Figure 1

Figure 1. Total number of nasopharyngeal swab specimens tested for SARS-CoV-2 by reverse transcription PCR by camp and reason for testing as part of enhanced surveillance for COVID-19 in displaced persons’ shelters,...

During May 2020–October 2021, camps submitted a total of 6,190 specimens collected as part of enhanced surveillance (i.e., not as part of an outbreak investigation) (Figure 1). Of these, 2,091 (34%) were specimens submitted from persons in travel quarantine, 3,791 (61%) were patients with ARI, 129 (2%) were patients with ILI, and 179 (3%) were PUIs. In addition, 13,586 specimens were collected as part of outbreak response activities; 4,350 (32%) were specimens from close contacts and 9,236 (68%) were specimens collected in the community as part of active case finding. Surveillance tests performed per person varied from 0.02 in Mae La to 0.13 in Tham Hin.

A total of 14 COVID-19 outbreaks were detected in the camps during the 18-month surveillance period for a total of 1,342 cases reported (Table 2). In 10 outbreaks, <10 cases were identified; 5 outbreaks were limited to a single case. Five outbreaks were detected by testing done during travel quarantine, and 9 were detected by testing patients with ARI symptoms. The index cases for all 14 outbreaks were identified and laboratory confirmed. Probable introduction of COVID-19 into the camp was estimated to have occurred 1–2 weeks before detection for all outbreaks.

The first outbreak with >10 cases was at Tham Hin camp, Ratchaburi Province, in April 2021. At the time, Alpha variant was the predominant strain in Thailand. Case investigation found that the index case-patient had been visited by family members who circumvented travel quarantine. The index case-patient was a religious leader and had close contact with nearly 100 persons during the infectious period. The large number of high-risk close contacts overwhelmed quarantine facilities, so a house quarantine approach was started. Community isolation facilities were used for all close contacts who tested positive, regardless of clinical symptoms. A lockdown of the camp was instituted for 4 weeks after detection of this outbreak, in which only 1 designated person in each nonquarantined household was allowed to move about the camp to pick up food rations and other necessary supplies. After 6 weeks of intensive contact tracing, 110 total confirmed cases were identified, and the outbreak was considered controlled.

The number of outbreaks detected increased during August 2021 and continued until the time of this report in November 2021, after the wave of community transmission across Thailand from the Delta variant (Table 2). When outbreaks were detected in camps and confirmed by RT-PCR, active case-finding using ATKs was performed. Movement restrictions in certain camp sections were implemented on the basis of evidence of transmission in the general community. As the outbreaks grew in size, house isolation was implemented for patients with asymptomatic or mild infections, and teams were deployed to provide hygiene materials and daily check-ups on clinical status. Contact tracing, home quarantine, and testing of high-risk contacts continued.

Over 18 months during 2020–2021, a novel COVID-19 surveillance system was launched in 9 refugee camps along the Thailand–Myanmar border; this system tested >6,000 specimens and detected 15 outbreaks. The system incorporated national surveillance recommendations and adapted them for the camp-based setting, where human and physical resources are more limited than in other parts of Thailand. To account for these limitations, laboratory testing was expanded and offered to patients demonstrating symptoms of ARI and ILI to increase sensitivity of the surveillance to detect COVID-19. In total, 9 outbreaks were detected through testing of symptomatic persons at the camps’ clinics. In addition, testing of residents under quarantine after travel outside the camp detected 5 outbreaks during this period. This system operated in parallel with and was complementary to the existing camp HIS and national COVID-19 surveillance systems, and all cases were reported in the relevant systems.

Although direct comparisons of COVID-19 surveillance across different humanitarian settings is challenging because of differences in disease detection, reporting, and local outbreak conditions, reports from other countries offer other examples of functional case detection. In Greece, during the initial 9 months of the pandemic in 2020, a total of 25 outbreaks were detected in 39 refugee and asylum-seeker reception facilities with a total population of ≈60,000 (13). In Yemen, a community-based surveillance system generated 91 alerts and detected 5 COVID-19 outbreaks in an internally displaced population of 1,806 persons over a 5-month period (25). At Cox’s Bazar in Bangladesh, 3,084 cumulative cases had been reported out of 63,776 total tests performed as of September 2021, for a positivity rate of ≈4.8% (26).

The establishment and conduction of laboratory surveillance in the camps themselves was critical. The remote locations of several camps necessitated special transportation arrangements to preserve cold chain requirements and reach laboratories in appropriate times. Relying on testing through official channels would have led to delays in detection and outbreak response because of the challenges in transport and the more stringent PUI case criteria for testing by MOPH laboratories. Some patients who were tested met PUI criteria, but they were a small subset (n = 146), and no outbreaks were detected from PUI testing. Additional patients would possibly have met PUI criteria, but their exposure risk was either not assessed or they were not forthcoming about potential exposure risks.

Figure 2

Figure 2. Epidemiologic curve of the total number of laboratory-confirmed COVID-19 cases per week by displaced person camp, Thailand–Myanmar border, November 8, 2020–October 31, 2021. For reference, population sizes of each camp...

Thailand did not have widespread community transmission until mid-2021, when the Delta variant became the predominant strain. This timing afforded camp-based healthcare providers time to plan, recruit and train staff, and bring the enhanced surveillance system into action. During July–October 2021 alone, 11 outbreaks were detected. This number corresponded roughly to the high level of community transmission that was seen across Thailand during that time (Figure 2). In November 2021, several camps were experiencing growing outbreaks. Community resistance to distancing measures, isolation, and testing has been a factor in controlling spread and has been similarly described in other refugee communities (27). To build support in this community, risk communication and community engagement activities are ongoing.

A previous modeling paper by Gilman et al. (28) identified that the application of control measures, such as efficient isolation of infected persons, use of face masks, and limiting movement of camp residents between sectors, would be effective in limiting COVID-19 transmission. Similar control measures were applied and appeared to have an effect in Tham Hin camp. The outbreak during April–May 2021 started from multiple contacts of an infected person, which nearly overwhelmed the quarantine facilities that had been prepared. Speedy adjustment to the situation and the decision to use house quarantine for close contacts was critical to ensure that existing facilities could accommodate persons who tested positive. Active case finding through systematic screening by camp sections served to identify and stop unknown chains of transmission. Diligent contact tracing, community participation, provision of support to quarantined and isolated persons through food aid, and daily healthcare visits to quarantined households limited transmission; the outbreak was declared over with a total of 110 cases detected after 2 months.

Commercial ATKs were not approved for use in Thailand until July 2021 but were rapidly adopted as an essential tool because of their lower cost, rapid turnaround time, and lack of cold chain requirements. ATKs were particularly helpful because diagnostic laboratories were often distant from the camps, and sample transport and processing required 3–5 days. As an example, a close contact with a positive ATK result could be rapidly isolated and contact tracing could begin while RT-PCR results were pending. False-negative results, however, are commonly experienced with ATK tests because of their lower sensitivity, so RT-PCR testing was still relied upon for confirmation.

The enhanced surveillance system was subject to several limitations. Camp medical staff did not complete comprehensive examination forms for patients seeking care at the ARI clinic, so we could not evaluate whether patients were correctly classified as ARI, ILI, or PUI. Because testing of patients in the ARI clinic was voluntary, uptake varied and the number of tests performed might not accurately reflect the overall incidence of ARI and ILI; some COVID-19 cases could have been missed. Surveillance testing per population was nearly 5-fold greater in Tham Hin camp than in Mae La camp; this difference was related to several factors, including community acceptance of testing.

Similarly, the number of tests performed on persons in travel quarantine might not indicate the total number of persons who returned to a particular camp. Lags in test results and reporting could have caused discrepancies in the total number of COVID-19 cases described in the camps in this study compared with official numbers reported by Thai MOPH. Because ATKs are not as highly sensitive or specific as RT-PCR testing, some COVID-19 cases could have been missed, and the incidence of COVID-19 in the camps might be underestimated.

Despite many humanitarian settings having robust surveillance, more published reports are needed that describe such systems (29). A review of the literature covering COVID-19 surveillance found 2 other studies that describe implementation and adaptation to a humanitarian setting, in Yemen and Sudan (25,30). In Sudan, healthcare providers were trained as rapid response teams (30), whereas in Yemen a community-based surveillance system approach was used (25). The surveillance system we describe includes elements of community- and healthcare-based surveillance, in which community-based assistants perform contact tracing, identify persons with recent travel history, and refer persons with compatible illness for testing. In addition, our enhanced surveillance system also has an element based in existing clinics, with testing provided for persons experiencing symptoms of ARI and ILI.

COVID-19 surveillance in refugee, migrant, and displaced person populations continues long-term as successive waves of SARS-CoV-2 transmission continue worldwide and vaccine campaigns gradually increase their coverage. Refugees and displaced persons frequently have reduced access to public health services because of language barriers, location in remote areas, and healthcare systems that exclude noncitizens or unofficial residents. Because mobile populations might be more likely to move informally within a country or internationally, establishing surveillance to detect pathogens of international significance and extending national surveillance systems to these groups are vital. The enhanced surveillance developed in displaced persons’ shelters on the Thailand–Myanmar border is one such example and has provided a functional solution to this ongoing challenge.