Personalized ratings app may improve patient’s choice of nursing home

Their study, entitled "When Patients Customize Nursing Home Ratings, Choices and Rankings Differ from the Government's Version," appears in the April issue of Health Affairs.

"Medical care is undergoing a cultural transformation, moving towards personalization and patient-centered care," said Dana B. Mukamel, professor in the Department of Medicine and research study leader. "Our work brings this philosophy to provider choice and quality reporting."

Most patients and their families rely on the five-star ranking system devised by CMS and featured on the Nursing Home Compare website. The rankings are calculated according to a fixed subset of quality measures, with each measure given a fixed importance weight. This five-star ranking, according to Mukamel, is a "one size fits all patients" measure that does not reflect individual patient needs and preferences.

"When patients and their families have the option of creating their own composite scores based on their personal medical needs and preferences, they make different choices," Mukamel said. "Our findings suggest that patients may benefit if the rating report cards are modified to also include an option for personalized ranking of quality measures."

In the study, patients used an iPad-based application, developed by Mukamel and the research team. The app, called Nursing Home Compare Plus, enabled patients to rank the importance of each performance measure according to their personal medical needs and preferences.

The application combined the performance measures published on the CMS Nursing Home Compare website with the preferences expressed by the patient and family. A list of nursing homes that combined personal preferences and the Nursing Home Compare performance measures was provided to the patient.

Researchers quantified the level of agreement between Nursing Home Compare Plus and the five-star rankings by measuring the level of agreement between the two for all nursing homes in each user's ZIP code-based choice set, on a scale of zero to one, with zero indicating random agreement and one indicating perfect agreement. The levels of agreement ranged from 0.22 -- 0.38, indicating substantial disagreement between the two ranking systems, indicating that the "one size fits all patients" approach to choosing a nursing home might not be best.

"Patients' medical needs and preferences vary," Mukamel said. "Instead of looking for the 'best' provider along all measures, patients should be looking for the best provider for them. Therefore, allowing patients to personalize the composite measures has the potential to improve their choices."

The randomized controlled study involved 146 patients who were discharged from the departments of medicine and surgery at UC Irvine Medical Center to nursing homes, from February 2014 to August 2015.