For the almost 50 million surgeries and procedures performed annually in the United States, preoperative medical evaluation serves to identify and optimize perioperative risk to decrease adverse outcomes and to prevent same-day cancellations of surgery [1]. Traditionally, preoperative evaluations are performed face-to-face in the clinic. Starting in July 2014, given its large catchment area, the Veterans Affairs Greater Los Angeles Healthcare System implemented a telemedicine preoperative medicine clinic using Clinical Video Telehealth (CVT). CVT is a technology that Veterans Affairs (VA) providers have used since the early 2000s. With CVT, clinicians can gather relevant history and conduct a limited physical exam using a camera and digital stethoscope. Since adoption, CVT has found increasing rates of use, especially for patients living in rural areas, who face significant barriers to completing in-person visits. In the wake of the COVID-19 pandemic, the importance of providing a telehealth option in outpatient care has become even more apparent [2].

Several studies in non-VA settings have demonstrated that preoperative evaluations done via telemedicine are associated with high patient/provider satisfaction, cost savings, and a lower rate of same-day cancellation when compared to in-person evaluations [3-6]. However, the potential limitations of telemedicine preoperative evaluation (eg, not performing a comprehensive physical exam may preclude clinical diagnoses) may lead to subsequent case-cancellation complications. Thus, our project aimed to evaluate how telemedicine preoperative evaluations using CVT impact hospital length of stay.

There were 62 patients included in this retrospective study. The cancellation rate was 1.74% for CVT versus 3.48% for in-person. Table 1 outlines the distribution of patient characteristics stratified by preoperative visit type. In this unadjusted analysis, there were no significant differences between the cohorts.

Table 2 outlines the negative binomial regression for the association of visit type with hospital length of stay. The age- and Elixhauser score–adjusted incidence rate for hospital length of stay was significantly shorter in patients who underwent preoperative CVT compared to an in-person visit (IRR 0.52, 95% CI 0.29-0.92, P=.02).

In summary, we show that preoperative CVT, while holding age and the Elixhauser comorbidity score constant in the model, has an IRR for hospital length of stay that is 0.52 times lower compared to in-person visits. This study found a significant difference in the IRR of postoperative length of stay between patients receiving telehealth versus in-person preoperative evaluations. This suggests that telemedicine can play a viable role in this clinical setting. Telemedicine has the potential to increase care access across all specialties and health care systems. Our findings had several limitations including that the study was retrospective, was conducted at a single center, and had a low sample size, leading to an increased risk of type II error. Length of stay may be affected by many factors. In the VA patient population, social reasons may affect length of stay more than the typical patient population. There likely is selection bias between those patients who were willing to do CVT versus those who wanted an in-person evaluation.

In our patient population, several patients were more interested in telemedicine compared to in-person visits and we hope to expand to other locations. We plan to apply biomedical informatics to the electronic medical records to extract granular patient data including but not limited to (1) demographic data (age, race, socioeconomic status, and zip code), (2) comorbidities and severity of each comorbidity, (3) postoperative complications, (4) telemedicine-specific data (cancellation rates, missed appointments, and scheduling delays), and (5) patient perceptions and experiences. We hope this research design will help us to identify the benefits and potential disadvantages of telemedicine in the perioperative period. Future studies should be prospective and adequately powered to limit type II error. In addition, future studies should explore how to appropriately triage patients as being “telehealth-appropriate” in the preoperative setting, as well as investigate the effects of preoperative telehealth on other patient-centered outcomes.