Article type
Year
Abstract
Introduction/Objective: Germany has a unique system of hospital based medical rehabilitation. Measuring and assuring outcome quality however is a worldwide problem with two principal solutions. The alternative is either to use a costly pre-post-design to observe change or a single effect measurement after rehab with direct questions for perceived change. We compare the two methods in a large sample of rehab patients.
Methods: From May to October 1996 a consecutive sample of 1070 rehab patients of all indications from three pension funds for blue collar workers completed a 4-week inpatient rehabilitation. 2 weeks before and about 6 weeks after rehab all received a postal questionnaire with some 80 items enquiring about various aspects of their subjective health and health related quality of life (response rate 72%) within the previous four weeks. The core of the pre- and post-questionnaire is formed by the German version of the SF 36 for chronic conditions. The post-instrument included also 15 direct questions for perceived change ("when you compare your situation before rehabilitation with today, has rehabilitation resulted in changing the following Problems ..?" with response categories: improvement (+1), no change (+-0), deterioration (-1)). In this report we concentrate on items for general health, pain (SF36) and functional capacity in vocational activities (NRS).
Results: 67% of the 1070 responders were males, the average age was 48 years. In the pre-post-design, all comparisons indicated an improvement with effect sizes of less than 0.8 standard deviations. General health e.g. increased from on average 2.0 to 2.8 (sd=1.0). Similar results were to be observed using direct (perceived) change measurement, with effect sizes between +0.31 to +0.66 sd-units. However, spearman correlations between the two change measures only ranged between 0.30 and 0.42. Kappas were always below 0.30, indicating low if any agreement between the two methods.
Discussion: a) In agreement with results from other pre-post-studies the German system of inpatient rehabilitation seems only slightly effective using subjective health indicators as outcomes. b) The two methods of change measurement differ markedly in terms of costs with clear advantages for the direct method. Its usefulness however seems unpaired by a low concurrent and criterion based validity. The question is should quality management in rehab medicine be driven by observation of patients or by their perceptions?
Methods: From May to October 1996 a consecutive sample of 1070 rehab patients of all indications from three pension funds for blue collar workers completed a 4-week inpatient rehabilitation. 2 weeks before and about 6 weeks after rehab all received a postal questionnaire with some 80 items enquiring about various aspects of their subjective health and health related quality of life (response rate 72%) within the previous four weeks. The core of the pre- and post-questionnaire is formed by the German version of the SF 36 for chronic conditions. The post-instrument included also 15 direct questions for perceived change ("when you compare your situation before rehabilitation with today, has rehabilitation resulted in changing the following Problems ..?" with response categories: improvement (+1), no change (+-0), deterioration (-1)). In this report we concentrate on items for general health, pain (SF36) and functional capacity in vocational activities (NRS).
Results: 67% of the 1070 responders were males, the average age was 48 years. In the pre-post-design, all comparisons indicated an improvement with effect sizes of less than 0.8 standard deviations. General health e.g. increased from on average 2.0 to 2.8 (sd=1.0). Similar results were to be observed using direct (perceived) change measurement, with effect sizes between +0.31 to +0.66 sd-units. However, spearman correlations between the two change measures only ranged between 0.30 and 0.42. Kappas were always below 0.30, indicating low if any agreement between the two methods.
Discussion: a) In agreement with results from other pre-post-studies the German system of inpatient rehabilitation seems only slightly effective using subjective health indicators as outcomes. b) The two methods of change measurement differ markedly in terms of costs with clear advantages for the direct method. Its usefulness however seems unpaired by a low concurrent and criterion based validity. The question is should quality management in rehab medicine be driven by observation of patients or by their perceptions?