Three week compared to seven week run-in period length and the assessment of pre-randomization adherence: A study within a trial

Collister D 1 , Mbuagbaw L 2 , Guyatt G 3 , Devereaux PJ 4 , Tennankore KK 5 , Reis G 6 Show all authors , Sola L 7 , Xavier D 8 , Jha V 9 , Gallagher M 10 , Dans AL 11 , Liu W 12 , de Zoysa J 13 , Félix C 14 , Mark PB 15 , Li Z 16 , Tyrwhitt J 17 , Wilkinson J 18 , Sheridan P 19 , Yuan F 20 , Walsh M 21 , ACHIEVE investigators

Affiliations 

  • 1 Population Health Research Institute, Hamilton, Canada; Department of Health Research Methods, Evidence and Impact, McMaster University, Canada; Department of Medicine, McMaster University, Canada. Electronic address: [email protected]
  • 2 Department of Health Research Methods, Evidence and Impact, McMaster University, Canada. Electronic address: [email protected]
  • 3 Department of Health Research Methods, Evidence and Impact, McMaster University, Canada; Department of Medicine, McMaster University, Canada. Electronic address: [email protected]
  • 4 Population Health Research Institute, Hamilton, Canada; Department of Health Research Methods, Evidence and Impact, McMaster University, Canada; Department of Medicine, McMaster University, Canada. Electronic address: [email protected]
  • 5 Division of Nephrology, Department of Medicine, Dalhousie University, Canada. Electronic address: [email protected]
  • 6 Departamento de Medicina, Pontifícia Universidade Católica de Minas Gerais, Brazil. Electronic address: [email protected]
  • 7 CASMU-IAMPP, Uruguay. Electronic address: [email protected]
  • 8 Department of Pharmacology and Division of Clinical Research, St. John's Medical College, India. Electronic address: [email protected]
  • 9 The George Institute for Global Health, New Delhi, India. Electronic address: [email protected]
  • 10 The George Institute for Global Health, Sydney, Australia. Electronic address: [email protected]
  • 11 University of the Philippines, College of Medicine, Philippines. Electronic address: [email protected]
  • 12 Hospital Sultanah Aminah, Malaysia
  • 13 Department of Medicine, The University of Auckland, Auckland, New Zealand. Electronic address: [email protected]
  • 14 Facultad de Ciencias Medicas Eugenio Espejo, Universidad Tecnológica Equinoccial, Ecuador
  • 15 Institute of Cardiovascular and Medical Sciences, United Kingdom. Electronic address: [email protected]
  • 16 Department of Nephrology, Peking University People's Hospital, Beijing, China
  • 17 Population Health Research Institute, Hamilton, Canada. Electronic address: [email protected]
  • 18 Population Health Research Institute, Hamilton, Canada. Electronic address: [email protected]
  • 19 Population Health Research Institute, Hamilton, Canada. Electronic address: [email protected]
  • 20 Population Health Research Institute, Hamilton, Canada. Electronic address: [email protected]
  • 21 Population Health Research Institute, Hamilton, Canada; Department of Health Research Methods, Evidence and Impact, McMaster University, Canada; Department of Medicine, McMaster University, Canada
Contemp Clin Trials, 2021 08;107:106466.
PMID: 34098039 DOI: 10.1016/j.cct.2021.106466

Abstract

BACKGROUND/AIMS: To examine how measuring adherence at 3 weeks by self-report and pill counts compares to measurements at 7 weeks in a pre-randomization run-in period.

METHODS: Study within a trial of an international parallel group randomized controlled trial (RCT) that compares spironolactone to placebo. Adults receiving dialysis enter an 8-week active run-in period with spironolactone. Adherence was assessed by both self-report and pill counts in a subgroup of participants at both 3 weeks and 7 weeks.

RESULTS: 332 participants entered the run-in period of which 166 had complete data. By self-report, 146/166 (94.0%) and 153/166 (92.2%) had at least 80% adherence at 3 and 7 weeks respectively (kappa = 0.27 (95% C.I. 0.16 to 0.38). By pill counts, the mean (SD) adherence was 96.5% (16.1%) and 92.4% (18.2%) at 3 and 7 weeks respectively (r = 0.32) with a mean (SD) difference of 3.1% (17.8%) and a 95% limit of agreement from -31.7% to +37.9%. The proportion of adherent participants by self-report and pill counts at 3 weeks agreed in 87.4% of participants (McNemar's p-value 0.58, kappa 0.11, p = 0.02) and at 7 weeks agreed in 92.2% (McNemar's p-value 0.82, kappa 0.47, p 

* Title and MeSH Headings from MEDLINE®/PubMed®, a database of the U.S. National Library of Medicine.

MeSH terms
Similar publications