Development and validation of interpretability cutoffs for the short forms of the Brain Injury Rehabilitation Trust Personality Questionnaires (BIRT-PQs).

BACKGROUND

Previous analyses demonstrated that Italian versions of the five short forms of the Brain Injury Rehabilitation Trust Personality Questionnaires (SF-BIRT-PQs) showed strong psychometric properties. However, evidence on their interpretability lacks.

AIM: To calculate cutoffs for each of the SF-BIRT-PQ to discriminate between the subject's behavior that deviates significantly (D+) or not (D-) from the average of healthy individuals.

DESIGN

Multicenter prospective study.

POPULATION: One hundred and fifty-four subjects with acquired brain injury, their 154 caregivers, and 120 healthy subjects.

METHODS

SF-BIRT-PQs and disability measures were administered to the included subjects. Moreover, subjects scored the Satisfaction Profile (SAT-P) and disability measures, while the Caregiver Burden Inventory (CBI) and Frontal Behavioral Inventory (FBI) were administered to caregivers. Cutoffs between scores indicating D- or D+ patients were calculated with Z-scores, and Mann-Whitney Tests were used to assess the scores of the comparator instruments across the two groups (i.e., D+ and D-) generated by the cutoff for each SF-BIRT-PQs separately for the patient (using SAT-P, DRS, LCF) and caregiver samples (using CBI, FBI, DRS, LCF).

RESULTS

Cutoffs to discriminate between D+ and D- were identified (SF-BMQ ≥25 points, SF-BREQ ≥19 points, SF-BSCQ ≥14 points, SF-BDQ ≥15 points, SF-BIQ ≥18 points indicated D+). Generally, for the patients' sample, between-group comparisons showed significant differences between all SF-BIRT-PQs (except BDQ and BIQ) and the SAT-P subscales (P<0.05) but not with disability measures (P>0.05). In contrast, for the caregivers' sample, between-group comparisons showed significant differences between all SF-BIRT-PQs and the CBI and FBI subscales (P<0.05) but not with disability measures (P>0.05). A digital ruler was devised to convert raw scores into interval measures.

CONCLUSIONS

We provided evidence in support of the construct validity of the proposed cutoffs for each SF-BIRT-PQ.

CLINICAL REHABILITATION IMPACT: These cutoffs make the short forms of the BIRT-PQs interpretable by quantifying the amount of the patient's behavior divergence from that of the average healthy individual. The electronic rulers for each BIRT-PQ provide several functions that greatly facilitate the SF-BIRT-PQ administration and interpretation.