https://orcid.org/0000-0003-0576-4520
Abstract
Mumps is a highly contagious disease spread by airborne droplets, making control especially difficult in congregate, crowded settings such as shelters and jails. A mumps outbreak in Honduras, starting in 2018 among adults who were unvaccinated, spread northward with Central Americans migrating to the United States. We describe 2 mumps outbreaks in Houston during 2019 among migrants at the Houston Contract Detention Facility (HCDF) and among inmates at the Harris County Jail (HCJ).
We investigated cases of acute onset parotitis. Three or more mumps cases in a facility was considered an outbreak. Confirmed cases had positive polymerase chain reactions (PCR). Probable cases were linked epidemiologically to a confirmed case in the same unit and a positive serology for serum anti-mumps immunoglobulin M (IgM) antibody. Outbreak control measures included enhanced surveillance, isolation of housing units, educational outreach, and immunization with Measles, Mumps, Rubella (MMR) vaccine.
At HCDF, during a 10-month period, we investigated 42 possible cases. Of the possible cases, 28 were lab-confirmed with 9 probable, 4 ruled out, and 1 vaccine reaction. All were migrants. At HCJ, during a 3-month period, we investigated 60 suspect cases; 20 cases were lab-confirmed, 13 probable and 27 ruled out. All but 2 were inmates. Only about a third of those offered MMR vaccination accepted.
Successful outbreak resolution required close cooperation with HCDF and HCJ with ongoing surveillance, isolation of units with cases and MMR vaccination. Such facilities will have outbreaks; regular communications with local public health could improve response.
Mumps is a highly contagious viral disease transmitted via respiratory drops and contaminated fomites [1]. Outbreaks are likely to occur among unvaccinated persons in crowded settings, including homeless shelters, jails, prisons, schools, factories, and other congregate settings [2, 3]. Parotitis occurs in about two-thirds of cases, and complications may include meningitis, encephalitis and orchitis. Mumps has an incubation period of 12–25 days. Infected persons are contagious during a time extending from one to two days before parotitis develops and continuing 4–5 days after symptoms start [1]. We describe 2 separate mumps outbreaks investigated by the Houston Health Department (HHD) during 2019 in a detainee facility and a jail.
METHODS
Case Definitions
We followed standard recommendations from the Centers for Disease Control and Prevention (CDC) and the Council of State and Territorial Epidemiologists (CSTE) to identify, characterize, and control the outbreaks [1]. All suspect cases investigated had clinical symptoms of mumps, specifically acute parotitis. Laboratory-confirmed cases had a positive polymerase chain reaction (PCR) [1, 3]. Probable cases had a positive serologic test for mumps immunoglobulin M (IgM) [1, 3] antibody and an epidemiological link to a laboratory-confirmed case [3] domiciled in the same housing unit. An outbreak was defined as 3 or more cases linked by time and place [3].
Control and Surveillance Measures
Meetings with administration and staff were held at both facilities to gather information about numbers of persons in custody and fluctuation in the population, numbers of staff (law enforcement, corrections, support staff, medical providers, and contractors), daily schedules (meals, sleeping, recreation, and other) and the physical layout of each facility. Transport policies for court appearances were also assessed. We also ascertained facility policies on vaccination against mumps with the measles, mumps, rubella (MMR) vaccine for those in custody and for staff.
After an initial assessment, we worked with the medical staff to implement daily, active surveillance measures to assess the numbers of infections, implement isolation of housing units with suspect cases, ensure appropriate use of masks and other personal protective equipment in the facility and during transport, provide appropriate sanitation and begin laboratory testing for suspect cases. We also conducted retrograde surveillance to look for cases that may occurred before HHD notification. Vaccinations were offered for staff and those in custody through various delivery measures, including onsite clinics for employees and in housing units with suspect cases.
Housing units where detainees/inmates were identified as suspect mumps cases were isolated and conducted daily routines (meals, sleep, recreation, and other) separately and without contact with other housing units. Recommended infection control measures included hand hygiene (handwashing with soap), cough etiquette, and bans on the sharing of drinks, food, utensils, and cigarettes. Those with suspect/confirmed mumps cases were confined to their rooms for the duration of their symptoms. A housing unit with a person with mumps symptoms was placed on isolation until 1 mumps incubation period of 25 days passed with no new suspect cases. An outbreak was defined as resolved after 2 incubation periods (50 days) passed with no new cases. Monitoring for reintroduction of cases was conducted at both facilities after the initial outbreaks were resolved.
RESULTS
Contract Detention Facility Outbreak
Notification and Facility
The outbreak was detected when migrants’ relatives called the City of Houston 311 Information Center to report possible mumps cases at HCDF. On 4 January 2019, the relatives were referred to HHD, and staff contacted the HCDF nurse manger on 7 January to obtain information on 7 cases. Staff physicians had made clinical diagnoses with no definitive laboratory testing ordered.
HCDF agreed to ongoing surveillance and implement control measures. A series of meeting over a 2-week period educated corporate staff and the HDCF staff about mumps as a reportable condition, transmission, and use of isolation measures, vaccination, and infection control.
The Houston facility houses a revolving detainee population with an average daily census of 950–1000 and a staff of 430, both employees and contractors. In 2019, HCDF had an average daily stay of 35 days [4]. After intake, the migrants are housed in one of 26 units; each follows a common daily schedule for eating, recreation, and educational activities. Each unit is configured dormitory style and houses between 20 and 60 detainees. A common day area in each unit is where residents eat, watch television, work on computers, and gather. Detainees may also be transported to hearings related to their case. Detainees are provided daily indoor and outdoor recreation.
At the time of the outbreak, MMR vaccination for the staff was not required. The facility did mandate tuberculosis testing for staff.
Cases
Seven mumps cases were reported to HHD on 7 January 2019 and announced at a press conference in early February 2019. The earliest case was had an onset date of 12 December 2018, 26 days before HHD was notified. During that time, 1–2 incubation periods (based on 12–25 days) passed, permitting possible spread within the facility. The February press conference was the first public notification of mumps cases in Texas; others had been reported [5] but not formally announced.
Of the 42 detainees investigated, 29 (69.1%) were lab confirmed by PCR; 9 (21.4%) were probable and 4 (9.5%) were ruled out. The mean of age of the 38 cases was 25.1 years, and all but 2 were male (39, 95.1%). Most cases (33, 86.6%) were from Central American counties, most frequently Honduras (15, 47.4%). Other birth countries were Guatemala (9, 23.6%), El Salvador (six, 16%), Mexico (four, 10.5%) and Spain (one, 2.6%). None of the cases had any documentation (written or self-report) of prior MMR vaccination.
The cases were detected during a 10-month period between December 2018 and November 2019 (Figure 1). During that time, 17 units were placed on isolation, 6s of them more than once. Based on available information on detainee custody dates, 27 (81.1%) cases were exposed in the HCDF during the incubation period before symptom onset (12–25 days).
Mumps cases, Houston Contract Detention Facility, 2018–2019.
Vaccination
A local pharmacy administered MMR during 2 clinics for employees where 134 (31%) were vaccinated. Immunizations were offered in detainee housing, but the number of doses offered and given were not consistently documented.
Jail Outbreak
Notification and Facility
On 11 June 2019, medical staff from the HCJ notified the HHD of a probable mumps outbreak with 12 suspect cases and requested assistance in outbreak control. A retrospective record review found the first case and occurred on 16 May, 26 days before HHD notification.
The facility is operated by Harris County but is located within the City of Houston.
HCJ houses up to 8000 inmates on an average day and has a 20–30% turnover every 2 or 3 days as inmates are booked and released. The average daily census in 2019 was 8831 according to information reported to the Texas Commission on Jail Standards [6]. Based on that estimate and reported turnover, an estimated 1766 and 2649 persons would cycle through the jail every 2–3 days. This rapid recycling is different than a prison where inmates may be in custody for periods lasting months to years.
Most persons in custody at the HCJ have been arrested by officers from the City of Houston or Harris County. Arrested persons who are taken to the Joint Processing Center where a hearing officer presides over a bail/probable cause session. Between the time of arrest and hearing, the arrestees are held in a common area. Those who make bail are released, and others remain in custody. After the hearing, an inmate is assigned to a unit, which may contain 15–60 persons. The exact unit placement depends upon available space, gang affiliation, racial tension, and fights.
Approximately 2500 persons are employed at the facility, including employees and contractors.
The jail administration agreed to work with HHD on mumps surveillance, unit isolation, and immunization of staff and detainees.
Cases
At HCJ, 60 suspect cases (57 inmates, 3 staff) were investigated between 11 June 2019, when HHD was notified and August 2019, when the outbreak was controlled. A comprehensive record review found the initial case occurred on May 16, 26 days before HHD was notified. Because of that delay, 1–2 incubation periods passed with possible spread in HCJ. Among 57 inmates, 18 (32%) cases were lab-confirmed, 13 (23%) were probable, and 26 (45%) were ruled out. Among staff, 2 cases were confirmed, and 1 was ruled out. Among inmates, the mean age was 34 years (range 20–57 years), and all were male. The 2 staff cases were both female, and both were 45 years old. Country of origin was not collected. Most cases (29/33, 87%) reported a history of childhood vaccination with MMR, but no records were available.
The cases were detected during a 3-month period between June and August 2019 (Figure 2). During that time, 37 units were placed on isolation as suspect cases were evaluated, and 8 were isolated more than once. A mean of 390 inmates were in daily isolation during the 3 months. Based on available information on inmate custody dates, 28 (84.8%) cases were exposed in the HCJ during their incubation period (12–25 days) before symptom onset.
Mumps cases, Harris County Jail, 2019.
Vaccination
During several vaccine clinics were held in the jail, HHD administered vaccine was administered to 549 inmates and 850 staff members, an acceptance rate of 31% of inmates and 34% of staff.
Links Between the Outbreaks
We were not able to identify any specific person who was in both facilities and were unable to determine if 1 or more local transfers may have seeded the jail outbreak or reintroduced mumps into HCDF.
DISCUSSION
The Houston mumps outbreaks followed the arrival of migrants from Honduras who were unvaccinated because mumps vaccine was not included in the immunization schedule [7] when they were children. In 2018, Honduras reported 14 761 mumps cases compared to 115 in 2017 [8], and the outbreak spread north with the surge in migration.
Two key lessons learned in these outbreaks are that public health officials need to have a working relationship with custodial facilities in their jurisdictions and vaccination alone will not control the outbreaks. The crowded conditions that promote outbreaks are not likely to change soon, and additional outbreaks are likely.
That knowledge about custodial facilities should include an assessment of who can make yes/no decisions about public health recommendations for outbreak control. Local health officials also need to understand medical screening requirements for new entrants, vaccine policies and infection control practices, and assess how isolation/quarantine measures might be implemented. In addition, assessment of the ability to implement a vaccination program in response to a specific disease merits consideration.
Timely notification with rapid intervention is essential in controlling any outbreak, but delays occurred in both outbreaks. HHD did not learn about either until almost a month had passed. The delays meant that up to 2 new generations of new mumps cases could have been incubating at HCDF and at HCJ.
Once reported, the detainee outbreak required 10 months of effort, and the jail outbreak required 3 months. Both facilities where cooperative once communication was established, but better baseline communication could have speeded up the intervention efforts and permitted tailoring advice to each facility’s circumstances.
HCDF implemented the recommended control measures but could not control the flow of migrants in and out of that detainee facility because local staff did not make the transfer decisions. HCJ faced the similar challenges with ongoing arrivals and departures of arrested persons but was able to make its own, local decisions about how inmates were placed, isolated, and released.
The impact of the mumps cases among migrants was significant. The HDCF cases were among the 898 mumps cases reported in 57 facilities scattered across 19 US states between 1 September 2018 and 22 August 2019 [5]. Private companies operated 34 facilities, 19 were county jails housing detainees, and 4 were ICE-operated. Most cases (758, 84%) were exposed while in federal custody [5].
When the mumps outbreaks began, neither Houston facility recommended or required MMR [1] for those in custody or staff. Both situations were like most jails, which offer minimal healthcare. Prisons where inmates have lengthy stays are more likely to offer immunizations, often for hepatitis B [9]. Vaccine acceptance in other outbreaks has been in the 30% range [10], like our results.
Persons in custody are often susceptible to vaccine-preventable disease; work in Australia has found more than half of those in custody were susceptible to measles and 40% to mumps [11]. The issues are even more difficult when assessing persons with unknown vaccination status [10]. The delay in notification to HHD meant that permitted further spread and incubation.
However, implementation of an immunization program in detainee/jail settings is problematic for several reasons. Recent work on coronavirus disease (COVID) vaccination has suggested that a highly efficacious vaccine will have suboptimal effects in high-spread, congregate settings where populations dynamics can reduce vaccine effectiveness and increase the reproductive number [12]. In addition, vaccine hesitancy in such populations may be up to 40% [13]. More than half of the 5110 participations in correctional/detention facilities would refuse (2318, 45.4%) or were hesitant (498, 9.8%) to be vaccinated for coronavirus disease 2019 (COVID-19) [14].
Any vaccine effort in an outbreak in these detainee/jail settings is likely to deal with an adult population whose immunization would not have been documented in an immunization registry, where data are often limited to children [15], and/or contains persons from other countries with unavailable vaccination information [10, 14].
Similar circumstances of overcrowding have led to reports of other infectious disease cases and outbreaks in both detainee centers and prisons. A sample of 22 facilities found 280 influenza cases, 1052 of varicella, and 301 of mumps [16]. Outbreaks of pneumococcal disease [17], measles [18], and mumps [19] have been reported in prisons. More recently, COVID outbreaks have been a major problem in both types of facilities [20, 21].
Other contributing factors to correctional outbreaks include inmate reluctance to report early symptoms of disease, limited staff capacity to deal with an outbreak and an emphasis of security over health needs [22].
The existing size of populations currently in custody at ICE facilities and in jails make it likely that local and state health departments will continue to deal with infectious disease outbreaks. In 2019, the average daily census in ICE facilities was 55 000 in 2019 [23], up from 42 000 in 2018 [24]. Local jails also remain crowded. IN 2019, local jails reported 10.3 million arrests, and at mid-year local jails contained 734 500 inmates [25].
As a practical matter, local public health officials need be aware of conditions and existing policies on immunization and medical screening in detainee centers and jails. Periodic communication between public health staff and officials at detainee centers and jails should include briefing on local outbreaks. Those discussions should include discussion on how isolation might be done and how vaccination plans could be implemented. Public health officials and detainee/correctional facilities should also realize that low vaccine acceptance rates could prolong an outbreak. Cooperative efforts may enable a more rapid intervention and earlier control of the outbreaks which seem likely to recur.
Notes
Financial support. No outside funding was obtained for this set of investigations. The authors conducted both investigations as part of the work of the Houston Health Department.
References
Author notes
Potential conflicts of interest. D. P. reports Substance Abuse and Mental Health Services Administration (SAHMSA) funding for Frontlines, an opioid and naloxone grant awarded to City of Houston ($1 million), unrelated to this work. All other authors report no potential conflicts.
All authors have submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Conflicts that the editors consider relevant to the content of the manuscript have been disclosed.
