The use of telemedicine for perioperative pain management during the COVID-19 pandemic

Abstract

Introduction

Using a human factors engineering approach, the Johns Hopkins Personalized Pain Program adopted telemedicine for perioperative pain management in response to the COVID-19 pandemic. This study examines the impact of telemedicine adoption on the quality and outcomes of perioperative pain management.

Methods

A mixed-methods study with a convergent parallel design was conducted. From June 2017 to December 2021, 902 patients participated in the Personalized Pain Program. Quantitative data on daily opioid consumption, pain severity and interference, physical and mental health status, and patient satisfaction and engagement were continuously collected with all patients using chart review and patient surveys. Beginning 23 March 2020, the Personalized Pain Program transitioned to telemedicine. A pre–post quasi-experimental design was used to examine the impact of telemedicine. In addition, qualitative interviews were conducted with 3 clinicians and 17 patients to explore their experience with telemedicine visits.

Results

The monthly number of new patients seen in the Personalized Pain Program did not significantly change before and after telemedicine adoption. Compared to patients having in-person visits before the pandemic, patients having telemedicine visits during the pandemic achieved comparable improvements in daily opioid consumption, pain severity and interference, and physical health status. While telemedicine helped overcome many challenges faced by the patients, the limitations of telemedicine were also discussed.

Conclusion

The COVID-19 pandemic stimulated the use of telemedicine. To facilitate telemedicine adoption beyond the pandemic, future research is needed to examine best practices for telemedicine adoption and provide additional evidence on the effectiveness of telemedicine.

Introduction

The opioid crisis persists in the United States (US) as overdose deaths escalate nationwide.¹ Surgery increases the risk of pain and opioid exposure.^2–6 Surgeons account for one-third of opioid prescriptions.⁷ Over 80% of surgical patients receive at least one opioid prescription after surgery, and more than 10% of surgical patients administered opioids experience opioid-related adverse drug events, resulting in higher rates of mortality and increased healthcare costs.^8,9

The COVID-19 pandemic, while crippling the US economic and healthcare systems and threatening millions of American lives, has created significant challenges to perioperative pain management for surgical patients.¹⁰ Efforts to prevent the spread of the virus have disrupted perioperative care (e.g. postoperative follow-up, medication maintenance) and limited access to social supports and medical resources (e.g. physical therapy, mental health service), which can manifest as increased rates of emergency department visits, hospital admissions, and drug overdoses.¹¹ Healthcare systems, therefore, must adapt to continuously provide needed services for surgical patients during the pandemic.

Telemedicine allowing synchronous and/or asynchronous communication between patients and clinicians can potentially address the challenges to perioperative pain management exacerbated by the pandemic.^12,13 Nascent literature has shown increased acceptability and utilization of telemedicine.^14–17 The benefits of telemedicine include decreased travel time and cost, fewer missed medical visits, and improved inter-clinician communication.^18,19 In addition, the quality of care has not been shown to be compromised with telemedicine.²⁰ However, similar to other health information technology, the application of telemedicine is neither straightforward nor spontaneous and can be influenced by various factors (e.g. technology infrastructure, capability and limitations of patients and clinicians, and reimbursement).^21,22

In response to the COVID-19 pandemic, the Johns Hopkins (JH) Personalized Pain Program (PPP) transitioned to telemedicine using a human factors engineering (HFE) approach.²³ In this article, we described the HFE approach to telemedicine adoption and assessed the impact of telemedicine adoption on patient outcomes (e.g. daily opioid consumption, pain severity and interference, physical and mental health status, and patient satisfaction and engagement).

Method

Context

In 2017, the JH PPP was developed to address an unmet need in pain management coordination for chronic opioid users undergoing surgery. The PPP consists of anesthesiologists and psychiatrists who manage patients’ pain while minimizing opioid exposure across the perioperative periods.²⁴ Preliminary data showed that PPP patients reduced their daily opioid consumption with improved pain and functioning.²⁵

Based within a tertiary academic medical center, the PPP continued receiving referrals throughout the COVID-19 pandemic. To continue providing specialized pain care, the PPP transitioned to telemedicine. Beginning 23 March 2020, 1 week after the US Secretary of Health and Human Services modified telemedicine requirements,²⁶ in-person PPP visits were restricted; most patients were seen through telemedicine.

Initially, telemedicine visits were conducted through Polycom Telecommunications, the default web-based telemedicine platform built into Epic. Patients and providers had to download the Polycom app and have access to cameras and microphones for communication. In August 2020, the PPP transitioned to Cisco Webex, a browser-based video platform that does not require a download, is compatible with most major internet browsers, and can be used on a computer, tablet, or smartphone. Patients and clinicians could click a link within Epic to join the video appointment. In case of poor or no connection, the provider would either continue video visits using the Doximity app or Zoom technologies or call patients from their private phones using the Doximity dialer or a blocked phone number.

An HFE approach to telemedicine adoption

An HFE approach guided by the Systems Engineering Initiative for Patient Safety (SEIPS) 2.0 model²⁷ was applied to facilitate the PPP's transition to telemedicine.²³ According to the SEIPS 2.0 model, telemedicine is a technology in the work system, which interacts with the other work system elements, including people (e.g. patients, clinicians), tasks (e.g. consultation, prescription), other tools and technologies (e.g. prescription drug monitoring program (PDMP), electronic medical records (EMR)), physical environment (e.g. clinics, patient home), and organizational conditions (e.g. policy, staffing). The entire work system needed to be optimized to facilitate telemedicine adoption.

Prior to the pandemic, PPP clinicians and the JH HFE team collaborated on several projects to improve the quality and safety of perioperative pain management. Over 80 observations of in-person PPP visits and 40 interviews with PPP patients and clinicians had been conducted to understand the clinical workflow.²⁸ Based on these findings, the team worked with different stakeholders (e.g. leadership, IT experts, clinicians, and patients) to proactively identify challenges to telemedicine adoption and adapt the PPP work system (Table 1).²⁹

Table 1.

Adaptation of the Personalized Pain Program (PPP) work system in response to the COVID-19 pandemic.

Work system elements		Adaptation
People	Clinicians	Authorized to electronically prescribe controlled substances
	Coordinator	Trained in establishing and troubleshooting telemedicine platform with patients
	Patients	Receiving instructions on how to install and launch telemedicine platform through email prior to health visit
Tasks	Before visit	Coordinator obtains patient consent to telemedicine visit Coordinator reschedules in-person visit to telemedicine visit Coordinator confirms patient access to telemedicine platform Coordinator emails PPP surveys to patients
	During visit	Clinician confirms patient identity and contact information (to call patient in case of technology failure) Clinician obtains medical history and completes modified physical exam Clinician electronically prescribes medications (including controlled substances)
	After visit	Patient receives summary in EMR Coordinator schedules follow-up visit and serves as liaison if the patient has follow-up questions
Tools and technologies		(Non-)HIPAA-compliant platform for telemedicine visit Access to PDMP Access to EMR
Physical environment		Patient and clinician not required to go to the clinic
Organization		Guidance on using (non-)HIPAA-compliant platforms Guidance on scheduling appointments for out-of-state patients

EMR: electronic medical records; PDMP: prescription drug monitoring program.

Evaluation of the impact of telemedicine

A mixed-methods study with a convergent parallel design³⁰ was conducted. Quantitative data were continuously collected by chart review and patient surveys. A pre–post quasi-experimental design was used to examine the impact of telemedicine. In addition, qualitative interviews were conducted with PPP patients and clinicians to explore their experience with telemedicine visits. Quantitative and qualitative data were mixed during data interpretation. The Johns Hopkins Medicine Institutional Review Board approved this study.

Data collection

Chart review

During each PPP visit, a clinician reviewed the patient's opioid prescriptions from the EMR and PDMP and did a pill count when available. Average daily opioid consumption was converted to morphine milligram equivalent (MME) using an opioid calculator app and recorded in the EMR.²⁵ We reviewed the EMR of PPP patients and manually extracted and verified their MME from each PPP visit. We also collected demographic and clinical information of PPP patients at the initial clinic visit and/or from the EMR.

Patient surveys

As a routine practice, all PPP patients completed the Brief Pain Inventory (BPI)³¹ and the 12-Item Short-Form Health Survey (SF-12 v2)³²/RAND 36-Item Health Survey (SF-36 v2)³³ prior to clinic visits. The BPI, a well-validated 13-item survey measuring pain severity and interference, were administered through Qualtrics until mid-September 2019, then REDCap. PPP patients were emailed a link to the survey one day before their clinic visit. They could complete the survey in advance or on the day of their PPP appointment.

The SF-12 v2 with proven validity and reliability was used to measure physical and mental health status from June 2017 through March 2020. The PPP had a license with Optum (Eden Prairie, MN) to administer the SF-12 v2 via the website, https://www.amihealthy.com/. Prior to the pandemic, patients accessed the website on an iPad provided during their clinic visits. Since patients could no longer access the website outside of clinic visits as of March 2020, the PPP administered the SF-36 v2 via emailed survey through REDCap.

Starting 22 August 2018, two additional questions were added to the survey to assess patient satisfaction with the PPP (How satisfied are you with the Personalized Pain Clinic?) and patient perceptions regarding their engagement (To what extent are you engaged in your perioperative pain management?). Both questions used a 5-point Likert-type scale.

Clinician and patient interviews

Phone interviews were conducted with a convenience sample of PPP patients and clinicians from July to October 2020. The sample size was determined by the theoretical saturation concept.³⁴ To facilitate the interviews, a semi-structured interview guide was developed, including questions on impacts of the COVID-19 pandemic on perioperative pain management, differences between telemedicine and in-person, advantages and disadvantages of telemedicine visits, and suggestions for improving telemedicine visits. All interviews were conducted by a team of two researchers and audio-recorded and transcribed for analysis.

Data analysis

Quantitative analysis

The analysis included adult patients having PPP visits between June 2017 (inception of the PPP) and December 2021. Given the limits on opioid prescriptions (e.g. a maximum 30-day supply per prescription) and the needs for continuous patient assessment and treatment adjustment, most patients visited the PPP every 2–4 weeks. Clinic visits more than 3 months apart were highly likely to be associated with different surgical procedures and referrals. Therefore, we defined episode of visits (EOVs) as a series of consecutive clinic visits with every two adjacent visits no more than 3 months apart. EOVs with only one clinic visit were excluded from the analysis. The remaining EOVs were categorized into three groups: EOVs with clinic visits only before telemedicine adoption (Group 1), EOVs with clinic visits before and after telemedicine adoption (Group 2), and EOVs with clinic visits only after telemedicine adoption (Group 3). Patient and EOV characteristics were described using frequencies with percentages or means with SDs. Chi-square tests/Fisher's exact tests and Wilcoxon rank-sum tests were conducted for group comparisons.

The impact of telemedicine on the monthly number of new patients joining the PPP and different patient outcomes were then assessed. A run chart showing the number of new patients participating in the PPP each month was created; a Wilcoxon rank-sum test was conducted to test if the number significantly changed before and after telemedicine adoption. Stratified by EOV groups, patient outcomes during the first and last PPP visits, as well as changes in these outcomes between the first and last PPP visits, were described using means with SDs. Wilcoxon rank-sum tests were conducted for group comparisons. To evaluate the impact of telemedicine adoption, multiple linear regression models were then conducted to compare changes in each outcome (difference between the last and first PPP visits) across EOV groups with adjustments of other covariates. The models regressed change in each outcome on the group dummy variables with clustered sandwich estimator to address clustering of EOVs from the same patient. The Bayesian information criterion and Akaike's information criterion were used to determine the final model specifications.

Finally, we examined whether the impact of telemedicine varied across patients with different demographic characteristics. Each demographic characteristic was tested separately by adding its interaction terms with the EOV groups to the regression models. The estimated coefficients of the “EOV group by patient subgroup” terms provided the magnitude of possible varying effect of telemedicine on patient subgroups.

Qualitative data analysis

We conducted a qualitative content analysis³⁵ of the interview data. Interview transcripts were reviewed in an iterative process to create a node structure with common themes.³⁶ Each transcript was coded independently by at least two researchers and discussed to reach a consensus. Different approaches (e.g. triangulation, looking for negative cases, detailed description of the context, and skeptical peer review) were used to ensure the credibility, transferability, dependability, and confirmability of the qualitative results.³⁷

Results

Quantitative results

From June 2017 to December 2021, 902 patients participated in the PPP with a total of 963 EOVs. The monthly number of new PPP patients did not change significantly before (mean = 16, SD = 5) and after (mean = 17, SD = 4) telemedicine adoption (Figure 1). Among the 963 EOVs, 334 with only 1 clinic visit were excluded from the analysis. The remaining 629 EOVs (from 569 patients) included 291 Group 1 EOVs, 69 Group 2 EOVs, and 269 Group 3 EOVs. Figure 2 shows the detailed distribution of included patients and their EOVs.

Figure 1.

Monthly number of new patients participating in the Personalized Pain Program (PPP).

Figure 2.

Distribution of included patients and episode of visits (EOVs).

Table 2 shows the demographic and clinical characteristics of patients with the three groups of EOVs. A significant difference was only identified in marital status. Compared to Group 2, patients in Group 3 were less likely to be single (p = 0.001). There was no significant difference in other characteristics across the three groups.

Table 2.

Demographic and clinical characteristics of Personalized Pain Program (PPP) patients with the three groups of episode of visits (EOVs).

	Group 1 N = 291 EOVs		Group 2 N = 69 EOVs		Group 3 N = 269 EOVs		P-value*
	n	%	n	%	n	%
Age							0.121
18–29	45	16	20	29	35	13
30–39	68	23	13	19	58	22
40–49	59	20	16	23	62	23
50–59	65	22	12	17	61	22
60+	54	19	8	12	53	20
Gender							0.527
Female	156	54	35	51	154	57
Male	135	46	34	49	115	43
Race							0.065
Caucasian	171	59	36	52	182	68
African American	102	35	28	41	70	26
Other	18	6	5	7	17	6
Marital status							0.008
Single	122	42	40	58	105	39
Married	120	41	20	29	120	45
Separated/divorced/widowed	43	15	7	10	44	16
Other	6	2	2	3	0	0
Education							0.194
High school or below	125	43	39	57	111	41
College	55	19	13	19	54	20
Professional or doctorate	30	10	7	10	37	14
Not reported	81	28	10	14	67	25
Employment status							0.058
Employed	81	28	19	28	101	38
Unemployed	123	42	31	45	85	31
Disabled	46	16	10	14	46	17
Retired	35	12	5	7	32	12
Other	6	2	4	6	5	2
Insurance							0.219
Private	171	59	45	65	178	66
Public	114	39	21	30	85	32
Self-pay or uninsured	6	2	3	5	6	2
On medications for opioid use disorder							0.117
No	259	89	55	80	233	87
Yes	32	11	14	20	36	13
Surgery type
Cardiac/Thoracic	22	7	2	3	21	8	0.565
Gastroenterology/General	49	17	9	13	46	17
Neurosurgery/Otolaryngology	28	10	7	10	14	5
Orthopedic/Trauma	107	37	31	45	100	37
Plastic/Vascular	16	5	5	7	17	6
No surgery	69	24	15	22	71	27
Surgery timeline							0.434
Surgery before first visit	165	57	41	59	162	60
Surgery after first visit	57	19	13	19	37	14
No surgery	69	24	15	22	70	26

*Using chi-square tests or Fisher’s exact tests for categorical variables.

Table 3 shows the characteristics of the three groups of EOVs. Compared to Groups 1 and 3, patients in Group 2 had significantly longer treatment duration, more PPP visits, and a higher portion of psychiatric visits. Compared to Group 1, patients in Group 3 had significantly longer treatment duration and more PPP visits.

Table 3.

Characteristics of the three groups of episode of visits (EOVs).

	Group 1 N = 291 EVOs		Group 2 N = 69 EOVs		Group 3 N = 269 EOVs		Group 1 versus Group 2	Group 1 versus Group 3	Group 2 versus Group 3
	Mean	SD	Mean	SD	Mean	SD		P-value*
Length of EOVs (days)	104	108	408	323	147	126	<0.001	<0.001	<0.001
Number of total visits	5	4	16	13	7	5	<0.001	<0.001	<0.001
Number of in-person visits	5	4	6	7	–	–
Number of telemedicine visits	–	–	10	8	7	5
	n	%	n	%	n	%		P-value*
Number of visits to psychiatrist							<0.001	0.139	<0.001
0	192	66	29	42	185	69
1	24	8	8	11	21	8
2–5	56	19	10	15	34	12
6–10	12	4	3	4	18	7
≥11	7	3	19	28	11	4

*Using Wilcoxon rank-sum tests for continuous variables and chi-square tests or Fisher's exact tests for categorical variables.

Impact of telemedicine on patient outcomes

Figure 3 shows the outcomes of the three groups of patients during their first and last PPP visits. All three groups achieved significant reduction in daily opioid consumption (Figure 3(a); p's < 0.001), pain severity (Figure 3(b); p < 0.001 for Groups 1 and 3; p = 0.004 for Group 2), and pain interference (Figure 3(c); p's < 0.001). Compared to patients in Group 1, Group 2 had significantly lower pain interference during the last PPP visit (p = 0.027); Group 3 had significantly less opioid use during the first (p = 0.008) and last (p = 0.022) PPP visits, significantly higher reduction in pain severity between the first and last PPP visits (p = 0.004), and significantly lower pain interference during the last PPP visits (p = 0.010). There was no significant difference in daily opioid consumption, pain severity, and pain interference between patients in Groups 2 and 3.

Figure 3.

Impact of telemedicine on patient outcomes. (a) Daily opioid consumption, (b) pain severity, (c) pain interference, (d) physical health status, (e) mental health status, (f) satisfaction, and (g) perceived patient engagement.

Additionally, patients in Group 1 achieved significant improvement in physical health status (Figure 3(d); p < 0.001; measured by SF-12). Patients in Group 3 achieved significant improvement in physical health status (Figure 3(d); p < 0.001; measured by SF-36) but reduction in mental health status (Figure 3(e); p = 0.003; measured by SF-36). No direct comparison of physical and mental health status could be made between patients in Groups 1 and 3, nor between the first and last PPP visits of patients in Group 2 since different instruments were used before and after telemedicine adoption.

Finally, patients in Group 1 reported significant improvement in satisfaction with the PPP (Figure 3(f); p = 0.007) and perceived patient engagement (Figure 3(g); p = 0.002). No change in satisfaction and perceived patient engagement between the first and last PPP visits was observed in patients in Groups 2 and 3. The satisfaction (p < 0.001) and perceived engagement (p = 0.006) of patients in Group 3 during the first PPP visit were significantly higher compared with Group 1.

Telemedicine impact across demographic groups

The impact of telemedicine on satisfaction with the PPP varied by race (Figure 4). Caucasian patients reported less improvement in satisfaction over time with the PPP after telemedicine adoption compared with before adoption. African-American patients reported decreased satisfaction over time with the PPP before telemedicine adoption and improved satisfaction over time with the PPP after telemedicine adoption (interaction term of EOV group and African-American: coefficient = 0.89; SE = 0.34; p = 0.009). There was no significant difference in the impact of telemedicine on other outcomes across demographic groups.

Figure 4.

Impact of telemedicine on changes in patient satisfaction with the Personalized Pain Program (PPP)—difference between Caucasian and African-American patients.

Qualitative results

Twenty interviews were conducted with 17 patients and 3 clinicians having telemedicine visits. Table 4 lists the main themes and illustrative quotes from the interviews.

Table 4.

Patient and clinician experience with telemedicine visits.

	Themes	Quotes
Strengths	Facilitating PPP access for patients with postoperative immobility	Q1: “When I was a fresh postoperative patient, it was very hard [for me to travel].” (Patient 15)
	Facilitating PPP access for patients with limited transportation options	Q2: “[Telemedicine] was good for me, because I have a problem with transportation getting around.” (Patient 2)
	Saving travel time	Q3: “I do enjoy the tele-visit, because it keeps me from driving an hour just for a 15- to 30-min doctor visit.” (Patient 6)
	Avoiding inconvenience of parking at hospital	Q4: “[Telemedicine] is convenient; you don’t have to pay for parking to come down.” (Patient 13)
	Alleviating patient apprehension of exposure to COVID-19	Q5: “Because I am immune-deficient, it would've been a little scary for me to go to the hospital.” (Patient 3)
	Providing same level of care as in-person visits	Q6: “I don't feel as though COVID affected my pain management at all.” (Patient 14)
Weaknesses	Preventing patients from conveying their emotions to clinicians	Q7: “Just basic feelings don’t come across as much in the telemedicine visits.” (Patient 9)
	Technical difficulties with setting up and using telemedicine platform	Q8: “We had trouble getting hooked up [on the telemedicine platform] a couple of times.” (Patient 4)
	Hindering physical examinations	Q9: “[The clinicians] can't really identify certain things from just talking to you. Sometimes, they need to physically see or touch you.” (Patient 15)
Suggestions for improvement	Providing technical support for patients	Q10: “Patients need more assistance on how to use the system.” (Clinician 3)
	Clinicians adapting their approaches for physical examinations	Q11: “I become very creative to keep watching the patient's body language and movement to know how they are doing.” (Clinician 1)
	Integrating telemedicine as a routine practice of PPP	Q12: “Offering tele-medicine as a choice, even after things settle down, would improve patient experience.” (Patient 3)

PPP: Personalized Pain Program.

In general, both PPP clinicians and patients were satisfied with telemedicine visits, which helped overcome challenges faced by PPP patients, including postoperative immobility (Q1), lack of access to transportation (Q2), lengthy travel times (Q3), inconvenience of parking at the hospital (Q4), and apprehension of exposure to COVID-19 (Q5). Telemedicine visits were perceived as effective as in-person visits and mitigated the impact of the pandemic (Q6).

The limitations of telemedicine visits were also discussed. Telemedicine visits were considered less personal and prevented patients from conveying their emotions to the clinicians (Q7). Some patients experienced technical difficulties with setting up and using the telemedicine platform (Q8); additional support for patients was required to facilitate telemedicine use (Q10). Also, physical examinations were limited by telemedicine (Q9); clinicians needed to adapt their approaches for physical examinations (Q11).

Given its strengths and weaknesses, telemedicine service was suggested to be integrated as a routine PPP practice beyond the pandemic (Q12).

Discussion

The JH PPP adopted telemedicine for perioperative pain management in response to the pandemic. Based on an established partnership between PPP clinicians and HFE experts and the pre-pandemic preliminary data on PPP workflow, an HFE approach engaging different stakeholders to proactively identify and address work system challenges to telemedicine adoption was applied to facilitate a seamless and prompt transition of the PPP to telemedicine.

Our quantitative data showed that, using telemedicine, the PPP was able to maintain patient access to and quality of perioperative pain management. Despite decreased numbers of surgical procedures performed both nationally³⁸ and at our own institution during the pandemic, the PPP retained its patient volume through telemedicine. This could be partially explained by our qualitative data, which showed that telemedicine could reduce barriers to accessing the PPP and, therefore, increase the participation rate of eligible patients. Another potential explanation was the high volume of urgent and emergent surgical procedures during the pandemic.

Compared to patients having in-person visits before the pandemic (Group 1), patients having telemedicine visits during the pandemic (Group 3) achieved comparable improvements in daily opioid consumption, pain severity and interference, and physical health status. Some patients with needs for prolonged psychiatric treatment (most patients in Group 2) were able to use telemedicine to continue their care in the PPP. Since telemedicine helped address various patient challenges, patients reported initial and persistent satisfaction and engagement with telemedicine visits. Especially, African-American patients who reported less satisfaction with in-person visits reported greater satisfaction with telemedicine visits. This result was consistent with findings from a recent systematic review of telemedicine use among minorities.³⁹

However, our qualitative data also highlighted some limitations, which may hinder telemedicine's broader application. Telemedicine, while abridging the physical distance between patients and clinicians, may broaden their psychological distance. To mitigate the limitation of telemedicine in conveying personal emotions, effective patient-centered communication skills (e.g. planning participants, opening the visit, elicit patient's perspective, and demonstrate active listening) need to be translated and used in the telemedicine setting.^40,41 Difficulties in technology access (e.g. limited access to computer and broadband internet) and use (e.g. poor usability of telemedicine platform) are other common challenges to telemedicine adoption, which may disproportionately affect certain patient populations (e.g. elderly, racial minorities).^42,43 Additional studies are needed to better understand the impact of telemedicine on disparities in care access, delivery, and outcomes. Finally, our results confirmed the limitation of telemedicine in conducting physical examinations,⁴⁴ which may contribute to the variation in its uptake across specialties.^45,46 Pain care may benefit from adapting virtual physical examinations as previously reported for neurological,⁴⁷ musculoskeletal,^48,49 spine,⁵⁰ and hip⁵¹ examinations. The validity of virtual versus in-person physical examinations, however, still needs to be demonstrated.⁵²

Limitations

This study had several limitations. First, this study was conducted in an outpatient pain clinic at an academic hospital; findings may not be generalizable to other settings. However, the HFE approach can be adapted to facilitate telemedicine adoption in other settings. Second, this study used a pre–post quasi-experimental design, which precludes assessment of the causal relationship between telemedicine adoption and differences in patient outcomes. We attempted to mitigate this limitation by showing that patients with EOVs before and after telemedicine adoption had similar demographic and clinical characteristics and controlling their outcome measures during the first PPP visit in multiple linear regression analyses. Finally, this study examines the adoption of telemedicine during a pandemic. As telemedicine becomes a routine PPP practice, its use and impact will need to be reassessed.

Conclusion

While the idea of telemedicine can be traced back to the early 20th century,⁵³ the COVID-19 pandemic stimulated its adoption in different clinical settings.^54–56 The JH PPP was able to continuously provide a high and consistent level of care for its patients during the pandemic by transitioning to telemedicine using an HFE approach. Future research is needed to examine best practices to facilitate telemedicine adoption and effectiveness beyond the pandemic.