Rheumatology Advance Access originally published online on March 1, 2005
Rheumatology 2005 44(5):633-637; doi:10.1093/rheumatology/keh566
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Quality of care for NSAID users: development of an assessment tool
Department of Rheumatology and 1 Pharmacy, Selly Oak Hospital, University Hospital Birmingham Foundation Trust, 2 Kingsfield Medical Centre, 3 Harborne Medical Practice, Birmingham, 4 Cannock Chase Hospital NHS Trust, Cannock and 5 St Mary's Hospital, Newport, Isle of Wight, UK.
Correspondence to: P. Jobanputra, Department of Rheumatology, University Hospital Birmingham Foundation Trust, Selly Oak Hospital, Raddlebarn Road, Selly Oak, Birmingham B29 6JD, UK. E-mail: P.Jobanputra{at}bham.ac.uk
Objective. Assessments of NSAID use based on authoritative guidelines typically overlook patients' views and nuances of medical history. Our objective was to develop an assessment tool that incorporates these aspects, and technical items, for quality of care assessments in NSAID users.
Methods. Patients newly referred to a university hospital were interviewed by a nurse using an agreed template. A multidisciplinary group of rheumatologists, nurse specialists, primary care physicians and a pharmacist reviewed current guidance and systematic reviews on NSAID use, and a series of interview transcripts. The group agreed, by informal consensus, important determinants of effective and safe NSAID use. Technical aspects of medical care and items that reflected interpersonal care were included in an index for assessing quality of care for individual patients. Interview transcripts of 100 patients were scored by panel members and reliability of scores was tested by calculating weighted percentage agreement and the 
statistic.
Results. Our final index had five domains: medical risk factors; steps taken to reduce risk; knowledge of adverse effects; NSAID dose; and cost efficiency. Each item was scored 0, 1 or 2. Scores were summed, giving a maximum of 10 (low scores indicating low quality). Intra-rater agreement was >90%; was 0.47–0.87 for individual domains and 0.59 for overall score. Inter-rater agreement for overall score was 95%; was 0.25–0.78 for domains and 0.48 for overall score. Patients with especially low scores were identified using the mode of scores for five assessors; obvious clinical concerns were identified, supporting index face validity.
Conclusions. A simple index to evaluate quality of care for NSAID users based on a patient interview is described. This may be used by one or more assessors to examine care standards and highlight deficiencies in relation to NSAID use in practice.
KEY WORDS: Quality of care, NSAIDs, Assessment, Rheumatic diseases