Exploring the Factor Structure of the Glasgow Children's Benefit Inventory: New Recommendations for Reporting Results.

Objective: Patient-reported outcomes can be useful for reporting benefit from non-life-saving interventions, but often they report a single overall score, which means that much information on the specific areas of benefit is lost. Our aim was to perform a new factor analysis on the Glasgow Children's Benefit Inventory (GCBI) to create subscales reflecting domains of benefit. Further aims were to assess the internal consistency of the GCBI, and to develop guidelines for reporting both a total score and sub-scales in future studies.

Methods: We collected 4 existing datasets of GCBI data from children who have undergone tonsillectomy, ventilation tube insertion, pinnaplasty, and submucous diathermy to the inferior turbinates. We performed exploratory factor analysis with principal axis factoring with varimax rotation, we sought redundancy in question items, and we measured internal consistency.

Results: Using the combined dataset of 772 cases, we found 4 factors which accounted for 64% of the variance and which we have labeled "Psycho-social," "Physical health," "Behavior," and "Vitality." Subscale results varied in predictable ways depending on the nature of the intervention. Cronbach's alpha was 0.928. Item-total correlations were high, and no item could be deleted to improve alpha. Floor effects were apparent for various questions but were not consistent between different interventions.

Conclusions: The GCBI contains a range of questions which each add value in different clinical interventions. We can now make recommendations for reporting the results of the GCBI and its 4 new subscales.