Article Text

Download PDFPDF

Close encounters of the US kind: illness and injury among US athletes at the PyeongChang 2018 Winter Olympic Games
  1. Dustin Nabhan1,2,
  2. Johann Windt1,2,
  3. David Taylor1,2,
  4. William Moreau3
  1. 1 Sports Medicine, United States Olymic Committee, Colorado Springs, Colorado, USA
  2. 2 US Coalition for Prevention of Illness and Injury in Sport, Colorado Springs, Colorado, USA
  3. 3 Sports Medicine, Southern California University of Health Sciences, Whittier, California, USA
  1. Correspondence to Dr Dustin Nabhan, Sports Medicine, United States Olymic Committee, Colorado Springs, CO 80909, USA; Dustin.Nabhan{at}


Aim To describe injury and illness incidence during the 2018 Winter Olympic Games (WOG) by Team USA.

Methods A descriptive observational study. We used registered Team USA Olympic athletes’ electronic medical records to review preparticipation health histories and medical encounters immediately prior to and throughout the 2018 WOG. Medical encounters were defined as all medical services provided by a healthcare provider, including evaluation, treatment and prophylactic services. All medical conditions were described according to International Olympic Committee injury and illness reporting criteria.

Results Team USA included 134 men and 108 women, aged 18–39 years, who represented 17 sport federations. The 47 Team USA medical staff documented 1744 medical encounters on 242 registered athletes (7.2 medical encounters per athlete). Forty-seven illnesses (194.2/1000 athletes) and 32 time loss injuries (132.2/1000 athletes) were recorded during the Games.

Conclusions An injury surveillance programme consisting of an electronic preparticipation health history and surveillance of medical encounters during the WOG was used to describe the health status of Team USA. We noted limitations to the surveillance process that can be addressed at future events.

  • epidemiology
  • injury prevention
  • electronic medical records
  • informatics

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.


Medical surveillance is the first step in Professor Willem van Mechelen’s sequence of injury prevention research model.1 The International Olympic Committee (IOC) has implemented injury and illness surveillance programmes at all Olympic Games since 2008.2–7 The IOC’s surveillance programme has evolved from pen and paper surveys to electronic medical records (EMRs) during the Pyeongchang 2018 Winter Olympic Games (WOG). The IOC and its National Olympic Committee's (NOC) aim to better understand injury and illness patterns during the Olympic Games.7

The NOC uses injury and illness trends to strategically develop medical staffing plans by allocating medical credentials to the provider types that are most appropriate for specific sports.8 9 Surveillance data at the national governing body level have also informed injury prevention strategies in preparation for mass sporting events.10

Few publications describe the incidence of health problems of a nation’s team (NOC) during the WOG. Reporting of injury and illness incidence at the NOC level will help key stakeholders better understand the burden of illness and injury during the WOG. Therefore, we describe how epidemiological data collected from an EMR facilitated medical service delivery before, during and after the 2018 Pyeongchang WOG.


This was an observational study describing the United States Olympic Committee’s Medical Surveillance programme findings before, during and after the 2018 WOG. Confirmed Team USA athletes competing at the 2018 WOG were registered prior to competition within a cloud-hosted EMR system (GE Centricity Software; General Electric, Fairfield, Connecticut, USA). The registration process required each athlete to access a secure, web-based EMR patient portal where they provided demographic information, a preparticipation health history and consent regarding the release of their medical information.

The WOG included eight precompetition days (1 Feb–8 Feb) and 17 competition days (9 Feb–25 Feb). Our observation period included 3 days after the WOG (26 Feb–28 Feb) to ensure all medical encounters associated with the WOG were included. During this time, Team USA medical staff entered all medical encounters into athletes’ EMRs. Preparticipation health history forms were electronically delivered to each athlete on their selection to Team USA and before 15 January 2018 through a secure web-based patient portal. The preparticipation health history identified athletes with two primary characteristics: (1) any active health conditions (eg, injuries or illnesses, allergies or pre-existing medical conditions), to ensure that active health issues were addressed and that the athlete would be ready for games participation, and (2) current medication use, to determine the potential need or presence of a related therapeutic use exemption (TUE). Preparticipation health history responses were reported to the Team USA Sports Medicine team through a secure web-based visual analytics dashboard (Tableau Server V.10.3.1; Tableau Software, Seattle, Washington, USA) to identify and analyse important findings and to provide additional medical care or communication where indicated.

Medical encounters were defined as all medical services provided by a Team USA credentialled healthcare provider, including evaluation, treatment and prophylactic services. Team USA medical staff documented medical encounters at all venues, including the Team USA clinics, high performance centres, WOG polyclinic and regional medical facilities.

Team USA’s integrated multidisciplinary medical team included 47 credentialled healthcare providers with diverse medical specialties, including orthopaedics, family medicine, internal medicine, physical medicine and rehabilitation, cardiology, athletic training, chiropractic, massage and physical therapy. Credentialled providers were either assigned to (1) all Team USA athletes (n=12) or (2) to a specific sport discipline (n=35).

Team USA maintained medical clinics within the Mountain and Coastal Olympic Villages. These Team USA clinics evaluated, diagnosed and managed athlete health conditions, and provided therapeutic services when indicated. The service line included physical examinations, diagnostic ultrasonography, formulary, physiotherapy, manual therapy, taping and bracing. Athletes presenting with upper respiratory or gastrointestinal (GI) complaints with symptoms consistent with an infectious aetiology were evaluated with PCR rapid diagnostic testing (RDT) (Biofire, BioMerieux, Salt Lake City, Utah, USA). Medical treatment details were recorded in the EMR system.

Following the WOG, all medical encounters that occurred during the Games period were deidentified and extracted for analysis (Tableau V.10.3.1, Tableau Software). Medical encounters were classified by IOC illness and injury survey criteria.2 We defined time loss injury as a neuromusculoskeletal condition that resulted in time lost from sports participation. Encounters with a healthcare provider for neuromusculoskeletal conditions that an athlete sought medical attention for but did not result in loss of participation were classified as non-time loss medical encounters. This definition included all ongoing medical treatment and performance therapy encounters, including manual therapy, stretching, passive therapeutic modalities, physiotherapy and exercise therapy for pre-existing conditions or injury prevention. All injuries required medical attention. Illness was defined as a non-neuromusculoskeletal condition requiring medical evaluation, regardless of time lost.

Study procedures were approved by the institutional review board of Southern California University of Health Sciences.

Data analysis

Following the WOG, we extracted deidentified data from the EMR database. Medical encounters were filtered to include only athletes on the final Team USA WOG roster, treated by credentialled WOG providers during the precompetition and competition WOG days.

Basic descriptive analyses described total and relative frequencies for all medical encounters, injuries and illnesses, stratified by gender and sport discipline. Medical encounters were described as total and per-athlete encounters. Injuries were described as total injuries and cumulative incidence proportion (%). Illnesses were represented as total illnesses and per 1000 athlete days. Injury and illness incidence proportions were described per 1000 athletes.11 12

Differences in medical encounters between different sporting disciplines and genders were calculated using Poisson regression. Poisson regression was chosen to model medical encounter ‘counts’. Robust standard errors and incidence rate ratios were calculated to compare different sporting disciplines and men/women. All statistical analyses were carried out in R V.3.4.2 using the stats, sandwich and msm packages.


Team USA included 242 athletes (134 men and 108 women) with an average age of 27 years old (range 18–39). Registered athletes represented eight international federations: ski/snowboard, ice hockey, bobsleigh/skeleton (BOB), speed skating (SSK), figure skating, biathlon, curling and luge (LUG). Each international federation included several sport disciplines as described in table 1.

Table 1

Number of male and female athletes within each sporting discipline

Preparticipation health histories

All athletes received a link to complete an electronic health history questionnaire through a secure online portal by 15 January 2018 and completed the health histories prior to 1 February 2018. Five athletes reported active health issues for which a physician had previously excluded them from exercise. Twenty-eight athletes reported ongoing conditions that required medication or treatment. The United States Olympic and Paralympic Committee Sports Medicine Division acted as a ‘quarterback’ for all of these self-reported medical conditions and facilitated medical evaluations and referrals by each athlete’s assigned medical staff in preparation for the WOG.

Of the 242 athletes, 110 reported current medication use. There were 63 distinct substances reported that included both over-the-counter and prescription medication. Eight active TUEs were reported, and on nine occasions, athletes reported that they did not know whether a TUE was required. In these instances, the medical team contacted the athlete directly to advise and educate them on the WADA list and TUE application process.

There were 123 allergies reported, broadly categorised as medical (n=40), food (n=47) or environmental allergies (n=36). Sixty-six athletes reported at least one existing allergy, and 26 reported multiple allergies. Allergies reported by sport federation and gender are found in the supplementary material (online supplementary table S1).

Supplemental material

Daily reporting and medical encounters

In total, 47 Team USA medical staff members documented 1744 medical encounters on 175 of the 242 (73%) registered athletes, an average of 7.20 medical encounters per registered athlete. Women presented more frequently than men, with 9.1 vs 5.7 encounters per athlete (incident rate ratio (IRR)=1.9, p<0.001), respectively. Medical encounters broken down by sport discipline and gender are displayed in table 2. Compared with CUR as a reference category, gender-adjusted IRRs ranged from 1.1 to 30.6 (table 2). The highest relative encounter rates, adjusted for gender, were seen in BOB and LUG.

Table 2

Total and per-athlete medical encounters stratified by sport and gender


Thirty-two time loss injuries and 47 illnesses were documented during the WOG. Total injuries stratified by sport, location type and cause are displayed in table 3. Total injury counts by gender, as well as cumulative incidence proportion, are displayed in online supplementary table S2. Twelve per cent (n=28/242) of registered athletes sustained an injury resulting in time loss from sport (four athletes sustained two time loss injuries). Additional information about injury location, cause and type is displayed in the supplementary material (online supplementary table S2).

Table 3

Time loss injury characteristics (location, type and cause) stratified by sport


Nineteen per cent (n=46) of athletes received medical attention for an illness, and one athlete reported had two bouts of illness. Respiratory (34) and dermatological (6) complaints were the most frequent illnesses. There was one case of acute appendicitis. This required urgent hospitalisation for appendectomy and resulted in time lost from training, but not from competition. There were no other time loss illnesses. Illnesses encountered by each sport are presented in table 4, and the PCR testing results are displayed in table 5. Illnesses by sport and gender are included in online supplementary table S3.

Table 4

Total illnesses, illnesses per 1000 athlete days and illness characteristics stratified by sport federation

Table 5

Rapid diagnostic testing results

Medical treatments

Medical treatments included dispensing over the counter and prescription medications, providing manual therapy, taping and bracing, and prescribing therapeutic physical therapy modalities. Thirty prescription medications were provided to 22 Team USA athletes during the Games. Medications included non-steroidal anti-inflammatories (40%), antihistamines (17%), antimicrobials (10%), respiratory system drugs (10%), antipyretic/analgesics (17%), ear/nose/throat drugs (3%) and skin medications (3%).


Illness rates reported by Team USA were higher than those reported during previous WOGs, while injury incidence was within the range of previous reports. The incidence proportion was 194 illnesses and 132 injuries per 1000 athletes. Our data extend the knowledge base for medical teams planning for future WOG and for researchers in sports injury prevention.

Health conditions reported by preparticipation health history

The preparticipation health history identified 12% of athletes entering competition with an active health condition. Entering an event with a health condition is a known risk factor for sustaining an injury during a major championship.13 It is not known whether self-reported health status is the most valid or reliable way of collecting information on an athlete’s current condition. This method should be compared with other approaches (oral interview and medical chart review) to determine the best approach to screen for active health problems in preparation for a major event.

The self-reported use of medication (45%) by Team USA WOG athletes was similar to previous reports in athletes. A 2006 survey of Finnish elite athletes reported that 34% of elite athletes were taking medication at the time of the survey, and 74% were prescribed medication in the last calendar year.14 Of the 110 athletes who reported medication use, there were 9 cases of Team USA athletes reporting medication use without knowing the World Antidoping Agency Status. Inappropriate use of medication can result in inadvertent adverse analytical findings and/or side effects that have deleterious athletic performance.14 Healthcare providers are encouraged to review medication lists and to provide antidoping resources for athletes to minimise the risk of medication-related adverse events.

The high incidence of self-reported allergies in this population warrants additional investigation. Up to 50% of athletes have aeroallergen sensitivity, and symptoms of allergic rhinoconjunctivitis mimic those of upper respiratory infection.15 16 Allergy screening has been proposed as part of the preparticipation exam in elite athletes.17 Appropriate management of allergic burden may prevent symptoms arising that are commonly confused with upper respiratory infection.18

Injury incidence

Injury incidence reported by Team USA was similar to that reported at previous WOGs, 132 injuries per 1000 athletes. The IOC reported 111.8 injuries per 1000 registered athletes in Vancouver, and 140 injuries and per 1000 athletes in Sochi.6 The only NOC injury data available are Team Great Britain’s reports of 482 injuries per 1000 during the 2014 WOG.10 The sport disciplines with the highest injury rates were ice hockey, alpine skiing, freestyle skiing and snowboarding. This finding is consistent with previous IOC injury surveillance data, which have shown contact sports, high-velocity sports and acrobatic sports have relatively higher injury rates than other events in the Winter Olympic Programme.4 6 Injury surveillance projects such as the IOC’s Medical Surveillance programme provide international federations with information on risk of participation allowing them to make informed decisions on rule and policy changes to improve athlete health.

Illness incidence

The proportion of Team USA athletes that sustained an illness was 194/1000 athletes. This rate is in the upper range of previous reports of illness incidence at WOGs, which range from 51 to 196/1000 athletes. The IOC reported 72.1 illnesses per 1000 registered athletes in Vancouver and 89 illnesses per 1000 athletes in Sochi.6 Previous NOC reports of illness incidence during the WOG include Team Great Britain’s report of 196 illness per 1000 during the 2014 WOG and Norway’s 2010 WOG report of 51 illnesses per 1000 athletes.10 13 Variation in illness rates across WOG and NOCs may stem from variable compliance with surveillance programmes, environmental and regional factors at each WOG, technology use (eg, EMR and PCR testing) and different preventative medicine practices. Consistent collection of illness data with standardised definitions for reporting is needed to better understand the risks of participation for WOG athletes.

Illnesses were more prevalent than injuries for Team USA during the WOG, which is relatively uncommon in the previous WOG reports. Travel across multiple time zones increases risk of illness.19 Illness prevention efforts included evidence-driven hygiene kits that included hand sanitiser, mouthwash, soap, probiotic supplementation and advice on jet lag prevention. Nonetheless, illness significantly burdened this NOC’s athlete population. Further research may inform more effective illness prevention methods in athletes travelling across time zones for competition.19 20

Two specific infectious disease concerns threatened athletes and support staff at the 2018 WOG, influenza and norovirus. In the week preceding the WOG, many Team USA delegates arrived with upper respiratory infections and complaints of a sore throat. Using PCR RDT, we identified several infections to be influenza, despite an uncharacteristic symptom profile. Patients with influenza were isolated from the WOG community to minimise the spread of influenza to WOG participants and staff. The medical team believes that the absence of influenza diagnoses later in the games suggests this strategy may have succeeded in preventing widespread infections. Accurate diagnoses also enabled us to treat athletes with viral upper respiratory infections without antibiotic prescriptions.

The second infectious disease threat was a norovirus outbreak primarily isolated to the organising committee’s volunteer security staff outside the villages. This outbreak was identified by the local organising committee medical commission and was communicated to all delegations. Following this announcement, Team USA implemented a low threshold for PCR testing to diagnose patients with GI complaints during the WOG. All symptomatic GI patients identified through EMR were asked to participate in voluntary stool sample testing. The eight resultant GI tests identified one case of norovirus in a noncredentialled support staff. The infected individual was medically managed and isolated with no secondary cases reported.

Medical encounters for non-time loss conditions

Team USA medical staff documented 1744 medical encounters on 73% registered athletes, an average of 7.20 medical encounters per athlete. Medical encounters included many services performed for prophylactic or performance reasons, such as manual therapy and clinician-guided therapeutic exercise. We determined the high use of preventative and performance services inflates medical encounter records. In this population, the medical attention definition is a poor indicator of injury burden, and therefore we concluded the time loss definition better describes injury incidence. This distinction may need to be made at elite levels of sport, where there is high use of medical services in healthy individuals, as opposed to events where medical teams serve in a first aid/first responder role.

Women had significantly more medical encounters than their male counterparts, 9.1 vs 5.7 encounters per athlete (IRR=1.9, p<0.001). We are not aware of any other literature on this topic and we believe it is an interesting finding worthy of further investigation. If gender differences in medical care use are identified, this information could be used to prepare medical staffing plans for future events.

Limitations and challenges

The preparticipation health history depends on athletes self-reporting their health history. This method may not be the most accurate and athletes may have withheld information when completing these forms. Future research should investigate the accuracy of methods of health history data collection.

There are limitations in the use of electronic surveillance during sporting events. It can be challenging to train healthcare providers to understand and implement unfamiliar EMR and diagnostic coding systems during the Games. The authors suspect some providers may not have regularly reported injuries in a timely manner and did not always follow standards of care by recording patient encounters. Additionally, encounters provided by non-credentialled healthcare providers were not captured.

The WOG are 25 days of a 4-year training and competition cycle. This limited time window of capture represents only athletes who survived a training season to compete. This may under-represent the true burden of health conditions in each sporting discipline. Surveillance data that span the entire competitive seasons or athlete career provide better information on the risks of sport participation than competition data.

Data aggregation efforts depended on athletes having identical demographic and sport details in multiple electronic systems (eg, athlete roster and EMR). This was not always the case, for example, when athletes change their name after marriage. The EMR database categorised athletes by a code for international federation, not the sport discipline they compete in. It was challenging to describe injury and illness by specific event type (giant slalom, half pipe and big air) or physiological characteristics associated with their event (power, endurance and technical). These limitations further reinforce the need for standardised databases, definitions, coding systems and provider training as important components for medical surveillance efforts at mass participation events.

Future directions and conclusion

Medical surveillance data were collected in two ways: (1) athlete self-reported preparticipation health histories and (2) clinician-documented encounters from the EMR. Advances in direct patient communication may reduce communication barriers for athletes participating in mass sporting events. Secure mobile applications and messaging platforms may facilitate communication between athletes and their medical team regarding any changes in health status.21 Injury surveillance efforts relying on athlete self-reports can be effective and may be supplemented by provider feedback.21 22 Integrating these types of technology into EMR software may improve data collection capabilities.23

What is known?

  • Medical surveillance methods have been used to understand the risks of participation during the Winter Olympic Games (WOG).

  • There is a significant illness burden on WOG participants.

  • High-velocity and contact sports have relatively high injury incidence during WOG.

What is new?

  • Preparticipation health surveys reveal active health conditions that should be managed in preparation for major competition.

  • Rapid diagnostic testing (RDT) can be used to guide the management of athletes stricken with illness during competition.

  • The medical attention definition may result in inflated reporting of injury and illness occurrence as compared with the time loss definition.

How might it impact on clinical practice in the future?

  • Preparticipation health histories may be tailored to identify the need for specific resources to be targeted to athletes registered for participation in major competitions.

  • RDT may be considered as a supplementary diagnostic tool for events with high illness burden.

  • International sport federations and National Olympic Committee planning for participation in Winter Olympics may consider plans to mitigate injury risk in high-velocity and team sports.


The authors recognise the International Olympic Committee Medical and Scientific Commission and the US Coalition for the Prevention of Illness and Injury in Sport for supporting this paper.



  • Contributors DN, JW and WM conceived, planned and collected data for the project. DT and JW completed the data analysis. All authors participated in writing the manuscript. DN and WM were in charge of the direction of the entire project.

  • Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.

  • Competing interests General Electric is mentioned in the manuscript as the electronic medical records vendor. The company is also a US Olympic Committee and International Olympic Committee sponsor. This relationship has been described in a competing interests statement.

  • Patient consent for publication Not required.

  • Ethics approval This project was approved by the ethics board of Southern California University of Health Sciences.

  • Provenance and peer review Not commissioned; externally peer reviewed.

  • Data availability statement No data are available.