Introduction

This concept identifies and briefly describes the measures of comorbidity investigated in the Cancer Data Linkage in Manitoba: Expanding the Infrastructure for Research deliverable by Lix et al. (2016). Some of these measures have been used extensively in MCHP research over time, and some are used for the first time in this research. The concept also provides links to more detailed information where available, as well as the relevant section in Lix et al. (2016) that describes the methodology for developing these measures of comorbidity.

Measures of Comorbidity Used in MCHP Research

Over time, several measures of comorbidity have been used in MCHP research. Three comorbidity measures that have been extensively used in MCHP research include:

1. Charlson Comorbidity Index

The Charlson Comorbidity Index is based on a list of 17 categories/conditions identified from diagnoses codes in hospital abstracts and medical services (physician claims) data. Each condition is assigned a weight from 1 to 6. The index score is the sum of the weights for all identified conditions. An index score of 0 indicates no comorbid conditions, while higher scores indicate a greater level of comorbidity.

For more information on the Charlson Comorbidity Index measure used in Lix et al. (2016), please see the Charlson Index definition in the deliverable’s Technical Appendix.

For more detailed information on the Charlson Comorbidity Index and how it is used at MCHP, please read the Charlson Comorbidity Index concept.

2. Elixhauser Comorbidity Index

The Elixhauser Comorbidity Index is based on 31 individual conditions identified from diagnoses in hospital and physician data. Binary indicator variables are used to ascertain the presence/absence of each condition in the data source(s). The index score is based on the cumulative number of conditions present.

For more information on the Elixhauser Comorbidity Index measure used in Lix et al. (2016), please see the Elixhauser Index definition in the deliverable’s Technical Appendix.

For more detailed information on the Elixhauser Comorbidity Index and how it is used at MCHP,, please read the Elixhauser Comorbidity Index concept.

3. Johns–Hopkins Adjusted Clinical Group® (ACG®) Indices

With the Johns–Hopkins Adjusted Clinical Group (ACG) system, each patient is allotted to a single ACG group. Patients in each group have the same type and degree of comorbidity. The case–mix system identifies common combinations of morbidities that determine an individual’s need for health services. This study focused on specific comorbidity measures within the ACG system, including Resource Utilization Bands (RUBs) and Aggregated Diagnostic Groups (ADGs®).

For more information on the Aggregated Diagnosis Group ™(ADG®) measure used in Lix et al. (2016), please see the Aggregated Diagnosis Group ™(ADG®) definition in the deliverable’s Technical Appendix.

For more information on the Resource Utilization Bands (RUBs) measure used in Lix et al. (2016), please see the Resource Utilization Bands (RUBs) definition in the deliverable’s Technical Appendix.

For more detailed information on the ACG related comorbidity measures and how they are used at MCHP, please read the:


• Adjusted Clinical Group (ACG) - Overview concept; and the following sections of this same concept:
  
  
  • Resource Utilization Bands (RUBs) and
    
  • Aggregated Diagnostic Groups (ADGs®)

Additional Measures of Comorbidity Used in Lix et al. (2016)

4. Chronic Disease Score

The Chronic Disease Score comorbidity measure is based on the presence and complexity of prescription medications for select chronic conditions (categorized by ATC classification codes) and contains 24 categories. A single summary score is produced, which has a lower bound of zero but no upper bound.

For more information on the Chronic Disease Score comorbidity measure used in Lix et al. (2016), please see the Chronic Disease Score definition in the deliverable’s Technical Appendix.

5. Number of Diagnoses

The Number of Diagnoses comorbidity measure is defined as an aggregate count of diagnoses codes using the International Classification of Diseases (ICD) coding systems for hospital and physician data, based on three–digit ICD–9–CM codes. ICD–10–CA codes are converted to ICD-9-CM codes for consistent measurement.

For more information on the Number of Diagnoses comorbidity measure used in Lix et al. (2016), please see the Cumulative Number of New Diagnoses definition in the deliverable’s Technical Appendix.

6. Number of Prescription Drugs Dispensed

The Number of Prescription Drugs Dispensed comorbidity measure is defined as an aggregate count of prescriptions filled in the year prior to diagnosis, based on fourth-level Anatomic Therapeutic Chemical (ATC) classification codes.

For more information on the Number of Prescription Drugs Dispensed comorbidity measure used in Lix et al. (2016), please see the Cumulative Number of New Prescription Drugs definition in the deliverable’s Technical Appendix.

Further Information

For more information on how these measures were investigated and applied in Lix et al. (2016), please read: Chapter 4: Demonstration Project 2: Measuring Comorbidity in Cancer Patients in the deliverable.