Shifting From In-Person to Virtual Care Had Small, Short-Term Impacts on Health Care Practices

A rapid shift from in-person to virtual care did not compromise health care-related practices in the long-term. These results from a recent study were published in Medical Care.

As a reaction to the COVID-19 pandemic, health care delivery was shifted to virtual platforms, where possible. This shift, while protecting patients and providers from infection risk was not without its own unique risks. For example, moving to virtual care has the potential to disrupt pharmacotherapy, health care-related practices, such as measurement-based care, and to affect provider or patient engagement.

At Kaiser Permanente (KP), a virtual-care-first model was rapidly deployed in March 2020 when a worldwide pandemic was declared. In this study, data from 3 KP health systems: Mid-Atlantic States (KPMAS), Georgia (KPGA), and Colorado (KPCO) were used to evaluate what the effects on health care-related practices the shift to virtual care had among patients with a new diagnosis of major depression. Data from January 2019-March 2020 (prepandemic), April 2020-June 2020 (peak pandemic), and July 2020-June 2021 (recovery) were retrospectively assessed. To balance for cohort differences, an inverse of probability of treatment weighting (IPTW) approach was used.

During the entire study period (January 2019-June 2021), 12,266, 8238, and 8227 virtual first follow-up visits occurred after an incident diagnosis of major depression at KPMAS, KPGA, and KPCO, respectively. Stratified by study site, significant patient characteristics for age, gender, race and ethnicity, comorbidities, and area deprivation were observed.

Between prepandemic and peak pandemic periods, antidepressant medication orders significantly decreased from 20.3% to 19.0% (P =.0235) at KPMAS and 17.6% to 13.0% (P =.0011) at KPCO and decreased from 32.9% to 29.7% (P =.0654) at KPGA. These declines rebounded between peak pandemic and recovery from 19.1% to 21.9% (P =.0053) at KPMAS and 13.1% to 15.3% (P =.0380) at KPCO but not at KPGA (29.6% vs 27.3%; P =.0715), respectively.

No significant trends in fulfillment of antidepressant medication orders were observed at any site at any time period, however, fewer orders tended to be filled between prepandemic and peak pandemic periods at KPCO (78.4% vs 70.3%; P =.0650), respectively.

At all 3 sites, completion of behavioral health symptom screeners increased between prepandemic and peak pandemic periods (all P <.0001) and again between peak pandemic and recovery (all P ≤.0003). The final rate of symptom screener completion was 81.0% at KPMAS, 46.4% at KPGA, and 35.5% at KPCO.

The findings of this study may not be generalizable for smaller health care systems that do not have the infrastructure that KP has.

These data indicated that the rapid shift to virtual behavioral health care did not appear to significantly impact health care-related practices negatively. Although there was an initial significant decline in antidepressant orders at 2 of the 3 sites, a rapid rebound was observed. In addition, the shift to virtual care may have caused clinicians to adopt symptom screeners more than before.

We spoke with Nancy S. Weinfield, PhD, of Kaiser Permanente Mid-Atlantic Permanente Research Institute and lead author of the study.

What were the motivations for this study?

The study is 1 piece of a larger study of virtual care at Kaiser Permanente that was already underway when the pandemic was declared. Other articles in the same journal issue represent different aspects of that larger study. The onset of the pandemic accelerated the use of virtual care, and changed some of our focus. This particular article came from the idea that virtual care for depression was already becoming more common prepandemic. Although virtual care offered some clear benefits for treatment of major depression such as reducing travel time, there were some care practices such as medication orders, filling prescriptions, and completing symptom screeners that seemed easier to do when everything was on-site. When the pandemic caused a rapid shift away from in-person care, physicians, therapists and patients needed to pivot to overcome those challenges of virtual care. This study addresses how those challenges were met.

Were you surprised by any of the findings?

I was somewhat surprised by 2 things: the steep increases in completion of the symptom screeners over a short period of time at the onset of the pandemic, and the lack of a significant decrease in fulfillment of medication orders during the peak pandemic period. They are good surprises, and indicate a high ability to adapt to the change in circumstances from both the behavioral health professionals and patients.

Do you envisage the success of transitioning to virtual care to be replicable at other institutions? What systems that Kaiser Permanente has in place are needed to replicate these results?

Kaiser Permanente had a virtual care program prepandemic, and each region invested substantial resources into shifting physicians, therapists and patients into virtual care who hadn’t used it before. There was a lot of training, and the effort was backed up by strong information technology. Replicating these results may be difficult for smaller health systems that have limited experience and resources.

Do you think telehealth is the future of major depression maintenance and management? What is missed in exclusive virtual mental health care, or conversely, what is gained?

I think telehealth is and will continue to be an important strategy for MDD treatment, but not the only strategy. Each patient, physician and therapist will work best under different circumstances. The aspects of virtual care that are most and least beneficial for treatment outcomes, and for whom, is an important topic for future research.

What is the biggest take-home message from this work?

For me, the take-home message is one of resilience of physicians, therapists, patients, and the health systems in the face of challenging circumstances for behavioral health care. Everyone had to adapt to pivot to virtual care without notice, so virtual care was taking place in some cases between providers and patients who were both new to the modality. The findings from this study show that the efforts supporting that pivot maintained the quality of virtual care, and even increased the use of measurement-based virtual care. It’s an impressive example of the type of change that can happen when everyone is committed to making it work.