Outcome
Risk adjustment for congenital heart surgery: the RACHS-1 method

https://doi.org/10.1053/j.pcsu.2004.02.009Get rights and content

Abstract

The new health care environment has increased the need for accurate information about outcomes after pediatric cardiac surgery to facilitate quality improvement efforts both locally and globally. The Risk Adjustment for Congenital Heart Surgery (RACHS-1) method was created to allow a refined understanding of differences in mortality among patients undergoing congenital heart surgery, as would typically be encountered within a pediatric population. RACHS-1 can be used to evaluate differences in mortality among groups of patients within a single dataset, such as variability among institutions. It can also be used to evaluate the performance of a single institution in comparison to other benchmark data, provided that complete model parameters are known. Underlying assumptions about RACHS-1 risk categories, inclusion and exclusion criteria, and appropriate and inappropriate uses are discussed.

Section snippets

Applying RACHS-1 to specific research questions

RACHS-1 can be used to evaluate differences in mortality among groups of patients within a single dataset, such as variability among institutions. Adjusted mortality rates can be calculated, by correcting observed mortality rates for more or less complex case mixes, in comparison to the average within the dataset. Alternatively, standardized mortality ratios can be calculated as the ratio of the observed deaths over the expected deaths, based on the average performance with similar cases within

Clarification of RACHS-1 risk categories

A major component of RACHS-1 is the grouping of individual types of cardiac procedures with similar risks for in-hospital death together into six risk categories. The risk categories were created by consensus of the panel using collective judgment and actual mortality data from two large datasets. Several important points regarding this component of RACHS-1 should be emphasized. First, a major goal in the development of RACHS-1 was the creation of a tool that would be useful in the assessment

Clarification of other RACHS-1 components

Similar reasoning was used for other case mix exclusions incorporated into the RACHS-1 methodology. Because the purpose of RACHS-1 was to allow comparisons of mortality caused by congenital heart surgical procedures, cases with significant risk factors for death from noncardiac surgical causes, such as premature infants who underwent patient ductus arteriosus ligation, were not included. Similarly, situations where in-hospital mortality differences were less relevant as a basis for comparison,

Limitations in interpretation of analyses using RACHS-1

To appropriately use RACHS-1, it is important to understand that the method was not developed to do true “predictive modeling” for individual cases, but rather to allow meaningful comparisons of mortality for groups of patients undergoing congenital heart surgery. Predictive modeling attempts to assess the actual likelihood that individual patients will die, and might be expected to include numerous distinguishing features of individual cases that alter risk for death. Conditions that occur

RACHS-2

The RACHS-1 method is an excellent research tool with acceptable measurement properties. Because the tool merely requires that certain variables be incorporated into analyses in certain ways, without specifying model parameters, it can be flexibly used to allow risk adjustment in more recent data sets, even though the initial validation was completed using data from 1993 to 1995. Even if risk groups converge as mortality rates decrease, use of the RACHS-1 method will retain validity because

References (6)

There are more references available in the full text version of this article.

Cited by (259)

  • Congenital cardiac surgery and artificial intelligence

    2023, Intelligence-Based Cardiology and Cardiac Surgery: Artificial Intelligence and Human Cognition in Cardiovascular Medicine
View all citing articles on Scopus

Supported in part by grant no. National Heart, Lung, and Blood Institute K08-HL02936.

View full text