Re: Poor description of non-pharmacological interventions: analysis of consecutive sample of randomised trials
Hoffmann and colleagues (1) and Cook and colleagues (2) add to the growing body of evidence that demonstrates inadequate reporting of non-pharmacological interventions. However, abstracts are screened in evidence synthesis before accessing full-texts, are more widely circulated than full-text articles (3) and are generally the most widely available parts of articles (4). We have gathered systematic evidence suggesting that the abstracts of non-pharmacological interventions lack important details that are available in abstracts of pharmacological interventions. Specifically, the active ingredients (intervention components responsible for effectiveness) of non-pharmacological interventions are often missing from abstracts (5). As well as compromising faithful replication of studies, inadequate specification has important consequences for the screening phases of a systematic review. Studies with inadequately specified interventions may be inappropriately rejected, therefore compromising the validity of the review.
The reporting of comparator interventions is typically worse than the reporting of the trialled intervention and the problem of labelling as usual or standard care with no further explanation deserves further emphasis. The CONSORT guidance for abstracts recommends that the essential features of the procedures used in the control group be clearly identified in the abstract (6). Our study found that 19% of the included non-pharmacological comparator interventions were described in the abstracts as usual or standard care with no further details, compared to 2% of the pharmacological comparator interventions (McCleary and colleagues, unpublished data). As Cook and colleagues (2) allude to, in this situation readers cannot compare the intensity of usual care with the active intervention and with what is usually provided in their own setting (6). Further, de Bruin and colleagues (7) have shown that the active content of control group care predicts patient outcomes and can explain effects of trialled interventions. Freedland and colleagues (8) identified the range of control group methods used, but, based on their review of studies involving a usual care group, Smelt and colleagues concluded that there is no consensus on the use of the term usual care: in some studies, care providers involved in the trial tailor the intervention to the individual, while in others, one particular method is used to treat all control participants (9). Additionally, authors rarely describe the information and instructions used within usual care groups (9) but this data can be made available on request (7). Smelt and colleagues recommended that another item be added to the CONSORT statement describing the information that should be reported on the treatment of participants in these groups. This might be particularly beneficial for non-pharmacological interventions, where active control treatments and usual care treatments are more often mentioned (10). A similar update to the CONSORT guidance for reporting abstracts would also be welcomed.
Two things which might improve reporting of non-pharmacological interventions are increasing the word limit for abstracts and developing a shared language for reporting non-pharmacological interventions. The quality of reporting has been found to be higher with greater allowed abstract word count for structured abstracts (11). The CONSORT guidelines for abstracts state that 250 to 300 words are sufficient to report all recommended items (6). It might be beneficial for journals with limits lower than this to consider increasing their abstract word limits in line with these guidelines to facilitate higher-quality abstract reporting, particularly for non-pharmacological interventions which likely require more words to specify than pharmacological interventions.
Although a shared language is used to describe pharmacological interventions, no such language exists for non-pharmacological interventions. As a result, different terms can be used to describe the same intervention, while the same terms can be used to describe different interventions. The extensive work that has been conducted so far into developing a taxonomy for describing the techniques used in behaviour change interventions should serve to improve their reporting (12). Further work of this nature across the spectrum of non-pharmacological intervention types is essential for improving intervention specification across the board.
(1) Hoffmann TC, Erueti C, Glasziou P. Poor description of non-pharmacological interventions: analysis of consecutive sample of randomised trials. BMJ 2013;347.
(2) Cook A, Douet L, Boutron I. Descriptions of non-pharmacological interventions in clinical trials. BMJ 2013;347.
(3) Berwanger O, Ribeiro RA, Finkelsztejn A, Watanabe M, Suzumura EA, Duncan BB, et al. The quality of reporting of trial abstracts is suboptimal: survey of major general medical journals. J Clin Epidemiol 2009;62(4):387-392.
(4) von Elm, E. Writing the abstract: completeness and accuracy matter. Eur J Anaesthesiol 2011;28(7):483-484.
(5) McCleary N, Duncan E, Stewart F, Francis J. Active ingredients are reported more often for pharmacologic than non-pharmacologic interventions: an illustrative review of reporting practices in titles and abstracts. Trials 2013;14(1):146.
(6) Hopewell S, Clarke M, Moher D, Wager E, Middleton P, Altman DG, et al. CONSORT for reporting randomized controlled trials in journal and conference abstracts: explanation and elaboration. PLoS Med 2008;5(1):e20.
(7) de Bruin M, Viechtbauer W, Schaalma HP, Kok G, Abraham C, Hospers HJ. Standard care impact on effects of highly active antiretroviral therapy adherence interventions: A meta-analysis of randomized controlled trials. Arch Intern Med 2010;170(3):240-250.
(8) Freedland KE, Mohr DC, Davidson KW, Schwartz JE. Usual and unusual care: existing practice control groups in randomized controlled trials of behavioral interventions. Psychosom Med 2011;73(4):323-335.
(9) Smelt AF, van der Weele GM, Blom JW, Gussekloo J, Assendelft WJ. How usual is usual care in pragmatic intervention studies in primary care? An overview of recent trials. Brit J Gen Pract 2010;60(576):e305-18.
(10) Boutron I, Tubach F, Giraudeau B, Ravaud P. Methodological Differences in Clinical Trials Evaluating Nonpharmacological and Pharmacological Treatments of Hip and Knee Osteoarthritis. JAMA 2003;290(8):1062-1070.
(11) Burns KEA, Adhikari NKJ, Kho M, Meade MO, Patel RV, Sinuff T, et al. Abstract reporting in randomized clinical trials of acute lung injury: An audit and assessment of a quality of reporting score. Crit Care Med 2005;33(9):1937-1945.
(12) Michie S, Richardson M, Johnston M, Abraham C, Francis J, Hardeman W, et al. The Behavior Change Technique Taxonomy (v1) of 93 Hierarchically Clustered Techniques: Building an International Consensus for the Reporting of Behavior Change Interventions. Annals of Behavioral Medicine 2013;46(1):81-95.
Competing interests: No competing interests