Tool for Predicting Discharges to Skilled Nursing Could Lead to Further Diversions Home

In a reimbursement landscape where hospitals increasingly look to cut nursing homes out of the continuum of care entirely, a recent study found that it’s possible to predict “with good accuracy and clinical usability” which patients will end up being discharged to a skilled nursing facility.

The index — which was developed by researchers from Massachusetts General Hospital, Harvard Medical School, and the University of California — was able to predict whether a patient will eventually use the services of a SNF on the first day of hospitalization, according to a study published in the June issue of JAMDA.

Specifically, the researchers found that problems with mobility or the impaired ability to bathe upon hospital admission were among the most significant predictors of a patient being discharged to a SNF. That suggests that factors other than a given resident’s clinical needs are significant in determining whether or not a patient goes to a SNF.

Advertisement

“There is increasing evidence that discharge to post-acute facilities is often influenced by non-skilled needs (e.g. availability of caregiver, marital status and living alone),” the researchers wrote. “It therefore follows that the successful identification of patients at risk for SNF discharge will require more than the evaluation of traditional administrative and clinical data; rather, it requires an assessment of the patient’s functional status, social support, and living situation as well.”

Clinicians, researchers. and administrators need to use that data — particularly as electronic medical records (EMRs) increasingly capture nursing assessments — to inform patient care needs upon hospital discharge, the researchers argued. And those care needs have long been a sticking point for hospitals and SNFs alike.

For example, Henry Ford Health System in Detroit found that it had little timely data on its patients in post-acute facilities, which prompted it to turn to a partner with CarePort Health, a care coordination software company, to follow its patients.

Advertisement

The issue of patients disappearing from hospitals’ radar after they leave the acute setting had been a notable issue for some time, CarePort CEO Lissy Hu told Skilled Nursing News last month.

“I think there was a time when there was a mentality of: We send our patients out from the hospital to the SNF, and that was it,” she told SNN. “It was just a black hole.”

But the index now allows some insight into what might make patients more likely to be sent to a SNF, and a surprising number of the variables were related to social factors, rather than specific conditions.

The risk tool described in the study includes several questions related to patient diagnosis, so clinical factors shouldn’t be ignored entirely. But significant demographic variables included age, marital status, insurance type, household size, residence, and distance from the hospital.

In some ways, the tool might end up diverting patients from the SNF; the researchers noted that identifying family and community resources early could allow a hospital to send patients directly home, if their primary reason for SNF discharge is the lack of home support.

And the constant push by accountable care organizations (ACOs) and other alternative payments to save money could end up playing a role. ACOs achieve much of their savings by cutting SNF usage, while mandatory bundled payment models have routinely led to drops in skilled nursing utilization — most recently illustrated in New Jersey, where discharges to SNFs dropped from 45% to 26% for certain common joint replacement episodes in a three-year span.

The researchers noted these trends as well.

“For example, hospitals enrolled in Medicare’s Bundled Payments for Care Improvement (BPCI) Model 2 are responsible for the total cost of patient care, including SNF care, within 90 days of discharge,” the researchers wrote. “The wide variation in SNF use and unclear association with outcomes indicate that it is a ripe area for efforts to improve value within the BPCI model … The model presented here may help further these efforts.”

Companies featured in this article: