EDITORIAL COMMENT
Prognostic IndicatorsUseful for Clinical Care?*
Stephen S. Gottlieb, MD, FACC*
Division of Cardiology, University of Maryland, Baltimore, Maryland
* Reprint requests and correspondence: Dr. Stephen S. Gottlieb, Division of Cardiology, University of Maryland, 22 South Greene Street, Baltimore, Maryland 21201 (Email: sgottlie{at}medicine.umaryland.edu).
Key Words: heart failure • prognosis • statistical models
Do we really need another study looking at prognostic indicators of mortality of heart failure patients? After all, innumerable studies demonstrate the predictive importance of a slew of natriuretic peptides, inflammatory markers, and electrocardiographic parameters, not to mention countless demographic and readily available blood tests. However, have they improved clinical care?
At this point, the answer is probably "no." Do we (and should we) treat a patient differently because the B-type natriuretic peptide concentration suggests a high mortality rate? Hopefully, we maximize medications for all patients. Should blood transfusions be administered because anemia is a poor prognostic sign? Such a reaction would assume that the anemia is the cause of the worse outcome, a questionable hypothesis. Do we place defibrillators because of decreased heart rate variability? At present, devices are placed because of the results of randomized studies and clinical judgment.
One could interpret the excellent study by Kalogeropoulos et al. (1) in this issue of the Journal as providing more reasons to not care about prognostic studies. For reasons we don't understand, prognostic factors in virtually all cardiovascular disease are different on the basis of race. Despite controlling for multiple factors (including risk factors, symptoms, demographic variables, and comorbidities) in regression models, many studies have shown an unexplained impact of ethnicity. For example, in the SWAN (Study of Women's Health Across the Nation) study, African Americans had elevated C-reactive protein concentrations as compared with whites, whereas Japanese and Chinese women had lower levels (2). Similarly, the likelihood of significant coronary artery disease and in-hospital mortality in the American College of Cardiology–National Cardiovascular Data Registry varied significantly by ethnicity and sex (3). Intracerebral bleeds were seen more in African Americans, Hispanics, and Asians with atrial fibrillation than in whites (4).
The reasons for these differences are not known. Genetic variation might be the explanation, and the significance of polymorphisms is actively being studied (5). However, it is also possible that we are just missing knowledge of important lifestyle characteristics affecting prognosis or neglecting to control for an important risk factor. Whatever the reason, these studies all point out the limitations of regression analyses, which do not (and cannot) control for the multiple factors that impact prognosis.
A prognostic indicator can also be limited by differences in the disease process. Perhaps the findings of Kalogeropoulos et al. (1) reflect that the physiology of acute heart failure is qualitatively different from that of chronic heart failure. Even the impact of proven (chronic) treatments might be altered in the acute state. These considerations should alert us to considering as many factors as possible (and realizing the differences among individuals) when evaluating and treating patients.
But the many limitations discussed in the preceding text and the common concern that patients all need to be treated as individuals should not blind us to the utility of the Seattle Heart Failure Model or similar tools. There is one area where prognosis should be paramount in our thinking and our knowledge has not been translated into action: the timing of transplantation and left ventricular assist device (LVAD) implantation. Most programs still use a peak oxygen consumption (usually with a cutoff of 14 ml/kg/min) as a major factor to define appropriateness of transplantation. This derives from a study from 1991 (!) (6) and is used despite the knowledge that many factors affect oxygen consumption (7).
There is increasing realization that data-driven conclusions might be more accurate than clinical judgment in many medical circumstances. In the emergency room, natriuretic peptide concentrations seem better than clinical judgment (8). In cancer, objective parameters predict mortality better than physicians (or committees) (9). Indeed, a meta-analysis of comparisons of clinical and mechanical (defined as formal, statistical) predictions in various circumstances showed the superiority of the mechanical (10).
This should not be surprising. Physicians are more likely to be influenced by a short-term perspective, whereas clearly a long-term perspective is needed when considering transplantation and LVAD use. In many circumstances, such as in patients after myocardial infarction or with chronic heart failure, the use of evidence-based medicine has clearly improved outcomes. Yet, the emotion associated with hospitalized patients with heart failure has prevented us from looking critically at data regarding these patients, even though there are signs that we might not be transplanting the correct patients. Shouldn't we be concerned that status 2 patients might do better with medical management than with transplantation (11)?
What message should we take from the study by Kalogeropoulos et al. (1) about prognostication? We could focus on the demonstrated limitations of prognostication. Yes, there are factors we do not understand. The calculated prognosis might be influenced by the specific patient population evaluated. Race and acuity seem to impact the prognostic ability and can change the expected mortality. The use of defibrillators might influence mortality slightly differently than predicted. Undoubtedly, there are modifying factors omitted in all models that attempt to quantify expected prognosis of patients with heart failure.
More appropriate, however, would be to look at how much information we can derive from well-designed and tested heart failure scores. In 2008 we can reliably predict the risk of mortality of heart failure patients. We will never know with certainty what will happen to an individual patient, but tested scores are better than clinical judgment in knowing the chances of death for that patient. Although future modifications will continue to improve predictive models, present imperfections should not prevent us from realizing that we can now identify black and white, acute and chronic patients who will have a better outcome with medical therapy than that expected with transplantation (or vice versa). It is time for transplantation selection committees and physicians implanting LVADs to use the most objective evidence to determine which patients are likely to be helped and which will be harmed by surgical intervention. We can reliably compare the expected medical prognosis with that of transplantation or LVAD implantation (12).
Whether the Seattle Heart Failure Model, a modification, or a different prognostic indicator is used, transplants should not be performed in people with a good predicted outcome. Left ventricular assist devices should not be implanted when the predicted mortality is worse with the intervention than without. There are too many patients who can benefit from transplantation and too much money spent on counterproductive procedures to continue to use outdated criteria. Physicians might all believe that their own judgment is exceptional, but medical care will improve and be more cost effective if advanced heart failure treatment is based upon data and logical thinking.
 |
Footnotes
|
|---|
* Editorials published in the Journal of the 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. Kalogeropoulos AP, Georgiopoulou VV, Giamouzis G, et al. Utility of the Seattle Heart Failure Model in patients with advanced heart failure J Am Coll Cardiol 2009;53:334-342.[Abstract/Free Full Text]2. Kelley-Hedgepeth A, Lloyd-Jones DM, Colvin A, et al. SWAN Investigators Ethnic differences in C-reactive protein concentrations Clin Chem 2008;54:1027-1037.[Abstract/Free Full Text] 3. Shaw LJ, Shaw RE, Merz CN, et al. American College of Cardiology-National Cardiovascular Data Registry Investigators Impact of ethnicity and gender differences on angiographic coronary artery disease prevalence and in-hospital mortality in the American College of Cardiology-National Cardiovascular Data Registry Circulation 2008;117:1787-1801.[Abstract/Free Full Text] 4. Shen AY, Yao JF, Brar SS, Jorgensen MB, Chen W. Racial/ethnic differences in the risk of intracranial hemorrhage among patients with atrial fibrillation J Am Coll Cardiol 2007;50:309-315.[Abstract/Free Full Text] 5. Yancy CW. The role of race in heart failure therapy Curr Cardiol Rep 2002;4:218-225.[Medline] 6. Mancini DM, Eisen H, Kussmaul W, Mull R, Edmunds Jr. LH, Wilson JR. Value of peak exercise oxygen consumption for optimal timing of cardiac transplantation in ambulatory patients with heart failure Circulation 1991;83:778-786.[Abstract/Free Full Text] 7. Elmariah S, Goldberg LR, Allen MT, Kao A. Effects of gender on peak oxygen consumption and the timing of cardiac transplantation J Am Coll Cardiol 2006;47:2237-2242.[Abstract/Free Full Text] 8. Green SM, Martinez-Rumayor A, Gregory SA, et al. Clinical uncertainty, diagnostic accuracy, and outcomes in emergency department patients presenting with dyspnea Arch Intern Med 2008;168:741-748.[Abstract/Free Full Text] 9. Gripp S, Moeller S, Bölke E, et al. Survival prediction in terminally ill cancer patients by clinical estimates, laboratory tests, and self-rated anxiety and depression J Clin Oncol 2007;25:3313-3320.[Abstract/Free Full Text] 10. Grove WM, Zald DH, Lebow BS, Snitz BE, Nelson C. Clinical versus mechanical prediction: a meta-analysis Psychol Assess 2000;12:19-30.[CrossRef][Web of Science][Medline] 11. Lietz K, Miller LW. Improved survival of patients with end-stage heart failure listed for heart transplantation: analysis of organ procurement and transplantation network/U.S. United Network of Organ Sharing data, 1990 to 2005 J Am Coll Cardiol 2007;50:1282-1290.[Abstract/Free Full Text] 12. Lietz K, Long JW, Kfoury AG, et al. Outcomes of left ventricular assist device implantation as destination therapy in the post-REMATCH era: implications for patient selection Circulation 2007;116:497-505.[Abstract/Free Full Text]
Related Articles
-
Utility of the Seattle Heart Failure Model in Patients With Advanced Heart Failure
- Andreas P. Kalogeropoulos, Vasiliki V. Georgiopoulou, Grigorios Giamouzis, Andrew L. Smith, Syed A. Agha, Sana Waheed, Sonjoy Laskar, John Puskas, Sandra Dunbar, David Vega, Wayne C. Levy, and Javed Butler
J. Am. Coll. Cardiol. 2009 53: 334-342.
[Abstract]
[Full Text]
[PDF]
-
Inside This Issue of JACC
J. Am. Coll. Cardiol. 2009 53: A20.
[Full Text]
[PDF]
|
Top
References