medwireNews: Over half of patients hospitalized for pneumonia are given a different diagnosis on initial presentation compared with their diagnosis at discharge, show data from 118 Veterans Affairs Hospitals in the USA.
Furthermore, a third of pneumonia patients with positive chest imaging plus a diagnosis of the condition at discharge lacked both an initial diagnosis and treatment, indicate the findings.
Patient clinical notes echo treatment ambiguity, with 10% of patients treated potentially unnecessarily with a combination of antibiotics, corticosteroids, and diuretics for diagnoses that “mimic or accompany” pneumonia, such as cardiac disease, sepsis, and chronic obstructive pulmonary disease, report the researchers.
“These analyses [suggest] that initial pneumonia diagnosis is a complex process with substantial uncertainty that may previously have been underappreciated,” remark Barbara Jones, from the University of Utah and Salt Lake City VA Healthcare System in the USA, and colleagues.
“‘Overdiagnosis’ and treatment ambiguity can have important consequences on patient outcomes, including adverse effects of unnecessary antibiotics,” they write in the Annals of Internal Medicine.
The team examined “diagnostic concordance” in 317,437 patients (96.0% men, average age 72.1 years) who were hospitalized with a pneumonia diagnosis between January 2015 and January 2022. The researchers collated each patient’s diagnosis at three specific timepoints: the initial emergency department (ED) encounter, the initial chest imaging, and discharge.
Pneumonia diagnoses were defined using natural language processing of clinician notes, diagnostic coding, and whether the patient received treatment with antibiotics or antivirals.
Just 38.1% of the entire cohort had a positive pneumonia diagnosis at all three timepoints, while 56.7% showed discordance between the diagnosis given during initial ED presentation and then at discharge. And of those diagnosed with pneumonia initially, 36.3% did not have a pneumonia diagnosis at discharge and 21.0% did not have positive initial chest imaging.
Among those whose discharge diagnosis of pneumonia matched their positive initial chest image showing the condition, 33.4% were not initially diagnosed with pneumonia, nor did they receive any treatment for pneumonia.
These patients had the highest 30-day mortality rates, at 14.4% versus 10.6% for those whose diagnosis timepoints were fully concordant, 6.8% for those who lacked diagnosis by chest imaging, and 11.0% for those who had a positive diagnosis in the ED but lacked a positive discharge diagnosis.
Jones and colleagues identified uncertainty and ambiguity in 58% of patients’ clinical notes from the ED, and in 49% of discharge summaries. Patients with any diagnostic discord had more frequent “assertions of uncertainty” from clinicians than those with concordant diagnoses, as well as higher rates of heart failure, renal disease, and receipt of treatment with corticosteroids and diuretics as well as antibiotics.
Overall, 89.7% of the cohort received antibiotics within 24 hours of presenting to the ED and 8.7% an antiviral. In addition, 27.0% were given diuretics and 36.0% corticosteroids.
In an accompanying editorial, Mark Metersky (University of Connecticut School of Medicine, Farmington, USA) and Grant Waterer (The University of Western Australia and Curtin University, Perth) note that discordant diagnoses may not be inaccurate, since the signs and symptoms of pneumonia are often nonspecific and overlap with other conditions.
“A patient may have both pneumonia and [chronic obstructive pulmonary disease] or heart failure, and, at the time of discharge, either diagnosis could be accurately selected,” they write.
The pair share the concerns of Jones et al, that some patients are harmed by receiving unnecessary treatment, or indeed by not receiving necessary treatment.
This “validates the long-standing clinical approach of treating once a threshold of suspicion is reached, rather than requiring absolute confirmation of the diagnosis,” they conclude.
medwireNews is an independent medical news service provided by Springer Healthcare Ltd. © 2024 Springer Healthcare Ltd, part of the Springer Nature Group
Ann Intern Med 2024; doi:10.7326/M23-2505
Ann Intern Med 2024; doi:10.7326/M24-0889
