EDITORIAL COMMENT
Mapping the Journey*
Lynne Warner Stevenson, MD, FACC* and
Eldrin Lewis, MD, FACC
Brigham and Women’s Hospital, Cardiovascular Division, Boston, Massachusetts
* Reprint requests and correspondence: Dr. Lynne Warner Stevenson, Brigham and Women’s Hospital, Cardiovascular Division, 75 Francis Street, Boston, Massachusetts 02115 (Email: lstevenson{at}partners.org).
To travel hopefully is a better thing than to arrive.
—Robert Louis Stevenson (1)
The paper by Cleland et al. (2) for the Carvedilol or Metoprolol European Trial (COMET) investigators, in this issue of the Journal, portrays how the dynamic area of patient well-being extends around the single dimension of survival time. Although the comparison of carvedilol and metoprolol recapitulates the previously published results (3), this article provides timely emphasis on incorporating how patients feel. The challenge remains to understand how symptoms, quality of life, and the values placed on survival vary on average for therapies and individually for patients. The decisions to be made should determine how the questions will be asked.
Focus on the quality factor becomes increasingly critical when both quality and survival are imminently limited. The classic studies of heart failure therapies focused on patients with mild-moderate heart failure, for whom the demonstrated benefits often have not detectably impacted quality of life. Recently, in more advanced disease, biventricular pacing, left ventricular assist devices, and the nitrate-hydralazine combination have substantially improved symptoms and function (4–6), using the same tools previously challenged as insensitive.
Assessment tools for heart failure can be disease specific, relating to expected symptoms of heart failure, such as the Minnesota Living with Heart Failure questionnaire and the Kansas City Cardiomyopathy Questionnaire (3,7) (Table 1). These may provide the most sensitive indicators of impact for a heart failure intervention (8). In this study, the New York Heart Association (NYHA) symptom classification was considered to be a disease-specific tool. Although it includes patients’ description of symptoms, it is an assessment made by the physician, subject to considerable interpretation and variation.
There are many generic instruments that describe general limitations that can encompass non-cardiac side effects and underlying co-morbidities and allow extrapolation to other disease states (8). Examples include the sickness impact profile (SIP), short-form health survey (SF-36), and the EuroQOL. In this study, the primary tool used for serial assessment was the question "How have you been feeling over the past week?" with a five-point scale from very good to very poor. This score with further validation could be a generic instrument.
The adjustment of actual survival time for perceived health-related quality requires a conversion factor between 0 and 1, with 1 being the best quality. This factor can be considered a "utility" measure (9). Both disease-specific and generic quality-of-life instruments have intrinsic validity, but neither reveals the utility value placed on survival by the patient. It has often been assumed that NYHA functional classes and patient assessments of how they feel can be translated into a utility for modification of total survival time. A true utility, however, includes direct estimates of the value of survival time perceived by patients, determined directly by what duration of survival or chance of survival they would be willing to give up (9).
The two major utility tools are the time trade-off and the standard gamble, sometimes referred to as patient preferences (9). For the time trade-off utility, patients are taken through sequential questions to determine how much time they would be willing to sacrifice in order to enjoy the remaining time in perfect health. That preferred time is divided by the total time horizon offered to yield a fraction, the utility, between 0 and 1. The fraction is progressively lower as NYHA functional class is higher, but there is wide variability. There is a significant correlation with the Minnesota scores but a weak one; many patients with the worst scores still express a desire to live all remaining life without any trade-off (10). The other tool, the standard gamble, asks the question of what risk of imminent death a patient would accept for the change to achieve perfect health. Results of the standard gamble and time trade-off correlate more closely with each other than either correlates with other assessments. However, unlike the time trade-off tool, the standard gamble includes the element of risk-seeking or risk-aversion, and may also be influenced by the patient’s previous outcomes from procedures. The EuroQOL thermometer has been validated in terms of ranking different health states for pharmacoeconomic comparisons (8), but not when choices are to be made between survival and non-survival.
Assumption of utility is further complicated by the distribution of responses, which is often bimodal, particularly in heart failure. Most patients express either a willingness to trade or risk almost all remaining time, or a willingness to trade or risk almost no time for better health. Furthermore, when utility changes over time for an individual patient, it is usually by a large fraction, often reversing the preference between survival and better health (11). This threshold effect challenges analyses, but in fact seems congruent when considering how most people describe the value they place on survival, and what would change it.
In the COMET study, many assumptions about the patient journey were made to produce quality-adjusted life years. The five-point scale and the New York Heart Association functional classes were used as ranks, which were then converted to utility values between 0 and 1, with the limitations as discussed above. Considerable care was taken to repeat the analyses using several different scales for this conversion, but each is limited by the two assumptions of a proportional relationship between ranks and utility values without accounting for the threshold effect, and a fixed relationship between symptoms and utility among different patients. Well-being rank was decreased by one level if diuretic dose was increased, although the increase might have been triggered (12) by changes in weight or physical examination before detectable symptoms. For each interval of time covered, these rank values were then multiplied by the days during that interval, which included those days for which the patient was alive without hospitalization. Integration of actual patient utilities with days alive out of hospital was a secondary end point, the "patient-preference adjusted survival," pre-specified for the Evaluation Study of Congestive Heart Failure and Pulmonary Artery Catheterization Effectiveness (ESCAPE) trial (11,13).
Comparing therapies for a population in which most patients survive with reasonable quality of life, the integral of quality-adjusted survival did not differ from the primary end point for this trial or for the ESCAPE trial (3,12). For the four years of the COMET study, the difference in total adjusted life lost was accounted for by the difference in survival as previously reported (3). The days lost to hospitalization were the same in both groups. The days lost to impaired quality as translated either by the five-point score or the NYHA functional class were the same with carvedilol and metoprolol. Although not showing additional benefit, the patient journey analysis does provide reassurance that the added survival days for the carvedilol arm were not diminished by worse quality than for metoprolol survivors.
The relevance of adjusted life-years depends on the questions being asked. The comparison of two therapies with very similar clinical benefits and side effects is unlikely to be enlightened by this integrated end point. In contrast, a patient-perceived utility measurement as a modifier of days alive may illuminate comparisons between two therapies with disparate benefits and risks, such as might be used in the future for ventricular assist devices (14) and home inotropic agents (15). Such comparisons would help to guide choices of therapy for individual patients with different survival preferences.
The use of the integrated adjusted life-year end point may also facilitate comparison of the cost-effectiveness of different interventions tested in different populations. For instance, the cost per quality-adjusted life year for carvedilol versus the short-acting metoprolol as used in this trial would be approximately $40,000, based on local pharmacy costs. This compares with an estimated $125 to $275 to save a quality-adjusted life year by treating heart failure with angiotensin-converting enzyme inhibitors and metoprolol in both developed and developing countries (16). Carvedilol has not been compared with the other beta-blockers shown to improve survival in heart failure, bisoprolol (17) or the long-acting metoprolol compound (18), for mortality or cost effectiveness.
Beyond implications for trial design, sharpening focus on the patient journey provides vital information on the importance that patients place on how well they live as well as how long they live. Over the course of this trial, approximately 25% of the total possible days were discounted for impaired quality. As heart failure becomes more advanced, patients often voice a preference to trade half or more of their remaining time in order to feel better (10,11). Hearing this message, can we afford to relegate health-related quality of life to a secondary end point? For some journeys, the risk of death may be more appropriately relegated to a "safety" end point, whereas the primary outcomes are those that enable comfort, contribution, and companionship.
 |
Footnotes
|
|---|
* Editorials published in the Journal of American College of Cardiology reflect the views of the authors and do not necessarily represent the views of JACC or the American College of Cardiology. 
|
References
|
|---|
1. World of Quotes. Available at: http://www.worldofquotes.com/author/Robert-Louis-Stevenson/1/index.html. Accessed February 21, 2006..2. Cleland JGF, Charlesworth A, Lubsen J, et al. COMET Investigators A comparison of the effects of carvedilol and metoprolol on well-being, morbidity, and mortality (the "patient journey") in patients with heart failurea report from the Carvedilol or Metoprolol European Trial (COMET). J Am Coll Cardiol 2006;47:1603-1611.[Abstract/Free Full Text] 3. Poole-Wilson PA, Swedberg K, Cleland JG, et al. Comparison of carvedilol and metoprolol on clinical outcomes in patients with chronic heart failure in the Carvedilol or Metoprolol European Trial (COMET)randomised controlled trial. Lancet 2003;362:7-13.[CrossRef][Web of Science][Medline] 4. Abraham WT, Fisher WG, Smith AL, et al. Cardiac resynchronization in chronic heart failure N Engl J Med 2002;346:1845-1853.[Abstract/Free Full Text] 5. Rose EA, Gellijns AC, Moskowitz AJ, et al. Long-term use of a left ventricular assist device for end-stage heart failure N Engl J Med 2001;345:1435-1443.[Abstract/Free Full Text] 6. Taylor AL, Ziesche S, Yancy C, et al. Combination of isosorbide dinitrate and hydralazine in blacks with heart failure N Engl J Med 2004;351:2049-2057.[Abstract/Free Full Text] 7. Rector TS. A conceptual model of quality of life in relation to heart failure J Card Fail 2005;11:173-176.[CrossRef][Web of Science][Medline] 8. Berry C, McMurray J. A review of quality of life evaluations in patients with congestive heart failure Pharmacoeconomics 1999;15:247-271. 9. Torrance GW. Utility approach to measuring health-related quality of life J Chron Dis 1987;40:593-603.[CrossRef][Web of Science][Medline] 10. Lewis EF, Johnson PA, Johnson W, et al. Preferences for quality of life or survival expressed by patients with heart failure J Heart Lung Transplant 2001;20:1016-1024.[CrossRef][Web of Science][Medline] 11. Stevenson L, Hellkamp A, Leier CV, et al. Beyond body counts to days of patient-preferred survival with advanced heart failurean analysis from the ESCAPE trial. (abstr) Circulation 2005;112:II640. 12. ESCAPE Investigators and Coordinators Evaluation study of congestive heart failure and pulmonary artery catheterization JAMA 2005;294:1625-1633.[Abstract/Free Full Text] 13. Shah MR, O’Connor CM, Sopko G, et al. Evaluation study of congestive heart failure and pulmonary artery catheterization effectiveness (ESCAPE)design and rationale. Am Heart J 2001;141:528-535.[CrossRef][Web of Science][Medline] 14. Moskowitz AJ, Weinberg AD, Oz MC, Williams DL. Quality of life with an implanted left ventricular assist device Ann Thorac Surg 1997;64:1764-1769.[Abstract/Free Full Text] 15. Stevenson LW. Clinical use of inotropic therapy for heart failure: looking backward or forward? Part II: chronic inotropic therapy Circulation 2003;108:492-497.[Free Full Text] 16. Gaziano T. Cardiovascular disease in the developing world and its cost-effective management Circulation 2005;112:3547-3553.[Free Full Text] 17. The Cardiac Insufficiency Bisoprolol Study II (CIBIS-II)a randomised trial. Lancet 1999;353:9-13.[CrossRef][Web of Science][Medline] 18. Effect of metoprolol CR/XL in chronic heart failureMetoprolol CR/XL Randomised Intervention Trial in Congestive Heart Failure (MERIT-HF). Lancet 1999;353:2001-2007.[CrossRef][Web of Science][Medline]
|
Top
References