Original Article
Changes to the subscales of two vision-related quality of life questionnaires are proposed

https://doi.org/10.1016/j.jclinepi.2005.04.007Get rights and content

Abstract

Background and Objective

Psychometrically sound questionnaires for the assessment of vision-related quality of life (QOL) are scarce. Therefore, the objective was to further validate two vision-related QOL questionnaires in a Dutch population of visually impaired elderly.

Methods

A total of 329 visually impaired older persons referred to low vision services completed the low vision QOL (LVQOL) and Vision-Related Quality of Life Core Measure (VCM1) questionnaires at baseline, after 1–4 weeks (retest), and after 5 months. Confirmatory factor analyses were performed on baseline data. The smallest detectable change (SDC) was assessed, based on the standard error of measurement (SEM). Change scores between the baseline and 5 months follow-up data were related to a general transition question to assess the minimal important change (MIC). Furthermore, the MIC was related to the SDC, to examine whether the MICs were detectable beyond measurement error.

Results

The original factor structures could not be confirmed. After omitting items and remodeling, adequate fits were obtained. SDCs comprised at least one quarter of the scale for all scales and subscales on the individual level and exceeded the MICs on every occasion.

Conclusion

We propose MICs of 5–10 points for the scales and subscales of the LVQOL and VCM1. The questionnaires are not useful in the follow-up of individual patients.

Introduction

During the last decades there has been increased attention for health-related quality of life (HRQOL) as an outcome measure in health-care [1], [2], [3]. The advantage over outcome measures at the impairment level is that HRQOL includes aspects of perceived health from a patient's perspective and the impact of health on a patient's life. The use of HRQOL as an outcome measure is especially appealing for use in populations with chronic conditions for which no cure exists. In these conditions, outcome measurements at the impairment level will show no or little improvement; however, improvement in HRQOL may occur through psychological adjustment, rehabilitation, or use of assistive devices.

The leading cause of severe visual impairment in the Western world is age-related macular degeneration (AMD). Other important causes are diabetic retinopathy (DRP) and glaucoma [4], [5], [6]. These are all chronic conditions for which no cure currently exists. Treatment is aimed at slowing down progression of the disease. For persons with severe visual impairment or blindness, rehabilitation programs exist that are aimed at helping them cope with their disability in daily life.

Not surprisingly, many quality of life (QOL) questionnaires have been developed specifically aimed at visually impaired people. In a recent review, as many as 31 such questionnaires have been identified [7]. Large differences in psychometric quality were found, and several areas of psychometric testing have been largely or totally neglected. One of these areas is the assessment of dimensions within a scale. Whereas a number of questionnaires were examined for factorial validity, none of these factor solutions have been confirmed in later studies, even when questionnaires were evaluated for different populations or in different languages [7]. A factor structure of a questionnaire that has been evaluated only once in one population cannot be assumed to be stable [8].

Other areas that need attention are the agreement between repeated measurements in reproducibility studies and the ascertainment of the minimal important change (MIC), which is also referred to as the minimal important difference [7]. Agreement measures refer to the absolute measurement error of an instrument, expressed in the unit of measurement of the instrument. Two well-known agreement measures are the 95% limits of agreement according to Bland and Altman [9] and the smallest detectable change (SDC), based on the standard error of measurement (SEM) [10].

The MIC can be defined as “the smallest difference in score in the domain of interest which patients perceive as beneficial and which would mandate, in the absence of troublesome side effects, a change in the patient's management” [2]. Assessment of a MIC is therefore very important for the interpretation of change scores in the target population. A logical next step is to link the MIC with the SDC, to assess whether the smallest difference perceived as beneficial from a patients' perspective can actually be detected beyond measurement error by the measurement instrument. Surprisingly, only a few authors have previously taken this final step in the evaluation of measurement instruments [11], [12].

Our main objective was to validate two vision-related QOL questionnaires in a Dutch population of persons with severe visual impairment (or low vision). Specific aims were (1) to confirm the factor structures of these questionnaires, (2) to assess the level of agreement between repeated measures by calculating the SDC, and (3) to assess the MIC and to examine if these questionnaires can be used to detect the MIC on individual and on group level.

Section snippets

Design

The study population consisted of participants of a 1-year follow-up study on the outcome of low vision services on the quality of life of visually impaired older men and women. Measurements were taken at baseline, 5 months later, and 1 year later. Confirmatory factor analyses were performed on the baseline data. The baseline and 5-month follow-up data were used to assess the MIC.

Participants who returned the baseline questionnaire and who had not yet had their appointment with low vision

Factor analyses

The questionnaires were completed by 329 participants at baseline. Mean age was 78.2 years (SD = 9.0); 62% of participants were female. The median Snellen visual acuity was 0.28 (interquartile range = 0.16–0.38). The main causes of visual impairment as recorded from the best eye were AMD (52%), DRP (13%), glaucoma (6%), corneal disease (6%), cataract (6%), and occluded vein (5%). Total percentage of missing values was 5.2% for the LVQOL and 0.2% for the VCM1. Of the 329 participants, 18 had

Factor analysis

The results show that the original factor structures of the VCM1 and the LVQOL could not be confirmed in our data set. The one factor structure of the original VCM1 is conceptually clear, except for the item on inability to do preferred activities. This item is clearly different from the other items, which are aimed at assessing feelings and perceptions associated with people's visual impairment. When the item was removed, the fit of the model increased considerably to a value > 0.9. The factor

Conclusion

In conclusion, we have revised the factor structure of two vision-related quality of life questionnaires using confirmatory factor analyses. The revised scales and subscales were then used to examine the smallest detectable change beyond measurement error for individuals and for groups and the minimal important change. The MIC was larger than the SDCind for all scales, meaning both questionnaires are not very useful in the follow-up of individual older visually impaired persons. There were

Acknowledgments

We would like to thank Herman Ader for his statistical advice concerning confirmatory factor analysis. Furthermore, we would like to thank the following Dutch foundations for financial support of this study: the ‘Landelijke Stichting voor Blinden en Slechtzienden’ (Michiel de Boer is LSBS-fellow), the ‘Stichting Blindenhulp’, the ‘Haak-Bastiaanse Kuneman Stichting’, and the ‘Stichting voor Ooglijders’.

References (32)

  • M.R. de Boer et al.

    Psychometric properties of vision-related quality of life questionnaires: a systematic review

    Ophthalmic Physiol Opt

    (2004)
  • F.J. Floyd et al.

    Factor analysis in the development and refinement of clinical assessment instruments

    Psychol Assess

    (1995)
  • J.M. Bland et al.

    Statistical methods for assessing agreement between two methods of clinical measurement

    Lancet

    (1986)
  • H. Beckerman et al.

    Smallest real difference, a link between reproducibility and responsiveness

    Qual Life Res

    (2001)
  • N.S. Jacobson et al.

    Clinical significance: a statistical approach to defining meaningful change in psychotherapy research

    J Consult Clin Psychol

    (1991)
  • K. Bruynesteyn et al.

    Determination of the minimal clinically important difference in rheumatoid arthritis joint damage of the Sharp/van der Heijde and Larsen/Scott scoring methods by clinical experts and comparison with the smallest detectable difference

    Arthritis Rheum

    (2002)
  • Cited by (65)

    • Reproducibility: Reliability and agreement of short version of Western Ontario Rotator Cuff Index (Short-WORC) in patients with rotator cuff disorders

      2016, Journal of Hand Therapy
      Citation Excerpt :

      High within group variability is known to result in lower ICC and consequently higher MDC.77,78 We noticed that the smaller value of MDC90group than MDC90individual aligns with the agreement parameters reported in the literature on psychometric characteristics of outcome measures.63,69,79 MDC90group has never been reported in the previous studies done on measurement properties of Short-WORC31 or WORC.6,21–30

    View all citing articles on Scopus
    View full text