The Veterans Administration system supports telemedicine (TM) to provide medical consultations between patients and physicians via videoconference. At present, little is known about the impact of such TM consultations on patient-physician communication and related health outcomes. Analyses of in-person (IP) medical encounters have shown that effective patient-physician communication is associated with improved health outcomes.
To determine whether the physical separation between patient and physician required during TM has an affect on patient-physician communication and related outcomes, including patient and physician satisfaction, patient compliance, and patient understanding of medical care.
In this clinical trial, 238 patients were randomized to receive either consultative care at the remote site via TM with a consultant physician located the Milwaukee VA (intervention) OR by an IP consultation with a consultant physician at the Milwaukee VA (control). The same group of consultant physicians provided both IP and TM consultations.
Patients in both study arms had their medical encounter video recorded. We compared patterns and quality of patient-physician communication for the TM and IP encounters, using the Roter Interaction Analysis System. Data on patient and physician satisfaction with the encounter and patients' understanding of their medical problems were collected at the end of each medical encounter. Patient compliance (medication refill behavior) was assessed at 90 days post visit. The frequency of communication behaviors during the TM and IP encounters was compared using the analysis of a Linear Mixed Model. Comparison of patient satisfaction, physician satisfaction, patient compliance, and patient knowledge measures between TM and IP groups were conducted with similar Linear Mixed Models.
We randomly assigned 221 subjects to receive TM (n = 111) or IP (n =110) visits. Of 221 patients enrolled, 14 patients were out of study and 7 were missing video data. Linear mixed models were used for analyses, with type of visit (TM or IP) as fixed effect and physician as random effect. Patients were less likely to participate during TM visits as reflected by a higher Verbal Dominance ratio (physician talk / patient talk) for TM as compared to IP visits (ratio = 1.54 vs. 1.30, p = 0.001). Patient-centered ratio (patient-centered communication categories / physician-centered communication) was similar for TM and IP visits (0.92 vs. 0.87 p = 0.39). Physician data gathering on biomedical topics (p = 0.41) was similar in TM and IP visits. However, physician data gathering for psychosocial topics was significantly lower during TM vs. IP visits (p = 0.02). Rapport building (positive talk and social talk) was higher during TM vs. IP (p = 0001). For each of these analyses a significant physician effect (p<0.001) was present. Patients were noted to be inhibited during TM, they were less likely to ask for psychosocial information during TM vs. IP visits (p = 0.001). Patients were also less likely to engage in shared decision-making and partnership building during TM vs. IP visits (p = 001).
This study identified effective and ineffective communication behaviors during TM consultations and has improved our understanding of how TM technology affects the patient-physician relationship and clinical outcomes. We hope that our findings will foster the development and testing of specific training interventions to maximize the quality of patient-physician interaction during clinical TM consultations.
None at this time.
Rural, Satisfaction (patient), Telemedicine