Original articles
Chronic Comorbidity and Outcomes of Hospital Care: Length of Stay, Mortality, and Readmission at 30 and 365 Days

https://doi.org/10.1016/S0895-4356(98)00160-7Get rights and content

Abstract

This article evaluates the behavior of an adaptation of the Charlson Index (CHI) applied to administrative databases to measure the relationship between chronic comorbidity and the hospital care outcomes of length of stay (LOS), in-hospital mortality, and emergency readmissions at 30 and 365 days. These outcomes were analyzed in 106,673 hospitalization episodes whose records are registered in a minimum basic data set maintained by the public health authorities of the community of Valencia, Spain. The highest comorbidity measured by the CHI was associated with greater LOS and in-hospital mortality and increased readmission at 30 and 365 days. The rate of readmissions at 1 year dropped, however, in the group with the greatest comorbidity, probably owing to an increase in mortality after hospitalization. While comorbidity does appear to increase the risk of adverse outcomes in general and mortality and readmission specifically, the second outcome is only possible if the first has not occurred. For this reason, information and selection biases derived from administrative databases, or from the CHI itself, should be taken into account when using and interpreting the index.

Introduction

Administrative databases are a very valuable source of information for research on the outcomes of hospital care 1, 2. However, any study conducted on the basis of information collected from hospitals’ administrative databases presents specific drawbacks, most pointedly because the cases under examination have not previously been assigned to random comparative groups. This aspect must be taken into consideration when information of this nature is used to compare the effectiveness of different technologies or the outcomes achieved by different providers. Another pitfall can be the confounding effect implied by different levels of severity of the patients’ conditions 3, 4. When attempting to use information from administrative databases for purposes of comparison, the researcher in fact encounters considerable methodological and conceptual problems. On the one hand, “severity” is a complex construct that includes a wide range of clinical dimensions such as the severity of the main diagnosis, number and severity of coexistent diagnoses and complications, antecedents, age, physiopathological stability, and functional status. A number of nonclinical aspects associated with the prognosis must be taken into account as well. These include the patient’s socioeconomic status, lifestyle, and other relevant variables [5]. Compounding this situation are the questionable quality of diagnostic data 6, 7, 8, the shortcomings of diagnosis coding systems 9, 10, 11 and difficulties in establishing statistical modeling on the basis of information found in these databases. All of these problems have led to generating much controversy with regard to the usefulness and the limitations involved in using administrative databases for the purpose of making evaluations and comparisons 12, 13, 14, 15, 16, 17.

Comorbidity influences many different outcomes of hospital care such as length of stay (LOS) 18, 19, 20, the development of complications 19, 20, 21, 22, 23, 24, 25, 26, surgical outcomes 18, 27, mortality in different time frames and in different types of patients 19, 24, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, functional status and quality of life 19, 41, 42, 43, 44, 45, and hospital readmission [46], although results are contradictory with regard to this last parameter [47]. As comorbidity influences such a wide range of outcomes, it is essential to take this parameter into account when comparing the effectiveness and efficiency of different medical technologies and providers. Various proprietary risk adjustment systems have attempted to account for comorbidity by incorporating certain comorbidities into their algorithms. Likewise, various comorbidity scoring systems have been developed, including the Cumulative Illness Rating Scale 41, 43, 48, the Kaplan-Feinstein Index [28], the Charlson Index (CHI)—in versions for medical records [29], administrative databases 19, 35, 36 or by interview [49]—the Index of Coexistent Diseases 32, 50, the Chronic Sub-Scale of the APACHE II System [51], the Rand Comorbidity Index [52], as well as other indices used for this purpose, such as the American Society of Anaesthesia (ASA) surgical risk classification [53], the Duke Severity of Illness Checklist [54], the Pre-Arrest morbidity Index [55] or, simple, the number of secondary diagnoses recorded [23].

These scoring systems share some features but vary in terms of many others. Among the different systems developed, the CHI offers a series of strong points. These include the fact that it is relatively easy to construct, versions for administrative databases based on the International Classification for Diseases are available, and the CHI is widely used. Hence, the objective of this study was to assess the behavior of an adaptation of the CHI for administrative databases with respect to LOS, in-hospital mortality, emergency readmission at 30 days, and, in particular, emergency readmission at 365 days. Our research hypothesis was that all adverse outcomes would increase as CHI scores rose, although in the case of readmission, in particular at 1 year, the behavior of the CHI would be determined by the competitive relationship between the risk of mortality and the risk of readmission, as greater comorbidity would be expected to be associated with an increased risk of both death and readmission. However, because the second outcome is only possible if the first has not occurred, we hypothesized that chronic comorbidity would not be linearly associated with 1-year readmission and that readmissions would begin to fall as mortality rose in groups with higher comorbidity scores.

Section snippets

Setting

The Valencia Health Service (VHS) is a public health care organization, similar in features to the British National Health Service. It provides free primary and hospital health care to the 3.84 million inhabitants of the Valencia region. The VHS contracts services with public and private centers, while it manages its own network of 19 acute hospitals, including several university hospitals. With a total of more than 7600 beds, the VHS is responsible for approximately 80% of the hospital beds in

Results

Table 2 shows the behavior of LOS and in-hospital mortality depending on levels of chronic comorbidity. The first of these variables, LOS, standardized by age and gender, significantly increased with each level of the CHI, from 8.6 days in patients with scores of 0, to 16 days in patients with scores over 4. Likewise, the in-hospital mortality rate rose from 3.9% in patients without comorbidity to 16.6% in patients with higher scores.

Table 3 shows the behavior of readmission at 1 month and 1

Discussion

The CHI was developed to classify death-at-1-year prognoses attributable to comorbidity in longitudinal studies [29] and has been used in numerous works to account for the confounding effect of comorbidity on survival or other outcomes 38, 62, 63, 64, 65, 66, 67, 68. Charlson used clinical records from a cohort of 559 patients, taken from medical services in an acute hospital and, employing a proportional hazards regression model, identified several chronic processes whose presence

Acknowledgements

This work was conducted as part of a research project funded by grant no. 92/1028 from the Fondo de Investigación Sanitaria (FIS), and complementary grant no. 06/005/1995 from the Institució Valenciana d’Estudis i Investigacions (IVEI) and the 1996 Bayer Grant of Health Economics. The Directorate for the Management of Specialized Care of the Valencia Health Service provided the databases used. The Care Activity Service of this organization and, in particular, Vicente Escoms have provided

References (77)

  • G.R. Parkerson et al.

    The Duke Severity of Illness Checklist (DUSOI) for measurement of severity and comorbidity

    J Clin Epidemiol

    (1993)
  • T.A. Dodds et al.

    A validation of the functional independence measurement and its performance among rehabilitation inpatients

    Arch Phys Med Rehabil

    (1993)
  • A.A. Jaeger et al.

    Functional capacity after cardiac surgery in elderly patients

    J Am Coll Cardiol

    (1994)
  • P.A. Robiolio et al.

    Predictors of outcome of tricuspid valve replacement in carcinoid heart disease

    Am J Cardiol

    (1995)
  • S. Peiró et al.

    Selection bias, indexes of “burden” and risk

    J Clin Epidemiol

    (1996)
  • M.E. Charlson

    Adapting a clinical comorbidity index for use with ICD-9-CM administrative dataA response

    J Clin Epidemiol

    (1993)
  • R.A. Deyo

    Adapting a clinical comorbidity index for use with ICD-9-CM administrative dataA response

    J Clin Epidemiol

    (1993)
  • M. Charlson et al.

    Validation of a combined comorbidity index

    J Clin Epidemiol

    (1994)
  • D.M. Smith et al.

    Nonelective readmissions of medical patients

    J Chron Dis

    (1985)
  • R.L. Evans et al.

    Identifying factors associated with health care usea hospital-based risk screening index

    Soc Sci Med

    (1988)
  • F.A. Connell et al.

    The use of large data bases in health care studies

    Ann Rev Public Health

    (1987)
  • M.P. Roos et al.

    A population-based health information system

    Med Care

    (1995)
  • S. Greenfield

    The state of outcomes researchAre we on target?

    N Engl J Med

    (1989)
  • L.I. Iezonni

    Risk adjustment for medical outcomes study

  • L.L. Roos et al.

    How good are the data? Reliability of one health care data bank

    Med Care

    (1992)
  • J. Green et al.

    How accurate are hospital discharge data for evaluating effectiveness of care?

    Med Care

    (1993)
  • D.C. Hsia et al.

    Medicare reimbursement accuracy under the prospective payment system, 1985 to 1988

    JAMA

    (1992)
  • L.F. McMahon et al.

    Can Medicare Prospective Payment survive the ICD9-CM disease classification system?

    Ann Intern Med

    (1986)
  • A.R. Feinstein

    ICD, POR, and DRG. Unsolved scientific problems in the nosology of clinical medicine

    Arch Intern Med

    (1988)
  • L.I. Iezzoni

    Using administrative diagnostic data to assess the quality of hospital careThe pitfalls and potential of ICD-9-CM

    Int J Technol Assess Health Care

    (1991)
  • S.F. Jencks et al.

    Interpreting hospital mortality data. The role of clinical risk adjusting

    JAMA

    (1988)
  • K.L. Kahn et al.

    Interpreting hospital mortality data. How can we proceed?

    JAMA

    (1988)
  • P.J. Sanazaro et al.

    A critique of the use of generic screening in quality assessment

    JAMA

    (1991)
  • J. Green et al.

    Analyzing hospital mortality. The consequences of diversity in patient mix

    JAMA

    (1991)
  • R.A. Hayward et al.

    An evaluation of generic screens for poor quality of hospital care on a general medicine service

    Med Care

    (1993)
  • J.P. Kassirer

    The quality of care and the quality of measuring it

    N Engl J Med

    (1993)
  • P.D. Cleary et al.

    Variations in length of stay and outcomes for six medical and surgical conditions in Massachusetts and California

    JAMA

    (1991)
  • Cited by (262)

    View all citing articles on Scopus
    View full text