Factors influencing the willingness to participate in medical research: a nationwide survey in Taiwan

Introduction

Advances in medical care depend on the participation of human participants. However, participation rate is one of the main challenges researchers face (Charlson & Horwitz, 1984). Evidence shows that the long-term participation rate declines gradually (Galea & Tracey, 2007; Morton, Cahill & Hartge, 2005). Low participation creates some concerns and challenges: (1) smaller sample size reduce the accuracy of estimates and statistical power of hypothesis tests, and increase the chance of making Type II errors (Quinn & Keough, 2002); and (2) the results may not be generalizable to the general population (Kho et al., 2009; Kypri et al., 2011; Maclennan et al., 2012; Madigan et al., 2000). To address the aforementioned concerns, understanding potential factors associated with low participation is important.

After reviewing past literatures, we found studies that involved large samples of Asian people to assess their participation in medical research and to determine the influencing factors were limited (Chu et al., 2015; Kim et al., 2008; Sugawara et al., 2015). Therefore, the purpose of this study was to examine how demographic background and trust in medical research affect the willingness to participate in medical research in Taiwan.

Materials and Methods

Results

About 12 percent of the respondents answered that they would be willing to participate in medical research. Table 1 shows the descriptive statistics of the survey participants. The respondents were divided into two groups: people who were willing to participate in medical research and people who were not. A univariate analysis showed that male participants were more willing to participate in medical research than were female participants (61.82% versus 49.51%).

Table 2 presents the results of a stepwise logistic regression analysis of factors that significantly affect the willingness to participate in medical research. The results in the adjusted model revealed that respondents who had college degrees or above (odds ratio (OR) = 0.43, 95% confidence interval (CI) [0.26–0.70]) and were married or lived with significant others (OR = 0.69, 95% CI [0.49–0.99]) were less likely to participate in medical research. On the other hand, respondents who were men (OR = 1.76, 95% CI [1.25–2.47]), and respondents whose household family members had biomedicine-related degrees or had one themselves (OR = 2.24, 95% CI [1.15–4.37]), were more likely to participate in medical research.

Table 3 indicates how the respondents’ trust in medical research affected their willingness to participate in medical research. After demographic factors, including their education level, sex, marital status, and whether they or their household family members held biomedicine-related degrees, were controlled, the following respondents were more likely to participate in medical research: those who expressed trust in doctors that conducted medical research (OR = 1.81, 95% CI [1.28–2.56]), those who agreed that doctors would never ask them to join medical research study that might harm them (OR = 1.45, 95% CI [1.02–2.07]), those who thought that it was safe to be in a medical research study (OR = 2.38, 95% CI [1.68–3.35]), and those who agreed that medical researchers had no selfish reasons for doing research (OR = 1.76, 95% CI [1.22–2.53]).

Discussion

In this study, we found that only 12 percent of the respondents would be willing to participate in medical research. In addition, people with a higher education level and those who were married or lived with significant others showed a lower willingness to participate in medical research. In comparison, male respondents, and respondents whose household family members held biomedicine-related degrees or had one themselves, were more likely to participate in medical research. Moreover, after significant demographic variables were controlled, people with a higher level of trust in medical research/researchers demonstrated higher willingness to participate in medical research.

Although our study seems to show low willingness from Taiwanese to participate in medical research, the participation rate was comparable to a household survey in the US (11% among adults) (Davis et al., 2013). In addition, potential financial conflict of interest (FCOI) as described in the research vignette of our study might also influence the participation rate. Previous investigations had shown that people had lower trust in studies with FCOI (Aitken, Cunningham-Burley & Pagliari, 2016; Caulfield et al., 2006; Critchley, Bruce & Farrugia, 2013; Liu & Li, 2017), and low public trust in medical research would in turn reduce the willingness to participate in medical research (Caulfield et al., 2006; DeAngelis, 2000; Gatter, 2003).

Our findings on the importance of trust in medical researchers are consistent with a report from Hall and his colleagues (Hall et al., 2006). They developed two measurement tools to study people’s trust in medical researchers, and they found a positive association between trust in medical researchers and the willingness to participate in hypothetical medical research study. The trust issue must be considered critically. Recently in Taiwan, two prominent biomedical-related researchers were involved in FCOIs, and these incidents aroused significant public concern (Pan, 2017; Chang, 2010). Additional studies are warranted to investigate the impact of these incidents on public trust and participation in medical research.

Notably, the negative effect of education level on participation in medical research is contrary to most of the findings from previous studies conducted in other countries (Bouida et al., 2016; Brewer et al., 2014; Brown & Moyer, 2010; Cobb, Singer & Davis, 2014; Davis et al., 2013; Robinson, Ashley & Haynes, 1996). Our findings are also somewhat different from the results of the three studies conducted in South Korea and Japan, in which the researchers had found no association between education level and participation, and no association between marital status and participation in clinical trials (Chu et al., 2015; Kim et al., 2008; Sugawara et al., 2015). Such discrepancies might be due to the difference between study population of each study. For example, the two studies in Korea and Japan recruited their study participants from hospital patients, and not from general population (Kim et al., 2008; Sugawara et al., 2015). Another explanation might be the trust in doctors. We performed an additional analysis and found that participants with a higher education level expressed lower trust in doctors who conducted medical research (Table S1). This finding was consistent with some previous studies (Freburger et al., 2003; Kayaniyil et al., 2009; O’Malley et al., 2004; Stepanikova et al., 2006). We postulate that our participants with a higher education level had lower trust in doctors; thus, they were less likely to participate in medical research.

It is likewise noteworthy that whether our participants or their close family members held biomedicine-related degrees would positively influence their willingness to join medical studies. This might imply that providing background knowledge would be necessary to increase people’s willingness to participate in medical research. Improving the participants’ awareness regarding why medical research is necessary and how it is conducted might be important (Du et al., 2008). In fact, that is exactly why the National Research Program for Biopharmaceuticals, funded by Taiwanese government, included a clinical trial awareness project (Tsai, 2017).

Our study revealed that women were less likely to participate in medical research than men. The findings of previous studies from other countries were inconsistent on whether sex might influence people’s willingness to participant in medical research. While some studies showed that women expressed lower willingness to participate in medical research because of concerns with potential reproductive adverse effects (Parekh et al., 2011; Wizemann & Pardue, 2001), a review concluded that women were more likely to participate in medical research (Galea & Tracey, 2007). While on the other hand, several nationally representative surveys found no difference on sex for participation rate (Chu et al., 2015; Cobb, Singer & Davis, 2014; Davis et al., 2013; Ewing et al., 2015). Nevertheless, evidence did demonstrate that clinical trials had not always adequately enrolled women (Liu & DiPietro Mager, 2016). For example, according to the US FDA statistics, women made up less than 33% of the participants in phase I trials between 2006 and 2007 (Pinnow et al., 2009). Since there are not enough empirical data on this issue in Taiwan, more studies are required to identify why Taiwanese women may be less willing to participate in medical research.

Our study has the following advantages: (1) limited studies, such as the present study, have examined factors influencing the willingness of Asians to participate in medical research; (2) the present study was conducted as a nationwide survey with a representative sample of adult population in Taiwan; and (3) this study involved face-to-face interviews that were conducted by well-trained interviewers.

Despite the aforementioned merits, the results should be interpreted with caution because of the following limitations: (1) the cross-sectional design of this study created difficulty to distinguish between causes and outcomes, and the reasons behind the factors affecting the willingness to participate in medical research are still undetermined; more follow-up studies, such as examining whether a clinical trial awareness program may lead to better participation rate, are still needed; (2) the clinical trial described in the research vignette was an investigation of a medicinal product, but the respondents might have answered differently if it was another type of medical research; and (3) some medical studies might focus on specific populations, such as children or elders aged above 70 years, which were not included in the survey data we used; the results might be different if the same survey was conducted on the population not included or specified in this study.

Conclusions

In conclusion, our study identified several demographic factors associated with the willingness to participate in medical research. In addition, trust in medical research seemed to be an independent factor relating to the participation rate. These findings are useful for developing strategies to improve the recruitment of (Asian) participants. Relevant discussions on research ethics and policies, such as the importance of public trust in medical researchers, could also be based on this study.

Supplemental Information