Impact of substance use disorder on pneumonia hospital outcomes in chronic lymphocytic leukemia: An inpatient nationwide analysis.

Abstract

e19034 Background: CLL is associated with immune dysfunction and severe infections, particularly pneumonia (PNA). Substance use disorders (SUD) increase respiratory infection risk and impair healthcare engagement. The impact of SUD on pneumonia outcomes in CLL is unknown. Methods: Retrospective National Inpatient Sample analysis 2016 to 2023 of CLL hospitalizations for infectious non-COVID PNA with or without documented SUD. Primary outcome was in hospital mortality. Secondary outcomes included mechanical ventilation, complications, and discharge disposition. Multivariable logistic regression models adjusted for demographics, hospital factors, income, insurance, and comorbidities were developed. Chi-squared tests for independence were performed. Propensity matching (PSM) was performed for age, sex, race, income, insurance, hospital factors, and comorbidities. Results: Our cohort included 127,915 hospitalized patients with CLL and PNA, of whom 1,990 (1.56%) had documented SUD. SUD was more prevalent among patients aged 50 to 69 years (54.77% vs 24.71%, p<0.001), those with median household income below USD 57,488 (37.44% vs 25.44%, p<0.001), and Medicaid insurance (22.36% vs 3.39%, p<0.001). Patients with CLL/PNA and SUD were more likely to smoke (55.03% vs 39.17%, p<0.001), have chronic pulmonary disease (52.26% vs 39.81%, p<0.001), and depression (22.11% vs 12.03%, p<0.001), whereas those without SUD more frequently had hypertension (70.04% vs 63.88%, p<0.001), diabetes (30.54% vs 23.87%, p<0.001), and dementia (9.01% vs 3.27%, p<0.001). After PSM, discharge against medical advice remained higher in the CLL/PNA with SUD group (7.29% vs 1.26%, p<0.001). In hospital mortality was lower in the CLL/PNA with SUD group (7.29% vs 9.20%), but this difference was not statistically significant after multivariable adjustment (aOR 1.05, 95% CI 0.70 to 1.58, p=0.816) or after PSM (aOR 0.84, 95% CI 0.50 to 1.40, p=0.497). SUD was associated with higher odds of invasive mechanical ventilation (15.83% vs 8.94%, aOR 1.42, 95% CI 1.05 to 1.93, p=0.024), pneumothorax (2.26% vs 0.73%, aOR 2.70, 95% CI 1.35 to 5.41, p=0.005), and vasopressor use (5.03% vs 2.59%, aOR 1.80, 95% CI 1.06 to 3.05, p=0.029). Conclusions: In hospitalized patients with CLL and PNA, concomitant SUD was associated with higher odds of in hospital respiratory complications and discharge against medical advice. No statistically significant difference in in hospital mortality was observed after multivariable adjustment and propensity score matching. These results suggest an association between SUD and increased inpatient clinical complexity, without evidence of a difference in short term mortality in this cohort.