Reporting on quality of life in randomised controlled trials: bibliographic study
BMJ 1998; 317 doi: https://doi.org/10.1136/bmj.317.7167.1191 (Published 31 October 1998) Cite this as: BMJ 1998;317:1191All rapid responses
Rapid responses are electronic comments to the editor. They enable our users to debate issues raised in articles published on bmj.com. A rapid response is first posted online. If you need the URL (web address) of an individual response, simply click on the response headline and copy the URL from the browser window. A proportion of responses will, after editing, be published online and in the print journal as letters, which are indexed in PubMed. Rapid responses are not indexed in PubMed and they are not journal articles. The BMJ reserves the right to remove responses which are being wilfully misrepresented as published articles or when it is brought to our attention that a response spreads misinformation.
From March 2022, the word limit for rapid responses will be 600 words not including references and author details. We will no longer post responses that exceed this limit.
The word limit for letters selected from posted responses remains 300 words.
Editor. Sanders et al reported that the presentation of results of
quality of life (QoL) data in clinical trials was often flawed [1]. We are
creating a database of QoL questionnaires used in clinical trials (CT) and
we assess the quality of the trials with a checklist that we have
elaborated [2].
The major biases we encounter through our reading are the following :
trial not comparative [3] ; no justification of the number of patients to
include, the number can be too small to achieve enough power to detect a
difference between two treatments, or the number can be too large leading
to a statistically significant but non clinically pertinent difference [4]
; the questionnaire of QoL is not validated and its responsiveness has not
been tested ; no description of the follow-up of patients during the study
is provided ; no description of withdrawals and handling of missing data
is available ; analysis of QoL data is performed on a per protocol basis
instead of on an intent to treat basis, e.g. less than 530 patients are
analysed instead of the 812 patients randomized in a trial comparing
ranitidine vs. placebo in gastroesophageal reflux [4] ; flawed
presentation of QoL results, e.g. no QoL value is presented but only
graphs and standard deviation of scores in the different domains of QoL
are missing. Ideally, the confidence interval of the difference between
treatment groups and/or the size-effect should also be presented ; the
level of significance is not adapted to the number of statistical
comparisons leading to a raise of a error ; the clinical relevance of the
results of QoL is not discussed by authors, e.g. when only 2 or 3 among 8
or 9 domains of the QoL questionnaire improve significantly in one group
vs. the other group [4, 5] or when the differences in QoL scores between
groups seem rather slight [4]. For example, in using a specific
questionnaire (CLAU-S) as a primary end-point, only 2 of the 9 domains
improved significantly with naftidrofuryl vs. placebo at 6 months. These 2
domains were pain and daily life, which are the minimal goals to reach for
a treatment claiming to alleviate symptoms of arteriopathy [5].
Like any assessment criterion, QoL requires the same rigorous methodology,
follow-up of patients and statistical analysis. Thus reporting on QoL in
clinical trials should be significantly improved and should follow the
CONSORT guidelines like any clinical trial.
References
1. Sanders C, Egger M, Donovan J, Tallon D, Frankel S. Reporting on
quality of life in randomised controlled trials: bibliographic study. BMJ
1998;317:1191-4.
2. Chassany O, Bergmann JF, Chwalow J, Laville M, Caulin C. Database of
quality of life instruments applied to clinical trials. Submitted for the
Sixth Annual Symposium of contibuted papers : quality of life evaluation.
Drug Information Association, March 28-30, 1999, South Carolina, USA.
3. Lukacs B, McCarthy C, Grange JC, QOL BPH Study Group in General
Practice. Long-term quality of life in patients with benign prostatic
hypertrophy: preliminary results of a cohort survey of 7093 patients
treated with an alpha-1-adrenergic blocker, alfuzosin. eur Urol
1993;24(Suppl.1):34-40.
4. Rush DR, Stelmach WJ, Young TL et al. Clinical effectiveness and
quality of life with ranitidine vs placebo in gastroesophageal reflux
disease patients : a clinical experience network (CEN) study. J Fam Pract
1995;41:126-36.
5. Liard F, Benichou AC, Gamand S, Lehert P. The effects of naftidrofuryl
on quality of life. Dis Manage Health Outcomes 1997;2(Suppl.1):71-8.
Olivier Chassany1, Jean François Bergmann2, Charles Caulin2.
Service de Médecine Interne, hôpital Lariboisière, 75010 Paris, France.
1 Senior lecturer in Therapeutics, 2 Professor of Therapeutics
Dr Olivier Chassany
Service de Médecine Interne A
Hôpital Lariboisière
2 rue Ambroise Paré
75475 Paris Cedex 10 - France
Tel : (33) 1 49 95 81 27
Fax : (33) 1 49 95 84 46
e-mail : olivier.chassany@lrb.ap-hop-paris.fr
Conflict of interest : none
Word count : 405
References : 5
Competing interests: No competing interests
Editor. Sanders et al reported that the presentation of results of
quality of life (QoL) data in clinical trials was often flawed [1]. We are
creating a database of QoL questionnaires used in clinical trials (CT) and
we assess the quality of the trials with a checklist that we have
elaborated [2].
The major biases we encounter through our reading are the following :
trial not comparative [3] ; no justification of the number of patients to
include, the number can be too small to achieve enough power to detect a
difference between two treatments, or the number can be too large leading
to a statistically significant but non clinically pertinent difference [4]
; the questionnaire of QoL is not validated and its responsiveness has not
been tested ; no description of the follow-up of patients during the study
is provided ; no description of withdrawals and handling of missing data
is available ; analysis of QoL data is performed on a per protocol basis
instead of on an intent to treat basis, e.g. less than 530 patients are
analysed instead of the 812 patients randomized in a trial comparing
ranitidine vs. placebo in gastroesophageal reflux [4] ; flawed
presentation of QoL results, e.g. no QoL value is presented but only
graphs and standard deviation of scores in the different domains of QoL
are missing. Ideally, the confidence interval of the difference between
treatment groups and/or the size-effect should also be presented ; the
level of significance is not adapted to the number of statistical
comparisons leading to a raise of  error ; .the clinical
relevance of the results of QoL, e.g. when only 2 or 3 among 8 or 9
domains of the QoL questionnaire improve significantly in one group vs.
the other group [4, 5] or when the differences in QoL scores between
groups seem rather slight [4]. For example, in using a specific
questionnaire (CLAU-S) as a primary end-point, only 2 of the 9 domains
improved significantly with naftidrofuryl vs. placebo at 6 months. These 2
domains were pain and daily life, which are the minimal goals to reach for
a treatment claiming to alleviate symptoms of arteriopathy [5].
Like any assessment criterion, QoL requires the same rigorous methodology,
follow-up of patients and statistical analysis. Thus reporting on QoL in
clinical trials should be significantly improved and should follow the
CONSORT guidelines like any clinical trial.
References
1. Sanders C, Egger M, Donovan J, Tallon D, Frankel S. Reporting on
quality of life in randomised controlled trials: bibliographic study. BMJ
1998;317:1191-4.
2. Chassany O, Bergmann JF, Chwalow J, Laville M, Caulin C. Database of
quality of life instruments applied to clinical trials. Submitted for the
Sixth Annual Symposium of contibuted papers : quality of life evaluation.
Drug Information Association, March 28-30, 1999, South Carolina, USA.
3. Lukacs B, McCarthy C, Grange JC, QOL BPH Study Group in General
Practice. Long-term quality of life in patients with benign prostatic
hypertrophy: preliminary results of a cohort survey of 7093 patients
treated with an alpha-1-adrenergic blocker, alfuzosin. eur Urol
1993;24(Suppl.1):34-40.
4. Rush DR, Stelmach WJ, Young TL et al. Clinical effectiveness and
quality of life with ranitidine vs placebo in gastroesophageal reflux
disease patients : a clinical experience network (CEN) study. J Fam Pract
1995;41:126-36.
5. Liard F, Benichou AC, Gamand S, Lehert P. The effects of naftidrofuryl
on quality of life. Dis Manage Health Outcomes 1997;2(Suppl.1):71-8.
Competing interests: No competing interests
Quality of life in clinical trials
EDITOR -
Sanders et al's excellent bibliographic study on the frequency and detail
of quality of life data in Cochrane-listed randomised controlled trials is
both timely and disturbing. Despite increasing emphasis on patient-centred
outcomes in all aspects of clinical practice and research, less than 5% of
trials reported on quality of life, and even fewer comprehensively
reported quality of data using well-validated, familiar instruments.1
Deyo and Patrick discussed methodological, attitudinal and conceptual
barriers to the use of quality of life assessments in research in 1989.2
They noted the paucity of information regarding the responsiveness,
reliability, validity and psychometric characteristics of most
instruments. In the 1980s, many authors, including Deyo2 and Feldstein,3
noted the problem of a confusing array of instruments, including scales
with the same purpose. For example, Feldstein noted 43 activities of daily
living scales in 1986.3
Deyo and Patrick suggested increased reporting of studies comparing
different quality of life instruments in the same population, and
comparing the use of generic instruments in different diagnostic groups.
The authors also advocated the testing of any newly developed or ad hoc
instruments against well-established scales and the development of a
"quality of life research laboratory" to aid researchers with the analysis
and standardisation of quality of life data.2
In light of the insights provided by Sanders et al's review,1 all
trialists need to revisit the suggestions above. If this information was
available, the CONSORT guidelines could be expanded to include
recommendations of which tested, well-validated quality of life
instruments should be selected for use in different situations. This would
aid researchers' choice and improve standardisation and generalisability.
Susan P Wright
Cardiovascular Research Fellow
Department of Medicine
University of Auckland, Auckland, New Zealand
sp.wright@auckland.ac.nz
1. Sanders C, Egger M, Donovan J, Tallon D, Frankel S. Reporting on
quality of life in randomised controlled trials : bibliographic study. BMJ
1998;317:1191-4.
2. Deyo R, Patrick D. Barriers to the use of health status measures
in clinical investigation, patient care and policy research. Medical Care
1989;27(3):S254-S268.
3. Feinstein A, Josephy B, Wells C. Scientific and clinical problems
in indexes of clinical disability. Ann Intern Med 1986;105:413.
Competing interests: No competing interests