Guest Post by Dr. Helen Gelly: How “Big Data” Is Changing How We Practice Medicine

Our last guest post by the esteemed Dr. Helen Gelly, CEO of HyperbaRXS in Atlanta, GA, was so popular that I just had to have her back in this space! As always, her insights are both incisive and much-needed. Enjoy! –CF

It has been said that “you can’t manage what you can’t measure” and the utilization of big data to predict outcomes and enhance performance has impacted all facets of our lives. Retail big data analytics has influenced how we shop.  Now, the electronic health record as well as claims data has become the source material for Medicare’s big data analytics platform and will manage how we practice medicine. Much like Amazon, Medicare spends a lot of time analyzing our “habits” and for the most part, we unaware of the ways they utilize our data. The amount of data that is available on any one practitioner is mind-boggling. Much of it is publicly available, which has turned mining Medicare data into a new career path. Physicians need to be aware of what data is out there, and we need to check up on ourselves to make sure that it is correct. So let us see what is out there.

Open Payments tracks information about payments made to physicians and teaching hospitals by drug and device companies.  Travel, research, gifts, speaking fees and meals paid to doctors are tracked and reported. Nothing goes unreported — even that doughnut you had when you thought no one was looking.

Physician Compare does just what the name implies – it compares your quality measure performance to other physicians across the country, and provides data on hospital affiliations, amongst many other things. Physician Compare has made a deal with YELP so that the quality scores from MIPS participation will be translated to YELP star ratings. Make sure that you select quality measures that you are going to ace, since your star rating will depend on it. Think carefully about what quality measures you report in 2018.

Medicare Provider Utilization and Payment Data provides open access data on services and procedures provide to traditional Medicare beneficiaries by physician and other healthcare professionals. Detailed data includes a list of all the CPT codes ( > 10 claims) you have submitted in a calendar year.  Included is your gender, location, site of service, how many services you provided, how much you charged, and how much you were paid. Why is this important? Why should you check? Because if you are reporting a site of service “O” or office, and you are actually in a hospital-owned outpatient department “F”, you are putting you and your HOPD at risk. Let me say this another way. If you assert to Medicare that the place of service where you render care is your private office (“O”) but it’s actually the hospital outpatient department (‘F”), CMS can claw back all of the facility fees it paid to the hospital. Furthermore, since the physician payment rate for those services is higher for a doctor in his/her private office, you have been overpaid. (CMS reduces the payment rate to doctors in hospital outpatient clinics because they are not responsible for the cost of overhead or staffing). You have thus overbilled Medicare (also called fraud) and owe this money back with penalties. This information is public, so you should check it for accuracy.

Moving onto the Provider Summary Table. This details much of the same information as the Charge data, but expands the categories that are tracked.  Do you ever wonder about the average age of your patients, how many women versus men, what racial category they fall within, and what their comorbid conditions are?   According to my data, over 31 % of my patients have heart failure, 44% have chronic kidney disease, 47% have diabetes, and 75 % have hypertension. But none have osteoporosis, schizophrenia or stroke.  Is that possible??? Of course not, but what is reported is what I reported. I record the diagnoses that impact my treatment plan for non-healing ulcers. But, lately, I’ve been rethinking the way I record co-morbid conditions. This brings us to what may be perhaps the most under-appreciated “score” a doctor can have. If you think your credit score is important, that can’t hold a candle to the importance of your “Average HCC Risk Score of Beneficiaries.” To no one’s surprise, I am not a 10; not even close.

What are Hierarchical Condition Categories (HCCs)?

Established in 2004 (yes you read that correctly) they were originally used as a risk-based payment methodology for Medicare Advantage (MA) plans (Medicare HMOs run by private payers).  The higher the HCC score, the more complex the patient and thus the more dollars allotted to the MA plan for their care. This makes sense.  MA plans got more money for taking care of complex patients to reduce the likelihood that MA plans would cherry-pick the healthy Medicare patients to enroll in their HMOs. However, Medicare never passes up an opportunity to leverage data for other healthcare initiatives. We all need to understand what HCCs are and their impact on our “value.”

There are two different types of HCCs:

  1. The CMS model used by CMS for risk adjustment in the Medicare Advantage plans (elderly). There are two different components, CMS-HCC and CMS-RxHCC that addresses Medicare Part D.
  2. The HSS-HCC model maintained by the Department of Health and Human Services that addresses commercial payers and therefore covers all ages.

These HCC scores predict the future cost of care, so this year’s HCC scores predict what the costs will be to care for patients next year. These data are important for physicians who participate in ACOs and in shared savings contracts with MA plans. However, for physicians subject to MIPS, your HCC score impacts your value based payment modifier or “cost” component. It is likely that either your QRUR or your HCC scores will be used in the calculation of “cost.” Your HCC score will serve as the basis for determining how complex your patients are and what their expect cost of care will be (Risk Adjustment Factor or RAF).   In a recent CMS sponsored webinar, CMS mentioned that they were planning on giving extra points (up to 5) in the calculation of your MIPS score for higher HCC scores. It seems that they are leaning toward using HCC scores in these value calculations.

The average HCC score is 1. The average Medicare beneficiary cost is estimated to be a little less than $10,000. The proposed way that this may impact hyperbaric and wound care providers can be illustrated as follows:

The average hyperbaric patient will have 40 HBOT treatments at $113/ treatment (for the physician component). Total cost of HBOT supervision will be about $4,520. If your HCC score is 1, then you have consumed about 45% of the predicted cost of that patient for the entire year! If you have an HCC score of 3, then the HBOT you provided only accounted for 15% to the predicted cost. That’s the same patient, and the same protocol, but simply by raising the HCC score to 3, I will be perceived to be more “frugal” in contributing to the cost of care. Therefore, I am seen as providing better “value.”

The point is, Medicare EXPECTS that sick people will require more resources to be spent on them. Complicated patients cost more to care for. They require advanced modalities and even under the best of circumstances, they take a long time to get better. Since our cost of care is high, we  need to accurately reflect the complexity of our patients, otherwise Medicare will begin to claw money back using the logic that we are spending resources wastefully. We need to become more conversant with risk adjustment and think carefully how to accurately document the medical conditions, complications, and complexity that reflect our patients.

The next installment will address the HCC list, pitfalls in documentation, and the way forward.