Original article
Quality of life: an index for identifying high-risk cardiac patients

https://doi.org/10.1016/S0895-4356(01)00368-7Get rights and content

Abstract

A sample of 945 cardiac patients admitted under emergency conditions completed a quality of life questionnaire 4 months post-discharge. Half (471) were randomly allocated to a group used to develop a logistic regression model to predict mortality and cardiovascular morbidity 8 months later. Age 65–85 years, ever having heart failure, experiencing another cardiovascular event since discharge, and low global quality of life (QOL) score were found to be predictive of these outcomes; an interaction between QOL and heart failure was also found. The model was used to formulate a risk index which was validated in the remaining 474 patients. The index defines four levels of increasing risk of adverse outcomes, with rates in the development and validation groups, respectively, of: low risk 4% and 9%; moderate risk 13% and 15%; high risk 31% and 33%; very high risk 52% and 40%. Scores in the emotional, physical and social QOL domains were also found to be predictive of adverse outcomes, suggesting that interventions in any of these areas may prove beneficial. The index may be useful for follow-up evaluation of cardiac patients.

Introduction

While quality of life (QOL) is often used as a measure of outcome and for evaluation of treatment, it can also be important as a prognostic indicator for later health outcomes. Overall QOL, physical activity level, social support, depression, and various other health-related QOL dimensions have been reported to be significant predictors of survival and rehospitalization both in the general population 1, 2 and in a variety of disease settings such as cancer 3, 4, 5, pulmonary disease [6], renal disease 7, 8, 9, organ transplants [10], and for psychiatric patients [11]. Other studies have addressed the association between psychosocial variables and survival with regard to cardiac patients, often focussing on depression as a prognostic factor for mortality 12, 13, 14, but in recent years more general QOL measures have been examined.

The relationships between QOL and specific psychosocial factors predictive of later outcomes in cardiac patients, such as social support and depression, suggest that QOL might also be predictive of later outcomes and thus useful for prognosis and risk assessment. Recently, Rumsfield et al. showed that the physical component summary score of the Short Form 36 is an independent risk factor for mortality following coronary artery bypass graft (CABG) surgery [15]. Konstam et al. reported that baseline assessment of health-related QOL, in particular the domains of activities of daily living and self-reported general health, independently predicted mortality and rehospitalization in patients with congestive heart failure [16]. Lim et al. found a highly significant association between QOL and later adverse outcomes in myocardial infarction (MI) patients assessed using a heart-specific QOL instrument [17].

Several risk indices for CABG surgery 18, 19, severe heart failure [20], MI 21, 22, 23, 24, 25 and cardiac complications of surgery [26] have been proposed. These have included a wide range of demographic (e.g., age, sex), simple clinical (e.g., body mass, blood pressure, comorbidities), and more complex biochemical (e.g., peak VO2, serum creatinine levels) variables. They mostly rely on data collected in hospital, sometimes from invasive clinical tests, and are typically designed to be used on admission to predict the short-term outcomes of in-hospital morbidity and mortality. Although some of the short-term indicies can be adapted (e.g., [24]), few indicies for longer-term prognosis are available. Elmore and colleagues [27] found that the addition of post-hospital data improved the prognostic value of a 1-year mortality index for first MI patients, and recommended the inclusion of such data in subsequent studies. However, despite the reported association between QOL and morbidity/mortality, we were unable to discover any coronary risk indices that have considered QOL as a prognostic factor. We postulate that QOL may be an important factor for the post-discharge identification of cardiac patients at high risk of later morbidity and mortality.

The aims of this analysis were: (1) to test the hypothesis that a heart-specific QOL measure was independently predictive of mortality and morbidity in cardiac patients after controlling for other clinical and demographic variables; and (2) to conduct exploratory analysis into the possibility of using the QOL measure in combination with routinely collected variables to develop a simple risk index for identifying cardiac patients at risk of adverse outcomes in the medium term (up to 8 months from evaluation).

Section snippets

Subjects

Subjects are drawn from all 15 public and three of the seven private hospitals in the Hunter Region of Australia (adult population approx. 340,000). Between December 1, 1996 and January 31, 1998, patients between the ages of 20 and 85 who were discharged alive with diagnoses of acute myocardial infarction (AMI), unstable angina, angina pectoris, chronic ischaemic heart disease (IHD), or heart failure (ICD-9-CM codes 410, 411.1, 413, 414 and 428, respectively) were eligible to participate in a

Sample population

Questionnaire packages were sent to 2070 patients, all of whom had been admitted as emergency cases at index. Of these, 94 had died since discharge, 321 did not respond, 271 did not wish to participate, and 31 packages were returned name unknown. Thirty-eight patients were later found not to have eligible diagnoses and were excluded from the study. Of the 1300 questionnaires which were returned (a response rate of 1300/1907 = 68%), 1153 patients (89%) answered within the 3–5 month limit. Only

Discussion

Quality of life, measured using the heart-specific instrument MacNew, was found to independently predict mortality and morbidity after adjusting for other clinical and demographic variables. Using multivariate logistic regression modelling, four prognostic factors were found to be independently associated with the adverse outcome of death or emergency CV rehospitalization in this study. Age 65–85 years, ever diagnosed with heart failure, experiencing another CV event since index, and low global

Acknowledgements

This research is supported by a grant from the National Health and Medical Research Council of Australia. The authors thank Heather Powell, R.N. for review of medical records and providing clinical advice, and Janet Fisher, Data Manager of the Hunter Area Heart and Stroke Register, for readmission and mortality data. This work was performed at the National Centre for Epidemiology and Population Health, Australian National University, Canberra, and the Centre for Clinical Epidemiology and

References (35)

  • T. Killip et al.

    Treatment of myocardial infarction in a coronary care unita two year experience with 250 patients

    Am J Cardiol

    (1967)
  • J.G. Elmore et al.

    The evolving clinical status of patients after a myocardial infarctionthe importance of post-hospital data for mortality prediction

    J Clin Epidemiol

    (1996)
  • L.L.-Y. Lim et al.

    A self-administered quality of life questionnaire after acute myocardial infarction

    J Clin Epidemiol

    (1993)
  • T.K. Hillers et al.

    Quality of life after myocardial infarction

    J Clin Epidemiol

    (1994)
  • M.E. Charlson et al.

    A new method of classifying prognostic comorbidity in longitudinal studiesdevelopment and validation

    J Chron Dis

    (1987)
  • S.K. Inouye et al.

    Importance of functional measures in predicting mortality among older hospitalized patients

    JAMA

    (1998)
  • G.A. Kaplan et al.

    Perceived health and mortalitya nine-year follow-up of the Human Population Laboratory group

    Am J Epidemiol

    (1983)
  • Cited by (29)

    • The Prognostic Effect of Physical Health Complaints With New Cardiac Events and Mortality in Patients With a Myocardial Infarction

      2017, Psychosomatics
      Citation Excerpt :

      Future studies should address this issue. In contrast with some studies, which measured reported health almost immediately after the index MI7 or a year later,3 we chose, like others,5,9 to assess baseline HCS 3 months post-MI. We consider this an advantage, as the immediate effect of the initial stressful cardiac event and hospitalization is settled down and a more representative view of which complaints persist can be made.

    • Change in health-related quality of life in patients with coronary artery disease predicts 4-year mortality

      2014, International Journal of Cardiology
      Citation Excerpt :

      The use of the generic SF-36 Health Survey in the Schenkenveld et al. study [23] rather than a disease-specific HRQL instruments may also provide a partial explanation for the observed lack of an association between deteriorated HRQL and mortality in that study. A significant relationship between adverse events, both mortality and poorer HRQL, has been demonstrated using the disease-specific MacNew after the event in a non-intervention [13] and after the procedure in an intervention study [22]. As the MacNew has been validated as both a discriminative and an evaluative outcome measure in each of the major CAD diagnoses, MI, angina, and heart failure [29,35], the present study substantiates the potential of the MacNew as a predictive patient-reported outcome measure in patients with CAD [13,22].

    • Health status as a risk factor in cardiovascular disease: A systematic review of current evidence

      2009, American Heart Journal
      Citation Excerpt :

      The follow-up period across studies varied considerably, ranging from 1 month to 10 years. Eight studies used a short-term follow-up period between 1 and 6 months,6,7,14,25,26,29,39,40 seven studies having a mean follow-up of 12 months.8,13,15,22,28,31,34 Nine studies used a follow-up between 14 months and 2.7 years,5,6,12,16,19,32,33,36,38 seven studies between 3 and 3.5 years,17,18,20,21,23,24,35, and 4 studies between 5 and 10 years.4,27,30,37

    View all citing articles on Scopus
    View full text