During April 2018–March 2020, we enrolled participants on a rolling basis from 32 camel-owning households in Marsabit County, northern Kenya (Figure 1). We defined a camel handler as any person in the household who had contact with camels (Figure 2). This study was approved by the Scientific and Ethical Review Committee of Kenya Medical Research Institute (approval no. SSC3472), the Institutional Review Board of Washington State University (approval no. 16245), and the US Centers for Disease Control and Prevention (approval no. 7065). We obtained written informed consent from all participants.

We conducted monthly visits with the participants. At each visit, we collected nasopharyngeal and oropharyngeal swab samples from each participant. We also administered a questionnaire to each participating camel handler to identify signs and symptoms of possible respiratory illness during the previous 30 days. In addition, we recorded occurrences of respiratory illness among their household members. Participants belonged to 32 households with a median of 6 persons (interquartile range [IQR] 1–8 persons) and 32 camels (IQR 2–48 camels) at enrollment. The median age of these participants was 19 years (IQR 11–38 years). Most (67.2%) participants were male, of whom 39.3% were employed as camel workers and 38.2% were school going household members (Table 1). All participants handled camels. The most frequent interactions were herding (74.4%), cleaning barns (67.9%), feeding (67.6%), and milking (63.7%) (Figure 2).

We stored the swab samples in virus transport media and tested them for MERS-CoV by reverse transcription PCR (RT-PCR) at the Kenya Medical Research Institute (Nairobi, Kenya) as described previously (12). We conducted real-time RT-PCR selective for the upstream region envelope and 2 distinct regions of nucleocapsid genes on total nucleic acid extracted from 200 μL of sample. We defined a positive sample by positivity of all 3 PCRs.

We tested 1,369 samples from 262 camel handlers during the 2-year follow-up period. Participants had a median of 43.6% of monthly follow-up visits (IQR 8%–75%). Three (1.1%) participants (cases 1–3) tested positive for MERS-CoV by RT-PCR. The cycle threshold (C_t) values for case 1 were 38.9 for the upstream envelope, 37.7 for the nucleocapsid 2, and 39.3 for the nucleocapsid 3 genes; for case 2, the values were 39.7 for the upstream envelope, 36.9 for the nucleocapsid 2, and 39.8 for the nucleocapsid 3 genes; for case 3, the values were 35.6 for the upstream envelope, 36.0 for the nucleocapsid 2, and 36.8 for the nucleocapsid 3 genes. We detected all 3 cases during July–September 2019 (Table 2).

Case 1 was in a woman 20 years of age who enrolled in June 2019 and had 9 monthly follow-up visits. She participated in the study with 11 other members of the household, all of whom tested negative for MERS-CoV throughout the follow-up period. Case 2 was in a man 49 years of age who enrolled in May 2019 and had 7 monthly follow-up visits. He participated in the study with 6 of his 10 household members; all the participants in his household tested negative for MERS-CoV. Case 3 was in a man 22 years of age who enrolled in May 2018 and had 12 monthly follow-up visits. He participated in the study with 3 of his 9 household members; the participants in his household tested negative for MERS-CoV. None of the 3 with positive results tested positive for MERS-CoV in the subsequent months.

All of the 3 with positive results were asymptomatic at diagnosis and had no concurrent conditions or histo-ry of travel outside of the county or country in the previ-ous month. None of them or their household members had respiratory illness before or after diagnosis.

We report 3 PCR-confirmed cases of MERS-CoV in humans in Kenya; these cases met the World Health Organization case definition of MERS-CoV infection (13). All 3 persons were asymptomatic before and after diagnosis; this finding supports previous data suggesting that the virus causes no or mild disease in Africa compared with the Middle East and Asia, perhaps because of the younger age of most camel herders in Africa (4,8,9). Our findings are limited by the high C_t values (>35) of all cases, a level which some experts might not consider to be positive. However, because these patients had C_t values <40 for 3 distinct MERS-CoV genes, we feel confident that these are unlikely to be false positive results. Researchers have observed low upper respiratory tract RNA concentrations in asymptomatic patients and contacts of MERS-CoV patients (14). In contrast to studies conducted in the Middle East, we found no evidence of human-to-human transmission; a total of 20 household members of the 3 patients tested negative for MERS-CoV before and after their household member’s diagnosis. However, we might have missed some infections that occurred between follow-up visits. Furthermore, not all household members were enrolled in the study. In addition, serologic assessment of MERS-CoV T-cell responses might detect mild and asymptomatic MERS-CoV cases (15). Finally, the low (0.2%) seroprevalence among participants who had high exposure to camel herds with MERS-CoV circulation suggest a low level of zoonotic camel-to-human transmission. We previ-ously found no antibodies against MERS-CoV in camel herders despite high seroprevalence among camels in this community (9).

In conclusion, we confirmed zoonotic transmis-sion of MERS-CoV from camels to handlers in Kenya. Focused surveillance is needed to detect these rare in-fections when they occur.