Despite the use of a series of preventive measures, a high incidence of severe acute respiratory syndrome (SARS) was observed among health care workers (HCWs) during the SARS epidemic. This study aimed to determine which preventive measures may have been effective in protecting HCWs from infection, and which were not effective.
A retrospective study was performed among 758 'frontline' health care workers who cared for SARS patients at the Second Affiliated Hospital and the Third Affiliated Hospital of Sun Yat-sen University. The HCWs with IgG against SARS and those without IgG against SARS were respectively defined as the "case group" and the "control group", and logistic regression was conducted to explore the risk factors for SARS infection in HCWs.
After adjusting for age, gender, marital status, educational level, professional title, and the department in which an individual worked, the results of a multivariate logistic regression analysis indicated that incidence of SARS among HCWs was significantly and positively associated with: performing tracheal intubations for SARS patients, methods used for air ventilation in wards, avoiding face-to-face interaction with SARS patients, the number of pairs of gloves worn by HCWs, and caring for serious SARS cases.
Some measures, particularly good air ventilation in SARS wards, may be effective in minimizing or preventing SARS transmission among HCWs in hospitals.
Severe acute respiratory syndrome (SARS), a viral respiratory illness caused by the coronavirus SARS-CoV, was possibly the first globally significant occupational disease to emerge in the 21st century, making healthcare work potentially hazardous . This was indicated by the high incidence of SARS observed among health care workers (HCWs) in the epidemic of SARS, especially during its earlier stages [2-5]. In China, from a total of 5323 SARS cases, 966 (over 18%) were HCWs, and in the early period of the SARS epidemic, near 90% of the SARS patients were frontline HCWs . In Hong Kong, a total of 384 (22.1%) of 1739 suspected or confirmed SARS patients were hospital workers . Generally, SARS outbreaks first originated in hospitals where SARS patients were treated and subsequently spread to communities from there [8,9].
Several studies indicated that HCWs coming into direct or indirect contact with SARS patients in wards had a greatly increased risk of contracting SARS-Cov, despite some strict infection control measures being taken [7,10,11]. A similar situation also arose in the Second Affiliated Hospital and the Third Affiliated Hospital of Sun Yat-sen University during the epidemic of SARS in 2003. A total of 846 HCWs worked on the frontline of caring for SARS patients in the two affiliated hospitals and 112 of them contracted SARS during this time. Throughout the whole period of the SARS epidemic, a series of infection control and protective measures were employed in the two affiliated hospitals. But, why were some of HCWs infected by SARS, and some of them were not? The objective of this study was to determine which preventive measures used were effective in protecting HCWs from SARS, and which were not effective. To answer this question, we conducted a retrospective study of HCWs who worked at the frontline in the two affiliated hospitals during the SARS epidemic.
In mid-May 2003, about 4 months after the initial SARS outbreak in Guangzhou, a retrospective study was conducted in HCWs working at the frontline of the SARS epidemic, providing primary care in the Second Affiliated Hospital and the Third Affiliated Hospital of Sun Yat-sen University, where the first and second outbreak of SARS among HCWs occurred in the early stage of SARS epidemic in Guangzhou. Among a total of 846 frontline HCWs who tended to SARS patients from the two hospitals, 758 (89.2%) who were on duty during the investigation were surveyed, and they included HCWs from all departments involved in the care of SARS patients in the two hospitals. But, those who were off-duty during the survey were excluded. During the SARS epidemic, a total of 112 HCWs working on the frontline were diagnosed suffering from "SARS" according to a case definition of SARS by the Ministry of Health, China , and 90 of them were successfully interviewed, giving a response rate of 80.4% (90/112). Written informed consent was obtained from all the participants prior to enrollment after a detailed explanation of the study objectives and requirements of the survey. The Ethical Committee of the Sun Yat-sen University approved the study.
Definition of a SARS case
A SARS case was defined using the criteria for probable SARS cases provided by the Health Ministry of China . Criteria for probable and suspected SARS cases included travel to a SARS epidemic area in the 2 weeks before the onset of symptoms or close contact with a probable SARS patient; fever of ≥ 38°C; chest x-ray abnormalities; normal or decreased leukocycte count; and no response to treatment by antimicrobial drugs.
In the present study, 10 mL of peripheral venous blood was collected from all the subjects, and the serum was separated and stored at -70°C. Immunoglobulin (Ig) G against SARS-CoV was detected using an enzyme-linked immunosorbent assay (ELISA) . Among the 758 surveyed HCWs, 91 ones (80 cases were diagnosed suffering from "SARS" and 11 ones were not diagnosed suffering from "SARS") had IgG antibodies against SARS, and the prevalence rate of IgG antibodies against SARS was 12.01% for the total samples . Furthermore, the prevalence of IgG antibodies against SARS-CoV was 88.9% (80/90) for HCW with SARS [86.3% (63/73) in the Second Affiliated Hospital and 100.0% (17/17) in the Third Affiliated Hospital], and 1.6% (11/668) for HCW without SARS who worked on frontline of SARS [2.8% (8/288) in the Second Affiliated Hospital and 0.8% (3/380) in the Third Affiliated Hospital] .
A standardized interview with a structured questionnaire was used to obtain the following information in mid-May 2003, about 4 months after the initial SARS outbreak in Guangzhou. (1) Socio-demographic characteristics, including: age, gender, marital status, educational level, professional title, and in which department did you work? (2) History of SARS patient care, including: i) Did you receive any special training for how to handle SARS? ii) Did you ever perform a tracheotomy? iii) Did you ever perform tracheal intubations? iv) Did you ever care for "Super Spreading SARS cases"? (3) Use of personal protective and control measures, including: i) How many gowns did you wear while you cared for SARS patients? ii) How many multilayered cotton masks did you wear while you cared for SARS patients? iii) How many pairs of gloves did you wear while you cared for SARS patients? iv) With what frequency did you wear shoe covers while you cared for SARS patients? vi) With what frequency did you wear a cap while you cared for SARS patients? (vii) With what frequency did you wear a face shield while you worked in SARS wards? i) With what frequency did you wear goggles while you performed operations on SARS patients? (4) Health-related behaviors included: i) With what frequency did you wash uncovered skin after you cared for SARS patients? ii) With what frequency did you wash hands after you cared for SARS patients? iii) With what frequency did you wash your nasal cavity after you cared for SARS patients? iv) With what frequency did you wash your mouth after you cared for SARS patients? (5) Other relevant control measures were: i) What type of air ventilation system was used in your office and in SARS wards? ii) What type of hand-washing equipment was used in your office? More details about the name, definition and value of these variables are listed in Table 1.
Table 1. Variables and definition
HCWs who had IgG against SARS (91 cases = 80 cases with SARS and 11 cases without SARS) and those without both IgG against SARS and SARS (657 cases) were defined as the "case group" and "control group", respectively. 10 HCWs had been previously diagnosed as SARS, but their IgG against SARS test was negative, so that they were excluded from the data analysis. Logistic regression was conducted to explore the risk factors for SARS infection among HCWs and odd ratios (ORs) and 95% confidence intervals were used to assess the association of SARS infection with the factors studied. Univariate analysis was performed at first for each risk factor. Factors with P < 0.1 were included in a multivariate logistic regression analysis and analyzed using a forward-stepwise procedure. In the multivariate logistic regression analysis, age, gender, marital status, educational level, professional title, and the department in which the HCW worked were controlled as potential confounding factors. The entry and exit criteria were set at P = 0.05 and P = 0.10, respectively. List-wise deletion was used in the multivariate analyses. All the P values were two-tailed, and a P < 0.05 value was considered statistically significant, unless otherwise mentioned.
All the statistical analyses were performed using SPSS 11.0 for Windows .
Socio-demographic characteristic of the surveyed subjects
Table 2 presents general information about the surveyed subjects provided by the two affiliated hospitals.
Table 2. Socio-demographic characteristics of HCWs in the two affiliated hospitals
Logistic regression analysis
Table 3 presents the results of univariate logistic regression analysis. Among the eighteen surveyed risk factors, fifteen factors were significantly associated with SARS infection in HCWs, with the exceptions being "Frequency of wearing face shield in SARS ward", "Frequency of washing hands after caring for SARS patients", and "Frequency of washing nasal cavity after caring for SARS patients". See Table 3.
Table 3. Univariate logistic regression analysis
After adjusting for age, gender, marital status, educational level, professional title, and the department in which the individual worked, a multivariate logistic regression model identified five variables associated with altered risk of contracting SARS at a significance level of 0.05 (Table 4). They were: performing tracheal intubations for SARS patients, insufficient methods used for air ventilation in wards, avoiding face-to-face interaction while caring for SARS patients, the number of pairs of gloves worn by the HCW, and caring for "Super Spreading SARS Cases". The result of the Hosmer-Lemeshow goodness of fit for the model was: χ2 = 4.739, df = 7, and P > 0.05.
Table 4. Multivariate logistic regression analysis
Performing tracheal intubation for SARS patients and caring for "Super Spreading SARS Cases" significantly increased the risk of SARS infection among HCWs working on the frontline. In contrast, wearing multiple (2) pairs of gloves could protect HCWs from SARS infection. Compared with wards with artificial central ventilation, those with natural ventilation or with both natural ventilation and electronic exhaust fans at the same time significantly decreased the probability of HCWs being infected with SARS-CoV. A much lower incidence rate of SARS was found among HCWs who either usually or consistently avoided face-to-face contact with SARS patients in their care.
It was hypothesized that performing certain high-risk procedures, such as nasopharyngeal aspiration, bronchoscopy, endotracheal intubation, airway suction, and cardiopulmonary resuscitation, might increase the rate of SARS-Cov shedding occurring in a SARS patient's respiratory secretions, thereby increasing the risk to HCWs of contracting SARS while performing such procedures . The results of the present study demonstrate that performing tracheal intubations was highly associated with incidence of SARS among HCWs. Therefore, the results imply that adequate personal protective equipment is required when conducting certain high-risk procedures which may contribute to the presence of infectious droplets in the environment.
It was hypothesized that the primary mode of SARS transmission was via droplets spread through close person-to-person contact , and this was strongly supported by the occurrence of clusters of cases among HCWs caring for SARS patients and family members of SARS patients [16,17]. In the present study, we found that avoiding face-to-face contact with SARS patients while caring for them could significantly reduce the probability of HCWs contracting the virus. This may be due to decreased exposure to infected droplets resulting from this practice. This result implies that HCWs could use appropriate personal protective measures (such as avoiding face-to-face contact with SARS patients) to protect themselves from SARS infection while they are caring for patients with SARS.
There is some evidence that longer range airborne transmission may have played a role in the spread of the SARS virus in some settings, such as in the outbreak of SARS in wards with faulty ventilation in the Prince of Wales Hospital of Hong Kong , in the transmission of SARS on an aircraft , and in the community outbreak at Amoy Gardens in Hong Kong . The results of the present study also indicate that airborne transmission might have been a contributing factor in spread of SARS in 2003. Compared with ventilation through artificial central air-conditioners in the wards, natural ventilation alone and natural ventilation enhanced by an additional electronic exhaust fan at the same time could significantly reduce the risk of HCWs contracting SARS in the wards. In wards with artificial central ventilation, windows were closed which might lead to much lower air flow and much higher viral load in the wards, and HCWs were easily infected with the SARS virus while working in such an environment. By contrast, the windows of wards with natural ventilation and natural ventilation enhanced by an additional electronic exhaust fans were opened, and the air flow and the exchange rate of air in the wards were high, which might greatly decrease the density of the SARS virus in the wards and may also reduce the probability of HCWs contracting the virus.
SARS-Cov may be shed from a SARS patient's respiratory secretion and feces, and the latter may further contaminate objects in the ward. The protective gown, gloves, multilayered cotton mask, and head and foot coverings wore by HCWs may also be contaminated while caring for SARS patients. It has been shown that SARS-Cov may remain viable for considerable periods on a dry surface (up to 24 hours)  and is stable in feces and urine at room temperature for at least 1 to 2 days and 4 days in stool from patients with diarrhea . Hence, touching surfaces or objects that are contaminated with SARS-CoV may introduce the virus into the mucous membranes of the eye, nose, or possibly the mouth. It is believed that nominally 'clean' areas may be contaminated if an HCW wears a piece of protective clothing contaminated with SARS patients' secretions into the area. For this reason, HCWs must wear two layers of gown, gloves, multilayered cotton mask, head and foot covering in SARS wards and discard the outer layer before entering clean areas, in order to prevent fomite transmission to other areas . This study proved that wearing two layers of gloves significantly protected HCWs from SARS compared with wearing a single layer of gloves, but we did not find that wearing double layers of gowns, multilayered cotton masks, and head and foot coverings were associated with HCWs being protected from SARS. This might be due to the fact that almost of all the procedures involved in caring for patients were done with the hands; hence gloves were more highly contaminated by SARS patients' secretions.
A small number of severely infected patients or super-spreading patients appeared to play a disproportionate role in the spread of the disease to HCWs. For instance, several clusters of SARS outbreaks in hospitals can be traced to such patients in Hong Kong, Singapore, and Toronto [2,4,5]. It had been hypothesized that these patients might have a relatively depressed immune system with associated high viral loads and may be unduly facilitating transmission of the virus. In the present study, the same index patient led to the two clusters of SARS outbreaks among HCWs in the two affiliated hospitals. Statistical analysis showed that caring for a "Super-spreading Patient" significantly increased the risk of HCWs suffering from SARS. In light of this, a series of stringent infection control measures should be required when HCWs care for patients suspected of being SARS super-spreaders.
Several limitations of the study ought to be mentioned here. First, our investigation was limited to two affiliated hospitals of Sun Yat-sen University. This is not representative of all of the hospitals in which patients with SARS were admitted and cared for in Guangzhou. Therefore, this is a typical case investigation. Second, ventilation in the wards was not objectively assessed for some reason, meaning that we could not exactly evaluate the influence of the ventilation in the wards on the transmission of SARS among HCWs. Third, we could not trace the tree structure of the primary, secondary, and third class cases, which prevented us from clarifying the association of the HCWs infected by SARS with the index case directly or indirectly. Fourth, some factors, such as oxygen therapy and bi-level positive airway pressure ventilation were found to be related to nosocomial infection of SARS in other study , were not included in the present study, which indicated that we missing an opportunity to find some effective measures for protecting HCWs from SARS or to assess their effect. Fifth, in the early stage of SARS epidemic, the diagnosis of SARS was based on the history of epidemiology, signs and symptoms suggested by the Health Ministry of China , not on the directive biomarkers of SARS-CoV or antibodies against SARS-CoV, which might lead to over reporting "SARS" cases or missing identifying inapparent infection or subclinical infection. This might be the reason that 80 of 90 HCWs with "SARS" and 11 of 668 subjects without "SARS" were seropositive. Sixth, some prevention measures were usually employed at the same time in SARS wards, which meant that these measures were highly correlated. In this situation, multivariate statistical analysis might omit some effective measures in the final model due to multicollinearity. Seventh, 10.8% of frontline HCWs who cared for SRAS patients were not included in the present study, which was the reason that the number of HCWs involved in intubation in the present study was less than our previous study , which might cause to underestimate the association of the intubation with the nosocomial infection of SARS. Finally, although we identified several preventive measures which were effective for protecting HCWs from SARS, we could not eliminate the inefficiencies of other adopted measures, due to the fact that we utilized a retrospective rather than an interventional study design.
In summary, good air ventilation in wards and a series of simple control and preventive measures might decrease or prevent SARS transmission among HCWs in hospitals.
The authors declare that they have no competing interests.
W-QC contributed to study design, data analysis and paper writing. WHL participated in study design and administered the study. CYL, YTH, ZLL, LL were responsible for the field investigation and data collection. JH and GL organized the field investigation. GMY participated in study design and administered the study. All authors read and approved the final manuscript.
We would like to thank the management and healthcare workers in the Second and Third Affiliated Hospitals of Sun Yat-sen University. This project was supported by a research grant from both the Ministry of Science and Technology and the Ministry of Education of China.
MMWR 2003, 52:405-411. PubMed Abstract
Seto WH, Tsang D, Yung RWH, Ching TY, Ng TK, Ho M, Ho LM, Peiris JSM, Advisors of Expert SARS group of Hospital Authority: Effectiveness of precautions against droplets and contact in prevention of nosocomial transmission of severe acute respiratory syndrome (SARS).
MMWR 2003, 52:433-436. PubMed Abstract
MMWR 2003, 52:461-466. PubMed Abstract
MMWR 2003, 52:241-248. PubMed Abstract
Ann Intern Med 2003, 139:564-7. PubMed Abstract
Chen WQ, Lu CY, Wong TW, Ling WH, Lin ZN, Hao YT, Liu Q, Fang JQ, He Y, Luo FT, Jing J, Ling L, Ma X, Liu YM, Chen GH, Huang J, Jiang YS, Jiang WQ, Zou HQ, Yan GM: Anti-SARS-CoV immunoglobulin G in healthcare workers, Guangzhou, China.
Yu IT, Xie ZH, Tsio KK, Chiu YL, Tang XP, Hui DS, Lee N, Li YM, Huang ZT, Liu T, Wong ZW, Zhong NS, Sung JJ: Why did outbreak of severe acute respiratory syndrome occur in some hospitals nut not in others?
CID 2007, 44:1017-1025. Publisher Full Text
The pre-publication history for this paper can be accessed here: