LB 15 Predictors of Disease Severity and Mortality in Hospitalized Patients with Novel H1N1 Influenza

Sunday, March 21, 2010
Grand Hall (Hyatt Regency Atlanta)
Madhuri M. Sopirala, MD, MPH , The Ohio State University Medical Center, Columbus, OH
Douglas Haas, MD , The Ohio State University Medical Center, Columbus, OH
Naeem A. Ali, MD , The Ohio State University Medical Center, Columbus, OH
Brent Riscili, MD , The Ohio State University Medical Center, Columbus, OH
Jeff Pan, PhD , The Ohio State University Medical Center, Columbus, OH
Julie E. Mangino, MD , The Ohio State University Medical Center, Columbus, OH
Background: Identifying epidemiologic characteristics, predictors of disease severity and mortality in H1N1 patients (pts) may help during their initial evaluation and subsequent management.

Objective: To identify predictors of disease severity and death in pts with H1N1. To identify the clinical and demographic characteristics associated with lack of fever in pts hospitalized with H1N1.

Methods: A retrospective review of pts hospitalized with laboratory confirmed H1N1 from June to November 2009 was conducted to identify the predictors of the disease severity [requiring intensive care unit (ICU) stay] and death. Characteristics of pts who presented with fever, those who required ICU stay and those who died were compared to those of pts who did not have fever, did not require ICU and were survivors respectively. The differences between the 2 groups of pts in each of these categories were tested using 2 sample t-test for continuous variables and Chi-square test for categorical variables. Gender, race, functional status, age, body mass index (BMI), abnormal chest radiograph (CXR), chronic lung disease, immunosuppression, cardiovascular disease, diabetes mellitus (DM), end stage renal disease, pregnancy, home prednisone use, length of stay (LOS), ICU admission and death were evaluated for these pts.

Results: A total of 151 H1N1 pts were hospitalized during the study period. Women comprised 81/151 (53.8%) pts and 11 were pregnant. African Americans (AA) comprised 36% and Caucasians were 59%. Median age was 46; BMI ranged from 18 to 75; median 29.35; 53% (81/151 95% CI 45.4% to 61.8%) pts presented with fever. There was no significant difference between pts who presented with and without fever except for DM. Pts with DM tended to present without fever (66/81 vs 46/70; p<0.05). Forty one pts required ICU stay (30.5%, 95% CI 23.2- 38.5%). Pts who required ICU stay were older than those without an ICU stay (48.7 vs 42.1, p=0.03) and had abnormal CXR more often than those not requiring ICU (29/105 vs. 34/46, p<0.001). ICU pts had longer LOS (19.6 vs 6.1 day, p=<0.001). There was no association between fever and ICU admission. Twelve pts expired; overall mortality was 7.9% (95% CI 4.17% to 13.5%). Ten of twelve expired were men (p- <0.001); 8/10 men who expired were Caucasian (p <0.001). Expired pts were older (52.6 vs 43.3 p =0.039) and had longer LOS than survivors (21.2 days vs 9.4 days, p=.018). No deaths occurred in AAs or pregnant women. There was no significant difference in BMI in the tested categories.
Conclusions: Afebrile presentations of patients should not preclude aggressive management since 53% H1N1 pts presented without fever. Interestingly, pregnancy, AA race and obesity were not associated with increased disease severity or death. Older age, abnormal CXR at presentation, Caucasian race and male gender have surfaced as predictors of disease severity and death. H1N1 pts with these predictors should be aggressively treated.