Association between the cumulative dose of glucocorticoids before the development of pneumonia and death in patients receiving long-term glucocorticoids: a secondary analysis based on a Chinese cohort study

Abstract

Background: The present study aimed to evaluate the association between the cumulative dose of glucocorticoids (GCs) and case fatality in hospitalized patients who developed pneumonia while receiving glucocorticoid therapy. Methods: This retrospective cohort study included 625 patients receiving long-term GC treatment who were hospitalized with pneumonia (322 male and 303 female). Data were obtained from the Dryad Digital Repository and were used to perform secondary analysis. Multivariable Cox proportional hazard regression model and restricted cubic splines (RCS) were used to evaluate the association between the cumulative dose of GCs and case fatality. Sensitivity analyses and subgroup analyses were performed. Results: The 30-day and 90-day death rates were 22.9 and 26.2%, respectively. After adjusting for potential confounders, compared with those in the lowest quintile (≤ 1.5 g), the Cox proportional hazard regression model analysis showed that patients with different cumulative doses of GCs (1.5 to 2.95, 2.95 to 5, 5 to 11.5, and > 11.5 g) had lower risks for 30-day death, with respective hazard ratios of 0.86 (95% CI, 0.52 to 1.42), 0.81 (0.49 to 1.33), 0.29 (0.15 to 0.55), and 0.42 (0.22 to 0.79). The multivariable-adjusted RCS analysis suggested a statistically significant N-shaped association between the cumulative dose of GCs and 30-day death. A higher cumulative dose of GC tended to first lead to an increase in 30-day death within 1.8 g, then to a statistically significant decrease until around 8 g [HR for 1 g = 0.82 (0.69 to 0.97)], and again to an increase afterward. Similar results were found in the subgroup analyses and sensitivity analyses. Conclusion: N-shaped association between the cumulative dose of GCs and case fatality was observed in patients receiving long-term GC treatment who were hospitalized with pneumonia. Our findings may help physicians manage these patients.