Methods

The study was conducted at 3 hospitals that admitted >80% of the patients with suspected Nipah virus encephalitis from 26 December 1998 through 30 April 1999. We identified nurses and doctors exposed to patients hospitalized with outbreak-related encephalitis after 1 March 1999, by reviewing nursing duty rosters (for nurses) and ward rotation schedules (for doctors). Most outbreak-related encephalitis patients (110 [97%] of 113) were admitted after 1 March 1999. Exposure was defined as having provided direct patient care to an encephalitis patient from the Negeri Sembilan State who was admitted after 1 March 1999. We sought to enroll an equal number of unexposed nurses and doctors from wards that did not admit encephalitis patients, to minimize the risk of inclusion of HCWs who may have been exposed unknowingly. HCWs from these wards who reported working with encephalitis patients or having worked on wards that admitted patients with encephalitis after October 1998 were excluded from the unexposed group. All pathologists and their assistants in the pathology departments were included in the study and were asked whether they had any contact with the body of an encephalitis patient admitted after 1 March 1999. Study participants completed a questionnaire about types of exposures to patients with encephalitis and work practices. In addition, they were asked to provide 1 (unexposed HCWs) or 2 serum samples (exposed HCWs). The initial serum sample from both unexposed and exposed HCWs was obtained at the time the questionnaire was administered. To ensure that recent infections were not missed, exposed HCWs were asked to provide a second serum sample 3–4 weeks later

Infection was detected by an EIA for anti–Nipah virus IgM and IgG antibodies. IgM antibodies were detected by using an IgM capture EIA, and IgG antibodies were detected by using an indirect EIA. Hendra virus–derived antigen was used in both assays, because antibodies to Nipah virus react against Hendra virus antigens (CDC, unpublished data). Serum samples that were positive or that gave a borderline positive result for Nipah virus antibodies by the EIA were tested with a neutralization assay, to confirm the speci-ficity of the EIA result. For the neutralization assay, the heat-inactivated serum samples were tested in duplicate in a biosafety level 4 laboratory at dilutions of 1:10–1:640 against a median virus challenge of 30–300 TCID of prototype Nipah virus in Vero E-6 cells. Replication of virus was determined at 6 days by microscopy, to detect syncytial formation, after the cells were fixed with formaldehyde, and cell mono-layers were stained with crystal violet. The neutralization titer was considered to be the highest dilution with complete inhibition of syncytial formation

A standardized questionnaire was developed in English, translated into Bahasa Malay, and administered by public health nurses fluent in both languages. The questionnaire obtained information on demographics, illness, and types of exposures to encephalitis patients or patient body fluids and secretions. A second questionnaire, administered to the exposed cohort at the time of the collection of the second serum sample, obtained information about illnesses, exposures, and activities that had occurred after the administration of the first questionnaire

Data were entered into an Epi Info 6.02 data file (Centers for Disease Control and Prevention), and differences in responses among unexposed and exposed HCWs were tested for significance by using the χ² test. Because the 3 hospitals varied in the level of exposure by both number of patients admitted and the time of first outbreak-related admissions, we examined data from each hospital independently

Results

In all, 363 HCWs, pathology workers, and pathology assistants were recruited for the exposed cohort (>98% of HCWs and patholo-gists that records showed had been exposed to outbreak-associated cases of encephalitis) and 288 HCWs, patholo-gists, and pathology assistants were recruited for the unexposed cohort. Hospital A had 10 outbreak-related admissions, the first occurring on 2 March 1999; hospital B had 117 outbreak-related admissions beginning on 26 December 1999; and hospital C had 84 outbreak-related admissions after 22 February 1999 (table 1). The exposed and unexposed groups were similar in distribution of jobs, age, sex, and ethnicity. More than 60% of the exposed HCWs reported contact with an encephalitis patient before the institution of infection control measures on 19 March 1999 when Nipah virus was first identified (table 2). Many HCWs reported episodes of a high-risk exposure (e.g., needlestick injury [n=12] or splash of patient body fluids to mucosal membranes [n=39])

View larger version:

• In this window
• In a new window

• Download as PowerPoint Slide

Table 1

Date of first presumed Nipah encephalitis admission, number of infected patients admitted, and workers exposed by study hospital

View larger version:

• In this window
• In a new window

• Download as PowerPoint Slide

Table 2

Exposures reported by health care workers caring for or pathologists performing autopsies on patients with presumed Nipah virus infection

There were no reports of any serious illness, encephalitis, or hospital admission among any HCW or pathology worker. The proportion of HCWs reporting any illness during the month before the study was not statistically significant between exposed and unexposed HCWs (32% vs. 26%, respectively; P=.12), nor was the proportion of HCWs who reported a febrile illness (12% vs. 13%, respectively; P=.87) statistically significant

None of the first serum samples from either the exposed or unexposed groups tested positive by EIA for Nipah virus IgG or IgM antibody. Sixty-four nurses and doctors temporarily assigned to hospital 3 were unavailable for a second serum sample; they were on temporary duty at hospital 3 and had returned to their permanent duty station at the time the second serum sample was to be obtained. Most (293 of 299) of the remaining exposed HCWs and pathology workers gave a second serum sample and completed the second questionnaire. Three (1%) of the second serum samples from exposed HCWs and pathology workers were positive for Nipah virus IgG anti-bodies. It is likely that these 3 were false positives; no IgM response occurred, and the serum samples were negative for anti–Nipah virus neutralizing antibodies. These 3 serum samples were provided by 3 nurses: 1 reported a febrile illness before the first serum sample was obtained, another reported a febrile illness between the 2 serum samples obtained, and 1 reported a mucosal splash exposure. All 3 nurses had cared for outbreak-related encephalitis patients >30 days, compared with 10 days in nurses with negative IgG

Discussion

The results of this study of >300 HCWs exposed to patients presumed to be ill with Nipah virus encephalitis suggest that the risk of nosocomial transmission of Nipah virus is very low. Despite substantial contact with patients with outbreak-related encephalitis (or potentially infectious body fluids from these patients) by many HCWs or pathology workers, none had clini-cal evidence of infection (i.e., encephalitis or other serious illness), and none had definitive serologic evidence of infection, although the sensitivity of these assays is not completely known. Although 3 HCWs had a positive EIA result for Nipah virus IgG antibodies in their second serum sample, none developed an IgM response or Nipah virus neutralizing antibodies. Thus, it is likely that these 3 serum samples contained antibodies that were not induced by Nipah virus infection but did react against the Hendra virus antigen preparation used in the EIA. The lack of nosocomial transmission of Nipah virus to HCWs is somewhat unexpected, given reports of potentially high-risk exposures: many HCWs cared for patients before they were advised to use contact and respiratory precautions, and a substantial number reported needlestick injuries, splash exposure, and skin exposure to body fluids from encephalitis patients

The lack of nosocomial transmission is consistent with the fact that most Nipah virus–infected patients had direct contact with pigs [7], and few, if any, were likely to have been infected by contact with household members. Our findings are also consistent with the lack of transmission of Hendra virus to HCWs caring for Hendra virus–infected patients in Australia [8]. This difference in risk of infection between exposure to pigs and humans follows from differences in the extent of infection in the respiratory tract between pigs and humans. Immunohisto-logic studies demonstrate high levels of Nipah virus antigen in the lungs and upper respiratory tract of infected pigs, with substantially lower levels of Nipah antigen in these tissues from autopsy specimens from Nipah virus–infected humans [9]

In conclusion, Nipah virus–infected patients do not appear to present a high risk of transmitting the virus to HCWs. However, it is likely that there is some risk for human-to-human transmission, since viral antigens have been detected in human kidney and respiratory tract tissues, and since the virus has been isolated from human urine and respiratory secretion speci-mens [6]. Given the potential severity of illness, we feel that HCWs should use standard and droplet precautions (e.g., strict attention to good hand washing practices, wearing a mask, and wearing gloves when coming in contact with secretions, excretions, and body fluids of patients) while caring for patients infected with, or likely to be infected with, Nipah virus [10]

Nipah Virus Nosocomial Study Group Members

Abu Bakar Suleiman, Mohamad Taha Arif, Ah Sian Tee, Deven Kurup, and Marzukhi Mohamad Isa (Ministry of Health, Kuala Lumpur, Malaysia); Wan Rozita Wan Mahiyuddin and Muhammad A. Kamaluddin (Institute for Medical Research, Kuala Lumpur, Malaysia); Flora Ong (Sarawak State Health Department, Kuching, Malaysia); Ghazali Othman (Negeri Sembilan State Health Department, Negeri Sembilan, Malaysia); and Patrick C. Stockton (Centers for Disease Control and Prevention, Atlanta)