In an insurance world where financial underpinnings are tied to risk, the industry tries to create a science of risk assessment (in the generic sense) so they can price against it with confidence. This has worked historically for life insurance, for example, and in the indemnity health insurance markets, as well. But in today’s health care world, there are points of confusion about risk adjustment coding used to calibrate premiums, as in Medicare Advantage, or to redistribute them, as in the health insurance exchanges.
When I mention the idea of “areas of grey” in risk adjustment coding, this is not to say that the coders I met are terribly confused. The risk adjustment coders have their heads down and are working energetically towards the Holy Grail of comprehensiveness and accuracy. Indeed, they are all quite certain that their approaches are reliable and accurate. The problem is that there are areas of disagreement when you get these coders all together in a room, as we notice in our workshops and conferences. And where there is uncertainty, there is likely to be inconsistency. Even where there is relative certainty in any one coding shop, there is still inconsistency, which is known as the problem of inter-rater reliability. So much for the fine science of insurance certitude.
We came across a large degree of coder agreement across the broad scope of items. But on certain points, we had a lot of disagreement. RISE is working collaboratively to isolate the causes of disagreement so we can narrow down on a set of principles that iron out the differences.
One source of this differing of opinions stems from the statutes and regulations that come into play when submitting a claim vs. retrospective supplemental coding. When filing a claim, a provider must ensure that all diagnoses submitted are properly documented and have been monitored, evaluated, assessed and treated (MEAT). However, a health plan is afforded the opportunity to retrospectively review medical records for diagnoses that need to be added or deleted in the list of active codes.
There is a different opinion regarding use of “MEAT” by many others. As voiced by Brian Boyce of ionHealthcare®, “Fee-For-Service (FFS) works well with using MEAT, because one cannot upcode a level of service (such as a higher E&M code) using a diagnosis as justification unless there was MEAT supporting the diagnosis. FFS is a retrospective payment model, where a payment or service is supported by diagnosis for payment AFTER the service was performed. Boyce reminds coders that Risk Adjustment is a brand new payment methodology that is trying to move away from FFS, and supports a prospective payment model based on what may be predicted as needed in the FUTURE based on known/ confirmed diagnoses in the year prior. Risk adjustment participant guides encourage coding of all current diagnoses in a face-to-face encounter by an approved physician, and avoiding the reporting of diagnoses which are resolved or past medical history (PMH) in nature”.
While some confuse the use of MEAT and the other acronym of TAMPER as being equivalents of one another, TAMPER was really created by Boyce to assist coders when faced with diagnoses listed as PMH or Medical History, as a way to help determine if those diagnoses should or should not be submitted. The PMH review called TAMPERTM (created and trademarked by Brian Boyce, ionHealthcare®), stands for Treatment, Assessment, Monitor/Medicate, Plan, Evaluate, or Referral. This school of thought does not rely on billing regulations, but comes from an HCC coding perspective that refers to the CMS guidance for valid risk adjustment coding. It is grounded in the principle that there must be evidence supplied to substantiate the most comprehensive picture of the patient’s true and current health conditions that are accepted as predictive of future health care cost liabilities of the MAO plan.
The TAMPERTM thought process requires that the diagnoses technically documented as historical in nature need to be digested to reveal what was documented and taken into consideration by the physician in face-to-face encounters with the patient during the reporting period, not just an isolated encounter at a point in time. This process involves dissecting the chart for admissible evidence and knitting them together to tell that comprehensive and accurate story. In our HCC Coding Accuracy for Coders workshop, we walk through the elements of the chart to identify which are usable and under what circumstances as substantiation for particular diagnostic scenarios.
This approach promoted by Brian Boyce and others is oriented to retrospective chart reviews, taking into account the whole scope of on-going and future care for a patient. This is markedly different from the Fee for Service billing world in physician offices where there are billing rules constraining them as to what is permissible for a specific encounter in the past. Note that in a RADV audit under CMS, the task is to produce any face-to-face encounter within the year being audited that supports the risk adjustment value, while HHS and Medicaid audits reflect back to a particular date of service encounter for accuracy. However, even apart from that, there seems to be some controversy over different examples discussed using the TAMPERTM thought process, depending on whom you ask. We think that we can help narrow the differences by working closely with a risk adjustment advisory group (see below).
By contrast, Todd and Kameron Gifford of ERM Consulting Inc. advocate a different approach. ERM utilizes an evidence based prospective process to create a complete an accurate medical record at the time services are rendered. This “Rapid Practice Innovation” approach to coding and documentation accuracy involves educating the frontline and creating “best practices” that align with current workflow.
“Rapid Practice Innovation” involves a systematic “up-training” of all the staff working in the clinics, from check-in to check-out, from back office to billing office. It includes a comprehensive annual wellness exam in addition to any regular visits for chronic care management. Some practices’ billing systems limit the number of submittable diagnoses to only four, which is too restrictive for patients with more complex conditions to be fully documented via claims in a single visit. Over all, managing this process to optimize diagnostic code capture is something that requires training, tools and perseverance. Todd and Kameron Gifford will be presenting a case study of just such an effort in a success story played out in Corpus Christi, Texas, at the RISE California Summit on July 21.
RISE has formed an initial risk adjustment advisory faculty to help us to do two things. First, beyond the level of medical coding that is captured in the AAPC CPC certification, for example, we seek to determine what additional course material HCC coders need to master. And at what level of difficulty? AAPC has proposed to partially answer that question with their new certification for risk adjustment coders, the CRC© (Certified Risk Adjustment Coder). But RISE will be taking this further, including the financial dimensions of risk adjustment not part of the AAPC certificate. Second, RISE wants to tease through the “grey zone” of coding that seems to divide the professional community. We will be pushing forward on that agenda so we can integrate it into our workshops and educational programs.
We are certainly interested in hearing from coders and their management around these issues. Additional faculty is welcome, also, as we seek to put together more workshops to make the material more accessible regionally to more local audiences.