Bernsten 2001.
| Methods | Cluster‐randomised trial | |
| Participants | 2454 general/elderly patients (Intervention 1290; Control 1164) (86 control sites and 104 intervention sites) Community pharmacies in Denmark, Germany, The Netherlands, Northern Ireland (co‐ordinating centre), Portugal, Republic of Ireland and Sweden Year of study: Not stated. |
|
| Interventions | Community pharmacists provided pharmaceutical care to patients in the intervention group including patient assessment, identification of actual and potential drug‐related problems (e.g. poor compliance, poor knowledge, adverse drug reactions). Data sources included (i) the patient (by informal questioning); (ii) the patient’s general practitioner (GP); and (iii) pharmacy‐held records. Pharmacy interventions included: (i) educating the patient about drug regimen and medical condition(s); (ii) implementing compliance‐improving strategies such as drug reminder charts; and (iii) rationalising and simplifying drug regimens in collaboration with the patient’s GP. Continuous process Duration 18 months |
|
| Outcomes | Hospitalisations over past 18 months | |
| Notes | Funding source: European Commission, under the BIOMED 2 programme for medical research, funded the coordination of this multicentre study Conflict of interest: Not stated |
|
| Risk of bias | ||
| Bias | Authors' judgement | Support for judgement |
| Random sequence generation (selection bias) | Unclear risk | Half of the recruited sites were randomly assigned as control sites and half as intervention sites and, where possible, control and intervention sites were matched as closely as possible according to size (e.g. total number of patients served), situation (e.g. city centre vs village) and type (e.g. owned by a single proprietor vs part of a national chain). |
| Allocation concealment (selection bias) | Unclear risk | No information provided |
| Blinding of participants and personnel (performance bias) All Outcomes/Outcome 1 | Unclear risk | Not stated but unlikely due to intervention pharmacist training |
| Blinding of outcome assessment (detection bias) All Outcomes/Outcome 1 | Low risk | Hospitalisations are an objective measure |
| Incomplete outcome data (attrition bias) All outcomes | High risk | Between group attrition < 10% but, large changes in sample size due to some arms only running for 6 or 12 months. |
| Selective reporting (reporting bias) | Unclear risk | Major results are reported. Unclear why some results presented by country and some averaged across all |
| Other bias | Unclear risk | Intervention was not the same across all countries Quote: "Each country adapted the manual, translating and modifying sections where appropriate, according to differing national practices." |