Reporting of lost to follow-up and treatment discontinuation in device and pharmacotherapy trials in chronic heart failure: a systematic review

Campbell, R. T., Willox, G. P., Jhund, P. S. , Hawkins, N. M., Huang, F., Petrie, M. C. and McMurray, J. J. V. (2016) Reporting of lost to follow-up and treatment discontinuation in device and pharmacotherapy trials in chronic heart failure: a systematic review. Circulation: Heart Failure, 9(5), e002842. (doi: 10.1161/CIRCHEARTFAILURE.115.002842) (PMID:27162229)

Abstract

Background—Premature treatment discontinuation and loss to follow-up (LTFU) with unknown outcomes leave uncertainty about the true efficacy and safety of a treatment and a lack of confidence in the results of any trial. We reviewed the extent of (and trends over time in) reporting LTFU and treatment discontinuation in large studies in chronic heart failure published since 1990. Methods and Results—Online databases were systematically reviewed to identify randomized controlled clinical trials (RCTs) in chronic heart failure with >400 participants and utilizing all-cause mortality as a component of the primary or secondary end point. Assessments were made of documentation of treatment discontinuation, LTFU, inclusion of and completeness of a Consolidated Standards Of Reporting Trials (CONSORT) diagram, and whether LTFU was differentiated from withdrawal of consent. Sixty-eight trials were identified, with >154 000 participants. Reasons for treatment discontinuation in pharmacotherapy trials were infrequently reported (35%), particularly in a CONSORT diagram (20%). Eighty-three percent of trials reported LTFU, although only 34% of these differentiated LTFU for vital status from withdrawal of consent. Use of a CONSORT diagram increased over time, although reporting of LTFU in the CONSORT diagram remained low overall at 35%. Conclusions—Participant flow through RCTs in chronic heart failure has not been uniformly reported, and the use of a complete CONSORT diagram has been low, although it seems to be improving. All study participants should be accounted for within a CONSORT diagram in any RCT to enable the practicing cardiologist to interpret how the results should influence his/her clinical practice.

• Item Type: Articles
• Status: Published
• Refereed: Yes
• Glasgow Author(s) Enlighten ID: Willox, Mr Gage and Jhund, Professor Pardeep and Petrie, Professor Mark and Campbell, Dr Ross and McMurray, Professor John
• Authors: Campbell, R. T., Willox, G. P., Jhund, P. S., Hawkins, N. M., Huang, F., Petrie, M. C., and McMurray, J. J. V.
• College/School: College of Medical Veterinary and Life Sciences > School of Cardiovascular & Metabolic Health
• Journal Name: Circulation: Heart Failure
• Publisher: American Heart Association
• ISSN: 1941-3289
• ISSN (Online): 1941-3297