Allard J, Hebert R, Asselin J, et al. Efficacy of a clinical medication review on the number of potentially inappropriate prescriptions prescribed for community-dwelling elderly people. Canadian Medical Association Journal 2001;164:1291-96.
PubMed link: https://www.ncbi.nlm.nih.gov/pubmed/11341138
Full text article: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC81019/
| Methods | Study design: Cluster randomized controlled study
Number of groups: Two groups |
| Participants | Number of participants: 266 randomized, 196 completed
· Intervention deprescribing group: 136 randomized, 80 completed · Control group: 130 randomized, 116 completed Age: 80.4 ± 4.3 years Sex: 180 female, 86 male Participants with dementia: No Inclusion criteria: · Over 75 years of age · Living in the community · At risk of losing their autonomy · Taking more than three drugs per day Exclusion criteria: · None stated Concomitant medicines: 6.05 ± 1.76 total medicines per person Country: Canada Setting: Community – recruitment mail-out to people registered with the Quebec Health Insurance Board to identify eligible people |
| Interventions | Medicine: Polypharmacy
Withdrawal Schedule: Not described Comparator: Usual care Method to identify targets: Patient-specific intervention Medication review: Recommendation by a team comprising 2 physicians, a pharmacist and a nurse. Recommendation to participant’s treating physician Tool to identify deprescribing targets: list of potentially inappropriate prescriptions developed by the Quebec Committee on Drug Use in the Elderly |
| Outcomes | Total number of potentially inappropriate medicines per person
Total number of medicinesprescribed per person Number of subjects with at least one potentially inappropriate medicine |
| Dates | Dates: Not described
Follow-up duration: One year |
| Funding sources | Department of Family Medicine of the University of Sherbrooke for financial support |
| Notes | Authors stated: “Competinginterests: None declared.” |
Risk of bias table
| Bias | Authors’ judgment | Support for judgment |
| Random sequence generation (selection bias) | Unclear risk | The allocation method was only described as “were randomly assigned to an experimental and a control group” but no detail was given. |
| Allocation concealment (selection bias) | Unclear risk | Method used to conceal allocation sequence was not described. |
| Blinding of participants and personnel (performance bias) | High risk | Not blinded. |
| Blinding of outcome assessment (detection bias) | High risk | Personnel were not blinded at the start, but the outcome was unlikely to be influenced by the lack of blinding at this stage. The nurse calculating the primary measurement (potentially inappropriate medicines) was blinded, and blinded personnel performed the global assessment of change.
The potentially inappropriate medicines themselves were generated from an independentlist. Given the level of inter-rater agreement on studies assessing potentially inappropriate medicines has ranged from approximately a kappa of 0.6 to 0.8, this may introduce bias. |
| Incomplete outcome data (attrition bias) | Low risk | Attrition and exclusions were reported, and reasons were reported. It is unlikely that incomplete data would affect the outcome. It was an intention-to-treat analysis. |
| Selective reporting (reporting bias) | Low risk | The primary and secondary outcomes are reported as pre-specified. The primary outcome is also reported using subsets that were not pre-specified, but this was additional reported and not as a replacement to the pre-specified measurement. |
| Other bias | High risk | The study did not control for multiple prescribers or for reasons why recommendations were not followed. The number of prescribers was acknowledged in the article as a major confounder. There was no description of how the physician to contact was chosen for each patient. This would not matter if there were only one prescriber, but where there are multiple prescribers, there is no way of evaluating whether the primary prescriber or the prescriber of most of the medications was contacted. To truly test this intervention, results from subgroups should have been compared or else the inclusion criteria be modified so that patients all have comparable numbers of prescribers (ideally one). |
Leave a Reply