Abstract
Background:
Opportunistic infections (OI's) are defined as infections that are more frequent or more severe because of immunosuppression. Conditions like HIV infections, Hematopoietic stem cell transplantation, autoimmune diseases like SLE and RA have been established as risk factors for increased mortality in OI. Studies suggest that lymphomas are a risk factor for mortality associated with OI's like Tuberculosis and candida (Silva et al, 2005; Chen et al, 2014) probably due to underlying immunodeficiency and therapeutic exposures inducing prolonged lymphopenia. Data describing association of underlying Non-Hodgkin Lymphoma (NHL) on mortality associated with OI's is limited. Our objective was to study impact of NHL on in‐hospital mortality of patients hospitalized with OI's.
Methods:
We analyzed data from the Healthcare Cost and Utilization Project's (HCUP) National Inpatient Sample, between 2010 and 2014 using the ICD-9 codes for Opportunistic infections (OI) which included tuberculosis (010-018), non-tubercular mycobacteria (031), cytomegalovirus (078.5), Herpes Zoster( 053), Candidiasis(112.4), Toxoplasmosis (130), Pneumocystis (136.3), Cryptococcosis(117.5), Listerosis (027.0), Nocardiosis (039), Aspergillosis (117.3) coccidiomycosis (114), histoplasmosis (115), blastomycosis (116.0) . The subset of this cohort with NHL excluding lymphoid leukemias listed as a secondary diagnosis were collected for analysis. Baseline characteristics including age, race, gender, Charlson comorbidity index, and infection with HIV etc. were analyzed for patients with admitted with OI's, with and without underlying diagnosis of NHL. Our primary outcome was in-hospital mortality. Characteristics associated with in‐hospital mortality were identified using multivariable logistic regression.
Results
Between 2010 and 2014, a total of 1,81,016 admissions were identified with a primary diagnosis of OI. A total of 4442 admissions had an underlying diagnosis of NHL. Mean age of patients hospitalized with OI's with NHL, was higher (65.4 ±0.52 years vs 58.9±0.14years, p<0.05 ), and 54.7% of the patients were males. Racial distribution of the patients with NHL was similar to those suggested in previous studies (Li et sl, 2016), with 73.19% patients being whites. Patient's median income, hospital bed-size and teaching status of hospital all were found to be significant risk factors (p<0.05) for mortality. When adjusted for the above mentioned factors, as well as HIV or Transplant status of patients, NHL was found to be an independent factor for mortality with an odds ratio of 1.34 (p=0.03).
Conclusion
OI's constitute a major cause of morbidity and mortality in immunocompromised patients. There is limited data from a nationwide analysis which evaluates association of opportunistic infections and lymphoma as an independent marker of mortality. Our study suggests that NHL is an independent risk factor for mortality in patients who were admitted in the hospital for OI's. Further studies should be considered to the role of primary prophylaxis for conditions like PCP to avoid morbidity and mortality associated with these conditions. This will help reduce health care cost burden associated with such hospitalizations.
No relevant conflicts of interest to declare.
