Thapa PB, Meador KG, Gideon P, et al. Effects of antipsychotic withdrawal in elderly nursing home residents. Journal of the American Geriatrics Society 1994;42:280-86
PubMed link: https://www.ncbi.nlm.nih.gov/pubmed/7907098
Methods | Study design: Non-randomized controlled study
Number of groups: Two group |
Participants | Number of participants: 271 participants
· Active deprescribing group: 64 enrolled · Control group: 207 enrolled Age: 82.6 ± 7.7 years Sex: 199 female, 72 male Participants with dementia: Yes, baseline MMSE 11.73 Inclusion criteria · Antipsychotic users · Aged 65 years and over · Residents at baseline and follow-up Exclusion criteria: · Not described Concomitant psychotropic medicines: 18.7% used benzodiazepines, cyclic antidepressants, or other anxiolytics/sedatives Country: United States of America Setting: Residential aged care facilities – 12 residential aged care facilities |
Interventions | Medicine: Antipsychotics, baseline dose (chlorpromazine equivalence): 89.4mg ± 193.2
Withdrawal schedule: Not described Comparator: Continued treatment compared to withdrawn treatment |
Outcomes | Psychotropic medicine use
Behavioral problems assessed using the Nursing Home Behavior Problem Scale Psychiatric symptoms assessed using the Brief Psychiatric Rating Scale Function Activities of daily living assessed using Lawton’s Physical Self-Maintenance Scale Cognition assessed using MMSE Geriatric Depression Scale Abnormal Involuntary Movement Scale |
Dates | Dates: July 1991 to December 1992
Follow-up duration: Six months |
Funding sources | Grant 89308-G from the John A Hartford Foundation, New York, New York and by a cooperative agreement (#FD-U-000073-08) with the Food and Drug Administration, Bethesda, Maryland. Dr. Thapa received support from a CIBA-GEIGY fellowship in Pharmacoepidemiology award. |
Notes |
Risk of bias table
Bias | Authors’ judgment | Support for judgment |
Random sequence generation (selection bias) | High risk | Not a randomized study design. |
Allocation concealment (selection bias) | 5 out of 5 | Open study. |
Blinding of participants and personnel (performance bias) | 5 out of 5 | Open study. |
Blinding of outcome assessment (detection bias) | 5 out of 5 | Open study. |
Incomplete outcome data (attrition bias) | 3 out of 5 | The researchers took this in to account and both accounted for the reasons for missing outcome data, ad imputed missing data using appropriate methods.
“Because many residents could not complete all of the planned assessments, we minimized loss of information and potential bias from exclusions by conducting separate analyses for each of the different sets of measurements. We abstracted demographic and drug use data and sought nursing home staff ratings of behavior problem frequency for all baseline antipsychotic users, assessed psychiatric symptoms in assenting subjects who were not blind, deaf, or critically ill (defined as someone with a terminal disease or recovering from a major illness), and obtained several standard measures of function for residents who could follow simple directions and provide informed consent.” The group that the drop-outs were allocated to was not reported in the paper. Given the association between antipsychotic use and mortality, more detail would have been informative: “death (n = 22), hospitalization (n = 20), and transfer to other nursing homes ( n = 7) were the primary reasons for the attrition.” |
Selective reporting (reporting bias) | 5 out of 5 | Mean change reported rather than absolute numbers. Study objectives not clearly described, so not possible to know if the study is free of selective outcome reporting.
All the outcomes were reported but not in sufficient detail that could be included in a meta-analysis: Nursing Home Behavior Problem Scale, Brief Psychiatric Rating Scale, or Function scales. |
Confounding (non-randomized) | 5 out of 5 | All antipsychotics grouped together i.e. low potency and high potency. Hard to judge what affect the education program would have.
Educational intervention facility-wide rather than simply deprescribing as an intervention. This holistic approach makes it. |
Other bias | Low risk | |
Newcastle-Ottawa scale | ||
Selection bias | Representativeness of the exposed cohort | Somewhat representative of the average older adult inpatients with dementia. |
Selection of the non-exposed cohort | Selection of the non-exposed cohort: drawn from a different source.
|
|
Ascertainment of exposure | Ascertainment of exposure was by secure record. | |
Demonstration that outcome of interest was not present at start of study | Yes. | |
Comparability bias | Comparability of cohorts on the basis of the design or analysis | The study controlled for cessation of antipsychotics. |
Outcome bias | Assessment of outcome | Assessment of outcome: record linkage. |
Was follow-up long enough for outcomes to occur | Follow-up was long enough for outcomes to occur. | |
Adequacy of follow-up of cohorts | Follow-up rate:
Death (n =22) Hospitalization (n = 20) Transfer to other nursing homes ( n = 7) Other reasons not stated. Group withdrawals were from not stated. |
Leave a Reply