What Is Known
Before the COVID-19 pandemic , telemedicine was available but not widely utilized. During the COVID-19 pandemic , telemedicine became a staple part of health care, and there is a need for more in-depth study of the implications of telemedicine in physical medicine and rehabilitation.
What Is New
Specific patient populations in need of targeted study of telemedicine ’s utilization have been identified, including patients of color. In addition, the need for investigation into prescribing behavior, specifically centered around opioids, and a need to understand why emergency care usage was the same between telemedicine and office visit users have been highlighted.
In general terms, Moore’s Law states that technology will double in capacity and half in size every 2 yrs.1 telemedicine has become an accessible form of care. Here, telemedicine is defined as providing health care, including the diagnosis, prevention, and treatment of disease and injury, via video visit.2 telemedicine , until recently. Before the COVID-19 pandemic , the lack of rapid adoption of telemedicine centered on concerns for security and lack of coverage from most major insurers.3,4 telemedicine was particularly true in physical medicine and rehabilitation (PM&R), enhanced by the perceived importance of evaluating physical function during examinations, impairment status of some patient populations, language barriers, and technological difficulties.3
The COVID-19 pandemic presented a unique set of circumstances: owing to public health concerns, the Centers for Disease Control and Prevention strongly encouraged in-person activities be minimized, and the United States government mandated telehealth benefits be expanded via Medicare.5 telemedicine and subsequent switching of large numbers of patients over to telemedicine . One primary limitation of the existing literature on telemedicine is the lack of large cohorts in original research studies evaluating the effectiveness of telemedicine while adequately controlling for confounding variables.6 telemedicine during the pandemic has provided an opportunity to review telemedicine ’s implications in PM&R more closely.
In this study, the authors first aimed to describe demographic differences in patients who used telemedicine during the COVID-19 pandemic . Second, the authors aimed to answer the following question: What are the differences in clinical utilization between patients who used telemedicine and those who used office visits during the COVID-19 pandemic ? In this case, clinical utilization is used as a partial measure of disease management and treatment efficacy. It was hypothesized that emergency department and urgent care visits would be lower in patients who used telemedicine in comparison to patients who did not. It was also hypothesized that the number of drug prescriptions and bedded hospitalizations would be higher in patients who used telemedicine in comparison to patients who did not. The goal in testing these hypotheses is to (1) describe patient populations who used telemedicine and office visits in PM&R during the COVID-19 pandemic and (2) quantify the differences in clinical utilization between these groups. The implications of these findings are to identify areas in need of further study.
METHODS
A retrospective chart review was conducted on PM&R patients seen before and during the COVID-19 pandemic at the University of Pittsburgh Medical Center (UPMC) Department of PM&R. The University of Pittsburgh’s Institutional Review Board (STUDY20060306) approved this study and waived the need for written informed consent. This study adheres to the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines7 https://links.lww.com/PHM/B648
Patient data were collected from the UPMC PM&R Clinical Analytics Application (PM&R Clinical App). The PM&R Clinical App is a database used by UPMC healthcare providers to summarize the state of the patient population and compiles data from the electronic health record. This database is limited by information collected within the UPMC Healthcare Network. Upon exportation of patient data, an honest broker removed all identifying information. Patients were included if they were seen for a musculoskeletal complaint in the outpatient setting, as defined as relating to the skeleton, surrounding muscle mass, and/or associated joint space (see Supplemental Digital Content 2, https://links.lww.com/PHM/B649 pandemic ) and April 1 to June 30, 2020 (COVID-19 pandemic ). This specific period was chosen because there were no operational changes during this time to the medical center’s workflow. Because the authors were interested in focusing on novel telemedicine use during the pandemic , patients were excluded from the COVID-19 pandemic group if they had used telemedicine prepandemic. Prepandemic was defined as before March 20, 2020, because this date is when operational changes were observed in the authors’ healthcare system.
Within the pre-COVID-19 pandemic and COVID-19 pandemic groups, patients were further subdivided into three groups: telemedicine , office visits, and cancellation. Patients were identified as being members of these groups by the Encounter Type and Status labels on the PM&R Clinical App. Patients who did not have one of these labels on their encounter were excluded. Patients with the encounter type “telemedicine ” included patients who had a video visit with their provider. Telephone visits were not included in this group. Patients who had both office visits and telemedicine visits during 2020 were placed into separate groups based on which visit was first (i.e., Groups Telemedicine First 2020 and Office Visit First 2020). Diagnoses were categorized by two of the authors (G.G. and N.J.) independently based on the area of the body, and categorizes with associated ICD codes are shown in Supplemental Digital Content 2 (https://links.lww.com/PHM/B649 Figure 1 .
FIGURE 1: Workflow of patient inclusion and exclusion criteria. Patients were excluded at A owing to being younger than 18 yrs, being seen for a non-musculoskeletal-related complaint, and/or having a visit that was not labeled as telemedicine , office visit, or cancellation. Patients were excluded at B and C owing to using telemedicine before the study period.
Insurance type was categorized as Medicaid, Medicare, commercial, and self-pay/other. Medicaid is a joint federal and state program in the United States that covers health insurance for people with disabilities, families who are low-income, individuals receiving Supplemental Security Incomes, and other qualified individuals.8 9
Clinical utilizations investigated here included emergency department visits, urgent care visits, bedded hospitalizations, mortality, and prescription medications. Visits were recorded only if they were within the electronic health record in the UPMC healthcare system. Clinical utilization was evaluated at 7, 30, 90, and 180 days after the initial appointment. Of note, during the study period, appropriate personal protective equipment was available, and the medical center was able to maintain bed availability and capacity.
Because individual provider attribution was undeterminable in the PM&R Clinical App, prescription medications were split into two groups: Medications Commonly Prescribed by Nonphysiatrists and Medications Commonly Prescribed by Physiatrists. The Medications Commonly Prescribed by Nonphysiatrists included diuretics, immunomodulators, aspirin, metformin, and statins. The Medications Commonly Prescribed by Physiatrists included antidepressants, muscle relaxants, nonsteroidal anti-inflammatory drugs, opioids, and prednisone.
Statistical Analysis
Statistical analyses were performed in IBM SPSS Statistics 26 Software. Descriptive analyses included evaluation of age, sex, race, provider type, comorbidities, type of insurance, diagnosis, number of visits, prescription medications, and clinical utilization. Given the lack of telemedicine usage in 2019, no statistical analyses were performed comparing groups in 2019. The Office Visit First 2020 group was excluded from statistical analysis because of the low number of patients in this group (n = 32). Thus, four groups were compared for demographic variables (Cancellation, Telemedicine Only, Office Visit Only, and Telemedicine First groups from the pandemic time period).
Given that there was no further information on the Cancellation group, three groups were compared for the evaluation of encounter descriptors, comorbidities, diagnosis, and prescription medication (telemedicine only, office visit only, and telemedicine first groups from the pandemic time period). Number of visits and clinical utilization were compared between the Telemedicine Only and Office Visit Only groups. Shapiro-Wilk tests for normality revealed that all variables were normally distributed. Age was evaluated using analysis of variance, and all other categorical variables were evaluated using chi-square tests. An alpha level was set to 0.05 a priori. A post hoc power analysis revealed that a power of 0.972 was achieved with the sample size of 2267. Because of the small number of patients who had clinical utilizations during the study period, matching and controlling for confounding variables were not able to be performed.
RESULTS
Results include descriptions of observed differences for variables that were statistically significantly different.
Demographics
In the 2019 study period, 2384 patients had office visits, 1 patient had a telemedicine visit, and 1186 patients had cancellations. In the 2020 study period, 849 patients were included in the Telemedicine Only group, 924 patients were included in the Office Visits Only group, and 999 patients were included in the Cancellation group. There were 215 patients who had a telemedicine visit first and then an office visit (Telemedicine First group), and 32 patients who had an office visit first and then a telemedicine visit (Office Visit First group).
Table 1 includes descriptors for age, sex, race, and provider type for the Telemedicine Only, Office Visit Only, Telemedicine First, and Cancellation 2020 groups, and Supplemental Digital Content 3 (https://links.lww.com/PHM/B649 Telemedicine Only, Office Visit Only, Telemedicine First, and Cancellation groups. Both telemedicine groups and the cancellation group contained significantly more patients of color than the Office Visit Only group. Interestingly, when examining return status of the patients, the Office Visit Only and Telemedicine First groups had statistically significantly more new patients than the Telemedicine Only and Cancellation groups.
TABLE 1 -
Demographic descriptors for age, sex, race, and provider type for the Office Visit Only 2020,
Telemedicine Only 2020,
Telemedicine First 2020, and Cancellation groups
COVID-19 Pandemic
Demographic Variables
Office Visits Only
Telemedicine OnlyTelemedicine FirstCancellations
Statistical Value
a
P
n
924
849
215
999
–
–
Age, mean ± SD, years
59 ± 71
60 ± 66
59 ± 65
61 ± 68
0.141
0.936
Sex (% female)
62.9
63
60.8
64
0.798
0.850
Race (% of patients)
b
417.410
0.000
White
88.3
83.4
85.6
86.7
Black
9.6
12.7
9.9
9.9
Asian
0.9
1.2
1.8
1.4
American Indian
0
0.4
0
0.2
Provider type (% of patients)
b
218.023
0.000
Physician
90.5
67.9
63.5
85.6
Physician assistant
9
26.2
32.9
12.9
Certified registered nurse
0.3
5.8
3.2
1.5
When comparing groups, there were statistically significant differences in race, and provider type.
a F for age, X 2
b
Encounter Descriptors
The Office Visit Only group was more likely to see a physician, whereas the telemedicine groups had a higher percentage of patients who were seen by a physician assistant. When examining the provider type, a chi-square test showed the distribution in provider type was statistically significantly different between groups. Table 2 includes insurance type (Medicare, Medicaid, commercial, self-pay/other) for the Office Visits Only, Telemedicine Only, and Telemedicine First groups, and Supplemental Digital Content 4 (https://links.lww.com/PHM/B649 telemedicine groups used Medicare. There was a statistically significant different distribution in insurance type between the telemedicine and office visit groups.
TABLE 2 -
Demographic descriptors for insurance type for the Office Visit Only 2020,
Telemedicine Only 2020, and
Telemedicine First 2020 groups
COVID-19 Pandemic
Demographic Variables
Office Visits Only
Telemedicine OnlyTelemedicine First
X
2
P
Type of insurance (% of patients)
a
15.887
0.001
Medicare
28.8
34.4
35.6
Medicaid
10.2
13.7
13.5
Commercial
56.3
47.4
47.3
Self-pay/other
4.5
4.3
3.6
When comparing groups, there were statistically significant differences in type of insurance.
a
Comorbidities
Table 3 contains comorbidities for the Office Visits Only, Telemedicine Only, and Telemedicine First groups, and Supplemental Digital Content 4 (https://links.lww.com/PHM/B649 telemedicine groups contained a higher percentage of patients with congestive heart failure, chronic obstructive pulmonary disease, diabetes, and hyperlipidemia than the Office Visit Only group. Of note, the Telemedicine First group had higher proportions of patients with chronic kidney disease, asthma, and hypertension than the Telemedicine Only and Office Visit Only groups. There were statistically significant differences in the percentage of patients with congestive heart failure, chronic obstructive pulmonary disease, diabetes, hyperlipidemia, chronic kidney disease, asthma, and hypertension between the Telemedicine Only, Telemedicine First, and Office Visit Only groups. There were no significant differences in the prevalence of atrial fibrillation, coronary artery disease, cancer, depression, and osteoporosis between groups.
TABLE 3 -
Demographic descriptors for comorbidities and insurance type for the Office Visit Only 2020,
Telemedicine Only 2020, and
Telemedicine First 2020 groups
Demographic Variables
COVID-19 Pandemic
X
2
P
Office Visits Only
Telemedicine OnlyTelemedicine First
Comorbidities (% of patients)
Atrial fibrillation (Afib)
2.6
3.2
3.6
0.920
0.631
Asthma
a
10.9
15.3
18
10.909
0.004
Coronary artery disease (CAD)
4.9
6.7
6.3
2.274
0.321
Cancer
6.8
7.5
5.9
0.677
0.713
Congestive heart failure (CHF)
a
2.2
4.1
4.1
5.960
0.051
Chronic kidney disease (CKD)
a
1.4
3.3
3.6
7.440
0.024
Chronic obstructive pulmonary disease (COPD)
a
4.5
7.8
7.3
8.144
0.017
Depression
12.4
13.7
17.6
3.390
0.184
Diabetes
a
8.4
12.6
12.6
8.178
0.017
Hypertension (HTN)
a
26.6
31.7
34.5
7.914
0.019
Hyperlipidemia
a
25.4
30.1
23.4
5.891
0.053
Osteoporosis
4.5
5.9
4.1
2.442
0.295
When comparing groups, there were statistically significant differences in asthma, CHF, CKD, COPD, diabetes, HTN, and hyperlipidemia.
a
Diagnosis
Table 4 contains the diagnostic category for the Office Visits Only, Telemedicine Only, and Telemedicine First groups, and Supplemental Digital Content 4 (https://links.lww.com/PHM/B649 telemedicine groups. Systemic diagnoses were defined as diseases whose pathogenesis affects the whole body and is not specific to one area (e.g., diabetes). The Telemedicine First group had higher proportions of patients with complaints involving the lumbar spine, pelvis/sacrum/hip, lower limb, upper limb, shoulder/scapula, medication management, multiple body parts, and multiple visits for different reasons than the Office Visit Only and Telemedicine Only groups. There were statistically significant differences in the percentage of patients with systemic, lumbar spine, pelvis/sacrum/hip, lower limb, upper limb, shoulder/scapula, medication management, multiple body parts, and multiple visits for different reasons between the Telemedicine First, Telemedicine Only, and Office Visit Only groups. There was no statistically significant difference in cervical spine/neck, thoracic spine/chest, and amputation-related diagnoses between groups.
TABLE 4 -
Demographic descriptors for diagnostic category for the Office Visit Only 2020,
Telemedicine Only 2020, and
Telemedicine First 2020 groups
Demographic Variables
COVID-19 Pandemic
X
2
P
Office Visits Only
Telemedicine OnlyTelemedicine First
Diagnostic category (% of patients)
Systemic
a
18.8
12.3
11.7
18.432
0.000
Cervical Spine/neck
13.9
15.8
18.5
2.767
0.251
Thoracic spine/thorax/abdomen
3.7
5.3
5.9
3.733
0.155
Lumbar spine
a
24
45
54
121.407
0.000
Pelvis/sacrum/hip
a
25.9
18.5
31.1
21.517
0.000
Lower limb
a
20.1
16.1
25.2
12.306
0.002
Upper limb
a
5.5
4.3
9.5
8.189
0.017
Shoulder/scapula
a
7.8
7.5
13.9
11.173
0.004
Multiple body parts
a
20.2
36.3
44.6
81.110
0.000
Amputation
3.2
4.1
3.2
1.065
0.587
Medication management
0.4
13.1
16.7
130.267
0.000
Multiple visits for different reasons
a
0.9
0.6
7.2
60.270
0.000
When comparing groups, there were statistically significant differences in systemic, lumbar spine, pelvic/hip/sacrum, lower limb, upper limb, shoulder/scapula, multiple body parts, and multiple visits for different reasons.
a
When looking specifically at the patients who had multiple body parts, there was a statistically significant difference in the number of body parts with complaints within groups. Specifically, the Telemedicine First group (n = 99) had significantly more body parts affected than the Office Visit Only group (n = 188) and the Telemedicine Only group (n = 309). Table 5 contains the number of body parts for the Office Visits Only, Telemedicine Only, and Telemedicine First groups.
TABLE 5 -
Demographic descriptors for number of body parts in the Multiple Body Parts Group and number of visits for Office Visit Only 2020,
Telemedicine Only 2020, and
Telemedicine First 2020 groups
Demographic Variables
COVID-19 Pandemic
X
2
P
Office Visits Only
Telemedicine OnlyTelemedicine First
No. of body parts in the Multiple Body Parts group
a
91.185
0.000
Two
89.3
91.6
69.7
Three
9.6
5.8
17.2
Four or more
2.1
3.6
16.2
No. of visits (% of patients)
a
89.438
0.001
One
86.9
75.5
–
Two
10.3
11.9
59.9
Three
2.5
10.8
26.3
Four or more
0.2
1.9
13.9
When comparing groups, there were statistically significant differences in number of body parts and number of visits.
a
Number of Visits
Table 5 contains the number of visits for the Office Visits Only, Telemedicine Only, and Telemedicine First groups, and Supplemental Digital Content 4 (https://links.lww.com/PHM/B649 Telemedicine Only and the Office Visit Only groups were compared. The Telemedicine Only group had statistically significantly more visits than the Office Visit Only group.
Prescription Medications
Table 6 and Supplemental Digital Content 5 (https://links.lww.com/PHM/B649 telemedicine groups were more likely to be taking diuretics, immunomodulators, metformin, prednisone, opioids, muscle relaxants, antidepressants, and statins than the Office Visit Only group. There were statistically significant differences in the percentage of patients taking diuretics, immunomodulators, metformin, prednisone, opioids, muscle relaxants, antidepressants, and statins between the Telemedicine Only, Telemedicine First, and Office Visit Only groups. There was no significant difference in nonsteroidal anti-inflammatory drug and aspirin prescriptions between groups.
TABLE 6 -
Percentage of patients taking prescription drugs for Office Visit Only 2020,
Telemedicine Only 2020, and
Telemedicine First 2020 groups
Prescriptions (% of Patients)
Telemedicine OnlyOffice Visit Only
Telemedicine First
X
2
P
Medications commonly prescribed by nonphysiatrists
Diuretics
a
19.2
15.4
19.8
8.394
0.015
Immunomodulators
a
2.9
1.2
1.4
10.926
0.004
Metformin
a
11.5
8.4
10.4
8.697
0.013
Aspirin
25.2
23.3
21.2
4.850
0.088
Statins
a
34.5
30.2
29.3
4.403
0.011
Medications commonly prescribed by physiatrists
Prednisone
a
10.3
8.3
10.8
6.209
0.045
Opioids
a
43.4
24.4
50.5
74.506
0.000
NSAIDs
52.3
52
55.4
4.246
0.120
Muscle relaxants
a
38.1
31.7
33.3
11.795
0.003
Antidepressants
a
43
37.1
48.6
9.551
0.008
When comparing groups, there were statistically significant differences in diuretics, immunomodulators, metformin, statins, prednisone, opioids, muscle relaxants, and antidepressants.
a
NSAIDs, nonsteroidal anti-inflammatory drugs.
Clinical Utilization
Figure 2 shows how clinical utilization changed postappointment over time. To prevent cluttering on the figure, statistical variables are included in Supplemental Digital Content 6 (https://links.lww.com/PHM/B649 Telemedicine Only and Office Visit Only groups (e.g., a patient may have had a telemedicine visit first, then an emergency department visit, and then an office visit, or a telemedicine visit first, then an office visit, and then an emergency department visit). There were no statistically significant differences in emergency department and urgent care usage between groups over all time periods. The Telemedicine Only group had significantly more bedded hospitalizations than the Office Visit Only group at 30-days post-appointment, but there were no statistically significant differences at other time points. There was no statistically significant difference in mortality (X 2 P = 0.726) between the Telemedicine Only (n = 2, 0.235%) and the Office Visit only groups (n = 3, 0.323%).
FIGURE 2: A, Percentage of patients who had a bedded hospitalization after their first appointment. There was a statistically significant difference at 30 days postappointment between the Telemedicine Only 2020 and Office Visit Only 2020 groups, but no statistical differences between these groups at other time periods. B, Percentage of patients who had an urgent care visit after their first appointment. There were no statistically significant differences between the Office Visit Only 2020 and Telemedicine Only 2020 groups. C, Percentage of patients who had an emergency department visit after their first appointment. There were no statistically significant differences between the Telemedicine Only 2020 and Office Visit Only 2020 groups. **Statistically significant difference on graph.
DISCUSSION
With the advent of the COVID-19 pandemic , the usage of telemedicine as a central part of health care has rapidly expanded. This study aimed to describe the differences in demographics between patients who used telemedicine and those who were seen in the office during the pandemic and identify differences in clinical utilization. With these results, the authors are aiming to identify patient populations in need of future research into long-term outcomes associated with the use of telemedicine and identify gaps in the understanding of how telemedicine affects patient care in PM&R.
It was observed that proportionately more people of color used telemedicine and canceled than office visits during the pandemic . This finding is in line with a recent survey conducted during the pandemic 10 11 12 telemedicine ?” In addition, this study further highlights the need to address racial disparities that currently exist in PM&R outcomes.13 telemedicine specifically in this population are of significant priority.
In addition, differences in the percentage of new patients per group were observed. Specifically, the Office Visit Only and Telemedicine First groups had more new patients than the Telemedicine Only and Cancellation groups. This finding is in line with that observed by Bhuva et al.14 Telemedicine First group is that the provider or patient felt that telemedicine would not adequately address these patients’ clinical needs after the encounter, or that a physical examination was needed, and the patient had to come to an office visit. Thus, it is worth future studies targeting telemedicine ’s use for intake versus for follow-up.
Significantly more comorbidities were also observed in the telemedicine groups than the Office Visit Only group. Eberly et al.12 telemedicine groups than office visit groups during the pandemic . These findings were likely driven by the fact that nearly all these comorbidities increased patients’ risk for COVID-19, thus placing a selection bias for the telemedicine groups. Although the authors do not anticipate that this finding would translate into a selection bias during nonpandemic time periods, this finding does place important confounding variables on the evaluation of clinical utilization.
Interestingly, the Telemedicine First group had higher rates of patients with lumbar spine, pelvis/sacrum/hip, lower limb, upper limb, shoulder/scapula, medication management, multiple body parts, and multiple visits for different reasons than the Office Visit Only and Telemedicine Only groups. As mentioned above, it is possible that the provider or patient who was seen in the Telemedicine First group felt that telemedicine would not adequately address these patients’ clinical needs after the encounter, or that a physical examination was needed, and thus had the patient come to an office visit. This finding highlights the need to study the use of telemedicine more specifically in patients with diseases and injuries related to the aforementioned diagnoses. Recently, approaches to remote physical examination techniques have been published, and adoption may modulate this finding.15
Of note, in the Multiple Body Parts sub-grouping, the Telemedicine First group had significantly more body parts involved than the Telemedicine Only and the Office Visit only groups. Although the addition of telemedicine to the care of medically complex children in a randomized control trial reduced the number of adverse of events,16 telemedicine for patients with multiple medical conditions and complaints.
The Telemedicine Only group had proportionally more visits than the Office Visit Only group, and this finding is in line with previous literature.17 telemedicine visit that require more frequent follow-ups or this finding is a result of the ease of scheduling and attending a telemedicine appointment versus an office visit appointment.
Of note, the telemedicine groups were more likely to see a physician assistant than a physician. In the United States, physician assistants diagnose illness and manage treatment plans under the supervision of a licensed physician. They can independently prescribe medications but cannot advertise or bill their services independent of a supervising physician. Thus, it is possible that the physician assistant identified issues during the telemedicine visit that then prompted a visit with a physician.18
In addition, it was observed that more patients were in the Telemedicine First group (n = 215) than in the Office Visit First group (n = 32). It is unclear why this distribution is occurring, but in future studies, it is worth further delving into these findings to discern why these visit patterns are occurring and how this is affecting long term outcomes.
The telemedicine groups were more likely to be on prescription medications than the Office Visit Only group. This trend did not seem to be unique to prescription medications commonly prescribed by physiatrists, as rates were also higher in prescription medications commonly prescribed by nonphysiatrists. However, although the telemedicine groups did have higher rates of comorbidities than the office visit groups, the rates of opioid use (telemedicine : 43.4% vs. office visit: 25.7%) are more disparate than the comorbidities observed here (average telemedicine : 11.8% vs. average office visit: 9.3%). This suggests that something else is driving the higher rates of prescription drug use. As highlighted below in the limitations, the authors were not able to discern whether the prescriptions in this study are new or already existing, thus limiting the conclusions that can be drawn from these data. Previous studies have, however, raised concerns that telemedicine changes physician prescribing behavior,19 telemedicine , specifically with controlled substances such as opioids.
Interestingly, there were generally no differences in clinical utilization of emergency services between the telemedicine and office visit groups. This finding was of interest, given that the telemedicine groups contained higher rates of comorbidities than the Office Visit only group. Despite these higher rates, these findings suggest that patients who use telemedicine are not more likely to use emergency clinical support than patients who use office visits. This finding suggests that further study is needed to understand long-term health outcomes and to determine whether this finding was related to avoidance of care altogether during the pandemic .
Although this study does add to an important, growing body of literature regarding the application of telemedicine in PM&R, it does exist with limitations. First, this is a retrospective analysis of data collected during a global pandemic , and applications of these findings to nonpandemic times are unclear. This study included only patients with musculoskeletal-related complaints and generalizability is limited to this specific patient population. The PM&R Clinical App contains information only from UPMC clinical sites; thus, if a patient was seen in an emergency department outside of this network, it would not be reported in the database. Although this database does contain information on the nature of each diagnosis, it does not contain key descriptors of the patients’ pain duration, severity, and intensity. In addition, this database does not report the reason for the clinical utilization visit, the exact date of this visit, or the return status of the patient (new patient versus follow-up visit), thus limiting some of the analyses and conclusions that can be drawn from this dataset. The PM&R Clinical App also does not contain information regarding education status, occupation level, need for inpatient rehabilitation, convenience of transportation, impairment status, disease severity, or habitual technology usage. These factors may influence the choice to use telemedicine but were not controlled or accounted for in this study. Because so few patients had clinical utilizations, the authors were not able to control for confounding variables, which include age, race, disease, pandemic , and comorbidities. Given that this study was conducted at a single academic institution, generalizability to other locations should be completed cautiously. Importantly, the database does not allow discerning whether the prescription medications evaluated were newly prescribed at the visits evaluated or whether the patients were already on these prescription drugs. Lastly, although this database does contain information about prescription usage, data on the rates of other interventions in this patient population, such as injections or physical therapy, were unavailable, and this is an important area for future study.
CONCLUSIONS
With telemedicine becoming a substantial part of clinical care during the COVID-19 pandemic , it is important for more targeted, prospective trials to understand the implications of telemedicine in PM&R. The findings of this study can help guide more targeted studies that may ultimately be useful for informing health policy, both during the pandemic and beyond. This study was geared at identifying gaps in the understanding of telemedicine in PM&R and identifying patient populations that need future study. Given the general lack of differences in clinical utilization between patients in the telemedicine groups and the office visit group, there is a need for future studies to understand why this trend was observed and to have an overall understanding of long-term outcomes of telemedicine in PM&R. This study’s findings suggest that it is worth further studying telemedicine in patients of color to discern barriers and accessibility. This study’s findings also highlight the need to understand why patients with lumbar spine, pelvis/sacrum/hip, lower limb, upper limb, shoulder/scapula, medication management, multiple body parts, and multiple visits for different reasons diagnoses were seen in an office visit after their telemedicine visit. Further studies are also needed to understand why the telemedicine groups had more visits than the office visit group and whether this is a sign of better or worse clinical efficiency and care. Importantly, given the higher opioid prescription rates in the telemedicine groups compared with the office visit group, there are urgent needs to understand how prescribing behavior changes with telemedicine . Lastly, given that this study focused on patients with musculoskeletal related complaints, there is a need to study other interventions, including physical therapy and usage of rehabilitation-based hospitals.
