Background: Legionella pneumophilia is typically transmitted via inhalation aerosols derived from water or soil during outbreaks or be acquired sporadically, commonly causing community-acquired or nosocomial pneumonia1,2. Death occurs through progressive respiratory failure, shock and multiorgan failure3,4. Epidemiological data for in-patient Legionella in the United States (U.S.) is lacking. The objective of the study is to evaluate inpatient epidemiology, association with hyponatremia and chronic obstructive pulmonary disease (COPD) of adult Legionella related hospital admissions.
Methods: We conducted a retrospective analysis using the 2016 National Inpatient Sample data. The NIS is the largest publicly available inpatient database in the U.S., representing a 20% stratified sample of all U.S. non-federal hospitals, and is sponsored by the Agency for Healthcare Research and Quality and the Healthcare Cost and Utilization Project (HCUP). We identified Legionella infection (A481), hyponatremia (E871), pneumonia (J12*, J13*, J14*, J15*, J16*, J17*, J18*) and COPD (J42*, J43*, J44*) using the International Classification of Diseases, Tenth Revision, Clinical Modification (ICD-10-CM). We examined the demographics, hospital settings, length of stay, costs of admission and in-hospital mortality rate. Multivariate regression analysis was performed using STATA 16.0 to determine the relationship of outcomes between groups of Legionella patients with or without coexistence of COPD. It was also used to evaluate the relationship of hyponatremia and Legionella infection in patients admitted with pneumonia. P-value <0.001 was used as the significance threshold.
Results: There was an estimate of 4735 adult admissions with a discharge diagnosis of Legionella for the year 2016 (15.68 per 100,000 adult admissions), among which 60.7% were male, 78.9% were over age of 50 (Mean age was 61.3), and 67.7% were Caucasian. Geographically, 25.2% were from the Northeast, 33.7% were from the Midwest, 28.5% were from the South, 12.6% were from the West. The average length of stay was 10.2 days and the total cost was $122 million in 2016. The in-hospital mortality rate was 6.5%. Among all patients admitted with Legionella, 36.2% of patients had a diagnosis of hyponatremia. After adjusting age, gender, race, income, comorbidities, hospital bed size, region, insurance, teaching status and location, the odds ratio (OR) of hyponatremia in all pneumonia patients with Legionella was 3.85 [95% confidence interval (CI) 3.33 – 4.45, p<0.001]. 21.8% of hospitalized Legionella admissions had a diagnosis of COPD. Coexisting Legionella and COPD were not associated with higher mortality, length of stay, or total cost (p=0.26, 0.93, 0.09, respectively).
Conclusions: Our study is the first population-based retrospective study about Legionella infection in a national hospital database in the U.S. Hospitalizations with Legionella infection is more common in patients with age greater than 50. COPD was not associated with higher mortality, length of stay, or total cost in Legionella admissions. Among all hospitalizations with pneumonia, patients with Legionella infection had significantly higher odds of having hyponatremia.