Table 1

Articles examining factors to be considered on return to school following a concussion

ArticleStudy design, duration, countryParticipants (n, age, sex)Exposure/intervention (Definition)Outcome (Definition)Results (including statistical outcomes)Main limitationsStudy quality assessment (DB score)Level of evidence
Baker et al
23
Retrospective, descriptive study (telephone interview of cohort)
2010–2012
USA
Student athletes (n=91),
aged 13–19 years
SRC observed by athletic trainer and assessed by sports medicine physician/telephone follow-up/SCAT2, BCTT and ANAM or ImPACTFactors associated with school difficulties following concussion and school days missed
  • Problems with RTS=35/91 (38.5%)

  • Age, gender, previous concussions not associated with school problems

  • Days to recover: >10=57%, >21=29%

  • Recovery <10 days=less likely to report school problems (P<0.01)

  • Students who reported school problems: (A) longer to become asymptomatic (P<0.005) or pass BCTT (P<0.03), (B) more symptoms (P<0.021), (C) higher severity scores SCAT2 (P<0.023), (D) blurred vision (2.5 times), (E) difficulty remembering (1.8 times), (F) ANAM/ ImPACT BS=49% vs no BS=26% (P<0.03)

  • Missed days of school: higher symptom severity scores (P<0.032)

  • Significant delay for phone follow-up (mean of 14.4±9.6 months): recall bias

  • Two different CNTs used

  • 31/91 (34%) of patients asymptomatic at time of first clinic appointment

174
Brown et al 24 Single-centre, prospective cohort
October 2009 to July 2011
USA
n=335
Mean age 15 (8–23) years
62% male
Patients seen at a sports concussion clinic ≤3 weeks after concussion. PCSS and cognitive activity scaleEffect of cognitive load on duration of PCS
  • Mean PCSS score at initial visit was 30

  • Mean duration of symptoms 43 days; no difference between age groups

  • Total symptom burden at initial visit and cognitive activity independently associated with duration of symptoms

  • Highest cognitive activity associated with prolonged symptom duration; no difference between mild/moderate cognitive activity and cognitive rest

  • Included non-SRC

  • Cognitive activity scale not validated

  • Included adults

  • Specialty sport medicine clinic

  • No direct evaluation of RTS

183
Corwin et al 25 Retrospective EMR review
July 2010 to December 2011
USA
n=247 (Note: same data set as ref 3 0 )
Aged 7–18 years
58% male
Patients with concussion seen at a tertiary paediatric hospital-affiliated Sports Medicine Clinic
Vestibular deficit=abnormalities on VOR and/or tandem gait
Prevalence and recovery of patients with concussion and vestibular deficits, and correlate with ImPACT results81% had vestibular abnormality on initial exam; took significantly longer to RTS (59 vs 6 days, P=0.001) and be fully cleared (106 vs 29 days, P=0.001); scored more poorly on ImPACT and took longer to recover from deficits
  • History ≥3 concussions had 100% prevalence of vestibular deficits and took longer to resolve

  • Relatively small sample size

  • External validity: referral of more severe concussions

  • Delayed presentation

  • Retrospective review

  • 23% non-SRC

184
Darling et al 26 Retrospective chart review and telephone follow-up
2010–2012
USA
Athletes (n=117)
Aged 13–19 years (75% male) for chart review
Patients and parents (n=91) (77.8%) for telephone follow-up
76.9% male
  • Patients with SRC seen at a university sports medicine clinic

  • Once asymptomatic (SCAT2), then CNT (ANAM or ImPACT), then BCTT, then Zurich Guidelines for RTP

Evaluate success of RTP and RTL
  • All athletes RTP without exacerbation of symptoms

  • Telephone follow-up indicated that 42.9% had difficulty readjusting to the classroom following concussion, with 38.5% reporting new or increased issues on RTL

  • Most common complaint was difficulty concentrating

  • 48% had one or more CNTs below average when asymptomatic

  • Performance on CNT was not predictive of RTL issues

  • 2 different CNTs

  • Last CNT prior to RTP used on average 3 weeks after injury—school problems could have already resolved

  • Heterogeneous time of initial evaluation (some day of injury, some not for weeks after injury)

  • University sports medicine clinic

  • 2-month period following RTP for phone follow-up: recall bias

  • Did not describe issues encountered on RTL or length of issues

  • No baseline data

174
Lovell et al 27 Case–control
Baseline data collected prior to 2000 and 2001 seasons
USA
Concussed high school athletes (n=64)
94% male
24 controls (67% male)
ImPACT before and after SRC (36 hours, days 4, 7) compared with non-injured controlsEvaluate memory dysfunction and self-reporting of symptoms in high school athletes with concussion
  • Significant decline in memory in concussed athletes compared with controls

  • Significant differences between preseason and postinjury memory test results at day 4 and day 7 after injury

  • Self-reported symptoms resolved by day 4

  • Duration of on-field mental status changes was related to memory impairment at 36 hours, days 4 and 7 after injury; also related to slower resolution of self-reported symptoms

  • Study and control groups not equivalent in number, gender or sport

  • Excluded those with LOC

  • Small sample size

  • Very short study duration (up to 7 days after injury)

  • Ages of athletes not specified

  • No direct evaluation of RTS or academic needs

154
Makki et al 28 Prospective cohort study
Dates not specified
USA
n=42
Aged 14–19 years
88% male
Student-athletes with SRC attending concussion clinic within 10 days of injury
Daily reports on symptoms (PCSS) and hours of school attendance for 14 days after first clinic attendance
Prospective evaluation of relationship of school attendance and symptoms after SRC
  • Mean age 15.4±1.3 years

  • Mean symptom severity score on first clinic day 32.7±23.3

  • Symptom severity scores decreased with time (β=−2.12, P<0.001)

  • Symptom severity score increased with hours of school (β=0.32, P<0.0137) with no significant effect of days from injury (β=−2.12, P=0.50)

  • Fast, moderate or delayed recovery speed groups did not moderate relationship between school hours and symptoms (P=0.44)

  • Hours of school increased with days from injury

  • Subjects not well described

  • No data regarding numbers of subjects in fast, moderate and delayed recovery speed groups

  • Short period of study (only 14 days)

  • Small sample size

  • No power calculation

  • Subjects seen at concussion clinic

  • Dates of study not identified

123
Purcell et al 29 Retrospective chart review
September 2009 to December 2012
Canada
n=198
Aged 8–17 years, 220 SRC
72.7% male
SRC assessed at a university sports medicine clinicTime to symptom-free RTL and RTP; comparison of children 8–12 years with adolescents aged 13–17 years
  • Symptom-free (days) 8–12 years old=12 vs 13–17 years old=14 (P=0.04)

  • RTL (days) 8–12 years old=4 vs 13–17 years old=2.5 (P=0.86)

  • RTP (days) 8–12 years old=14 vs 13–17 years old=19.5 (P=0.06)

  • 31.3% were aged 8–12 years

  • Initial SCAT2 symptom scores higher in adolescents (12.3 vs 10.6 in 8–12 years old) (P=0.07)

  • Symptom severity score higher in adolescents (19.0 in 8–12 years old vs 27 in 13–17 years old) (P=0.08)

  • 39.5% of patients symptom-free at 10 days

  • 16% still symptomatic 4 weeks after injury

  • 40.3% required academic accommodations

  • 81.8% of 8–12 years old RTP by 4 weeks versus 62.6% of 13–17 years old

  • RTL within 5 days: children 60%, adolescents 71.8%

  • Retrospective chart review

  • SCAT2 not validated in paediatric patients

  • External validity: patients seen at sports medicine clinic (care may not be applicable to community)

204
Ransom et al 30 Case review—structured school questionnaire
Dates not specified
USA
n=349
Aged 5–18 years
67% male
Concussion—symptomatic versus recovered/outpatient concussion clinic large regional medical centre, assessed within 28 days of injury
Parent and child report (69%), parent-only report (31%)
Nature and severity of symptoms; extent of adverse academic effects of concussion
Using PCSI, RBL, CLASS
  • Symptomatic students and parents reported higher levels of concern for impact of concussion on school performance (P<0.05); more school-related problems (P<0.001)

  • Symptomatic high school students reported more adverse academic effects than younger students (P<0.05)

  • Greater severity of PCS associated with more school-related problems and worse academic effects (P<0.001)

  • Higher frequency of impaired neurocognitive scores in symptomatic group (P<0.001)

  • High school students more concerned about academic effects (P<0.01)

  • Math as most problematic class in all grade levels followed by reading/language, arts, science, social studies

  • External validity: patients seen in a concussion clinic may not be representative of community

  • Time period of study not specified

  • Does not discuss how sample size was determined

  • No objective data on academic effects, only parent/child reports, which may be influenced by other variables

174
Ransom et al 31 Cohort study
Dates not specified
USA
n=142
Students aged 11–18 years
59% male
Students with concussion evaluated within 4 weeks of injury at an outpatient concussion clinic
  • PCSI, CLASS, modified BRIEF, ChEERS, EEI, MACS or ImPACT

Demonstrate utility of EBA approach to establish predictive measures to identify students at risk for perceived academic problems during concussion recovery
  • 56% reported low number of academic problems; 44% reported high number of academic problems

  • SRC: 87% in low group; 77% in high group

  • High academic problems associated with greater PCS severity, executive problems and exertional response (<0.001)

  • Cognitive measures did not predict school difficulty

  • Self-reported symptom ratings predicted perceived academic problems more strongly in high school students compared with elementary and middle school students (P<0.001 vs 0.04)

  • Parent reports did not account for additional significant variance

  • Dates not specified

  • Recall bias: students and parent asked to recall perceived academic problems

  • External validity: outpatient concussion clinic

  • No power calculation

  • 2 different CNTs used

163
Taubman et al 32 Prospective observational cohort study
9 December 2011 to 4 June 2013
USA
n=103
Aged 11–19 years; 95 patients completed study
57/95 (60%) male
Patients with concussion who presented within 7 days of injury to a private paediatric practice
  • Cognitive and physical rest

  • PCS scale

Recovery of concussion defined as RTS without symptoms or accommodations
(quick recovery=RTS in <30 days; prolonged recovery=RTS in >30 days)
  • 58/95 (61%) rested immediately after injury

  • Patients who delayed rest were significantly more likely to have prolonged recovery versus those who rested immediately (67% vs 35%, P=0.016)

  • SRC 62/95 (68%)

  • In patients who had a quick recovery, RTS was significantly less in those with immediate rest compared with those with delayed rest (P=0.029) and recovered earlier than those who delayed rest (10 days vs 13 days, P=0.050)

  • PCS score significantly higher in patients with prolonged recovery versus those with quick recovery (40 (3–70) vs 18 (0–100), P=0.002)

  • Only 68% SRC

  • Unclear what PCS score measured (number of symptoms or symptom severity)

  • Not clear if all potential patients were included

  • Determination of sample size not outlined

  • Very strict interpretation of rest recommendation from CISG: did not recommend school absence until all symptoms resolved rather gradual increase in cognitive activities such that symptoms are not significantly exacerbated

203
Zuckerman 
et al 33
Retrospective cohort study
January 2012 to May 2015
USA
n=282
Median age 15.8 years (range 11.6–22.2 years)
61.4% male
Middle school, high school and collegiate student-athletes seen at a regional comprehensive SRC centre, telephone interviewsEffect of SES on outcomes after SRC
  • Previous concussion in 34%

  • Most concussions in football (32.3%), soccer (16.3%), basketball (15.5%)

  • Median symptom duration 21 days (range 1–365 days)

  • Median missed school days, 2 (range 0–90 days)

  • Median missed days of practices, 10 (range 0–150 days)

  • No relationship between SES and symptom duration or days of missed practice

  • Student-athletes with private insurance missed more days of school than those with public insurance (HR 0.46, 95% CI 0.26 to 0.83, P=0.009)

  • Mixed age cohort

  • 58% of cohort unable to be contacted for SES info—external validity concerns

  • Patients seen at a regional SRC centre—external validity concerns

  • Retrospective attainment of SES status by phone interview

  • Very skewed insurance results (90.4% private insurance)

  • 85% white

  • Use of zip code as measure of SES

  • Recall bias—patients contacted a minimum of 3 months after injury, when recovered

173
  • ANAM, Automated Neuropsychological Assessment Metrics; BCTT, Buffalo Concussion Treadmill Test; BRIEF, Behaviour Rating Inventory of Executive Function; BS, borderline score; ChEERS, Children’s Exertional Effects Rating Scale; CISG, Concussion in Sport Group; CLASS, Concussion Learning Assessment and School Survey; CNT, computerised neuropsychological test; DB, Downs and Black checklist; EBA, evidence-based assessment; EEI, Exertional Effects Index; EMR, electronic medical record; ImPACT, Immediate Post-Concussion Assessment and Cognitive Test; LOC, loss of consciousness; MACS, Multimodal Assessment of Cognition and Symptoms for Children; PCS, postconcussion symptoms; PCSI, Post-Concussion Symptom Inventory; PCSS, Post-Concussion Symptom Scale; RBL, retrospective baseline; RTL, return to learn; RTP, return to play; RTS, return to school; SCAT, Sports Concussion Assessment Tool; SES, socioeconomic status; SRC, sport-related concussion; VOR, vestibular ocular reflex.