An evaluation of reporting of consent declines in three high impact factor journals

Abstract

Background

Informed consent confers upon participants the right to decline or accept participation in a study in equal measure. Consent declines can be used as a potential metric to assess the adequacy of the informed consent process. Limited literature is available on the reporting of consent declines in Clinical Research. We evaluated randomized controlled trials published in three high-impact factor journals for consent declines (four-year period) to assess the extent of exertion of autonomy by research participants.

Methods

CONSORT flow charts in RCTs published in New England Journal of Medicine (NEJM), The Lancet and Journal of the American Medical Association (JAMA) between 2012 and 2015 were evaluated. The total and per journal proportion of RCTs reporting consent declines was calculated. Percentage of consent declines was also compared based on number of centers (single centric vs. multicentric) and type of setting (developed vs. developing). Between-groups analysis was done using Chi Square, Chi Square for trend and a Crude odds ratio [95% C.I] (cOR) were calculated.

Results

Only 3.7% (1103/29,825) of the published articles were RCTs. Of the 1103 RCTs, only 532 (48.2%) reported consent declines. JAMA had the highest RCTs reporting consent declines (171/267.64%), the Lancet [129,150/781,924, 16.5%] had the most participants declining consent. Difference was significant using Chi square for linear trend [p < 0.00001]. Consent decline rate was higher in developed countries relative to developing countries [cOR(95% C.I): 2.27(2.23,2.32), p < 0.000001] and in single centric versus multicentric studies [cOR (95% C.I): 3.63(3.56,3.71), p < 0.0000001].

Conclusion

Both authors and journal editors need to ensure better reporting of consent declines.

Keywords

Autonomy consent decline informed consent research ethics

Background

The principle of autonomy described in the Belmont report¹ confers upon every participant the right to accept or decline participation in any research study. The informed consent process is a metric of autonomy and researchers must ensure that a potential participant’s autonomy is protected when he/she obtainsinformed consent. Informed consent is the autonomous authorization of an intervention by an individual and encompasses three fundamental components: disclosure of all relevant information, understanding by the potential participant of the information that is presented, and a voluntary decision based on this information.² When a potential participant declines consent, the same principle of autonomy is exercised and therefore the “numbers of consent declines” is a good metric to evaluate assertion of autonomy as much as acceptance to participate does.

Randomized controlled trials (RCTs) are considered the gold standard among clinical study designs and have the potential to change clinical practice³ and feature high on the evidence pyramid, just below systematic reviews and meta-analyses.⁴ The New England Journal of Medicine (NEJM), Journal of the American Medical Association (JAMA) and The Lancet are high impact factor general medical journals that address broad specialities and have a wide readership base. The papers published in these journals also address genuine clinical equipoise.

The Consolidated Standards of Reporting Trials (CONSORTs) statement [today]comprise a 32-item checklist and a flow diagram.⁵ CONSORT first evolved in 1993 when a multitude of experts including medical journal editors, clinical trialists, epidemiologists, and methodologists – met in Ottawa, Canada to discuss ways of improving the reporting of RCTs. This statement has been endorsed by World Association of Medical Editors (WAME), the International Committee of Medical Journal Editors (ICMJE) and more than 600 journals, which also include the three journals mentioned earlier.^6,7

The CONSORT (2010) guidelines recommend reporting of the details of eligible participants not recruited in a clinical trial which includes consent declines.⁸ There is very limited data on consent declines reported in the published RCTs. Hence, the present study was carried out with the objective of assessing the reporting rate of consent declines in RCTs published in the Lancet, NEJM (New England Journal of Medicine), and JAMA (Journal of American Medical Association)

Methods

Ethics

The study was initiated following Institutional Ethics Committee approval and consent waiver was obtained.

Study design

A retrospective audit of published RCTs in three journals.

Selection of journals

The journals selected were the New England Journal of Medicine (NEJM; IF 55.87), The Lancet (IF 39.2) and the Journal of the American Medical Association (JAMA, IF 35.28) and RCTs published between the years 2012 and 2015 were evaluated.

Selection criteria

RCTs published during 2012 and 2015 in the three journals were searched using PubMed. Inclusions were parallel design RCTs, whilecluster randomized trials and follow-up clinical trials including post hoc analyses were excluded.

Search strategy

The articles were searched using the following search terms “Randomized/Randomised/Random* Controlled Trials AND name of the journal [e.g.: NEJM][jour] AND date of publication [dp].” The CONSORT flowchart, texts and supplements of all the three journals between the years 2012 and 2015 were assessed. The journals’ archives for the four years were also searched in order to verify the search results.

Outcome measures

We assessed the overall rate of consent declines and reporting of consent declines in each journal. We compared consent declines within each of the following subgroups a. Type of journal, b. Type of country for the conduct of the study (developed, developing and mixed). The type of country was defined based on the annex developed by the development policy and analysis division (DPAD) of the department of economic and social affairs of the United Nations Secretariat⁹ and c. Number of centers involved in the conduct of the study (single centric vs. multicentric).

Statistical analysis plan

Demographic data of the published RCTs and consent declines are presented using descriptive statistics.

The comparison of consent declines within each of the subcategories was done using Chi square test and Chi square test for trend. The strength of association was measured using crude odds ratio (cOR) and 95% confidence intervals. All analysis was done using Epi Info 7.0 and GraphPad InStat 3.0 at 5% significance.

Results

Demographics

A total of n=29,825 articles were published in the three journals during the study period. Of these, 1103 (3.7%) were RCTs. The highest number of articles were published by the Lancet followed by NEJM and JAMA. However, the NEJM had the highest number of RCTs [n = 468], followed by the Lancet [n = 368] and JAMA [n = 267]. Developed countries had maximum number of RCTs (n=782,70.89%) ; 115 were single center studies, while 988 were multi-centric (Table 1).

Table 1.

Demographics of the published literature.

Journal	Articles published	RCTs published	Percentage of RCTs published
NEJM	6040	468	7.74
The Lancet	18,853	368	1.95
JAMA	4932	267	5.41
Articles as per number of centers
Type of center		RCTs published	Percentage of RCTs published
Unicentric		115	10.4
Multicentric		988	89.5
Articles as per region
Type of region		RCTs published	Percentage of RCTs published
Developed		782	70.89
Developing		85	7.70
Mixed		236	25.92

RCTs: randomized controlled trials.

Overall reporting of consent declines

A total of n = 1103 RCTs published in the three journals, only 532/1103 (48.2%) reported consent declines.

Overall participant consent declines from those RCTs which reported the declines

The 532 RCTs that reported consent declines, described a total of 37,53,700 participants. Of these, 3,22,271/37,53,700 had declined consent giving an overall consent decline rate of 8.6%.

Comparison of consent declines between journals

The highest number of articles reporting consent declines were in JAMA (171/267, 64%), followed by the Lancet (170/368, 46.2%) and NEJM (191/468, 40.8%). The highest number of participants declining consent was reported in the Lancet [129150/781,924, 16.5%] followed by JAMA [98,910/834,138, 11.8%] and the NEJM [94,211/2,137,638, 4.4%]. This difference was statistically significant (p < 0.000001; Chi square for trend).

Comparison of the consent declines between countries

The maximum number of consent declines were seen in trials conducted in developed countries (9.2%) and least in trials conducted in developing countries [cOR (95%): 2.27 (2.23,2.32), p < 0.0000001] (4.3%) (Table 2).

Table 2.

Consent declines based on journal, type of country and centers.

Type of country	Participants who declined consent		Participants screened	Percentage of consent declines	p *
Comparison of consent declines based on type of country
Developed	289,005		3,113,458	9.28	<0.0000001
Developing	10,707		248,752	4.30
Mixed	22,559		391,490	5.76
Comparison of consent declines based on type of centres
Unicentric	8775	325,585		2.69	<0.0000001
Multicentric	313,496	3,428,115		9.14	<0.0000001

*Using Chi square test for linear trend.

**Using Chi square test.

Comparison of the consent declines between the centers

Multicentric studies reported higher number of consent declines (9.1%) in comparison to the unicentric studies (2.7%) [cOR 95%, 3.63 (3.56–3.71), p < 0.0000001] (Table 2).

Discussion

Our study evaluated the reporting of consent declines in RCTs published in three broad specialty journals with high impact factors, namely the NEJM, JAMA and The Lancet. We found that as few as 48.2% of all RCTs actually reported consent declines over a four-year period. As a corollary, as many as 52% did not report consent declines, even though this is a mandatory component of the CONSORT flowchart. It appears that this ethical imperative has been missed by authors, reviewers and editors. The real consent decline rate seen among the published RCTs may actually be much higher; however, we can only get the true extent when every RCT reports it as mandated by CONSORT.

Our results are similar to that of Toerien et al.¹⁰ who found that 56.4% studies reported consent declines in six high-impact factor journals. The choice of the Lancet, NEJM and JAMA was based by us on the following criteria —the journals addressing broad specialities, had wide readership and high impact factors. We found a significant variation within the journals with respect to the “consent decline” reporting rate with the highest reporting in JAMA (64%) and lowest in NEJM (41%). The number of participants who ‘declined to participate’ was highest in The Lancet (16.5%) followed by JAMA (11.8%) and NEJM (4.4%). While it is difficult to explain the difference among the journals, a lack of adherence to the CONSORT statement has been bemoaned by Egger et al.¹¹ who found that 92% of RCTs from NEJM and 20% of RCTs from JAMA lacked the CONSORT flow chart. Similarly, Susvirkar et al.¹² found that there was a wide variation [59.38% to 97.14%] within high-impact factor journals in terms of adherence to the CONSORT statement.

We observed a lower consent decline rate (4.3%) in developing countries relative to the developed countries. Osamor et al.¹³ have suggested that the ‘capacity to make decisions are embedded within social and cultural practices’. Developing countries are faced with challenges like poverty, illiteracy, poor health infrastructure which could account for this difference. A study conducted by Doshi et al.¹⁴ in India which evaluated reasons which motivate patients to take part in non-therapeutic trials found that most of them participated because ‘My doctor asked me to’. There are, however, differences in the consent declines even within developing countries. For example, Thaker et al.¹⁵ assessed consent declines among Indian participants approached in several studies over a five-year period and found a consent decline rate of 21%, while Alei et al.¹⁶ from Iran found a consent decline rate of 26%. A true comparison of consent declines in developed versus developing countries can only be made when evidence becomes available in this area.

We also found that participants in multicentric studies had thrice the odds of consent decline as compared to those in single centre studies. In multicentric studies, it is possible that the overall oversight of the study is better and appropriate training is ensured to allow for participants to exercise their rights, thus leading to higher consent declines. However, this is our premise and not substantiated by evidence only.

Our study is limited by the fact that we did not assess the difference in consent declines on the basis of risks involved in the study, relative vulnerability of the participants, study design, complexity of the methodology and sample size of the studies. In addition, our data have been generated from only three journals representing the best of the published literature and much of the published research remians unassessed. There is more research needed to validate consent declines as a metric for assessment of adequacy of informed consent.

Conclusion

We found a low rate of reporting of consent declines in three high-impact factor broad multispecialty journals that impact evidence-based medical practice. We also found a higher rate of consent declines among developed countries relative to developing countries and a higher consent decline rate among multicentric relative to single centre studies. As consent is a metric of autonomy, reporting of this metric is crucial to assess the adequacy of the consent process. Hence, rigorous efforts are needed by journal editors, authors and reviewers to ensure adherence to CONSORT guidelines, in particular the ethical component of the guideline.

Footnotes

Author’s contributions

Figer BH: Contribution to design, acquisition of data, analysis, interpretation and manuscript writing.

Thaker SJ: Contribution to design, acquisition of data, analysis, interpretation and manuscript revision.

Boob M: Contribution to design, acquisition of data, analysis, interpretation and manuscript revision.

Gogtay NJ: Contribution to conception, design, acquisition of data, analysis, interpretation, manuscript writing, revision and approval of the manuscript.

Thatte UM: Contribution to conception, design, acquisition of data, analysis, interpretation, manuscript writing, revision and approval of the manuscript.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Ethics approval

Institutional Ethics Committee approval was obtained for the study and consent waiver obtained.

Availability of data and material

All the data related to the manuscript can be made available in the public domain on request.

References

The Belmont Report. Ethical principles and Guidelines for the protection of Human Subjects of Research, www.hhs.gov/ohrp/regulations-and-policy/belmont-report/index.html (accessed 19 March 2018).

Brock

DW.

Philosophical justifications of informed consent in research. In: Emanuel EJ, Grady CC, Crouch RA, et al. (eds) The Oxford Text Book of Clinical Research. 1st ed. New York: Oxford University press, 2008, pp.606–611.

Kabisch

Ruckes

Seibert-Grafe

et al . Randomized controlled trials: part 17 of a series on evaluation of scientific publications. Deutsches Ärzteblatt Int 2011; 108: 663–668.

Bondemark

Ruf

Randomized controlled trial: the gold standard or an unobtainable fallacy?

Eur J Orthod 2015; 37: 457–461.

The CONSORT Statement, www.consort-statement.org/ (accessed 19 June 2018).

Impact of CONSORT, www.consort-statement.org/about-consort/impact-of-consort (accessed 19 June 2018).

Endorsers Journals and Organizations, www.consort-statement.org/about-consort/endorsers1 (accessed 20 June 2018)

The CONSORT Flow Diagram, www.consort-statement.org/consort-statement/flow-diagram (accessed 20 June 2018).

Country Classification Data sources, country classifications and aggregation methodology. www.un.org/en/development/desa/policy/wesp/wesp_current/2012country_class.pdf. (accessed 10 April 2018).

10.

Toerien

et.al . A review of reporting of participant recruitment and retention in RCTs in six major journals. Trials 2009; 10: 52.

11.

Egger

Jüni

Bartlett

CONSORT Group (Consolidated Standards of Reporting of Trials). Value of flow diagrams in reports of randomized controlled trials. JAMA 2001; 285: 1996–1999.

12.

Susvirkar et al . An assessment of the compliance of the randomized controlled trials [RCTs] published in two high impact factor journals in 2013 with CONSORT statement 2010. Natl Med J India [Accepted for Publication on 11 July 2017].

13.

Osamor

Kass

Decision-making and motivation to participate in biomedical research in southwest Nigeria. Dev World Bioeth 2012; 12: 87–95.

14.

Doshi

Kulkarni

Ghia

et al . Evaluation of factors that motivate participants to consent for non-therapeutic trials in India. J Med Ethics 2013; 39: 391–396.

15.

Thaker

Figer

Gogtay

et al . An audit of consent refusals in clinical research at a tertiary care center in India. J Postgrad Med 2015; 61: 257–263.

16.

Alaei

Pourshams

Altaha

et al . Obtaining informed consent in an illiterate population. Middle East J Dig Dis 2013; 5: 37–40.