Summary Elixhauser score interally validated for mortality prediction in older adults

1. The Summary Elixhauser (S-Elixhauser) comorbidity score is internally validated for predicting 30-day and one-year mortality for older adults.

2. The S-Elixhauser comorbidity score has not been adequately externally validated.

Evidence Rating Level: 2 (Good)

Study Rundown: In observational studies, summary comorbidity scores are an important way to describe the effect of comorbidity burden on mortality. The S-Elixhauser score aims to predict in-hospital, 30-day, and one-year mortality in older adults with comorbidities. The most common comorbidities identified were uncomplicated hypertension, chronic pulmonary disease, deficiency anemia, complicated diabetes, and hypothyroidism. Further, 30-day and one-year mortality were strongly associated with metastatic cancer, dementia, congestive heart failure, weight loss, severe renal failure, leukemia, and weight gain. However, one-year mortality was underestimated in those with the highest S-Elixhauser scores. The S-Elixhauser summary score outperformed other comorbidity scores when examining a subset of conditions, including heart failure, chronic obstructive pulmonary disease, and diabetes. The S-Elixhauser summary score may outperform others as it was developed with different weights for each outcome, making it more specific. The limitation to summary risk scores is that they do not always discriminate the severity of disease, whether a patient is undergoing treatment or not, and whether the study population used to develop the scores are representative of the general population.

Click to read the study in AIM

Relevant Reading: Net reclassification improvement: computation, interpretation, and controversies: a literature review and clinician’s guide

In-Depth [retrospective cohort]: This retrospective cohort study examined a population- and outcome-specific weighted comorbidity score aimed at predicting in-hospital, 30-day, and one-year mortality in older adults. A 20% random sample of data from Medicare beneficiaries between 2017 and 2019 was used. The cohort included persons older than age 66 who were hospitalized at least once during the study period. The S-Elixhauser summary score was generated by applying the modified Agency for Healthcare Research and Quality (AHRQ) Elixhauser algorithm to define 38 different comorbidities. Separate multivariable logistic regression models were created for each of the three mortality outcomes. The baseline model for each outcome included race and ethnicity, age, and sex. Concordance statistics (c-statistics) were generated and bias-corrected calculations were completed with a 95% confidence interval (CI). The weights that were generated for the 38 comorbidities for 30-day mortality ranged from -5 for obesity to 12 for metastatic cancer. The S-Elixhauser summary score for 30-day mortality had a c-static of 0.667 (95% CI, 0.666 to 0.667). When external validation was completed, the c-statistic was 0.711 (95% CI, 0.709 to 0.713) for the Elixhauser comorbidity indicators. The individual Elixhauser comorbidity indicators had the highest c-statistic for predicting all three outcomes, 0.711 for 30-day mortality (95% CI, 0.709 to 0.713), 0.658 for in-hospital mortality (95% CI, 0.658 to 0.661), and 0.748 for one-year mortality (95% CI, 0.747 to 0.749). For patients with heart failure, chronic obstructive pulmonary disease, or diabetes, the S-Elixhauser summary score had a c-statistic ranging from 0.657 to 0.732 for predicting 30-day mortality. More research needs to be done to examine the external validity and utility of the S-Elixhauser summary score compared to existing comorbidity scores.

Image: PD

©2022 2 Minute Medicine, Inc. All rights reserved. No works may be reproduced without expressed written consent from 2 Minute Medicine, Inc. Inquire about licensing here. No article should be construed as medical advice and is not intended as such by the authors or by 2 Minute Medicine, Inc.