Acute Concussion Symptom Severity and Delayed Symptom Resolution

Page created by Isaac Barrett
 
CONTINUE READING
Acute Concussion Symptom Severity and Delayed
Symptom Resolution
        WHAT’S KNOWN ON THIS SUBJECT: Children are often evaluated             AUTHORS: Joseph A. Grubenhoff, MD,a,b Sara J. Deakyne,
        in the emergency department after a concussion. Although               MPH,c Lina Brou, MPH,a,b Lalit Bajaj, MD, MPH,a,b R. Dawn
        prolonged symptoms are associated with higher initial symptom          Comstock, PhD,a,d and Michael W. Kirkwood, PhDe
        severity when measured 2 to 3 weeks after injury, a similar            Departments of aPediatrics, and ePhysical Medicine and
        association with acute symptom severity has not been                   Rehabilitation, University of Colorado, Aurora, Colorado;
                                                                               bEmergency Department, and cDepartment of Research
        demonstrated.
                                                                               Informatics, Children’s Hospital Colorado, Aurora, Colorado;
                                                                               and dDepartment of Epidemiology, Colorado School of Public
        WHAT THIS STUDY ADDS: Higher acute symptom severity is not             Health, Aurora, Colorado
        associated with development of persistent post-concussion              KEY WORDS
        symptoms 1 month after injury, but persistent post-concussive          brain concussion, brain injury, acute, brain injury, traumatic,
        symptoms affect a significant number of children after                  post-concussion symptoms, post-concussion syndrome,
        concussion. Outpatient follow-up is essential to identify children     emergency medicine
        who develop persistent symptoms.                                       ABBREVIATIONS
                                                                               AUC—area under the curve
                                                                               CI—confidence interval
                                                                               DSR—delayed symptom resolution
                                                                               ED—emergency department
                                                                               GCS—Glasgow Coma Scale
abstract                                                                       IQR—interquartile range
                                                                               LOC—loss of consciousness
BACKGROUND AND OBJECTIVES: Up to 30% of children who have concus-              ICD-10—International Statistical Classification of Diseases and
sion initially evaluated in the emergency department (ED) display delayed      Related Health Problems, 10th revision
                                                                               OR—odds ratio
symptom resolution (DSR). Greater initial symptom severity may be an
                                                                               PCS—post-concussion syndrome
easily quantifiable predictor of DSR. We hypothesized that greater symp-        RR—relative risk
tom severity immediately after injury increases the risk for DSR.              Dr Grubenhoff conceptualized and designed the study, designed
METHODS: We conducted a prospective longitudinal cohort study of chil-         the database, oversaw data collection and analysis, drafted the
                                                                               initial manuscript, and reviewed and revised the manuscript; Ms
dren 8 to 18 years old presenting to the ED with concussion. Acute symp-
                                                                               Deakyne performed the primary statistical analysis, assisted
tom severity was assessed using a graded symptom inventory. Presence of        with study design and database design, managed study
DSR was assessed 1 month later. Graded symptom inventory scores were           personnel in recruitment and data acquisition, and co-authored,
tested for association with DSR by sensitivity analysis. We conducted a sim-   reviewed, and revised the manuscript; Ms Brou assisted with
                                                                               the cluster analysis and drafting and reviewed and revised the
ilar analysis for post-concussion syndrome (PCS) as defined by the
                                                                               manuscript; Dr Bajaj assisted with study design and study
International Statistical Classification of Diseases and Related Health         personnel management, oversaw data analysis, and critically
Problems, 10th revision. Potential symptoms characteristic of DSR were         reviewed and revised the manuscript; Dr Comstock assisted
explored by using hierarchical cluster analysis.                               with data analysis and interpretation and critically reviewed and
                                                                               revised the manuscript; Dr Kirkwood conceptualized and
RESULTS: We enrolled 234 subjects; 179 (76%) completed follow-up. Thirty-      designed the study, assisted with data instrument design, and
eight subjects (21%) experienced DSR. Initial symptom severity was not         co-authored, reviewed, and revised the manuscript; and all
significantly associated with DSR 1 month after concussion. A total of 22       authors approved the final manuscript as submitted.
subjects (12%) had PCS. Scores .10 (possible range, 0–28) were                 www.pediatrics.org/cgi/doi/10.1542/peds.2013-2988
associated with an increased risk for PCS (RR, 3.1; 95% confidence              doi:10.1542/peds.2013-2988
interval 1.2–8.0). Three of 6 of the most characteristic symptoms of DSR       Accepted for publication Apr 10, 2014
were also most characteristic of early symptom resolution. However,            Address correspondence to Joseph A. Grubenhoff, MD, 13123 East
cognitive symptoms were more characteristic of subjects reporting DSR.         16th Ave, B-251, Aurora, CO 80045. E-mail: joe.grubenhoff@
                                                                               childrenscolorado.org
CONCLUSIONS: Greater symptom severity measured at ED presentation
does not predict DSR but is associated with PCS. Risk stratification there-                                             (Continued on last page)

fore depends on how the persistent symptoms are defined. Cognitive
symptoms may warrant particular attention in future study. Follow-up is
recommended for all patients after ED evaluation of concussion to
monitor for DSR. Pediatrics 2014;134:54–62

54    GRUBENHOFF et al
                                Downloaded from pediatrics.aappublications.org by guest on May 23, 2015
ARTICLE

There are 630 000 emergency depart-             severity of acute concussion symptoms          a convenience sample of children ages
ment (ED) visits annually for mild trau-        may be a useful indicator of overall se-       8 to 18 years who sustained con-
matic brain injury (concussion) among           verity, and therefore may constitute an        cussions no .6 hours before present-
children ages 0 to 19 years.1 The ma-           easily measurable risk factor to predict       ing to Children’s Hospital Colorado’s
jority of those who have concussion             DSR. Recent research among youth               trauma center ED, which has ∼65 000
experience symptom resolution in a few          athletes who have sports-related con-          annual visits. Patients identified on the
weeks.2 However, a notable minority             cussion supports this concept.20               ED electronic track board presenting
experience persistent post-concussive           DSR is a defining feature of post-              with complaints of head injury or symp-
sequelae.                                       concussive syndrome (PCS). However, it         toms associated with concussion were
When examining post-concussive se-              is important to highlight that there is no     screened for enrollment 16 hours per
quelae with standardized performance-           universally accepted definition of PCS.         day, 7 days per week by professional
based cognitive and behavioral tests,           Indeed, whether the nonspecific symp-           research assistants who enrolled sub-
most prospective studies indicate that by       toms typically attributed to this condition    jects and administered all study proce-
2 to 3 months post-injury, deficits are no       constitute a syndrome with a common            dures. Subjects were contacted by
longer apparent.2–6 Fewer studies have          pathophysiological explanation is contro-      telephone 30 days after injury to com-
systematically examined outcomes using          versial.21,22 Nonetheless, DSR affects chil-   plete follow-up procedures. Subjects
                                                dren who have concussion, and evidence         were considered lost to follow-up if they
post-concussive symptom reports from
                                                suggests that the risk for experiencing        failed to respond after 3 attempts. The
children. However, available research
                                                persistent symptoms is modifiable.23–25         study was approved by the Colorado
suggests that some pediatric patients
                                                Identifying children at increased risk for     Multiple Institutional Review Board.
display more persistent symptoms than
might be expected if examining perfor-          DSR at the time of injury would allow          Subjects
mance-based test results alone.7,8 The          selective implementation of interventions
                                                                                               Children were considered to have
acute injury risk factors predictive of         earlier in the recovery phase.
                                                                                               concussion if they had a Glasgow Coma
delayed symptom resolution (DSR) in             The primary objective of this study was        Scale (GCS) score of 13 or 14 or at least
children are poorly understood.                 to determine whether greater symptom           2 of the following symptoms occurr-
Traditionally, determining the severity of      severity measured immediately after            ing after a direct blow to or rapid
concussion was predicated on the pre-           injury is associated with DSR. We hy-          acceleration/deceleration of the head:
sence of certain signs and symptoms at          pothesized that higher scores on               bystander-witnessed LOC; post-traumatic
the time of injury, most notably loss of        a graded symptom inventory immedi-             amnesia; disorientation to person,
consciousness (LOC).9,10 However, LOC           ately after injury would be associated         place, or time; subjective feelings of
occurs relatively infrequently after con-       with the DSR at 1 month in a pediatric ED      slowed thinking; perseveration; vomiting/
cussion and is no longer used to define          cohort presenting for acute evaluation         nausea; headache; diplopia/blurry vision;
injury severity as it is not consistently       of concussion. Given the lack of a uni-        dizziness; or somnolence. This clinical
associated with neuropsychological de-          versally accepted definition of PCS, we         definition of concussion has been used
ficits or DSR.11,12 In contrast, research        also evaluated the performance of              elsewhere.28,29 Children who had open
in adults who have concussion has               a graded symptom inventory for iden-           head injuries, intoxication with alcohol
found that post-traumatic amnesia as            tifying the risk for meeting clinical          or controlled substances, receipt of
                                                criteria for PCS laid out in the In-           narcotics for pain control, injuries re-
well as higher overall symptom levels
                                                ternational Statistical Classification of       sulting from child abuse, multisystem
(ie, both number and severity of symp-
                                                Diseases and Related Health Problems,          injuries, or underlying central nervous
toms) are associated with DSR.13–17 In
                                                10th revision (ICD-10).26 Although ICD-        system abnormalities were excluded.
pediatric patients, greater symptom
                                                10 has not yet been adopted in the
levels present a few weeks after injury
                                                United States for coding purposes, the         Measurements
are associated with a longer duration of
                                                clinical criteria for PCS have been in-
post-concussive symptoms.18 Greater                                                            At the ED enrollment visit, the following
                                                vestigated in concussion research.14,27
symptom levels have also been associ-                                                          demographic and injury character-
ated with objective signs of altered            METHODS                                        istics were obtained: mechanism of
mental status (eg, post-traumatic am-                                                          injury; parental report of previous
nesia) in pediatric ED patients immedi-         Study Design                                   concussion; GCS score as determined
ately after injury.19 Taken together, these     We conducted a prospective cohort study        by the treating provider; and presence
findings suggest that the number and             from October 1, 2010 to March 31, 2013 of      of abnormalities on head CT scan as

PEDIATRICS Volume 134, Number 1, July 2014                                                                                           55
                               Downloaded from pediatrics.aappublications.org by guest on May 23, 2015
reported by a board-certified pediatric         have concussion, we planned to divide        were conducted by using SAS 9.3 (SAS
neuroradiologist if obtained.                  the cohort into low and high acute           Institute, Inc, Cary, NC) and hierarchical
Self-reported concussion symptoms              symptom groups based on a defined             clustering was conducted by using
were quantified by using a graded               cut-off score on a graded symptom in-        SPSS 22.0 (IBM SPSS Statistics, IBM
concussion symptom inventory. The              ventory.32 We hypothesized that DSR          Corporation, Chicago, IL).
symptom inventory included the 12              would be more prevalent in the high
items from the Concussion Symptom              symptom group. Therefore, we esti-           RESULTS
Inventory30 plus 2 additional items re-        mated that a sample size of 202 subjects
                                                                                            Research assistants screened 1253
garding feeling irritable and sad. Sub-        would be necessary to demonstrate
                                                                                            patients for participation; 273 met in-
jects verbally rated to what degree they       a 15% absolute difference in prevalence      clusion criteria and 234 subjects con-
were experiencing 14 symptoms com-             of DSR between the low and high              sented to participate in the study. Of
mon to concussion. We modified the              symptom group, using 90% power and           those enrolled, 179 subjects (76%)
instrument for our pediatric popula-           2-tailed a of 0.05.                          completed the 30-day follow-up call and
tion from a 0 to 6 point scale to a 0 to 2     Pre-injury scores for individual symp-       comprised the study cohort (Fig 1).
point scale to ensure understanding            toms were subtracted from both the           Subjects who did not complete follow-
(range, 0–28). Parents rated their             initial and 30-day follow-up scores to       up were similar to those who did in
child’s symptom severity in the week           account for the presence of these non-       age, gender, initial GCS, mechanism of
before injury using the same in-               specific symptoms before injury. Symp-        injury, and history of previous concus-
strument to provide a pre-injury base-         toms present before injury but absent        sion. Subjects lost to follow-up had
line for these nonspecific symptoms.            post-injury were scored as 0 (ie, a neg-     significantly lower initial graded symp-
The primary outcome, DSR, was de-              ative score was not assigned). De-           tom inventory scores (median score, 7;
fined as the presence of 3 or more              scriptive statistics for demographic and     interquartile range [IQR], 4–12) com-
symptoms 1 month after injury that             acute injury data were calculated as         pared with subjects completing the
were absent or less severe in the week         proportions or medians with inter-           study (median score, 10; IQR, 7–13;
before injury reflective of findings in          quartile ranges and compared by using        P = .01).
similar cohorts.18                             a x 2 test and Wilcoxon rank sum as          Thirty-eight children (21%) from the
ICD-10 criteria for PCS require the            appropriate. Because there is no de-         study cohort met the study definition for
presence of 3 or more of the following 8       fined point separating low from high          DSR. The pre-injury baseline symptom
symptoms 1 month after injury: head-           symptoms, a sensitivity analysis was         score differed significantly in both
ache, dizziness, fatigue, irritability, dif-   performed by using the x2 statistic to       groups, but the scores were low for both
ficulty in concentration or performing          determine the best cut-point for initial     groups (Table 1). Two subjects in the
mental tasks, impairment of memory,            symptom severity scores to divide the        early symptom resolution (ESR) group
insomnia, and reduced tolerance to             low and high groups using DSR as the         and 1 in the DSR group experienced
stress, emotional excitement, or alco-         outcome. After sensitivity analysis, the     a subsequent concussion in the follow-
hol.31 The criteria do not adjust for          best cut-point score was used for mul-       up period. Forty subjects underwent
symptoms present before injury as our          tiple logistic regression, adjusting for     head CT scan with only 5 abnormal
clinical definition did. Patients were          gender and age, as symptom report            findings, all found in the ESR group
considered to have PCS if they reported        may vary by both age and gender.32–34        (Table 2).
3 or more symptoms on the Concussion           The same methods were used                   The results of sensitivity analysis of
Symptom Inventory that aligned with            substituting our clinical definition of       initial symptom inventory scores rang-
PCS diagnostic criteria. There is no           DSR with the PCS criteria. Results of x 2    ing from 8 to 14 are shown in Table 3.
corollary for “reduced tolerance to
                                               analysis were considered significant if P     The receiver-operator characteristic
stress, emotional excitement, or alco-
                                               , .05. Relative risk and odds ratios (OR)    curve is shown in Fig 2 (area under the
hol” on the Concussion Symptom In-
                                               were considered significant if the 95%        curve [AUC], 0.508; 95% CI, 0.475–0.683;
ventory, so that symptom was excluded
                                               confidence interval (CI) did not include      P = .14, rounded). The best cut point
during analysis.
                                               1. Lastly, we conducted hierarchical         was a score of 11 with a sensitivity of
                                               clustering with average linkage analysis     63% and a specificity of 50% for DSR.
Statistical Analysis                           to determine if certain symptoms were        However, a score of 11 was not signif-
Based on a previous study examining            more characteristic of delayed versus        icantly associated with DSR (P = .46). In
symptom severity in ED patients who            early symptom resolution.35 Analyses         multivariate analysis, adjusting for age

56    GRUBENHOFF et al
                                  Downloaded from pediatrics.aappublications.org by guest on May 23, 2015
ARTICLE

                                                                                                                   PCS in the high symptom versus low
                                                                                                                   symptom group was 3.7 (95% CI, 1.3–
                                                                                                                   10.6).
                                                                                                                   Figures 3 and 4 show the results of
                                                                                                                   cluster analysis. Three of the 6 most
                                                                                                                   characteristic initial symptoms in the
                                                                                                                   DSR and ESR groups (those in which
                                                                                                                   the relative linkage distance on the
                                                                                                                   x-axis are shortest) were similar (pho-
                                                                                                                   nophobia, photophobia, blurred, or dou-
                                                                                                                   ble vision). However, cognitive symptoms
                                                                                                                   (difficulty remembering, difficulty con-
                                                                                                                   centrating, or “feeling foggy”) were more
                                                                                                                   characteristic of the DSR group.

                                                                                                                   DISCUSSION
                                                                                                                   Our study of children 8 to 18 years old
                                                                                                                   presenting to an ED ,6 hours after
FIGURE 1                                                                                                           concussion demonstrated that initial
Study Participant Flow Diagram.                                                                                    symptom severity is not associated with
                                                                                                                   DSR. This is an important finding given
and gender, a score of 11 was still not                    in Fig 2 (AUC, 0.629; 95% CI, 0.509–0.748;              evolving knowledge of concussion symp-
associated with DSR (OR, 1.4; 95% CI,                      P = .03). The best cut point was 10 with                tom resolution. In 1988, Lishman pro-
0.7–2.8).                                                  a sensitivity of 77% and specificity of                  posed that symptoms appearing shortly
Twenty-two subjects (12%) met criteria                     51% for PCS (P = .02). The relative risk                after a concussion were primarily the
for PCS. Sensitivity analysis for initial                  for PCS in subjects who had an initial                  result of physiologic derangements di-
graded symptom scores ranging from 8                       symptom score .10 was 3.1 (95% CI,                      rectly related to the injury, whereas
to 14 is shown in Table 4. The receiver-                   1.2–8.0). In multivariate analysis, ad-                 protracted symptoms were more likely
operator characteristic curve is shown                     justing for age and gender, the OR for                  related to latent psychological factors.36
                                                                                                                   Two decades later, accumulated research
                                                                                                                   suggests that “physiogenic” and “psy-
TABLE 1 Demographic and Injury Characteristics for Early and Delayed Symptom Resolution                            chogenic” factors contribute to the con-
           Groups
                                                                                                                   stellation of symptoms present both
                                                             ESR (n = 141)          DSR (n = 38)           Pa      immediately after injury as well as
Demographics                                                                                                       throughout recovery.37
   Mean age, years (SD)                                        12.6 (2.5)            13.4 (2.2)            .79
   Male, %                                                         70                     66               .69     Intuitively, it is reasonable to assume
   History of previous concussion, %                               24                     29               .53     that more severe acute physiologic in-
Injury characteristics
                                                                                                                   jury will manifest as more severe
   Mechanism, %                                                   —                      —                 .79
      Sport                                                        48                     53               —       symptomatology and likely require
      Fall                                                         43                     34               —       a longer recovery period. Recent re-
      Assault                                                       3                      5               —       search supports this assumption. A
      Motor vehicle collision                                       1                      3               —
      Other                                                         5                      5               —       prospective cohort study of youth ath-
   LOC, %                                                          26                     29              0.68     letes evaluated in sports concussion
   Post-traumatic amnesia, %                                       26                     34              0.42     clinics in the first 3 weeks after a con-
   Received head CT scan in ED, %                                  21                     26              0.52
   Abnormal head CT scan results, %                                 4                      0              0.02
                                                                                                                   cussion demonstrated that increasing
   Initial GCS, medianb                                            15                     15              0.99     initial graded symptom inventory
   Preinjury graded symptom score, median (IQR)                   1 (0–2)               2 (1–4)           0.002    scores were associated with increased
   Initial ED graded symptom score, median (IQR)                  9 (6–13)           10.5 (7–15)          0.14
                                                                                                                   odds of symptom resolution occurring
a x 2 was used to compare proportions and Wilcoxon rank sum to compare medians. x2 analysis compared the overall
difference among mechanism for the ESR and DSR groups and only this single p-value is provided.
                                                                                                                   beyond 28 days.20 Similarly, pediatric
b IQR for GCS was 15–15 for both groups.                                                                           ED patients who had high symptom

PEDIATRICS Volume 134, Number 1, July 2014                                                                                                                57
                                   Downloaded from pediatrics.aappublications.org by guest on May 23, 2015
TABLE 2 Characteristics of Subjects Who Had Abnormal Head CT Scans                                          but our results differed from these
Age (y)        Gender         GCS     LOC (+/2)                              Injuries                       previous reports. This suggests that
10.8             F            15         –             Cerebral contusion                                   acute symptom report alone is not an
10.0             M            14         –             Subarachnoid and epidural hematoma, skull fracture   accurate reflection of the physiologic
14.0             M            15         +             Subdural hematoma
 9.4             M            14         –             Subarachnoid hematoma, skull fracture
                                                                                                            and psychological factors that ulti-
10.1             M            15         +             Subarachnoid hematoma                                mately lead to DSR.
                                                                                                            We defined DSR in terms relevant to
TABLE 3 Sensitivity Analysis of Graded Symptom Inventory Scores for Identifying Delayed
                                                                                                            clinical practice. Specifically, we de-
            Symptom Resolution                                                                              veloped a definition that would likely
Cut Point            Sensitivity, %          Specificity, %          NPV, %              PPV, %         P    prompt a primary care provider to refer
 8                           73                   33                  79                 26           .66   a child for specialist evaluation (at least
 9                           71                   33                  81                 22           .45   3 symptoms that are worse 1 month
10                           71                   35                  82                 23           .14   after injury than they were before in-
11                           63                   50                  84                 26           .46
12                           53                   55                  81                 24           .13
                                                                                                            jury). Although the clinical criteria for
13                           47                   66                  82                 27           .19   PCS also require the presence of at least
14                           37                   74                  81                 27           .10   3 symptoms 1 month after concussion,
                                                                                                            the diagnostic accuracy of this defini-
levels measured in the early weeks                       tation of concussion symptoms, there is            tion is a topic of scientific debate as it is
after a concussion had significantly                      some evidence that the acute injury                both subjective and imprecise.39–41
higher odds of symptoms persisting                       factors are stronger determinants of               There is also significant controversy as
for up to 1 year.18 In both studies, initial             symptom reports early in recovery,                 to whether the term “syndrome” is
symptom inventories were obtained an                     whereas non-injury factors contribute              appropriate, given that common con-
average of 11 days after concussion.                     more to persistent symptoms.38 A                   cussion symptoms are also found in
Although both physiologic and psycho-                    strength of our study is that we en-               patients who do not have concussion.42
logical factors contribute to manifes-                   rolled subjects within 6 hours of injury,          Regardless of these shortcomings, PCS

FIGURE 2
ROC curves displaying sensitivity analysis of concussion symptom inventory scores for identifying delayed symptom resolution or post-concussive syndrome.
Optimal scores (closest to upper left of graph) for each outcome are shown along with the AUC and associated P value for each curve.

58        GRUBENHOFF et al
                                       Downloaded from pediatrics.aappublications.org by guest on May 23, 2015
ARTICLE

TABLE 4 Sensitivity Analysis of Graded Symptom Inventory Scores for Identifying ICD-10 Post-               different criteria for PCS showed that
            Concussion Syndrome
                                                                                                           a subset of 6 symptoms common to all 3
Cut Point          Sensitivity, %           Specificity, %         NPV, %         PPV, %            P       criteria was specific to PCS owing to
 8                      77                       33                 91             14             .46      concussion among adults.44 In contrast,
 9                      77                       36                 92             14             .34
10                      77                       51                 94             18             .02
                                                                                                           other work calls into question whether
11                      64                       55                 92             17             .11      the symptoms included in various di-
12                      55                       66                 91             18             .10      agnostic criteria are specific to con-
13                      41                       73                 90             18             .21
                                                                                                           cussion. Our results resemble those of
14                      36                       80                 90             21             .10
                                                                                                           McCauley, Boake, and colleagues, who
                                                                                                           found wide variations in the prevalence
has been studied as an outcome mea-                     .10 was associated with a threefold                of PCS, depending on the criteria
sure in studies of persistent symp-                     increased risk for PCS in our cohort,              employed as well as a lack of specificity,
toms.27,43,44 Therefore, we repeated our                whereas there was no association with              because many adult patients who did
analysis using PCS as the outcome                       our clinical definition.                            not have head trauma also met PCS
rather than our clinical definition of                   One may conclude from these findings                criteria.45,46 Additionally, although the
DSR. We noted 2 important findings.                      that the smaller subset of symptoms                PCS ROC curve showed a statistically
First, we found a 43% relative decrease                 that meet criteria for PCS are more                significant association (P = .034) be-
in prevalence in the outcome (9% ab-                    representative of a specific clinical entity        tween an initial symptom severity score
solute difference) when applying this                   characterized by persistent symptoms               of 10 and PCS, the absolute AUC of 0.629
alternate definition. Second, we showed                  than those found on broader symptom                suggests no more than a modest re-
that a graded symptom inventory score                   inventories. An analysis comparing 3               lationship. Therefore, concluding that

FIGURE 3
Hierarchical cluster analysis for participants who had early symptom resolution. Shorter relative cluster linkage distances on the x-axis indicate symptoms
(y-axis) that are more characteristic of the group, whereas longer distances indicate symptoms that are less characteristic.

PEDIATRICS Volume 134, Number 1, July 2014                                                                                                              59
                                    Downloaded from pediatrics.aappublications.org by guest on May 23, 2015
FIGURE 4
Hierarchical cluster analysis for participants who had delayed symptom resolution. Shorter relative cluster linkage distances on the x-axis indicate symptoms
(y-axis) that are more characteristic of the group, whereas longer distances indicate symptoms that are less characteristic.

PCS criteria are more representative of               specific symptoms are more closely                     fying children who, at the time of their
a unique clinical syndrome is difficult to             associated.14,48 The exploratory na-                  concussion, are at risk for DSR. We did
justify and suggests that accurate risk               ture of cluster analysis prevents                     not perform serial assessments to
stratification is heavily dependent on                 drawing firm conclusions regarding                     determine the precise day of symptom
how the outcome is defined.                            the ability of these symptoms to pre-                 resolution or the range of symptom
Hierarchical cluster analysis is an ex-               dict DSR. However, cognitive symp-                    duration. It is possible that some sub-
ploratory method that aims to dem-                    toms may warrant particular scrutiny                  jects in the DSR group had resolution of
onstrate which features are most                      when present.                                         symptoms shortly after their 30-day
characteristic of a group. Three of the 6             We experienced a lost-to-follow-up rate               follow-up call and were misclassified.
most characteristic symptoms for both                 of 24%. The subjects lost to follow-up                However, the prevalence of DSR in this
the DSR and ESR groups were identical in              had a significantly lower median ini-                  study was 21% compared with studies
our cohort, suggesting that these symp-               tial symptom score than the final study                in other US pediatric ED cohorts 3
toms (phonophobia, photophobia, blur-                 cohort. It is plausible that most of these            months after concussion, which ranged
red or double vision) may not be useful in            subjects would have fallen into the ESR               from 15% to 29%, so this limitation is
identifying those at risk for DSR. However,           group but did not complete the study                  unlikely to have had a significant impact
we found it interesting that cognitive                owing to resolution of symptoms.49 If                 on our results.50,51
symptoms were more characteristic of                  true, the absence of these patients                   Finally, we did not include a control
subjects who had DSR. Although some                   from analysis would tend to bias our                  group who had injuries to body regions
authorshavefoundastrongerassociation                  results toward the null hypothesis.                   other than the head. Although children
between the number of initial symptoms                Our follow-up period was limited to 30                who have concussion tend to report
and DSR,17,47 others have shown that                  days, as we were interested in identi-                more post-concussive symptoms than

60     GRUBENHOFF et al
                                     Downloaded from pediatrics.aappublications.org by guest on May 23, 2015
ARTICLE

children with orthopedic injuries, there          specificity further emphasizes the need                sequelae and that accounts for the
is considerable overlap in symptom re-            for more accurate definitions related to               contribution of both physiologic and
port, highlighting the non-specific na-            sequelae after concussion.                            psychological processes. Given the in-
ture of post-concussive symptoms.29 We                                                                  ability to predict the resolution of post-
are therefore unable to evaluate what             CONCLUSIONS                                           concussive symptoms at the time of
proportion of symptoms is simply at-              Greater symptom severity at the time of               injury, outpatient follow-up and serial
tributable to injury in general and what          injury does not predict DSR among                     symptom assessment should be a cor-
proportion is attributable to concus-             children presenting to the ED for eval-               nerstone of concussion management
sion specifically. Our objective, how-             uation of concussion, but it is a risk                for all children after ED discharge.
ever, was to determine whether acute              factor for meeting criteria for PCS as
symptom severity could be used for                defined by ICD-10. These findings un-                   ACKNOWLEDGMENTS
risk-stratification among head-injured             derscore the need to refine the defini-                 We thank Kendra Kocher, BS, who pro-
children rather than as a diagnostic              tion of post-concussive syndrome to one               vided study coordination and database
tool. Nonetheless, the issue of symptom           that is truly representative of concussion            management.

REFERENCES
 1. Faul MXL, Wald MM, Coronado VG. Trau-          9. Leclerc S, Lassonde M, Delaney JS, Lacroix        18. Yeates KO, Taylor HG, Rusin J, et al. Longi-
    matic Brain Injury in the United States:          VJ, Johnston KM. Recommendations for                  tudinal trajectories of postconcussive
    Emergency Department Visits. Hospital-            grading of concussion in athletes. Sports             symptoms in children with mild traumatic
    izations and Deaths 2002–2006. Atlanta, GA:       Med. 2001;31(8):629–636                               brain injuries and their relationship to
    Centers for Disease Control and Pre-          10. Kelly JP, Rosenberg JH. The development of            acute clinical status. Pediatrics. 2009;123
    vention, National Center for Injury Pre-          guidelines for the management of concus-              (3):735–743
    vention and Control; 2010                         sion in sports. J Head Trauma Rehabil.            19. Grubenhoff JA, Kirkwood MW, Deakyne S,
 2. Carroll LJ, Cassidy JD, Peloso PM, et al;         1998;13(2):53–65                                      Wathen J. Detailed concussion symptom
    WHO Collaborating Centre Task Force on        11. Lovell MR, Iverson GL, Collins MW, McKeag             analysis in a paediatric ED population.
    Mild Traumatic Brain Injury. Prognosis for        D, Maroon JC. Does loss of consciousness              Brain Inj. 2011;25(10):943–949
    mild traumatic brain injury: results of the       predict neuropsychological decrements af-         20. Meehan WP III, Mannix RC, Stracciolini A,
    WHO Collaborating Centre Task Force on            ter concussion? Clin J Sport Med. 1999;9              Elbin RJ, Collins MW. Symptom severity
    Mild Traumatic Brain Injury. J Rehabil Med.       (4):193–198                                           predicts prolonged recovery after sport-
    2004; (43, suppl)84–105                       12. Erlanger D, Kaushik T, Cantu R, et al. Symptom-       related concussion, but age and amnesia
 3. Satz P, Zaucha K, McCleary C, Light R,            based assessment of the severity of a con-            do not. J Pediatr. 2013;163(3):721–725
    Asarnow R, Becker D. Mild head injury in          cussion. J Neurosurg. 2003;98(3):477–484          21. Iverson GL. Misdiagnosis of the persistent
    children and adolescents: a review of         13. King NS, Crawford S, Wenden FJ, Caldwell              postconcussion syndrome in patients with
    studies (1970-1995). Psychol Bull. 1997;122       FE, Wade DT. Early prediction of persisting           depression. Arch Clin Neuropsychol. 2006;
    (2):107–131                                       post-concussion symptoms following mild               21(4):303–310
 4. Satz P. Mild head injury in children and          and moderate head injuries. Br J Clin             22. Garden N, Sullivan KA. An examination of
    adolescents. Curr Dir Psychol Sci. 2001;10:       Psychol. 1999;38(pt 1):15–25                          the base rates of post-concussion symp-
    106–109                                       14. Yang CC, Hua MS, Tu YK, Huang SJ. Early               toms: the influence of demographics and
 5. Babikian T, Asarnow R. Neurocognitive             clinical characteristics of patients with             depression. Appl Neuropsychol. 2010;17(1):
    outcomes and recovery after pediatric TBI:        persistent post-concussion symptoms:                  1–7
    meta-analytic review of the literature.           a prospective study. Brain Inj. 2009;23(4):       23. Ponsford J, Willmott C, Rothwell A, et al.
    Neuropsychology. 2009;23(3):283–296               299–306                                               Impact of early intervention on outcome
 6. Belanger HG, Vanderploeg RD. The neuro-       15. Sheedy J, Geffen G, Donnelly J, Faux S.               following mild head injury in adults.
    psychological impact of sports-related            Emergency department assessment of mild               J Neurol Neurosurg Psychiatry. 2002;73(3):
    concussion: a meta-analysis. J Int Neuro-         traumatic brain injury and prediction of              330–332
    psychol Soc. 2005;11(4):345–357                   post-concussion symptoms at one month             24. Wade DT, King NS, Wenden FJ, Crawford S,
 7. Barlow KM, Crawford S, Stevenson A,               post injury. J Clin Exp Neuropsychol. 2006;           Caldwell FE. Routine follow up after head
    Sandhu SS, Belanger F, Dewey D. Epide-            28(5):755–772                                         injury: a second randomised controlled
    miology of postconcussion syndrome in         16. Drake AI, McDonald EC, Magnus NE, Gray N,             trial. J Neurol Neurosurg Psychiatry. 1998;
    pediatric mild traumatic brain injury. Pe-        Gottshall K. Utility of Glasgow Coma Scale-           65(2):177–183
    diatrics. 2010;126(2). Available at: www.         Extended in symptom prediction following          25. Mittenberg W, Tremont G, Zielinski RE,
    pediatrics.org/cgi/content/full/126/2/e374        mild traumatic brain injury. Brain Inj. 2006;         Fichera S, Rayls KR. Cognitive-behavioral
 8. Yeates KO. Mild traumatic brain injury and        20(5):469–475                                         prevention of postconcussion syndrome.
    postconcussive symptoms in children and       17. Lundin A, de Boussard C, Edman G, Borg J.             Arch Clin Neuropsychol. 1996;11(2):139–145
    adolescents. J Int Neuropsychol Soc. 2010;        Symptoms and disability until 3 months            26. Boake C, McCauley SR, Levin HS, et al.
    16(6):953–960                                     after mild TBI. Brain Inj. 2006;20(8):799–806         Limited agreement between criteria-based

PEDIATRICS Volume 134, Number 1, July 2014                                                                                                           61
                               Downloaded from pediatrics.aappublications.org by guest on May 23, 2015
diagnoses of postconcussional syndrome. J        35. Beckstead JW. Using hierarchical cluster           44. Laborey M, Masson F, Ribéreau-Gayon R,
      Neuropsychiatry Clin Neurosci. 2004;16(4):           analysis in nursing research. West J Nurs              Zongo D, Salmi LR, Lagarde E. Specificity of
      493–499                                              Res. 2002;24(3):307–319                                postconcussion symptoms at 3 months af-
27.   Iverson GL, Lange RT. Examination of “post-      36. Lishman WA. Physiogenesis and psycho-                  ter mild traumatic brain injury: results
      concussion-like” symptoms in a healthy               genesis in the ’post-concussional syn-                 from a comparative cohort study. J Head
      sample. Appl Neuropsychol. 2003;10(3):137–           drome’. Br J Psychol. 1988;153(10):460–469             Trauma Rehabil. 2014;29(1):E28–E36
      144                                              37. Silverberg ND, Iverson GL. Etiology of the         45. McCauley SR, Boake C, Pedroza C, et al.
28.   Taylor HG, Dietrich A, Nuss K, et al. Post-          post-concussion syndrome: physiogenesis                Postconcussional disorder: Are the DSM-IV
      concussive symptoms in children with                 and psychogenesis revisited. Neuro-                    criteria an improvement over the ICD-10? J
      mild traumatic brain injury. Neuropsychol-           Rehabilitation. 2011;29(4):317–329                     Nerv Ment Dis. 2005;193(8):540–550
      ogy. 2010;24(2):148–159                          38. McNally KA, Bangert B, Dietrich A, et al.          46. Boake C, McCauley SR, Levin HS, et al. Di-
29.   Moran LM, Taylor HG, Rusin J, et al. Do              Injury versus noninjury factors as pre-                agnostic criteria for postconcussional
      postconcussive symptoms discriminate in-             dictors of postconcussive symptoms fol-                syndrome after mild to moderate trau-
      jury severity in pediatric mild traumatic            lowing mild traumatic brain injury in                  matic brain injury. J Neuropsychiatry Clin
      brain injury? J Head Trauma Rehabil. 2011;           children. Neuropsychology. 2013;27(1):1–12             Neurosci. 2005;17(3):350–356
      26(5):348–354                                    39. Frances A. The new crisis in confidence in          47. Chrisman SP, Rivara FP, Schiff MA, Zhou C,
30.   Randolph C, Millis S, Barr WB, et al. Concus-        psychiatric diagnosis. Ann Intern Med.                 Comstock RD. Risk factors for concussive
      sion symptom inventory: an empirically de-           2013;159(3):221–222                                    symptoms 1 week or longer in high school
      rived scale for monitoring resolution of         40. Kendell R, Jablensky A. Distinguishing be-             athletes. Brain Inj. 2013;27(1):1–9
      symptoms following sport-related concussion.         tween the validity and utility of psychiatric      48. De Kruijk JR, Leffers P, Menheere PP,
      Arch Clin Neuropsychol. 2009;24(3):219–229           diagnoses. Am J Psychiatry. 2003;160(1):4–             Meerhoff S, Rutten J, Twijnstra A. Prediction
31.   World Health Organization. The ICD-10                12                                                     of post-traumatic complaints after mild
      Classification of Mental and Behavioral           41. Carroll CP, Cochran JA, Guse CE, Wang MC.              traumatic brain injury: early symptoms and
      Disorders: Clinical Description and Di-              Are we underestimating the burden of                   biochemical markers. J Neurol Neurosurg
      agnostic Guidelines. Geneva: World Health            traumatic brain injury? Surveillance of se-            Psychiatry. 2002;73(6):727–732
      Organization; 1992                                   vere traumatic brain injury using Centers          49. Blinman TA, Houseknecht E, Snyder C, Wiebe
32.   Grubenhoff JA, Kirkwood M, Gao D, Deakyne            for Disease Control International Classifi-             DJ, Nance ML. Postconcussive symptoms in
      S, Wathen J. Evaluation of the standardized          cation of Disease, 9th revision, Clinical              hospitalized pediatric patients after mild
      assessment of concussion in a pediatric              Modification, Traumatic Brain Injury Codes.             traumatic brain injury. J Pediatr Surg. 2009;
      emergency department. Pediatrics. 2010;              Neurosurg. 2012;71(6):1064–1070; discus-               44(6):1223–1228
      126(4):688–695                                       sion 1070                                          50. Eisenberg MA, Andrea J, Meehan W, Mannix
33.   Frommer LJ, Gurka KK, Cross KM, Ingersoll        42. Meares S, Shores EA, Taylor AJ, et al. Mild            R. Time interval between concussions and
      CD, Comstock RD, Saliba SA. Sex differences          traumatic brain injury does not predict                symptom duration. Pediatrics. 2013;132(1):
      in concussion symptoms of high school                acute postconcussion syndrome. J Neurol                8–17
      athletes. J Athl Train. 2011;46(1):76–84             Neurosurg Psychiatry. 2008;79(3):300–306           51. Babcock L, Byczkowski T, Wade SL, Ho M,
34.   Berz K, Divine J, Foss KB, Heyl R, Ford KR,      43. Yang CC, Tu YK, Hua MS, Huang SJ. The                  Mookerjee S, Bazarian JJ. Predicting post-
      Myer GD. Sex-specific differences in the se-          association between the postconcussion                 concussion syndrome after mild traumatic
      verity of symptoms and recovery rate fol-            symptoms and clinical outcomes for                     brain injury in children and adolescents
      lowing sports-related concussion in young            patients with mild traumatic brain injury. J           who present to the emergency department.
      athletes. Phys Sportsmed. 2013;41(2):58–63           Trauma. 2007;62(3):657–663                             JAMA Pediatr. 2013;167(2):156–161

(Continued from first page)
PEDIATRICS (ISSN Numbers: Print, 0031-4005; Online, 1098-4275).
Copyright © 2014 by the American Academy of Pediatrics
FINANCIAL DISCLOSURE: The authors have indicated they have no financial relationships relevant to this article to disclose.
FUNDING: Funding for the conduct of this study was provided by a Thrasher Research Fund Early Career Award to Dr Grubenhoff. Use of REDCap Database was
supported by NIH/NCATS Colorado CTSI grant UL1 TR000154. Contents are the authors’ sole responsibility and do not necessarily represent official NIH or Thrasher
Research Fund views.
POTENTIAL CONFLICT OF INTEREST: The authors have indicated they have no potential conflicts of interest to disclose.

62       GRUBENHOFF et al
                                       Downloaded from pediatrics.aappublications.org by guest on May 23, 2015
Acute Concussion Symptom Severity and Delayed Symptom Resolution
Joseph A. Grubenhoff, Sara J. Deakyne, Lina Brou, Lalit Bajaj, R. Dawn Comstock
                           and Michael W. Kirkwood
        Pediatrics 2014;134;54; originally published online June 23, 2014;
                         DOI: 10.1542/peds.2013-2988
Updated Information &               including high resolution figures, can be found at:
Services                            http://pediatrics.aappublications.org/content/134/1/54.full.ht
                                    ml
References                          This article cites 48 articles, 13 of which can be accessed free
                                    at:
                                    http://pediatrics.aappublications.org/content/134/1/54.full.ht
                                    ml#ref-list-1
Citations                           This article has been cited by 2 HighWire-hosted articles:
                                    http://pediatrics.aappublications.org/content/134/1/54.full.ht
                                    ml#related-urls
Permissions & Licensing             Information about reproducing this article in parts (figures,
                                    tables) or in its entirety can be found online at:
                                    http://pediatrics.aappublications.org/site/misc/Permissions.xh
                                    tml
Reprints                            Information about ordering reprints can be found online:
                                    http://pediatrics.aappublications.org/site/misc/reprints.xhtml

PEDIATRICS is the official journal of the American Academy of Pediatrics. A monthly
publication, it has been published continuously since 1948. PEDIATRICS is owned, published,
and trademarked by the American Academy of Pediatrics, 141 Northwest Point Boulevard, Elk
Grove Village, Illinois, 60007. Copyright © 2014 by the American Academy of Pediatrics. All
rights reserved. Print ISSN: 0031-4005. Online ISSN: 1098-4275.

              Downloaded from pediatrics.aappublications.org by guest on May 23, 2015
Acute Concussion Symptom Severity and Delayed Symptom Resolution
Joseph A. Grubenhoff, Sara J. Deakyne, Lina Brou, Lalit Bajaj, R. Dawn Comstock
                           and Michael W. Kirkwood
        Pediatrics 2014;134;54; originally published online June 23, 2014;
                         DOI: 10.1542/peds.2013-2988

 The online version of this article, along with updated information and services, is
                        located on the World Wide Web at:
           http://pediatrics.aappublications.org/content/134/1/54.full.html

  PEDIATRICS is the official journal of the American Academy of Pediatrics. A monthly
  publication, it has been published continuously since 1948. PEDIATRICS is owned,
  published, and trademarked by the American Academy of Pediatrics, 141 Northwest Point
  Boulevard, Elk Grove Village, Illinois, 60007. Copyright © 2014 by the American Academy
  of Pediatrics. All rights reserved. Print ISSN: 0031-4005. Online ISSN: 1098-4275.

              Downloaded from pediatrics.aappublications.org by guest on May 23, 2015
You can also read