BACKGROUND: Respiratory syncytial virus (RSV), parainfluenza virus (PIV) and influenza virus (Flu) are common respiratory viral infections (RVIs) implicated in hematopoietic stem cell transplant (HSCT) recipients. Despite their possible association with high rates of pneumonia and mortality, their clinical and economic burden has not been well studied.
METHODS: HSCT recipients with documented RVI who were treated at our institution between September 2012 and October 2015 were included in the study. We used Vizient (formerly University Health Consortium) clinical database to collect and compare total costs, including length of stay, ICU admission rates, intravenous immunoglobulin use, steroid use, and mortality rates among RVIs in HSCT recipients. Encounter-specific demographics, risk factors, underlying cancer, and outcomes were also collected. Multiple linear regression analyses were applied to identify predictors of higher total cost associated with RVI in HSCT recipients at MD Anderson.
RESULTS: Average total cost per encounter was $49,371 for RSV, $29,679 for PIV, and $15,077 for Flu. A total of 1,636 hospitalization days (d) were attributed to these RVIs with an average of 7 d per RSV, 8 d per PIV, and 5 d for Flu infection. The average ICU admission rate was 12% for RSV, 9% for PIV, and 4% for Flu. Around 11% of total RVI encounters had active graft-vs.-host disease at the time of their RVI. Out of the patients with upper respiratory infection, 20% RSV, 44% PIV, and 21% Flu progressed to pneumonia during the 28 d of the study period. Of the 246 total RVI encounters, overall all-cause mortality rate was 6% (RSV: 8% [8/98], PIV: 1% [1/70] and Flu: 8% [6/78]). Length of stay, ICU admission, and receiving intravenous immunoglobulin were strong predictors of higher cost for all RVIs.
CONCLUSION: This study underscores the significant impact of RVIs in terms of economic and clinical burden in HSCT recipients. Major differences in total costs per encounter across the three RVIs were observed. This cost and clinical data may be helpful for future cost effectiveness studies in this population.
