SUMMARY Observational studies can be used to evaluate treatment eeectiveness among patients with a broader range of illness severity than typically seen in randomized controlled clinical trials. However, there are several diiculties with observational evaluations including non-equivalent comparison groups, treatment doses and durations that vary widely, and, in longitudinal studies, multiple courses of treatment per subject. A mixed-eeects approach to the propensity adjustment for non-equivalent comparison groups is described that can account for each of these perturbations. The strategy involves two stages. First, characteristics that distinguish among subjects who receive various levels of treatment are examined in a model of propensity for treatment intensity using mixed-eeects ordinal logistic regression. Second, the propensity-stratiÿed eeectiveness of ordered categorical doses is compared in a mixed-eeects grouped time survival model of time until recovery. The model is applied in a longitudinal, observational study of antidepressant eeectiveness. Then a Monte Carlo simulation study indicates that the strategy has acceptable type I error rates and minimal bias in the estimates of treatment eeectiveness. Statistical power exceeds 0.90 for an odds ratio of 1.5 with N = 250 and 500, and is acceptable for an odds ratio of 2.0 with N = 100. Nevertheless, with N = 100, the models that had high intraclass correlation coeecients had greater tendency towards non-convergence. This approach is a useful strategy for observational studies of treatment eeectiveness. It is capable of adjusting for selection bias, incorporating multiple observations per subject, and comparing eeectiveness of ordinal doses.
[1]
Uldis Zariņš.
Copyright
,
2015,
European Neuropsychopharmacology.
[2]
John P. Rice,et al.
A 20-Year Longitudinal Observational Study of Somatic Antidepressant Treatment Effectiveness
,
2003
.
[3]
M B Keller,et al.
A dynamic adaptation of the propensity score adjustment for effectiveness analyses of ordinal doses of treatment
,
2001,
Statistics in medicine.
[4]
F B Hu,et al.
Random-effects regression analysis of correlated grouped-time survival data
,
2000,
Statistical methods in medical research.
[5]
D. Hedeker,et al.
MIXOR: a computer program for mixed-effects ordinal regression analysis.
,
1996,
Computer methods and programs in biomedicine.
[6]
D. Hedeker,et al.
A random-effects ordinal regression model for multilevel analysis.
,
1994,
Biometrics.
[7]
D. Rubin,et al.
Reducing Bias in Observational Studies Using Subclassification on the Propensity Score
,
1984
.
[8]
D. Rubin,et al.
The central role of the propensity score in observational studies for causal effects
,
1983
.
[9]
J. Fleiss,et al.
Statistical methods for rates and proportions
,
1973
.
[10]
W. G. Cochran.
The effectiveness of adjustment by subclassification in removing bias in observational studies.
,
1968,
Biometrics.