Performance of model-based vs. permutation tests in the HEALing (Helping to End Addiction Long-term
Cluster randomized trials
Covariate-constrained randomization
Model-based tests
Negative binomial regression
Permutation tests
Journal
Trials
ISSN: 1745-6215
Titre abrégé: Trials
Pays: England
ID NLM: 101263253
Informations de publication
Date de publication:
08 Sep 2022
08 Sep 2022
Historique:
received:
13
01
2022
accepted:
02
09
2022
entrez:
8
9
2022
pubmed:
9
9
2022
medline:
14
9
2022
Statut:
epublish
Résumé
The HEALing (Helping to End Addiction Long-term The primary outcome, the number of opioid overdose deaths, is count data assessed at the community level that will be analyzed using a negative binomial regression model. We conducted a simulation study to evaluate the type I error rates and power for 3 tests: (1) Wald-type t-test with small-sample corrected empirical standard error estimates, (2) Wald-type z-test with model-based standard error estimates, and (3) permutation test with test statistics calculated by the difference in average residuals for the two groups. Our simulation results demonstrated that Wald-type t-tests with small-sample corrected empirical standard error estimates from the negative binomial regression model maintained proper type I error. Wald-type z-tests with model-based standard error estimates were anti-conservative. Permutation tests preserved type I error rates if the constrained space was not too small. For all tests, the power was high to detect the hypothesized 40% reduction in opioid overdose deaths for the intervention vs. comparison group both for the overall HCS and the subgroup analysis of Massachusetts (MA). Based on the results of our simulation study, the Wald-type t-test with small-sample corrected empirical standard error estimates from a negative binomial regression model is a valid and appropriate approach for analyzing cluster-level count data from the HEALing Communities Study. ClinicalTrials.gov http://www. gov ; Identifier: NCT04111939.
Sections du résumé
BACKGROUND
BACKGROUND
The HEALing (Helping to End Addiction Long-term
METHODS
METHODS
The primary outcome, the number of opioid overdose deaths, is count data assessed at the community level that will be analyzed using a negative binomial regression model. We conducted a simulation study to evaluate the type I error rates and power for 3 tests: (1) Wald-type t-test with small-sample corrected empirical standard error estimates, (2) Wald-type z-test with model-based standard error estimates, and (3) permutation test with test statistics calculated by the difference in average residuals for the two groups.
RESULTS
RESULTS
Our simulation results demonstrated that Wald-type t-tests with small-sample corrected empirical standard error estimates from the negative binomial regression model maintained proper type I error. Wald-type z-tests with model-based standard error estimates were anti-conservative. Permutation tests preserved type I error rates if the constrained space was not too small. For all tests, the power was high to detect the hypothesized 40% reduction in opioid overdose deaths for the intervention vs. comparison group both for the overall HCS and the subgroup analysis of Massachusetts (MA).
CONCLUSIONS
CONCLUSIONS
Based on the results of our simulation study, the Wald-type t-test with small-sample corrected empirical standard error estimates from a negative binomial regression model is a valid and appropriate approach for analyzing cluster-level count data from the HEALing Communities Study.
TRIAL REGISTRATION
BACKGROUND
ClinicalTrials.gov http://www.
CLINICALTRIALS
RESULTS
gov ; Identifier: NCT04111939.
Identifiants
pubmed: 36076295
doi: 10.1186/s13063-022-06708-9
pii: 10.1186/s13063-022-06708-9
pmc: PMC9461200
doi:
Banques de données
ClinicalTrials.gov
['NCT04111939']
Types de publication
Journal Article
Randomized Controlled Trial
Langues
eng
Sous-ensembles de citation
IM
Pagination
762Informations de copyright
© 2022. The Author(s).
Références
Biom J. 2017 May;59(3):478-495
pubmed: 28128854
Clin Trials. 2004;1(3):297-305
pubmed: 16279255
Clin Trials. 2022 Apr;19(2):162-171
pubmed: 34991359
Stat Med. 2006 Feb 15;25(3):375-88
pubmed: 16143991
Clin Trials. 2005;2(2):130-40
pubmed: 16279135
Am J Public Health. 2004 Mar;94(3):423-32
pubmed: 14998806
Drug Alcohol Depend. 2020 Dec 1;217:108335
pubmed: 33248391
Biometrics. 2001 Dec;57(4):1198-206
pubmed: 11764261
J Clin Epidemiol. 1999 Jan;52(1):19-26
pubmed: 9973070
Stat Med. 1994 Sep 15;13(17):1715-26
pubmed: 7997705
Biometrics. 2001 Mar;57(1):126-34
pubmed: 11252587
Stat Med. 1993 Feb;12(3-4):329-38
pubmed: 8456215
Stat Med. 2016 May 10;35(10):1565-79
pubmed: 26598212
Drug Alcohol Depend. 2020 Dec 1;217:108327
pubmed: 33091843
Clin Pharmacol Ther. 1974 May;15(5):443-53
pubmed: 4597226
Stat Med. 2017 Oct 30;36(24):3791-3806
pubmed: 28786223
Biometrics. 1979 Jun;35(2):503-12
pubmed: 486683
Drug Alcohol Depend. 2020 Dec 1;217:108329
pubmed: 33075691
Clin Exp Pharmacol Physiol. 1994 Sep;21(9):673-86
pubmed: 7820947
Stat Med. 1996 Jun 15;15(11):1069-92
pubmed: 8804140