Development of the Knee Quality of Life (KQoL-26) 26-item questionnaire: data quality, reliability, validity and responsiveness

Andrew M Garratt¹^,2 , Stephen Brealey² , Michael Robling³ , Chris Atwell³ , Ian Russell⁴ , William Gillespie⁵ and David King⁶ for the DAMASK Trial Team

¹National Resource Centre for Rehabilitation in Rheumatology, Pb 23, Vinderen, 0319 Oslo, Norway

²Department of Health Sciences, York Trials Unit, Seebohm Rowntree Building, University of York, Heslington, York YO10 5DD, UK

³Department of Primary Care and Public Health, Cardiff University, Neuadd Meirionydd, Heath Park, Cardiff, CF14 4YS, UK

⁴Institute of Medical and Social Care Research, Brigantia Building, Penrallt Road, University of Wales Bangor, Gwynedd, LL57 2AS, UK

⁵Hull York Medical School, University of York, York, YO10 5DD, UK

⁶X-ray Department, York Hospital, Wigginton Road, York, YO31 8HE, UK

author email corresponding author email

Health and Quality of Life Outcomes 2008, 6:48doi:10.1186/1477-7525-6-48


Published:	10 July 2008

Abstract

Background

This article describes the development and validation of a self-reported questionnaire, the KQoL-26, that is based on the views of patients with a suspected ligamentous or meniscal injury of the knee that assesses the impact of their knee problem on the quality of their lives.

Methods

Patient interviews and focus groups were used to derive questionnaire content. The instrument was assessed for data quality, reliability, validity, and responsiveness using data from a randomised trial and patient survey about general practitioners' use of Magnetic Resonance Imaging for patients with a suspected ligamentous or meniscal injury.

Results

Interview and focus group data produced a 40-item questionnaire designed for self-completion. 559 trial patients and 323 survey patients responded to the questionnaire. Following principal components analysis and Rasch analysis, 26 items were found to contribute to three scales of knee-related quality of life: physical functioning, activity limitations, and emotional functioning. Item-total correlations ranged from 0.60–0.82. Cronbach's alpha and test retest reliability estimates were 0.91–0.94 and 0.80–0.93 respectively. Hypothesised correlations with the Lysholm Knee Scale, EQ-5D, SF-36 and knee symptom questions were evidence for construct validity. The instrument produced highly significant change scores for 65 trial patients indicating that their knee was a little or somewhat better at six months. The new instrument had higher effect sizes (range 0.86–1.13) and responsiveness statistics (range 1.50–2.13) than the EQ-5D and SF-36.

Conclusion

The KQoL-26 has good evidence for internal reliability, test-retest reliability, validity and responsiveness, and is recommended for use in randomised trials and other evaluative studies of patients with a suspected ligamentous or meniscal injury.