19 “if in doubt, sit them out” Investigations There are no reliable radiological or blood investigations that assist with the diagnosis of uncomplicated concussion15 . Where symptoms persist for a prolonged period of time (more than 14 days for adults and more than 4 weeks for children) or where there is evidence of deteriorating neurological function, CT or MRI of the head may be indicated to exclude other serious pathologies such as fracture or intracranial bleed. There are no biomarkers which indicate the presence or otherwise of uncomplicated concussion15 . Neurocognitive testing can form part of the clinical assessment process but such testing in isolation is not sufficient to diagnose or exclude concussion41 . Predictors of clinical recovery A number of ‘concussion modifiers’, including pre-injury factors and initial injury severity indicators such as loss of consciousness or amnesia, have been studied to assess their association with prolonged recovery of symptoms or worse outcome3,16 . The evidence relating to the impact of many of these factors on prolonged recovery is inconclusive16 . Overall, the most consistent risk factor for slower recovery is acute and subacute symptom burden. While there is no clear evidence that loss of consciousness or post-concussive seizure is associated with a worse outcome, prolonged loss of consciousness (more than one minute) or prolonged seizure (more than one minute) following head trauma should raise suspicion of a more serious injury such as intracranial haemorrhage. Several studies have shown previous concussion to be a risk for further concussion and the clinician should take a conservative approach where there is a history of repeated concussion or concussions occurring in close time proximity16,49 . Recurrent concussion is sometimes associated with reduced threshold for concussion and this should alert the clinician to potential increased vulnerability to further concussive episodes. Children take longer to recover from a concussion injury20,21 . Age-appropriate assessment and a more cautious approach to return to play or sport should be adopted with children and adolescents aged 18 and under19,46 . The assessing clinician should be mindful of comorbidities including pre-existing mental illness such as depression, migraine, learning disability, ADHD or sleep disturbances. Similarly, a high level of vigilance should be observed in a situation where the athlete is taking medication such as psychoactive drugs or anticoagulant medication. The majority of concussions in adults resolve within 10 – 14 days3 . Medical investigation and re‑evaluation to assess mental and cognitive health, and the development of an individualised management plan are required where there are a high number of concussive symptoms, the concussive symptoms are of high severity or the symptoms last longer than 14 days for adults or longer than 4 weeks for children3,50 . Special considerations in concussion The issue of concussion has received significant media attention in recent years. The focus of a large part of this attention has been on chronic traumatic encephalopathy (CTE), particularly in retired athletes from the National Football League in the USA23 . CTE can only be diagnosed by post-mortem observation, and it is described by McKee et al51 as a ‘progressive neurodegeneration clinically associated with memory disturbances, behavioural and personality change, Parkinsonism, and speech and gait abnormalities’51 . Neuropathological criteria for the diagnosis of CTE have recently been defined at a consensus meeting to be the abnormal accumulation of hyperphosphorylated tau protein in the brain52 . Some researchers have presented preliminary evidence that repeated head trauma causes the condition51 , but there is significant selection bias in many of the reported cases23 . The link between sport-related concussion and CTE is based on low-level evidence. Research is limited to case reports, case series and retrospective analyses which cannot adequately determine causality or risk factors. The potential contribution of confounders, such as genetic predisposition, psychiatric illness, alcohol and drug use or co-existing dementia, is not adequately accounted for in the current literature53 . While there is significant concern about CTE and its possible relationship with concussion, it is important to note that no causative link has been clearly established. Recent public health concerns about CTE and the desire to prevent these complications have, to a large extent, driven the increased focus on developing best-practice guidelines for the identification, diagnosis and management of sport-related concussion. Further research is needed to understand what type of trauma,