Patients and clinicians can use electronic health record (EHR) data to compare treatments and make decisions about care. In clinical trials, researchers randomize patients to treatments to account for patient characteristics. Incorporating mechanisms into EHR systems for randomizing patients to treatments may enable researchers to conduct comparative effective research (CER) in clinical settings. However, randomizing patients within EHR systems requires special software and consideration of ethical and consent issues for patients, and of potential impacts on healthcare processes.
(1) To develop clinical consent processes based on guidance from patients and clinicians; (2) To evaluate software that identifies and randomizes patients within EHR systems
|Software development, qualitative study and analysis, empirical analysis
|Data Sources and Data Sets
- 2 focus groups with 8–10 patients at DHMC
- 1 focus group with 19 staff and clinicians at DHMC
- Follow-up surveys on consent processes
EHR demonstration randomizations:
- 60 patients receiving care at DHMC WWC
- 37 patients receiving care at DHMC CPS
- Development of EHR randomization software
- Qualitative analysis of focus group data from patient and provider stakeholders
- Quantitative analysis of CPS and WWC demonstration randomizations
|EHR randomization software; feasibility, quality, and acceptability of software and consent procedures
The research team conducted separate focus groups with patients and with staff and clinicians at Dartmouth Hitchcock Medical Center (DHMC). The team asked participants about their perceptions of randomization processes, informed consent procedures, and use of CER in clinical settings. Informed consent options included point-of-care written or verbal consent, and global consent for future studies at the time of clinic entry.
Next, the research team evaluated new randomization software embedded within EHR systems. The software identifies patients who are eligible for a study and then notifies clinicians about the patients’ eligibility. Then during clinic visits, clinicians can invite eligible patients to participate in a study and obtain their consent. The software randomly assigns a treatment and inserts an assignment indicator in the patient’s EHR.
To evaluate the software and clinical process, the research team conducted demonstration randomizations in two clinical centers. At the DHMC Weight and Wellness Center (WWC), the demonstration randomized the order of initial dietician and health coach visits. At the Center for Pain and Spine (CPS), the demonstration randomized levels of activity restriction after spine surgery. The team measured patient participation rates in the demonstrations.
Patients, clinicians, and ethics leaders helped design and run the study.
Focus groups. Patients generally had positive views on research conducted in clinical settings. Patients and clinicians preferred written consent over other consent options. They expressed concerns about the duration of global consent; acceptable durations ranged from two months to five years. Clinicians expressed concern about the need for additional staffing to conduct research trials.
Demonstration randomizations. At WWC 60 of 459 eligible participants, or 13%, consented to participate in the randomization. At CPS, 37 out of 48 patients, or 77%, consented to randomization.
The research team evaluated software developed to use in a commonly used EHR system. Further research is needed to adapt and evaluate the software with other EHR systems. The demonstration clinics had experience with EHR-based research; results may differ for clinics with less experience.
Conclusions and Relevance
This study evaluated EHR-embedded software for identifying and randomizing patients for CER in clinical practice. Patients and clinicians preferred individual consent options over global consent. At WWC, telehealth visits due to COVID-19 pandemic restrictions may have affected participation rates.
Future Research Needs
Future research could evaluate the software and processes in other EHR systems and clinics and investigate ways to reduce the administrative burden.