Setting

Hospital A is a 368-bed tertiary acute-care facility and serves military staff and their families. Hospital B is a 500-bed tertiary MoH hospital with an associated but physically separate outpatient renal dialysis unit. Hospital C is a 250-bed MoH facility and is the MERS-CoV designated referral hospital for Taif. Hospital D is a private hospital.

Epidemiologic Investigation

We defined a case-patient as any patient from Taif Governorate who was reported with laboratory-confirmed MERS-CoV infection during August 1, 2014–February 1, 2015. In Saudi Arabia, reporting is required for all patients with clinical or radiologic evidence compatible with MERS-CoV disease and with a positive real-time reverse transcription PCR (rRT-PCR) on 2 specific gene targets: the region upstream of the E gene and open reading frame 1a (13). We reviewed available medical and public health records for all reported case-patients during the study period and conducted interviews with available hospital staff. We collected available demographic information, medical history, symptoms at onset, clinical course, preillness exposures, and evaluation and treatment locations. We grouped together case-patients whose illness onset occurred within 2–14 days of exposure (work- or treatment-related) to the same facility. Available residual patient specimens were analyzed at CDC.

Laboratory Investigation

Statistical Analysis

For reported demographic and clinical characteristics, differences were assessed for significance (p = 0.05) by using χ² test, Fisher exact test, and t-test, as appropriate. All data were analyzed by using SAS 9.3 (SAS Institute, Cary, NC, USA).

Epidemiologic Investigation

Figure 1

Figure 1. Middle East respiratory syndrome coronavirus (MERS-CoV) case-patients reported in Taif, Saudi Arabia, during September 2014–January 2015. Indicated are time of symptom onset or first positive laboratory testing. Healthcare setting where transmission...

During August 1, 2014–February 1, 2015, the MoH received reports of 38 patients with laboratory-confirmed MERS-CoV (Figure 1). Twenty-eight (74%) were men, 22 (58%) were of Saudi nationality, and median age was 51 (range 17–84) years (Table 1). Thirteen (34%) patients were HCP: 7 nurses, 2 physicians, 2 cleaning personnel, 1 administrative professional, and 1 clerk. The most common underlying medical conditions were diabetes, reported by 16 (47%), and renal failure requiring dialysis, reported by 12 (33%). At illness onset, 35 (92%) patients reported >1 respiratory symptom. Two patients, both HCP identified through routine testing of contacts of previously identified patients, reported no symptoms (Table 1).

Twenty-one (55%) of the 38 patients died, all in the hospital. Deceased patients were significantly older than survivors (median age 60.4 vs. 39.4 years; p = 0.001) and were more likely to be men (90% vs. 53%; p = 0.023) and Saudi nationals (76% vs. 35%; p = 0.020). Median time from onset to death or discharge was 17 (range 1.0–84.0) days. HCP patients were more likely than non-HCP patients to be women (54% vs. 12%, p = 0.016), non-Saudi (92% vs. 16%, p<0.001), and younger (median age 37 vs. 65 years; p<0.001); they were also more likely to survive (85% vs. 24%, p<0.001). Two of the 13 HCP patients died. Both were non-Saudi men: a 40-year-old physician with no underlying medical conditions and a 46-year-old information technologist with a history of smoking and hypertension.

Of the 38 MERS-CoV patients reported and investigated during the outbreak period, 33 were associated with 4 facilities (Figure 1). We were unable to link 5 patients epidemiologically to other patients (Table 2).

Hospital A

The first patients in this outbreak were reported from hospital A. Of 10 patients associated with this hospital, 6 had illness onset during September 5–October 2, and 4 had onset during November 20–December 10; patients were tightly clustered in time during these 2 periods (Table 2). The initial patient reported was a 45-year-old male military employee with an unspecified exposure to an outlying farm. Onset of cough, shortness of breath, and fever began on September 5, 2014; he was admitted with respiratory compromise to hospital A on September 10. During September 13–October 2, five additional patients were reported, including 2 HCP employed by the hospital and 2 long-term care inpatients with no community exposures. During November 20–December 10, four additional patients were reported; 3 were HCP. One long-term care patient was admitted to the hospital on June 30, had MERS symptom onset on October 3, and died on December 15, 2014; his hospital course spanned both periods of clustered patients at this facility. Five HCP were among this hospital’s clusters: a 37-year-old male clerk who was married to an intensive-care nurse; a 40-year-old male cleaner; a 29-year-old female nurse; and a 40-year-old male physician and a 46-year-old male working in information technology, both of whom died. Of this hospital’s 10 reported patients, only the 29-year-old nurse had recognized contact with a known MERS-CoV patient before her illness onset. Six (60%) of the 10 patients died during their hospital course: 5 (83%) of 6 patients during the first transmission period and 1 (25%) of 4 patients during the second period.

Hospital B Dialysis Unit

Hospital B reported 15 patients from its outpatient renal dialysis unit, which was located in a building separate from the acute-care facility. When the outbreak occurred, the dialysis unit had 58 dialysis machines in 8 common rooms, 71 nursing staff, and 377 registered patients receiving periodic hemodialysis. For the 15 patients reported in this cluster, onsets occurred during October 4–27, 2014. Eleven were dialysis patients, and 4 were dialysis unit HCP. The first recognized patient associated with this setting was a 53-year-old man with onset of MERS-related symptoms on October 4, 2015. He underwent dialysis on October 4, 6, 8, and 9 in a 9-bed common room while he was symptomatic. During October 13–28, ten dialysis patients and 3 HCP became ill, and their NP specimens tested positive for MERS-CoV. A fourth HCP reported no symptoms, but his NP specimen was confirmed by rRT-PCR to be MERS-CoV positive on October 25, after RT-PCR screening of NP specimens from identified HCP contacts. Of the 4 MERS-CoV–confirmed HCP, 2 reported working in the dialysis unit while symptomatic on October 18, 20, and 27, just before their MERS-CoV confirmatory testing.

On October 22, infection-control practices were changed on the basis of an onsite assessment by MoH Infection Prevention and Control staff. The changes included screening patients for fever and respiratory symptoms before admission to the dialysis unit; eliminating waiting and prayer areas; discouraging early arrival for dialysis; enforcing a no-visitation policy; increasing distance between patients undergoing dialysis (by reducing number of beds from up to 9 to 6 per room); establishing isolation of dialysis patients with respiratory symptoms; and providing additional infection-control training for staff.

The 15th patient reported from this cluster had illness onset on October 27, after changes were implemented. Of patients in this cluster, 8 (72.7%) of 11 non-HCP died; the 4 HCP survived. Besides the 15 patients reported from this facility, a 17-year-old man who underwent dialysis at this facility on October 4, 6, 8, 11, and 13 reported symptom onset on October 14; his NP specimen was confirmed positive on October 18, after he traveled to Riyadh and was admitted to a hospital there.

Hospital C

On October 3, a 60-year-old man was transferred to hospital C from an outlying hospital in Taif Governorate after respiratory symptoms developed on October 1 and laboratory testing of his NP specimen confirmed MERS-CoV on October 3. He was transferred to a hospital in Jeddah on October 5 and died there on December 25. On October 11, 15, and 17, three HCP (2 nurses and 1 physician) became ill and were hospitalized at hospital C. Each eventually recovered and was discharged, and no further cases were reported from hospital C.

Hospital D

On November 1, a 75-year-old woman was transferred to hospital D and admitted to the intensive care unit. She had been evaluated at hospital C on October 22 and November 1 for respiratory complaints and fever. Laboratory testing at hospital D confirmed her NP specimen as MERS-CoV positive on November 3; she was transferred back to Hospital C on November 4 for MERS-CoV treatment and died there on November 9. During HCP contact screening on November 4, an NP specimen from the cleaner of her room at hospital D on November 1–4 was confirmed as positive for MERS-CoV by RT-PCR. He denied symptoms consistent with MERS. On November 8, respiratory symptoms developed in the patient’s 22-year-old grandson; his NP specimen tested positive for MERS-COV on November 11, and he died on December 14. On November 11, an 81-year-old inpatient staying on the same floor where the initial patient received care had onset of respiratory symptoms, and her NP specimen tested positive for MERS-CoV. She died on November 20.

Additional Cases

Five cases were unlinked to cases reported from the 4 hospitals. The first case-patient was a 65-year-old male retiree with a history of diabetes, heart disease, smoking, and hypertension. After shortness of breath and fever developed on September 13, he sought care at a private hospital on September 17. His NP specimen tested positive for MERS-CoV, and he was referred to hospital C the same day. He was discharged on October 8.

The second case-patient was a 72-year-old male taxi driver with a history of smoking. On November 15, fever developed, followed by sore throat, vomiting, and respiratory failure 2 days later. On November 19, he was admitted to hospital C, where pneumonia was diagnosed, and his NP specimen tested positive for MERS-CoV. He died on December 8.

The third case-patient was a 76-year-old male farmer with a history of diabetes, heart disease, and hypertension. Fever and respiratory symptoms developed on December 27, and he was admitted to hospital C on December 31. MERS-CoV was confirmed by laboratory testing on January 1, and he died on March 3, 2015.

The fourth case-patient was a 33-year-old man who had a history of diabetes and worked as a security guard for a private home. Cough, fever, and headache developed on December 28; on January 7, 2015, he was admitted to hospital C, where pneumonia was diagnosed and his laboratory specimen was MERS-CoV positive. He was discharged on January 19.

The fifth case was a 73-year-old male retiree with diabetes and hypertension. Fever, shortness of breath, nausea, vomiting, and gum bleeding developed on January 4, 2015, and he was admitted to hospital D on January 9 and transferred to hospital C on January 13, 2015. NP specimens collected on January 11 and 19 were positive for MERS-CoV. He died on January 20.

Laboratory Investigation

Molecular Detection and Spike Gene Sequencing

CDC performed laboratory confirmation of MERS-CoV by rRT-PCR on acute-phase serum samples from 17 patients whose NP specimens had been previously confirmed positive for MERS-CoV by RT-PCR (Technical Appendix Table). Median number of days from symptom onset to serum collection was 3.5 (range 0–18 for 16 patients). Serum samples from 15 (88.2%) patients, including 4 samples collected 9–18 days after symptom onset, were confirmed to be positive by rRT-PCR by at least 2 independent assays. The mean rRT-PCR cycle threshold (C_t) value for a region upstream of the E gene from 9 respiratory specimens was 26.4 (range 17.5–37.9), compared with 34.8 (range 31.1–38.1) for acute serum samples collected on the same day. In general, patients with low C_t values (proxies for virus load) in respiratory specimens also had low C_t values in serum samples.

Figure 2

Figure 2. Midpoint-rooted phylogenetic tree inferred from multiple-sequence alignment of 10 new cases of Middle East respiratory syndrome coronavirus (MERS-CoV) spike open reading frame sequences (4,062 nt) from Taif, Saudi Arabia (brackets). Colored...

Because of limited available serum volume and generally low virus loads, we focused sequencing efforts on the MERS-CoV spike gene, which has been shown to be a reliable proxy for virus genotyping (16) and encodes the receptor-binding domain responsible for attachment to host cells. Sequencing of the spike gene coding region was attempted on all rRT-PCR–positive specimens; complete sequences were obtained from serum samples of 10 patients. The mean N2 rRT-PCR C_t value of serum samples that were successfully sequenced was 32.8 (range 31.1–35.6), compared with 37.8 (range 35.6–40.6) for samples with failed sequencing. Phylogenetic analysis of the 10 Taif spike sequences showed that the viruses formed a single, discrete cluster located within the Hafr-Al-Batin clade (17) and were most closely related to MERS-CoV viruses circulating in Riyadh during 2013 and 2014 (Figure 2). Sequences from 6 patients (1 from hospital A, 4 from hospital B’s dialysis unit, and 1 from hospital D) were identical, and all 10 sequences possessed 2 defining base substitutions at positions 3,670 (G>A) and 3,840 (C>T) (Technical Appendix Table). Sequences from 2 epidemiologically linked cases (patients 27 and 30 in the Technical Appendix Table) associated with hospital D formed a subcluster among the Taif viruses on the basis of 2 defining base substitutions at positions 1,679 (C>T) and 3,496 (G>A).

Five unique nucleotide substitutions conferring predicted amino acid changes were identified among the 10 sequences, of which 3 (E536K, D537E, T560I) were located in the spike protein receptor-binding subdomain that directly interacts with the dipeptidyl peptidase 4 receptor (18) (Technical Appendix Table). Random coding changes in the MERS-CoV spike protein may be functionally inconsequential or may confer selective advantage by enabling greater adaptation to the host and possibly enhanced virus transmission (18).