Original Article
A new comorbidity index: the health-related quality of life comorbidity index

https://doi.org/10.1016/j.jclinepi.2010.01.025Get rights and content

Abstract

Objective

To derive and validate the health-related quality of life comorbidity index (HRQL-CI).

Study Design and Setting

Of 261 clinical classification codes (CCCs) in the 2003 Medical Expenditure Panel Survey (MEPS), 44 were identified as adult, gender-neutral, chronic conditions. The least absolute shrinkage and selection operator (LASSO) procedure identified CCCs significantly associated with the Short Form-12 physical component summary (PCS) and mental component summary (MCS) scores. Regression models were fitted with the selected CCCs, resulting in two subsets corresponding to PCS and MCS, collectively called the HRQL-CI. Internal validation was assessed using 10-fold cross-validation, whereas external validation in terms of prediction accuracy was assessed in the 2005 MEPS database. Prediction errors and model R2 were compared between HRQL-CI models and models using the Charlson-CI.

Results

LASSO identified 20 CCCs significantly associated with PCS and 15 with MCS. The R2 for the models, including the HRQL-CI (0.28 for PCS and 0.16 for MCS) were greater than those using the Charlson-CI (0.13 for PCS and 0.01 for MCS). The same pattern of higher R2 for models using the HRQL-CI was observed in the validation tests.

Conclusion

The HRQL-CI is a valid risk adjustment index, outperforming the Charlson-CI. Further work is needed to test its performance in other patient populations and measures of HRQL.

Introduction

What is new?

Key findings

  1. A new risk adjustment index was derived for use in health-related quality of life (HRQL) studies.

What this adds to what is known?
  1. The new HRQL comorbidity index (HRQL-CI) provides better predictive ability for HRQL measures than the commonly used Charlson-CI.

What is the implication, what should change now?
  1. Researchers have a comorbidity risk adjustment index to control for the potential influence of other illnesses when studying the effect of specific diseases or other risk factors associated with HRQL measures, which outperforms commonly used Charlson-CI.

Comorbidity is the existence or occurrence of any distinct additional disease or diseases during the clinical course of a patient who has an index disease under study. A comorbidity index (CI) is a weighted measure that, when conducting statistical analyses, will control for the potential influence of those illness on an outcome of interest [1], [2]. A relatively uncomplicated method of controlling for comorbidity is to use the simple count of illnesses [3]. Diagnosis methods often use data available from a medical record derived from hospital stays. These are known as discharge-based CIs [4], [5], [6], [7], [8], [9], [10]. A common diagnosis-based method was derived by Charlson et al. [11]. This method was initially developed as a chart-based weighted index that provides a simple and valid method for estimating the risk of death associated with comorbid illness, taking into account both the number and seriousness of comorbid illnesses. Other versions of the Charlson-CI use data derived from electronic claims data, electronic hospital records, and patient self-report [4], [12].

CIs are used in health services research to control for confounding in observational studies and for risk adjustment in studies of health care quality [13], [14], [15], [16]. Comorbidity affects mortality [17], [18], [19], [20], [21], health resource utilization [22], [23], [24], admission and readmission to hospital [20], [22], [25], and health-related quality of life (HRQL) or functional status [26], [27], [28]. Without adequate measures to adjust for intervening comorbidity differences, valid comparisons of health status and HRQL outcomes in population studies cannot be made [15]. Large nationally representative data sets, such as the National Health and Nutrition Examination Survey and the Medical Expenditure Panel Survey (MEPS) include data derived from health status or HRQL instruments. These measures have and will continue to be used as primary or secondary outcome variables in analyses. It is therefore of interest to control for the presence of comorbidity when using these measures to study HRQL attributed to specific illnesses.

HRQL refers to the physical, emotional, and social impact of disease and related treatments and is distinct from physiologic measures of disease [29], [30]. Generally, HRQL decreases with increasing comorbidity [31], [32], [33], [34], [35]. Two types of questionnaires are used to measure HRQL, general and disease- or intervention-specific. General measures assess concepts that are relevant to a wide range of people, including ability to function in everyday life and emotional well-being [36]. They are not specific to any age, disease, or treatment group and are designed to be broadly applied across different populations to allow for comparisons across many conditions [37], [38], [39], [40].

Researchers have validated measures of comorbidity by how well they predict mortality, health resource use, and expenditures, either as a predictor themselves or to adjust for the contributing effects of other diseases when studying the association of a specific illness. Comparatively little research has been conducted validating existing CIs with HRQL or health status as the primary outcome variable. Moreover, few studies have been conducted to develop and validate HRQL- or health status-specific CI.

The purpose of this study was to derive and validate a CI using diseases that have the greatest association with HRQL. The secondary goal was to compare the results of explanatory models that use the new index derived specifically for HRQL with a CI originally derived to predict mortality and health care resource use, the Charlson-CI using MEPS database.

Section snippets

General description of the data sets used for the study

The MEPS is a nationally representative, public domain data set maintained by the Agency for Healthcare Research and Quality. The MEPS data set was considered an appropriate data source because of the reliability and validity of the data, in particular the disease, treatment, and HRQL data. Further specific information may be found at http://www.meps.ahrq.gov/mepsweb/.

The 2003 Full-Year Consolidated Data File (HC-079) and the 2003 Medical Conditions File (HC-078) were used to derive the

Results

Descriptive statistics for subject characteristics and HRQL measures are provided in Table 1 for both 2003 and 2005 data sets. The table shows that the characteristics of the study population are very consistent across the 2 years and there has not been any temporal shift in the population characteristics.

The model selection and the model building procedure were entirely carried out on 2003 data set using 12,713 respondents. Figure 2 presents a summary of the model selection procedures where

Discussion

Controlling for comorbidity is essential in studies using general HRQL instruments. We have proven enormous gain in explaining the variation in HRQL measures by these new HRQL-CIs (from 2- to 16-fold increase in R2 across the models we studied) over the Charlson-CI. A researcher has superior power to characterize true associations between other predictors and such HRQL outcome measures by using this index as a controlling covariate in a regression model. One strength of this study was the use

Conclusion

The new HRQL-CI is a valid risk adjustment index. It outperforms the Charlson-CI when predicting health status for the SF-12 PCS and MCS, the two core single-item health status measures as well as in an asthma specific population. Further work is necessary to test its performance in other patient populations.

References (54)

  • J. Xuan et al.

    Effects of comorbidity on health-related quality of life scores: an analysis of clinical trial data

    Clin Ther

    (1999)
  • R.J. Adams et al.

    Coexistent chronic conditions and asthma quality of life: a population-based study

    Chest

    (2006)
  • J.G. VanManen et al.

    Added value of co-morbidity in predicting health-related quality of life in COPD patients

    Respir Med

    (2001)
  • J.E. Byles et al.

    Single index of multimorbidity did not predict multiple outcomes

    J Clin Epidemiol

    (2005)
  • A. Schneider et al.

    Depression and panic disorder as predictors of health outcomes for patients with asthma in primary care

    Respir Med

    (2008)
  • H.A. Wijnhoven et al.

    The influence of co-morbidity on health-related quality of life in asthma and COPD patients

    Respir Med

    (2003)
  • S. Greenfield et al.

    Recent developments and future issues in the use of health status assessment measures in clinical setting

    Med Care

    (1992)
  • P. Hitchcock Noel et al.

    Depression and comorbid illness in elderly primary care patients: impact on multiple domains of health status and well-being

    Ann Fam Med

    (2004)
  • D.A. Deyo et al.

    Adapting a clinical comorbidity index for use with IOCD-9-CM administrative databases

    J Clin Epidemiol

    (1992)
  • A.C. Elixhauser et al.

    Comorbidity measures to use with administrative data

    Med Care

    (1998)
  • S. Greenfield et al.

    The importance of co-existent disease in the occurrence of postoperative complications and one-year recovery in patients undergoing total hip replacement. Comorbidity and outcomes after hip replacement

    Med Care

    (1993)
  • B.S. Linn et al.

    Cumulative illness rating scale

    J Am Geriatr Soc

    (1968)
  • C.J. Roe et al.

    Comorbidities and prediction of length of hospital stay

    Aust N Z J Med

    (1998)
  • D.C. Voaklander et al.

    Self report co-morbidity and health related quality of life—a comparison with record based co-morbidity measures

    Soc Indic Res

    (2004)
  • L.I. Iezzoni

    Dimensions of risk

  • W.A. Ghali et al.

    Risk adjustment using administrative data. Impact of a diagnosis-type indicator

    J Gen Intern Med

    (2001)
  • R.A. Incalzi et al.

    The interaction between age and comorbidity contributes to predicting the mortality of geriatric patients in the acute-care hospital

    J Intern Med

    (1997)
  • Cited by (56)

    View all citing articles on Scopus

    This project was funded by the Agency for Health Care Research and Quality, R03 grant 1 R03 HS017461-01A1 “Developing a Comorbidity Index for Health-Related Quality of Life Studies,” Steven R. Erickson PI, from September 1, 2008 to August 31, 2009.

    Abstract of this article was accepted for poster presentation for the October 2009 meeting of the International Society for Quality of Life Research, New Orleans, LA.

    View full text