Group comparisons on the measures of family burden and parental distress are summarized in . Although the main effect of group was not significant for the FBII, planned contrasts showed that the mild TBI with LOC group reported higher overall family burden than the OI group (d = .37, p = .04). Neither the overall main effect of group nor the planned contrasts were significant for the BSI.
Analyses of Group Differences in Family Burden and Parental Distress
The analyses involving measures of PCS as predictors of family burden and parental distress are summarized in and . The three-way interaction terms involving group, PCS, and time were not significant for either the FBII or BSI. When the three-way interaction term was trimmed from the models, the PCS by time interaction was significant for the FBII when parent ratings on the PCS Interview and both HBI symptom scales were used to measure PCS. No other interactions between parent-rated PCS and time were significant. The interactions between child ratings of PCS and time were not significant for either the FBII or BSI.
Analyses of the Family Burden of Injury Interview with Group and PCS as Predictors
Analyses of the Brief Symptom Inventory with Group and PCS as Predictors
In all instances, when the PCS by time interaction was significant, higher PCS were associated with greater perceived family burden, but the relationship was stronger at the baseline assessment, for example (r = .50, p = .001) for the parent-rated PCS Interview, than at the 3-month assessment (r = .22, p = .001). Put another way, differences in perceived family burden between children with high versus low levels of baseline PCS tended to decrease over time (e.g., for the parent-rated PCS Interview, F = 42.05, η2 = .13, p = .001 at baseline, and F = 7.40, η2 = .03, p = .007 at 3 months). An example of a significant PCS by time interaction is provided in .
Post-concussive symptoms by time interaction.
The main effect for PCS was significant for both the FBII and BSI when using parent ratings on all three measures of PCS. However, when using child-rated measures of PCS, the main effect for PCS was significant for only the BSI. In general, higher ratings of PCS as reported by both parents and children shortly after injury were associated with greater perceived family burden and parental distress at 3 months.
The group difference between children with mild TBI accompanied by LOC and those with OI on the FBII was no longer significant when any of the parent ratings of PCS or children's ratings of somatic symptoms on the HBI were included in the analyses as independent variables. However, the group difference on the FBII remained significant when the children's ratings on the PCS Interview and their ratings of cognitive symptoms on the HBI were included as independent variables.
The PCS by group interaction was significant for the BSI for parent ratings of somatic symptoms on the HBI, but not for any other measures of PCS (). Contrary to expectations, the parents of children with OI reported greater parental distress than the parents of children in the two mild TBI groups, and the differences were more pronounced when parents reported higher levels of PCS on the HBI Somatic Symptoms scale.