The Carbon Footprint of Residency Interview Travel

In recent years, the residency application process has placed an increasingly large burden on the limited time and finances of fourth-year US medical students. Recent studies have reported significant increases in the average number of applications submitted, length of rank order lists, and costs incurred by medical students over the past decade.^1–3  With an increase in the number of interviews attended and associated travel, we must consider the additional burden of the carbon footprint of this process and its impact on climate health.

The impact of travel (flights in particular) on the carbon footprint of academia is well documented^4–8 ; however, there is a paucity of data regarding the impact of residency interviews. Whether by airplane, train, or automobile, the mode of transportation has different effects on the environment. Commercial aviation is the fastest-growing source of greenhouse gas emissions,⁹  with total CO₂ emissions from all commercial air flights totaling 918 million metric tons in 2018, representing 2.4% of global emissions.¹⁰  Assessing the carbon footprint of flights is unique compared to other forms of travel in that the relationship between carbon footprint and distance traveled is nonlinear; the ascent and descent of planes use considerably more fuel than cruising at altitude.¹¹  One study estimated the carbon intensity of flights to be 3.69 to 5.39 ounces CO₂ per revenue-passenger-mile.¹⁰  While these numbers may seem small, the length of travel makes the effect of domestic travel up to 0.75 metric tons of CO₂ per passenger per flight.

The fuel efficiency of motor vehicles varies widely between make, model, and year of manufacture; thus, it is difficult to make broad comparisons about the carbon footprint per distance traveled by car and distance traveled by plane or train. However, regardless of the baseline fuel efficiency of the car, carpooling can drastically reduce the carbon footprint attributed to each passenger; sharing a ride with just one other person will halve the carbon footprint of each rider.¹¹  Increased occupancy contributes to a lower carbon footprint per passenger of buses and trains, which have lower total carbon footprint per distance traveled than planes and more passenger capacity than cars. However, commercial train and bus routes in the United States are not as extensive or widely available for many travel needs; therefore, flights and single occupancy car trips remain the most common options for long distance travel.

This study seeks to elucidate the travel patterns of medical students applying to residency in order to estimate the total carbon footprint of this process.

Graduating fourth-year medical students from the University of Michigan Medical School (UMMS) were surveyed in 2020 to gather information regarding their travel related to residency interviews. The survey was designed to obtain information related to interview travel in order to estimate each individual's carbon footprint (provided as online supplementary data). Data gathered included intended specialty, number of interviews completed, start and end points for all flights, trains, buses, and personal vehicle trips as well as the make, model, and year of vehicles used. Students were each provided a $20 gift card for completion of the survey, provided by an institutional research grant, and all data were deidentified after collection. This study was approved as exempt by the University of Michigan Medical School Institutional Review Board. The survey was reviewed by an education faculty member (W.P.) and a content expert faculty member (J.W.) to assess for clarity, sensibility, and completeness. It was initially administered to a small group of medical students for feedback. The carbon footprint of individual flights and car trips was determined using the Carbon Footprint Calculator for Individuals and Households¹² ; the carbon footprint of intercity train trips was calculated using the Greenhouse gases, Regulated Emissions, and Energy use in Transportation (GREET) model.¹³  These values were then summed to determine each student's personal carbon footprint. The Carbon Footprint Calculator is independently audited by the quality insurance standard and complies with methodology outlined by the UK government.¹⁴  Verification of air travel carbon footprint data was completed using the ICAO Carbon Emissions Calculator.¹⁵ 

In order to calculate individual carbon footprints, the total footprint of one trip by plane, train, or car is divided by average occupancy. The carbon footprint of a trip from Detroit, Michigan, to Denver, Colorado, for example, varies widely by mode of transportation. Traveling by plane produces 0.25 metric tons CO₂ per passenger; by train, 0.19 metric tons CO₂ per passenger; and by car, assuming a 2010 Toyota Camry (a common car reported by students in this survey), 0.52 metric tons CO₂ for one passenger. Class averages of total carbon footprint per student and carbon footprint per interview were calculated. Responses were grouped by specialty type: procedural specialties included all surgical subspecialties, obstetrics and gynecology, emergency medicine, anesthesiology, and dermatology; medical specialties included internal medicine, pediatrics, neurology, physical medicine and rehabilitation, family medicine, psychiatry, and radiology. Further breakdown by travel for preliminary and transitional year positions was also completed. Group averages were compared using a t test in Microsoft Excel.

The response rate was 59% (103 of 174), and Table 1 shows the breakdown of students by intended specialty. Specialty demographics are included for the respondents and the overall graduating class in the online supplementary data. Average number of interviews per applicant varied widely by intended specialty (9.75–27.33) and individually (3–45); total class average for interviews completed was 14.39. The class average for total carbon footprint per student was calculated as 3.07 metric tons CO₂, making the class average carbon footprint per interview 0.21 metric tons CO₂ (Table 2). On average, 76% of the average student's carbon footprint was attributed to flights, followed by car trips (23%) and train trips (1%). Of the 2424 one-way trips to and from interviews completed by applicants, 1326 (55%) were flights, 969 (40%) were car trips, and 129 (5%) were train trips. Most car trips were to cities within 400 miles of Ann Arbor, Michigan, such as Detroit, Chicago, Cleveland, and Pittsburgh. Common travel destinations can be found in Table 3.

There was no statistically significant difference in number of interviews attended (P = .41), average total CO₂ per student (P = .77), or average CO₂ per interview (P = .99) between the procedural and medical specialties; this finding was robust even after outliers, defined as students with number of interviews greater than 3 standard deviations above the mean, were removed. Averages for each specialty are displayed in Table 2. Of the 19 students applying to specialties requiring a transitional or preliminary year such as anesthesia, neurology, dermatology, or physical medicine and rehabilitation, 14 students provided detailed information regarding their travel to interviews dedicated to transitional or preliminary year positions. The average carbon footprint of interviews for these 14 students was 3.36 metric tons CO₂ with an average of 0.61 metric tons of CO₂ (13%) attributed to transitional or preliminary year interview travel alone. There was no statistically significant difference in number of interviews (P = .17), average total CO₂ per student (P = .56), or average CO₂ per interview (P = .99) between this group and all other applicants.

In this study, we determined that the carbon footprint of residency interview travel is considerable, with the majority accrued from air travel. Our results are timely given that as a result of the COVID-19 pandemic, all interviews during the 2020–2021 academic year have been virtual. If we extend the results of this study to estimate the total CO₂ emissions saved by this decision, using the population of all 18 925 US seniors who participated in the 2019 Match¹⁶  and assuming an average of 13 interviews¹⁷  and 0.21 metric tons CO₂ per interview, the resulting value would be 51 665 metric tons CO₂. This is equivalent to the amount of CO₂ produced by 11 162 passenger cars in 1 year.¹⁸  Given these results, we argue that reducing the travel burden on residency applicants during the eventual transition back to in-person interviews is a relatively simple tangible way to contribute to the goal of reducing the carbon footprint of academic medicine and the health care system as a whole.

This work adds to the growing literature on the need to reduce the burden of excessive interviewing. Some recent proposals to address this problem, such as the early result acceptance program (ERAP),¹⁹  may also serve to reduce the environmental effects of the residency application process by reducing travel. The ERAP proposes to allow students to apply to, interview at, and rank up to 5 programs before the traditional Match. If 25% of US allopathic seniors (4731 of 18 925) participated in and successfully matched through an early Match system, assuming a reduction in interviews for those applicants from 13 (median reported from NRMP)¹⁷  to 5, we would expect to see a 15% decrease in overall carbon footprint for the Match (43 717 metric tons CO₂). The decrease of 7948 metric tons CO₂ is equivalent of taking 1717 passenger cars off the road for 1 year.¹⁸  In the 2020 Match, all obstetrics and gynecology residency programs adhered to a standardized calendar for application deadlines as well as interview offer and rejection dates in order to simplify the process for applicants and programs.²⁰  With a more standardized scheduling process, students would be able to more efficiently cluster interviews by geographic location and to judiciously choose which interviews to attend, thus potentially reducing overscheduling and total travel.

While some alternatives to the current Match process may require broader institutional change, there are many relatively simple changes that can be made by residency programs to reduce interview-related travel and thus carbon emissions. Prior to the COVID-19 pandemic, many programs offered transitional/preliminary year candidates virtual interviews or added brief interviews to their categorical program interview day. If virtual interviews for transitional/preliminary year positions were continued, it could reduce the total carbon footprint of this population of 13 140 students²¹  from 35 872 metric tons CO₂ to 31 208 metric tons CO₂, assuming 13 interviews per applicant and a 13% reduction as suggested by the data, which is equivalent of taking 1008 passenger cars off the road for 1 year.¹⁸  Many specialties encourage or require students to complete visiting rotations at other institutions before applying. A survey of US applicants for residency in the 2014–2015 academic year suggested that approximately 60% of all fourth-year medical students completed at least one away rotation.²²  If all such programs interviewed candidates during their away rotations, it could also serve to reduce students' carbon footprint and travel burden.

This study is limited in its ability to extrapolate the results of this analysis to the total population of US allopathic seniors, as data were collected from only one medical school class, which is unlikely to be completely representative of all graduating US medical students. The University of Michigan Medical School is among the top 25 programs in the nation per US News & World Report,²³  and many students are offered interviews across the country. Additionally, the UMMS program is located in the Midwest, and therefore coastal destinations require increased travel compared to students already situated on the coasts. Since many UMMS students interview at one or more coastal programs, the resulting average carbon footprint calculated from this data may be an overestimation. Future similar studies from other schools would be invaluable to accurately assess the national environmental impact of this process. In addition, there were insufficient sample sizes in many individual specialties represented by the survey, which limited comparisons that could be made about differences in travel habits and carbon footprint among applicants to different specialties.

The environmental impact of residency interviews must be considered, and this work adds to multiple calls to improve residency application processes. As we transition to in-person interviews in future cycles, viable alternatives that decrease the number of interviews attended by applicants need to be prioritized. Reforming the residency application and interview process can serve multiple beneficial roles, reducing the burden on both applicants and the environment.

Converting all interviews during the 2020–2021 academic year to a virtual format reduced the carbon footprint of the residency application process by 51 665 metric tons CO₂. Practices to reduce the number of interviews attended and thus the total carbon footprint include an early result acceptance program, standardizing interview invitation schedules, and maintaining virtual interviews for some positions.