[Health valuations for patients with chronic ischemic heart disease]

The study deals with the question of how patients with chronic ischemic heart disease assess different health situations that can be achieved by rehabilitation. Furthermore it examines which factors influence these health valuations and whether the predictors vary depending on the level of education.

The health valuations of n = 331 patients with chronic ischemic heart disease are compiled using visual analogue scales (VAS). In addition to sociodemographic questions, generic and illness-specific scales (SF-12, MacNew) for the health-related quality of life (HRQOL) are used as potential predictors of the health valuations. Additional basic medical data were provided by the physician. Hierarchical regression analyses are conducted; the sociodemographic, medical and HRQOL variables are included stepwise. Since many variables are observed for the regression models, an imputation of missing values is made.

The health dimensions "Self-care and domestic life" and "Mobility" are assigned the highest values on the VAS. The lowest preference is assigned to the dimensions "Reduction of symptoms" and "Information about the disease". The differences between the health dimensions are statistically significant. Sociodemographic variables explain up to 3.6% of the variance of health valuations, with level of education and living with a partner being the most important predictors. The medical variables included in the second step explain between 2.1 and 6.8% incremental variance; the most important predictor is the operation performed prior to rehabilitation (bypass, heart valve). The HRQOL variables in the third block provide 7.1-24.9% incremental explanation of variance, by far the highest percentage. This is mainly achieved using the 3 MacNew scales (emotional, social and physical functioning). The overall explanation of variance for the health valuations is 17.1-28.8%. For patients with a higher level of education, the total explanation of variance is about 9.2% higher on average than for the total sample and lies between 21.2 and 44.1%.

The health dimensions examined are assessed quite differently by the patients. The most significant factor of influence is the HRQOL. In the subgroup of patients with a higher level of education, the predictive strength of the variables examined is considerably higher in comparison with the total sample. Since the health valuations can be predicted to a limited extent only using other data arising during routine care, it is necessary to implement special methods for compiling health valuations.