Capture the Complete Clinical Picture With Precision

Capture the Complete Clinical Picture With Precision

Clinical validation could upend your current risk adjustment policies.

The rise of risk adjustment-based managed care programs in the past 15 years within Medicare, Medicaid, and Affordable Care Act (ACA) plans has led to remarkably good news: There has been a significant improvement in the accuracy of the diagnosis codes abstracted from medical documentation among many providers. But many more providers still have a lot to learn. As reimbursement is increasingly tied to multiple comorbidities, providers and coders must learn to capture more completely the clinical picture of their patients and do so with specificity.

Not All Growth Is Good

Most of us work in a business setting where a growth mindset prevails. Practices might think, “Last year, the practice’s revenue grew by 12 percent. This year’s budget requires 15 percent growth.” But since computerization has already optimized so much clinical work, there are fewer efficiencies to be found. To avoid financial stagnation, what are provider groups doing?

Managed care programs, payers, and physician practices are getting creative. And that’s not necessarily a good thing. In fact, there’s a trend we are seeing in risk adjustment today that could lead many practices down a path laden with liabilities for the practice, the providers, and the patients. The trend can be linked back to an ICD-10-CM guideline added Oct. 1, 2016:

Body I.B.19. Code assignment and Clinical Criteria

The assignment of a diagnosis code is based on the provider’s diagnostic statement that the condition exists. The provider’s statement that the patient has a particular condition is sufficient. Code assignment is not based on clinical criteria used by the provider to establish the diagnosis.

For 2023, the following was appended to this guideline: “If there is conflicting medical record documentation, query the provider.”

The purpose of this guidance was to differentiate between the skillset of a provider and a coder. It is not within a coder’s purview to question the clinical knowledge of the provider or to change a diagnosis based on evidence in a note. Should there be a concern, the coder may query the provider, but the provider alone has the knowledge and authority to diagnose a patient. This guidance is intended to remind coders to “stay in their own lane.” We see this guidance providing ammunition to clinical documentation improvement (CDI) professionals, who help to improve documentation standards, to also push for incremental changes to documentation so that risk-adjusting diagnoses can be captured instead of non-paying diagnoses. On many occasions we have heard coding educators and physician influencers saying, “If you would just add these words to the record, you will get a risk score bump that leads to profits for your practice.” This may not be a compliant process and may influence cherry-picking of selective high-paying diagnoses.

CMS’ Current “Hit List”

Until very recently, audits from risk adjustment data validation professionals (RADVs), the Office of Inspector General (OIG), and the Department of Justice (DOJ) were focused on code validation, but the Centers for Medicare & Medicaid Services (CMS) has broadened the scope of these audits. Primarily, CMS has developed a significant “hit list” of targeted diagnoses in which errors are common. Diagnoses on the list include:

  • Aortic atherosclerosis from an incidental radiological finding;
  • Chronic respiratory failure from the presence of oxygen
  • Myocardial infarction from an old EKG;
  • Acute myocardial infarction or acute stroke when a hospitalization does not occur;
  • Acute embolism or acute deep vein thrombosis (DVT) in lieu of prophylaxis of DVT;
  • Opioid dependence noted with prescriptive narcotics for pain control;
  • Vascular claudication in the presence of medication for neurogenic claudication;
  • Major depressive disorder absent therapy, medication, or appropriate history; and
  • Coding active cancer in cases in which the cancer has been excised and treatment has been completed.

This list is a mixed bag of errors in chronicity (e.g., coding acute or active when illness is no longer acute or active) or clinical misdirection from CDI professionals ambitiously working to pump up risk adjustment factor (RAF) scores. Keep in mind that even histories and incidental findings must be pertinent to the current encounter or should affect patient care in order to be reportable diagnoses, according to Guideline IV.J., which reads:

IV.J. Code all documented conditions that coexist

Code all documented conditions that coexist at the time of the encounter/visit and that require or affect patient care, treatment or management.

In several current lawsuits, CMS is proposing clinical validation reversals based on the medical record. For example, Kaiser Permanente was accused by the DOJ of distributing standardized queries in which patients, identified through a data search as having a prescription for oxygen and diagnosed with hypoxia (no RAF), should be assigned an additional code: acute respiratory failure (with RAF). DOJ notes in the lawsuit that not everyone with a prescription for supplemental oxygen has respiratory failure.

Levels of Coding and Auditing

Keep in mind that the continuum of clinical care, documentation, coding, billing, and payment has many complex levels, and most of us fall into only one or two. Many risk adjustment programs deploy additional layers of analytics that involve looking at past, present, and future diagnoses for patients as they perform predictive modeling on what diagnoses are missing or yet undiagnosed. Often times, these analytics exercises are driven by financial — not clinical — needs and may lead to coding incentive programs that may perpetuate false upcoding.

For the purposes of this discussion, we can divide our tasks into the three we may have control or influence over: coding, CDI, and clinical validation.

  1. Coding is straightforward, with little physician interaction except occasional queries for specificity. Coders use the entire medical record and follow the guidelines to abstract codes based on support in the record.
  2. Clinical documentation improvement addresses shortfalls in documentation and may include education and queries that further specify the patient’s diagnosis or addresses information in non-codable parts of the chart (e.g., problem list, past medical history, medications, and durable medical equipment) that may be pertinent to the encounter. The goal of CDI is to improve the overall comorbidity capture for the patient so that the entire encounter is better captured. CDI requires more pathophysiology and anatomical understanding than does straightforward coding, and the CDI role is often given to nurses or others with clinical backgrounds, although seasoned coders can handle this role.
  3. Clinical validation is the most complex level of a coding audit and addresses the medical validity of the diagnosis assigned by the provider: looking at the labs, the medications, the medical history, and the physical exam to review the accuracy of the diagnosis. In many of the recent OIG findings, major depressive disorder was found to be simple or episodic depression, failing clinical validation for the diagnosis of major depressive disorder.

For those who work in coding or CDI, it’s important to focus on the guidelines and CMS guidance for risk adjustment compliance and to resist any institutional campaign to increase RAF scores that does not include an increase in the clinical care of the patient. Better preventive care and management of existing comorbidities was the original goal of risk adjustment, specifically, “to advance health equity, drive comprehensive person-centered care, and promote affordability and the sustainability of the Medicare program,” according to CMS. With more than 40 percent of Americans choosing Medicare Advantage plans over traditional Medicare, it’s no wonder that RAF scores have taken center stage for so many organizations focused on profits.

What Other Targets Might Be On the Horizon?

In our work, we see many high-risk score diagnoses which, upon inspection, seem to have dubious clinical validity, but which are reported with frequency by some payers. If your organization has coding policies for capturing high-risk diagnoses, stay alert. If considering similar policies, ensure your leadership weighs the pros and cons of accepting the associated compliance risks. What may seem acceptable risk for the deep pockets of some of the large insurers may not be safe for a small- to medium-sized Medicare Advantage plan, managed care organization, or provider group. Engage finance, medical leadership, CDI, coding management, and the entire coding and quality assurance team in discussions about coding, clinical documentation improvement, and clinical policies that affect risk scores.

Remain circumspect as you create policies and consider the following “soft targets” that could be problematic in an audit:

Homocysteinemia/homocystinuria

Homocystinuria is a rare, inherited genetic disorder tested for at birth, similar to phenylketonuria (PKU) testing. Recently, elevated homocysteine in blood chemistry has been found to be a predictor of heart disease, and some providers have been documenting homocysteinemia in their Medicare patients to report the abnormality. As a result, homocystinuria, with a disease incidence of 1 in 100,000, has been reported with an incidence of 1 in 4,000 in Medicare fee-for-service claims.

The ICD-10-CM index entry until 2022 directed readers to E72.11 Homocystinuria from the entry “homocysteinemia, homocystinuria.” Therefore, this code has been used not only for an inborn defect of metabolism, but also for the diagnostic marker linked to heart disease. Because E72.11 maps to hierarchical condition category (HCC) 23 Other Significant Endocrine and Metabolic Disorders, the classification has allowed a risk-adjusting diagnosis to be used to report a symptom that could lead to heart disease.

The index entry Homocysteinemia classified in 2022 to a new code, R79.83 Abnormal findings of blood amino-acid level. While E72.11 as a genetic disorder continues to risk adjust, R79.83 reports a symptom and carries no risk value. Ensure your CDI professionals are not querying for E72.11 when the appropriate code is R79.83.

Diabetes with unspecified, other specified complication

E11.8 Type 2 diabetes mellitus with unspecified complications is an example of catch-22 coding in risk adjustment. CMS expects the medical record to support any diagnosis that is submitted for risk adjustment payment, including E11.8. What’s the catch? If there is support in the record for a complication of diabetes, then the actual complication would be specified, and a specific code would be assigned instead of E11.8. A provider simply parroting the code description for E11.8 — other specified complication — does not lead a coder to E11.8, but to default code E11.9 Type 2 diabetes mellitus without complications, because for risk adjustment, the complicating diagnosis must be supported in documentation.

E11.69 Type 2 diabetes mellitus with other specified complication maps to diabetic osteomyelitis in the Index. There may be other appropriate complications, but most would fall under oral, integumentary, circulatory, or other organ systems already addressed in the classification.

Importantly, the etiology of type 2 diabetes should be considered. Dysmetabolic syndrome X is considered a predictor for diabetes and is defined as a cluster of insulin resistance, obesity, hypertension, and dyslipidemia. Clearly, risk factors for diabetes should NOT be reported as complications of diabetes. It would therefore be inappropriate, except in rare instances, to state that the patient’s dyslipidemia, hypertension, or obesity was the product of diabetes. Documented “obesity due to diabetes,” “diabetic hyperlipidemia,” or “hypertension due to diabetes” would likely fail a clinical validation audit.

Major depressive disorder

Though this disorder is an OIG and DOJ target already, it bears repeating: Major depressive disorder is a serious illness. It is larger than the sadness created by the death of a family member or loss of a job. It elicits a significant shift in the brain’s chemistry that requires pharmaceutical intervention and management by behavioral health professionals.

An error of omission in ICD-10-CM, lack of a unique code for unspecified depression, has been corrected this year with the implementation of F32.A Depression, unspecified. In previous years, unspecified depression was indexed to F32.9 Major depressive disorder, single episode, unspecified. But even with the Index error, the OIG is still disallowing the use of F32.9 in its actions because there is not clinical validity. This is another code that CDI professionals involved in RAF recapture should be very cautious about.

Secondary hypercoagulable state

Some organizations have created policies for patients who have been diagnosed with atrial fibrillation, stating that any patient on anticoagulants for atrial fibrillation have a comorbidity of secondary hypercoagulable state, because atrial fibrillation can sometimes cause patients to develop clots in the atria. There is a potential that these clots could break loose and travel to the brain, causing a stroke. This advice encourages plans and providers to grab an additional HCC for every patient with atrial fibrillation: HCC 96 Specified Heart Arrhythmias, plus HCC 48 Coagulation Defects and Other Specified Hematological Disorders for the secondary hypercoagulable state.

This blanket application of secondary hypercoagulable state with atrial fibrillation diagnosis is lucrative, but not necessarily accurate. One reason that atrial fibrillation risk adjusts is because of the risk associated with formation of blood clots that increases the risk of stroke. But many patients do not go on to develop any clots. Many patients never go on anticoagulants.

Morbid obesity with a BMI under 40

When the provider documents morbid obesity with a body mass index (BMI) of less than 40, there should be ample evidence of the clinical evaluation that went into this diagnosis. What conditions exacerbate the obesity, and in what way? Details matter if a case is to be made that the comorbidities are putting the obese patient’s life at risk. It is not acceptable for a coder to simply look for a comorbidity that could be applied; the provider must link the comorbidity and provide a rationale as to why the diagnosis should be “upgraded” from overweight or obesity to morbid obesity.

Your Best Route to Compliance

Compliance in the current risk adjustment environment is quite a bit easier than you may think, if you just rewind to its original intent: to drive comprehensive, person-centered care. It’s all about ensuring that the patient’s chronic conditions are well managed, that complications are prevented, and that the Medicare population can maintain its health and well-being for as long as possible.

Return to that idea during your workday and your patients and practice will thrive. Ask yourself these questions with each chart:

  • Did the diagnoses assigned to the patient reflect the medical decision making and care provided (e.g., examination for intact skin on the contralateral foot of a patient with a leg amputation)?
  • Is the patient informed and in agreement with the diagnoses in their chart (e.g., patient is celebrating being cancer-free for five years, but she has a diagnosis of breast cancer because she continues tamoxifen for prophylactic care)?
  • Do the diagnoses and services reported support good patient care or are they designed to capture greater RAF value (e.g., reporting incidental finding of aortic atherosclerosis in a patient who already takes a statin)?

It’s time to focus on patient outcomes rather than financial outcomes. But if we properly adhere to compliance rules, we can win with both.


Authors:

Sheri Poe Bernard, CPC, COC, CDEO, CRC, CPC-I, CCS-P, and Khushwinder Singh, MD, MHA, CPC, CPCO, CPMA, CRC, CDIP, lead the risk adjustment team at Granite GRC Consulting. They improve reimbursement, patient care, and patient experiences by engaging with Medicare Advantage plans and provider groups to analyze claims data, review risk adjustment processes, and audit medical documentation to identify gaps and address opportunities for improvement in the accuracy and completeness of clinical data. Contact SPB@GraniteGRCConsulting.com or KSS@GraniteGRCConsulting.com.

5 Responses to “Capture the Complete Clinical Picture With Precision”

  1. Robin Sewell says:

    Great read!

  2. Wanda Gordon says:

    Medical coders do not have a high level of education. Yet you seem to want the doctor to constantly monitor the mc and patients. If there is no code for a particular diagnoses then doctor and staff have to spend enormous time researching because of lack of help from any of your organizations. No one ever contacts the patient to ask what has been diagnosed in your case and do you agree with your doctor.

  3. Mary Hanson says:

    Awesome reminder of the goal of RAF scores: Excellent patient care, not exaggerated RAF capture!! Please keep these valuable articles coming, we need the guidance on AMBULATORY HCC capture. Thank you both for your commitment to excellent–and ACCURATE–diagnosis coding.

  4. Joslin Baker OGuinn says:

    Well-written and informative article. I wish my doctors would ask me if I agree with their dx codes assignments for me. ha

  5. Stephen Marquess, CRC, CDEO says:

    This is a great article. I work for an FQHC where profit is not necessarily the priority. Thus, coding is by the book and we don’t go searching, but code what is there (at least that is the goal). Before I came to this FQHC, I was with a private group that had purchased our small practice that did echocardiograms on all patient’s that they could just to capture pulmonary hypertension, wanted to diagnose every diabetic with E11.8 and then list all the complications, and we did lots of bloodwork for the Homocystinuria code. I lasted less than a month, after 20 years with the practice. I ahve thirty years of clinical background, am a CRC and CDEO, yet any argument I made came to naught. Hopefully, many will read this and pay attention.

Leave a Reply

Your email address will not be published.