Skip to main content
. 2016 Feb 12;2016(2):CD009095. doi: 10.1002/14651858.CD009095.pub3

Zermansky 2006.

Methods RCT (randomised by patient)
Total study duration: 6 months
Participants 661 (331 intervention, 330 control) care home residents, 65 care homes
Setting: Nursing and residential homes for older people
Age: Intervention mean 85.3 (IQR 81‐90); control mean 84.9 (IQR 80‐90)
Gender: Intervention 75 (22.7%) male; control 79 (23.9%) male
Country: UK
Date of study: 2002
Interventions Medication review by a single pharmacist
A clinical medication review was conducted by the study pharmacist who held a postgraduate qualification in clinical pharmacy, within 28 days of randomisation. It comprised a review of the GP clinical record and a consultation with the resident and carer. The pharmacist formulated recommendations with the resident and carer and passed them on a written proforma to the GP for acceptance and implementation. GP acceptance was signified by ticking a box on the proforma. Control patients received usual GP care
Outcomes Measured at baseline and six months ± three weeks post‐randomisation:
Hospital admissions (non‐elective)
Mortality
Medication‐related problems
Medicine costs (cost of 28 days of repeat medicines per participant)
Not used for this review:
Number of changes in medicines per participant
Number of medicines per participant
Recorded medication reviews
Falls
SMMSE
Barthel index
Number of GP consultations
Notes Funded by The Health Foundation, 90 Long Acre, London WC2 9RA (Registered Charity Number 286967)
Risk of bias
Bias Authors' judgement Support for judgement
Random sequence generation (selection bias) Low risk Patients were randomised in randomly sized blocks of 2 to 8 patients using an algorithm written in Visual Basic in Microsoft Access
Allocation concealment (selection bias) Low risk Not reported in paper. Allocation was concealed to the research pharmacist and nurse data collector by statistician [Author contacted]
Blinding of participants and personnel (performance bias) 
 All outcomes High risk Open design, no blinding attempted
Blinding of outcome assessment (detection bias) 
 Subjective outcomes High risk No blinding conducted
Blinding of outcome assessment (detection bias) 
 Objective outcomes Low risk No blinding conducted, however outcomes not likely to be influenced by lack of blinding
Incomplete outcome data (attrition bias) 
 Primary outcomes Low risk Similar attrition in both groups with similar reasons for dropouts. Described as intention‐to‐treat by authors
Incomplete outcome data (attrition bias) 
 Secondary outcomes Low risk Similar attrition in both groups with similar reasons for dropouts. Described as intention‐to‐treat by authors
Selective reporting (reporting bias) Unclear risk Insufficient information to permit judgement
Similar baseline outcome measurements Low risk Similar baseline measurements for hospital admissions and medicine costs
Similar baseline characteristics Low risk Similar baseline characteristics reported
Reliable primary outcome measure Low risk Number of changes in medication
Adequate protection against contamination High risk Randomised by patient therefore contamination possible
Other bias Unclear risk Sample size calculation indicated that 1600 residents were required, however, only 661 residents were recruited

IQR: Interquartile Range

MMSE: Mini‐Mental State Examination

SD: Standard Deviation

I5D: 15 Dimensional Instrument of Health‐related Quality of Life