Although slightly fewer than 1% of hospitalizations for chronic obstructive pulmonary disease (COPD) are complicated by sepsis, this complication increases the risk for in-hospital mortality fivefold, investigators who studied a representative national sample found.
Among nearly 7 million hospitalizations in which the primary diagnosis was COPD, nearly 65,000 (0.93%) patients experienced sepsis as a complication. In all, 31% of patients with COPD and sepsis were discharged from the hospital to another care facility, and 19% of patients died in hospital, report Harshil Shah, MD, from Guthrie Corning (N.Y.) Hospital and colleagues.
“Our study highlights the need for better risk stratification in patients with COPD developing sepsis to improve the outcomes. Further studies are warranted to consider factoring some of the modifiable factors into account and to ameliorate the outcomes of sepsis during COPD hospitalizations,” Dr. Shah and colleagues write in a poster presented during the at the annual meeting of the American College of Chest Physicians, held virtually this year.
COPD has been associated with increased risk for sepsis because of the use of corticosteroids, underlying comorbidities, and, potentially, because of impaired barrier function, the authors note.
To determine the effects of sepsis and predictors of poor outcomes among patients hospitalized for COPD, the investigators used standard diagnostic codes to identify patients with a primary diagnosis of COPD from the Nationwide Inpatient Sample for the period 2007 through 2018 and sepsis from codes in secondary fields in the International Classification of Diseases (9th/10th Editions) Clinical Modification.
They identified a total of 6,940,615 hospitalizations in which the primary diagnosis was COPD; in 64,748 of those cases, sepsis was a complication.
As noted, the in-hospital death rate, one of two primary outcomes, was 19% for patients with COPD and sepsis, and the rate of discharge to other facilities was 31%.
In analysis adjusted for confounding factors, sepsis was associated with an odds ratio for mortality of 4.9 (P < .01) and an OR for discharge to a facility of 2.2 (P < .01).
With regard to trends, the investigators saw that, although the adjusted odds for in-hospital mortality remained stable over time, discharge to facilities increased significantly. In 2007, the adjusted OR was 2.2, whereas in 2018, it was 2.6 (P for trend = .02).
Predictors of in-hospital mortality among patients with sepsis included increasing age (OR, not shown), White ethnicity (OR, 1.2), treatment in the Northeast region (OR, 1.4), disseminated intravascular coagulation (OR, 3.7), pneumococcal infection (OR, 1.2), congestive heart failure (OR, 1.2), and renal failure (OR, 1.4; P < .01 for all comparisons).
Mortality risk for many patients
A COPD specialist who was not involved in the study told this news organization that sepsis is an uncommon but serious complication, not just for patients with COPD but also for those with other severe illnesses.
“Sepsis has a high risk for mortality whether a person has COPD or not,” commented David M. Mannino III MD, FCCP, FERS, professor of medicine at the University of Kentucky, Lexington, and a cofounder and co–medical director of the COPD Foundation.
“It’s not surprising that sepsis is lethal in this population; the question is, if you have COPD, are you more likely to have sepsis? And I think the answer is probably yes. The connection there is that people with COPD have a higher risk for pneumonia, and pneumonia itself is probably one of the biggest risk factors, or certainly an important risk factor, for the development of sepsis,” he said in an interview.
It would be interesting to see the relationship between sepsis and in-hospital mortality for patients with other chronic diseases or people without COPD, he said, and he would have liked to have seen more detailed information about trends over time than Dr. Shah and colleagues provided.
No funding source for the study was reported. Dr. Shah and colleagues and Dr. Mannino have disclosed no relevant financial relationships.
A version of this article first appeared on.