A Decade Later, Lessons Learned From the Hospital Readmissions Reduction Program

Medicare's Hospital Readmission Reduction Program reduced fall-related health care use: An unexpected benefit?

OBJECTIVE

To assess whether Medicare's Hospital Readmissions Reduction Program (HRRP) was associated with a reduction in severe fall-related injuries (FRIs).

DATA SOURCES AND STUDY SETTING

Secondary data from Medicare were used.

STUDY DESIGN

Using an event study design, among older (≥65) Medicare fee-for-service beneficiaries, we assessed changes in 30- and 90-day FRI readmissions before and after HRRP's announcement (April 2010) and implementation (October 2012) for conditions targeted by the HRRP (acute myocardial infarction [AMI], congestive heart failure [CHF], and pneumonia) versus "non-targeted" (gastrointestinal) conditions. We tested for modification by hospitals with "high-risk" before HRRP and accounted for potential upcoding. We also explored changes in 30-day FRI readmissions involving emergency department (ED) or outpatient care, care processes (length of stay, discharge destination, and primary care visit), and patient selection (age and comorbidities).

DATA COLLECTION

Not applicable.

PRINCIPAL FINDINGS

We identified 1.5 million (522,596 pre-HRRP, 514,844 announcement, and 474,029 implementation period) index discharges. After its announcement, HRRP was associated with 12%-20% reductions in 30- and 90-day FRI readmissions for patients with CHF (-0.42 percentage points [ppt], p = 0.02; -1.53 ppt, p < 0.001) and AMI (-0.35, p = 0.047; -0.97, p = 0.001). Two years after implementation, HRRP was associated with reductions in 90-day FRI readmission for AMI (-1.27 ppt, p = 0.01) and CHF (-0.98 ppt, p = 0.02) patients. Results were similar for hospitals at higher versus lower baseline risk of FRI readmission. After HRRP's announcement, decreases were observed in home health (AMI: -2.43 ppt, p < 0.001; CHF: -8.83 ppt, p < 0.001; pneumonia: -1.97 ppt, p < 0.001) and skilled nursing facility referrals (AMI: -5.95 ppt, p < 0.001; CHF: -3.19 ppt, p < 0.001; pneumonia: -10.27 ppt, p < 0.001).

CONCLUSIONS

HRRP was associated with reductions in FRIs, primarily for HF and pneumonia patients. These decreases may reflect improvements in transitional care including changes in post-acute referral patterns that benefit patients at risk for falls.

Natural language processing for prediction of readmission in posterior lumbar fusion patients: which free-text notes have the most utility?

BACKGROUND CONTEXT

The increasing volume of free-text notes available in electronic health records has created an opportunity for natural language processing (NLP) algorithms to mine this unstructured data in order to detect and predict adverse outcomes. Given the volume and diversity of documentation available in spine surgery, it remains unclear which types of documentation offer the greatest value for prediction of adverse outcomes.

STUDY DESIGN/SETTING

Retrospective review of medical records at two academic and three community hospitals.

PURPOSE

The purpose of this study was to conduct an exploratory analysis in order to examine the utility of free-text notes generated during the index hospitalization for lumbar spine fusion for prediction of 90-day unplanned readmission.

PATIENT SAMPLE

Adult patients 18 years or older undergoing lumbar spine fusion for lumbar spondylolisthesis or lumbar spinal stenosis between January 1, 2016 and December 31, 2020.

OUTCOME MEASURES

The primary outcome was inpatient admission within 90-days of discharge from the index hospitalization.

METHODS

The predictive performance of NLP algorithms developed by using discharge summary notes, operative notes, nursing notes, physical therapy notes, case management notes, medical doctor (MD) (resident or attending), and allied practice professional (APP) (nurse practitioner or physician assistant) notes were assessed by discrimination, calibration, overall performance.

RESULTS

Overall, 708 patients were included in the study and 83 (11.7%) had 90-day inpatient readmission. In the independent testing set of patients (n=141) not used for model development, the area under the receiver operating curve of NLP algorithms for prediction of 90-day readmission using discharge summary notes, operative notes, nursing notes, physical therapy notes, case management notes, MD/APP notes was 0.70, 0.57, 0.57, 0.60, 0.60, and 0.49 respectively.

CONCLUSION

In this exploratory analysis, discharge summary, physical therapy, and case management notes had the most utility and daily MD/APP progress notes had the least utility for prediction of 90-day inpatient readmission in lumbar fusion patients among the free-text documentation generated during the index hospitalization.

Association between Medicare's Hospital Readmission Reduction Program and readmission rates across hospitals by medicare bed share

Background

Medicare’s Hospital Readmissions Reduction Program (HRRP), implemented beginning in 2013, seeks to incentivize Inpatient Prospective Payment System (IPPS) hospitals to reduce 30-day readmissions for selected inpatient cohorts including acute myocardial infarction, heart failure, and pneumonia. Performance-based penalties, which take the form of a percentage reduction in Medicare reimbursement for all inpatient care services, have a risk of unintended financial burden on hospitals that care for a larger proportion of Medicare patients. To examine the role of this unintended risk on 30-day readmissions, we estimated the association between the extent of their Medicare share of total hospital bed days and changes in 30-day readmissions.

Methods

We used publicly available nationwide hospital level data for 2009–2016 from the Centers for Medicare and Medicaid Services (CMS) Hospital Compare program, CMS Final Impact Rule, and the American Hospital Association Annual Survey. Using a quasi-experimental difference-in-differences approach, we compared pre- vs. post-HRRP changes in 30-day readmission rate in hospitals with high and moderate Medicare share of total hospital bed days (“Medicare bed share”) vs. low Medicare bed share hospitals.

Results

We grouped the 1904 study hospitals into tertiles (low, moderate and high) by Medicare bed share; the average bed share in the three tertile groups was 31.2, 47.8 and 59.9%, respectively. Compared to low Medicare bed share hospitals, high bed share hospitals were more likely to be non-profit, have smaller bed size and less likely to be a teaching hospital. High bed share hospitals were more likely to be in rural and non-large-urban areas, have fewer lower income patients and have a less complex patient case-mix profile. At baseline, the average readmissions rate in the low Medicare bed share (control) hospitals was 20.0% (AMI), 24.7% (HF) and 18.4% (pneumonia). The observed pre- to post-program change in the control hospitals was − 1.35% (AMI), − 1.02% (HF) and − 0.35% (pneumonia). Difference in differences model estimates indicated no differential change in readmissions among moderate and high Medicare bed share hospitals.

Conclusions

HRRP penalties were not associated with any change in readmissions rate. The CMS should consider alternative options – including working collaboratively with hospitals – to reduce readmissions.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12913-021-06253-2.