Factors associated with nosocomial SARS-CoV transmission among healthcare workers in Hanoi, Vietnam, 2003

Abstract

Background: In March of 2003, an outbreak of Severe Acute Respiratory Syndrome (SARS) occurred in Northern Vietnam. This outbreak began when a traveler arriving from Hong Kong sought medical care at a small hospital (Hospital A) in Hanoi, initiating a serious and substantial transmission event within the hospital, and subsequent limited spread within the community.

Methods: We surveyed Hospital A personnel for exposure to the index patient and for symptoms of disease during the outbreak. Additionally, serum specimens were collected and assayed for antibody to SARS-associated coronavirus (SARS-CoV) antibody and job-specific attack rates were calculated. A nested case-control analysis was performed to assess risk factors for acquiring SARS-CoV infection.

Results: One hundred and fifty-three of 193 (79.3%) clinical and non-clinical staff consented to participate. Excluding job categories with < 3 workers, the highest SARS attack rates occurred among nurses who worked in the outpatient and inpatient general wards (57.1, 47.4%, respectively). Nurses assigned to the operating room/intensive care unit, experienced the lowest attack rates (7.1%) among all clinical staff. Serologic evidence of SARS-CoV infection was detected in 4 individuals, including 2 non-clinical workers, who had not previously been identified as SARS cases; none reported having had fever or cough. Entering the index patient's room and having seen (viewed) the patient were the behaviors associated with highest risk for infection by univariate analysis (odds ratios 20.0, 14.0; 95% confidence intervals 4.1-97.1, 3.6-55.3, respectively).

Conclusion: This study highlights job categories and activities associated with increased risk for SARS-CoV infection and demonstrates that a broad diversity of hospital workers may be vulnerable during an outbreak. These findings may help guide recommendations for the protection of vulnerable occupational groups and may have implications for other respiratory infections such as influenza.

Figure 1

Epidemic curve of the SARS outbreak among Hospital A staff, Hanoi, 2003.

Figure 2

Diagram representing criteria for selection of case and control subjects to evaluate risks for SARS-CoV infection stemming from hospital exposure to the Hanoi index patient. (*) SARS cases were confirmed by serologic testing, viral culture, or RT-PCR performed on specimens obtained from persons with clinically compatible illness. (†) Excluded as study cases were SARS cases among the staff who were unlikely to have contracted infection from the index case (i.e., illness onset after March 5^th, 2003 or seroconversion > 18 days after last exposure to the index patient); included as study cases are those SARS cases among the staff who had illness onset on or before Mar 5th, or seroconversion within 18 days of last exposure to the index case. (‡) Included as study controls were non-cases demonstrated to be negative for SARS-CoV antibody at least 18 days after last exposure to the index case. Potential control subjects were excluded from the analysis if no serologic specimen was collected from them or if the specimen was collected too early to assess final outcome status from exposure to the index case (i.e., within 18 days of last exposure).

Figure 3

Subjective symptoms of illness reposted among staff at Hospital A (n = 27 SARS cases, n = 115 non-cases), Hanoi, 2003. The presence of an asterisk indicates that the symptom was significantly associated with SARS case status (p < 0.05, Fisher's exact, two-sided). (#) indicates sample sizes for cases and non-cases of 8, and 58 respectively.