Implementation of the Patient-Driven Payment Model (PDPM) significantly increased coding intensity among skilled nursing facilities, while Medicare expenditures grew by $665 per beneficiary on average, despite the model’s intended budget neutrality.

Increased spending was observed especiallyamong patients with medically complex health care needs, according to a study published on Monday in JAMA Internal Medicine. Meanwhile, there were no significant changes among rehospitalizations, functional outcomes or mortality rates, the study found. 

“Overall, SNF consistently demonstrated lower coding intensity compared with hospitals during both the pre- and post-PDPM periods,” study authors said. “However, as the PDPM implementation date approached, we observed an increase in the SNF coding intensity, whereas the hospital coding intensity remained stable.”

The mean comorbidity index in nursing homes increased from 9.9 to 10.3 between Jan. 1, 2018 and Feb. 29, 2020, the study found, with the steepest changes in the 100 days leading up to PDPM implementation on Oct. 1, 2019.

PDPM implementation shifted from therapy volume-based payment to reimbursement based on clinical and functional characteristics.

It appears that the incentive structure of PDPM influenced SNF coding and spending patterns without affecting short-term outcomes. This trend highlights the need for continued monitoring of PDPM to ensure an alignment between payment policy and high-quality care.

While coding intensity varied by primary diagnoses and conditions, the most pronounced rise was seen among patients hospitalized for congestive heart failure. In terms of diagnosis-specific coding intensity, comorbid conditions with the largest increases in frequency from PDPM implementation were tied to weight loss, complicated hypertension, obesity and chronic pulmonary disease.

In other words, coding intensity increased especially for diagnoses with higher payment weights like malnutrition, obesity and complicated hypertension. PDPM either incentivized upcoding or improved diagnostic documentation, the study found.

The wide variation in coding intensity particularly among for-profit operators raises concerns about equity and consistency in payment practices, according to the study. Authors suggested more monitoring to assess PDPM’s impact on care quality access and facility-level disparities.

The rise in Medicare spending was notable among clinically complex patients, including those with congestive heart failure, Alzheimer’s-related dementia and chronic obstructive pulmonary disease. Authors see this change as a realignment toward more medically complex patients and away from therapy-intensive cases like stroke.

The study included more than 2 million Medicare beneficiaries.

An alternative model

The study suggests that changes to PDPM are needed in order to meet key goals such as patients’ return to pre-hospitalization baseline status, completion of therapies requiring skilled nursing care, and prevention of postacute complications, according to an invited commentary on the findings that also was published in JAMA Internal Medicine. 

To achieve these goals, the discharging hospital must be held accountable in certain areas, including setting realistic goals before discharging patients and providing complete clinical information about the hospital stay, the commentary authors wrote.

They also argued that SNF reimbursement should be tied more directly to staffing levels that match clinical complexity of patient populations.

“For example, patients who require intravenous antibiotics to complete treatment or frequent venipuncture to monitor for drug toxic effect require more nursing time, and SNF payment linked to staffing could address the higher needs,” they wrote.

Finally, Medicare could require that SNF patients see a physician or advanced practice clinician within three days, and increase reimbursement to these clinicians for seeing SNF patients. This change would address currently inadequate access to such clinicians, which increases the risk of postacute complications by making early diagnosis and management less likely.

The commentary was written by authors affiliated with the Perelman School of Medicine and the Leonard Davis Institute of Health Economics, University of Pennsylvania; and the Division of Hospital Medicine, University of California, Los Angeles.