51. A reliability study of an instrument for measuring general practitioner consultation skills: the LIV-MAAS scale
- Author
-
Ian Enzer, Maggie Pearson, Jude Robinson, Tom Walley, and Stuart Barton
- Subjects
Adult ,Psychometrics ,Interview ,Attitude of Health Personnel ,Pilot Projects ,Social skills ,Nursing ,Rating scale ,Patient-Centered Care ,Surveys and Questionnaires ,Content validity ,Humans ,Medicine ,Medical History Taking ,Competence (human resources) ,Quality Indicators, Health Care ,Physician-Patient Relations ,Medical education ,Primary Health Care ,business.industry ,Item analysis ,Health Policy ,Public Health, Environmental and Occupational Health ,Physicians, Family ,Videotape Recording ,General Medicine ,Middle Aged ,United Kingdom ,Checklist ,Inter-rater reliability ,Clinical Competence ,business - Abstract
Objective. To evaluate the reliability of a new tool, the LIV-MAAS, in assessing consultation competence in UK generalpractice.Design. These were pilot studies, with small numbers of participants. Videoed general practitioner (GP) consultations wereanalysed by trained lay and professional raters, using the LIV-MAAS. The inter-rater reliabilities were assessed. Four videoswere assessed by Wve raters in a pilot study. After this, 71 consultations from eight doctors were assessed by sets of three raters.Main measures. Inter-rater reliabilities and inter-consultation reliabilities.Results. For the pilot study, the estimated inter-rater reliability ranged from 0.69 (one rater) to 0.91 (Wve raters). For the mainstudy, the estimated inter-rater reliability for the LIV-MAAS checklist using two raters was 0.71, and using three raters it was0.78. Mean differences in reliability within each series of nine consultations were 0.20 (three raters) and 0.42 (two raters).Conclusions. As a measure of ‘consultation competence’, administered by trained raters (medical or lay) to real GP consult-ations, the LIV-MAAS instrument shows adequate reliability and stability but would beneWt from considerable shortening.Further development of the LIV-MAAS and testing with larger samples are required.Keywords: consultation competence, general practitioner assessment, inter-rater reliability, LIV-MAAS, MAAS-GPInterviewing and consultation skills are key in the clinicalpractice of medicine [1]. In recent years, there have beenattempts to make use of psychometric techniques and thedevelopment of speci Wc instruments to measure the attributesof the consultation, both in the UK and internationally [2–7].The expression ‘consultation competence’ is used in the UKliterature to describe the range of skills a medical practitionerneeds in a meeting with a patient [8–11].Medical interviewing involves content and process [1,4].The process requires communication skills to promoteinformation Xow and interpersonal skills to establish rapportwith the patient. A recent ‘functional’ approach has threeaspects: to collect information, to respond to the patient’semotions, and to educate and inXuence behaviour andimplement treatment plans [3,12].As part of a wider study of general practitioner (GP, familypractitioner) consultation skills, we wished to use a reliableinstrument to measure competence in the process of con-sultation. A literature search found one systematic review ofsuch instruments [5], which recommended two ‘as best Wttingthe criteria of reliability, validity and practicality’: the ArizonaClinical Interview Rating Scale [7] (developed in the USA)and the MAAS-GP [4] (developed in The Netherlands, undera medical system much closer to that of the UK). Other toolsalso exist, including the recent UK Leicester AssessmentPackage (LAP) [8], but neither the MAAS-GP nor the LAPhad been adequately tested with samples of real patients; forinstance the MAAS-GP [4] was tested with simulated patientsand medical student raters.The MAAS-GP is based on the presence or absence ofspeciWed behaviours. A technical review identiWed severalproblems: MAAS-GP used an old scaling model with a verysmall group for item analysis, had a mixed item format, hadsome very low subscale reliabilities, and had limited normdata. Despite these shortcomings, the 68-item MAAS-GP hasbeen claimed to best Wt the criteria of reliability, validity, andpracticality for evaluation instruments for medical interviewingskills [4,5]. We therefore chose to develop it with an additional27 items added to include a patient perspective and UK focus.The new 95-item scale, the LIV-MAAS, is divided into sixsubscales and the full instrument can be reviewed on a web-site [13]. The content validity of this new tool has been dem-onstrated for the UK [14]. Like the MAAS-GP scale, itdepends on a rater recording the presence or absence ofAddress reprint requests to Professor T. Walley, Prescribing Research Group, Depart ment of Pharmacology and Therapeutics, University of Liverpool, L69 3GF, UK. E-mail: twalley@liv.ac.uk
- Published
- 2003
- Full Text
- View/download PDF