Patients with culture-confirmed community-acquired E. coli pneumonia, abbreviated CAECP, had higher mortality within 90 days than patients with community-acquired pneumonia caused by other typical pathogens, the comparison group known as Other-CAP, according to a small retrospective hospital study. Death by 90 days occurred in 49% of the E. coli group, compared with 16% of the comparison group (P=.005).
The association remained after researchers adjusted for age and malignancy. The adjusted odds ratio was 7.413, with a 95% confidence interval from 1.850 to 29.712 (P=.005). An adjusted odds ratio expresses the strength of an association after selected factors are taken into account; the wide confidence interval means its exact size is uncertain. Because the data came from a retrospective observational study, the finding does not establish that E. coli pneumonia caused the deaths.
Similar at the outset
The result stood out because the two groups looked similar on the initial measures reported by the researchers. Demographic, clinical, laboratory and selected radiological presentation measures were comparable, including age, sex distribution, comorbidity burden, body temperature, heart rate, C-reactive protein, procalcitonin, multilobar involvement and air bronchogram findings. None of those reported comparisons was statistically significant.
Two pneumonia severity scores also showed no statistically significant difference. Mean CURB-65 was 1.73 in CAECP and 1.45 in Other-CAP (P=.253). Mean Pneumonia Severity Index scores were 108.6 and 95.4, respectively (P=.143).
The difference emerged during follow-up
The groups separated more clearly in their hospital course. Mean hospital stay was 15.7 days for CAECP patients, compared with 9.6 days for Other-CAP patients (P=.047). Empirical treatment was expanded for 16 CAECP patients, or 43%, compared with 6 patients, or 19%, in the Other-CAP group (P=.036).
That treatment comparison describes a difference in care, not a demonstrated benefit or harm. The study did not establish whether treatment expansion improved outcomes.
In-hospital mortality was 14% in CAECP and 0% in Other-CAP, but the difference was not statistically significant (P=.058). The small number of deaths limits how confidently that in-hospital comparison can be interpreted.
A narrowly defined study population
The study drew on 286 screened files and included 68 patients: 37 with CAECP and 31 with Other-CAP. It was a single-center retrospective observational cohort of adults with community-acquired pneumonia and positive sputum cultures; hospital-acquired pneumonia, ventilator-associated pneumonia and other respiratory infections were excluded.
The researchers did not treat a positive culture as sufficient on its own. E. coli isolation was interpreted together with the clinical and radiological diagnosis of community-acquired pneumonia.
The analysis used a targeted binary logistic model to assess the association between CAECP and 90-day mortality after adjustment for age and malignancy. Malignancy was associated with 90-day mortality in that model, while age was not statistically associated with the outcome.
The authors reported the study's single-center retrospective design, small sample and limited number of mortality events as limitations. The selected hospitalized, culture-positive cohort may not represent all patients with community-acquired pneumonia, and the wide confidence interval means the mortality estimate should be treated cautiously.
The authors conclude that CAECP can resemble other CAP at presentation and still be associated with higher 90-day mortality despite similar baseline severity scores. They call for standardized clinical guidance, while the study's observational design means the association is not proof of cause.
Paper data and sources
Original title: Community-acquired Escherichia coli pneumonia: clinical, radiological, laboratory, and prognostic characteristics in a retrospective single-center study.
Authors: Kücük S, Mutlu S, Dogan C, Soydan S
Journal/Repository: Annals of Saudi medicine
Status: Peer-reviewed
First online: 2026-10-01
DOI: 10.5144/0256-4947.2026.358
Original paper · Full text