Muir AJ, Sanders LL, Wilkinson WE, et al. Reducing medication regimen complexity – A controlled trial. Journal of General Internal Medicine 2001;16:77-82
PubMed link: https://www.ncbi.nlm.nih.gov/pubmed/11251757
Full text article: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1495168/
| Methods | Study design: Non-randomized experimental trial with a parallel-design
Number of groups: Two groups |
| Participants | Number of participants: 836 enrolled
· Intervention deprescribing group: 368 enrolled · Control group: 468 enrolled Age: 65.2 years Sex: 10 female 826 male Participants with dementia: No Inclusion criteria: · Patients at the participating Veterans Affairs medical center Exclusion criteria: · Not described Concomitant medicines, median: 9 medicines per person Country: United States of America Setting: Hospital – Veterans Affairs medical center |
| Interventions | Medicine: Polypharmacy
Withdrawal schedule: Not specified Comparator: Usual care Method to identify targets: Patient-specific intervention: medication review: Medication reconciliation and medicine list undertaken by investigators. The resident in the General Medicine Ward Team was instructed to review medication regimens. General medicine ward teams led by second- and third-year medicine residents. Teams generally consisted of a resident, an intern, 1 or 2 students, and an attending physician. Tool to identify deprescribing targets: Doctor’s choice (no list, criteria, or tool used) |
| Outcomes | Mean change (SD) in medications and doses
Median number of Admission and discharge medications by drug class Median proportions of patients taking individual medications |
| Dates | Dates: July 1997 and May 1998
Follow-up duration: Five to seven weeks |
| Funding sources | Geriatric Research Education and Clinical Center of the Durham VAMC, and by the Duke Claude D. Pepper Older Americans Independence Center, no 1P60AG11268-02 |
| Notes |
Risk of bias table
| Bias | Authors’ judgment | Support for judgment |
| Random sequence generation (selection bias) | High risk | Not a randomized study. |
| Allocation concealment (selection bias) | High risk | There was an attempt to do this, but it is unclear how successful it was.
“Attending physicians were general internists and subspecialists, each of whom supervised two ward teams. Attending physicians rotated every 4 weeks and did not cross collection periods for the intervention or control groups.” Those physicians in the intervention groups then received the grids. |
| Blinding of participants and personnel (performance bias) | 5 out of 5 | This was not a blind study. |
| Blinding of outcome assessment (detection bias) | 5 out of 5 | As above |
| Incomplete outcome data (attrition bias) | Unclear risk | The physician was used as the unit of analysis but considered the medicines they prescribed for individual patients. Could not reconcile the design of the study (intervening with physicians) and the reported results as numbers of patients |
| Selective reporting (reporting bias) | Low risk | The researchers had a pre-specified protocol that was approved by a Human Research Ethics Committee. It is unclear to what degree they adhered to a pre-specified analysis plan. The paper stated only one intended objective but did not clearly spell out how they intended to measure this. However, the simple outcome was reduction in complexity,and this was reported.
“The objective of the study was to determine if the medication grid significantly reduced the number of medications and doses prescribed by physicians in an intervention group compared with the control group.” |
| Confounding (non-randomized) | 1 out of 5 | The method was sound. The unit of analysis was the team of physicians, so the intervention was applied to the unit.
It is unclear to what degree confounding between the control and intervention group of physicians occurred, and the pharmacists attended to both groups. However, a large list of confounders was considered. |
| Other bias | Low risk | |
| Newcastle-Ottawa scale | ||
| Selection bias | Representativeness of the exposed cohort | The intervention group was a selected group of users. |
| Selection of the non-exposed cohort | Concurrent control group was drawn from the same community as the exposed cohort. | |
| Ascertainment of exposure | Ascertainment of exposure was through secure record.
|
|
| Demonstration that outcome of interest was not present at start of study | It was demonstrated that the outcome of interest was not present at start of study. | |
| Comparability bias | Comparability of cohorts on the basis of the design or analysis | The study was appropriately controlled for the deprescribing intervention. Note: the unit of analysis was the team of physicians,so the intervention was applied to the unit. |
| Outcome bias | Assessment of outcome | Assessment of outcome was by record linkage. |
| Was follow-up long enough for outcomes to occur | Follow-up was probably long enough for outcomes to occur. (It assessed medicine use at discharge, so use at discharge was reasonable.) | |
| Adequacy of follow-up of cohorts | There was no statement about follow-up of the cohorts. | |
Leave a Reply