Utilizing Information Systems to Decrease Occupational Exposure to H1N1 Influenza

Background: Many pediatric patients are admitted to hospitals with “influenza-like illness” (ILI), including cough, fever and congestion.  Infection Control (IC) guidelines stipulate that symptomatic patients be isolated promptly, instead of waiting for positive diagnostic tests, to avoid occupational exposure of healthcare workers (HCW) and nosocomial exposure of patients to influenza (FLU).  Data from 9/09 at Children’s Memorial Hermann Hospital (CMHH), during the H1N1 pandemic, demonstrated that only 14/45 (31%) of patients admitted with ILI and testing positive for FLU were isolated at time of admission.  Many HCW were exposed as a result.  Automated information system alerts have been utilized to improve quality of care. Data regarding the utility of information systems to improve compliance with IC guidelines are not available.

Objective: Improve compliance with isolation of patients with ILI by instituting an automated isolation order, triggered by diagnostic testing for FLU. Secondary aims included decreasing risk of occupational HCW FLU exposure and nosocomial FLU transmission.

Methods: CMHH is a 240 bed academic tertiary care.  Diagnosis of FLU is performed via nasopharyngeal (NP) wash specimen for viral culture or rapid antigen testing.  If a NP specimen is ordered, the patient is assumed to have ILI.  Contact and droplet isolation are required for ILI. From 5/09-1/10, a physician order was needed for isolation. From  2/10-7/10, an automated isolation order was generated in the medical record whenever a NP viral test was ordered. The number of patients with an isolation order on the date of FLU testing divided by total number of FLU tests was determined pre and post intervention.  Numbers of HCW occupationally exposed to FLU along with Tamiflu prescriptions dispensed by occupational health (OH) were analyzed pre and post intervention.  Nosocomial FLU infections, defined as positive influenza test >48 hours after admission in patients admitted without ILI, were also examined.

Results: From 5/09-1/10, only 543/1,101 (54%) of patients tested for FLU were isolated appropriately.  From 2/10-7/10, this increased to 430/448 (96%), p <0.0001.  68 HCW were evaluated by OH from 5/09-1/10 for occupational exposure to FLU, and 56 Tamiflu prescriptions were purchased for exposed HCW. A Tamiflu course cost the hospital $108.  The hospital spent $6,048 for post exposure prophylaxis during this period.  Post intervention, no exposures occurred.  One patient acquired noscomial FLU pre intervention. There were no cases of nosocomial FLU post intervention.

Conclusions: Automated computer isolation orders in response to diagnostic testing for pathogens such as FLU significantly increase implementation of isolation precautions. Reducing isolation delays decreases risk of occupational exposures and associated costs. The applicability of automated reflex orders for isolation of other pathogens should be investigated.