Use of vaccination in a large outbreak of primary varicella in a detention setting for African immigrants

Background

Primary varicella (PV) presents a public health risk for adults in closed residential settings, especially for immigrants from tropical areas where infection during childhood is less likely.

Methods

In this study, an outbreak of PV at a detention facility for illegal immigrants from Eritrea and Sudan in southern Israel is described. Basic demographic information and clinical course for all cases were obtained.

Results

One hundred and nine cases of PV, all in young adult men aged 18–40 years, were diagnosed over a 7-month period (June to December 2012). Diagnosed patients were placed in quarantine until the resolution of illness without other public health measures being implemented. The Israeli Ministry of Health was notified of the outbreak in early December and recommended two doses of varicella vaccine for all susceptible detainees and staff. Within 2 weeks of completion of the first dose of vaccine, there was only one additional case in a detainee immunized 13 days prior to diagnosis. The effectiveness of vaccination in halting the outbreak was immediate, despite the fact that 15.6% of detainees refused to be immunized.

Conclusions

The possible roles of vaccination or natural infection in achieving herd immunity and thereby ending the outbreak in this population are discussed. We recommend considering early vaccination for all when an outbreak or a series of connected cases is detected in a closed-residential setting such as the detention facility described here.

Adults, Outbreak, Prison, Varicella

Introduction

Varicella zoster virus (VZV) causes a diffuse vesicular exanthem known as chickenpox. Infection occurs primarily by direct contact with infected persons and via airborne transmission. As a herpesvirus, VZV can also reactivate locally after a primary infection to cause the disease known as herpes zoster (shingles).¹ In temperate climates, primary varicella (PV) occurs with a seasonal peak in the late winter and early spring, with high rates of exposure during childhood. In tropical areas, for reasons that are not completely clear, many VZV infections occur during the second or third decades of life resulting in a higher proportion of young adults who are susceptible.² While usually mild in children, VZV can be severe in adults³ and immune-compromised hosts.⁴ The incubation period can range from 10–21 days, but usually lasts 14–16 days. Asymptomatic infection is also possible. Immunity after primary infection is long-lasting, and recurrent cases of disseminated varicella in adults are rare.

Since the introduction of a varicella vaccine in 1995 (in the United States) the epidemiology of the disease continues to change with increasing rates of vaccine coverage. One dose of vaccine is effective in preventing 95% of severe disease and 78% effective in preventing all disease in children, but its effectivity is lower in adults.⁵ A two-dose regimen of vaccine is effective in preventing 95% of all disease.⁶ Vaccination may also be effective as post-exposure prophylaxis (PEP) both in sporadic and outbreak settings, although there is less experience of using the vaccine for this indication.⁷ Recommendations exist for starting PEP in outbreaks of children, especially within 3 days of exposure.⁸ However, a Cochrane review from 2008 found few studies in children and no controlled trials in adults providing evidence for recommending PEP.⁹ Adults are known to develop less of an immune response after one dose of varicella vaccine.⁵ In this study we describe an outbreak of PV in a detention facility in southern Israel, as well as the control measures that were implemented.

Materials and methods

In Israel, PV is a disease requiring collective notification when a cluster of cases are identified. The District Health Office (DHO) is responsible for the investigation of outbreaks and for issuing recommendations for the prevention of further disease spread. Outbreak investigation includes on-site assessment of living conditions, possible methods of spread and validation of clinical diagnosis. All cases of PV that are presented for medical attention are collected in an electronic medical record system. Demographic information and details about clinical course are recorded for each case.

Population

More than 51 000 people from Eritrea and Sudan have illegally crossed Israel's borders with Egypt, between 2007 and 2012, for reasons of economic opportunity or political asylum.¹⁰ In Israel, a detention facility was designed to accommodate this part of the population and currently houses about 6000 detainees under the authority of the Israel Prison Service. Two complexes were involved in the outbreak described in this study: Ketziot (K) housed 800 and Saharonim (S) 1700 detainees. The detainees live in a variety of conditions including tents and permanent facilities, all with full access to sanitary and medical needs. Complex K consisted exclusively of men aged between 18–40 years, while complex S housed 100 women and 65 children (under the age of 18 years) in a separate ward. The facilities are divided into smaller wards that are separate, although detainees from one ward can be exposed to detainees from other wards during clinic visits and legal hearings, which are held in a central location with a common waiting area. Additional exposure can occur when detainees are transferred from S to K and within facilities for security and safety considerations.

Description of outbreak

A case of PV was defined by clinical evaluation on the basis of a characteristic vesicular rash with multiple lesions at different stages of progression. Biological material was not sampled. In the K facility, the first cases of PV were identified on 11 November 2012, with one case in each of the two different wards. After a 2-week interval, additional cases were diagnosed as shown in Figure 1. Another wave of cases was reported 2 weeks later, consistent with the average incubation period of PV. The spread of infection occurred in the same wards as the first two index patients. The disease was mild in all cases and none of the patients were referred to the hospital. Patients were quarantined from diagnosis until the complete crusting of lesions, as per standard recommendations for infection control in a closed setting,¹¹ and then returned to the wards. Individual data on length of quarantine were not available, however, there were no cases requiring prolonged isolation.

Figure 1.

Epidemic curve of primary varicella (PV) in facility K.

In the S facility, there were 57 cases of PV since the middle of June 2012, as shown in Figure 2. Other than quarantine of cases upon diagnosis, no other outbreak control measures were implemented. The cases occurred exclusively in the adult male population and there were three emergency department visits, including one admission to hospital.

Figure 2.

Epidemic curve of primary varicella (PV) by week in facility S.

The Southern DHO of the Israeli Ministry of Health was notified of the outbreak on 4 December 2012. The next day both facilities were visited by an epidemiologic investigation team including a public health physician and nurse epidemiologist from the DHO. Detailed information about the residential arrangements and clinical and demographic data for each case were obtained. The importance of isolating cases with PV and providing individual notification to the DHO were emphasized. After analyzing the available data, the DHO recommended vaccination with two doses of varicella vaccine for all detainees and staff that did not have a documented history of PV or a contraindication to the vaccine. Given language, cultural and cooperation barriers with regard to detainees, a subjective history of PV was not considered an appropriate method of screening.¹² Serological testing was not performed due to the urgency of the situation, as well as for logistic and cost-effectiveness considerations. The study was exempt from institutional review board (IRB) approval because it was part of outbreak management and part of routine public health practice of the DHO in response to a major event.

Vaccine implementation

The Israel Prison Service began immunizing detainees on 27 December 2012. All members of staff reported a history of varicella therefore no staff members were immunized. Given the language barrier and complexities regarding detainees’ cooperation, varicella history or lack thereof by subject report were not considered reliable. Arab-speaking staff presented the intervention to the detainees including the expected benefit of the vaccine, its safety and for who the vaccine was indicated.

Results

There were 11/800 cases (1.5%) at the K facility and 98/1700 cases (5.7%) at the S facility. By the end of the outbreak, 109/2500 cases of PV had been reported, an infection rate of 4.4% at both facilities combined. The median age was 25 years in the S facility and 28 years in the K facility. At the K facility, 460/800 detainees (57.5%) received the vaccine and 242/800 (30.2%) of detainees refused. At the S facility, 1459/1700 detainees (85.8%) were vaccinated and 150/1700 detainees (8.8%) refused the vaccine. Overall, the refusal rate was 15.6% (392/2500) for both facilities. Two detainees had a contraindication to the vaccine due to advanced HIV infection with CD4 counts of less than 200. There were no indications for the administration of varicella zoster immune globulin (VZIG) or intravenous immunoglobulin.

We estimated that the remaining susceptible population consisted of 147/2500 persons (5.8%). This estimation was calculated by subtracting the number of documented cases and expected vaccine responders to one dose⁵ (82%) from the overall population and multiplying the result by the assumed seronegativity rate (20%).¹³ Results were tabulated using Microsoft Excel (Microsoft Corporation, Redmond, WA, USA).

Discussion

This paper presents a rare description of an outbreak of varicella among refugees. Primary varicella outbreaks in prison or refugee settings have previously been described in the medical literature.^14–19 These reports were of several cases in each facility, while the presently described outbreak included 109 cases. The magnitude of the outbreak in this study could be attributed to the unexpected circumstances of varicella among adults from tropical climates and the delay in reporting. Previous studies of immigrants and refugees from Africa found higher rates of seronegative subjects in these populations as compared to Western populations. Barnett et al. reported the seroprevalence of various refugee populations in Canada and found that 21% of East African refugees aged between 13–20 years lacked detectable antibodies to varicella.²⁰

Although PV can result in severe disease in adults, nearly all cases in this large outbreak were mild. However, the extent of the infection was substantially higher than expected when compared to a Western population in a similar setting. This may be consistent with the epidemiology of varicella in tropical areas as opposed to temperate climates,²¹ and due to differential genetic susceptibility. Our results are similar to findings described among American military recruits, which demonstrated lower rates of evidence of past varicella infection among African-Americans.²²

There are conflicting reports of varicella vaccination in outbreak settings.²³ Most of the experience comes from school or daycare-based outbreaks. Lopez et al. described the failure of one dose of varicella vaccine to halt a school outbreak of PV, despite 99% coverage rates.²⁴ Arnedo-Pena et al. also described vaccine efficacy in a Spanish town, but since vaccine rates were low, an outbreak still occurred.²⁵ The outbreak described in this study stopped within one incubation period after vaccine initiation despite an overall 15.6% refusal rate.

This episode of continuing PV incidence in East African immigrants afforded an opportunity to observe the effect of varicella vaccine as PEP in an outbreak setting, while working to protect the health of the detainees. The cessation of the outbreak upon initiation of vaccine raises the question of herd immunity. It is likely that the described population had relatively low natural immunity, no vaccination history, and continuous exposure under crowded living conditions, which abruptly ended with mass vaccination. One possible explanation is that the natural course of the outbreak depleted the susceptible group. However, one would expect in such a case, that the epidemic curve would appear more Gaussian (bell-shaped) and would not cease abruptly at the height of disease incidence (Figure 3). Also, according to our estimation, as many as 6% of the detainees remained susceptible.

Figure 3.

Primary varicella (PV) cases in facilities K and S (by week), from the 7 October 2012 until the end of outbreak.

Alternatively, the vaccine afforded the needed addition to the naturally protected percentage, thus achieving the threshold for PV herd immunity. We do not believe that a change in reporting could account for the decrease in morbidity. Vigilance of the staff to identify cases of PV was actually increased following the report to the DHO and the initiation of the vaccine campaign. Serological testing was not performed because it was felt that the vaccine should be provided immediately given the number of cases and logistic considerations in performing blood tests on a large number of subjects.

Our study had some limitations. Given the urgent nature of the outbreak, there was limited opportunity to obtain information about movement between wards and facilities. The lack of cooperation made it difficult to engage the detainees both in regard to ascertainment of past varicella disease and explaining the importance of vaccination. Because of the above concerns, serological studies were not obtained, which limited our ability to calculate the number of susceptible individuals and the vaccine efficacy.

Conclusions

In summary, we describe a large varicella outbreak in which one dose of varicella vaccine rapidly curtailed the progression of new cases. It appears that earlier initiation of the vaccine may have had a higher impact on the prevention of secondary cases and ongoing morbidity. Based on the experience reported here, we believe that early immunization is recommended when an outbreak or cluster of connected cases is detected in a closed-residential setting.

Authors' contributions: EJH and MG conceived the study; EJH and MG designed the study protocol; ML and LG carried out the clinical assessment; EJH, LD and MG carried out the analysis and interpretation of these data. EJH and MG drafted the manuscript; LG, LD and ML critically revised the manuscript for intellectual content. All authors read and approved the final manuscript. EJH and MG are guarantors of the paper.

Acknowledgements: None.

Funding: None.

Competing interests: None declared.

Ethical approval: Ethical approval was not sought as this was acute outbreak management as part of routine public health practice of the DHO.