TY - JOUR

T1 - The Ability of Comorbidity Indices to Predict Mortality After Heart Transplantation

T2 - A Validation of the Danish Comorbidity Index for Acute Myocardial Infarction, Charlson Comorbidity Index, and Elixhauser Comorbidity Index

AU - Bonnesen, Kasper

AU - Mols, Rikke E.

AU - Løgstrup, Brian

AU - Gustafsson, Finn

AU - Eiskjær, Hans

AU - Schmidt, Morten

N1 - Copyright © 2023 The Author(s). Transplantation Direct. Published by Wolters Kluwer Health, Inc.

PY - 2023

Y1 - 2023

N2 - UNLABELLED: Advanced heart failure patients often have comorbidities of prognostic importance. However, whether total pretransplantation comorbidity burden predicts mortality in patients treated with heart transplantation (HTx) is unknown. We used population-based hospital and prescription data to examine the ability of the Danish Comorbidity Index for Acute Myocardial Infarction (DANCAMI), DANCAMI restricted to noncardiovascular diseases, Charlson Comorbidity Index, and Elixhauser Comorbidity Index to predict 30-d, 1-y, 5-y, and 10-y all-cause and cardiovascular mortality after HTx.METHODS: We identified all adult Danish patients with incident HTx from the Scandiatransplant Database between March 1, 1995, and December 31, 2018 (n = 563). We calculated Harrell's C-Statistics to examine discriminatory performance.RESULTS: The C-Statistic for predicting 1-y all-cause mortality after HTx was 0.58 (95% confidence interval [CI], 0.50-0.65) for a baseline model including age and sex. Adding comorbidity score to the baseline model did not increase the C-Statistics for DANCAMI (0.58; 95% CI, 0.50-0.65), DANCAMI restricted to noncardiovascular diseases (0.57; 95% CI, 0.50-0.64), Charlson Comorbidity Index (0.59; 95% CI, 0.51-0.66), or Elixhauser Comorbidity Index (0.58; 95% CI, 0.51-0.65). The results for 30-d, 5-y, and 10-y all-cause and cardiovascular mortality were consistent.CONCLUSIONS: After accounting for patient age and sex, none of the commonly used comorbidity indices added predictive value to short- or long-term all-cause or cardiovascular mortality after HTx.

AB - UNLABELLED: Advanced heart failure patients often have comorbidities of prognostic importance. However, whether total pretransplantation comorbidity burden predicts mortality in patients treated with heart transplantation (HTx) is unknown. We used population-based hospital and prescription data to examine the ability of the Danish Comorbidity Index for Acute Myocardial Infarction (DANCAMI), DANCAMI restricted to noncardiovascular diseases, Charlson Comorbidity Index, and Elixhauser Comorbidity Index to predict 30-d, 1-y, 5-y, and 10-y all-cause and cardiovascular mortality after HTx.METHODS: We identified all adult Danish patients with incident HTx from the Scandiatransplant Database between March 1, 1995, and December 31, 2018 (n = 563). We calculated Harrell's C-Statistics to examine discriminatory performance.RESULTS: The C-Statistic for predicting 1-y all-cause mortality after HTx was 0.58 (95% confidence interval [CI], 0.50-0.65) for a baseline model including age and sex. Adding comorbidity score to the baseline model did not increase the C-Statistics for DANCAMI (0.58; 95% CI, 0.50-0.65), DANCAMI restricted to noncardiovascular diseases (0.57; 95% CI, 0.50-0.64), Charlson Comorbidity Index (0.59; 95% CI, 0.51-0.66), or Elixhauser Comorbidity Index (0.58; 95% CI, 0.51-0.65). The results for 30-d, 5-y, and 10-y all-cause and cardiovascular mortality were consistent.CONCLUSIONS: After accounting for patient age and sex, none of the commonly used comorbidity indices added predictive value to short- or long-term all-cause or cardiovascular mortality after HTx.

U2 - 10.1097/TXD.0000000000001438

DO - 10.1097/TXD.0000000000001438

M3 - Journal article

C2 - 36935871

VL - 9

JO - Transplantation Direct

JF - Transplantation Direct

SN - 2373-8731

IS - 4

M1 - e1438

ER -