# Gothenburg direct observation tool for assessing person-centred care (GDOT-PCC): evaluation of inter-rater reliability

## TL;DR

This study evaluates how consistently healthcare professionals can use a tool to assess person-centred care in simulated consultations.

## Contribution

The study provides an evaluation of the inter-rater reliability of the GDOT-PCC tool for assessing person-centred care.

## Key findings

- Cronbach’s α was acceptable for all evaluated domains, indicating good internal consistency.
- ICC values were high for 11 of 13 domains, suggesting good to excellent inter-rater reliability.
- Reliability for the patient perspective domain was uncertain due to wide confidence intervals.

## Abstract

To assess the inter-rater reliability of the Gothenburg direct observation tool-person-centred care in assessing healthcare professionals’ competency in delivering person-centred care (PCC).

Observational, fully-crossed inter-rater reliability study.

The study was conducted between October and December 2022 at the participants’ homes or offices.

Six health professionals individually rated 10 video-recorded, simulated consultations against the 53-item, 15-domain tool covering four major areas: PCC activities, clinician manner, clinician skills and PCC goals. Cronbach’s α was used to assess internal consistency. Intraclass correlations (ICC) and 95% CI were computed for the domains.

Two domains (planning and documentation and documentation) were excluded from analyses due to insufficient evaluable data. Cronbach’s α was acceptable (>0.70) for all evaluated domains. ICC values were high (ICC ≥0.75) for 11 of the 13 domains; however, CIs were generally wide and the lower bounds fell within the good range (ICC=0.60–0.74) for six domains and fair agreement (ICC=0.40–0.59) for the remaining six. The ICC for the domain patient perspective was non-informative due to its wide CIs (ICC=0.74 (0.39–0.92)).

ICC estimates for most domains were comparable to or exceeded those reported for similar direct observation tools for assessing PCC, suggesting that they may reliably be used in, for example, education and quality improvement applications. Reliability for the domains planning and documentation and documentation needs to be assessed in studies sampling more documentation behaviours. Reliability for the patient perspective domain may owe to methodological issues and should be reassessed in larger, better-designed studies.