Level of Agreement Between Child and Parent Reporting on the Child Sport Concussion Assessment Tool (SCAT5)

METHODS

Participants

A total of 157 youth male football athletes aged 8 to 12 (10.7 ± 1.3) years and their parent or legal guardian (59% male, age = 40.8 ± 8.1 years) volunteered to participate in the study while attending a university-sanctioned youth football camp in June 2022 open to youth football athletes from around the United States. Regarding relation to the child, 82% were the biological parents, with 5.1% being the legal guardian (Table 1). A prior single concussion was reported by 9 children. A total of 15 parents described a history of concussion, with 10 reporting 1, 3 reporting 2, and 2 reporting 3 or more concussions in themselves. Full demographics as supplied by both the child and parent are shown in Table 1. Athletics directors and football coaching staff provided a letter of support for the study to be conducted at the university-sanctioned youth football camp. Institutional review board approval from The University of Alabama was granted, and a parental permission form was given to the parent and child; proceeding with completion of the Child SCAT5 was considered agreement to participate.

Table 1. Child and Parent Demographics as Reported for Themselves (N = 157)

View Fullscreen

RESULTS

The Child SCAT5 had high internal consistency for both the child (Cronbach α = 0.916) and parent (α = 0.923) symptom reports. Agreement on medical and injury background ranged from 67% (6 child/9 parent reports), with 1 additional child indicating being diagnosed with a learning disability, to 85% (18 child/21 parent reports), with 2 additional children describing being diagnosed with ADD/ADHD (Table 2). Agreement on being hospitalized for a head injury was 0%, based on 14 children and 0 parents. The level of agreement for individual symptoms ranged from 70.7% (111/157) for gets distracted easily, difficulty remembering what they’re told, and forgets things to 94.9% (149/157) for going to faint and double vision. The total number of symptoms yielded agreement of 35.0% (55/157), whereas agreement on symptom severity scores was 35.6% (56/157).

Table 2. Parent and Child Reporting and Agreement on Medical and Injury Background

View Fullscreen

Mean scoring and frequency distributions for the child and parent reports are provided in Table 3. Between 75% (118/157 on headaches) and 98% (154/157 on going to faint) of children and 72% (113/157 on gets distracted easily) to 98% (154/157 on double vision) of parents supplied a symptom score of 0 (no symptoms). Interestingly, an individual item symptom score of ≥1 exceeded 95% upper CIs. Therefore, any nonzero symptom reporting was deemed abnormal. Abnormal child report symptom items ranged from 1.9% (3/157 on going to faint) to 25% (39/157 on headaches), whereas abnormal parent report symptom items ranged from 1.9% (3/157 on double vision) to 28% (44/157 on gets distracted easily), with an abnormal percentage of 27% (43/157) on headaches (Table 4). Children indicated an average of 2.80 symptoms; their parents rated 2.60 symptoms. Similar findings existed in symptom severity scores, with an average child severity of 3.73 and parent rating of 3.25. Abnormal scores were nearly identical for the total number of symptoms (29.9% [47/157] for the child and 28.6% [45/157] for the parent) and symptom severity scores (26.7% [42/157] for the child and 25.4% [40/157] for the parent).

Table 3 Symptom Reporting, Frequencies, and Agreement Between Child and Parent Reporting

View Fullscreen

Table 4. Confidence Intervals and Abnormal Scoring Between Child and Parent Reporting^a

View Fullscreen

DISCUSSION

The purpose of our study was to determine the level of agreement between child and parent medical history and symptom reporting and quantify their performance on the Child SCAT5 among male football athletes. As anticipated at baseline, children and their parents agreed fairly well on the diagnosed medical history. We believe we are the first to quantify the agreement between the child and parent on medical history, as only 1 individual must report that information on the Child SCAT5, unlike symptom reporting, for which both the child and parent (if available) do so. Additionally, it does not specify which individual should complete the medical history for the most accuracy. Agreement, defined as the percentage of matched correct diagnoses between child and parent, ranged from 67% (learning disability) to 85% (ADD/ADHD), with 75% agreement for headaches or migraines and 82% for a concussion history. Therefore, most children were able to accurately report their diagnosed modifying factor, with matched parent confirmation, and may be reliable to report this section if comfortable doing so. Interestingly, inconsistencies were noted with 14 additional children reporting a hospitalization for a head injury that was not a response match with the parent as well as 6 additional children providing a diagnosis of headaches or migraines that was not matched. These findings are similar to those of Wojtowicz et al,¹⁷ who observed that 15.9% of boys 13 to 18 years old inaccurately described their concussion history. Although our results suggested 0 additional children who reported a history of concussion that was not confirmed by the parent, they reflected minor, general inaccuracies from younger athletes. It may be that the 14 athletes in our study had a misconception of hospitalization as opposed to seeing a pediatrician or family physician for an orthopaedic evaluation.

Previous researchers demonstrated low baseline correlations between child and parent reports for symptom severity scores.¹⁵ These findings are likely due to the expectation that most athletes without diagnosed modifiers should be asymptomatic at baseline.^21–23 Postconcussion correlations were similar at a 4-week follow-up in athletes aged 7 to 21 years,¹³ which may reflect a return to the earlier or a new baseline level. We attempted to show the level of agreement, rather than correlation, to further characterize the similar scoring across individual symptom Likert-scale ratings, the total number of symptoms, and the severity scores. It is clinically important to recognize the prevalence at which children and parents report identical symptom ratings rather than attempting to understand the linear relationship between them or the validity of the ratings to detect what the symptom evaluation is intended to do. Agreement was roughly 70% to 95% between individual symptom items, mostly reflecting that the child and parent both accurately identified 0 symptoms at baseline, given that 75% to 98% of both children and parents reported an individual symptom item of 0, which can also be seen in the previous literature on pretest symptoms during symptom provocation concussion tasks.²¹ One group examined symptom reporting in children aged 9 to 12 years on the 17-item Post-Concussion Symptom Inventory, which rates symptoms on a 3-point Likert scale: 0 (none), 1 (a little), or 2 (a lot). Asymptomatic prevalences for boys ranged from 85.8% (hard to pay attention) to 97% (blurry vision) for similar symptom items.³⁰ Our rates were comparable with those of Hunt et al³⁰ for headache (75.2% versus 89.9%), nausea (82.8% versus 91.7%), dizziness (92.4% versus 94.1%), difficulty paying attention (83.4% versus 85.8%), difficulty remembering (78.3% versus 94.7%), and blurry vision (92.4% versus 97.0%).

Using the Child SCAT3 from 2015, earlier authors revealed that Canadian youth male ice hockey athletes between 9 and 12 years old described 7.9 total symptoms with a severity score of 11.4, while parent ratings indicated 7.2 symptoms and a 9.8 severity score.⁸ Babl et al²⁹ noted similar child report symptom scores but lower parent reports (4.78 symptoms and 5.93 severity score). Our results suggest that youth males endorsed fewer symptoms and less severity, with approximately 2.5 total symptoms and a 3.5 severity score. Even when compared with male-only values, Kelshaw et al¹⁹ determined that 12-year-old male children reported 7.5 symptoms and a severity score of 10.4 on the Child SCAT5, whereas Brooks et al¹⁸ observed that 5- to 13-year-old boys described 9.9 symptoms and a 15.1 severity score on the Child SCAT3. With the release of the Child SCAT3 in 2012 and the SCAT5 in 2016, the only difference between the symptom inventories is the inclusion of neck pain as a symptom in the latest version.⁶ Despite a difference of 1 additional symptom and 3 severity scores, our male youth football athletes reported fewer symptoms altogether. The difference in symptom reporting between children and parents may be due to geography (children and parents coming from around the country to a youth football camp rather than researchers going to specific youth sports organizations or associations), based on sport type. Additionally, Kelshaw et al¹⁹ investigated only athletes as young as 11 years old, so younger ages, such as those between 8 and 10 years old, may not report as many symptoms due to development and maturation. Further, because the children and parents may not have had a connection to the research staff or university, they may have underreported or inaccurately reported their symptoms.^31–33 Our findings more closely relate to those of Glaviano et al,⁷ who stated that 12-year-old male athletes in middle school indicated 1.4 symptoms with a severity score of 2.2 using the SCAT2 symptom scale despite using the nonchild version, which includes similarly themed individual symptoms but on a 7-point Likert scale, ranging from 0 (none) to 6 (severe). Thus, regardless of the number of Likert ratings, most youth athletes at 12 years of age reported few to no symptoms and minor severity of the items that they did select. We classified values exceeding the upper CIs only, as asymptomatic individuals at baseline would naturally fall below the lower threshold. According to our CIs, any reporting of an individual symptom item was considered abnormal, along with 4+ total symptoms and a 5+ severity score, as children and parents can only rate symptoms on a whole-digit scale. Roughly 3% to 40% of individual symptom items, 45% of the total number of symptoms, and 40% of severity scores were abnormal, warranting further research into modifying factors and comparisons with other baseline measures, such as balance and vestibular-ocular assessment. Youth athletes aged 8 to 14 years have shown an abnormal scoring rate of 9% to 13% on baseline vestibular-ocular motor screening, which may require more investigation for us to better understand why athletes 12 years of age and younger report symptoms during provocation and nonprovocation.³⁴

This study was not without limitations. First, this was a 1-time assessment deployed at a university-sanctioned youth football camp in the southeastern United States. Although the camp was open to registration of youth athletes across the country, most athletes were likely from the region or state. We did not track residing location, which should be explored in future research to ensure geographic representation. Further, only youth males were included in this study, as no females who attended the youth camp participated. Additional research is needed to validate our findings in female athletes, as 12-year-old females have been noted to report more symptoms than their male counterparts at baseline.⁷ We assumed that all participating youth athletes were cleared for full physical activity and sport without restrictions and that results might vary postinjury. Future researchers should examine the level of agreement between child and parent reports postconcussion and during recovery, especially given that parents rate their children’s symptoms differently from baseline. On the day of injury, parents are to rate symptoms according to how their child appears now; on all subsequent days after injury, they must rate symptoms based on how they perceive the child to have felt over the past 24 hours. Participants who were diagnosed with a previous concussion only had 1 such injury. Future authors should address the effects of a dose response³⁵ (eg, 1, 2, 3+ prior concussions). Lastly, symptom factor structures have not been established for the symptom evaluation on the Child SCAT5 in children aged 5 to 12 years considering that affective or emotional symptoms (eg, sadness, nervousness, irritability) are not evaluated. Cross-comparison with the adult SCAT5 can help identify factors, yet investigators should conduct exploratory factor analytic methods to better understand cognitive-sensory, sleep-arousal, vestibular-somatic, and affective factor structures at baseline and postconcussion.

CONCLUSIONS

Our results indicated upward of 85% agreement between child and parent reporting on the medical history and 70% to 90% agreement on individual symptoms in male youth football athletes. Additionally, even though the Child SCAT5 symptom checklist displayed high internal consistency for both child and parent reports, abnormal scores in the range of approximately 3% to 30% existed at baseline. Mean scores were similar between child and parent reports, but agreement between symptoms ranged between 70% and 95%, suggesting that children and parents may report symptoms similarly, with the vast majority describing zero symptoms. Understanding that most child athletes do not endorse many symptoms at baseline is pivotal; thus, any symptom provocation and reporting on symptom assessment may provide valuable clinical information for athletic trainers, pediatricians, and other health care professionals.