When I ask facilities what their most common denials are, invariably, pneumonia makes the list. That was my personal experience when I handled clinical validation denials as a physician advisor. Most of them I deemed appropriate denials and declined to appeal. Recently, I’ve been working on a project where I’ve been assessing encounters resulting in DRG (Diagnosis-Related Group) downgrades, and I would like to specifically address pneumonia today.
The typical scenario is a patient with underlying comorbidities of heart failure or chronic obstructive pulmonary disease (COPD) and presents to the emergency department complaining of shortness of breath and cough. The patient may have no or low-grade fever. The chest X-ray is not compelling – no distinct infiltrate is seen. The radiologist may be hedgy, noting markings that “could be pneumonitis or pulmonary edema or atelectasis; clinical correlation necessary.”
The emergency physician or hospitalist empirically treats for pneumonia and makes an uncertain diagnosis (e.g., possible, probable, likely). The issue arises when the condition is neither definitively ruled in, during the course of the encounter nor conclusively ruled out. The uncertain diagnosis is propagated through copying and pasting. There are also concomitant issues, with hypoxia being labeled acute hypoxic respiratory failure, or sepsis being diagnosed without evidence of organ dysfunction.
As a reviewer, I am able to take in the entire admission through the lens of the retro-spectoscope. At the end of the stay, it may be that much easier to ascertain that the entire clinical presentation was due to acute-on-chronic heart failure (HF) due to fluid overload (e.g., missed dialysis) or exacerbation of COPD from viral bronchitis. The definition of principal diagnosis, which establishes the DRG, is “that condition which, after study, was chiefly responsible for occasioning the admission of the patient to the hospital for care.” If I can tell that the derangement causing the hospitalization was the HF or COPD, and I doubt the validity of the pneumonia, so can the payor, and they will want to revise the DRG accordingly.
The initial providers are not going to know how the encounter is going to play out. It is reasonable to empirically treat for pneumonia. The error occurs when the subsequent providers fail to evolve, resolve, remove, and recap. The diagnosis should evolve as more information becomes available, like culture results. Resolving is documenting that the condition has been successfully managed and is no longer an active problem. “Remove” is what should be done when a condition has been ruled out. All clinically significant diagnoses should be recapped in the discharge summary to encapsulate the hospital course and tell the story of the encounter.
Elements that imperil the diagnosis of pneumonia are:
- Lack of fever or elevated white blood cell count, normal vital signs; no evidence of the patient being in any respiratory distress, or appearing ill or toxic;
- No infiltrate ever materializes on either chest X-ray or CT scan;
- The provider hesitantly diagnoses pneumonia in the setting of organ dysfunction (e.g., metabolic encephalopathy, acute hypoxic respiratory failure, acute kidney injury, Type 2 myocardial infarction), but does not diagnose sepsis;
- Consultants such as pulmonologist or infectious disease don’t make a diagnosis of pneumonia in the same patient with the same information;
- The assessment and plan (A&P) is unchanging, and the diagnosis remains uncertain; and
- Pneumonia is absent from the discharge summary.
Here is what the provider can do to prevent pneumonia denial jeopardy:
- Paint the picture of a patient with pneumonia – providers should call out fever, tachypnea, increased shortness of breath, chest pain, and/or hypoxemia, and link the signs/symptoms with the pneumonia, especially if there are other pulmonary confounding diagnoses;
- Rethink the diagnosis if there are no radiological findings suggestive of pneumonia. It is hard to sell pneumonia to a payor without an infiltrate;
- As the encounter progresses in a patient with an uncertain diagnosis of pneumonia, either definitively rule in, rule out, or intentionally maintain uncertainty. Each day, you should have more information and data on which to base your diagnosis;
- Link pneumonia with acute hypoxic respiratory failure or with sepsis if those diagnoses are concurrent. If pneumonia gets ruled out but the provider is retaining sepsis as a diagnosis, ensure there is an alternate infection as a source;
- Consider the consultants’ opinions. Either agree and incorporate their diagnoses and specificity into your documentation, or disagree and document your rationale. Discuss it with them so everyone can be on the same page while taking care of the patient. Avert internal inconsistency;
- If pneumonia is ruled out, document the reasoning, and declare it ruled out. The next day, pneumonia should no longer be on the impression list. Don’t just drop it without explanation. It will not be clear whether it was ruled out or just accidentally forgotten; and
- The discharge summary should tell the story of the patient encounter. If they really had pneumonia, it should appear in the list of discharge diagnoses. If the course of antibiotics is not complete, prescribe it in the discharge instructions.
One of the most important things to do is to perform clinical validation prior to final coding and billing. Make sure clinical documentation integrity specialists are vigilant when reviewing records of patients with pneumonia, and query when indicated. It may not be feasible to do a mandatory review of all patients with pneumonia, but it may be possible to review all pneumonia from a specific provider who has a history of being lax at diagnosing it.
It is understandable, even desirable, to make the diagnosis of pneumonia preliminarily and have it ruled out over the course of the next few days. However, the provider needs to make sure their documentation tells the story of the patient encounter accurately. An ounce of prevention is worth hours of fighting clinical validation denials!
