SHM Comments on Draft Hospital-Wide Readmission Measure for MIPS

Download Letter

Center for Outcomes Research & Evaluation (CORE)
Yale New Haven Health Services Corporation
1 Church Street, Suite 200
New Haven, CT 06510

Dear Inpatient Outcome Measures for the MIPS Development Team:

The Society of Hospital Medicine (SHM), on behalf of the nation’s hospitalists, is pleased to provide comments on the draft Hospital-Wide All-Cause Unplanned Readmission Measure (HWR Measure) as specified for reporting under the Merit-based Incentive Payment System (MIPS). The HWR Measure was reviewed by SHM’s Performance Measurement and Reporting Committee, a group consisting of practicing hospitalists and hospitalist leaders who are experts in measurement and assessment. 

Hospitalists specialize in providing care to the nation’s hospitalized patients and are the front-line providers in America’s hospitals. They have a unique position in the healthcare system, having a hand in the performance of both the individual physician-level and hospital-level performance agendas. As such, hospitalists have a longstanding relationship with the hospital-level readmissions measures. Our comments on the HWR Measure are informed by this experience. 

Attribution 

The HWR Measure uses a novel attribution methodology to assign a single readmission case to up-to-three providers. Hospitalists are very likely to be a majority of the discharging and primary inpatient care clinicians. A multi-attribution approach encourages team-based care and prioritizes handoffs between providers during hospitalization and at discharge. SHM broadly supports a multi-attribution methodology. 

We believe it is important for outpatient providers to be engaged with this measure, as handoffs to and patient follow-up with outpatient providers are critical to reducing unplanned readmissions. As structured, the measure would attribute cases to the primary care provider who provides the plurality of primary care services over the 12 months prior to the admission, with precedence given to primary care specialties. Many patients may see specialists in a primary care role, particularly those that are dealing with chronic conditions or conditions like cancer. These specialists may also be more clinically relevant to the hospitalization and therefore more appropriate for follow-up post-discharge. We encourage the measure development team to reevaluate whether a specialty-neutral approach to the outpatient attribution may encompass more relevant patient-provider relationships.  For example, in patients with readmission for CHF, a large proportion of unplanned readmissions, attribution to the patient’s cardiologist may better target providers involved in the post-discharge care of the patient. 

We acknowledge the discussion in Appendix D.4 Excluded Attribution Rules about why the Outpatient PCP+ approach was not used. However, we believe the clinicians who are attributed cases in this measure should be relevant to the patient’s needs at the time of discharge. These may be primary care providers or may be specialists. The current attribution methodology prioritizes only primary care providers. As such, the Outpatient PCP+ may be more appropriate. Another potential approach for attributing outpatient providers to cases may be to look at the plurality of outpatient Evaluation & Management (E&M) services billed in the readmission window. 

Group Reporting 

The Measure Methodology Report for Public Comment contains a recommendation for the measure to be reported at the level of eligible clinician groups with at least 100 patients in the measure. We are broadly supportive of this measure being used for group-level reporting, particularly as the analysis shows that few eligible clinicians (about 0.7%) have 200 cases to meet the minimal value of test-retest reliability. We also note that group-level reporting further encourages team-based care and shared accountability. Given that group level reporting can meet minimal test-retest reliability with 100 cases, we support the recommendation for group level reporting with at least 100 patients. 

We are concerned with the relatively low number of eligible clinician groups who meet the 100-admission threshold. According to the analysis in Table 19, only 14.1% of eligible clinician groups meet or exceed that threshold, despite including more than 96% of patients. While the measure may be reliable at that volume cutoff and include nearly all patients, it does exclude a large number of potentially eligible provider groups. We encourage CMS and the measure development team to consider strategies for future measures and measure specifications that include more clinicians and clinician groups. 

Concern about a 30-day readmission window 

The HWR Measure uses a 30-day window for identifying readmissions. This is consistent with the current hospital-level readmission measures. This measure development period yields an excellent opportunity to explore the utility and value of using 30 days for the readmission window. 

We disagree with using 30 day as a window for measuring readmissions and encourage the measure development team to consider implementing a shorter readmission period. Recent research indicates that a shorter readmission window, such as 7 days, may be more reflective of between-hospital variation in performance.^[1] The shorter window may also be more reflective of the impact of discharge care coordination and follow-up, while a longer window incorporates confounding external factors, such as patient social and community impacts, which are beyond the providers’ or hospitals’ control. 

We believe a readmission measure should target the impact of the hospital and clinicians associated with the hospital stay, discharge process and post-discharge follow-up. Performance assessment should be focused on the areas in which providers have actual influence or direct control. A narrowed readmission window may provide the most effective means to detect true variations in actionable data, providing a better opportunity to develop effective solutions to reduce preventable readmissions. It may also provide more meaningful information to patients about the discharge and post-discharge work of a hospital, inpatient clinicians, and outpatient partners. 

SHM appreciates the opportunity to provide feedback on the HWR Measure as it continues to undergo development and testing. If you have any questions or we can provide more information, please contact Josh Boswell, Director of Government Relations, at jboswell@hospitalmedicine.org. 

Sincerely,

Gregory B. Seymann, MD, SFHM

Chair, Performance Measurement and Reporting Committee

Society of Hospital Medicine