Quantitative interpretation of utterances and movements from videoed interactions between rheumatology nurses and patients commencing methotrexate: a pilot study

Abstract

Background

Educating patients about methotrexate is a core role of rheumatology nurses. We have previously reported the scoring of videoed interviews of rheumatology nurses educating patients prior to commencing methotrexate in comparison with the Calgary-Cambridge consultation model, and the qualitative analysis of the transcripts (Robinson et al. Musculoskeletal Care 2021). We were interested to investigate what could be learned from a more quantitative analysis of utterances and movements in these consultations and how they related to the qualitative interpretations.

Aim

To investigate the frequency of utterances and body movements during interactions between rheumatology nurses and patients commencing methotrexate and to relate these to the qualitative interpretations of the interviews.

Methods

Video-recordings of ten patients receiving methotrexate education from four different rheumatology nurses were available from the previous study. They were analysed using the Medical Interaction Process System (MIPS). This involved coding all utterances and body movements minute-by-minute by multiple inspections of the recordings. The first 10 min of each consultation was coded. The utterances and movements of the nurses and patients were compared. The thematic analysis based on the structure and content of the Calgary-Cambridge (C–C) consultation model was available from the previous study. This enabled the results from the MIPS to be compared between the interviews that scored higher on the C–C model and those scoring lower.

Results

The inter-rater reliability between 2 raters for one video was satisfactory (80–100% agreement). Numerically, giving information dominated the nurse contribution and assent by positive utterances and head nodding dominated for the patients. The results were consistent with the nurse agenda dominating the interaction with little opportunity for patient involvement. Nurses in high-scoring interviews made more illustrative gestures and fewer batonic movements while patients did the opposite. Nurses in high-scoring consultations asked more open questions, with more checking of understanding and summarising but fewer interruptions. Patients in low-scoring consultations were much more animated with head movements and illustrative gestures. Patients also checked and interrupted more.

Conclusions

In this pilot study, the MIPS was usable and demonstrated verbal and non-verbal behaviours consistent with the qualitative assessments. It also showed some behaviours that are not intuitive but may indicate how effectively the interview was progressing. Some nurse behaviours identified that were associated with the higher scoring interviews may be useful indictors for training including making illustrative rather than batonic gestures and checking understanding. Patient behaviours, such as greater animation, were exhibited in low-scoring consultations, and could indicate that the interview was not addressing the patient perspective. Quantification of utterances and movements can be done and may give insights into the consultation process.

Key Points • Interview data can be quantified using the Medical Interaction Process System (MIPS). • Quantification of qualitative data can give insights into the nurse/patient consultation process.