Patient hesitancy in perioperative clinical trial enrollment during the COVID-19 pandemic

Abstract

The COVID-19 pandemic has caused tremendous disruptions to non-COVID-19 clinical research. However, there has been little investigation on how patients themselves have responded to clinical trial recruitment during the COVID-19 pandemic. To investigate the effect of the COVID-19 pandemic on rates of patient consent to enrollment into non-COVID-19 clinical trials, we carried out a cross-sectional study using data from the Nitric Oxide/Acute Kidney Injury (NO/AKI) and Minimizing ICU Neurological Dysfunction with Dexmedetomidine-Induced Sleep (MINDDS) trials. All patients eligible for the NO/AKI or MINDDS trials who came to the hospital for cardiac surgery and were approached to gain consent to enrollment were included in the current study. We defined "Before COVID-19" as the time between the start of the relevant clinical trial and the date when efforts toward that clinical trial were deescalated by the hospital due to COVID-19. We defined "During COVID-19" as the time between trial de-escalation and trial completion. 5,015 patients were screened for eligibility. 3,851 were excluded, and 1,434 were approached to gain consent to enrollment. The rate of consent to enrollment was 64% in the "Before COVID-19" group and 45% in the "During COVID-19" group (n = 1,334, P<0.001) (RR = 0.70, 95% CI 0.62 to 0.80, P<0.001). Thus, we found that rates of consent to enrollment into the NO/AKI and MINDDS trials dropped significantly with the onset of the COVID-19 pandemic. Patient demographic and socioeconomic status data collected from electronic medical records and patient survey data did not shed light on possible explanations for this observed drop, indicating that there were likely other factors at play that were not directly measured in the current study. Increased patient hesitancy to enroll in clinical trials can have detrimental effects on clinical science, patient health, and patient healthcare experience, so understanding and addressing this issue during the COVID-19 pandemic is crucial.

Fig 1. Patient screening and enrollment for the NO/AKI and MINDDS trials.

All patients presenting for cardiac surgery were screened for enrollment. Patients were excluded from the trials if they were found to be ineligible based on inclusion or exclusion criteria, enrolled in another study, or for various other reasons.

Fig 2. The rate of consent to enrollment into the NO/AKI and MINDDS trials dropped significantly during COVID-19.

The rate of consent to enrollment among approached eligible patients was 64% (n = 1,001) in the “Before COVID-19” group and 45% (n = 333) in the “During COVID-19” group. Statistical analysis revealed a statistically significant difference in enrollment inquiry responses between the two time frames (n = 1,334, P<0.001) and a difference in proportions of 19% (95% CI 12.88% to 25.12%).

Fig 3. General demographic factors did not differ significantly between the “Before COVID-19” and “During COVID-19” groups of the NO/AKI and MINDDS trials.

(A.) There was no statistically significant difference in mean age (n = 1,402, P = 0.34), (B.) sex frequency (n = 1,405, P = 0.47), or (C.) race and ethnicity frequency (n = 1,272, P = 0.27) between the two time frames.

Fig 4. Residential SES indicators did not differ significantly between the “Before COVID-19” and “During COVID-19” groups of the NO/AKI and MINDDS trials.

(A.) There was no statistically significant difference in the median ZCTA MAHI (n = 1,397, P = 0.05), (B.) median ZCTA MEAL (n = 1,398, P = 0.51), or (C.) mean ZCTA unemployment rate (n = 1,398, P = 0.64) between the two time frames.

Fig 5. Responses to the healthcare experience survey questions did not differ significantly between the “During COVID-19” and “Before and COVID-19” groups of the NO/AKI trial.

(A.) There was no statistically significant association between responses to question eight (n = 90, P = 0.68) or (B.) question nine (n = 85, P = 0.61) and time frame. (C.) There was no statistically significant difference in the median response to question 10 (n = 88, P = 0.58) between the two time frames.

Fig 6. Responses to the trial satisfaction survey questions did not differ significantly between the “During COVID-19” and “Before and COVID-19” groups of the NO/AKI trial.

(A.) There was no statistically significant difference in the median response to question 11 (n = 91, P = 0.75) between the two time frames. (B.) There was no statistically significant association between responses to question 12 (n = 91, P = 1) or (C.) question 13 (n = 91, P = 1) and time frame.