Table 4. .

Summary of Studies Evaluating the Effectiveness of Rest After Concussion Continued on Next Page

Study	Design	Patients	Intervention	Outcome Measures	Main Findings	Conclusions	Level of Evidencea
Brown et al30 (2014)	Prospective cohort	N = 335; 15 ± 2.6 y old (8 to 23 y old); 62% male; 21.8% ice hockey players; 20.6% football playersConcussion specialty clinic	No interventionPatients categorized by CA daysCA level assessed via a 5-point scale developed by authors (0 = complete cognitive rest, 1 = minimal cognitive activity, 2 = moderate cognitive activity, 3 = significant cognitive activity, 4 = full cognitive activity)	Duration of concussion symptoms	Overall mean duration of symptoms = 43 ± 53 dPatients in fourth quartile of CA days had delayed recovery compared with those in quartiles 1–3CA days associated with symptom duration (hazard ratio = 0.9942; 95% confidence interval = 0.9924, 0.9960)	Patients engaged in highest levels of CA took longest times for symptoms to resolveSimilar recovery trajectory noted for lower 3 quartiles, which suggests that complete cognitive rest may not be needed	3
Buckley et al33 (2016)	Prospective cohort	N = 50; rest group = 19.8 ± 1.2 y old and no-rest group = 19.4 ± 1.3 y oldCollege	No-rest group not prescribed any rest but was withheld from sport participationRest group prescribed cognitive and physical rest on day of concussion and next day was instructed to not attend class, team meetings, or study hall; to not use electronics; and to refrain from team and personal exercise	Concussion clinical battery, including graded symptom checklist, ImPACT computerized neurocognitive test, BESS, and SAC	Rest group symptomatic longer than no-rest group (5.2 ± 2.9 d versus 3.9 ± 1.9 d; P = .047)No differences found on SAC, BESS, or ImPACTNo difference in time until clinical recovery	2 d Cognitive and physical rest not effective in altering recovery time as measured by clinical assessment battery	3
de Kruijk et al35 (2002)	RCT	N = 103; 51 in no–bed-rest group (39.9 ± 14.5 y old; 52% male; 13% sport-related mechanism of injury) and 52 in full– bed-rest group (34.1 ± 6.5 y old, 60% male, 6% sport-related mechanism of injury)Emergency department and outpatient clinic	No–bed-rest group instructed to be mobile and take ≤4 h of bed rest/d, which decreased until day 5, when they were to resume normal daily activitiesFull–bed-rest group instructed to take full bed rest during first 6 d after injury, then by progressive decrease in bed rest, as followed by no–bed-rest group initially	Severity of 16 post-traumatic complaints at 2 wk, 3 mo, and 6 mo after injuryPosttraumatic complaints categorized into cognitive, dysthymic, vegetative, and physical subgroupsMeasured with visual analog scaleMedical Outcomes Short Form-36	No–bed-rest group had 17 h rest, whereas full–bed-rest group reported 57 h of bed restDuring first 4 d, patients in full–bed-rest group reported less dizziness but no differences seen for headache or nauseaAt 2-wk follow-up, full–bed-rest group had lower visual analog scale scores for feeling faint, but no other symptom complaints were different than in no-rest groupNo differences on any SF-36 subscale score	6 d of Complete bed rest resulted in no differences between no–bed-rest and full–bed-rest groups in posttraumatic symptoms or general health status	2
Gibson et al32 (2013)	Retrospective cohort	N = 184; 15 ± 3 y old (8 to 26 y old); 72% maleConcussion specialty clinic	No interventionTreatment plans reviewed to determine whether cognitive rest was recommendedPatients grouped by symptom duration (≤30 d versus >30 d)	Recovery, defined as being symptom free at rest and with exertion, return to or above baseline on computerized neurocognitive testing and balance assessment	Cognitive rest recommended to 85 patients; a higher percentage of those were younger than age 15 (58% versus 37%; P < .01)Mean symptom duration longer in patients to whom cognitive rest was recommended (57 d versus 29 d; P < .001)After adjusting for covariates, no association found between rest recommendation and symptom duration (odds ratio = 0.50; 95% confidence interval = 0.18, 1.37)	No relationship between cognitive rest and duration of symptoms	4
Majerske et al29 (2008)	Retrospective cohort	N = 95; 15.9 ± 1.4 y old; 84% male; 59% football players; 41% had concussion history; no learning disabilities, attention-deficit/ hyperactivity disorder, or seizure disordersConcussion specialty clinic	No interventionPatients grouped into 1 of 5 activity levels by level of cognitive and physical activity on activity-intensity scale from information within medical chart	Neurocognitive function (ImPACT)PCSS total symptom score	Patients in 2 highest activity levels (school and sports game and school and sports–practice) performed worse on visual memory (P = .003) compared with patients who participated in school activity and light activity at homePatients in highest activity-level performed worse on reaction time (P < .001) compared with patients who participated in school activity and light activity at homeTrend between activity intensity level and total symptom score (P < .08)No differences between 2 lowest activity levels and moderate activity level	Patients engaged in highest levels of activity, including full school and game participation in sport, demonstrated impairments in 2 neurocognitive domainsModerate levels of activity (school activity and light activity at home) might be beneficial to patients	4
Maerlender et al31 (2015)	RCT	N = 28; standard group (12 female, 3 male) and exertion group (8 female, 5 male)College	Participants randomized to standard care or exertionStandard-care group instructed to engage in no systematic exertion beyond normal activities for schoolExertion group rode stationary bike at mild to moderate level for 20 min daily and wore activity monitor	Recovery, defined as return to baseline on neurocognitive, balance, and symptom scores	Median days to recovery did not differ by groups (P = .705)No group difference found in number of patients who took longer than 2 wk to recover (P = .464)Symptom increase after bike protocol not related to length of recovery (P = .252)Average amount of daily vigorous exertion related to recovery; more activity resulted in increased recovery time (P = .039)	Early mild to moderate exercise did not delay recoveryVigorous exercise associated with longer recovery	2
Moor et al34 (2015)	Prospective observational	N = 56; 15.2 ± 1.7 y old; 53.6% male; 32.7% football playersHospital-based sports medicine center	No interventionParticipants surveyed at discharge regarding recommendations for physical and cognitive rest, how receptive they were to recommendations, and how frequently they followed each recommendation	Recovery, defined as date of physician clearance	All patients received recommendations for physical rest; 71.3% were receptive to this recommendation and 87.5% reported adherence to recommendation92.9% of Patients received recommendations for mental rest with restrictions from electronics; 67.3% were receptive to and 76.9% adhered to recommendation92.8% of Patients received recommendations for mental rest with school restriction; 82.9% were receptive and 90.2% adhered to recommendation	None of variables for assessing adherence to rest were significant predictors of recovery	3
Moser et al27 (2012)	Retrospective cohort	N = 49; 15 ± 2.6 y old (14 to 23 y); 67% male; 27% ice hockey playersConcussion specialty clinic	1 wk of Full rest, defined as no school, homework or tests, travel or shopping, driving, social activities, watching visually intense television, computer use, texting, reading, or physical exercise as well as increased sleepPatients grouped by length of time between sustaining concussion and first postconcussion assessment (1 to 7 d, 8 to 30 d, >31 d)	Neurocognitive function (ImPACT)PCSS	Improvement in cognitive function in all composite scores (P < .001 to P < .008)Total symptom score decreased after rest (P < .001)No main effect for time since concussion (P = .44)	1 wk of Cognitive and physical rest decreased symptoms and increased cognitive scores, regardless of time between concussion and onset of rest	4
Moser et al28 (2015)	Retrospective chart review	N = 13; 15.1 ± 1.5 y old with persistent symptoms after concussion; 57% male; 77% self-reported attention-deficit/hyperactivity disorder, learning disability, or history of 2+ concussions	Patients instructed to complete 1 wk of full rest, defined as no school, homework or tests, travel or shopping, driving, social activities, watching visually intense television, computer usage, texting, reading, or physical exercise as well as increased sleep	Neurocognitive function (ImPACT)PCSS	Cognitive function for all 4 composite scores improved after rest (P = .002 to P = .017)Total symptom score decreased after rest (P = .02)	1 wk of Prescribed rest decreased symptoms and improved cognition in 61.5% of patients	4
Thomas et al36 (2015)	RCT	N = 88; strict-rest group = 14.7 y old (13 to 15.5 y old); usual-care group = 13.1 y old (12.1 to 14.5 y old); 66% male; 71% sport mechanism; 27% football playersPediatric emergency department	Strict-rest (intervention) group was to maintain 5 d of strict rest at home, refraining from school, work, and physical activity followed by stepwise return to activityUsual-care (control) group received verbal recommendations as seen fit by treating physician	Activity diaryNeurocognitive function: computerized (ImPACT) and pencil-paper batteryBESSPCSS	Usual-care group reported more hours of moderate and high mental activity (8.33 versus 4.86 h; P = .03) from days 2–5 after injuryStrict-rest group reported higher PCSS total symptom score during follow-up period (187.9 versus 131.9; P < .03) and higher number of endorsed symptoms (70.4 versus 50.2; P < .03)No differences noted for computerized neurocognitive test scores or balance scores on days 3 or 10 postinjuryStrict-rest group performed better on Symbol Digit Modalities Test on day 3 (67.6 versus 59.9; P < .01) and worse on day 10 (67.6 versus 71.5; P = .04) but differences not found with other pencil-paper tests	Strict-rest group reported more symptoms and had slower symptom resolution than usual-care groupStrict rest may not be more beneficial than usual care	2

Abbreviations: BESS, Balance Error Scoring System; CA, cognitive activity; ImPACT, Immediate Post-Concussion Assessment and Cognitive Test; PCSS, PostConcussion Symptom Scale; RCT, randomized controlled trial; SAC, Standardized Assessment of Concussion.

^aLevels of evidence are based on the Oxford Centre for Evidence-Based Medicine taxonomy.¹⁷