Do Orthopaedic Virtual Clinic Visits Demonstrate Cost and Time Efficiencies Compared With In-person Visits?

Abstract

Background

There are numerous reasons for the increased use of telemedicine in orthopaedic surgery, one of which is the perception that virtual visits are more cost-effective than in-person visits. However, to our knowledge, no studies have compared the cost and time investment of virtual versus in-person visits using the time-driven activity-based costing (TDABC) method. Unlike methods that estimate cost based on charges for services rendered, TDABC provides a more precise measurement of costs, which is essential for assessing cost-effective innovations and moving to value-based healthcare.

Questions/purposes

(1) Are virtual visits less costly than analogous in-person visits, as measured by TDABC? (2) Does TDABC yield cost estimates that are lower or higher than the ratio of costs to charges (RCC), which is a simple, frequently used costing method? (3) Do the total time commitments of healthcare personnel, and that of the surgeon specifically, vary between the virtual and in-person settings?

Methods

Patients for this prospective, observational study were recruited from the practices of the highest-volume virtual-visit surgeons of three subspecialties (joints, hand, and sports) in a multihospital, tertiary-care academic institution in a metropolitan area in the Midwestern United States. Each surgeon had at least 10 years of clinical practice. Between June 2021 and September 2021, we analyzed both in-person and virtual return visits with patients who had an established relationship with the surgeon, because this represented the most frequent type of virtual visits and enabled a direct comparison between the two settings. New patients were not included in the study because of the limited availability of new-patient virtual visits; such patients often benefit from in-person physical examinations and on-site imaging. Additionally, patients seen for routine postoperative care were excluded because they were primarily seen in person by a physician assistant. Data were acquired during this period until 90 in-person and 90 virtual visits were collected according to selection criteria; no patients were lost to follow-up. Distinct process maps, which represent the steps involved in a clinic visit used to measure healthcare personnel time invested, were constructed for in-person and virtual clinic visits and used to compare total personnel and surgeon time spent. To calculate TDABC-derived costs, time allocated by personnel to complete each step was measured and used to calculate cost based on each personnel member’s yearly salary. From the accounting department of our hospital, we acquired RCC cost data according to the level of service for a return visit.

Results

The total median cost, as measured by TDABC, was USD 127 (IQR USD 111 to 163) for an in-person visit and USD 140 (IQR USD 113 to 205) for a virtual visit (median difference USD 13; p = 0.16). RCC overestimated TDABC-calculated direct variable cost in five of six service levels (in-person levels 3, 4, and 5 and virtual levels 3 and 5) by a range of USD 25 to 88. Additionally, we found that virtual visits consumed 4 minutes less of total personnel time (in-person: 17 minutes [IQR 13.5 to 23.5 minutes], virtual: 13 minutes [IQR 11 to 19 minutes]; p < 0.001); however, this difference in personnel time did not equate to cost savings because surgeons spent 2 minutes longer on virtual visit activities than they did on in-person activities (in-person: 6 minutes [IQR 4.5 to 9.5 minutes], virtual: 8 minutes [IQR 5.5 to 13 minutes]; p = 0.003).

Conclusion

Orthopaedic virtual visits did not deliver cost savings compared with in-person visits because surgeons spent more time on virtual visits and participated in virtual visits at the clinical site. Additionally, as anticipated, RCC overestimated costs as calculated by TDABC. These findings suggest that cost is not a primary advantage of transitioning to virtual visits, and that factors such as patient preference and satisfaction should be considered instead.

Level of Evidence

Level II, economic and decision analysis.

Introduction

Telemedicine has been proposed as a mechanism to decrease healthcare costs [9]. Studies using various traditional cost accounting methods have suggested that virtual care offers cost savings to the orthopaedic surgeon and patient for outpatient consultations [4, 7, 12, 25]. Despite relative underuse, virtual care has been of great interest to hospitals and practitioners; 90% of healthcare executives surveyed in 2017 mentioned telemedicine as a growth objective [8, 30, 34]. The coronavirus-19 pandemic catalyzed the widespread rapid adoption of virtual clinic visits to mitigate the spread of infectious disease [3]. In March 2020, 63% of surveyed orthopaedic surgery departments reported they were offering virtual care, with 83% of those surveyed listing coronavirus-19 as the primary impetus [27]. Telemedicine offers numerous advantages beyond potential cost reduction. It empowers patients with debilitating conditions by eliminating transportation obstacles, minimizing wait times, and enhancing communication among multiple treatment centers. Furthermore, it can improve patient adherence to treatment regimens and quality of life and reduce reliance on hospitals [6, 15, 24]. However, telemedicine can give rise to certain negative consequences, including unequal access to orthopaedic care based on factors such as race, ethnicity, primary language, and insurance coverage [36].

Healthcare professionals must have a clear understanding of the cost effectiveness of telemedicine in orthopaedic care to improve resource allocation and ensure accurate reimbursement models. However, we question whether estimates using traditional costing approaches, such as the ratio of costs to charges (RCC) method, can realistically estimate whether telemedicine is cost effective. RCC takes a top-down approach and divides the total department cost by the total department charges [2, 33]. This ratio is multiplied by the charge of a certain procedure to determine its cost [2, 33]. RCC is advantageous because it fully accounts for the total cost of a department and is relatively easy to implement, because no per-case evaluation is required [33]. However, it often provides oversimplified and overestimated costs that do not account for the individual expenditures of medical procedures [33]. Recent studies have suggested implementing time-driven activity-based costing (TDABC) method to measure cost with increased accuracy [10, 18, 19, 28]. TDABC measures the cost of each procedure using two variables: the time spent per personnel on a given procedure and the cost per time for each personnel member [18, 21, 32]. In contrast to the top-down approach of traditional costing methods such as RCC, which estimate costs at a macro, hospital-wide level, TDABC uses a bottom-up approach that provides a more precise understanding of costs at the level of resource consumption [17]. This approach presents several notable advantages. TDABC facilitates the analysis of cost-effective innovations and equips surgeons with the ability to glean insights from their resource allocation in specific activities relative to others [23]. Additionally, because of its cost calculation at the patient level, TDABC enables the linkage of episode-of-care expenditures with patient outcomes, thus enhancing the ability to assess the financial implications of healthcare interventions [18]. Previous studies have suggested that the true cost of a given activity lies between the calculated costs using TDABC and traditional methods of accounting [11, 26]. However, we found no publications that compared the RCC and TDABC methods in terms of surgeon virtual care cost analysis in orthopaedics.

We therefore asked: (1) Are virtual visits less costly than analogous in-person visits, as measured by TDABC? (2) Does TDABC yield cost estimates that are lower or higher than the RCC, which is a simple, frequently used costing method? (3) Do the total time commitments of healthcare personnel, and that of the surgeon specifically, vary between the virtual and in-person settings?

Patients and Methods

Study Design and Setting

Patients for this prospective, observational study were recruited from the practices of three surgeons in a multihospital, tertiary-care academic institution in the Midwestern United States. The surgeons with the highest volume from each of the top three subspecialties in virtual visits (joints, hand, and sports) were recruited for prospective TDABC data collection. The orthopaedic subspecialists in joints, hand, and sports held their clinics at two medical centers with similar practice characteristics in an urban, metropolitan area, in affiliation with the academic institution.

Patients

Between June 2021 and September 2021, three orthopaedic surgeons saw 1101 patients in person and virtually. Of those, we analyzed both in-person and virtual return visits with patients who had an established relationship with the surgeon, because this represented the most frequent type of virtual visits and enabled a direct comparison between the two settings. Of the total visits, 41% (446 of 1101) were identified as established in-person return visits, while 11% (124 of 1101) were categorized as virtual return visits. This analysis excluded 48% of other visit types, including new patients and those seen postoperatively. New patients were not included in the study because of the limited availability of new-patient virtual visits; such patients often benefit from in-person physical examinations and on-site imaging. Additionally, patients seeking routine postoperative care were excluded because they were primarily seen in person by a physician assistant. Data were acquired in this period until 90 in-person and 90 virtual visits were collected according to the selection criteria; no patients were lost to follow-up. All included patients were 18 years or older.

Descriptive Data

The surgeons participating in this study (TRB and EM—acknowledged—and CSD) specializing in joints, hand, and sports each have more than 10 years of clinical experience. All three surgeons are men, graduates of allopathic medical schools, and have completed fellowships in their respective subspecialties. On a typical clinic day, each subspecialist sees between 20 and 30 patients. Additionally, each surgeon incorporates virtual visits into their schedule throughout the day.

Breakdown of Cost Categories

The total cost of each visit was broken into three categories: direct fixed, direct variable, and indirect costs [14]. All costs associated with providing services on the day of the visit were factored into total direct costs (direct fixed and direct variable) [31, 35]. Direct fixed costs are nonvariable costs associated with completing tasks in the hospital, such as building equipment, water, and facility rent [21]. These costs do not vary based on patient activity [21]. In contrast, direct variable costs were derived from the salary of personnel involved in providing care on the date of service [14, 21]. Lastly, indirect costs correspond to components that are not directly related to patient care but still allow the patient-surgeon interaction to occur [14, 21]. Examples of these costs include marketing, administrative staff, and health records.

RCC Cost Calculation

RCC data were received directly from the hospital's accounting department, which divided the costs into direct fixed, direct variable, and indirect. Categorical costs were determined using designated ratios according to standardized internal hospital protocol, facilitated by a computer program (EPSi, Allscripts). These costs varied based on whether the respective surgeon coded the associated visit as a Level 3, 4, or 5 service. According to Centers for Medicare and Medicaid Services guidelines, higher levels of service should indicate a more complex or longer procedure, and thus an increased charge [5]. Although RCC values have inherent accuracy limitations, their acquisition from the hospital and inclusion in this study were driven by two factors. First, because TDABC only measures direct variables costs, RCC estimations of direct fixed costs were essential to provide an estimate of and subsequently compare the overall visit cost. Second, RCC values were gathered specifically to compare the RCC-estimated direct variable cost with the direct variable cost measured by TDABC.

TDABC Process Map

To use the TDABC methodology, separate process maps, which represent the steps involved in a clinic visit used to measure healthcare personnel time invested, were created through the direct observation of each surgeon’s workflow from the time of patient check-in until check-out [22]. The term “healthcare personnel” encompasses all individuals engaged in any aspect of the process map, including the surgeon, physician assistant, resident, medical assistant, secretary, and athletic trainer. Because the study focused exclusively on established return visits, the surgeon workflows exhibited a high degree of similarity, resulting in negligible differences between surgeon subspecialties. Consequently, one virtual care and one in-person visit process map were created that encompassed all meaningful variations among the three surgeons, such as specific healthcare personnel used. Detailed step-by-step process maps were used to record the time taken to complete each step (Fig. 1).

Fig. 1.

These process maps represent an in-person clinic and virtual visit. The median time in minutes to complete each step of the visit is recorded on the map.

TDABC Cost Analysis

Using the TDABC method, the direct variable cost was calculated by multiplying the average time each personnel member allocated at each step of the process map by their per-minute cost [10, 11, 18]. The aggregate annual personnel salaries from our study’s institution were converted to a cost-per-minute unit. We used an established method that approximates an 80% theoretical working capacity to account for personnel inefficiencies such as breaks and other activities not related to patient care [1, 16]. Additionally, 4 weeks per year were retracted from an individuals’ working capacity for assumed vacation [1, 16]. The resulting working capacity was multiplied by each personnel member’s cost-per-minute unit and process map time spent to determine each individual’s direct variable cost. All personnel costs were summated to provide a final direct variable cost per visit. Because the TDABC method cannot precisely calculate direct fixed costs, we obtained these values using RCC data from the hospital’s accounting department, which were separated by service level [21]. To develop an average direct fixed cost for each surgeon, we used a weighted average by multiplying the cost of each RCC level by the frequency the three surgeons charged each level. Surgeons 2 and 3 (EM and CSD) practiced exclusively at one location (Clinic 1). Surgeon 1 (TB) practiced at two locations (Clinic 1 and Clinic 2), but there was an identical ratio of in-person to virtual visits at each site in this study. Considering the similarities between the clinics, the limitations in the accuracy of RCC, and the study's scope, only direct fixed cost data from Clinic 1 were used to standardize direct fixed cost among surgeons. Finally, indirect costs were estimated as a fixed proportion (40%) of the total direct costs. This proportion was determined using the average ratio of indirect to total direct costs for an in-person and virtual outpatient visit, a method described by studies using TDABC [1, 21, 26].

Primary and Secondary Study Outcomes

Our primary study goal was to compare the overall cost and healthcare personnel time investment of in-person and virtual return visits using the TDABC method. To compare cost, we used the described TDABC cost analysis process, which combines direct variable cost measured by TDABC, direct fixed cost estimated by RCC, and indirect cost, which incorporates both methods. The total cost per visit was calculated according to this method for all three surgeons and was subsequently compared by surgeon and overall. Visit time was compared in two ways: the healthcare team and patient perspectives. From the healthcare team perspective, all staff activities including all personnel were considered. This includes surgeon note dictation and check-in to video software time. From a patient perspective, we defined “direct professional time” as the time a patient spent directly interacting with one of the professionals on the care team (surgeon, physician assistant, or resident) and “direct surgeon time” as the time a patient spent directly with the surgeon.

Our secondary study goal was to compare RCC estimated and TDABC calculated direct variable cost. To compare with RCC values by level of service, TDABC direct variable costs were divided into the respective service level recorded for each visit and subsequently averaged. The comparison between methods specifically examined the direct variable cost because it is the component of the total cost that is computed using TDABC [1, 21].

Ethical Approval

Institutional review board approval was sought and granted (number 13841).

Statistical Analysis

An a priori power analysis demonstrated that a minimum sample size of 30 patients per cohort per surgeon would be necessary to elucidate a hypothesized 20% difference in cost per group. The Shapiro-Wilk test was employed to determine the normality of cost and time data used for statistical comparison between in-person and virtual visits. Descriptive statistics were calculated for cost and time data (median and IQR for non-normal distributions and mean and SD for normal distributions). Total and surgeon-specific visit cost and time spent for in-person and virtual visits were compared using two-tailed t-tests for normal distributions and the Mann-Whitney U test for non-normal distributions. The TDABC direct variable cost was compared with the RCC-estimated direct variable cost by level of service using one-sample t-tests for normally distributed data and one-sample Wilcoxon tests for non-normally distributed data. To reconcile potential variations in cost among surgeons, we conducted two sensitivity analyses to assess the generalizability of our overall cost findings. The first analysis involved a generalized estimating equation of the overall cost based on visit environment, which accounts for correlations within each surgeon. The second analysis used a modified Mann-Whitney U test, controlling for the median difference in TDABC-calculated surgeon cost for each surgeon, which we believe to be the primary factor influencing differences in costs based on visit environment. Statistical significance was defined as a p value less than 0.05. All statistical analyses were performed using GraphPad Prism 9 (GraphPad Software).

Results

TDABC-derived Cost Comparison of Virtual and In-person Visits

Using the TDABC accounting method, there was no difference in total cost between in-person and virtual visit modalities (in-person: USD 127 [IQR USD 111 to 163], virtual: USD 140 [IQR USD 113 to 205]; median difference USD 13; p = 0.16) (Fig. 2). When segregating individual surgeon data, virtual visits cost USD 54 more for Surgeon 2 (in-person: USD 164 [IQR USD 127 to 191], virtual: USD 218 [IQR USD 179 to 248]; p < 0.001) (Fig. 2). Surgeon 1 (in-person: USD 135 [IQR USD 105 to 152], virtual: USD 114 [IQR USD 103 to 140]; p = 0.08) and Surgeon 3 (in-person: USD 113 [IQR USD 103 to 127], virtual: USD 122 [IQR USD 100 to 144]; p = 0.36) had no difference in total cost. Because TDABC explicitly measures the direct variable cost, this was further investigated. There was an increase of USD 9 in direct variable cost associated with virtual visits (in-person: USD 45 [IQR USD 33 to 66], virtual: USD 54 [IQR USD 39 to 86]; p = 0.043) (Fig. 2). Only Surgeon 2’s direct variable cost was different between environments, with an increase of USD 29 in the virtual visit setting (in-person: USD 61 [IQR USD 35 to 80], virtual: USD 90 [IQR USD 62 to 111]; p < 0.001).

Fig. 2.

In this comparison of in-person and virtual visit costs according to the TDABC method, the total cost is broken down into indirect, direct fixed, and direct variable costs. The error bars represent the IQR of the entire column; ^ap < 0.05. A color image accompanies the online version of this article.

To reconcile with the generalizability of our overall findings, considering the variations among surgeons, sensitivity analyses were used. The generalized estimating equation found no difference in overall cost based on visit environment when accounting for correlations within each surgeon (p = 0.27) (Supplemental Table 1; http://links.lww.com/CORR/B213). Additionally, when controlling for the median difference in each surgeon’s TDABC-calculated surgeon cost between environments, in-person and virtual visit costs were similar among all three surgeons (Supplemental Fig. 1; http://links.lww.com/CORR/B214).

Comparing RCC-derived with TDABC-derived Cost Estimates

TDABC direct variable costs were lower than those estimated by RCC for five of six total service levels including in-person and virtual visits. For in-person visits, TDABC-calculated costs were lower than their RCC-estimated counterparts by USD 32 for Level 3 (TDABC: USD 42 [IQR USD 32 to 61], RCC: USD 74; p < 0.001), USD 46 for Level 4 (TDABC: USD 63 ± 27, RCC: USD 109; p < 0.001), and USD 82 for Level 5 (TDABC: USD 82 ± 47, RCC: USD 164; p = 0.003). For virtual visits, TDABC-calculated costs were lower than with RCC by USD 25 for Level 3 (TDABC: USD 48 [IQR USD 36 to 62], RCC: USD 73; p < 0.001) and USD 88 for Level 5 (TDABC: USD 77 ± 34, RCC: USD 165; p < 0.001). There was no difference between the TDABC-calculated and RCC-estimated cost for Level 4 virtual visits (Fig. 3).

Fig. 3.

This figure shows the RCC estimated and TDABC calculated direct variable costs associated with the level of service for an in-person and virtual patient return visit. ^ap < 0.01. A color image accompanies the online version of this article.

Time Differences Between Virtual and In-person Visits

Overall, we found that virtual visits consumed 4 minutes less of total personnel time than in-person clinic visits (in-person: 17 minutes [IQR 13.5 to 23.5 minutes], virtual: 13 minutes [IQR 11 to 19 minutes]; p < 0.001) (Fig. 4). However, surgeons spent longer on virtual visit activities (with the patient, note dictation, and video software login) than on in-person visits (with the patient and note dictation) (in-person: 6 minutes [IQR 4.5 to 9.5 minutes], virtual: 8 minutes [IQR 5.5 to 13 minutes]; p = 0.003). The time in which a patient directly interacted with the surgeon, residents, and physician assistants during each visit was also compared. In total, direct professional time had a decrease of 3 minutes in the virtual setting (in-person: 8 minutes [IQR 4.5 to 12 minutes], virtual: 5 minutes [IQR 3 to 9 minutes]; p < 0.001) (Fig. 5). Although patient interaction time with all professionals decreased virtually, there was no difference in patient time spent directly with the surgeon, termed direct surgeon time (in-person: 4.5 minutes [IQR 2.5 to 8 minutes], virtual: 5 minutes [IQR 3 to 9 minutes]; p = 0.25) (Fig. 5).

Fig. 4.

This is a comparison of the average personnel time allocated per patient for in-person and virtual clinic visits. All activities involved in patient care are included. The error bars represent IQR of the entire column; ^ap < 0.05; ^bp < 0.01.

Fig. 5.

This graph shows a comparison of professional time spent directly with the patient for the in-person and virtual settings. The error bars represent the IQR of the entire column; ^ap < 0.01 for professional time comparison; ^bp < 0.05 for surgeon time comparison; ^cp < 0.01 for surgeon time comparison. A color image accompanies the online version of this article.

Discussion

Orthopaedic surgeons are increasingly incorporating telemedicine into their practices to expand the range of clinical options available to patients. Telemedicine is also seen as a potential means of reducing healthcare expenses. Consequently, it is crucial for practitioners to have a comprehensive understanding of the cost of telemedicine in orthopaedic care to improve resource allocation. However, we question the ability of traditional costing methods such as RCC to estimate whether telemedicine is cost effective. Our study discovered that the costs associated with in-person and virtual return visits were comparable when assessed using TDABC. In fact, virtual visits were more expensive when considering the direct variable costs. This stems from virtual visits consuming more time for the surgeon to complete, despite demanding less time from the healthcare team overall. Furthermore, the RCC accounting method overestimated the direct variable costs calculated by TDABC in five of six levels of service, which raises doubts about the suitability of RCC for determining the impact of cost-effective innovations and appropriately linking cost to patient outcomes.

Limitations

First, our study involved only three surgeons based on TDABC feasibility constraints. However, including multiple surgeons from different subspecialties as opposed to only one, as other TDABC studies have done [13, 21], was important to showcase the variation in surgeon practice characteristics such as time spent and personnel used, which influences cost. Considering this limitation, the generalized estimating equation sensitivity analysis, which accounts for correlations within each surgeon’s cost, also found no difference in overall cost between in-person and virtual visits. Although it is challenging to determine whether a given surgeon would have practice characteristics identical to those of the specific surgeons in our study, the results still provide valuable insights into the factors influencing the overall visit cost. Because direct fixed cost between in-person and virtual visits performed at the same clinic site should be comparable, the cost difference between an individual in-person and virtual visit at the same clinic primarily depends on the time spent by the surgeon on virtual visit activities. For instance, Surgeon 2 had a higher direct variable cost for virtual visits than for in-person visits because they spent more time on virtual visit activities. Surgeon time spent is also likely influenced by healthcare personnel used, because Surgeons 2 and 3 spent more time virtually than in-person, and both used physician assistants and residents. This suggestion is supported by the sensitivity analysis that found no difference in total cost for each surgeon when controlling for each surgeon’s median difference between the in-person and virtual TDABC-calculated surgeon cost, which reflects the difference in time spent between environments by each surgeon. Therefore, it is crucial for surgeons to carefully evaluate their own practice characteristics in relation to virtual visits.

The second limitation is the use of RCC data to estimate the direct fixed cost, which uses a fixed ratio to estimate the direct fixed cost from the overall visit cost. The inability of TDABC to measure the direct fixed cost is an acknowledged limitation of the study method, and RCC direct fixed data have been used previously [21]. However, RCC offers a generalized estimate of cost that does not have a logical connection to resource consumption; it is used by hospitals because it is easy to use and can account for all costs in a department. Additionally, although studies have expanded their indirect costs to focus on patient time costs associated with medical care [37, 38], our study focused on the cost of the visit from a surgeon’s perspective. Therefore, our cost analysis calculates indirect costs in accordance with other studies using the TDABC method as an average ratio of indirect to total direct costs for an in-person and virtual outpatient visit [1, 20, 21, 26]. However, because in-person and virtual visits at the same clinical site should yield comparable direct fixed and indirect costs, direct variable costs are likely more reflective of cost differences between visit modalities.

A third limitation of this study is that we did not measure chart review by surgeons, physician assistants, and residents from the process map because of the sporadic nature in which these individuals conducted this action. We believe this is a minor limitation because of its equal application to both visit groups, thereby nullifying its effects on the outcomes. A fourth minor limitation is the estimated 20% cost difference used for our power analysis, which may not be able to measure smaller differences in cost. However, this is a standard cost difference at our institution for power analyses in studies of this type. A final limitation of this study is the potential for the healthcare professionals to have changed their behavior because they were aware they were being observed (sometimes called the Hawthorne Effect). However, we believe this is likely negligible because of its equal application to both visit groups.

TDABC-derived Cost Comparison of Virtual and In-person Visits

The available evidence on cost differences between virtual and in-person outpatient visits in orthopaedics is limited. Nonetheless, our finding that orthopaedic virtual visits do not offer cost savings differs from other studies showing there are decreased costs associated with outpatient telehealth visits [4, 7, 12, 25]. However, none of the other studies used the TDABC method. One study from an academic institution in Norway estimated that a virtual orthopaedic visit was 35% less costly than a traditional in-person visit [4]. Their analysis measured costs associated with implementing and running a video service, such as personnel salary, videoconferencing equipment, and rent or equipment prices; however, their cost-per-visit calculations in both settings were not delineated. Another study at a university hospital in the United Kingdom reported a 41% surgeon cost savings for total joint arthroplasty follow-up appointments [7]. They received calculated visit costs from their hospital business finance department, which included direct and indirect costs of healthcare personnel time, radiography, rent, and administrative costs. However, the specific cost accounting methods that were used were unclear. In contrast to both studies’ findings, our results showed there was no difference in cost between virtual and in-person visits. This is likely because of the cost accounting methods we used, because TDABC allows a more accurate estimation of individual appointment expenditures [19]. Interestingly, a study in a large academic institution found no difference in general surgery and urology outpatient virtual visit costs using the TDABC method [29], which aligns with our findings. However, these virtual visit cost analysis findings should be regarded as a recommendation because they may not be universally applicable to all orthopaedic surgeons, particularly those with lower virtual visit frequencies. Nonetheless, considering the similarity in cost, perhaps the implementation of virtual visits should instead prioritize alternative factors such as patient preference and outcomes data.

Comparing RCC-derived With TDABC-derived Cost Estimates

With respect to RCC and TDABC comparison, previous studies of RCC overestimated TDABC-calculated costs by 41% and 36% for hip and knee arthroplasty expenditures, respectively [1, 26]. Similarly, we found that RCC overestimated TDABC cost in five of six service levels by USD 25 to 88, depending on the level of service. Interestingly, RCC costs provided by our hospital’s accounting department were almost identical between in-person and virtual visits, which reflects the on-site location of virtual visits. Thus, we would anticipate decreased direct fixed costs if surgeons used the virtual platform outside the hospital. Although the findings may raise doubts about the utility of RCC, it is crucial to use both accounting methods. TDABC offers a precise cost evaluation for innovations, but there may be a disparity between TDABC-calculated costs and overall departmental expenses. Hence, RCC is indispensable from a systemic standpoint to ensure proper book balancing. Because each cost accounting method serves distinct purposes, both are essential.

Time Differences Between Virtual and In-person Visits

When examining appointment time by visit modality, we found virtual visits were 24% shorter than in-person visits. In contrast to our findings, a previous study found virtual visits to be 26% longer in total than in-person visits [29]. However, surgeons in their study did not use physician assistants and residents during in-person visits. Additionally, the authors measured outpatient visit times in urology and general surgery, which contain subject matter differences from orthopaedics that may affect virtual visit discussions. Although the overall visit time may be shorter in the virtual environment, our data demonstrated no difference in surgeon time spent interacting with the patient. Even if the individual visit time and therefore cost for a given surgeon are identical between in-person and virtual visits, the true cost per visit when viewed on a daily scale would largely depend on the visit volume per day. Virtual visits consume less time overall, allowing surgeons to potentially conduct more virtual visits in a day while using a leaner staff.

Conclusion

Our study revealed that costs for in-person and virtual return visits were comparable using TDABC. We also found that the RCC accounting method overestimated direct variable costs calculated by TDABC in five of six service levels. In-person visits took longer, but surgeons spent a similar amount of time with patients in both environments. Because of a lack of cost differences between in-person and virtual visits, our study suggests that expenditures should not be a primary determinant of administrative decisions to implement virtual visits, and perhaps other factors such as patient preferences, ease of access, and patient satisfaction should be considered instead. Additionally, the overestimation of direct variable costs by RCC in contrast to TDABC creates an opportunity for broader adoption of TDABC measurement. This could involve the integration of more-detailed electronic medical record time stamps in clinical settings, similar to those already in use in operating rooms. A future direction of study would be to measure the value of virtual visits rather than only cost, which would require a patient-reported outcome or satisfaction measure.