Community-Acquired Pneumonia is Being Over-Diagnosed

By Erica Remer, MD, FACEP, CCDS, ACPA-C

I continue to find multiple instances of misdiagnosis and miscoding of respiratory conditions.

A common scenario was the emergency physician interpreting outside or portable chest X-rays as demonstrating a possible right lower lobe infiltrate indicative of pneumonia, but no additional confirmatory or exclusionary test was performed. Or the emergency physician was the only one who mentioned pneumonia until the clinical documentation integrity specialist elicited subsequent documentation with a query.

Another situation was the patient originally being treated empirically as having pneumonia; it was ultimately not thought to be present, but was never eliminated from the problem list. In retrospect, the patient had some alternate respiratory diagnosis, such as acute exacerbation of chronic obstructive pulmonary disease (COPD) or heart failure, but the pneumonia was never declared ruled out in the record.

The typical chief complaint was shortness of breath, which can herald pneumonia…or an exacerbation of COPD or heart failure. What dissuaded me from thinking these patients had pneumonia?

If they didn’t have a fever, a cough (especially productive), or increased respiratory rate, pneumonia seemed less likely to me. If an outside facility’s chest X-ray was not repeated and there was no official radiology overread, if a CT was done and the lung fields were commented on as being clear, or if the pulmonologist did not draw a conclusion of pneumonia on their assessment, I felt the clinical validity was in question.

A recent article published in the Journal of the American Medical Association (JAMA) Internal Medicine validated my impression and sought to characterize the inappropriate diagnosis of pneumonia in hospitalized patients. Authors from a collaborative of hospitals in Michigan asserted that the diagnosis of community-acquired pneumonia (CAP) was inappropriate if the patients had fewer than two signs or symptoms of CAP or negative chest imaging. Signs or symptoms include new or increased cough, change in sputum production, new or increased dyspnea, hypoxemia, abnormal lung sounds, increased respiratory rate, fever or hypothermia, and abnormal white blood cell count. This metric to identify inappropriate diagnosis of CAP has been endorsed by the National Quality Forum.

A total of 17,290 patients were included in the study, and 12 percent were felt to have been inappropriately labeled as having CAP. A total of 88 percent received a full course of antibiotics. Patients who were inappropriately diagnosed with CAP were more likely to be older, suffer from dementia, or present with altered mental status.

Lower respiratory tract infection, including pneumonia, is the most common infectious cause of hospitalization in the United States. It is a common source of sepsis. It is reasonable to empirically treat it if pneumonia is initially a serious concern, but there are perils in not ruling out the condition if it is not actually present. A delay in ruling it out promptly can prevent recognition and appropriate treatment of another condition, which in many cases is really causing the signs and symptoms. Unnecessary antibiotic usage can result in adverse effects, allergic reactions, and development of antibiotic resistance.

The most common presentation was dyspnea and/or cough. Inappropriate diagnosis of CAP was also associated with being on public insurance, having decreased mobility on admission, or having had an inpatient hospitalization within the previous 90 days, in addition to the variables noted earlier. These patients were more often discharged to a skilled nursing facility (SNF).

The study speculated on why physicians might be prone to inaccurately diagnosing CAP. Some of their explanations are that common diagnoses are, well, common, so providers may just settle on the convenient CAP diagnosis; CAP symptoms are also nonspecific and can mimic and overlap with other cardiopulmonary conditions. Moreover, providers are trained to treat early to avoid falling out of historical quality metrics.

What can you do about any of this?

1. Share this article with your medical providers so they are aware it is a problem.

2. Perform clinical validation queries, when appropriate

   • Here are the clinical indicators that should make you suspicious that the respiratory issue is less likely to be pneumonia and more likely to be another condition: no fever, no cough, normal vital signs, normal lung examination, normal white blood cell count;

   • If there is no reading of imaging other than the emergency provider’s (i.e., no overread by radiology of outside films);

   • If the radiologist’s read indicates the lungs are clear or states “no infiltrate” or “no pneumonia;”If the diagnosis perpetually remains in an uncertain diagnosis format; and

   • If the diagnosis perpetually remains in an uncertain diagnosis format; and

   • If there is conflicting or contradictory provider documentation (e.g., the pulmonologist doesn’t include pneumonia in their impression list).

3. Reconcile CDI/coder Diagnosis Related Group (DRG) mismatches to ensure that patients with more than one respiratory condition documented end up in the correct DRG.

4. Your facility may want to set up a task force, system, or quality improvement project to try to reduce misdiagnosis of pneumonia.

Efforts to decrease the incidence of misdiagnosis are not just for clinical documentation integrity; they should improve the quality of care of patients.

And shouldn’t that be our ultimate goal?

What is Chronic Atrial Fibrillation, Anyway?

By Erica Remer, MD, FACEP, CCDS, ACPA-C

The American College of Physician Advisors’ CDI Committee (disclosure: I am Chair) has published materials on numerous CDI topics. Each set of materials consists of a document detailing the points that a physician advisor should be aware of enabling them to educate their medical colleagues, and another downloadable document of a CDI tip for dissemination to the medical staff with pointers aimed at the clinician.

We are currently revising the tip on atrial fibrillation and flutter, and I found some interesting updates for you.

In 2023, the American College of Cardiology and American Heart Association published clinical practice guidelines on the diagnosis and management of atrial fibrillation (link to article here). For our and my committee’s purposes, the important update was the definitions of the types of atrial fibrillation (AF).

AF is the most common sustained dysrhythmia, which is characterized by disorganized, chaotic wiggling of the upper chamber/s of the heart causing irregularly conducted beats in the ventricles. It is clinically significant because patients seek medical attention, and it is associated with increased risk of stroke, heart failure, and death.

• Paroxysmal AF terminates within 7 days or less of onset. It usually does not require intervention. It often is recurrent and intermittent.

• Persistent AF is atrial fibrillation which is continuous and lasts for at least 7 days. It requires intervention to convert to normal sinus rhythm. It may recur. The article recommends that if a patient starts off with persistent AF, they should continue to be characterized as persistent even if the pattern of the AF subsequently becomes intermittent and “paroxysmal.”

• Long-standing persistent AF is persistent AF that lasts for more than 12 months. The patient and provider have not yet given up the hope that someday this patient may be converted into normal sinus rhythm.

• Permanent AF is persistent AF of some duration wherein the provider and patient make a shared decision to abandon efforts to convert the patient into a normal rhythm. There is no inherent difference in the character of the AF; it is just a therapeutic choice. These patients are often given medication long-term to control the rate of their AF and anticoagulation to reduce the risk of stroke or other issues from blood clots.

The new guidelines advise that the expression, “Chronic AF” is historical and should no longer be used. They explain “it has been replaced by the “paroxysmal,” “persistent,” “long-standing persistent,” and “permanent” terminology.” In an American Hospital Association Coding Clinic from 2019, guidance for the diagnostic statement of “chronic persistent atrial fibrillation,” was given that the coder should only use I48.1, the code prior to 2020 for persistent AF because “chronic atrial fibrillation is a nonspecific term” that could be referring to that laundry list of AF.  

We’ll come back to this in a few moments.

When revising the materials, we delved deeply into the use of “history of” in the context of AF. As we know, providers and coders have different understandings of what “history of” means. To a coder, it means “old, resolved, in the past, no longer present.” To a provider, it means, “the past medical history includes,” and they do not really make a mental distinction between historical conditions and chronic conditions. The ICD-10-CM code used for verbiage of “history of AF,” Z86.79, is titled Personal history of other diseases of the circulatory system.” Can’t get more vague than that, and it bundles in with “history of” heart failure, coronary artery disease, or aortic aneurysm.

It is disconcerting to a provider to use verbiage indicating a current condition if it is not currently manifesting. This action will require explanation to them. They need to understand that if a patient has had AF in the past and they are still receiving any kind of work-up or treatment for it presently, the patient should be diagnosed with some type of AF. Compare it to diagnosing hypertension in a patient on antihypertensives whose blood pressure is 100/54. They still have hypertension, albeit chronic and controlled, not a “history of.”

On rate control for recurrent short-lived, self-terminating AF = paroxysmal AF. On antiarrhythmic to prevent recurrence of AF which required cardioversion after several months = persistent AF. What if the patient is on anticoagulation for AF, but they are not sure what the duration or type of AF it was, and you can’t access their records? That may be a situation where documenting “chronic AF, type unknown” would be clinically appropriate.

I do not recommend using “chronic AF” indiscriminately. Persistent, long-standing persistent, and chronic AF are all comorbid conditions or complications (CCs) whereas unspecified and paroxysmal AF are not. Using “chronic AF” for all AF could be interpreted as fraud or abuse, an attempt to capture a CC when it isn’t warranted.

Another point to teach your providers is descriptive diagnoses, although they feel like they are being specified, often result in an unspecified code. “Atrial fibrillation with a rapid ventricular response” and “new-onset AF” convey urgency and clinical significance to the practitioner, but to the coder, they mean I48.91, Unspecified atrial fibrillation. In the case of the new-onset AF, there may be no more specificity to be had yet, but the former term may have opportunity to specify.

The last addition to our materials relates to atrial fibrillation as a potential sepsis-defining organ dysfunction. Clinicians are aware that new-onset AF may be heralding serious pathology, such as sepsis or diabetic ketoacidosis. The literature supporting this stance can be found here.

ACPA has many excellent benefits for the physician advisor member including the CDI resource materials (https://www.acpadvisors.org/). I strongly recommend you/your physician advisor check us out.

CMS Focuses on SDoH, BH in Proposed 2025 OPPS Ruling

By Tiffany Ferguson, LMSW, CMAC, ACM

In a move to support individuals at high risk of suicide or overdose, the Centers for Medicare & Medicaid Services (CMS) is proposing changes in the payment structure and services to expand telehealth options for care management and the inclusion of social determinants of health (SDoH) risk factors in outpatient treatment programs. These changes aim to provide better financial backing for practitioners who deliver critical interventions and follow-up care for at-risk populations.

Among the proposed changes, CMS is introducing separate payments for safety planning interventions. This new payment structure, GSPI1, will compensate practitioners for developing and implementing safety plans, which are vital for individuals identified as having a high risk of suicide or overdose. This code is in addition to or separate from the evaluation and management (E&M) visit or psychotherapy services.

Additionally, post-discharge follow-up contacts will now have separate payment, ensuring continuous care and support during a critical recovery phase. The proposal is looking to create a monthly billing code to provide post-discharge follow-up contacts that are performed in conjunction with a discharge from the emergency department for a crisis encounter. This would be a bundled service describing four calls in a month, each lasting between 10-20 minutes, GFCI1, pricing would be a direct crosswalk with principal care management - 99426.

CMS is also focusing on the integration of telehealth into more behavioral health treatment services. The new payment and coding proposals will facilitate the use of technology to enhance treatment delivery. This initiative is part of a broader effort to modernize and improve access to behavioral health services. The proposal is specifically looking at the expansion of telehealth for General Behavioral Health Integration, 99484 and Principal Care Management, 99424-99427, as well as potential reimbursement for FDA-approved digital mental health treatment (DMHT) devices (See page 377).

CMS is proposing updates for Opioid Treatment Programs (OTPs). In relation to SDoH, CMS is making efforts and opening comment for inclusion of the health-related social needs tool for those with opioid use disorder in OTPs. Although the reimbursement for providing the SDoH risk assessment was added last year, there is a recommended expansion to include this same add-on for OTP inclusive in the bundled intake. The enhanced payments will support a more thorough assessment of patients’ needs, including health-related social needs, harm reduction interventions, and recovery support services.

These proposed changes by CMS represent a proactive approach to addressing some of the most pressing issues in behavioral health and opioid treatment. For further details see the 2025 OPPS proposed ruling, starting around page 606.

Sepsis 2.5

By Erica Remer, MD, FACEP, CCDS, ACPA-C

In the old days, you walked into a patient’s room, visually surveyed them, and exclaimed, “(Spicy expletive deleted!) This patient is septic!” and then you rushed out to start aggressive treatment and transfer to the intensive care unit. The reason the pundits tried to specify what constituted sepsis in the 1990s was to avoid missing cases and incurring deaths.

The problem that we have with sepsis right now is that all the attempts to codify and operationalize it left out the most important characteristic. The patient is sick. I don’t mean sick like with a cold or a stomachache; I mean SICK, in all-caps.

I believe sepsis is part of the progression of non-self-limited untreated infections. It doesn’t happen with an adenoviral upper respiratory infection. It also isn’t present in every elderly patient with cystitis.

Sepsis is the penultimate stop prior to dying from an infection. When I used to present at mortality conferences, if I saw a patient who died from an infection, but sepsis was never recognized or explicitly documented, I referred the case to quality for education of the clinical personnel.

But how to recognize it? The general criteria of the systemic inflammatory response syndrome (SIRS) of fever/hypothermia, tachycardia, tachypnea, and abnormal white blood cell count was a start. These were a cue that something was amiss, but they were too nonspecific. Zillions of conditions could result in abnormal vital signs or an abnormal white blood cell count.

“Life-threatening organ dysfunction caused by dysregulated host response to infection” was closer. In writing this, I realized the problem is in the construction of this phrase. The organ dysfunction doesn’t really have to be immediately life-threatening. The condition of sepsis is what is life-threatening; we recognize it is present by the organ dysfunction it has caused.

Howard Rodenberg et al. published an article called, “Sepsis-2.5: Resolving Conflicts Between Payers and Providers,” on behalf of the Society of Critical Care Medicine. Putting aside the fact that I wish they had invited me to be a co-author, because I have been preaching what they wrote for a long time now, I really like their no-nonsense conceptual definition of sepsis:

Sepsis is present when a patient with infection exhibits evidence of organ dysfunction at a site external to the seat of infection, or more than what is routinely expected from a localized infection.

Septic shock is present when a patient with sepsis exhibits persistent hypotension following initial fluid resuscitation.

I used to say that sepsis is when a patient is sicker than the average patient with that underlying infection, heralded by organ dysfunction. A colleague disagreed with the sicker-than-average characterization – he pointed out that if a patient is in the ICU, an intensivist might not have the same definition of “sicker than average,” because the intensivist’s patients are all sick. I mean sicker than all others with that same underlying infection of pneumonia or cellulitis or UTI, not only in the cohort of patients for whom you are caring personally.

They propose that sepsis is identified by “an ill-appearing patient, documented or suspected infection, and evidence of organ dysfunction.” I like it!

I am not sure where some of the specific guidelines stem from (e.g., hyperlactatemia being > 3.0 mmol/L); had I been a co-author, I would not have set strict criteria. Hypoxia may be sufficient; does it have to meet the threshold of acute hypoxic respiratory failure? I would not endorse a specific change from the baseline. If someone normally has platelets of 120,000 and they have thrombocytopenia of 70,000 in the proper setting, I don’t think it is reasonable to reject that as clinically significant thrombocytopenia because the decrease from baseline is less than 50 percent.

I don’t want providers to have to tick off checkboxes and grab their calculator to see if a patient qualifies. I know this was crafted with payor input, but not all payers play by the same rules. I always think that a competent provider using sound clinical judgment should be able to make a diagnosis without being pigeonholed into a specific level or change from a baseline.

Their conceptual definition does explicitly resolve the issue of organ dysfunction involving the infected system. It says “or more than what is routinely expected from a localized infection.” Acute hypoxic respiratory failure is not routinely expected in pneumonia. Acute kidney injury is not typical for urinary tract infection.

The other thing that doesn’t sit well with me is the name. I don’t like “Sepsis-2” or “Sepsis-3” or “Sepsis-2.5.” That implies that the condition of sepsis has been changing and evolving. It has not! Sepsis has always been and will continue to be a real medical condition. The only thing changing is the words we use to describe and characterize the condition.

How about “life-threatening progression of infection identified by causing organ dysfunction?” I propose we don’t give clinicians more hoops to jump through making sure the patient meets specific thresholds and rigid criteria, but let’s allow providers to make the diagnosis and treat the patient, aggressively and with alacrity.

Is this Sepsis-3.1?

Should it be?

Social Admissions: Differences Using ABN and Preadmission HINN

By Tiffany Ferguson, LMSW, CMAC, ACM

There seem to be variances, particularly regionally, regarding the use of the preadmission Hospital-Issued Notice of Non-coverage (HINN) and the Advance Beneficiary Notice of Non-coverage (ABN). Both notices serve essential roles in informing patients about potential non-coverage by Medicare, but they differ significantly in their application and implications, especially in the context of social admissions.

The regulations surrounding HINNs, particularly the preadmission HINN, were last revised as Version 508 in April 2012, before the implementation of the Two-Midnight Rule and the Medicare Outpatient Observation Notice (MOON) in October 2013. This historical context may explain why preadmission HINNs were once more commonly used and may still be used more frequently in certain regions.

A preadmission HINN is issued by hospitals when they determine that an admission is not medically necessary, is inappropriate, or is custodial in nature. The primary purpose of a preadmission HINN is to notify the patient that the hospital, typically the utilization review committee (URC), does not consider inpatient admission to meet the criteria for Medicare coverage. This occurs when the attending physician supports inpatient admission; however, there is concern that the inpatient admission may not be medically necessary, is inappropriate, or can be considered custodial care. This notice is issued before or at the time of admission. The hospital (URC) is not required to obtain concurrence from the patient or the attending physician before issuing a preadmission HINN. This also applies to direct admissions to swing beds and in situations in which the hospital determines that skilled nursing facility (SNF) services are not needed.

An ABN, on the other hand, serves a broader function. In the case of custodial care, there are nuances to recognize, particularly since ABNs are provided when the hospital intends to provide care in the outpatient setting. The ABN is intended to inform Medicare beneficiaries in advance that Medicare may not cover certain items or services. The notice allows healthcare providers to shift financial liability to the beneficiary if Medicare denies the claim. An ABN must be provided before the service is rendered, giving the beneficiary time to make an informed decision about whether to proceed.

ABNs are used in a wide range of scenarios, typically involving services that may be deemed not reasonable and necessary under Medicare guidelines (or considered custodial care).

When considering social admissions made primarily for non-medical reasons, such as convenience or social support, the differences between using an ABN and a preadmission HINN become particularly relevant. The preadmission HINN focuses on the hospital’s assessment that the inpatient admission does not meet Medicare’s criteria for medical necessity or appropriateness. This notice serves to inform the beneficiary of their right to appeal to their Quality Improvement Organization (QIO) should they disagree with this decision.

The QIO will either concur with the member or the hospital on the medical necessity of the inpatient status. Should the QIO side with the hospital, the patient is financially liable for their Part A services.

If the hospital instead provides the patient with an ABN, the hospital is making a decision to keep the patient in an outpatient designation. The patient must provide a signature of acknowledgement prior to the services, if the hospital intends to bill the patient for any outpatient services that may be rendered while bedded.

Reviewing the use of the preadmission HINN, it appears to be most common in the Northeast, specifically New York. According to Livanta’s 2023 annual report, Region 2, which includes New York, New Jersey, Puerto Rico, and the Virgin Islands, reviewed the preadmission HINN 174 times, with an 81-percent concurrence rate with the hospital. The breakdown is interesting: 167 HINNs were given in New York, while only seven were given in New Jersey, and none in Puerto Rico and the Virgin Islands. In fact, New York last year provided significantly more notices than any other region. In Region 3 (Maryland, Delaware, Pennsylvania, and Virginia), this preadmission HINN was reviewed seven times by Livanta. In Region 5 (Illinois, Indiana, Michigan, Minnesota, Ohio, and Wisconsin), it was reviewed a total of six times, and in Region 7 (Iowa, Kansas, Missouri, and Nebraska), it was reviewed once in the entire year. The trends were similar in Acentra Health’s (formerly Kepro’s) annual reports for 2023. The highest number of preadmission HINNs, at 22, was found in Region 1 (Connecticut, Maine, Massachusetts, New Hampshire, Rhode Island, and Vermont).

This disparity suggests that New York hospital policies pertaining to utilization review seem to define the delivery of a preadmission HINN as the standard of practice for social admissions, with a decision to place patients in inpatient care, while other hospitals across the country are either not providing the preadmission HINN appropriately or choosing up front to keep patients in an outpatient designation.

In writing this article, I reached out to the Centers for Medicare & Medicaid Services (CMS) for clarification on why certain hospitals use the preadmission HINN versus the ABN for social admissions. I requested that the instructions for the HINNs be updated to reflect more current standards of practice, given that they are over 10 years old.

The response I received stated: “Thank you for your suggestion to clarify. We will take this into consideration with the next manual change.” One can only hope! 

Preadmission Screening Annual Resident Review (PASRR) and Long Hospitalizations

By Tiffany Ferguson, LMSW, CMAC, ACM

One thing I was never taught in my master’s in social work (MSW) program was the hospital requirement to complete a PASRR screening for every patient discharging to a skilled nursing facility (SNF).

The Preadmission Screening and Resident Review (PASRR) was created as part of the Omnibus Budget Reconciliation Act of 1987. PASRR requirements, added to the statute as sections 1919(b)(3)(F) and 1919(e)(7) of the Social Security Act, required states to create a system to assess the needs of individuals with mental illness or intellectual disability applying to or already residing in Medicaid-certified nursing facilities.

This system ensures that individuals are not being placed in such facilities unnecessarily or without adequate supports.

PASRR requires that Medicaid-certified nursing facilities:

• Evaluate all applicants for serious mental illness and/or intellectual disability;

• Offer all applicants the most appropriate setting for their needs (in the community, a nursing facility, or acute-care settings); and

• Provide all applicants with the services they need in those settings.

PASRR is an important tool for states to use in rebalancing services away from institutions and towards supporting people in their homes. To comply with the 1999 U.S. Supreme Court decision in Olmstead vs. L.C., under the Americans with Disabilities Act, individuals with disabilities cannot be required to be institutionalized to receive public benefits that could be furnished in community-based settings.

In brief, the PASRR process requires that all applicants to Medicaid-certified nursing facilities be given a preliminary assessment to determine whether they might have a serious mental illness or intellectual disability. This is called a “Level I screen.” Those individuals who test positive at Level I are then evaluated in-depth, called “Level II” PASRR. The results of this evaluation result in a determination of need, determination of appropriate setting, and a set of recommendations for services to inform the individual’s plan of care. Although this process is a federal regulation, it is managed through each state agency, often in different ways in each state.

In 2020, the PASRR regulations opened for public comment in efforts to acknowledge and attempt to reduce some of the inefficiencies. The most common compliant was about the time delay; patients are required to remain hospitalized for 7-9 business days while they await Level II evaluations. This means that every patient in a hospital awaiting a Level II PASRR evaluation must remain in the hospital during that time until they are approved for discharge to a SNF and/or nursing home facility. This process had become so burdensome that in California, Medicaid will cover administrative days for the hospital while patients wait for their Level II PASRR evaluation.

The conundrum I have with this regulation is multifaceted. It is unclear why, with technological advances and such a high focus on nursing home requirements, the proposed ruling has never been finalized or revised – and thus remains untouched since 2020. It is difficult to understand, with today’s staffing limitations and lack of hospital beds, how it can be considered perfectly acceptable for patients to sit in the hospital an additional 7-9 days while they wait for an outside evaluation.

The final argument is that this law misses the mark on the intention of inappropriately housing individuals in a care setting when a community setting would be more appropriate. This raises concerns about the growing number of patients who remain in the hospital for custodial reasons, often for extended lengths of time.

By PASRR standards, this would be greater than 30 days. Per PASRR requirements, an evaluation must be completed if the patient is going to a nursing home or SNF for more than 30 days with a serious mental illness or intellectual disability diagnosis. However, that same patient could remain in the hospital for months to years, without such evaluation or support from the state, for community-based services.

There are many arguments for why the hospital is not the best setting for custodial patients. These include risk of exposure to infections, the impact on other patients receiving medically necessary care by holding beds, and staff burnout. There is also a lack of socialization, sunlight, physical activity, and rehabilitation for the mind and body.

Hospitals are not built or trained to care successfully for patients with long lengths of stay, especially for patients without medical needs, and thus these patients are often neglected and ignored, isolated in their hospital rooms.

It appears that this is a void in our medical system and federal regulations, where prolonged custodial hospitalization is a tolerated practice, often due to lack of alternatives; however, transitioning to a SNF requires extra red tape because of federal and state attention.  

Why Medicare Advantage Denials for Patients with 2MN Still Exist

By Juliet Ugarte Hopkins, MD, ACPA-C

It has nearly been six months since 42 CFR 422.101(b)(2) within the Code of Federal Regulations was officially set into motion on Jan. 1, obligating Medicare Advantage (MA) plans to follow the Medicare Two-Midnight Rule and Medicare Inpatient-Only (IPO) lists when it comes to patient status.

Have all the postulated outcomes come to fruition? Hardly. Let’s consider two of them:

1. Denials from MA plans for inpatient-designated hospitalizations will drop precipitously. This most certainly has not happened, even anecdotally, because the Two-Midnight Rule does not simply involve passage of a second midnight in the hospital. It involves medical necessity of that second midnight. As much as it seems we have been talking about medical necessity since October 2013, when the Two-Midnight Rule was introduced by the Centers for Medicare & Medicaid Services (CMS), many continue to apply It inappropriately. More on this in a bit. While CMS made it clear in 2023 that MA plans must utilize the Two-Midnight Rule, they did not include any consequences or penalties if they do not. While some have recommended filing complaints to CMS about violations of the federal regulation by sending an email to your CMS regional office with the number of cases and the payors involved, this is only a suggestion, and not a formal notification or grievance process established by CMS. As such, there is technically nothing – at this point – to stop MA plans from electing not to comply with the Two-Midnight Rule.

2. Scores of MA plan medical directors will lose their jobs on Jan. 1 due to the simplicity of status determination related to application of the Medicare Two-Midnight Rule and subsequent decline of MA plan denials. Once again, application of the Two-Midnight Rule for hospitalizations covered by MA plans has not resulted in fewer denials. As such, there likely has been no decline in employment opportunities for MA plan medical directors. In fact, a cursory review of LinkedIn, and other job search engine sites results in plenty of postings for this type of position.

Let’s get back to the point about the passage of two midnights with medical necessity.

The problem is the term itself – “medical necessity.” In actuality, the concept is “medically necessary care which can only take place in the hospital setting.” Physical and occupational therapy services (including initial assessments), administration of oral medications, and assistance with activities of daily living (ADLs) and ambulation can all take place outside of the hospital. As such, none of these patient needs qualify as “medically necessary,” in the context of the Two-Midnight Rule.

This does not mean they are not necessary.

They most certainly are, and the lack of any of these services could very well lead to a disastrous or even deadly outcome for the patient. This conflict of concepts escalates the struggle clinicians and case/utilization management staff experience when making patient status decisions. 

Passage of a second midnight related to a delay in care also does not fit the bill for “medical necessity.” However, this can be trickier to assess. If discharge is delayed today because the hospitalist is waiting for cardiology to read the echocardiogram, which was performed yesterday – that’s a no-brainer.

Same goes for the patient who is hospitalized Friday and remains hospitalized Monday, because the hospital doesn’t perform stress tests on the weekends. But what if a patient with appendicitis presents to the emergency department at noon, an uncomplicated laparoscopic appendectomy takes place the following day at 4 p.m., and the patient discharges at 7 a.m. on the third day following an unremarkable recovery? Was the passage of two midnights medically necessary? It depends. Was there a medical condition that needed to be addressed or corrected before the surgery could take place, pushing it back to later in the day on the second day? Or was the patient ready for the procedure the afternoon of the first day, but there was no room in the surgical schedule until the following afternoon?

If the documentation is not clear about which was the case, you can be sure the payer will assume there was simply a scheduling delay and deny inpatient status.

Finally, let’s talk about urinary tract infections (UTIs). Maybe not just urinary tract infections – let’s utilize UTIs as a general stand-in condition when considering medical necessity.

You know the drill: “IV Ceftriaxone until urine culture and sensitivities return” is the common history and physical plan refrain, followed by a statement indicating that two midnights are expected.

Why? Because it generally takes at least two midnights, sometimes three, for a urine culture to demonstrate the infective pathogen and its sensitivities to various antibiotics. But why can’t the patient be placed on an oral antibiotic and discharged with follow-up on the urine culture results by their primary care provider? Is there a history of complicated UTIs for anatomical or instrumentation reasons, or has the patient suffered from highly resistant organisms in the past?

If these points are not applicable or are not documented in the record, medical necessity of two midnights cannot be assured as proven. 

It is imperative for clinicians, physician advisors, utilization managers, and yes, MA plan medical directors and case managers to understand the Medicare Two-Midnight Rule.

While a much simpler approach, I believe applying the Rule to any patient who remains hospitalized for at least two midnights is a non-compliant practice.

Don’t create confusion and frustration for your medical staff and utilization/case management colleagues. Utilize the Rule correctly, request additional clinician documentation when warranted to support medical necessity, and find other ways to address discharge delays in your hospital – other than attempting to pass them as appropriate inpatients.

CY25 Proposed Updates to ESRD Prospective Payment System (PPS)

By Tiffany Ferguson, LMSW, CMAC, ACM

On June 27, the Centers for Medicare & Medicaid Services (CMS) issued a proposed rule to update payment rates and policies for renal dialysis services for Medicare beneficiaries.

The updates, effective from Jan. 1, 2025, include changes to the End-Stage Renal Disease (ESRD) Prospective Payment System (PPS), with most notable adjustments being made to support acute kidney injury (AKI) dialysis payment rates, as well as the operationalization of oral-only drug inclusion in the ESRD PPS.

Here are some key highlights from the proposed rule:

1. ESRD PPS Base Rate Increase: Understanding that this population continues to rise in number, as does the corresponding cost of care, CMS is preparing for another year of influx. CMS has projected that for the 2025 calendar year (CY), Medicare will be paying $7.2 billion to approximately 7,700 ESRD facilities for renal dialysis services. CMS proposes increasing the ESRD PPS base rate to $273.20, up by $2.18 from the current $271.02. Total payments to all ESRD facilities are projected to rise by approximately 2.2 percent. Specifically, hospital-based ESRD facilities will see a 3.9-percent increase, and freestanding facilities will see a 2.1-percent increase.

2. Wage Index Changes: CMS proposes using a new methodology for the ESRD PPS-specific wage index that would be used to adjust ESRD PPS payment for geographic differences in area wages. By combining data from the Bureau of Labor Statistics and freestanding ESRD facility cost reports, CMS hopes to replace the existing hospital wage index values in this payment methodology.

3. Outlier Policy Updates: CMS proposes to expand the existing list of ESRD outlier services to include more drugs and biological products that traditionally have been included in the composite rate prior to establishment of the ESRD PPS. CMS is also proposing some slight technical adjustments in methodologies for calculating fixed-dollar loss (FDL) and Medicare allowable payment (MAP) amounts. For CY 2025, the FDL amount for pediatric beneficiaries would increase, while the amount for adults would decrease.

4. Low-Volume Payment Adjustment (LVPA): CMS is proposing to modify the LVPA policy to create a two-tiered system, in efforts to better align payment methodology with resource utilization. The two-tiered system designates that facilities with fewer than 3,000 treatments will receive a 28.3-percent adjustment, and those with 3,000 to 3,999 treatments will receive an 18.0-percent adjustment. Tiering is based on the median treatment count over the prior three years.

5. Inclusion of Oral-Only Drugs: CMS will include oral-only renal dialysis drugs in the ESRD PPS bundled payment, starting Jan. 1, 2025. This is expected to increase access to these drugs, particularly benefiting those without Medicare Part D coverage.

Updates for AKI Dialysis: CMS proposes extending Medicare payment for home dialysis for AKI patients, allowing them more choices in their treatment. ESRD facilities will be able to bill for home and self-dialysis training for AKI patients. The proposed payment rate is set to be equal to in-center dialysis treatment.

The provision for AKI dialysis treatment in ESRD facilities was first released in CY 2017; however, in-home treatment was excluded due to the nature of AKI typically being of a short duration, and the work for set-up and training raised concerns at that time. However, since 2020, there have been significant efforts to promote home peritoneal and hemodialysis for patients, rather than in-center.

Currently, CMS feels there is enough data to expand this to AKI patients. The expansion of ESRD facility conditions for coverage to include home dialysis for AKI is an intentional approach to increase home dialysis access and utilization, while also promoting greater opportunities to increase AKI treatment and management in the outpatient setting. 

The proposed rule by CMS introduces several updates aimed at improving payment rates and policies under the ESRD PPS for CY 2025. Key proposals include increasing the base payment rate, revising the wage index methodology, updating outlier policies, and extending Medicare payments to home dialysis for AKI patients.

These changes reflect CMS’s commitment to enhancing patient care, expanding access to essential medications, and ensuring that payment adjustments align with the resource use of low-volume facilities.

Warding off Clinical Validation Denials with Better Documentation

By Erica Remer, MD, FACEP, CCDS, ACPA-C

Clinical validation (CV) denials are plaguing us lately. When I work on projects entailing medical record review, I must admit that it is not unusual for me to agree with the payer. However, if a legitimate diagnosis is denied, the blame often stems from the documentation.

I recently had an epiphany as to a methodology for providers to put “mentation” into their documentation and ward off denials.

I recommend a macro in the setting of sepsis. It goes like this:

Sepsis due to (infection) with acute sepsis-related organ dysfunction as evidenced by (organ dysfunction/s).

If the coder can’t identify an infection, either the provider is not documenting it properly (e.g., “decubitus ulcer” is not equivalent to “an infected decubitus ulcer with surrounding cellulitis”), or there is no sepsis. There has to be an infection in order to progress to sepsis.

If the practitioner can’t provide any organ dysfunction in that field, either they missed the organ dysfunction (organ dysfunction is not only SOFA, Sequential Organ Failure Assessment), there is no sepsis, and/or there is increased risk of denial.

When I thought about it, I realized that this sentence composition works for lots of conditions, and could be prophylactic against other CV denials:

• Acute-on-chronic hypoxic and hypercapnic respiratory failure due to severe exacerbation of chronic obstructive pulmonary disease (COPD) as evidenced by oxygen saturation in mid-80s and increased CO2 with pH of 7.32. Placed on BiPAP, steroids, nebulizers, and antibiotics. Will monitor closely.

• Pneumonia, probably aspiration, as evidenced by fever, cough, shortness of breath, elevated WBCs, and RLL infiltrate. Was known to have vomited in NH two nights prior. Being treated with antibiotics and supplemental oxygen.

• Severe malnutrition due to cachexia from pancreatic CA as evidenced by BMI 14.7, loss of weight of 10 percent in last month, and muscle wasting. Appreciate dietitian consult. Will implement appetite stimulation and dietary supplementation.

The elements of this construction are:

1. The condition being diagnosed affirmatively or uncertainty (e.g., possible, probably, suspected, likely, etc.);

2. The etiology, using linkage (e.g., from, due to, caused by, as a result of, etc.);

3. The manifestations (clinical indicators); laboratory, imaging, or other supporting evidence (e.g., as evidenced by…); and

4. Plan of treatment.

Maybe they should think about it as: condition…due to…as evidenced by…treated with…

The provider doesn’t need to do this every time they are discussing the diagnosis (i.e., it doesn’t need to be copied and pasted from day to day). It needs to be done upon the initial diagnosis. From that point on, the practitioner just needs to document whatever is relevant that day. Is it getting better? Did some of the manifestations resolve? How is the plan changing? Did the diagnosis evolve from uncertain to definitive? Having multiple mentions of the ongoing condition will demonstrate that it is a clinically valid and significant diagnosis. A best practice is to have it reappear in the discharge summary as well.

Teach this to residents and onboarding providers, and when you are educating service lines on best practice. If you give feedback, use this construction to model good documentation practice.

Some of you might be thinking, “but the payers have specific, ridiculous, unattainable criteria that they demand for the diagnosis of that condition.” That is a different problem. If that is baked into your contracts, address it. If they are just quoting silly criteria to justify unjust denials, fight it. Good documentation doesn’t completely eliminate denials, but it does help the person responsible for appeals, and helps me assess clinical validity when I perform audits.

Have your provider practice good medicine and produce good documentation, expressing what they thought and why they did what they did. It won’t make all denials go away, but it should decrease the number of them. As a bonus, this pattern of construction may help the clinicians organize their thoughts and communicate their thought process to their colleagues.

And that might just improve the quality of care delivered to your patients!

2024 Q1 Coding Clinic Reinforces “As Many Codes as it Takes” Notion

By Erica Remer, MD, FACEP, CCDS, ACPA-C

I am overdue to give my comments on the 2024 American Hospital Association (AHA) Coding Clinic published for the first quarter. I really appreciate this Coding Clinic, because it gives reminders of general coding rules.

One of the things I have always loved about ICD-10-CM reflects my mantra: “as many codes as it takes.”

• The revisions to the ICD-10-CM Coding Guidelines include the following:

  • The additions of “other surgical site” for post-procedural and obstetrical surgery in the context of sepsis. This means that there should be at least four codes for sepsis in the setting of a procedure:

    • A code for the site of infection, specifying the area involved, like superficial or deep incisional surgical site;

    • The code to specify that there is post-procedural sepsis, T81.44 or O86.04;

    •  A code to identify the infectious agent, if possible;

    • R65.2- for severe sepsis, because all sepsis is now “the condition formerly known as severe sepsis;” and

    • And at least one code to detail what the sepsis-related organ dysfunction was.

• The answer to a question about how to code “acute prostatitis and cystitis” reminds us that you need both the code for prostatocystitis, which establishes the site, and N41.0, Acute prostatitis, to establish the acuity.

• On page 20, the guidance explains that a postprocedural intra-abdominal abscess would take 2 codes – T81.43XA, Infection following a procedure, organ and space surgical site, initial encounter, and K65.1, Peritoneal abscess – to fully flesh it out.

There were several recent questions that arose regarding what “other” or “other specified” codes are meant for conditions for which the provider gives details and often linkage, but with no specific code for the etiology. For instance, chorioamniotic separation goes to O41.8X30, Other specified disorders of amniotic fluid and membranes, third trimester, not applicable or unspecified; and E27.49, Other adrenocortical insufficiency, is one of the codes needed to capture the condition of hypothalamic pituitary adrenal axis insufficiency.

A question on page 17 regarding the coding of the verbiage “rheumatoid arthritis with inflammatory polyarthropathy” points out to me that you need to take into consideration where in the classification a code appears. M06.4, Inflammatory polyarthropathy, lives under the umbrella of M06, titled Other rheumatoid arthritis. It probably would have been better if it were titled Inflammatory rheumatoid polyarthropathy. However, I (not Coding Clinic) would also suggest that if the provider had specified “rheumatoid factor positive RA with inflammatory polyarthropathy,” the more accurate code would be M05.79, Rheumatoid arthritis with rheumatoid factor of multiple sites without organ or systems involvement.

There was a set of questions regarding dural tears that I am sure is driving some of you crazy. My experience is that some quality and clinical documentation integrity (CDI) teams try to get their providers to perform contortions to get out of triggering patient safety indicators and complications. They think that the magic words to preclude a complication from being considered a complication are “inherent to” or “integral to.” However, on pages 20 and 21, Coding Clinic is pointing out that regardless of whether the circumstances would have led to a dural tear for anyone performing that procedure, or whether there is thinning, scarring, adhesions, or stenosis, the condition is clinically significant and should be documented and captured as an accidental puncture or laceration. I agree with their assessment.

Pages 23-24 posed an interesting question. A patient presented to the ED with an exacerbation of asthma and did not have access to their albuterol inhaler. Coding Clinic explained that you don’t use an underdosing code for as-needed medications. The code they recommend is Z91.198, Patient’s noncompliance with other medical treatment and regimen for other reason. It makes me think that I might have been using the wrong code for underdosing of antipyretics (to reduce fever). It is still clinically significant and should be recorded (it might be the reason that mom dragged the baby out in the middle of the night – giving too low a dose of acetaminophen), but since it isn’t a long-term drug or prescribed course, like 10 days of antibiotics, it isn’t a “medication regimen.”

There was also a question about transaminitis and hyperbilirubinemia being documented as “acute liver injury due to metastatic liver disease and chemotherapy.” The indexing of “acute liver injury” goes to a trauma code in S36. The coder recognized that this is not correct. Coding Clinic advises using K71.8, Toxic liver disease with other disorders of liver and then codes for the metastasis and adverse effect of the antineoplastic drugs. Since the etiology is known and there is no distinct combination code, K71.8 is indicated, not K71.9, Toxic liver disease, unspecified. The next question also tackles the use of a trauma code – if there is no trauma, you shouldn’t use a trauma code (an S-T code). Medical intervention misadventures are not considered “trauma.”

The next one is a head-scratcher to me. On page 28, a patient with coronary artery disease and a bypass graft (CABG), presents to the ED with chest pain over three days and is diagnosed with a non-ST-segment elevation myocardial infarction (NSTEMI), likely from the stenosis of a vein graft. They recommend I24.4, NSTEMI, as the principal diagnosis and I25.10, Atherosclerotic heart disease of native coronary artery without angina pectoris, as the codes. Their reasoning is that “it would be inappropriate to assign a code for angina in the setting of an MI.”

Am I to infer that if the patient hadn’t suffered an MI, you would have used the code indicating “of CAD graft” with “unstable angina pectoris?” What if the patient had chronic stable angina prior to the MI? Are you not allowed to include that information in the coding? I think what they were trying to convey is that the NSTEMI was the progression of the atherosclerosis of autologous vein coronary artery bypass graft with unstable angina, and that once you had the MI, it superseded the CAD in that distribution. In addition, the patient was found to have other CAD, which was not felt to be the etiology of the chest pain.

I know that coding is based on documentation. We want the codes to reflect the conditions the patient has, so if the coding based on documentation doesn’t tell the story accurately, either the documentation or the coding must be adjusted. And…with as many codes as it takes.

As always, I recommend you read the Coding Clinic in its entirety on your own.

Programming Note: Listen live to Dr. Erica Remer as she reports this story during today’s Talk Ten Tuesdays broadcast at 10 a.m. EST.

It Takes Failure to Have Respiratory Failure

By Erica Remer, MD, FACEP, CCDS, ACPA-C

I have been performing a lot of chart reviews in my consulting capacity, making clinical validation determinations. Whether I am hired by the payor or the provider, I assess the clinical course and documentation and try to make a fair appraisal. One of the conditions that is particularly irksome is acute hypoxic respiratory failure.

For the newly updated pediatric sepsis definition, a survey of practitioners was undertaken to suss out what the consensus was for the medical condition. A scenario was posed, and the respondents weighed in on whether they thought it constituted sepsis or not. I decided to try that approach with acute hypoxic respiratory failure.

I posted this case on LinkedIn:

O2 sat of 89 percent on room air. 2L by n/c gets pulse ox to 95 percent. PE: No acute distress. Not toxic appearing. Exacerbation of COPD.

Do you think this is acute hypoxic respiratory failure?

Almost all of those who responded gave me a resounding “no.” A couple of people wanted a little more information, to try to determine if the patient had chronic respiratory failure. One person cited Pinson & Tang’s position that a pO2 < 60 mmHg (SpO2 < 91 percent) on room air was sufficient to conclude hypoxemic respiratory failure.

This highlighted the fact that in clinical documentation integrity (CDI) practice, we often depend on dogma put forth by respected individuals in the field (including me!) We teach our providers and create internal clinical guidelines to support our positions. If there is established evidence-based literature, great! If not, we sometimes fall back on expert opinion. One of the pitfalls of this is we sometimes spurn common sense and rely too heavily on medicine by checkbox.

I believe I have sorted out why this condition unsettles me and causes us all such grief. I am going to try to lay it out for you all, clinicians, coders, and payers. Please note that I am specifically honing in on adult respiratory failure.

The ICD-10-CM code J96.01 is titled “Acute respiratory failure with hypoxia.” Why is it written that way, and not “acute hypoxic respiratory failure?” Let’s deconstruct this.

I posit that first, you need to determine if there is acute respiratory failure.

Acute respiratory failure implies a rapidly developing, severe, and life-threatening impairment of the body to intake or absorb adequate oxygen – or to expel sufficient carbon dioxide. If left untreated, the individual is at high risk of dying from inadequate oxygenation of tissues (e.g., heart or brain) or profound acidosis, which is incompatible with life.

Chronic respiratory failure is more insidious. It often develops more gradually, so the body has time to compensate. The patient learns to deal with shortness of breath if they walk too fast or to stop exerting themselves before ischemia gives them chest pains. If the person comes to equilibrium in chronic respiratory failure, they will have a chronically reduced oxygen level, persistently elevated carbon dioxide level, or both, but their pH will be near-normal. Their status is tenuous, and a new stressor can disrupt the delicate equilibrium, setting off the same life-threatening cascade of events as acute respiratory failure. This is termed acute and (or on) chronic respiratory failure.

The second part of the title is “with hypoxia.” Having a pO2 of < 60 mmHg constitutes hypoxemia, a reduced oxygen level in the bloodstream. Hypoxia is defined by Merriam-Webster as “a deficiency of oxygen reaching the tissues of the body whether due to environmental deficiency or impaired respiratory and circulatory organs.” Hypoxia is the life threat, although it often follows and is heralded by hypoxemia. This is why the code isn’t “with hypoxemia.” Practically speaking, however, we use the terms interchangeably.

In order to have acute hypoxic respiratory failure, you need to have all components: acuity, low oxygen delivery to tissues, and the threat to life.

People are (mis)interpreting Pinson and Tang’s guidance to mean that simply having a pO2 < 60 (SpO2 < 91 percent) or a P/F ratio of less than 300 is adequate to diagnose acute respiratory failure with hypoxia. Even with the caveat that one cannot apply the criteria to patients with chronic respiratory failure, this is not sufficient. This level of oxygen establishes hypoxemia, but more is needed to satisfy the acute respiratory failure piece.

Acute respiratory failure is symptomatic and requires substantial treatment. The symptoms/signs may be pulmonary in nature, like shortness of breath, tachypnea, or intercostal retractions. They may relate to the consequences of the hypoxia or hypercapnia, like anxiety, lethargy, or confusion. These patients typically appear acutely ill.

Treatment is necessary. When I practiced clinically, there were patients on whom I would toss a nasal cannula at a couple of liters per minute. I was trying to make them more comfortable, but I didn’t necessarily think they would succumb without it. Patients in acute respiratory failure need treatment to survive.

For hypoxia, they need significant oxygen supplementation. Most pundits expect to see four or more liters per minute (whatever you consider “high-flow”). A patient does not need to be intubated, but if they are intubated to address air exchange defects (as opposed to “airway protection”), they are in respiratory failure.

Let us remember that when we are reviewing records, we are not assessing the patient directly, but the documentation. It makes it so much easier if the provider produces good documentation to arrive at a valid conclusion. Here are the things I look for:

• A history describing development over a short period of time;

• Signs or symptoms of respiratory distress (e.g., shortness of breath, tachypnea or bradypnea, air hunger, supraclavicular/intercostal/subcostal retractions, accessory muscle use, abnormal vital signs and/or findings on lung exam) or end-organ dysfunction from hypoxia/hypercapnia (e.g., restlessness, anxiety, reduced level of consciousness, metabolic encephalopathy, diaphoresis, dysrhythmias);

• Consistency in documentation;

  • The history of present illness depicting a patient in distress with symptoms consistent with etiology. Description of baseline status if pertinent;

  • A general description including some degree of distress, ill-appearing, or in extremis (e.g., “in no acute distress, non-toxic, not ill-appearing” would be inconsistent);

  • Re-evaluation of the patient’s condition at reasonable intervals;

  • The declaration of an impression of acute hypoxic respiratory failure at the time (e.g., not days later for the first time, when the patient is no longer exhibiting the signs and symptoms). I expect it to be high on the impression list, not number 9 as an afterthought;

  • If the diagnosis is prompted by a query, it better be supported with the thought process by the provider. I don’t want to see a random diagnosis just pop up without clinical support and only appear once;

  • The diagnosis carried throughout the record by all providers. When the condition resolves, it should be resolved in the impression list. It can even be removed if the length of stay is prolonged, but it should reappear in the discharge summary; and

  • No unresolved conflicting documentation amongst clinicians.

• Critical care time being claimed in the ED, the drawing of an arterial blood gas, a pulmonary consult, and/or admission to the ICU are supportive (but not necessary) clinical indicators; and

• Treatment suggesting the diagnosis is clinically valid and significant

  • E.g., high-flow oxygen, CPAP, BiPAP, intubation; and

  • Aggressive treatment of causative conditions (e.g., antibiotics, pulmonary toilet, steroids, respiratory treatments as indicated).

How about that case? The patient meets criteria for hypoxemia, but is in no acute distress, nor is toxic-appearing. They have a reasonable etiology (COPD), but they only need 2 L/min by nasal cannula. These clinical indicators do not seem to support acute or acute-on-chronic respiratory failure with hypoxia. I would diagnose this as “acute exacerbation of COPD with hypoxemia.” If I were the documenter, I would also have discussed the patient’s baseline to support chronic respiratory failure, if I felt they had it.

Clinicians, if a patient has a life-threatening respiratory issue, think acute or acute-on-chronic respiratory failure. Document the historical and physical points and data that support the diagnosis. Specify whether it is hypoxic, hypercapnic, or both. Be precise and consistent in your documentation. Don’t downplay it or attribute the signs and symptoms solely to alternate diagnoses.

I would recommend linking the acute or A/C respiratory failure to the underlying etiology. I also like linking the manifestations. Detail your thought process and make it hard for the payor to deny the existence of the condition:

Acute-on-chronic hypoxic respiratory failure from pneumonia complicating acute exacerbation of COPD, with pO2 86 percent on baseline home O2 2 L demonstrating air hunger, accessory muscle use, and diffuse wheezing.

If the patient has a low oxygen level, but is not exhibiting signs of respiratory failure, and only requires low-flow oxygen supplementation, call it hypoxia or hypoxemia. If they have acuity, and severity, and require sufficient treatment, call it acute respiratory failure with hypoxia.

Payers do not misinterpret this article and cite that the documentation has to have all of the elements above, or else acute respiratory failure is ruled out. There is always an underlying condition, but there are very few diagnoses in which acute respiratory failure is considered inherent (e.g., acute respiratory distress syndrome, ARDS). Don’t try to gaslight the provider that acute respiratory failure is inherent to pneumonia or heart failure. Clinicians must take in all the facts and data and use their clinical judgment to render a diagnosis.

If you are going to take anything from this article out of context, don’t use it at all!

Physician advisors, bookmark this article and send it to your providers when you discover opportunity in your medical staff’s documentation of respiratory failure. Feel free to condense it and make a tip sheet for them.

Finally, CDI specialists, don’t try to force the provider to use the diagnosis of acute respiratory failure because it is a major comorbid condition or complication (MCC). Our job is to make sure the medical record is accurately depicting how sick and complex the patient is. They need to look as sick in the EHR as they do in real life: no less, but no more.

(For more education on optimal documentation practices, check out Dr. Remer’s Documentation Modules with CME).

Reference: Richard Pinson, Cynthia Tang, Acute Respiratory Failure – All There Is To Know, Pinson & Tang, September 5, 2023.

Navigating the Complexities of Outpatient in a Bed (OPIB)

By Tiffany Ferguson, LMSW, CMAC, ACM

The implementation of outpatient in a bed (OPIB) classifications in hospital settings presents unique challenges, particularly regarding billing, patient care, and regulatory compliance. These challenges may vary across hospitals, depending on electronic medical record (EMR) classifications and internal hospital bylaws or policy guidelines for managing this population.

With the looming continued aging of baby boomers requiring more healthcare services and increased social awareness of patient healthcare needs, compounded by provider moral distress, and the administrative pressures of doing more with less, social hospitalizations are continuing to rise.

Understanding and Documenting OPIB

OPIB is defined as an outpatient medical record designation that identifies a patient bedded in the hospital who initially or no longer warrants medically necessary services, from a billing perspective. These patients are often placed in the hospital setting because their social situation leaves them with no other alternative, or they remain in the hospital until an alternative location can be secured, despite the completion of observation services. To date, the Centers for Medicare & Medicaid Services (CMS) has not recognized this level-of-care designation in hospitals.

Since these patients are outpatients, the full admission requirements for hospital inpatients are not required. Thus, documentation requirements should fall to the hospital bylaws and internal policies and procedures.

Extended Observation versus OPIB

It is rare and unusual for a patient receiving observation services to remain in the hospital for a prolonged period. Since observation is a service, the review of medically necessary observation services is either completed, and the patient’s continued stay is due to social considerations, or the patient’s condition warrants hospital admission. For patients who remain purely for custodial reasons, they should be notified that observation services have ended.

There should also be a clear transition documented in the EMR, with an end to the observation services ordered or a change in the EMR encounter type to indicate that observation hours have been completed. This will help avoid confusion regarding any changes or decrease in services due to the patient’s lack of medical necessity.

Key Considerations for OPIB Classification

• Defining Outpatient Classification: Clearly define what constitutes an outpatient classification. This involves setting clear criteria and guidelines for when a patient is placed in the hospital for an OPIB designation. Consider incorporating this into the patient classification guidelines to guide the utilization review and physician advisor team in reviewing these patients.

• Documentation Requirements: If this designation is a part of your hospital’s EMR, then it will be helpful to ensure that policies and procedures reflect the standards for provider and clinical staff documentation.   

• Acceptance and Decision Tree: Develop an acceptance or decision tree protocol for determining when a patient qualifies for OPIB. This protocol should outline specific criteria and decision points for healthcare providers to ensure that medical necessity does not exist or has ceased. This decision tree should also include review of alternative locations that may be better-suited for these individuals, rather than the hospital.

• Patient Communication: Inform patients about their OPIB classification, including how it affects their care and billing. Clear communication helps manage patient expectations and reduces confusion regarding their treatment plan.

• Billing and Ancillary Services: Determine if and how the hospital plans to charge for services rendered to OPIB patients, particularly ancillary/professional services. Establish transparent billing practices to ensure that patients are aware of potential costs. Should the hospital determine that they are not going to bill the patients, it will be key to maintain appropriate reporting (see Dr. Ronald Hirsch’s comments and articles on A2970).

• Internal Tracking and Reporting: Implement a system to track OPIB cases internally. This tracking is crucial for future prevention strategies, charity reporting, and maintaining compliance with regulatory standards.

Avoiding Social Admissions

To prevent social admissions, consider leveraging agreements with hospital-owned or community partners. This might include post-acute care facilities, community housing settings, or other properties the hospital owns or partners with. Establishing a fast track from ED to post-acute care may be a lower-cost alternative to taking up medically necessary resources in the hospital (which is likely a whole other article as to why the hospital is not the best fit for social care).

Effectively managing OPIB classifications requires clear definitions, documentation, and proactive communication with patients. Acknowledging the new norm and creating internal guidelines for addressing this population, hospitals can navigate the complexities of OPIB, maintain regulatory compliance, and provide appropriate care while managing resources efficiently.

Properly implemented, these strategies will help avoid legal challenges and improve overall patient satisfaction and outcomes.

NIH Study Finds Link between Treatment-Resistant HTN, SDoH

By Tiffany Ferguson, LMSW, CMAC, ACM

A recent study published in the Journal of the American Medical Association (JAMA) and funded by the National Institute of Neurological Disorders and Stroke (NINDS), part of the National Institutes of Health (NIH), has found that adverse economic and social conditions significantly increase the likelihood of developing treatment-resistant hypertension.

Such hypertension is defined as blood pressure that remains above 140/90 mmHg despite optimal use of three antihypertensive medications of different classes, including a diuretic. This condition significantly raises the risk of stroke, coronary heart disease, heart failure, and all-cause mortality.

The study analyzed data from 2,257 Black and 2,774 white adults participating in a larger study that includes over 30,000 Americans. Approximately half of these participants reside in the “Stroke Belt” in the southeastern United States, where stroke mortality rates are higher than the national average.

The social determinants of health (SDoH) factors in the study examined education level, income, social support, health insurance status, and neighborhood conditions. The study identified that several of these factors were linked to an increased risk of developing treatment-resistant hypertension, such as having less than a high-school education, a household income under $35,000, lack of social interactions, absence of someone to provide care during illness, lack of health insurance, residing in a disadvantaged neighborhood, and living in states with poor public health infrastructure.

Over a span of 9.5 years, the study observed that 24 percent of Black adults developed apparent treatment-resistant hypertension, compared to 15.9 percent of white adults. While adverse SDoH increased the risk for both groups, Black adults were disproportionately affected by these adverse conditions, leading to a higher incidence of the condition.

The researchers suggest that addressing the SDoH could mitigate the racial disparities observed and subsequently reduce the higher risks of stroke and heart attack among Black Americans. Effective interventions suggested included improving access to education, increasing household income, enhancing social support networks, expanding health insurance coverage, improving neighborhood conditions, and bolstering public health infrastructure.

In response to these findings, the NINDS Office of Global Health and Health Disparities is developing strategies to promote health equity. In August 2023, a supplement featuring 10 manuscripts was published, providing recommendations for addressing SDoH.

Furthermore, the NINDS’s Mind Your Risks® campaign, launched in 2016, highlights the link between high blood pressure and dementia, particularly targeting Black men ages 28-45. This campaign offers strategies for preventing and managing high blood pressure to improve brain and cardiovascular health.

Although the SDoH risk factors will take more than publications, manuscripts, and educational campaigns to change, this research is significant in acknowledging these disparities for future opportunities of targeted funding to high-risk communities.

Understanding Billing Requirements for Caregiver Training Services

By Tiffany Ferguson, LMSW, CMAC, ACM

As part of the Jan. 1, 2024 Outpatient Prospective Payment System (OPPS) guidelines, caregiver training services (CTS) codes were listed as billable services if provided by physicians and non-physician practitioners (NPPs).

I thought I would revisit these codes with greater specificity, as I am seeing some questions come up related to lack of provider awareness, and some confusion regarding providing support in an individual versus a group setting. The intent behind this ruling was to provide practitioners with the opportunity to be reimbursed for their time in providing treatment planning and supporting services to caregivers who support the management of their patients. 

Pursuant to the Recognize, Assist, Include, Support, and Engage (RAISE) Family Caregivers Act of 2017, a caregiver is defined as an adult family member or other individual who has a significant relationship with, and who provides a broad range of assistance to, an individual with a chronic or other health condition, disability, or functional limitation. This includes family members, friends, or neighbors who provide unpaid assistance to a person with a chronic illness or disabling condition.

CTS will be reimbursed when provided by qualified practitioners, including but not limited to: physicians, nurse practitioners, clinical nurse specialists, certified nurse-midwives, physician assistants, clinical psychologists, and therapists (to include physical therapists, occupational therapists, and speech-language pathologists).

CTS is covered for patients under an individualized treatment plan or therapy plan of care, even when the patient is not present. Patient or representative consent is required for the caregiver to receive CTS, and this consent must be documented in the patient’s medical record.

The CPT codes for CTS include the following:

• 96202: Initial 60 minutes of multiple-family group behavior management/modification training for parent(s)/guardian(s)/caregiver(s) of patients with a mental or physical health diagnosis, administered by a qualified healthcare professional, without the patient present;

• 96203: Each additional 15 minutes of multiple-family group behavior management/modification training;

• 97550: Initial 30 minutes of caregiver training in strategies and techniques to facilitate the patient’s functional performance in the home or community;

• 97551: Each additional 15 minutes of caregiver training; and

• 97552: Group caregiver training in strategies and techniques to facilitate the patient’s functional performance in the home or community, with multiple sets of caregivers.

It is important to note that the CPT codes are broken down into appropriate group codes or individual codes based on the number of patients represented for caregiver training.

For instance, if the clinician is working with a patient’s daughter who is providing in-home support to her father with dementia, and the clinician and the daughter discuss the diagnosis and regularly review care plan needs and medication concerns to support the patient’s health condition, then CPT codes 97550 would be appropriate. However, if the clinician is hosting a dementia education and support group with caregivers, the code would be 96202.

It is important to note that if multiple caregivers are present for the same patient in the group setting, the time is still the same for the individual patient. Thus, the codes are based on training time related to the patient, not the number of caregivers.

The Centers for Medicare & Medicaid Services (CMS) provides these coding guidelines under Health Equity Services, to support providers for the time they are spending not only to care for patients, but also to support their natural support system. CTS provides holistic support and recognizes the time providers spend talking to family members and caregivers regarding their patients. 

Healthcare Coverage Expanded to DACA Recipients

By Tiffany Ferguson, LMSW, CMAC, ACM

In a significant move aimed at enhancing healthcare access for Deferred Action for Childhood Arrivals (DACA) recipients, the Biden-Harris Administration has finalized a rule that expands eligibility for health coverage under the Patient Protection and Affordable Care Act (PPACA) and Basic Health Program (BHP).

DACA is a program enacted in 2012 that provides relief from deportation and work authorization for immigrants brought to the United States as children. This policy shift, announced by the U.S. Department of Health and Human Services (HHS) through the Centers for Medicare & Medicaid Services (CMS), represents a step in the right direction toward ensuring equitable healthcare for residents in the United States.

Specifically, the ruling makes a minor but important change to the definition of “lawfully present,” which is a requirement for eligibility enrollment in a qualified health plan. The current definition of “lawfully present” does include certain immigration classifications, such as Green Card holders, Cuban and Haitian entrants, and so many other classifications; however, it specifically excludes DACA. With this modification, DACA recipients will no longer be excluded from that definition, making it possible to participate in the next enrollment period starting Nov. 1, 2024.  

It is important to note that the change of definition for eligibility under “lawfully present” does not apply to Medicaid or the Children’s Health Insurance Program (CHIP) at this time. The current eligibility for Medicaid and CHIP allows states to include coverage for children and pregnant individuals who are lawfully residing in the United States, including those within their first five years of having certain legal status.  

To fill the gap, the PPACA will allow noncitizens who are ineligible for Medicaid because of their immigration status to access financial assistance, such as the Premium Tax Credits and Cost Sharing Reductions through a Marketplace plan, even if their income is below 100 percent of the federal poverty level.

This measure is expected to extend health coverage to approximately 100,000 previously uninsured DACA recipients, addressing a pressing need within the community. The news briefing on this change from CMS stresses the belief that healthcare coverage is a right, not a privilege.

CMS expressed a commitment to providing comprehensive education and technical assistance to support the implementation of this rule, ensuring that immigrants and other communities receive the necessary guidance to navigate the healthcare enrollment process effectively.

CMMI Independence at Home Program Falling Short After 8 Years

By Tiffany Ferguson, LMSW, CMAC, ACM

In late April the Center for Medicare & Medicaid Innovation Center (CMMI) released the evaluation of their Year 8 Independence at Home (IAH) demonstration. IAH is a Congressionally mandated initiative that seeks to evaluate the efficacy of incentivizing home-based primary care, reducing healthcare spending, and enhancing the quality of care for high-cost, high-need Medicare beneficiaries. This evaluation has provided valuable insights into the impact of such interventions, particularly in the context of the COVID-19 pandemic.

The IAH initiative commenced in 2012 with 18 participating practices, aimed at testing whether payment incentives could drive improvements in healthcare outcomes. Over subsequent years, the number of participating practices decreased, with only seven practices remaining by Year 8, amid the challenges posed by the pandemic for in-home services.

Some of the decline in participating primary care clinics is attributed to the program’s strict requirements, such as being available for primary care home visits at all hours, pending patient need, and the requirement to achieve success in cost reduction at least once in three consecutive years. The beneficiary requirements include that Medicare Fee-for-Service (FFS) patients have at least two chronic conditions, require assistance with at least two activities of daily living, and have been hospitalized and received acute or subacute rehabilitation services in the 12 months prior to program enrollment.  Beneficiaries must also not be in any long-term care or hospice program at the time of enrollment in the demonstration.

The evaluation focused on assessing the effects of IAH on total Medicare spending per beneficiary per month (PBPM) and other relevant outcomes, such as the number of ambulatory visits compared to acute-care services the members received during the demonstration period. In Year 8, although there was a potential reduction in total Medicare spending, it was not statistically significant. Notably, the incentive payments made to IAH practices exceeded the estimated spending reduction, raising questions about the cost-effectiveness of the program.

The report noted that while IAH beneficiaries experienced 16 percent more ambulatory visits compared to their counterparts, primary care remained the central service to their healthcare delivery needs. The breakdown of results was mixed: although inpatient spending was down, hospital admissions increased in Year 8, as did readmissions. 

The findings suggest that while the IAH initiative may seem theoretically appropriate to enhance the patient-PCP relationship, the sole mechanism of home-based services did not yield significant results (and likely was near-impossible during the pandemic). In addition, this is likely difficult to scale, given the already known shortage of primary care physicians and the efficiencies that telemedicine can provide, which are outside model expectations.

Additionally, the role of extenders to support home-based services such as chronic care management and community health workers may be better-suited to address in-home care for patients with chronic conditions and in need of home assistance than pulling primary care providers out of the clinic for home-based services.    

In conclusion, this model makes splitting the value-based and FFS payment structures difficult, as program design was incentivizing services on top of a FFS reimbursement structure, rather than a replacement via other capitation or value-based payment methodologies.

Coding Questions Arising From NPAC, Answered

By Erica Remer, MD, FACEP, CCDS, ACPA-C

I attended the 2024 National Physician Advisor Conference (NPAC) hosted by the American College of Physician Advisors (ACPA) in the middle of April. Its theme was “Safeguarding Patients Amidst Shifting Currents of Healthcare.” I love this conference because it is so rich with information and expands my knowledge base beyond clinical documentation integrity (CDI).

I gave an intro to CDI talk at the Essentials and Fundamental course on the first day, where we trained new physician advisors (PAs) and prepared veterans to sit for the ACPA-C. This is a new certification developed by the premier physician advisor organization to signify that an individual has the working knowledge to serve as a PA.

I also had the honor of kicking off the second day with a plenary session called Physician Advisors: Lifeguards Innovating Solutions to Diverse CDI Problems, in which I shared some details about projects I implemented while I was a physician advisor. I received some questions after the fact, and thought this might be a good venue in which to answer them.

How can you get organizations to move beyond only using MS-DRGs as the tool of risk-adjusting metrics?

Originally, CDI was meant to seek out and capture comorbid conditions and complications (CCs) and major CCs (MCCs), thus optimizing each MS-DRG and tier and improving reimbursement and length of stay. When the All-Patient Refined (APR)-DRG methodology was introduced, optimizing the severity of illness (SOI) and risk of mortality (ROM) metrics was a companion goal.

Then came value-based purchasing and other related quality metrics, again establishing other risk-adjustment-type models. And we can’t forget the Hierarchical Condition Categories (HCC) model for Medicare Advantage (MA) and other Accountable Care Organizations (ACOs).

If an organization is only looking at MS-DRGs, they are missing opportunities in multiple other risk-adjusted models. It’s a good place to start for burgeoning CDI programs, but it is only a jumping-off point.

My philosophy always has been to encourage providers to deliver excellent medical care, foster excellent documentation, have CDI specialists (CDISs) pick up the slack when needed, and ensure that coders accurately and compliantly code – and your reimbursement and quality metrics will fall where they belong in every model.

This segues into another question asking if I have any comment on documentation for getting out of PSIs (patient safety indicators). I attended two great talks on the topic of PSIs and quality metrics, by my Vice Chairs (Waldo Herrera-Novey and Adriane Martin) and two other members of my committee (Neelima Divakaran and Joey Cristiano). We all agree: We want folks not to get dinged when the problem is not really a post-procedural condition, and to address medical quality issues if it is. I never condone doing excessive contortions to avoid legitimate post-operative complications. Providers need to be taught that post-op is not a temporal reference, and to ensure that proper linkage is present. If the acute hypoxic respiratory failure is due to an exacerbation of underlying chronic obstructive pulmonary disease (COPD), say that, not “post-operative respiratory failure.”

Another PA asked if I could suggest resources through which you can access how CC/MCCs have financial impact on reimbursement. MS-DRGs are tiered by whether there are CCs or MCCs. Having a single CC is equal to having five of them. The relative weight is a measure of the average resource consumption, severity, and complexity of the patient. When you multiply the relative weight by the hospital’s blended base rate, you determine the reimbursement. If you refer to Table 5 of the current fiscal year’s Inpatient Prospective Payment System (IPPS) Final Rule (https://www.cms.gov/medicare/payment/prospective-payment-systems/acute-inpatient-pps/fy-2024-ipps-final-rule-home-page), you can find the specific DRGs’ relative weights. You can see the difference between a base DRG without a CC or MCC versus with one. You have to ask your administration for your hospital’s base rate.

Our health system has chosen to use Sepsis-2 as our criteria. Are we wrong? and Why not use severe sepsis if it’s on the list of CMS-approved codes and there’s a recognized definition?

First, let me get the terminology straight. SEP-1 is the terminology that indicates The Severe Sepsis and Septic Shock Management Bundle, which is the National Quality Forum sepsis quality measure that the Centers for Medicare & Medicaid Services (CMS) use. Its aim is to standardize sepsis treatment to reduce morbidity and mortality from the condition.

(To find the most up-to-date version, search for Specifications Manual for National Hospital Inpatient Quality Measures and the year of interest. It is listed by quarter. For Q3 and Q4 2024: https://qualitynet.cms.gov/files/65972a96d4b704001df0ae89?filename=HIQR_SpecsMan_v5.16.zip)

In the 1990s, esteemed medical groups came together to try to define the condition of sepsis and operationalize its treatment. Multiple iterations of the literature came out every few years, but Sepsis-2 (not SEP-2) was essentially defined as a presumed or confirmed infection with some systemic inflammatory response syndrome (SIRS) criteria. Although there were always other clinical indicators that could be seen in sepsis, the medical community gravitated towards using the general SIRS criteria, which were ubiquitously collected in all patients in whom the diagnosis was being considered. These were:

• Fever or hypothermia;

• Tachycardia;

• Tachypnea; and

• Elevated or low white blood cell count.

In Sepsis-2, there were three buckets of sepsis: sepsis as defined above, severe sepsis, which was sepsis with organ dysfunction, and septic shock, which was severe sepsis with hypotension not reversed with fluid resuscitation. ICD-10-CM has codes to indicate all three gradations.

Eventually, it was recognized that this was too broad a definition – there were many patients with infections who exhibited SIRS criteria but did not have sepsis, and some patients who had sepsis but didn’t demonstrate these particular signs.

Hence Sepsis-3. In 2016, The Third International Consensus Definitions for Sepsis and Septic Shock (Sepsis-3) were rolled out. The definition was established as “life-threatening organ dysfunction caused by a dysregulated host response to infection.” It was recognized that SIRS was sometimes an appropriate response to infection, and therefore, might not indicate sepsis. Furthermore, the presence of organ dysfunction is the critical ingredient for the diagnosis of sepsis.

This contracted the buckets to two: sepsis (the condition formerly known as severe sepsis, which includes organ dysfunction) and septic shock. ICD-10-CM did not change the coding schema. As a result, all sepsis now should have a code for the sepsis, another code indicating the presence of organ dysfunction, with or without shock, and additional codes for the causative infection and the resultant organ dysfunction.

In early 2017, Surviving Sepsis Campaign adopted the Sepsis-3 definition. They concurred with the International Consensus folks that “The current use of two or more SIRS criteria to identify sepsis was unanimously considered by the task force to be unhelpful” because “SIRS criteria do not necessarily indicate a dysregulated, life-threatening response.”

During the reign of Sepsis-2, many organizations put in place an alert system, and they utilized those basic SIRS components as the trigger. It is still useful as a warning signal, but SIRS does not establish sepsis. Many providers are resistant to change and cling to SIRS. I am supportive of using SIRS to identify patients who might have sepsis, but I strongly caution against using it as the diagnostic criteria.

The question I would ask the first organization is: do they want to catch in their net patients who do not end up having sepsis? Identify potential sepsis patients preliminarily and then have the clinicians rule it out over the course of the encounter. Then, the diagnosis should be removed before being coded and final-billed.

If you don’t, you frontload the system with patients who are erroneously assigned into a sepsis DRG, and you will be inundated with clinical validation denials on the back end. Not only that, but the U.S. Department of Health and Human Services (HHS) Office of Inspector General (OIG) has added Medicare inpatient hospital billing for sepsis to their workplan.

Is it wrong to use Sepsis-2 as your criteria? I would not recommend it. The current definition has been around for seven years. It’s time to transition.

Why shouldn’t providers use “severe sepsis” if it is on the list of CMS-approved codes and there’s a recognized definition? I don’t believe clinicians should alter their practice of medicine to satisfy coding. They may need to include some antiquated terminology, however, for the coders to capture legitimate conditions, due to the peculiarities of coding rules. Fortunately, there is coding guidance instructing coders that if providers diagnose sepsis with related organ dysfunction, the coder may compliantly pick up R65.20, Severe sepsis without septic shock, without the doctor using the obsolete qualifier “severe.” The condition formerly known as “severe sepsis” is now just “sepsis.”

Consider inserting my macro into your electronic health record (EHR):

Sepsis due to (infection) with acute sepsis-related organ dysfunction, as evidenced by (organ dysfunction/s).

This allows the clinician to avoid having to use old terminology, with the bonus of helping them ensure that the condition is indeed present and clinically valid.

The last question was, “How do you message the importance of thorough and complete documentation when physicians/providers are thinking the E&M 2023 changes have now relieved them of ‘note bloat?’”

I don’t endorse conveying the message that documentation is for billing. I believe thorough and complete documentation is important to tell the story of each patient encounter.

The professional fee Evaluation and Management (E&M) changes made in 2023, when Inpatient and Observation Hospital Care merged, are only based on medical decision-making (MDM) or time, which now should reduce note bloat. The history and physical examination should be performed and documented “as medically appropriate.” No more gratuitous review of systems just to check a billing box. There is no reason to embed a radiology report in the progress note at all, let alone every day of the admission. The provider only gets credit for it once. Rather than copying and pasting the assessment and plan daily, maybe spend a few moments thinking about what needs to be in the note today and how it informed your decisions.

Reducing note bloat means streamlining the documentation to make it really count. Does “no acute events overnight, no acute distress this morning, reportedly at baseline” explain why the patient in Room 204 is still here today? Does a list of signs and symptoms relay what your impression is? Does a hodgepodge of copied and pasted paragraphs really constitute a discharge summary? Remember that your “copying and pasting saves me time” is someone else’s “I hate slogging through other people’s copying and pasting to find the essential points.”

It is crucial to not buy into the “documentation is an unwanted burden on the provider” nonsense. Documentation is part of the process of delivering quality medical care to the patient and enabling others to recognize it. If providers put “mentation” into their documentation, the medical record and the patient will be better off for it.

Who Can Perform Clinical Validation?

By Erica Remer, MD, FACEP, CCDS, ACPA-C

There has been a kerfuffle on LinkedIn I would like to expound upon today. A colleague of mine, Siraj Khatib, was recently expressing his exasperation at clinical validation audits.

He referred to the Recovery Audit Program Statement of Work (SOW). In 2011, there was a paragraph in the section titled DRG Validation vs. Clinical Validation, which read, “clinical validation is a separate process, which involves a clinical review of the case to see whether or not the patient truly possesses the conditions that were documented. Clinical validation is beyond the scope of DRG (coding) validation, and the skills of a certified coder. This type of review can only be performed by a clinician or may be performed by a clinician with approved coding credentials.”

Dr. Khatib went on to posit that “only a bedside clinician is able to understand, interpret, (and) judge a patient’s clinical condition,” and pondered how “a medical director can have clinical acumen … (having) relinquished bedside medical practice for the comfort of an office, awaiting a bonus to churn out denials.”

There is a lot to unpack here.

First, I would like to validate the frustration of incurring unfounded denials. It requires a great deal of time, effort, and money to combat denials. Sometimes, there is legitimacy – the medicine is questionable, or the documentation is weak. Often, it is infuriating – the medicine is solid, the documentation impeccable, and the insurer is just throwing spaghetti against the wall to see what sticks.

Who is qualified to perform clinical validation, determining if a patient possesses a condition that is documented? The only one who truly can perform clinical validation is a clinician who is taking care of the patient. Did the patient have an exacerbation of their chronic systolic heart failure or not?

However, that is not how the system works. People who are not responsible for the patient’s medical care review the documentation and must make determinations based on the way the encounter is portrayed in the medical record. That is referred to as “clinical validation.”

Who is sufficiently competent to perform this role? The 2011 SOW expressed the decision of the Centers for Medicare & Medicaid Services (CMS) that they were going to only permit clinicians to make clinical validation determinations. I have always asserted that their decision to only utilize clinicians, such as nurses or physicians, is not a universal mandate. It was the stated practice of CMS.

Two of the best clinical documentation integrity specialists (CDISs) I have ever known were non-clinicians, and came from the health information management (HIM) world (you know who you are, Colleen and Kathy!) I do not think it is out of the question that someone from HIM would be capable of doing clinical validation, but I do contend that not all HIM individuals would be able to do so. They need to have experience in the clinical setting and long-term exposure to the medical record. Although clinicians (e.g., nurses, advanced practice practitioners, foreign medical graduates, etc.) are more commonly employed as CDISs, knowledgeable HIM folks are allowed to be CDISs. If they can serve as a CDIS, they can perform the task of clinical validation.

Institutions and systems can make the determination as to whom they deem competent to perform clinical validation. They do not need to insist on clinician credentials. By the same token, commercial payors have the latitude to make this same decision for their organization.

I completely missed the memo about this: in 2017, CMS revised its SOW to read, “clinical validation is prohibited in all RAC (Recovery Audit Contractor) reviews.” CMS no longer specifies that a clinician must perform clinical validation; they say that they are no longer doing it at all. I’m not sure that all the RACs got that memo, either!

But I want to address Siraj’s last contention. When I became a physician advisor for a large multi-hospital system, the system chief medical officer (CMO) advised me to continue practicing clinically. As an obstetrician, he missed operating. I, on the other hand, believed that for the safety of my patients, there was a minimum threshold of hours to remain clinically competent, especially in terms of procedures. If a patient needed an emergency thoracotomy or tracheostomy, I was not the right person for that job. However, I was really good at my non-clinical physician advisor job, despite no longer practicing at the bedside. I am really good at deciphering documentation, and I have 25 years of clinical experience to back it up. I am more than capable of determining medical necessity and quality of care from the medical record, without seeing patients on a daily basis. I suspect that medical directors in insurance companies also have years of experience behind them.

That is not really the fundamental issue.

How people wield their knowledge and generate denials is the problem. Having artificial intelligence (AI) generate a zillion denials in a matter of seconds is a problem. Working on contingency, whereby throwing spaghetti on the wall is profitable, is highly questionable. Rejecting an appeal without weighing its merits is an issue. Having a bonus based on productivity and not on merit is a problem.

One of my superpowers is being able to see things from all sides. I believe the system as it is designed is important, with checks and balances. Providers shouldn’t be able to engage in fraud and abuse, and payors should have to pay for services legitimately rendered without throwing up roadblocks. The government is the biggest payor, and they get their money from me and you. We don’t want them to be squandering our taxpayer dollars, but we don’t want our hospitals to go bankrupt fighting ridiculous denials, either.

If providers deliver excellent medical care and document their thought process well, then payors should pay for medically necessary care of their beneficiaries. If any of these elements is not present, there should be consequences. Clinical validation is one cog in that process, and you don’t have to be an actively practicing practitioner to make that judgment.

Pediatric Sepsis Defined: Experts Endorse Organ Dysfunction

By Erica E. Remer, MD, FACEP, CCDS, ACPA-C

Those of you who live in the adult world may not even be aware that the sepsis conundrum (i.e., Sepsis-2 vs. Sepsis-3) didn’t really pertain to pediatrics. That matter was tabled whilst the so-called grown-ups squabbled. Their previous criteria hailed from 2005, specifically from the International Pediatric Sepsis Consensus Conference. But it was already clear then that sepsis needed to be updated and better defined for the pediatric population as well, so an international consensus panel was convened in 2019, and the results were published online on Jan. 21, 2024.

The International Consensus Criteria for Pediatric Sepsis and Septic Shock (hereafter referred to as the International Consensus) were derived by 35 clinical experts from diverse pediatric practice in 12 different countries. The first thing that struck me was that they refer to a sentinel article titled Guidance for Modifying the Definition of Diseases: A Checklist, published in the Journal of the American Medical Association (JAMA) Internal Medicine in July 2017. It didn’t occur to me that there is a process to make adjustments in disease definitions. I wonder if it was crafted in response to the mess that arose after Sepsis-3 was introduced. The article explains that there is a fine line to walk between over-diagnosing conditions and capturing them early and often, versus harming patients by not entertaining a diagnosis quickly enough.

Something specific in that article struck a chord with me: “Diseases do not generally have discrete boundaries, and (my inserted word: clinical) judgment is required to determine the thresholds for diagnoses.” It seems as though it is more prudent to have many multidisciplinary representatives with expertise in the condition being addressed, making thoughtful decisions about diagnostic criteria than it being the Wild West, where everyone uses whatever criteria suit them that day – and for their own purposes.

The approach they took for the pediatric sepsis criteria was first to administer a global survey of 2,835 clinicians, published in The Current and Future State of Pediatric Sepsis Definitions: An International Survey, asking what the condition sepsis constituted and what the word should mean. I found this curious until I realized before there was even a Sepsis-1, we used to diagnose sepsis by gestalt. You walked in a room and said, “Uh-oh, this patient is really sick!”

The ultimate conclusion, supported by the majority of respondents, was that sepsis is an “infection with associated organ dysfunction.” The International Consensus authors noted that this was preferable to “infection-associated SIRS (Systemic Inflammatory Response Syndrome),” and that this evolved definition indicates “widespread adoption of the Sepsis-3 conceptual framework.”

Next, they undertook a systematic review of more than 3 million pediatric encounters, letting that inform the design of a derivation and validation study. The original concept was an eight-organ system model, but they ultimately dropped the renal, hepatic, endocrine, and immunological dysfunction criteria, seemingly to reduce “requirements for laboratory investigation and data collection.” I watched a presentation by the Children’s Hospital Association, and their commentary was that isolated hepatic, renal, endocrine, or immunological dysfunction was unusual, and it was far more common to be seen in conjunction with dysfunction of one of the other included systems, thus rendering those systems redundant.

The International Consensus settled on a composite four-organ system model, which was coined the Phoenix Sepsis Score (PSS). It seems as though the PSS was originally intended to be a prognosticating tool for mortality, similar to the Sequential Organ Failure Assessment (SOFA) score, but ultimately, attaining a score of 2 or more on the PSS was defined as identifying sepsis in an “unwell child with suspected infection.” Accruing ≥ 1 point from cardiovascular dysfunction establishes septic shock.

The PSS includes respiratory, cardiovascular (CV) (with age-based mean arterial pressure values), coagulation, and neurological systems. They are variably weighted, with CV potentially offering up to 6 points, respiratory up to 3 points, and the other two only up to 2 points. The score was designed to be able to be utilized even in low-resource areas.

There are a few more points regarding the International Consensus:

• If a child manifests organ dysfunction remote from the site of localized infection, they are recognized as being at higher risk of mortality than if they only have a localized, single-organ system impairment (e.g., respiratory failure in pneumonia).

• The PSS is assessed in the first 24 hours of presentation to the hospital. The article notes that the PSS “is not intended for early screening or recognition of possible sepsis and management before organ dysfunction is overt.”

• Due to the difficulty of defining organ dysfunction in neonates born < 37 weeks’ gestation and the contribution of perinatally acquired infection, term newborns remaining in the hospital after delivery and neonates whose postconceptional age is younger than 37 weeks are excluded.

Here are some of my concerns regarding the International Consensus:

• They use mortality as their only endpoint (morbidity is also a major concern post-sepsis);

• They limited the development of the PSS to data from the first 24 hours of hospitalization (again, with the primary endpoint of predicting risk of mortality). Clearly, children with infections and organ dysfunction discovered later on, or who acquire infections during the encounter, can suffer from sepsis. I asked the corresponding author, and he asserted that they expect the PSS to perform similarly whenever during the hospitalization the condition crops up; and

• If the Phoenix-8 score had comparable performance to the Phoenix-4 score (PSS), shouldn’t we diagnose sepsis if a patient has an infection and organ dysfunction not included in the PSS (e.g., hepatic or renal failure)? I’m going to hope that this is a rare occurrence, but what is the role for clinician judgment? Are payors going to deny claims of sepsis even if the clinician believes there is life-threatening organ dysfunction of a non-Phoenix-4 organ system?

This International Consensus statement certainly demonstrated rigor in development. The most important sentence to me in the paper is “SIRS should no longer be used to diagnose sepsis in children, and because any life-threatening condition is severe, the term severe sepsis is redundant.”

Fortunately, we have guidance permitting the capture of the code for severe sepsis if organ dysfunction is linked to the sepsis, even if the provider doesn’t document the word “severe.” Until the World Health Organization (WHO) and Centers for Disease Control and Prevention (CDC) eliminate the ICD-10/-CM code for severe sepsis without shock (R65.20), all patients with sepsis and sepsis-related organ dysfunction should receive at least four codes:

• A code indicating sepsis (e.g., unspecified or organism-specific sepsis; perinatal, obstetrical, or postprocedural sepsis);

• An R65.2- code indicating severe sepsis without or with septic shock;

• A code specifying the underlying localized infection that is the source of the sepsis; and

• At least one code detailing the organ dysfunction.

Pediatric experts have expressed their agreement: there is no such thing as sepsis without organ dysfunction.

Now, if we can only get the adult practitioners to buy in, too.

OIG Sets Sepsis in its Sights

By Erica Remer, DM, FACEP, CCDS, ACPA-C

I am going to begin a two-part look at sepsis, starting with the U.S. Department of Health and Human Services (HHS) Office of Inspector General (OIG) focus on Medicare Inpatient Hospital Billing for Sepsis, brought to our attention by Dr. Ronald Hirsch; next time, I am going to write about updates to pediatric sepsis.

The introduction to the OIG’s plan to analyze Medicare claims for sepsis says some very impactful things. It asserts that “sepsis is the body’s extreme response to an infection,” and that “it is a life-threatening, emergency medical issue that often progresses quickly and responds best to early intervention.” It acknowledges that “the definition of and guidance for sepsis have changed over the years,” in an attempt to capture sepsis better. It identifies the issue that the Centers for Medicare & Medicaid Services (CMS) and the Centers for Disease Control and Prevention (CDC) “currently recognize an older, broader definition,” and the OIG expresses a concern that hospitals may take advantage of the broader definition because they are financially incentivized to land patients in the relatively higher-weighted sepsis Medicare-Severity Diagnosis Related Groups (MS-DRG).

Their study will analyze patterns in inpatient hospital billing for 2023 and assess the variability of sepsis billing among hospitals. They plan to compare costs using the broader definition (i.e., Sepsis-2, according to SIRS, or systemic inflammatory response syndrome, criteria) versus the narrower definitions of sepsis, that is, Sepsis-3.

On Feb. 23, 2016, The Third International Consensus Definition for Sepsis and Septic Shock (Sepsis-3) was published in the Journal of the American Medical Association (JAMA), written by Mervyn Singer, Clifford Deutschman, et al. Surviving Sepsis Campaign followed up in March 2017, with their acceptance of the definition as “life-threatening organ dysfunction caused by a dysregulated host response to infection.” Whereas Sepsis-3 defined the condition, Surviving Sepsis Campaign operationalized how to treat sepsis, issuing best-practice statements and utilizing the GRADE (Grading of Recommendations Assessment, Development, and Evaluation) system. Prior to 2016, Surviving Sepsis Campaign (SSC) was where we had derived Sepsis-2.

This was over seven years ago. Why are we still adjudicating this? If the sentinel organizations use the same definition, why don’t all hospitals everywhere use it? Why hasn’t SIRS been put to bed?

There are multiple reasons, and the reason the OIG is leveraging is the money. The relative weight (RW) for sepsis without mechanical ventilation > 96 hours with a major comorbid condition (MS-DRG 871) in 2024 is 1.9826, and its “without MCC” counterpart (MS-DRG 872) has a RW of 1.0299. For comparison, Respiratory Infections and Inflammations with MCC (MS-DRG 177) has a RW of 1.6964, with CC (MS-DRG 178) is 0.9867, and MS-DRG 179 without CC or MCC has a RW of 0.7633. The MS-DRG for urinary tract infections w/wo MCC (MS-DRGs 689 and 690) have RWs, respectively, of 1.1744 and 0.8069. Hence, the most profitable DRG for a patient who is admitted with an infection is in the sepsis set.

What other reasons enter into this persistent utilization of SIRS? Clinician ignorance or clinicians clinging to “the way it has always been done” are obvious factors. The fact that the core measures bundle and New York has its own criteria that don’t align with either Sepsis-2 or Sepsis-3 are others. There are also practitioners who, out of an abundance of caution, would rather err on the side of picking up “early sepsis” than missing the boat and having a patient die, so they would rather liberalize the criteria to catch cases that turn out not to be sepsis. I am supportive of making a tentative diagnosis early, but the key to compliance is to remove the diagnosis once it has been ruled out.

We all designed our sepsis alerts to use the general variable SIRS criteria because it was easy, convenient, and ubiquitous, but the diagnosis of sepsis always included other clinical indicators. For instance, altered mental status, hyperbilirubinemia, thrombocytopenia, and coagulopathy were included in the Surviving Sepsis Campaign 2012 table of diagnostic criteria for sepsis. But everyone gets vital signs taken and a white blood cell (WBC) count drawn if they have a potential infection, so those were attractive as a screening diagnostic tool.

SIRS is a great marker for clinically significant disease; however, it is very non-specific. Conditions not infectious in etiology may demonstrate tachycardia, tachypnea, fever, or elevated WBCs. Patients with infections may demonstrate those symptoms without having progressed to sepsis. It may represent an appropriate response to the infection.

I once was rounding with a provider who documented sepsis as her clinical impression on a patient who was on the fourth day of their admission and was sitting in bed smiling and eating a sandwich. I asked the provider if the patient met the criteria of sepsis – were they “sick” with a capital S? The provider replied that this wasn’t part of the definition. I disagreed, saying it was so integral to the definition of sepsis that the experts didn’t think they needed to explicitly say it. It is my opinion that the indication of being “sick” with a capital S is organ dysfunction.

I had a reader ask me once why I don’t want providers to diagnose “sepsis without organ dysfunction. Don’t you think that it is better to catch it early than to miss it?” My response was that patients who have infections should be treated aggressively and appropriately, whether or not they have sepsis. If you nip the infection in the bud and avert the development of sepsis, good for you!

Dr. Hirsch used a great analogy that I would like to reuse. He said lots of patients have chest pain without enzyme markers for heart attack. They will be monitored and might be catheterized and stented. We don’t make the diagnosis of impending myocardial infarction (MI) and get paid in an MI DRG.

When I review records in the context of clinical validation denials, invariably, most of the cases I find righteously denied are billed as sepsis. Sepsis without organ dysfunction is…pneumonia or urinary tract infection or cellulitis. It doesn’t belong in the sepsis DRG, and I am going to predict that the OIG is going to agree with me.

If your hospital still uses SIRS criteria, it’s time for them to transition to Sepsis-3. It’s time for the state of New York to transition to Sepsis-3. If your institution uses Sepsis-3, but your providers document poorly or inconsistently, they should be educated and monitored.

Clinical documentation integrity specialists (CDISs) should put a program in place to perform clinical validation of the diagnosis of sepsis. Clinical validation denials are predictable, and a pain, but nothing compared to an unfavorable OIG determination.