JSTOR

Methods

Results

During the 36-month study period, there were 3 distinct outbreaks of infection with influenza A virus subtype H3N2, 1 in each ICU. Outbreak 1, in April 2005, occurred in the medical ICU and involved 7 HCWs. Outbreak 2, in December 2005, occurred in the coronary care unit and involved 4 HCWs. Outbreak 3, in July 2006, occurred in the surgical ICU and involved 6 HCWs (Table). For each outbreak, the HCW attack rate ranged from 18% to 24%; each outbreak occurred after HCW exposure to the respective index patient infected with influenza A (H3N2) virus. There was a median of 3 days of exposure (range, 2–5 days) between the identification of the index patient and the identification of the infected HCWs, and all cultures of samples obtained from HCWs yielded influenza A (H3N2) virus. No HCW reported recent exposure (ie, within the past 2 weeks) to family members or friends with an influenza-like illness, and all HCWs who had influenza were furloughed from work.

Table 1.  Nosocomial Influenza A Virus Infection and Outbreaks Among Healthcare Workers (HCWs) in 3 Intensive Care Units of a Thai Tertiary Care Hospital During the 2004–2006 Influenza Seasons

Open New Window

All relevant hospital personnel (eg, hospital staff, bed and nursing management, clinical, and other supportive service personnel) were informed about the outbreak; the remaining HCWs in the ICU received an influenza vaccine (Fluarix; GlaxoSmithKline) and were observed for 2 weeks for fever and influenza-like illness. In each outbreak, the infection control team was notified within 1 day of the identification of the index patient. Each influenza outbreak was extinguished within 2 days (range, 1–2 days) after the notification of the infection control team (Figure).

(33 KB)

Figure.  Epidemic curves for outbreaks of influenza A virus infection among healthcare workers and at-risk intensive care unit patients in a medical intensive care unit (A), a coronary care unit (B), and a surgical intensive care unit (C) in a Thai medical center. IC, infection control.

Open New Window

Antiviral prophylaxis was administered to all remaining inpatients. The investigation of potential secondary horizontal influenza transmission led to the identification of 1 patient (1 [12.5%] of 8) with nosocomial influenza A during outbreak 1; 1 patient (1 [12.5%] of 8) during outbreak 2; and 1 patient (1 [12.5%] of 8) during outbreak 3. All 3 cases of hospital-acquired influenza occurred in patients who were located in close proximity (ie, in the next bed) to the ICU index patient, who had rapid test results and viral cultures negative for influenza A (H3N2) virus on admission, and who had subsequent rapid test results and viral cultures positive for the virus within a median of 2 days (range, 2–3 days) after the identification of the index patient. During the 36-month period when these 3 influenza outbreaks among HCWs occurred, there were 15 patients with influenza who were admitted to the ICUs, and an additional 30 HCWs from these ICUs acquired influenza A infection during the nonoutbreak periods (Table). Notably, there were no seasonal patterns associated with these influenza outbreaks.

The mean estimated costs were $2,710 (range, $2,245–$3,321) per outbreak investigation and $256 (range, $220–$330) for the annual influenza vaccination of all 77 ICU HCWs (Table). The benefits of influenza vaccination of these ICU HCWs include the reduced risk of influenza infection, the minimization of risk for influenza transmission to at-risk ICU patients, and the cost savings, in comparison with the costs of redirected infection control measures and HCW furloughs.

Discussion

Although occupational health programs in developing countries seem long overdue, understanding how to best prioritize HCW protection with resource-limited healthcare delivery systems remains a challenge. Although a panoply of interventions have been recommended to prevent and control the spread of influenza in healthcare settings,^8-10 translating these recommendations into practice in resource-limited settings requires prioritizing the use of the existing infrastructure and available resources. Our findings revealed influenza A attack rates above 10% among unvaccinated HCWs, despite compliance with infection control measures.8 Although influenza outbreaks in spatially confined units have been reported worldwide, along with estimates for excess costs of care,^11-13 our data provide additional evidence for the incorporation of annual influenza vaccination into ICU occupational health programs. Notably, all at-risk ICU HCWs in our study agreed to influenza vaccination. This level of compliance may not be generalizable to nonoutbreak situations or other geographic areas, but these findings suggest the potential benefits of focused education concerning the impact of vaccination on the safety of HCWs and the added protection for at-risk patients.

Influenza infection occurs year round in tropical countries.14 Given the potential threat for the emergence of pandemic influenza in southeast Asia, these 3 outbreaks of influenza A (H3N2) infection among unvaccinated HCWs highlight the relevance of vaccines in the promotion of global standards for occupational health. Additionally, the outbreak investigations incurred costs more than 10-fold higher than the estimated costs of HCW vaccination.5 Although indirect costs related to increased hospitalization among inpatients were not included, our estimates suggest that there is a fiscal benefit to annual influenza HCW vaccination, along with additional protection of at-risk ICU patients from influenza A infection.