Article Text

Evidence-based approach to revising the SCAT2: introducing the SCAT3
  1. Kevin M Guskiewicz1,
  2. Johna Register-Mihalik2,
  3. Paul McCrory3,
  4. Michael McCrea4,
  5. Karen Johnston5,
  6. Michael Makdissi6,
  7. Jiří Dvořák7,
  8. Gavin Davis8,
  9. Willem Meeuwisse9
  1. 1Department of Exercise and Sport Science, Matthew Gfeller Sport-Related Traumatic Brain Injury Research Center, University of North Carolina, Chapel Hill, North Carolina, USA
  2. 2Clinical Research Unit, Emergency Services Institute, WakeMed Health and Hospitals, Raleigh, North Carolina, USA
  3. 3The Florey Institute of Neuroscience and Mental Health, Heidelberg, Australia
  4. 4Brain Injury Research Institute, Departments of Neurosurgery and Neurology, Medical College of Wisconsin, Wisconsin, USA
  5. 5Neurosurgeon, Division of Neurosurgery, University of Toronto; Concussion Management Program Athletic Edge Sports Medicine, Toronto, Canada
  6. 6The Florey Institute of Neuroscience and Mental Health, Melbourne Brain Centre, Austin Campus, Melbourne, Australia
  7. 7FIFA Medical Assessment and Research Center, Zurich, Switzerland
  8. 8Department of Neurosurgery, Austin and Cabrini Hospitals & The Florey Institute of Neuroscience and Mental Health, Melbourne, Victoria, Australia
  9. 9Sport Injury Prevention Research Centre, Kinesiology and Hotchkiss Brain Institute, University of Calgary, Calgary, Alberta, Canada
  1. Correspondence to Dr Kevin M Guskiewicz, Department of Exercise and Sport Science, Matthew Gfeller Sport-Related Traumatic Brain Injury Research Center, University of North Carolina at Chapel Hill, Chapel Hill, NC 27599-8700 USA; gus{at}


The Sport Concussion Assessment Tool 2 (SCAT2), which evolved from the 2008 Concussion in Sport Group (CISG) Consensus meeting, has been widely used internationally for the past 4 years. Although the instrument is considered very practical and moderately effective for use by clinicians who manage concussion, the utility and sensitivity of a 100-point scoring system for the SCAT2 has been questioned. The 2012 CISG Consensus Meeting provided an opportunity for several of the world's leading concussion researchers and clinicians to present data and to share experiences using the SCAT2. The purpose of this report is to consider recommendations by the CISG, and to review the current literature to identify the most sensitive and reliable concussion assessment components for inclusion in a revised version—the SCAT3. Through this process, it was determined that important clinical information can be ascertained in a streamlined manner through the use of a multimodal instrument such as the SCAT3. This test battery should include an initial assessment of injury severity using the Glasgow Coma Scale, immediately followed by observing and documenting concussion signs. Once this is complete, symptom endorsement and symptom severity, neurocognitive function and balance function should be assessed in any athlete suspected of sustaining a concussion. There is no evidence to support the use of a composite/total score; however, there is good evidence to support the use of each component (scored independently) in a revised assessment tool.

  • Athletics
  • Concussion
  • Contact sports
  • Head injuries
  • Neurology

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.


One of the major challenges in the medical management of concussion is that there is no single ‘gold standard’ for assessing and diagnosing the injury. At its core, concussion firmly remains a clinical diagnosis based on a combination of symptoms (somatic, cognitive and neurobehavioral), physical signs and impairment in cognitive function. To make things more complicated, the clinical features of concussion are often variable and can present in a delayed fashion following injury. Despite evolving imaging techniques (diffusion tensor imaging and functional MRI (fMRI)), concussion is still considered a functional injury, and not one that can yet be structurally identified on imaging.

Prior to the current era, concussion severity scales utilised classical injury characteristics such as loss of consciousness (LOC) and/or post-traumatic amnesia to establish diagnosis and prognosis, but recent studies and recommendations have diminished the value in this approach.1 Instead, over the past decade, expert groups have recommended a multifaceted approach to concussion management that emphasises the use of objective assessment tools aimed at capturing the spectrum of clinical signs and symptoms, cognitive dysfunction and physical deficits.1–4 More contemporary studies have concluded that these components in the concussion evaluation are more sensitive to the injury than using any one component alone.5–8 It is essential for clinicians to fully understand the necessary components of a multifaceted concussion evaluation, both at the time of the concussion and during serial assessments, prior to making a return-to-play decision. Overlooking elements of the evaluation process may result in making a premature return-to-play decision and exposing the athlete to repeat injury or postconcussion syndrome.

In 2001, the Concussion in Sport Group (CSIG) recommended a new paradigm of individualised assessment of recovery9 and at the 2004 meeting in Prague proposed the Sport Concussion Assessment Tool (SCAT) based on expert consensus of the best measures to assess concussion available at that time.4 In part, the aim was to standardise the assessment across sports internationally. In 2008, the CISG modified this tool based on available evidence and developed two forms (1) the SCAT2—designed for medical practitioners to assess concussion in greater detail and (2) the PocketSCAT2—for lay use to assist with the sideline recognition of screening for concussion.1

The purpose of this report was to review the current literature to identify the most sensitive and reliable concussion assessment components for inclusion in the revised version—the SCAT3. This review is focused on adult concussion and the SCAT2 tool. The paediatric SCAT3 assessment will be the subject of a separate report.


PubMed, MEDLINE, Psych Info and Cochrane Library databases were reviewed using the MeSH keywords ‘concussion’ combined with ‘athletics’ and refined adding the keywords ‘assessment ’, ‘balance’, ‘cognitive’, ‘symptoms’, ‘Sport Concussion Assessment Tool’ and ‘prognosis’ to yield articles relevant to updating and reviewing the components of the SCAT2.

Search results

The initial search terms of ‘concussion’ and ‘athletics’ yielded 1126 articles. When these terms were refined using the term ‘assessment’ the total was 254 articles. These 254 articles formed the basis of the review and were further refined by topics specific to the components of the SCAT2 including: ‘Sport Concussion Assessment Tool’ (7 articles), ‘balance’ (36 articles), ‘symptom scales’ (12 articles), ‘SAC’ (9 articles) and ‘prognosis’ (15 articles). An additional 20 articles were added through hand searching of reference lists from relevant articles and input from content experts. These articles were reviewed for application to updating and providing evidence for the SCAT 2, resulting in the 77 articles that were used in the current evidence review.

Evidence review for components of the SCAT2

The SCAT2 combines aspects of several previously published concussion tools into eight components designed to assess concussion symptoms (number endorsed and severity), cognition (Sideline Assessment of Concussion or SAC and Maddocks questions), balance (firm conditions of the Balance Error Scoring System or BESS), Glasgow Coma Scale (GSC) and some neurological signs (physical signs, coordination). Each of the eight components are scored and recorded. The SCAT2 total score ranges from 0 to 100 points, with lower scores indicating poorer performance (does not include the Maddocks questions or the symptom severity score). The tool is designed to be used serially after a concussion and includes a ‘score card’ designed to track the athlete's performance on each of the components (except for the Maddocks score, which pertains to initial sideline diagnosis only) over time during the recovery period. The SCAT2 embeds the SAC and modified-BESS testing within its matrix, which was felt to be important given the extensive use of these particular tools in North America and validation studies supporting their use in concussion assessment. Each component of the SCAT2 was carefully considered from both an evidence-based and practical standpoint prior to recommending its inclusion in the tool.1

Glasgow Coma Scale

The GSC (range 3–15) is a commonly used initial assessment score that has been widely accepted in neurosurgery, emergency medicine and acute care/trauma as a marker of the severity of head injury in adults.10–13 Because of the scored items on the scale, it is often not included in mild brain injury studies, as the very nature of a mild injury would cause a patient's GCS score to be at least 14, and in most cases 15. It is widely accepted as one of the best predictors of outcome following more moderate to severe injury and as such is appropriate for identification of more severe injury.14 ,15

Maddocks score/questions

The Maddocks questions are questions of orientation, particularly recall of recently acquired events. These questions have been shown to be more sensitive to concussion than standard orientation questions and can give important clinical information at the time of injury.16 The Maddocks questions are not included in the total SCAT2 score.

Physical Signs score

The Physical Signs score (max=2) is based on the presence of balance deficits or instability and/or LOC observed at the time of injury. There is no validation of the physical signs score being linked to deficits following concussion, with LOC alone often debated as a severity marker following concussion.17–19 However, recent studies have illustrated that LOC17 and amnesia20 may also be prognostic indicators for concussion. Balance deficits or instability are often observable in patients following concussion and the presence of these deficits may be an indicator of vestibular disruption.6 ,7 ,21 ,22 Recently, dizziness reported on field has also been associated with protracted recovery.23

Coordination score

The coordination score (max=1) is based on the patient's ability to perform a finger-to-nose coordination task as observed by the medical personnel scoring the SCAT2. This score provides the examiner with the ability to assess the patient's overall coordination and ability to follow instructions. Schneiders et al24 recently published a study examining three motor tasks in a sample of 172 healthy men and women aged 16–37 years. They measured time to complete five finger-to-nose repetitions, as measured on the SCAT2 and found small trial and gender effects for the finger-to-nose task. Intrarater reliability calculations made using the intraclass correlation coefficient (ICC) demonstrated higher reliability for the finger-to-nose and tandem gait tests than for the single-leg stance tests used in this study.

Symptom Assessment—Post Concussion Symptom Score (PCSS)

The symptom score is comprised of a 22-item postconcussion symptom scale using a seven-point Likert scale rating. This scale allows for assessment of symptoms endorsed, along with a severity score. The scale has been used in a variety of settings to assess concussion25–27 and a graded symptom checklist is the most commonly used postconcussion assessment tool.28 This symptom scale has been shown to be reliable and valid for the assessment of both symptom presence and severity (table 1).5 ,29–34 The symptom assessment is important both initially and in serial assessments following concussion to understand the course of symptom presentation and resolution following the injury. The SCAT2 total score includes the number of symptoms endorsed, but not the severity of symptoms. The maximum symptom score included in the SCAT2 total score is 22 (which is obtained by 22 minus the number of symptoms endorsed). The symptom severity is obtained by summing the rated symptom score for each symptom (maximum score of 132).

Standardised Assessment of Concussion

The SAC is a short test of mental status that has been validated in several studies for use in the assessment of sport-related concussion.7 ,35–42 and has a maximum score of 30. The SAC has four major scored components included to assess orientation, immediate memory, concentration and delayed recall. The SAC is shown to be sensitive to concussion within the first 48 h following injury7 ,36 ,38 and can be easily administered on the sideline or in a clinical setting (table 1). Because of its ease of use and practicality, the SAC provides an objective measure of changes in cognitive functioning following concussion.

Balance examination

The balance component in the SCAT2 includes only the three firm conditions of the BESS, and not the three conditions on foam. Most of the research to date on the BESS includes all six of the original conditions. The BESS in its entirety has been studied extensively in the concussed population21 ,38 ,43–45 and has been shown to have low to high intertester and intratester reliability, depending on the source of the study.46–49 It has also been shown to be sensitive to concussion, especially when used with the first 3 days following injury (table 1).38 ,50 Future research is needed on the firm conditions alone to assess if these conditions are an adequate assessment of balance.

Additionally, assessing balance in challenging environments (ie, ice hockey, figure skating, etc) should be further explored. Finnoff et al48 recently investigated the intra-rater and interrater reliabilities of the BESS in 30 non-concussed athletes. Intra-rater and interrater reliability ICC scores were 0.74 and 0.57, respectively, for the total BESS score and varied for individual BESS subcategories. Minimum detectable change scores of 7.3 (intra-rater) and 9.4 (interrater) were calculated for the total BESS score. The intrarater reliability ICC scores in this study were lower than those previously reported. The authors of this study concluded that total BESS score was not reliable, and that only firm single-leg, firm tandem and foam double-leg subcategories were valid when the same scorer is used, while only the firm single-leg subcategory is valid when different scorers are used.

Schneiders et al24 also examined time in single-leg stance on firm and foam surfaces, which is similar to the single-leg component of the BESS although scored differently, and time to complete a 3 m tandem gait task, which is not included specifically as part of any formal sideline assessment tool, but can reasonably be incorporated into the sideline examination of a concussed athlete. This study identified gender and order effects for the single-leg stance and tandem gait tasks. These results also highlight the need to consider individual factors such as age and gender when interpreting postinjury test results, as well as environmental factors that can affect the reliability of test comparisons. While Schneiders et al24 found poor reliability for single-leg postural stability tests, their timed single-legged stance tests differed in scoring from the BESS, which is used more commonly in the assessment of concussion.

A large body of literature exists to support the use of BESS testing in the evaluation of concussed athletes; however, these new reliability measures provide some evidence as to the most valid components to use postinjury.

Factors affecting SCAT2 measures

Although to date no studies have been conducted on the many factors that may affect the SCAT2 as a whole, various studies have addressed factors that may affect the components included in the SCAT2. Specifically, there are numerous studies suggesting that baseline rates for many of symptoms (ie, base rates) included in the checklist are relatively high among healthy, non-concussed athletes.51 ,52 In addition, factors such as gender52 dehydration53 and oral contraceptive use in women54 may play a role in symptom reports. Furthermore, mode of administration (interview vs checklist) of the symptom checklist is also a factor. As against using a concussion-symptom-specific checklist compared with just asking the athlete ‘how they feel’ will result in a higher number of symptoms endorsed.55 ,56 Clinicians should be mindful of administration mode of the checklist when considering both baseline and postinjury scores.

Concerning balance (the firm conditions of the BESS), another body of the literature suggests that training57 fatigue,58–60 number of administrations,47 ankle injury,61 sport played62 and testing environment63 can have a significant effect on the number of errors an individual commits during this task. Schneiders et al24 also observed gender and order effects for the single-leg stance and tandem gait tasks. These results underscore the need to consider individual factors such as age and gender when interpreting postinjury test results, as well as environmental factors that can affect the reliability of test comparisons. Clinicians should be mindful of these factors prior to performing the BESS stances and consider these factors in the interpretation of the score. The SAC is not affected by external factors in as many ways as symptoms and balance. In addition, practice effects are not as prominent for the SAC as they are the BESS.64 Age may, however, affect total score on the SAC (eg, child vs adolescent vs adult results). The GSC score, because it is typically an indicator of more severe trauma, has not been studied with respect to the above variables, but due to the nature of the scale would likely not be affected. Although the physical signs score has also not been studied concerning the effects of these factors outside of concussion, LOC could occur with heat illness, and balance problems/unsteadiness may result from a variety of issues that may present in a sport setting. The coordination examination, however, would likely remain unaffected, except in the presence of an associated upper limb injury. Each of the factors mentioned may complicate the concussion assessment and medical personnel should be aware of these potential confounding factors during the evaluation and management process.

Table 1

Psychometric properties of scaled components of the SCAT2

Setting-specific modifications to the SCAT2

Other sports, such as rodeo and bull riding,67 have attempted to modify the SCAT2 for their specific needs. A modified Ebel procedure was employed to content validate the rodeo SCAT2. Content validation using this method includes experts agreeing on the importance of each item that comprises the rodeo SCAT. This three-stage process involved: (1) face validation by a local committee; (2) initial expert consensus measurement via distance and (3) a face-to-face discussion for items that did not originally achieve 80% consensus of the group. Experts were chosen from the Canadian Professional Rodeo Sport Medicine Team (Canada) and the Justin Sports Medicine Team (USA). Twenty-seven of a total possible 68 items achieved 80% consensus in the second stage. In the third stage, 4 of the 68 items were removed with consensus from the expert group. All remaining items achieved 80% consensus for inclusion. In summary, the rodeo SCAT is content valid and thus, appropriate for use in the sport of rodeo context or environment.

Other authors have queried specific aspects of scoring and/or the serial assessment template.68 This paper, describing the SCOAT or ‘Sports Concussion Office Assessment Tool’, is based on the experiences of a South African sports physician group.68 The authors propose a simpler scoring system to that of the SCAT2 which would only allocate points for significant findings (eg, symptoms that are present or errors/mistakes made during the tasks). The SCOAT would also weigh symptoms on severity; include space for recording how the concussed player was managed (including prescribed medications and referrals); and include a final checklist before allowing return to play. The authors recommend the use of this modified instrument for conducting serial assessments, but conclude that it still needs to be validated for this use.

SCAT and SCAT2 normative data studies

Shehata et al69 published baseline values on the original SCAT for 260 healthy collegiate athletes. Of these athletes, 58.8% reported the presence of at least one symptom at baseline, and the overall mean PCSS score was 4.29. The most commonly reported symptoms were fatigue/low energy (37%), drowsiness (23%), neck pain (20%), difficulty concentrating (18%) and difficulty remembering (18%). Nearly 96.4% of the athletes had Standardised Assessment of Concussion immediate recall scores of 5, but only 36.9% had a delayed recall score of 5. Of the athletes, 91.6% were able to recite the months in reverse order and 51% were able to complete strings of six digits in reverse order. Gender differences were present, as well as differences between those athletes with and without prior history of concussion, for each component of the test. These results highlight the need to consider postinjury results in the context of an athlete's own baseline, as many common symptoms of concussion are non-specific and are present in healthy athletes at baseline.

In an adolescent sporting population, Valovich-McLeod et al70 looked at baseline normative data. Interscholastic athletes were administered the SCAT2 during a preseason concussion baseline testing session. There were 1134 high school athletes (872 men and 262 women) who participated. The SCAT2 total score across all participants was 88.3±6.8 (range 58–100); skewness was –0.86±0.07 and kurtosis was 0.73±0.14. Male athletes scored significantly lower on the SCAT2 total score (p=0.03; 87.7±6.8 vs 88.7±6.8), and 9th graders (86.9±6.8) scored significantly lower than 11th (88.7±7.0) and 12th (89.0±6.6) graders (p<0.001). Athletes with a self-reported concussion history scored significantly lower on the SCAT2 total score than those with no concussion history (p<0.001; 87.0±6.8 vs 88.7±6.5). These data provide representative scores on the SCAT2 in adolescent athletes and show that male athletes, 9th graders, and those with a self-reported concussion history scored significantly lower than female athletes, upperclassmen or non-concussed peers.

Discussion and recommendations

Based on the available evidence and practical considerations, the SCAT2 is a useful and practical concussion assessment tool. However, several modifications should be considered for refinement and inclusion in the SCAT3. There is no evidence to support the use of a composite/total SCAT2 score (eg, 82/100); however, there is good evidence to support the use of the following components (scored independently) in a revised assessment tool.

The order and content of the SCAT2 should be modified to include the following in the SCAT3:

  • Physical or Objective Signs: In addition to LOC and balance/instability, other objective markers may improve concussion recognition. These should include amnesia, disorientation, blank/vacant look and other visible signs of head/facial trauma. As developed and recommended by the NHL/NHLPA Concussion Working Group, 2011 (personal communication, Ruben Echemendia, Chair NHL Concussion Working Group).

  • GCS: This should be conducted immediately at the time of suspected injury to increase the likelihood of recognising moderate to severe brain injury. This need not be completed if the athlete has walked to the sideline and is clearly conscious and responsive.

  • Maddocks Questions/Score: These should be asked to assess orientation and memory at the time of injury only. This becomes less important for serial evaluations of the SCAT3.

  • Symptom Assessment—PCSS: This may be administered by the clinician or completed by the athlete, with or without parental input for the younger athletes. It is important to capture the method of recording, the number of symptoms endorsed, as well as the symptom severity, to establish a resolution pattern following the concussion.

  • SAC: No modifications are recommended to the SAC embedded in the SCAT3, but slight modifications will be made to the SAC in a child version, the ChildSCAT3 (see accompanying article).

  • Modified BESS: In addition to the three stances on a firm surface (double-leg, single-leg and tandem); adding the foam stances as indicated on the SCAT3 will improve sensitivity. In the absence of the three foam stances, adding a timed tandem gait test24 ,71 ,72 (with a pass/fail score) may improve the sensitivity of balance testing. However, more research is needed in concussed athletes to establish the efficacy of the tandem gait.

The Concussion Advice Section should be revised to reflect recent evidence and for clarification.

  • Because research has shown that recovery time is variable, we recommend replacing, ‘It is expected that recovery will be rapid’ with ‘Recovery time is variable across individuals’.

  • Because research has recently recommended placing more emphasis on mental/cognitive rest, we recommend changing the sentence concerning rest to: ‘Rest (physically and mentally), including training or playing sports until medically cleared’.

  • Because of the emphasis on a graduated return-to-play progression, we recommend changing the sentence concerning return to play to: ‘When returning athletes to play, they should follow a stepwise medically managed program, with stages of progression’.


Our review of the literature on the brief sideline assessment tools suggests that a variety of measures should be employed following concussion to provide a more complete clinical profile for the concussed athlete. Important clinical information can be ascertained in a streamlined manner through the use of a multimodal instrument such as the SCAT3. This test battery should include an initial assessment of injury severity using the GCS, immediately followed by observing and documenting concussion signs. Once this is complete, symptom endorsement and symptom severity, neurocognitive function and balance function should be assessed in any athlete suspected of sustaining a concussion. It is recommended that these latter steps be conducted following a minimum of 15 min rest period on the sideline to avoid the influence of exertion or fatigue on the athlete's performance.

Future research should consider the efficacy for inclusion of vision tests such as the King Devick Test73 ,74 and reaction time tests such as the clinical reaction time test.75–77 Recent studies suggest that these may be useful additions to the sideline assessment of concussion; however, the need for additional equipment may make them impractical for sideline use.


View Abstract

Supplementary materials

  • Supplementary Data

    This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.

    Files in this Data Supplement:


  • Contributors All the coauthors contributed to the conceptual design and writing of the manuscript.

  • Competing interests See the supplementary online data for competing interests (

  • Provenance and peer review Commissioned; internally peer reviewed.

  • ▸ References to this paper are available online at