April 2014 Edition Vol.11, Issue 4

UnitedHealthcare: Measuring Quality Outcomes in Oncology

UnitedHealthcare: Measuring Quality Outcomes in Oncology

By John McCleery & Adrian Barfield

In his opening remarks at a recently held American Journal of Managed Care symposium, Lee Newcomer, MD, MHA, senior vice president for Oncology, Genetics and Women’s Health at UnitedHealthcare, expressed that when he tried to have a discussion with a cancer care provider in 2008 on quality measures, he was told, “I know quality when I see it and cancer is too complex to measure, so just trust me, we’ll tell you when there’s quality.” 

According to Newcomer, that time has since past, and he focused his discussion on his views of quality measures, what that means, and what are the qualities that should be measured. 

“Everything that has been measured in oncology up until the last couple of years was not outcomes, it was process measures, and this is a really rocky transition for me,” he said. 

Newcomer says he splits the world of quality measures into 2 sections – formal, meaning measures that may be used for accreditation or payment purposes; and those measures that are actually used for quality improvement. 

Of the formal measures, Newcomer says he doesn’t place much stake in those since money or accreditation is involved.

On the other hand, quality measures that are used for quality improvement, although at times may be imperfect with embarrassing results they are nonetheless organic and can be corrected to yield results that reflect more accurately the right numbers to implement change in care.

Through the use of different databases and methods, Newcomer says quality measures can be quick and fast, and results can identify gaps of where improvement is needed.

“Accuracy for a quality improvement measure has to use what I call the hand-grenade horseshoe criteria – close is good enough.” 

He explained how UnitedHealthcare built a database of information from questionnaires sent to oncology providers asking for specific information on their breast, colon, lung, and prostate cancer patients. The insurer wanted to know patient histology, stage of diagnosis, some specific gene tests that were relevant, patients’ current status, relapse information, and if the patients were on adjuvant therapy or in remission. 

“And to the physician community’s credit, we got 70% of our facts forms back, on a voluntary basis, that says a lot of good things about our oncologists, and we now have 65,000 patients in that database with that information loaded.”

 

UnitedHealthcare is using the data to spot trends and identify areas for improvement. For instance, they discovered they had the wrong measure when examining for drug cost in each of the cancer categories. Some physicians were not using the Taxotere and Cyclophosphamide (TC) therapy combination. They were using Adriamycin-containing regimens for their high-risk patients, which were creating a cost difference.

Since discovering this two-fold difference in cost, “we were able to correct the measure and now have a new measure that splits into low-risk and high-risk patients.”

A review at hospitalization readmission rates revealed one hospital, in particular, had a very high hospitalization rate, “and they didn’t know it,” Newcomer said.

What UnitedHealthcare’s data showed was that when patients were hospitalized for toxicity, upon discharge, they were told to call the clinic and get a return appointment. However, the average return appointment wait time was about 30 to 40 days out. When these patients got sick again, they landed back in the hospital.

Armed with this kind of information, that hospital was able to change its process and now makes follow-up appointments within 48 hours before the patient leaves the hospital.

“That’s what quality improvement measures are all about. This was an embarrassing number for the hospital, but if they didn’t know it, how could they do anything about it, which is the whole reason for doing, in my mind, quality improvement types of things.” 

Quality improvement measures on the fly can also be dead wrong. In reviewing hospice days and looking at the standard measure that “everybody looks at”, United was told their numbers were completely erroneous by the provider groups. One group in particular told Newcomer that they could prove that United’s numbers were off by a factor of 10, and they did.

“So we dropped this measure, because our United claims database doesn’t pick up hospice very well and as a result we don’t have accuracy, but that is what we learned by having a discussion about a quality improvement measure.”

He ended his discussion by emphasizing that collaborating on quality measures can only be a win-win situation. “Part of collaboration is pooling data, but the most important part, is being transparent, so that there are not any hidden black boxes.”

“Finally, you can measure improvement,” he said, “but you need very instantaneous data to help the groups improve. So when we looked at that group with the high hospitalization rate, they needed a list of all the patients and they needed it fast so they could go back and start reading the charts that didn’t work.”  

Unfortunately, it will take another year to a year and a half of data to know if things are getting better with their readmission rates. “So there is a need for instantaneous i

nformation to manage the problem, but when you’re measuring the change, that takes time, both to accumulate patients and to get the data uploaded. That’s a tough juxtaposition, but its well worth doing,” he said.   

Newcomer doesn’t think accreditation makes much difference. “I don’t think those measures are actually going to get patient care better. The kind of work that we’re talking about, measuring for quality improvement and having an open dialogue about it, is what’s going to get us to where we need to be much faster.” 

If insurers can look at outcomes, and gather data on hundreds, if not thousands of patients with a specific clinical diagnosis, treat it with several different regimens, Newcomer says that the chemotherapy regimens should be compared and not the doctors. “And if we find that a particular chemotherapy regimen has much higher toxicity for the same outcome, much higher cost for the same outcome, you have to pay more to get that drug, or it shouldn’t be covered at all.” 

“And if we can find regimens that may in fact be getting much more superior results, I don’t know if that’s true or not, but, those regimens should now be given preference in our coverage documents.”

So as more data becomes rapidly available for payers to start profiling chemotherapy regimens, Newcomer wonders “if we shouldn’t instead be measuring the regimens and using that information to decide what we do and don’t cover in real world patients.”

 

 

Post a Comment

Your email is never shared. Required fields are marked *