Measuring disability in patients with chronic low back pain: the usefulness of different instruments

Measuring disability is an important topic in rehabilitation research in patients with chronic low back pain (CLBP). Due to the major impact of CLBP on functioning in both daily living and work, measuring disability in patients with CLBP is best described in terms of limitations in activities and restrictions in participation in daily living and work. A frequently used measurement instrument in rehabilitation medicine to measure self-reported limitations in activities of daily living (ADL) in patients with CLBP, is the Roland Morris Disability Questionnaire (RMDQ). A validated Dutch language version is available (RMDQ-Dv), but test-retest reliability and limits of agreement have not been investigated previously in the RMDQ-Dv. Previous studies found different responsiveness statistics, dependent on the external criterion used to measure change. A frequently used measurement instrument in rehabilitation medicine to measure performance of work-related activities is the Isernhagen Work Systems Functional Capacity Evaluation (IWS FCE). The IWS FCE consists of 28 work-related tasks based on the Dictionary of Occupational Titles, including lifting, carrying, pushing, pulling, forward bending, squatting, crouching, etc. Several subtests of the IWS FCE have proven good reliability in patients with CLBP. In this thesis, the RMDQ-Dv and the IWS FCE are examined on their usefulness to assess, evaluate and predict disability and functioning in patients with CLBP. The thesis mainly focuses on limitations in ADL and limitations in work-related activities. Furthermore, physical and psychosocial factors are studied on their ability to assess and predict restrictions in work participation in patients with CLBP. In chapter 1, the design and outline of the thesis, used measurement instruments and research questions are described. In chapter 2, the reliability and stability of the RMDQ-Dv is presented. Thirty patients with CLBP filled out the RMDQ-Dv with a two-week interval and before starting the rehabilitation program. Intra Class Correlation (ICC) was used as a measure for reliability and the limits of agreement were calculated for quantifying the stability (natural variation over time) of the RMDQ-Dv. The RMDQ-Dv showed good reliability to assess limitations in ADL in patients with CLBP (ICC=0.91). A natural variation of 6 points was found to the total scoring range of 0 to 24. In chapter 3, the consequences of using different external criteria on the responsiveness of the RMDQ-Dv is presented. Eighty-three patients with CLBP filled out the RMDQ-Dv. Four different external criteria were used: 1. Global perceived change in complaints. 2. Global perceived change in ability to take care of oneself. 3. Change in rating of pain intensity. 4. Smallest Real Difference. Standardised response means ranged from 1.33 to 3.45, pooled effect sizes ranged from 1.50 to 2.81 and areas under curves ranged from 0.76 to 1.00. Responsiveness of the RMDQ-Dv ranges from good to very good, dependent on the used external criterion. Chapter 4 describes to what extent self-reporting can replace performance based testing. Seventy-two healthy subjects were tested. Three different self-reports and a performance test were used to measure lifting high. The construct of the self-reports and performance tests covered the same components to enable an adequate comparison. Results showed that all lifting tasks could be predicted, though not solely via self-reports. Performance testing can be predicted with a margin of error of +/- 5-kg for at least 79% of the healthy subjects on the basis gender, self-reports and subject’s participation in fitness. In chapter 5, the performance on the lifting task of the IWS FCE is compared with RWL of the NIOSH lifting guideline for this task. Ninety-two patients with CLBP performed the FCE lifting task. RWL was calculated for this task. Mean difference between performance and RWL was 15.0 kg (SD 14.7; range –8.8 to 59.2). Performance on the FCE lifting task and RWL of the NIOSH for this task produced different safe lifting weights in individual patients with CLBP. This finding may result in contradictory recommendations about need for rehabilitation and return to work. In chapter 6 a systematic review is presented, performed to provide an overview of predictors for sickness absence divided into predictors for the decision to report sick and predictors for the decision to return to work in patients with CLBP. Medical and psychological databases were searched, as well as citations from relevant reviews. In– and exclusion criteria were applied. Two reviewers assessed independently the methodological quality of the papers. The only consistent evidence found was that patients with higher expectations had less sickness absence at the moment of follow-up measurement. Predictors varied with the decision to report sick or to return to work, the used measurement instruments, timing of follow-up measurements and definitions of outcomes. No core set of predictors exists for sickness absence in general in patients with CLBP. Chapter 7 describes results of a cross sectional study about factors related to work status in patients with CLBP, classified according to the international classification of functioning, disability and health (ICF). Ninety-two patients with CLBP filled out questionnaires and performed tests to assess factors related to different ICF domains. Results showed that non-working patients had a lower self-reported physical and mental health, lower physical fitness, more self-reported limitations in ADL, lower education, more depressive symptoms and higher psycho neuroticism than working patients (univariate analyses). Patients with a low educational level, a low self-reported physical or mental health were more likely to be non-working (logistic linear regression analysis). The relation between work status and CLBP is multidimensional, as was illustrated by using the bio-psychosocial model of the ICF. Self-reported limitations and physical and mental health are more strongly related to work status than objective measurements of performance. In chapter 8, an explorative prognostic cohort study with a one-year follow-up (4, 8 and 12 months after baseline) is presented (n=18). The aim of this study was to explore to what extent the standardized IWS FCE could be matched with observed work demands in workers with CLBP and secondly, to what extent this match could predict sick leave one year after rehabilitation treatment. Seven out of eleven analysed FCE activities could be directly matched with work demands (carrying, pushing, pulling, crouching, kneeling, forward bending static, and dynamic bending and rotating). For some workers, difficulties existed in matching of three activities (pushing, pulling and crouching). One activity (lifting) could be indirectly matched with work demands. One activity (walking) could not be matched with WPA data. Two activities (sitting, standing) were excluded from analyses due to practical limitations. No relation was found between performance on FCE activities, work demands, and sick leave during follow-up. It was concluded that seven FCE activities could be directly matched with work demands in the 18 occupations studied. It was however not possible to match all observed work demands with FCE activities. In the general discussion in chapter 9, the main findings about the usefulness of the RMDQ-Dv and IWS FCE to measure disability and functioning in patients with CLBP are discussed. Measuring disability is much more complicated than just measuring limitations in ADL, work-related activities or restrictions in work participation. The difficulty in measuring disability and clinical implications for Occupational and Social Insurance medicine are discussed. Recommendations for further research are presented. The RMDQ-Dv is a useful and valuable instrument when determining need for rehabilitation and evaluation of change in self-reported limitations in ADL. However, the large natural variation of the instrument should be taken into account when using this instrument in individual patients. Prospective research is needed to support of refute the hypothesis that the RMDQ-Dv can be used in predicting disability and functioning in work. The IWS FCE can be used for the assessment of performance of work-related activities and to assess whether the performance of work-related activities is sufficient to perform those activities at work. Because of the large natural variation and the unknown responsiveness, this instrument is less suitable to measure change in performance of work-related activities. In addition, the current standardized IWS FCE protocol is less suitable to predict work ability, and to predict sickness absence by matching work ability and work demands. Instead of using the IWS FCE as outcome measure instrument, the IWS FCE can be used as a tool in treatment programs, to improve limitations reported by the patients, by showing them that they can perform more than thought.