Triple Aim and the Hospital Readmission Reduction Program

AI Models for Predicting Readmission of Pneumonia Patients within 30 Days after Discharge

A model with capability for precisely predicting readmission is a target being pursued worldwide. The objective of this study is to design predictive models using artificial intelligence methods and data retrieved from the National Health Insurance Research Database of Taiwan for identifying high-risk pneumonia patients with 30-day all-cause readmissions. An integrated genetic algorithm (GA) and support vector machine (SVM), namely IGS, were used to design predictive models optimized with three objective functions. In IGS, GA was used for selecting salient features and optimal SVM parameters, while SVM was used for constructing the models. For comparison, logistic regression (LR) and deep neural network (DNN) were also applied for model construction. The IGS model with AUC used as the objective function achieved an accuracy, sensitivity, specificity, and area under ROC curve (AUC) of 70.11%, 73.46%, 69.26%, and 0.7758, respectively, outperforming the models designed with LR (65.77%, 78.44%, 62.54%, and 0.7689, respectively) and DNN (61.50%, 79.34%, 56.95%, and 0.7547, respectively), as well as previously reported models constructed using thedata of electronic health records with an AUC of 0.71–0.74. It can be used for automatically detecting pneumonia patients with a risk of all-cause readmissions within 30 days after discharge so as to administer suitable interventions to reduce readmission and healthcare costs.

Association Between Racial Disparities in Hospital Length of Stay and the Hospital Readmission Reduction Program

Background:

On average Black patients have longer LOS than comparable White patients. Longer hospital length of stay (LOS) may be associated with higher readmission risk. However, evidence suggests that the Hospital Readmission Reduction Program (HRRP) reduced overall racial differences in 30-day adjusted readmission risk. Yet, it is unclear whether the HRRP narrowed these LOS racial differences.

Objective:

We examined the relationship between Medicare-insured Black-White differences in average, adjusted LOS (ALOS) and the HRRP’s implementation and evaluation periods.

Methods:

Using 2009-2017 data from State Inpatient Dataset from New York, New Jersey, and Florida, we employed an interrupted time series analysis with multivariate generalized regression models controlling for patient, disease, and hospital characteristics. Results are reported per 100 admissions.

Results:

We found that for those discharged home, Black-White ALOS differences significantly widened by 4.15 days per 100 admissions (95% CI: 1.19 to 7.11, P < 0.001) for targeted conditions from before to after the HRRP implementation period, but narrowed in the HRRP evaluation period by 1.84 days per 100 admissions for every year-quarter (95% CI: −2.86 to −0.82, P < 0.001); for those discharged to non-home destinations, there was no significant change between HRRP periods, but ALOS differences widened over the study period. Black-White ALOS differences for non-targeted conditions remained unchanged regardless of HRRP phase and discharge destination.

Conclusion:

Increased LOS for Black patients may have played a role in reducing Black-White disparities in 30-day readmission risks for targeted conditions among patients discharged to home.