Development of a simple, reliable and universally accepted classification system for concussive injuries has long been a challenge. While more than 35 concussion severity scales have been published, none have been scientifically validated.1

At the First International Conference on Concussion in Sport the strengths and weaknesses of existing injury severity scales were considered; however, none of the scales were endorsed.2 The expert consensus recommendation arising from the meeting was that combined clinical measures of recovery should be used to assess injury severity and prognosis.2

At the Second International Conference on Concussion in Sport the concept of “simple” and “complex” concussion was introduced in an attempt to create a simple, clinically useful classification system to guide injury management.3 The distinction between simple and complex concussion related predominantly to the duration of symptoms, with simple concussion lasting less than 10 days and complex concussion lasting over 10 days. Other features, such as concussive convulsions, prolonged loss of consciousness (LOC), prolonged cognitive impairment or multiple concussions were also included in the definition of complex concussion.

We review the validity and clinical usefulness of the simple versus complex classification system for concussion in sport.

METHODS

Searches of Medline (ISI Web of Science, Pubmed) and SportDiscus databases were undertaken using keywords: “concussion”, “mild traumatic brain injury”, “head injury”, “sport”, “classification”, “simple concussion”, “complex concussion”, “pathophysiology”, “recovery”, “clinical features”, “symptom”, “balance”, “postural control”, “cognition”, “cognitive deficit”, “neuropsychological”, “neurocognitive”, “prognostic factors”, “recurrence”, “cumulative effect” and “injury severity”. The search was limited to English language and focused on articles published in the past 10 years. Searches were supplemented by reference lists from retrieved articles and related reviews as well as from the author’s own personal archives of references.

PATHOPHYSIOLOGY OF CONCUSSION IN SPORT AND IMPLICATIONS FOR CLASSIFICATION

The pathophysiology of concussion in sport remains poorly understood. Data derived from animal models of head injury suggest that linear acceleration or rotational shearing forces may result in short-lived neurochemical, metabolic or gene-expression changes.4

While the current expert consensus is that concussion represents a functional injury that recovers spontaneously,2 it remains unclear whether the condition reflects a unitary phenomenon with a linear spectrum of injury severity or whether different concussion subtypes exist.2 Although the current classification system divides concussion into two main injury subtypes, it assumes a linear spectrum of injury with an arbitrary cut-off at 7–10 days.3 An improved understanding of pathophysiology of concussion in sport is important to the future development of concussion severity scales.

IS THE SIMPLE VERSUS COMPLEX CLASSIFICATION SYSTEM FOR CONCUSSION IN SPORT VALID?

For the purpose of the discussion, the term “validity” is used to refer to “the degree to which the construct measures what it is suppose to measure”.5

The classification system proposed at the Second International Conference on Concussion in Sport was based on the expert panel’s collective clinical experience with concussive injuries.3 The term “simple concussion” was used to describe the common presentation of concussive injuries, where athletes are observed to recover uneventfully within 7–10 days of injury.3 The term “complex concussion” was used to describe cases that were delayed in recovery (that is, longer than 10 days), displayed specific clinical manifestations that may be associated with higher grades of injury severity (that is, concussion convulsions or prolonged LOC) or were complicated by factors such as repeated head injury.3

The proposed concussion classification system, therefore, provides a measure of concussion severity. This measure predominantly reflects the time course of recovery following concussive injury but also includes other clinical factors, which may alert the treating practitioner to a higher grade of injury or the potential for adverse outcomes.

What is the typical time course of recovery following concussion in sport?

The basic premise of the concussion classification system is that injuries typically resolve without complication over a period of 7–10 days. In considering the validity of the classification system, it is important to assess the literature regarding the time course of recovery following concussive injury. However, the main limitation is that currently there are no direct measures of recovery of brain function. Consequently, we are reliant on indirect clinical measures to monitor brain function following concussive injury. The most widely used and studied of these indirect measures include symptoms, balance deficits and cognitive impairment.

1. Recovery of symptoms

Monitoring of symptoms following concussion in sport is facilitated by the use of post-concussion symptom (PCS) scales.3 6 The scales consist of commonly reported concussion symptoms, which are typically rated on a 7-point scale from zero (absent) to six (severe). The results of prospective studies that have used PCS scales to monitor recovery following concussion in sport are summarised in table 1. The results consistently demonstrate resolution of symptoms in approximately 90% of concussed athletes by 7 days post injury.

2. Recovery of balance deficits

Increased postural sway has been demonstrated in concussed athletes using both sophisticated force plate technology14 and simple clinical tests of balance.12 14 Postural deficits have been shown to be most pronounced within 24 hours of injury and return to baseline levels over a period of 3–5 days post concussion.12 14 The results of prospective studies monitoring resolution of postural deficits following concussion in sport are summarised in table 2.

3. Recovery of cognitive impairment

Impairment of cognitive function is a hallmark feature of concussion in sport.2 Deficits are commonly observed in a range of domains, including information processing and decision making,15 motor function and reaction time,16 17 and memory.18 19

Following the independent pioneering work of Barth et al20 and Maddocks et al,15 limited neuropsychological test batteries have become a key component in the assessment of cognitive recovery following concussion in sport.2 Initially, paper-and-pencil tests such as the Digit Symbol Substitution Test and Trail-Making Test were used;15 20 21 however, more recently, these have been supplanted by computerised test platforms.19 22 23

The time course of recovery of cognitive function following concussion in sport is inconsistently reported in the literature. Some studies show transient neuropsychological deficits,11 19 24 while others demonstrate a prolonged decline in cognitive function.7 25 However, the common methodological flaw in these studies is that they estimate a player’s pre-concussion ability and compare their post-concussion performance with matched control players. This may reduce the sensitivity for detecting subtle changes, which may still be within the normative range.26 The results of prospective studies where cognitive performance is compared with the athlete’s own pre-injury baseline reveal a more consistent time course of recovery. These studies are summarised in table 3 and demonstrate recovery of cognitive function in the majority of athletes over a period of 7–10 days post injury.

Overall pattern of recovery

When taken together, the results of prospective cohort studies suggest that in the majority of cases of concussion in sport, clinical features resolve within 7–10 days of injury. However, in a small proportion of cases, symptoms and/or cognitive deficits may be prolonged. These findings are consistent with a recent systematic review of recovery following concussion in sport,28 and provide support to the classification of concussion as “simple” and “complex” based on time course of injury.

It is important to note, however, that the data on time-course of injury is derived largely from single injury sports in adult male participants. It remains unclear whether the model of recovery holds true for sports such as boxing, where repetitive head trauma is the norm rather than the exception, or specific populations such as young (that is, <18 years) or female athletes. Furthermore, studies using more direct measures of brain function, such as sophisticated functional neuroimaging techniques, may provide important information on recovery of concussive injuries in the future.

Are specific clinical manifestations of concussive injury associated with higher grades of injury severity?

Specific clinical manifestations of concussion that were included in the definition of complex concussion were concussion convulsions and prolonged LOC. The implication is that these factors are associated with higher grades of injury severity following concussion in sport.

1. Association between convulsions and concussion severity

Concussive convulsions typically occur within seconds of impact and last for a short period of time.29 30 In their full form, concussion convulsions comprise an initial period of tonic stiffening followed by myoclonic jerks of all limbs.29 30 Other motor phenomena, such as asymmetrical posturing and versive head movements, have also been described. 29 30 Few studies exist in the literature investigating the clinical outcome of concussive convulsions. In a large-scale retrospective series in Australian football, concussive convulsions were demonstrated to be benign with no adverse clinical, cognitive or neuro-imaging outcomes observed either at the time of injury or at long-term follow-up.29

2. Association between LOC and concussion severity

Traditionally, LOC has featured prominently in published, concussion severity scales as a marker of more severe injuries.31 32 The association between LOC and injury severity has largely been extrapolated from models of moderate-severe traumatic brain injury, where immediate impairment of consciousness is a hallmark of primary diffuse head injury, and the depth and duration of LOC correlate with the severity of brain injury.33 However, prospective cohort studies on concussion in sport have consistently demonstrated that brief LOC (that is, <60 s) does not reflect injury severity or predict time to recovery.8 34

More recently, Pellman and colleagues analysed cases of prolonged recovery (that is, ⩾7 days out from sport) from a large-scale concussion programme conducted in professional American football.35 The results demonstrated that LOC for a period >1 minute was associated with prolonged time to return to play. This study provides preliminary evidence for caution in cases of concussion with prolonged LOC; however, further research is required.

3. Association of other clinical factors and concussion severity

Prospective cohort studies on concussion in sport have identified a number of other clinical factors that may be associated with higher grades of injury severity:

• a) presence and duration of post-traumatic amnesia35 36

• b) total number of post-concussion symptoms36^–38

• c) presence of specific post-concussive symptoms such as headache39 and fatigue or fogginess.35 40

At present, further research is needed to better elucidate the relation between various clinical factors and severity of concussion in sport.

Is repeated concussive injury associated with poor outcome following concussion in sport?

The possible inclusion of repeated concussive injury in the category of complex concussion relates to a longstanding and widely held belief that repetitive head trauma is associated with cumulative deficits in brain function, which may manifest as chronic traumatic brain injury.41 42

The results from the current literature are conflicting regarding the effect of repetitive head injury on brain function. While some studies demonstrate impaired cognitive performance in athletes reporting a history of two or more concussive injuries when compared with controls,24 others have failed to demonstrate any significant effects of self-reported past history of concussion on cognitive performance.43 However, the major limitation of these studies is that they rely on retrospective recall of previous concussive injuries, which is known to be inaccurate and unreliable.44

Furthermore, many of the studies demonstrating an association between concussion history and prolonged deficits in cognitive function are cross-sectional studies in which retired players are compared with “normal” cohorts.45 46 The use of a cross-sectional design fails to control for important variables that can impact on cognitive performance, such as age, education, alcohol use and other socio-economic factors26 and, thereby, makes accurate interpretation of the results difficult.

Few well-designed prospective studies exist in the literature. In 2003, Porter monitored a small cohort of amateur boxers over a period of 9 years and found no evidence of cognitive decline on a limited paper-and-pencil test battery.47 Similarly, in a prospective study using a computerised test battery to monitor brain function, Moriarity and colleagues demonstrated no evidence of short-term cognitive impairment in 82 amateur boxers participating in a 7-day tournament.17 Conversely, in a recent prospective study of collegiate athletes, recovery of balance function on sophisticated visual field motion tests were significantly delayed in athletes following a second concussive injury.48

Overall, the relations between repeated head trauma and potential for cumulative deficits in brain function remains unclear. The association may be complicated by genetic factors specific to the individual athlete. In preliminary research, apolipoprotein-epsilon 4 genotype has been demonstrated to be associated with outcome following repeated head trauma in boxing49 and professional American football.50 Despite the methodological limitations of these studies, they provide preliminary evidence of a complex inter-relationship between concussion, genetics and the risk of cumulative damage.

IS THE SIMPLE VERSUS COMPLEX CLASSIFICATION SYSTEM FOR CONCUSSION IN SPORT USEFUL?

The main purpose of the concussion classification system was to provide a clinical framework to assist in the identification of more severe injuries or potential complications and, subsequently, guide management of concussive injuries. For “simple” concussions, no specific interventions were recommended over-and-above ensuring that the athlete rested until all clinical features of the concussive injury had resolved.3 Conversely, for “complex” concussions a multidisciplinary approach to management was recommended, including involvement of doctors “with specific expertise in the management of concussive injuries” and consideration given to the use of formal neuropsychological testing.3

For the classification system to facilitate clinical management of concussion in sport, the distinction between simple and complex should be made around the time of injury. However, in its current form, the classification system lacks prognostic markers of injury severity. Consequently, concussive injuries can only be classified retrospectively (that is, once they have recovered or 10 days has elapsed). Therefore, any specialised approach to the more difficult cases (that is, complex concussion) would be expected to delayed by at least 7–10 days.

CONCLUSION AND RECOMMENDATIONS REGARDING CLASSIFICATION OF CONCUSSION IN SPORT

Although there is some validity to a simple versus complex division, with clinical features of concussion resolving 7–10 days after injury in the majority of cases, the critical feature lacking is the ability to predict injury severity at the time of injury. Further concerns relate to the use of the term “simple concussion” particularly as the long-term effects of repeated head injury remain unclear.

It is recommended that until further information about the pathophysiology of injury, prognostic factors for more severe injuries and risk factors for poor outcome is known, concussion in sport should be considered as a single entity. Consequently, the terms “simple” and “complex” should be removed and may be replaced by a list of potential “modifying” factors. The modifiers, when present at the time of injury, may alert the treating practitioner to the possibility of a more severe injury or increased risk of poor outcome. Examples of potential concussion “modifiers” may include intrinsic factors (for example, age, gender, previous concussive injuries, specific genotype), injury specific factors (for example, prolonged LOC, post-traumatic amnesia, concussive convulsions, presence of fatigue/fogginess) or extrinsic factors (for example, type of sport, position played).