Factors Associated with Unprotected Exposure to 2009 H1N1 Influenza A Among Healthcare Workers, May-July 2009

Background: Healthcare workers (HCWs) are at risk for occupational influenza infection and can serve as vectors for healthcare-associated influenza transmission. Use of respiratory personal protective equipment (PPE) is one of four types of controls that can protect HCWs from occupational influenza infection.  Factors associated with unprotected exposure to 2009 H1N1 influenza among HCWs during the early stages of the pandemic are unknown.

Objective: To identify patient and HCW related factors associated with unprotected HCW exposures to 2009 H1N1 influenza.

Methods:   Infection Control (IC) records were reviewed to identify patients with confirmed 2009 H1N1 influenza infection admitted to the hospital through the Emergency Department (ED) from 5/09-7/09.  At the time of infection confirmation IC practitioners performed exposure investigations.  Involved HCWs were interviewed and PPE use was assessed.  Information about case patients was obtained from hospital records.  Influenza like illness (ILI) was defined as fever and either cough or sore throat.  According to hospital protocol patients with ILI were to be given a surgical mask and HCWs caring for these patients were to wear respiratory PPE.  An unprotected HCW exposure was defined as providing care to a patient with influenza without using recommended PPE.

Results:   44 patients with confirmed 2009 H1N1 influenza infection were evaluated.  37 (84%) met the definition of ILI and 26 (59%) cases resulted in 1 or more unprotected HCW exposures.  In total, 277 HCW exposures were identified.  The highest number of exposures was among nurses (41% of exposures) and most occurred in the ED (57% of exposures). Patient factors associated with a higher mean number of unprotected HCW exposures per case patient were absence of reported fever, absence of cough, and absence of fever on initial clinical exam (p < 0.05 for each).  Patients that did not meet criteria for ILI were associated with a higher mean number of unprotected HCW exposures (p < 0.05).  There were no significant associations between the mean number of unprotected HCW exposures and other symptoms, patient medical history or radiographic findings.

Conclusions:   In the early stages of the 2009 H1N1 influenza pandemic, many HCWs reported lack of use of recommended respiratory PPE while providing care to patients subsequently determined to be infected.  Unprotected exposures were more extensive among HCWs providing initial care to patients without classic ILI due to delays in initiating recommended IC measures.  Since 16% of patients did not have classic ILI, strategies that do not rely on the presence of classic signs and symptoms to trigger the implementation of IC measures appear to be needed to prevent unprotected HCW exposures.  Further study of factors that influence PPE adherence, interventions to improve PPE use, and strategies to enhance early implementation of IC measures are warranted.