Inter-Rater and Intra-Rater Reliability of the INSPECT (Interactive Nutrition Specific Physical Exam Competency Tool) Measured in Multi-Site Acute Care Settings

Development and initial evaluation of a treatment integrity measure for low-intensity group psychoeducational interventions

BACKGROUND

Despite the importance of assessing the quality with which low-intensity (LI) group psychoeducational interventions are delivered, no measure of treatment integrity (TI) has been developed.

AIMS

To develop a psychometrically robust TI measure for LI psychoeducational group interventions.

METHOD

This study had two phases. Firstly, the group psychoeducation treatment integrity measure-expert rater (GPTIM-ER) and a detailed scoring manual were developed. This was piloted by n=5 expert raters rating the same LI group session; n=6 expert raters then assessed content validity. Secondly, 10 group psychoeducational sessions drawn from routine practice were then rated by n=8 expert raters using the GPTIM-ER; n=9 patients also rated the quality of the group sessions using a sister version (i.e. GPTIM-P) and clinical and service outcome data were drawn from the LI groups assessed.

RESULTS

The GPTIM-ER had excellent internal reliability, good test-retest reliability, but poor inter-rater reliability. The GPTIM-ER had excellent content validity, construct validity, formed a single factor scale and had reasonable predictive validity.

CONCLUSIONS

The GPTIM-ER has promising, but not complete, psychometric properties. The low inter-rater reliability scores between expert raters are the main ongoing concern and so further development and testing is required in future well-constructed studies.

Item-Level Analysis of a Newly Developed Interactive Nutrition Specific Physical Exam Competency Tool (INSPECT) Using the Rasch Measurement Model

The Interactive Nutrition Specific Physical Exam Competency Tool (INSPECT) is a tool designed specifically to observe and measure registered dietitian nutritionists’ (RDNs) nutrition-focused physical exam (NFPE) competence in authentic acute care settings. The initial INSPECT items were generated and tested for content and face validity using expert RDNs’ input. The INSPECT was further examined for inter-rater, intra-rater, and internal consistency using clinical supervisor observations of RDNs performing NFPE on patients in real-life acute care settings. These previous studies showed the INSPECT to have excellent content validity, acceptable face validity, good inter-rater reliability, moderate to strong intra-rater reliability, and excellent internal consistency. In the current study, the Rasch measurement model was applied to examine the item-level properties of the INSPECT. Results confirm that the INSPECT measured a single construct. All items fit the established criteria for clinical observations of >0.5 and <1.7, had positive point measure correlations, met the Wright Unidimensionality Index criteria of ≥0.9, exhibited one latent construct with >40% variance explained by the Rasch dimension as well as a sub-dimension based on item difficulty from the principal component analysis of the first contrast Rasch residuals. Rasch rating scale analysis revealed that the rating scale and majority of the items (39/41) fit the Rasch model. Rasch item hierarchy analysis matched the a priori hypothesized hierarchy for the top-most and bottom-most items. Ceiling effects were seen for three items (hand hygiene, personal protective equipment, and patient position) and one item (handgrip using hand dynamometer) reached the floor effect. Rasch reliability assessment demonstrated high person reliability (0.86), high item reliability (0.96), and person separation of 3.56 ability levels. The principal component analysis of residuals revealed two factors based on item difficulty, one for micronutrient exam and another for macronutrient exam, initial steps, and bedside manner. The resulting two factors may likely be due to a sub-dimension of the latent NFPE trait. Overall, the INSPECT items were found to have good item-level psychometrics. Continued testing of the INSPECT with RDNs at different ability levels will help to determine cut-off scores ranging from novice to expert. Establishing cut-off scores for the INSPECT will further enhance the utility of the tool.