The Overall Hospital Star Rating system from the Centers for Medicare & Medicaid Services (CMS) is meant to give consumers an easy way to pick the best hospitals for them and their loved ones. Yet, due to the methodology that drives the ratings, consumers don't always have a complete and accurate picture of the quality of care being offered by the more than 4,500 hospitals nationwide noted in the CMS system.
Recent updates to address issues to the rating methodology were previewed to the hospitals in June. However, after significant pushback from hospitals and organizations, including Vizient, CMS delayed the scheduled July release to reexamine the proposed changes.
At issue is a change to the methodology that uses a statistical weighting approach for the key quality measures that has proven unreliable. For the July update, complications for total hip and knee procedures were the dominant safety factor comprising the safety score and minimizing any other major complications, including hospital-acquired deep vein thrombosis. The methodology also used a weighting system (latent variable modeling) that actually gave a higher ranking to some hospitals with more complications, compared to hospitals with fewer complications. Vizient has reached out to CMS to assist in refining the methodology.
While CMS is considering the issues recently raised about methodology, here are five recommendations to help the Star Rating earn five star reviews from the consumers it's trying to serve.
Recommendation 1: Smarter hospital groupings
To provide a simple, baseline quality metric for hospitals and consumers, CMS currently lumps together community hospitals, large health systems, specialty hospitals and teaching/academic hospitals on many key metrics, like mortality.
CMS should instead group hospitals by subtype, comparing hospitals treating advanced cancers, neurosurgical cases and cardiology cases to other hospitals specializing in the same areas. This stratification would help make comparisons based on the Star Rating more accurate for providers and more relevant to consumers.
Recommendation 2: Group primary procedures by specialty
Because the current Star Rating does not differentiate between various specialties, consumers cannot utilize the rating for specific needs. For example, a consumer facing a knee replacement surgery wants the best hospital for knee replacement, not the best heart center.
Instead, similar hospitals should be grouped together and offer more details on specialty procedures.
Recommendation 3: Use better data, faster
Consumers need near real-time performance data, yet based on research at Vizient, about 64 percent of the overall Star Rating score is measured with data that is at least two years old.
Another major limitation is that data is derived from Medicare patients who represent only a subset of the hospital’s patients.
Recommendation 4: Create data transparency
In addition to speed, having more transparent and useable data in the Star Rating is key.
It's unclear from the current rating system what major areas hospitals should focus on to improve care, as CMS is regularly changing the values of some of its measures.
The ratings should be more than a report card. They should also be a tool to guide performance improvement. As it stands, it’s impossible to draw meaningful and accurate conclusions about the quality of care for hospitals, physicians and the patients they serve.
Recommendation 5: Revise the weighting of measures of quality
The importance CMS currently places on the four key criteria used to calculate the Star Rating – patient safety, patient experience, readmission and mortality risk – should be reconsidered.
These four criteria each are worth 20 percent of the total score, with other measures receiving less attention. While these are the right buckets, they do not tell a complete and accurate story about quality. The weighting should be adjusted to give greater merit to criteria like effectiveness of care and timeliness of care.
Also, a better system needs to be put in place for hospitals that do not have enough volume to report on a specific metric. Currently in these cases, the metric is not counted and the remaining metrics are weighted higher. For example, specialty hospitals only report half of the current parameters, but they’re still included in the Star Rating with comprehensive hospitals. This results in hospitals being compared against each other but not all of the same metrics or relative weights of the metrics are being used. It may appear that a hospital with five stars is better than the four-star hospital, but they were compared differently. A better approach would be to cluster hospitals, as noted earlier, to create a more fair and useful comparison.
The Star Rating system is a great idea. With the tweaks outlined here, it can become a positive feedback loop between hospitals, physicians and patients to drive performance improvement.
Originally published in The Hill 6/8/18.
