2005 HSR&D National Meeting Abstract
3091 — Measuring Team Functioning to Enhance Patient Outcomes
Strasser DC (Atlanta VA; Emory Univ)
Burridge A (Atlanta VA)
Herrin JS (Biostatistican)
Roth DL (U Alabama - Birmingham)
Falconer JA (Northwestern Univ)
Uomoto J (Seattle Pacific Univ)
Bowen SE (Atlanta VA)
Stevens AB (U Alabama - Birmingham)
To propose a model and method for evaluating the quality of interdisciplinary team functioning in inpatient rehabilitation.
An initial survey consisting of 130 items was administered to 46 VA inpatient rehabilitation teams, inclusive of 530 members (12% MD, 21% RN, 26% PT, 23% OT, 12%ST, 8%SW). Items represented four dimensions of hypothesized team functioning: leadership, managerial skills, social climate and interprofessional relations. Social climate dimension consisted of 45 true/ false items adapted with permission from Moos; interprofessional relations consisted of 9 true/ false items adapted from Ducanis and Golin; Leadership and managerial practices dimensions were adapted with permission from Shortell and supplemented with new items developed and pilot tested by the project team (66 likert scale items). The underlying structure of the data was examined using factor analyses and confirmatory factor analysis (CFA). Factors were selected based upon criteria of Cattell’s scree test and theoretical interpretability. CFA was used to generate measures of fit and to refine the factors. Coefficient alpha was used to estimate internal consistency of the factors. The resultant factor structure was partially tested and supported by CFA on a second data set, obtained 3 years later and consisting of 314 members of 29 teams on 82 of the items retained from the initial survey.
The best fitting model for the managerial practices and leadership domains of the revised 82 item questionnaire contained 4 factors (coefficient alpha): Team Goal Directed Activities (0.95), Utility of Quality of Information (0.36), Coordination/Communication (0.95) and Supervisor Expectations (0.89). The original measures of social climate and interprofessional relations were retained as originally used. Items representing the hypothesized leadership dimension did not factor as a separate construct. The final CFA and the analysis of the resultant factors in identifying differences across professional groups, and differentially predicting change in patient total motor FIM scores and subscales will be presented.
The model and measures offer teams a way of describing and assessing team functioning.
Results from this study may be useful in targeting quality improvement initiatives intended to improve treatment process and patient outcomes.