Community-Acquired Pneumonia is Being Over-Diagnosed

I spoke about this subject in March (previous article), but during another chart review, I continued 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?

Program note:

Listen to Dr. Erica Remer today as she cohosts Talk Ten Tuesday with Chuck Buck at 10 Eastern.