A new study highlights the uncertainty of pneumonia diagnoses made in emergency departments (EDs) and hospitals.
The study, published yesterday in the Annals of Internal Medicine, found that in more than half of all Veterans Affairs (VA) patients hospitalized and treated for pneumonia over an 8-year period, the diagnosis of pneumonia changed between initial presentation in the ED and hospital discharge. In addition, uncertainty was often reflected in physician notes and prescription of treatments used for other conditions.
The authors of the study say the findings warrant attention from healthcare systems, clinicians, and patients.
Discordant diagnoses
For the study, a team led by researchers from the University of Utah and the Salt Lake City VA Healthcare System analyzed electronic health record data on patient encounters at EDs that resulted in hospitalization at a VA medical center from 2015 through 2022.
They targeted patients hospitalized with a pneumonia diagnosis, using diagnostic codes and an artificial intelligence-based technology called natural language processing to determine whether the diagnosis of pneumonia was made at ED admission, at initial chest imaging, or at discharge from the hospital. Their aim to was determine the frequency of discordant diagnoses, i.e., cases where pneumonia was diagnosed at ED admission but not at discharge, or vice versa.
The also looked for signs of diagnostic uncertainty in the clinical data, including alternative diagnostic codes for other conditions, clinician notes that expressed uncertainty, and tests and treatments used for diagnoses that might mimic or accompany pneumonia, such as cardiac conditions, chronic obstructive pulmonary disease, and asthma exacerbations.
"Pneumonia can seem like a clear-cut diagnosis, but there is actually quite a bit of overlap with other diagnoses that can mimic pneumonia," lead author Barbara Jones, MD, a pulmonary and critical care physician at University of Utah Health, said in a press release.
Of the 2,383,899 hospitalizations at 118 VA medical centers analyzed, 13.3% were diagnosed and treated for pneumonia, with 9.1% receiving an initial pneumonia diagnosis, and 10% receiving a discharge diagnosis. Of the hospitalizations with a diagnosis of pneumonia at either admission or discharge, 38.1% had complete concordance.
But discordance between the initial and discharge diagnosis occurred in 57% of hospitalizations. Among those who were diagnosed with pneumonia initially, 36% did not have a pneumonia diagnosis at discharge, and 21% did not have positive initial chest imaging. Among the patients who were discharged with a pneumonia diagnosis and positive initial chest imaging, 33% were not diagnosed on admission.
Pneumonia can seem like a clear-cut diagnosis, but there is actually quite a bit of overlap with other diagnoses that can mimic pneumonia.
The clinical notes also showed uncertainty, with 58% of ED notes and 49% of discharge summaries expressing uncertainty about the diagnosis. Analysis of treatments showed that 10% received antibiotics, corticosteroids, and diuretics within the first 24 hours, an indication of simultaneous treatments for multiple potential diagnoses. Meanwhile, 22% of patients who lacked an initial diagnosis but were diagnosed with pneumonia at discharge received delayed antibiotic treatment.
Analysis of outcomes showed that patients with diagnostic discordance had greater uncertainty and received more additional treatments. Patients who lacked an initial pneumonia diagnosis and were later discharged with pneumonia had a greater illness severity than those with concordant diagnoses, as reflected by higher intensive care unit admission (20.6% vs 14.8%) and 30-day mortality (14.4% vs 10.6%).
"The sum of these analyses suggests that initial pneumonia diagnosis is a complex process with substantial uncertainty that may previously have been underappreciated," the study authors wrote.
Implications for care
Jones and her colleagues say that while previous research has documented overdiagnosis and misdiagnosis of pneumonia, this is the first study to characterize the frequency of discordance between initial and discharge diagnosis of pneumonia, and it has important implications for current clinical care. They note that most of the evidence that informs pneumonia care comes from studies that identify pneumonia patients using discharge diagnoses.
"If more than a third of patients discharged with pneumonia do not start out with this diagnosis, and another third of patients initially treated for pneumonia are not included in the research, the mismatch between study and practice populations threatens the generalizability of existing observational evidence to real-world pneumonia care," they wrote.
They add that the findings also suggest clinicians and patients should anticipate and respond accordingly to diagnostic uncertainty with pneumonia.
"Both patients and clinicians need to pay attention to their recovery and question the diagnosis if they don't get better with treatment," Jones said.
