How do study design features and participant characteristics influence willingness to participate in clinical trials? Results from a choice experiment

Thomas, Caitlin; Mulnick, Sarah; Krucien, Nicolas; Marsh, Kevin

doi:10.1186/s12874-022-01803-6

Table 4 Participant characteristics predicting willingness to participate (logistic regression results)

From: How do study design features and participant characteristics influence willingness to participate in clinical trials? Results from a choice experiment

Characteristics	Reference	Levels	MLE (SE) ^a
Characteristics	Reference	Levels	Model 1: All	Model 2: Disease & Country	Model 3: All except disease & country
Constant		-	-1.042 (0.873)	-0.067 (0.334)	-0.286 (0.652)
Disease-related variables
Time since first diagnosis, years	0–5	≥ 6	-0.498 (0.277) ^*	-0.527 (0.240) ^**	-
Number of medical conditions	1–2	≥ 3	0.850 (0.547)	0.794 (0.481) ^*	-
Anemia	No	Yes	0.100 (0.417)	0.038 (0.365)	-
Anxiety	No	Yes	-0.007 (0.399)	-0.002 (0.357)	-
Depression	No	Yes	-0.808 (0.445) ^*	-0.491 (0.380)	-
Diabetes (Type I or Type II)	No	Yes	-0.206 (0.354)	-0.024 (0.304)	-
Heart disease	No	Yes	-0.394 (0.352)	-0.040 (0.312)	-
Migraine	No	Yes	0.300 (0.358)	0.363 (0.321)	-
Multiple sclerosis	No	Yes	0.130 (0.492)	0.404 (0.431)	-
Rheumatoid arthritis	No	Yes	-0.022 (0.332)	0.179 (0.303)	-
Cancer	No	Lower mortality ^b	0.055 (0.413)	0.300 (0.345)	-
	No	Higher mortality ^c	-0.103 (0.407)	0.249 (0.357)	-
Treatment-related variables
Number of previous lines of treatment	0–2	≥ 3	0.320 (0.276)	-	0.531 (0.251) ^**
Affordability of current medication	I can easily afford my medications	I can afford my medications	-0.357 (0.288)	-	-0.276 (0.263)
	I can easily afford my medications	I struggle to pay for my medications	-0.224 (0.501)	-	-0.020 (0.467)
Experience of treatment side effects	None	Mild	0.558 (0.288) ^*	-	0.594 (0.271) ^**
	None	Severe	0.756 (0.431)^*	-	0.753 (0.390)^*
COVID vaccine	No	Yes	0.269 (0.332)	-	-
Direct experience of clinical trial	No	Yes	0.409 (0.403)	-	-
Health-related variables
I feel fatigued	Not at all/ A little bit	Somewhat/Quite a bit/Very Much	0.192 (0.335)	-	-0.025 (0.288)
I have had difficulty with mobility	Not at all/ A little bit	Somewhat/Quite a Bit/Very Much	0.115 (0.375)	-	0.168 (0.329)
I have had pain or discomfort	Not at all/ A little bit	Somewhat/Quite a Bit/Very Much	-0.120 (0.337)	-	0.115 (0.288)
I have been anxious or depressed	Not at all/ A little bit	Somewhat/Quite a Bit/Very Much	-0.551 (0.375)	-	-0.553 (0.303)^*
I have to limit my social activity because of my health	Not at all/ A little Bit /Somewhat	Quite a Bit/Very Much	0.795 (0.410)^*	-	0.808 (0.384)^**
I have had difficulty with performing my usual activities	Not at all/A little Bit /Somewhat	Quite a Bit/Very Much	-0.577 (0.462)	-	-0.704 (0.435)
Current QoL (Classification) ^d	0–7	≥ 8	0.151 (0.465)	-	-
QoL vs. last year	Improve	Same/Deteriorate	-0.213 (0.315)	-	-0.393 (0.289)
Self-perceived likelihood to live to average age ^e	Neutral/High	Low	0.716 (0.367)^*	-	0.266 (0.345)
Self-perceived future quality of life	Improve	Same/Deteriorate	0.048 (0.427)	-	0.035 (0.389)
Sociodemographic variables
Age, years	18–50	51–65	-0.459 (0.401)	-	-1.014 (0.355)^***
	18–50	≥ 66	-0.916 (0.557)	-	-1.779 (0.427)^***
Sex at birth	Male	Female	-0.177 (0.263)	-	-0.158 (0.243)
Country of residence	USA	Poland	0.482 (0.471)	0.759 (0.249)^***	-
	USA	China	1.899 (0.516)^***	2.162 (0.357)^***	-
Paid job	Yes	No	-0.105 (0.354)	-	-0.198 (0.334)
Education	Other than university	University	-0.468 (0.318)	-	-0.162 (0.286)
Living alone	Yes	No	0.116 (0.373)	-	0.138 (0.347)
Looking after dependent(s)	Yes	No	0.250 (0.318)	-	0.198 (0.294)
Attitudinal variables
Trust in medical researchers ^f	Low (1–3)	High (4–5)	0.275 (0.341)	-	0.523 (0.327)
Altruism ^g	Low (1–3)	High (4–5)	1.074 (0.341)^***	-	1.017 (0.321)^***
Other variables
Survey version	Version 1	Version 2	0.979 (0.261)^***	-	0.825 (0.242)^***
BIC			695.2	590	625.3
PV			72.92%	71.88%	73.54%

Abbreviations: BIC, Bayesian Information Criterion; CP, % of correctly predicted choices; MLE, maximum likelihood estimation; PV, % of unseen choices correctly predicted (holdout sample); QoL, quality of life
^*p < 0.1; **p < 0.05; ***p < 0.01
^a The constant is the reference group of all dummy variables (all dummy variables = 0). Each dummy coded variable coefficient represents the difference from the reference
^b Lower mortality cancer = breast cancer, melanoma / skin cancer, prostate cancer, thyroid cancer
^c Higher mortality cancer = colorectal cancer, leukemia or another blood-related cancer, liver cancer, lung cancer, multiple myeloma, ovarian cancer, stomach cancer
^d The 6 QoL domain-related questions (fatigue, social activity, mobility, impact on usual activities, discomfort, anxiety/depression) were rated on 0–4-point rating scales (where 0 = Not at all, and 4 = Very much). Total scores were computed (ranging from 0–24) and then classified into one of three groups: 0–7; 8–16; and 17–24
^e Participants asked the likelihood they will live for the next 50 years if aged 18–35, the next 40 years if aged 36–50, the next 25 years if aged 51–65, or the next 10 years if aged ≥ 66
^f Trust in medical researchers assessed by asking participants to rate their agreement with item q3 of the Research Attitude Questionnaire [30]: ‘Medical researchers can be trusted to protect the interests of people who take part in their studies’. Rated on a scale of 1 (strongly disagree) to 5 (strongly agree)
^g Altruism assessed by asking participants to rate their agreement with item q4 of the Research Attitude Questionnaire [30]: ‘We all have some responsibility to help others by volunteering for medical research’. Rated on a scale of 1 (strongly disagree) to 5 (strongly agree)

Back to article page

ISSN: 1471-2288

Contact us

General enquiries: journalsubmissions@springernature.com

BMC Medical Research Methodology

Contact us