted, before the service can be offered widely methods for targeting need to be explored further.