Outcomes and predictors of hospital mortality of patients with Philadelphia-negative myeloproliferative neoplasms and sepsis: A Nationwide Inpatient Sample Database analysis

e19038

Background: Sepsis is associated with increased mortality in patients with hematologic malignancies, especially in those receiving cytotoxic chemotherapy. The development of new therapies also increases the risk of infections in Philadelphia negative myeloproliferative neoplasms (MPN). However, the clinical implications and costs of sepsis in patients with polycythemia vera (PV), essential thrombocytosis (ET), and primary myelofibrosis (PMF) have not been reported at a national level. Therefore, we aimed to determine the outcomes, hospitalization costs, and predictors of mortality in patients with MPN and sepsis. Methods: We retrieved adult patients with MPN from the Nationwide Inpatient Sample database from 2016-2018. We used the ICD-10 codes to identify and compare patients with and without sepsis. The main outcomes were hospital mortality and predictors of mortality. We computed the chi-squared test and the Mann-Whitney U-test. Mortality predictors are estimated using multivariate logistic regression and logistic fixed-effect methods to control for admission cohort and hospital time invariant characteristics. All analyses were performed using Stata Statistical Software version 14 (StataCorp, College Station, TX). Results: A total of 86,723 patients with MPN were identified. The majority of them were white (66.9%), female (54.3%), with a median age of 62 (IQR 47-76). The most common MPN was ET (84.3%), followed by PV (14.2%). Sepsis was diagnosed in 16,146 (18.6%) of the MPN patients. There was a significantly higher mortality in the sepsis group vs. the non-sepsis group (7.4% vs. 1.8%; p < 0.001), longer LOS (8 vs. 4 days; p < 0.001), and higher median hospitalization cost (US$74,128 vs. US$39,987; p < 0.001). In the multivariable analysis, sepsis was associated with higher mortality (OR: 4.74; CI 95%: 4.33-5.18; p < 0.001). Among the MPN, those with PMF and sepsis had a higher risk of death (OR: 2.21; CI 95%: 1.72-2.82; p < 0.001). Other significant variables associated with mortality were age > 65 (OR: 2.51; CI 95%: 2.14-2.95; p < 0.001), concomitant chronic heart failure (CHF) (OR: 1.57; CI 95%: 1.39-1.77; p < 0.001), chronic kidney disease (CKD) (OR: 1.22; CI 95%: 1.10-1.35; p < 0.001), and weight loss (OR: 2.12; CI 95%: 1.79-2.51; p < 0.001). There was no significant association with sex and race/ethinicity. Conclusions: On patients with MPN, sepsis was associated with higher mortality, hospitalization costs, and LOS. Additionally, those with increased age, CHF, CKD, weight loss, and PMF were also at increased risk of death. There should be more emphasis on assessing the risk of sepsis on MPN to prevent worse outcomes and higher costs. Further studies should focus on identifying the specific causes of sepsis in MPN and promoting early recognition.

Clinical and financial implications of central venous catheters bloodstream infections in patients with multiple myeloma in the United States

e20033

Background: Multiple Myeloma(MM) has an increased exposure to intravascular catheters due to the increased risk of severe infections and due to the route of treatment administration. Central Venous Catheter Bloodstream Infections(CVCBI) have significant mortality, imply a delay in treatment and increased cost as well. There is no prior report on the risk factors for poor outcome in MM patients that develop this complication. We aim to describe the predictors of mortality as well as the changes in cost that CVCBI implies. Methods: We retrieved adult patients with MM from the Nationwide Inpatient Sample database from 2016-2018. We used the ICD-10 codes to identify and compare patients who developed CVCBI with those that did not developed it. The main outcomes were hospital mortality and predictors of mortality. We computed the chi-squared test and the Mann-Whitney U-test. Mortality predictors are estimated using multivariate logistic regression and logistic fixed-effect methods to control for admission cohort and hospital time invariant characteristics. All analyses were performed using Stata Statistical Software version 14 (StataCorp, College Station, TX). Results: A total of 58,838 patients with MM were identified. The majority were white (63.5%), male (55.3%), with a median age of 70 (IQR 62-78). Most MM were not in remission (99.1%), followed by those in remission (1.3%) and relapse (0.3%). CVCBI was diagnosed in 264 (0.4%) of the MM patients. There was significantly higher mortality in the CVCBI group vs. the non-CVCBI group (8.7% vs. 5%; p < 0.01), longer LOS (10 vs. 5 days; p < 0.001), and higher median hospitalization cost (US$86,168 vs. US$43,511; p < 0.001). In the multivariable analysis, CVCBI was associated with higher mortality (OR: 1.69; CI 95%: 1.14-2.52; p < 0.001). Among patients with MM and CVCBI those that had achieved remission had a higher risk of death (OR: 2.87; CI 95%: 2.17-3.8; p < 0.001). Other variables associated with mortality were age > 65 (OR: 1.84; CI 95%: 1.59-2.15; p < 0.001), concomitant chronic heart failure (CHF) (OR: 1.46; CI 95%: 1.29-1.65; p < 0.001), chronic kidney disease (CKD) (OR: 1.43; CI 95%: 1.32-1.56; p < 0.001), and weight loss (OR: 2.31; CI 95%: 1.91-2.8; p < 0.001). When compared to medicare patients with higher mortality were more likely to be under medicaid(OR: 1.25; CI 95%: 1.02-1.55; p < 0.05) and private insurances(OR: 1.31; CI 95%: 1.15-1.49; p < 0.001). There was no significant association with sex, race/ethinicity or household income. Conclusions: In patients with Multiple Myeloma the development of Central Venous Catheter Bloodstream Infections was associated with a higher overall mortality, length of stay and cost of hospitalization. Age, CHF, CKD and weight loss were independent risk factors for poor outcome in this patient population. Further studies are required on developing strategies for the prevention of this complication.