Patients’ Views on Returning to the Office for In-Person Care During the COVID-19 Pandemic: A Real-Time Survey to Inform Institutional Healthcare Planning Decisions

The COVID-19 pandemic has affected clinical practice in a variety of ways. In the ambulatory setting, perhaps the most significant has been the sudden transition of telemedicine from an experimental and occasional event to the currently dominant way to deliver care in many settings.(1) At the time of this project (April–May, 2020) our primary care group achieved conversion to telemedicine for about 85% of our visits in about one week. In addition to payment for telemedicine services by CMS, many have noted other benefits from this experience, including new glimpses of patients’ homes during these electronic “home visits” and greater ease and efficiency for homebound, chronically ill, and travel-challenged elderly patients. Telemedicine visits have been assisted by effective use of increased technology for performing components of physical examinations remotely, such as the following(2):

• Photos of rashes or edematous legs;

• On-line screening questionnaires;

• Home blood pressure readings; and

• Home international normalized ratio (INR) and pulse oxygen monitoring equipment.

However, reported evidence shows that deferred care due to patients’ anxieties about possible virus exposure if they visit the clinic may have both lethal health consequences and economic costs.(3) Early reports have estimated that the frequency of patients not seeking care for COVID-19 and dying as a result has added to the number of excess deaths associated with this pandemic. More recent data also suggest that patients may be avoiding care for other chronic conditions and potentially suffering more morbidity and mortality as a result.(4) As a result of these findings, many are asking whether and how much telemedicine is appropriate and how much will be here to stay even after the COVID-19 pandemic wanes.(5)

Understanding patients’ concerns about returning for in-person care, more specifically, is valuable in helping to address those concerns, so that patients will return to receive appropriate care for both acute and chronic health conditions. At the time of this writing, many healthcare institutions have shuttered significant components of their facilities, both to make way for patients with COVID-19 and to protect their other patients.(6) These institutions have suffered significant financial losses as a result, because they have canceled margin-contributing elective surgeries and reduced their in-patient census to clear wards for potential increases in COVID-19 inpatients.(7) Now, along with the rest of the business community, these institutions also have begun planning to open back up carefully and cautiously. This requires budget forecasting based on predictions of post–COVID-19 caseloads, occupancy rates, and ambulatory visit rates. For example, in our institution, ambulatory visit rates and, therefore, income initially were estimated to be substantially lower for the upcoming year, affecting organizational planning for faculty salaries and clinic staffing needs. These estimates of reduced caseloads were based on a number of potential factors, including predictions of loss of employment-based health insurance affecting patients’ ability to receive services and patient anxieties about coming back into clinics for in-person care.

Thus, our survey was undertaken to determine our patients’ concerns about returning to clinic so we could try to address those concerns and maximize their comfort regarding coming in for appropriate care. Additionally, we wanted to assess the effectiveness of using the secured patient Web portal within our EHR to carry out this survey quickly and efficiently, thus serving as a potential practice-based research tool in the future.

Methods

Study Design and Subjects

All adult patients who were seen by a primary care physician in the previous two years were eligible to receive a survey. This included all patients 18 years of age or over, and seen in Family Medicine, Internal Medicine, and Geriatric clinics at least once by a physician in clinic since January 1, 2019. There was a total of 54,626 patients in Epic with a PCP visit, and of these 47,435 (87%) had active MyChart accounts (secure patient portal for access to their medical record and communication with the health team). These 47,435 patients were sent invitations via MyChart to complete and return a survey.

These patients were divided into aliquots of 5000 patients each. A separate aliquot was sent out every few days between May 12 and June 10, 2020 in order to not overwhelm the system. Over a period of about three weeks, all eligible patients received an invitation to complete the survey.

Measure

The two primary authors (WJS, RK) created a list of questions based on clinical experience and a review of the literature regarding concerns patients were expressing about returning to the clinic. Twelve questions addressing concerns about places, sources, and methods of exposure to COVID-19 were asked in a series of 5-point Likert scale items. Two additional, open-ended questions were asked: “What one change would make you most comfortable coming back into clinic for care?” and “What one thing makes you most reluctant to see your provider in clinic?” One question asked patients if they had any desire for psychological or emotional support during these stressful times of the pandemic. Patients who stated they did have an interest were directed to ask their PCP for a referral to our integrated behavioral health care team. In response to a review of the initially completed surveys (after the first 10,000 surveys had been sent out) a final question was added: “If you have a medical need for a face-to-face visit with your doctor, would you be willing to come into the office for that visit?” This addition demonstrated the ease with which the survey could be modified, adding an essential question overlooked initially, while recognizing the slightly diminished statistical representation of this additional question. A final set of four questions surveyed patients’ experience with telemedicine visits and their interest in continuing to have access to this technology in the future. (The final survey is shown in Appendix A.) These questions and a cover letter that explained the project and invited patients to participate were then reviewed and approved by leadership in the other two divisions participating (General Internal Medicine and Geriatric Medicine). UCSD’s Institutional Review Board approved a waiver of informed consent.

In an attempt to increase response rate, no questions about personal health information were asked; we asked about age only to assess trends in the concerns expressed given evidence of age as a key risk factor for COVID-19 consequences. These initial survey items were then circulated for review to members of the clinical leadership team across primary care clinics. The cover letter (and other communications) informed patients of steps clinics had taken to lower the risk of patient exposure to COVID-19 when they are in clinic.

Method

The MyChart portal was used to invite patients to participate. This methodology was developed by the primary author (WJS) and has been used by investigators across the UCSD School of Medicine. It allows the patient population to be queried on given criteria (e.g., patient age, diagnoses, treatments received) as long there is a member of the research team who is a member of the primary care health care team. This allows patients to be invited to participate in the research without allowing any research personnel who are not part of the healthcare team to become aware of patient information, until the patient provides consent to be contacted. In this project such criteria were met, because both the first and second authors are part of the primary care clinical leadership. Given the large number of patients to be sent a survey, the cohort was separated into aliquots of approximately 5000 patients. Once loaded into the system, the second author (RK) then sent out the letter of invitation and a URL/hot link for patients to complete the survey via Red Cap, supported through UCSD’s Clincal & Translational Research Institute.

Response rates to the survey were calculated by age group. Data analysis was performed with R software version 4.0. Data analysis began with descriptive frequencies of patients’ responses. Percentages of responses of 4 or 5 were used to classify high levels of concern and then order the relative severity of concern for each survey item. Ordinal logistic and binary regression analyses (Harrel F E, 2021; Rms: Regression Modeling Strategies. R package, version 6.2-0.) were performed comparing polytomous and binary responses to each survey question by different age groups, with those under the age of 30 serving as the reference group. To facilitate evaluation of the impact of observed differences in responses between age groups, we computed the estimated mean item response for each age group by summing the value of each item response option (1 through 5) multiplied by the model estimated probability of each response option for each age group. Bootstrap 95% confidence intervals for each mean from 200 samples also are displayed. Qualitative analysis of text responses to the two open-ended questions will be conducted in the future.

Results

Table 1 shows the breakdown by age group of the surveys sent out and those completed. Of the patients sent surveys, 29,978 actually opened the invitation message. The overall response rate was 17.5%, with 8503 patients completing the survey. More than half of the patients who were sent surveys were over 50 years of age, and the older patients completed the survey at a significantly higher rate than the younger age groups: with response rates of 6%, 9%, 20%, and 29% differing significantly (X[squared] = 3564.4, df = 3, p <.01) between each increasing age group. There were 197 (2%) respondents who did not report their age, and all regression models were restricted to the remaining 8306 respondents. Patients over the age of 50 were at least three times more likely to complete the survey than those under the age of 30.

The highest level of patient concern was found to be contracting COVID-19 from other patients (33% reported a 4 or 5). This was closely followed by concerns about using elevators (27% rated 4 or 5), and coming in contact with surfaces in waiting rooms, restrooms, and check-in areas (18%–21% rated 4 or 5). The requirement to wear masks was a high concern (rated 4 or 5) for about one in five patients. Table 2 shows the percentage of patients who responded, with levels of concern (rated 1 to 5) on each survey item.

Using ordinal logistic regression we observed small but statistically significant (p <.01) differences in the probability of reporting greater concerns across age groups for most survey questions. Mean expected item responses are displayed for COVID sources of risk items (Figure 1). Although the pattern of responses across groups to survey items was statistically significant, as can be seen in Figure 1, the mean differences appear quite small and likely are not clinically meaningful. Estimated mean ratings of video visits for each age group had larger differences than other questions and are presented in Figure 2. In general, respondents in the 66+ age group reported lower ratings to questions about concern contracting COVID from providers or the clinic environment. Respondents in the 66+ age group had higher visit concerns about wearing a mask, using the restrooms, and use of assisted transportation to get to the office than respondents in other age groups. Video visits were reported by 41% of all respondents (see Table 2). Among those who attended a video visit (n = 3431), about one third of respondents found video visits more efficient, whereas one third of respondents rated them as less efficient. Most indicated they would like to have video visits after COVID restrictions are removed (48% rated 4 or 5), and low levels of technical problems were reported (15% rated 4 or 5). Respondents aged 31 to 50 (OR = 0.87; 95% CI: 0.77-0.98, p <.01), and 51 to 65 (OR = 0.82; 95% CI: 0.74-0.91, p <.01) were less likely than those over 66 to have attended a video visit in the past two months (see Figure 2). Respondents aged 18 to 30 (OR = 0.88, 95% CI: 0.71-1.10, p = .27) were similarly less likely than those over 66 to have attended a video visit, although the estimate was not significant statistically. Respondents in the over-66 age group consistently reported lower ratings of usefulness (p values <.01), efficiency (p values <.01), and liking to have video visits in the future (p values <.01) than each younger respondent age group.

Figure 1. Concern about exposure in clinic. Dots represent mean response for that age group.

Figure 2. Experience with video visits. Dots represent mean response for that age group.

In logistic regression models, the rates of those wanting psychological support (4%) were low (see Table 2). Respondents aged 18 to 30 (OR =3.44; 95% CI: 2.24-5.29), 30 to 50 (OR = 2.49; 95% CI: 1.85-3.35), and 51 to 65 (OR = 1.40; 95% CI: 1.04-1.88) each had significantly higher odds (p values <.01) of wanting psychological support than respondents aged 66 or older. Rates were highest among ages 18 to 30 (8%), with rates of 6%, 3%, and 2%, respectively, for each subsequently older age group. With analyses restricted to the last two cohorts of respondents (n = 2853) who were asked if they were willing to come into the office if they had a medical need, 93% responded positively, and there was no significant difference by age group (X[squared] = 3.91, df = 3, p = .27).

Patient responses to the open-ended questions of what would help patients feel most comfortable and what conditions would make them most uncomfortable about returning to clinic were analyzed separately, used in communications to patients, and may be reported in the future.

Discussion

Our results clearly show that across all age groups our patients are generally not afraid to return to in-person care. There appears to be less concern over exposure to the virus via healthcare providers and other sources than there is about exposure to fellow patients, whom they perceive as presenting a bigger threat. Of the substantial proportion who experienced a telemedicine visit, patients found them roughly equivalent to in-person visits in effectiveness and efficiency, with a substantial number of them expressing interest in continuing to use this technology in the future—with a slightly greater desire among 31- to 50-year-olds and less so among those over 65 years of age.

The significantly greater overall response rate from the older patients is seen as a benefit, given they are the population generally viewed not only as being more vulnerable to viral exposure and severe outcomes, but also as having more chronic disease burden that continues to need ongoing care. This survey, completed by most patients in late May 2020, also showed relatively little interest in psychological services. Given that older patents are less likely to request psychological services than younger patients,(8) this could be expected. Yet it is somewhat surprising given the anxiety levels we have seen being expressed at an individual level regarding COVID-19 in general. Perhaps the interest and need for such services will change as quarantine and other consequences of public health interventions continue.

The possibility of a substantial number of telemedicine visits continuing into the future may have significant impacts on our practice structures and functioning. It may well be that we will need fewer exam rooms and more “telemedicine” rooms. It may be that we will need less support staff, although some argue that taking care of patients requires specific actions that do not disappear when the patient is seen via telemedicine. Technology support and sophistication as well as costs will certainly be factors. All these effects should be studied and reported in the literature.

Leveraging the EHR for both clinical outcome data as well as its patient Web portal as a source of patient recruitment provides key support for carrying out practice-based research.

In the academic primary care setting there is growing emphasis on faculty and residents being involved in significant scholarly activity. We believe that practice-based research should be one of the main scholarly activities for family medicine/primary care faculty. The current project demonstrates the effective use of the EHR’s patient Web portal to elicit the opinions of a large number of patients very quickly. Without any offered incentive to complete the survey, over 8500 patients provided their concerns over returning to the clinic for care, let us know their experience with video visits, and told us whether they needed psychological support during these very stressful times. This information was quickly included in plans for shifting resources and planning for delivery of primary care in the fast-changing world of primary care during the COVID-19 pandemic.

We think this study, conceived and completed in about five weeks and providing timely data to help practice-based decision making, is a very good example of practice-based research. Leveraging the EHR for both clinical outcome data as well as its patient Web portal as a source of patient recruitment provides key support for carrying out practice-based research. Linking the recruitment offer with an external patient-reported outcomes tool such as RedCap, Qualtrix, or others, is an efficient way to capture data that does not need to reside in the medical record. If the latter is desired, then building questionnaires within the EHR that the patient can access from the Web portal is a very workable approach. Having available department or institution-based EHR technical support is critical to being able to use these sophisticated EHR tools to accomplish such scholarly practice-based research projects.

This survey was limited to primary care patients who actively used their MyChart accounts. As a result, we may have missed some of the most vulnerable patients in our practices. Because we did not collect other sociodemographic or healthcare variables, it is unclear how the 17.5% who responded differed from those who did not respond. The survey was mostly done in late May 2020 when the San Diego community was starting to open up and the community was beginning to be optimistic about decreasing COVID-19 spread. The potential for further surveys of patient attitudes as the pandemic continues has now been demonstrated.

Acknowledgment: The authors wish to acknowledge Dr. Marleen Millen for having suggested to Dr. Sieber that a survey of patients’ concerns be conducted.

References

1. Bokolo A. Use of telemedicine and virtual care for remote treatment in response to COVID-19 pandemic. J Med Syst. 2020;44(7):132.

2. Ansary AM, Martinez JN, Scott JD. The virtual physical exam in the 21st century. J Telemed Telecare. 2019. doi: 10.1177/1357633X19878330. Online ahead of print.

3. Lange SJ, Ritchey MD, Goodman AB, et al. Potential indirect effects of the COVID-19 pandemic on use of emergency departments for acute life-threatening conditions — United States, January–May 2020. Morb Mortal Wkly Rep. 2020;69(25):795-799.

4. Woolf SH, Chapman DA, Sabo RT, Weinberger DM, Hill L. Excess deaths from COVID-19 and other causes, March-April 2020. JAMA. 2020;324:510-513.

5. Kichloo A, Albosta M, Dettloff K, et al. Telemedicine, the current COVID-19 pandemic and the future: a narrative review and perspectives moving forward in the USA. Fam Med Community Health. 2020; Aug 8(3):e000530; doi: 10.1136/fmch-2020-000530.

6. Murphy K, Koski-Vacirca R, Sharfstein J. Resilience in health care financing. JAMA. 2020;324:126-127.

7. Diaz A, Sarac BA, Schoenbrunner AR, Janis JE, Pawlik T. Elective surgery in the time of COVID-19. Am J Surg. 2020;219:900-902.

8. Wetherell JL, Kaplan RM, Kallenberg G, Dresselhaus TR, Sieber WJ, Lang AJ. Mental health treatment preferences of older and younger primary care patients. Int J Psychiatry Med. 2004;34:219-233.

Appendix A. Cover Letter and Survey Items

Dear Valued Patients,

We hope you are faring well under the stresses and challenges of the COVID pandemic. We are continuing to work hard to make our UCSD Health system function at the top of its game to serve you, your families, and our San Diego community. In Primary Care (Family Medicine, General Internal Medicine and Geriatrics) we have fully adopted video and telephone visit technology to provide care while keeping everyone at a safe distance.

This has allowed us to stay in touch with you and serve your needs, but increasingly we are recognizing the limits of video and telephone medicine and are becoming concerned about your needs that we may be missing by not seeing you in person. Thus, we are now turning our attention to planning for opening up our ambulatory primary care clinics to more in-person, face-to-face visits with our patients.

We feel that our clinics are already low-risk environments for COVID. We are continuing to refer our patients with potential COVID symptoms to other areas of UCSD Health (Urgent Care and Emergency Dept.). Our staff are all following required use of masks and we have developed distancing protocols for our waiting areas and cleaning protocols for our exam rooms between patients. Screening of our healthcare workers in other areas at UCSD Health at significantly greater risk than our offices have yielded near zero % of positive testing in those without symptoms.

However, we want to hear more about YOUR perceptions of the safety of coming in to see us in our Primary Care offices and what YOU are still concerned about. In this way we can work to develop policies and procedures to allay those concerns. We are asking you to fill out the brief survey attached to the link below to let us know your thoughts and concerns. It will only take you about 5 minutes to complete.

Thank you for helping us to bring back the most important element of continuity in primary care—our personal interaction with our patients.

Please click on the URL below which will take you to the brief survey we would like you to fill out.

We would appreciate your response by June 5, 2020.