85
Views
0
CrossRef citations to date
0
Altmetric
Original Research

Optimal look back period and summary method for Elixhauser comorbidity measures in a US population-based electronic health record database

, , , , &
Pages 1-13 | Published online: 31 Jan 2017

Figures & data

Table 1 Patient demographic and encounter characteristics, for the complete, derivation, and validation datasets

Table 2 Prevalence of comorbid conditions by look back period, N = 3,273,298

Table 3 Adjusted odds of inhospital mortality and relative points by health condition(s) and look back period, n = 1,637,572

Table 4 Adjusted odds of inhospital mortality at 1 year and relative points by health condition(s) and look back period, n = 1,637,572

Table 5 Predictive performance by mortality outcome, look back period, and comorbidity summary method, n = 1,635,726

Figure S1 Comparisons of ROC curves across comorbidity summary methods by mortality outcome and look back period.

Abbreviations: ROC, receiver operating characteristic; AUROC, area under the receiver operating characteristic.
Figure S1 Comparisons of ROC curves across comorbidity summary methods by mortality outcome and look back period.

Figure S2 Comparisons of ROC curves across look back periods by mortality outcome and comorbidity summary method.

Abbreviations: ROC, receiver operating characteristic; AUROC, area under the receiver operating characteristic.
Figure S2 Comparisons of ROC curves across look back periods by mortality outcome and comorbidity summary method.