From 3M Health Information Systems
The population health elephant
Last month, I blogged about long held tenants of medical necessity and how those standards impact today’s creative thinking on actions to improve health outcomes.
I received interesting questions from colleagues. I took from our discussions this reaffirmation: Population health is defined differently depending on the professional hat of the wearer—sort of like the story of the blind men and the elephant. According to Wikipedia, “The story of the blind men and an elephant originated in the Indian subcontinent from where it has widely diffused. It is a story of a group of blind men (or men in the dark) who touch an elephant to learn what it is like. Each one feels a different part, but only one part, such as the side or the tusk. They then compare notes and learn that they are in complete disagreement.”
So, what are we hearing from different parts of the industry as healthcare professionals wrestle with the population health “elephant”? Here’s my take:
As I was pondering my colleague’s responses, I saw an interesting article in HealthData Management “HIT Think: Why providers need a consistent definition for population health management.” The author attended HIMSS 2017 and did an unscientific ad lib study to try and determine what attendees and vendors believe population health management to be. He asked individuals at HIMSS to define population health management:
- 83 percent replied chronic disease management
- 82 percent said general preventive care
- 70 percent said it is workplace wellness programs
The article concluded: “There is more confusion surrounding population health management than we as an industry would like to accept.” The author believes healthcare IT must “help design a strategic vision of a healthcare ecosystem that delivers quality care to all patient populations.” While I respect the IT perspective, in my opinion the operative word here is help. Ask any frustrated physician if they believe enough doctors were involved in the design of electronic health records—you can likely predict their response.
The next article to catch my eye was published by The Center for Health Care Strategies on March 16, 2017 titled “Integrating Community Health Workers (CHW) into Care Teams: Lessons from the Field.” In this instance, trained non-clinical community health workers are assigned to “patients with complex medical and social needs—issues such as housing, transportation, food insecurity, addiction and social support.”
The CHWs “build relationships with patients based on mutual understanding and respect and help them navigate complex health and social service systems to access necessary care and services.” Providers are beginning to recognize the unique skills of the CHW because they have a real knowledge of the cultural and community norms of the population they work with. I believe identifying patients that could benefit from intervention and support by a CHW would be improved if clinical providers reported ICD-10 codes that represent social determinants—the Z55-Z65 range. Currently, reporting these codes does not impact reimbursement—which could be a reason they are rarely found on claims. This is valuable data to programs such as this. Perhaps payers should add extra reimbursement for reporting these codes since someone is performing more work gathering the data?
Next I saw the OIG Advisory Opinion No, 17-01. A critical access hospital (CAH) petitioned OIG to review and reply with an advisory “regarding a hospital system’s proposal to provider free or reduced-cost lodging and meals to certain financially needy patients.” The provider requested that the OIG review their plan and provide guidance on whether (or not) the proposed support could be sanctioned under the Federal Anti-kickback statute. The hospital is a level I trauma center and serves patients including those in rural areas who have both transportation and financial challenges. The hospital contracted with a nearby local hotel in order to provide reduced cost (or no cost based on need) overnight lodging for patients undergoing treatment one night prior and up to two nights post treatment. In addition, the hospital would provide free or low cost meals in the cafeteria not to exceed a value of $15 per day.
OIG reviewed the laws and stated:
“We recognize that there are socioeconomic, educational, geographic, mobility or other barriers that could prevent patients form getting necessary care (including preventative care) or from following through with a treatment plan. Our interpretation of items or services that promote access to care encompasses giving patients the tools they need to remove those barriers. We believe that the lodging and meals would promote access to care, meaning they would improve a beneficiary’s ability to obtain items and services payable by Medicare or Medicaid.”
OIG listed criteria including specific patient selection required to participate in this program. The program would not be considered susceptible to prosecution under anti-kickback laws. The hospital staff would determine which patients meet eligibility criteria for this service. Active documentation of social determinants in the EHR and ICD-10 Z codes on claims would help identify those patients who may quality for services of this type.
Evolving healthcare data concepts
I was talking with some smart, creative thinkers at 3M. The discussion centered on how clinical data is actually more than what is reported on a claim.
Retrospective data gathered from claims is good, but what is equally important is access to data that represents the patient’s condition today, the moment they are receiving care. We’ve all heard the phrase “big data,” but could data be more useful if it was not dependent on what is reported in a CPT or ICD-10 code on a claim?
I have done my share of audits—both concurrent and retrospective—and I am often amazed by the amount of information contained in the medical record compared to what is reported on a claim. There is often much more there that impacts care.
(A quick example; I once reviewed the medical records for a man who was diagnosed with a traumatic hip fracture and admitted for a short stay. But when I read the medical record there was so much more to his story than a fracture. He was 65 years old, had fallen off a duck-blind in the woods that he was preparing for hunting season. He crawled three plus miles to the highway to flag down a motorist to call an ambulance since he could not find his cell phone in the brush under the blind. And he could not use his truck since his keys joined his cell phone in the brush. So, not only did he have a fracture, he was dehydrated, suffered exposure (this accident happened in Texas), and was bitten by a tick while crawling to the highway!)
To be fair, historically CPT and ICD codes on claims are there for reimbursement purposes. And once those needs are met with the correct codes, the coder moves on to the next case. But would it better to be able to access and collect all the additional patient information that could be vitally important to support population health beyond what is on the claim? Do we rely too much on data used for reimbursement purposes?
Could the additional unstructured data present in the medical record be useful to the disease registries? Cardiac, trauma, tumor, chronic diseases etc? Wouldn’t access to this data be useful for reporting electronic Quality Measures (eCQMs) for ambulatory and inpatient hospital quality reporting programs? And with MACRA (Medicare Access and CHIP Reauthorization Act) looming, to clinical providers as well? Successfully reporting MIPS (Merit-based Incentive Payment System) in the MACRA environment can make the difference between a 4 percent yearly loss for an eligible provider and up to a 22 percent bonus for excellent work.
I agree with my colleagues—clinical data really is more than what is reported on the claim. Being able to gather and harness this data is an exciting concept. And perhaps it is a far better idea to view the entire healthcare encounter since the unstructured data often reveals a larger story. I think approaching data in this manner will also help providers improve their documentation by asking the often missed social determent questions that can dramatically impact utilization. Sort of like getting to know the whole elephant.
Barbara Aubry is a regulatory analyst for 3M Health Information Systems.