Elsevier

Journal of Critical Care

Volume 28, Issue 1, February 2013, Pages e20-e21
Journal of Critical Care

Abstract 40
Risk-adjustment of patient subpopulations in the intensive care unit using oasis, a novel severity score

https://doi.org/10.1016/j.jcrc.2012.10.056Get rights and content

Section snippets

Objectives

Risk adjustment of patient populations is a useful technique for identifying hospitals with outlying performance, allowing for variability in patient disease severity in clinical trials and for benchmarking. Owing to varying demographics, admission health status, and care facilities available, comparison of outcomes between hospitals with distinct patient populations requires adjustment for the severity of illness of the cohort. For patients admitted to the intensive care unit (ICU), several

Methods

The OASIS requires 10 distinct physiologic measurements (features), a substantial reduction when compared with the Simplified Acute Physiology Score III (39 features), Mortality Probability Model III (23 features), or APACHE IV (140 features), yet has a similar discrimination to APACHE IV (area under the receiver operator curve of 0.87 vs 0.89). A patient's OASIS is calculated by assigning a point value for each feature dependent upon the range in which it falls. This is identical to the method

Results

Results show that using only 20% of the data, OASIS attained an R2 of 0.986, whereas APACHE IV did not display sufficient correlation to fit a regression.

Conclusions

The OASIS provides a simple alternative to more complex models when adjusting for patient severity across distinct patient groups. Here we have shown that it is easily recalibrated to subpopulations of patients and requires less data than current models. This is advantageous when the economic cost and data collection burden of implementing such a more complicated model is a limiting factor.

References (0)

Cited by (0)

View full text