Rapid, Safe, and Efficient Reduction in the Carbon Footprint of Preoperative Travel: A Nationwide Study

To the Editor: Evidence-based recommendations to mitigate the carbon footprint of general anesthesia exist,1 but there is a paucity of research on the footprint of preoperative care. Travel is a leading cause of emissions, and patients routinely travel to preoperative clinics for in-person evaluation before common procedures.2,3 Single-center European studies show that such travel has a substantial ecological impact,3–5 but there is a lack of large-scale data on the impact of travel to preoperative clinics in the United States. In addition, a recent meta-analysis, also based on single-center studies, concluded that preoperative telemedicine is a safe and efficient alternative to in-person evaluations before low-risk procedures and among healthier patients.3–5 However, there is also a paucity of multicenter data on the safety and efficiency of preoperative telemedicine among older adults with multiple comorbidities undergoing standard procedures like total knee arthroplasty (TKA).3–5 We address these gaps in knowledge, capitalizing on a sharp increase in telemedicine during the COVID-19 pandemic within one of the largest integrated healthcare networks in the United States: the Veterans Health Administration (VHA).6,7 Our aims were to (1) examine the pandemic’s impact on telemedicine in anesthesia-run preoperative clinics within the VHA, (2) the climate impact of travel for in-person evaluations versus travel avoided with telemedicine, and (3) the safety and efficiency of telemedicine versus in-person evaluations. After approval by the Durham Veterans Affairs Health Care System institutional review board, and building on our prior research,8,9 we used longitudinal time-stamped data contained in the VHA’s harmonized nationwide repository of administrative and electronic health records (see the supplemental content for details; https://links.lww.com/ALN/D927) to identify patients who had a telemedicine (telephone or video) versus in-person encounter in an anesthesia-run preoperative clinic within 30 days prior to inpatient TKA between January 2016 and June 2021. We compared the use of telemedicine, carbon emissions from travel, percentage of patients who traveled more than 80 miles roundtrip, baseline patient attributes (age and comorbidities summarized using the Van Walraven score), facility complexity,8,9 and rurality,10 among individuals undergoing procedures before versus after onset of the pandemic (January 2016 to March 2020 vs. April 2020 to June 2021, respectively). We determined distances between veterans’ homes and the VHA facility where care was delivered based on geocodes (latitude and longitude coordinates, data maintained by the VHA Planning Systems Support Group; see supplemental content, https://links.lww.com/ALN/D927).9 The roundtrip distance was converted into carbon dioxide emissions based on Environmental Protection Agency calculators (404 to 461 g CO2 per mile), assuming that gasoline-powered passenger vehicles were used). The safety outcomes were death or readmission within 30 days, emergency department visits within 7 days, and in-hospital complications (measured based on diagnosis codes). The efficiency outcomes were case cancellations at any time and in the preoperative holding area. Associations between telemedicine (vs. in-person evaluation) and these outcomes was determined using multivariable logistic regression models to adjust for baseline patient characteristics, time periods, and facility complexity (supplemental content, https://links.lww.com/ALN/D927). Statistical significance was determined based on a P value threshold of 0.005 (given the large sample size). Analyses were conducted in the secure VHA computing infrastructure using SQL Management Studio (Microsoft Corporation, USA) and SAS version 9.2 (SAS Institute, USA). The study cohort included 20,467 patients evaluated in 88 facilities nationwide (see supplemental fig. 1, https://links.lww.com/ALN/D928). Overall, 94% (n = 19,095) had in-person evaluations, and 6% (n = 1,372) had telemedicine evaluations (4.6%, n = 1,077, telephone evaluations; 1.4%, n = 295, video-based evaluations). At the facility level, 39 and 16 facilities, respectively, had telephone and video-based evaluations (table 1). The average roundtrip distance traveled for an in-person encounter was 135 miles, amounting to between 54 and 62 kg in carbon emissions (fig. 1). On average, the patients with telephone and video evaluations would have traveled 104 and 137 miles, respectively, which amounts to the avoidance of 42 to 48 kg and 55 to 63 kg in carbon emissions (fig. 1). Patients evaluated post pandemic were slightly older, less rural, and more likely to be treated in high-complexity facilities versus prepandemic patients (table 1). The overall burden of comorbidities was comparable in the two time periods. Telemedicine nearly quadrupled after March 2020 (to 17% vs. 4.5% pre pandemic), the percentage of patients traveling more than 80 miles dropped (to 33% vs. 43% pre pandemic), and emissions dropped by nearly 50% (all P < 0.005; table 1). The number of facilities using telephone evaluations increased from 24 to 43%, but the change was not statistically significant (P = 0.01). Compared to counterparts with telemedicine encounters, patients with in-person encounters were slightly younger, less rural, and less likely to be treated in high-complexity facilities (table 1), but the overall burden of comorbidity did not differ between recipients of telemedicine versus in-person evaluation (table 1). When controlling for covariates using multivariable logistic regression models, telemedicine was not associated with statistically worse safety or efficiency outcomes (table 1; supplemental tables S2 through S7, https://links.lww.com/ALN/D929). Table 1. - Preoperative Evaluation for Total Knee Arthroplasty across 88 Health Systems across the United States before versus after Onset of the COVID-19 Pandemic and with Telemedicine versus In-person Encounters Characteristics Compared January 2016–March 2020 (n = 17,140) April 2020–June 2021 (n = 3,327) P Value No. of procedures per month 336 222 < 0.005 Telemedicine**, % 4.5% 17% < 0.005 Travel > 80 miles roundtrip**, % 43% 33% < 0.005 Carbon dioxide emissions, metric tons/mo** 17.2–19.8 9.5–11 < 0.005 No. of facilities conducting TKAs 87 76 — Facilities using telephone-based evaluations*, No. (%) 21 (24%) 33 (43%) 0.01 Facilities using video-based evaluations, No. (%) 9 (10%) 14 (18%) 0.14 Baseline patient attributes Mean age, yr** 66.2 67.0 < 0.005 Mean Van Walraven index 6.5 6.6 0.66 Patients living in rural locations, %** 38% 35% < 0.005 Patients treated in high-complexity facilities, %** 74% 79% < 0.005 Characteristics Compared In-person (N = 19095) Telemedicine (total n = 1,327, 1,077 telephone and 295 video) P Value Mean roundtrip distance traveled vs. avoided, miles 135 traveled 104–137 avoided (telephone and video evaluations respectively) — Carbon emissions released vs. avoided, kg 54 released 42–55 avoided (telephone and video evaluations) — No. of facilities 88 39 (telephone), 16 (video) — Baseline patient attributes Mean age, yr** 66.3 66.7 < 0.005 Mean Van Walraven index 6.5 6.5 0.48 Patients living in rural locations, %** 37% 42% < 0.005 Patients treated in high-complexity facilities, %** 73% 95% < 0.005 Safety outcomes 30-day mortality, No. (%) 30 (0.16%) 3 (0.22%) 0.58 Adjusted odds ratio [95% CI] ref 1.24 [0.3, 3.7] 0.73 30-day readmissions, No. (%)* 3,172 (16.6%) 176 (12.9%) < 0.005 Adjusted odds ratio [95% CI] ref 0.84 [0.71, 1] 0.05 7-day emergency department visits, No. (%) 805 (4.2%) 74 (5.4%) 0.04 Adjusted odds ratio [95% CI] ref 1.24 [0.96,1.59] 0.09 Complications during hospitalization 1,107 (5.8%) 62 (4.5%) < 0.05 Adjusted odds ratio [95% CI]** ref 0.72 [0.59, 0.88] < 0.005 Efficiency outcomes Cases cancelled (all reasons), No. (%) 2,969 (15.5%) 244 (17.8%) 0.02 Adjusted odds ratio [95% CI] ref 0.98 [0.84, 1.13] 0.75 Cases cancelled in preoperative holding area, No. (%) 129 (0.68%) 6 (0.44%) 0.29 Adjusted odds ratio [95% CI] ref 0.65 [0.25, 1.37] 0.31 Overall encounters declined during the pandemic, telemedicine quadrupled among patients and nearly doubled among facilities, while age and comorbidities remained unchanged. Fewer rural veterans were seen during the pandemic. When comparing recipients of telemedicine versus in-person evaluations, age and comorbidities were almost identical, and outcomes were comparable, except for readmissions (telemedicine may be associated with slightly decreased readmissions). *P < 0.05, **P < 0.005.ref, reference group. Fig. 1.: Carbon emissions avoided by telemedicine evaluations versus released by travel for in-person assessments: Projections based on post-pandemic patterns of care. CO2, carbon dioxide.This nationwide study among older adults undergoing TKA within the largest health system in the United States shows that (1) the carbon footprint of travel to preoperative clinics was substantial (and significantly greater than the impact of general or regional anesthesia for TKA; see supplemental content, https://links.lww.com/ALN/D927), (2) telemedicine increased rapidly after onset of the pandemic, and (3) telemedicine was comparable to in-person evaluation in terms of the odds of adverse safety and efficiency outcomes. These results are consistent with prior studies.3–5 The rapid expansion of preoperative telemedicine also aligns with data from primary care and mental health clinics.6 The findings have at least two major implications. First, climate change poses an existential threat, and it appears very unlikely that preoperative telemedicine is inferior to in-person evaluations in its safety and efficiency, even among older adults with multiple comorbidities. Thus, rapid expansion of preoperative telemedicine appears to be a feasible strategy to reduce the carbon footprint of perioperative care. Second, it is notable that telemedicine evaluations are primarily telephone based and, on average, more common among patients living closer to VA facilities. Telemedicine was also, almost completely, limited to high-complexity facilities. Barriers to telemedicine in small hospitals and among rural patients need to be determined. This study is underpowered to detect a difference in mortality, and fewer readmissions can result from several factors (including social support). However, the lack of a difference in the odds of emergency department visits and lower odds of in-hospital complications is reassuring. Additional limitations include our assumptions about distances (we computed haversine distances rather than using road-mapping software), exclusion of internal medicine–run preoperative clinics, and lack of data on patient self-selection and provider-level differences in care. We also did not study the carbon impact of telemedicine.2 Research is needed to address these limitations. Still, the nationwide scope of this study, use of geocoded data,8 and analysis accounting for important confounders are significant strengths of this study.8 In conclusion, preoperative telemedicine seems to be safe and efficient and rapidly implementable. Facilities that are currently low users of telemedicine must consider the adoption of procedures and policies being used in preoperative clinics with established telemedicine programs. Acknowledgments The authors acknowledge the services provided by SporeData Inc. (Durham, North Carolina). Research Support This study was supported by the Veterans Health Administration (Washington, D.C.), Office of Connected Care, and Virtual Care (VC) Consortium of Research (CORE). This study was among several operational (non-research) activities funded by an award from the VHA Office of Connected Care in 2021 to 2022. The funding sources had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and the decision to submit the manuscript for publication. Competing Interests The authors declare no competing interests. Supplemental Digital Content Supplemental Content, https://links.lww.com/ALN/D927 Supplemental Figure, https://links.lww.com/ALN/D928 Supplemental Tables, https://links.lww.com/ALN/D929