Oldridge, N. PhD
From the schools of Allied Health Sciences and Medicine, Indiana University; IU Center for Aging Research; and Regenstrief Institute for Health Care, Indiana University, Indianapolis, Indiana.
Address correspondence to: N. Oldridge, PhD, Center for Urban Population Health, 945 N. 12th Street, PO Box 342, Milwaukee, WI 53201-0342.
There is measure in all things. - —Horace
Worldwide interest in quality of life as an outcome measure in the assessment of healthcare has been driven by the extended survival of patients with previously acute and terminal conditions, the consequent increase in chronic conditions where mortality and other traditional outcome measures do not adequately measure treatment effectiveness, the rapidly increasing aging population, and the increasing costs of healthcare. These factors have led to a move away from a mortality reduction at all costs perspective toward one of a cost-effective reduction of morbidity. This, in turn, has led to the development of quality of life instruments that meet the psychometric criteria necessary for reliable and valid outcome measures in research, health policy decision-making, and increasingly in clinical practice.
A PubMed search on the term “quality of life” between 1995 and today identified 2522 randomized clinical trials; the number identified between 1985 and 1994 was 692; between 1975 and 1984 it was 57 and before 1975 it was 0. The respective numbers of trials using “quality of life,” “cardiac rehabilitation,” and “exercise training” as search terms were 292, 156, 12, and 0 and, for “cardiac rehabilitation” only, they were 12, 4, 0, and 0. This extraordinary increase in the number of published trials with quality of life outcomes (but still a small number of cardiac rehabilitation trials) is an indicator of its relevance and importance as an attribute of clinical investigation and patient care. 1,2 Yet, at the same time, quality of life was defined in only 15% of a series of articles selected at random and published before 1992 with a patient-centered rating of quality of life measured in only 17% of the articles. 1
Evidence suggests that cardiac rehabilitation plays a significant role in enhancing recovery from an acute event and also in encouraging behaviors aimed at secondary prevention of coronary heart disease. 3 A multifactorial and multidisciplinary approach increasingly is considered the most effective approach to cardiovascular risk factor reduction, with the suggestion that the standard cardiac rehabilitation exercise-based programs need to focus more on comprehensive risk factor management efforts in the secondary prevention setting. Perhaps the most useful definition still is the 1964 World Health Organization one where cardiac rehabilitation is considered to be the sum of activity required to ensure cardiac patients the best possible physical, mental, and social conditions so that they may by their own efforts regain as normal as possible a place in the community and lead an active and productive life.4 This definition focuses on the efforts of both patient and clinician and on a return to normal life, in other words, the patient’s quality of life. In 1986, we suggested that assessing health-related quality of life (HRQL) was going to be increasingly important when evaluating the effectiveness of cardiac rehabilitation. 5 This has been reinforced by the goals of contemporary cardiac rehabilitation, which include the promotion of secondary prevention, risk factor management, and improving (HRQL). 6
A major issue is whether we can define quality of life and, if we cannot, how do we measure it? Although general agreement exists that the concept of quality of life is subject to numerous interpretations with no universal definition, agreement is found in the value of measuring the patient’s perspective of the value of what is being done to him or her. 2 Although the use of the term HRQL has raised concerns on the part of some investigators, 2 a useful operational definition of it as “representing the functional effects of an illness and its consequent therapy upon a patient, as perceived by the patient” has been proposed. 7 Two basic approaches have been developed to quantify and asses HRQL. Generic HRQL instruments are designed to permit comparisons across different population, but may neither discriminate within disease groups nor be responsive to the small but important changes seen in HRQL. Two types of generic HRQL instruments exist: health profiles that measure various aspects of HRQL and preference measures that reflect the preferences of patients for health states, ranging from full health to death. On the other hand, specific HRQL instruments are used in specific populations and not used to compare outcomes in different populations (eg, myocardial infarction and cancer patients) and are claimed to be more responsive as they are designed to include items relevant to the designated patient population.
As a consequence of the different measurement properties inherent in generic and specific HRQL instruments, should both be used, where possible? A recent review of generic HRQL instruments 8 suggests that the SF-36 health survey 9 is the most commonly used of the generic instruments reviewed and that no uniformly “worst” or “best” instrument exists. Is the SF-36 able to evaluate change in patients undergoing an intervention where the expected improvement may be small but important? Is the SF-36 able to discriminate between groups of patients differing, for example, in levels of symptom severity?
The article by Hawkes, Nowak, and Speare in this issue of JCR (pages 22–25) assesses the impact of coronary artery bypass surgery and subsequent cardiac rehabilitation on quality of life using the generic SF-36 health survey as the HRQL outcome measure. Following surgery, HRQL improved by 6 months on 7 of the 8 SF-36 scales in the 149 responders with no further improvement reported over the next 6 months. No significant differences were found in any of the eight SF-36 scales at baseline or 6 months between the 38 patients who reported participating in cardiac rehabilitation and the 111 patients who did not participate. Although the authors correctly state that the results of their study need to be treated with caution, they also suggest that the SF-36 is a useful measure of return to normal function following coronary artery bypass surgery. However, in terms of the SF-36 scores, little additional benefit was seen of cardiac rehabilitation and the authors suggest that the SF-36 may not be sufficiently responsive as an outcome measure with cardiac rehabilitation. As only 38 of the initial cohort of 253 patients reported that they had participated in the cardiac rehabilitation program, with no adherence data given, the authors are correct to suggest caution. However, a number of questions need to be raised about the selection of generic health status or HRQL instruments, such as the SF-36 health survey, as the only outcome measure in a study in a specific patient population.
The appropriateness of the term “quality of life” is controversial. Do we know whether the SF-36 measures health status, quality of life, or HRQL? An argument is raised that without patient input about their own values, we assess health status and not quality of life or HRQL. 1 The SF-36 health survey purports to measure health status and consists of 36 items with eight scales, which can be combined into two component scales. However, should it be used as a measure of quality of life? Or, is it HRQL? Further, how valid are the two component scales? In light of the fact that no gold standard exists for quality of life or HRQL, investigators need to consider making a priori predictions about how the instrument will perform in a particular study because, without these predictions, they can rationalize any finding. What is the relevance of the items in the generic SF-36 to patients having coronary bypass surgery or cardiac rehabilitation when compared with the relevance of items in specific HRQL questionnaires? What is the ability of the SF-36 to discriminate between groups of patients?
As recently shown 10 and supported in the paper by Hawkes et al, the SF-36 can assess change in health status after coronary bypass surgery and the extent of the improvement appears to be clinically important. The SF-36 has been associated with greater improvement with 6 months of home-based rehabilitation than hospital-based rehabilitation after coronary bypass surgery. 11 In the report by Hawkes et al, six of the eight SF-36 scales did not differ between patients who had participated in cardiac rehabilitation when compared to those who had not, although rehabilitation patients did report a greater improvement in physical functioning and bodily pain. What is the interpretation of the clinical importance of what may be a small but statistically significant difference in HRQL? As this question has clinical relevance, it needs more attention. Are the observations of little or no observed improvement with cardiac rehabilitation because the SF-36 items are not as relevant to these patients as they are to patients having coronary bypass surgery where the major goal is to improve symptoms and so improve perceived heath status? Could it be that a 6-week cardiac rehabilitation intervention, with 1 hour of exercise and 1 hour of education per week (no adherence rates provided), simply was not effective in changing HRQL? Might specific HRQL instrument have been even more responsive to changes after surgery and responsive to the subsequent cardiac rehabilitation? Theoretically, specific HRQL instrument should be more responsive and discriminative and one of the recommendations made by Hawkes et al is to use specific HRQL instruments with items that are relevant to the patient.
Despite all the uncertainties and caveats suggested above, assessing health status, quality of life, or HRQL is important in cardiac rehabilitation. Whether a generic HRQL instrument is used, specific HRQL instruments, where possible and feasible, are recommended when evaluating the effectiveness of cardiac rehabilitation in both the research and the clinical settings as this will have an impact on healthcare policy decisions about cardiac rehabilitation at local and national levels.
1. Gill TM, Feinstein AR. A critical appraisal of the quality of quality-of-life measurements. JAMA. 1994; 272:619–626.
2. Leplege A, Hunt S. The problem of quality of life in medicine. JAMA. 1997; 278:47–50.
3. Ades PA. Cardiac rehabilitation and secondary prevention of coronary heart disease. N Engl J Med. 2001; 345:892–902.
4. WHO Expert Committee. Rehabilitation of patients with cardiovascular disease. Technical Report Series No. 270.
Geneva: WHO; 1964.
5. Oldridge NB. Cardiac rehabilitation, self-responsibility, and quality of life. J Cardiopulm Rehabil. 1986; 6:153–156.
6. Wenger NK, Froelicher ES, Smith LK, et al. Cardiac Rehabilitation. Clinical Practice Guideline #17.
Rockville, MD: US Department of Health & Human Services, Public Health Service, Agency for Health Care Policy & Research and the National Heart, Blood & Lung Institute; 1995. AHCPR # 96-0672.
7. Schipper H, Clinch J, Olweny CLM. Quality of life studies: definitions and conceptual issues. In: Spilker B, ed. Quality of Life and Pharmacoeconomics in Clinical Trials.
2nd ed. Philadelphia: Lippincott-Raven Publishers; 1996:11–23.
8. Coons SJ, Rao S, Keininger DL, Hays RD. A comparative review of generic quality-of-life instruments. Pharmacoeconomics 2000; 7:13–35.
9. Ware Jr. JE SF-36 health survey update. Spine. 2000; 25:3130–139.
10. Rumsfeld JS, Magid DJ, O’Brien M, et al. Changes in health-related quality of life following coronary artery bypass graft surgery. Ann Thorac Surg. 2001; 34:1544–1550.
11. Arthur HM, Smith KM, Kodis J, McKelvie R. A controlled trial of hospital versus home-based exercise in cardiac patients. Med Sci Sports Exerc. 2002; 34:1544–1550.
© 2003 Lippincott Williams & Wilkins, Inc.