By Betty Tran, MD, MSc, Editor

SYNOPSIS: This secondary analysis of data from five tertiary care centers found that among patients hospitalized for community-acquired pneumonia, very few deaths potentially were related to a lapse in in-hospital quality of pneumonia care.

SOURCE: Waterer GW, et al. In-hospital deaths among adults with community-acquired pneumonia. Chest 2018 May 30. pii: S0012-3692(18)30801-8. doi: 10.1016/j.chest.2018.05.021. [Epub ahead of print].

Although hospitalization for and subsequent mortality related to community-acquired pneumonia (CAP) is common, it is unclear whether improvements in inpatient pneumonia-related care can affect pneumonia-related mortality. A prior study showed that only about half of all deaths in patients with CAP were attributable to their acute illness.1

Waterer et el conducted a secondary analysis of the Etiology of Pneumonia in the Community (EPIC) study of adults hospitalized for CAP at five tertiary care hospitals (three in Chicago, two in Nashville). Notably, patients with recent prior hospitalizations, tracheotomies/gastric tubes, cystic fibrosis, neutropenic cancer, transplant, and HIV with CD4 counts < 200/mm3 were excluded. Treating physicians made management decisions for each patient. For patients who died during their index CAP hospitalization, a five-physician panel at each study city (with expertise in emergency medicine, infectious diseases, pulmonary, and critical care) made several determinations. These included: cause of death based on a priori criteria; whether the cause of death was directly, indirectly (major or minor contribution), or not related to CAP; whether management was consistent with current recommendations in care quality metrics (antibiotics administered per Infectious Diseases Society of America and American Thoracic Society guidelines, antibiotics delivered within six hours of presentation or one hour [if shock present], using arterial blood gas or pulse oximetry to assess oxygenation); and whether end-of-life limitations in care existed. All five panelists discussed the cases, with complete medical records available for review until the physicians reached a consensus (four out of five or five out of five in agreement).

The authors included 2,320 adults hospitalized for CAP in the final study population. Fifty-two patients died during their CAP hospitalization. The most common causes of death were hypoxemic respiratory failure (25.0%) and septic shock (23.1%). Compared with patients who survived hospitalization, those who died were older and exhibited more comorbidities.

The physician panel attributed 27 deaths directly to CAP. Panelists attributed 10 deaths to situations in which CAP played an indirect role with major contribution. Further, physicians found that nine deaths occurred when CAP played a minor role. Finally, the physicians ruled that CAP was unrelated to the other six deaths. There were DNR orders for a significant number of patients who died (21 of 52 patients). Ten had DNR orders in place prior to admission, eight after admission but > 48 hours prior to death, and three within 48 hours of death. Sixty-seven percent of in-hospital deaths occurred within the first 10 days of admission.

Among the 52 patients who died in the hospital, the physician panel identified nine who had a lapse in quality of in-hospital CAP care. However, the physicians judged five of these nine deaths to be unrelated to the lapse in care quality. Further, the physicians judged two of the nine deaths were patients who had end-of-life care limitations in which decisions were made not to pursue ICU care at the time of admission. Therefore, only two patients who were not DNR were identified to have a lapse in quality in-hospital pneumonia care, potentially contributing to in-hospital death. In one patient, there was difficulty finding intravenous access, with a subsequent delay in antibiotics. For the other patient, medical staff thought there was an intra-abdominal infection, based on an admit chest X-ray that did not show signs of pneumonia, and administered ciprofloxacin only; however, a CT scan the next day was consistent with pneumonia.


This authors of this prospective, multicenter study of more than 2,000 adults hospitalized for CAP found a low in-hospital mortality rate of 2.2% (52 patients) and identified only two patients for whom a lapse in in-hospital pneumonia care potentially contributed to death. As such, the authors concluded that most in-hospital deaths among adults with CAP would not have been preventable with improved quality of in-hospital care.

The study has several strengths, including review of patient cases by multiple physicians with clinical expertise. These physicians paid careful attention to whether CAP could be an indirect contributor (minor or major) to death based on a broad view of how acute pneumonia could lead to extrapulmonary complications (e.g., new cardiovascular disease, stroke, renal failure, secondary infection after initial stabilization of CAP). Also, the authors captured end-of-life limitations on care, which affects whether patients die in the hospital.

The most important limitation to this study is its generalizability. The five hospitals were academic, urban, U.S. facilities that maintain extensive training programs and employ clinician scientists who are dedicated to studying CAP and providing high-quality care. The relatively low mortality rate and high compliance with quality pneumonia care are reflective of this. Thus, the study’s findings likely are not generalizable to other institutions, where in-hospital deaths due to CAP may be reduced by following recommended guidelines for CAP and sepsis management. This study and its case report template (available in Appendix 1 in the online supplement) would be a helpful starting point for individual hospitals to evaluate their own outcomes and guide quality improvement initiatives related to CAP hospitalizations.


  1. Mortensen EM, et al. Causes of death for patients with community-acquired pneumonia. Results from the Pneumonia Patient Outcomes Research Team Cohort Study. Arch Intern Med 2002;162:1059-1064.