No abstract

Severe inflammatory response, acute respiratory distress syndrome, and death are established serious consequences of the acute phase of severe viral pneumonia. However, the long-term respiratory outcomes of severe viral pneumonia, including its association with pulmonary fibrosis, are less known. Objective: To determine whether viral pneumonia is associated with an increased incidence of post-inflammatory pulmonary fibrosis. Design: We performed two retrospective observational cohort studies using longitudinal hospitalization records from the States of California (2005-2011) and Florida (2009-2015) for the discovery and validation studies, respectively. Patients who were 85-years-old and younger with at least two hospital encounters but without a prior diagnosis of pulmonary fibrosis were included. International Classification of Diseases-9 (ICD9) codes of primary and secondary diagnoses and procedures were used to identify the exposure: diagnosis of viral pneumonia; the outcome: incidence of post-inflammatory pulmonary fibrosis [PIPF, ICD9: 515]; and the confounders. Methods: Chronological age was used as the study time scale. Non-parametric Kaplan-Meier (KM) estimator and semiparametric Cox Proportional Hazard modelling were used to assessing the risk of PIPF. P-values < 10-3 were considered significant. Results: Among 9,802,565 patients from California and 8,741,345 patients in Florida cohorts, the prevalence of PIPF was 0.61% and 0.62% over 7 and 6.75 years, respectively. Patients with incident PIPF were older than those without [68(SD: 11) vs. 40(22) years]; among patients with PIPF, those with viral pneumonia diagnosis were younger than those without [63(12) versus 68(11) years]. Incidence of PIPF was higher for those with viral pneumonia diagnosis versus those without [1.6 (CI:1.51-1.69) vs. 0.91 (CI:0.86-0.96)] cases per 1000 person-years in California and in Florida [1.11 (CI:1.06 -1.16) vs, 0.93 (CI:0.89-0.98)]. Viral pneumonia was associated with increased risk of incident PIPF in both California aHR = 1.49 (1.38, 1.61), and Florida aHR of 1.26 (1.20, 1.33) cohorts (Table). Among patients who developed PIPF, the median time to diagnosis was 7.41 (6.16 -8.66) and 4.80 (4.34 - 5.26) years earlier for patients with viral pneumonia versus without in California and Florida cohorts. The association of viral pneumonia was not found for idiopathic pulmonary fibrosis [ICD9: 516.3]. Our findings suggest that patients hospitalized with viral pneumonia may have long term respiratory sequela that is often overlooked and suggest a need for additional studies focusing on phenotyping susceptible patients. This finding is especially important in light of the current COVID-19 pandemic because viral pneumonia is the most common manifestations of the disease, which could lead to subsequent fibrosis.

No abstract