This study showed that impaired health status, as measured by the EQ‐5D before discharge, is associated with a 2–3‐fold increased risk of all‐cause mortality in patients with established CAD. After adjustment for other prognostic variables, including age, risk factors, comorbidity and admission diagnosis, impaired health status remained an independent predictor of 1‐year mortality.
Several studies have reported on the predictive value of HRQL and health status questionnaires in relation to adverse clinical outcomes in patients with cardiovascular diseases.15,16,18,23
To our knowledge, this study is the first to use the EQ‐5D, a brief generic self‐perceived health status questionnaire, to predict short‐term mortality (ie, after 1 year) independently of established biomedical risk factors in patients with CAD with a relatively low overall risk. We identified reduced self‐care as the most powerful predictor of mortality. Of note, this dimension is strongly related to patients' abilities to care for themselves and adequately manage their condition. As a consequence, targeting and improving self‐care behaviour in intervention programmes could not only lead to improved HRQL but also enhance survival in this subset of patients.24,25
In addition, a major advantage of the EQ‐5D is that it is a brief and valid measure of health status that can be easily used in clinical practice.
Our findings support the recommendations of Krumholz et al17
to include health status measurements in clinical practice as an additional tool to identify patients who are at high risk for adverse outcomes. These patients may consequently benefit from a more aggressive treatment, including invasive, pharmacological and/or behavioural interventions or a combination hereof. An earlier report on the EHS‐CR showed that there is adequate room for improvement in the medical treatment of these patients, especially with respect to adjunctive pharmacology (glycoprotein IIb/IIIa inhibitors, statins and ACE inhibitors).20
Another important issue for advocating the use of health status assessment in clinical practice relates to the issue of discrepancy between patient‐rated and physician‐rated health status.26
As clinicians frequently underestimate patients' health status as reported by their patients,27
it is paramount that patients' evaluation of “how they feel” is taken into account. In addition, health status is an important patient‐centred outcome, with patients emphasising health status over prolonged survival.14
Hence, entering health status into the equation when discussing treatment options with patients may also be considered an ethical obligation.
Although this study clearly showed that the EQ‐5D provides prognostic information, little is known about the “how and why” impaired health status predicts mortality, independently of biomedical risk factors. It should be noted, however, that health status involves a much broader range of the effect of disease as experienced by the patient (ie, symptoms, functional limitation, and discrepancy between actual and desired function) compared with the focus of clinicians (ie, symptoms, signs and diagnosis).19
Further research is warranted into the mechanisms that may be responsible for the relationship between health status and mortality, as this could guide treatment with or the development of effective interventions. Emphasis should also be placed on the identification of the determinants of impaired health status, which has been advocated as a means to close the gap between research and clinical practice.17
Both depression and the distressed (type D) personality have been shown to predict impaired health status adjusting for measures of severity of disease and other risk factors.28,29
The question is whether these psychosocial risk factors are more important determinants of individual differences in clinical outcome than health status.
This study is the first to use the EQ‐5D as a predictor of mortality. Although other generic and disease‐specific health status questionnaires have been found to predict mortality, one of the major advantages of the EQ‐5D is its brevity. It comprises only six questions, whereas most of the other questionnaires ask many more questions (range 19–36) and are more taxing to patients.12,15,16,23
In addition, it is important to note that in patients with CAD a simple questionnaire such as EQ‐5D is able to discriminate between patients who have a higher mortality risk and those who do not. By contrast, we acknowledge that a lack of familiarity with the concept of health status, the perception of many clinicians that health status is a soft end point in evaluating a treatment19
and the high workload of physicians in clinical practice may be identified as barriers for implementing self‐perceived health status in every day clinical practice. However, it should be noted that it takes less than 5 min to complete the questionnaire, and health care professionals other than physicians can become involved in the assessment.
The current study has several potential limitations. Firstly, patients who did not complete the EQ‐5D questionnaire or who had missing follow‐up data had to be excluded from analyses. However, a comparison between responders and non‐responders did not show major differences. Secondly, it cannot be excluded that ill health conditions, other than cardiovascular diseases, could have had an effect on the results, as only “classical” risk factors and comorbidities were included in the database. Thirdly, health status was assessed only once, and at that time not all patients had undergone a revascularisation procedure. Finally, we used a generic rather than a disease‐specific instrument to evaluate health status; it is well known that generic measures may be less sensitive than disease‐specific measures to tap dimensions pertinent to clinical populations. Future research is needed to consider issues such as the predictive value of a single measurement as compared with serial measurements, the effect of changes in health status over time on outcomes, and comparing the results of the EQ‐5D with disease‐specific instruments. Despite these limitations, strengths of this study were the relatively large number of patients included from multiple hospitals across Europe. We were also able to adjust for a number of classical demographic and cardiovascular risk factors, showing that impaired health status is an independent predictor for mortality. Lastly, the enrolled patients are representative of “real life” practice, across a wide spectrum of European hospitals.
In conclusion, this study showed the strong incremental value of the EQ‐5D for the prediction of mortality in patients admitted with CAD, independently of other demographic, clinical and angiographic risk factors. Our results highlight the importance of including patients' subjective experience of their own health status to optimise risk stratification in clinical practice.