To the Editor:—  The World Health Organization issued a global alert on atypical pneumonia and severe acute respiratory syndrome (SARS) and the risk to healthcare workers on the March 12, 2003, after reports from Hong Kong, China, and Vietnam.* This letter reports the changes in anesthetic practice in our hospital during the 2-month outbreak in Singapore.

All patients who had fever and dyspnea were treated as suspect SARS patients, even if they had no SARS contact history. All elective surgery was postponed during the outbreak; only emergency and cancer operations were performed. The operating complex reception staff completed a checklist (table 1) and checked the temperature of all patients. Patients with unexplained temperatures 37.5°C or higher had surgery delayed until a cause was established. Patients were categorized as low-, medium-, and high-risk patients (table 2).

All staff were provided with thermometers to check and report their temperature thrice daily. Anyone with a temperature 37.5°C or higher had to stop work and have a checkup at the staff clinic. All staff kept a log of all patients they had contact with, to facilitate contact tracing, and the information was collated daily centrally. The anesthesiology and operating room (OR) staff were grouped into two teams working alternate stretches, so that if one team became infected, the other team could take over and maintain the service. Contact between the two teams was strongly discouraged, and all department meetings were postponed.

All staff had to pass an assessment on the correct use, removal, and disposal of personal protection equipment. At mask-fitting and leak-testing exercises, different makes of masks were tried, and those who were unable to achieve a good mask fit were posted out of the OR to areas with low risk of SARS exposure. For low-risk patients undergoing surgery in the general ORs, all OR staff donned N95 masks, goggles, splash-proof gowns, and gloves. The anesthesiologists practiced double gloving and removed the outer gloves immediately after any invasive or airway procedure before handling clean equipment. Eight medium- and high-risk patients were considered SARS exposed or suspect SARS patients, and they underwent surgery in the isolation OR. These patients were transferred into the operating complex wearing an N95 mask if they had no dyspnea, and the anesthesiologists used additional protection of powered air-purifying respirator systems (3M Jupiter Air filter unit and HT-101 hood; 3M UK PLC, Bracknell, United Kingdom; or T4 Personal Protection System; Stryker Instruments, Kalamazoo, MI) and surgical gowns. Probable SARS patients were not operated on in this hospital but were transferred to a designated SARS hospital.

The OR at the end of the last bank of ORs, closest to an entry point into the operating complex, was designated the isolation OR. Access to this OR was reduced to two points with double doors, and the other doors were sealed. All equipment not used during a particular operation was removed before patient entry. To minimize turbulence, all equipment needed for the procedure was kept within the OR rather than in the preparation room.

Our ORs are usually at a higher pressure (0.038–0.075 mmHg) with respect to the scrub room, preparation room, and corridors so that air flows from a clean to a less clean area. Although this system may help to prevent surgical infection, it may hinder containment of a virulent infection within the isolation OR. To reduce the spread of potentially contaminated air out of the isolation OR, its ventilation system inflow was reduced and the exhaust was increased, making its pressure 0.038 lower than its scrub and preparation rooms, 0.075 lower than the other ORs, and neutral with the corridor. With this rebalancing, the air changes were reduced from 334 changes/h over the operating zone and 56 changes/h for the whole OR to 192 and 36 changes/h, respectively. Although the shielding over the surgical zone was no longer as efficient, the laminar airflow still ensured that there was minimal turbulence over the surgical zone. Pressure measurements and smoke tests performed in accordance with Center for Disease Control guidelines showed that there was no escape of air out of the isolation OR. 1 

To reduce aerosol production and transmission, and equipment and environmental contamination, we used regional anesthesia whenever possible, and when general anesthesia was required, we used low fresh gas flows, including during bag–mask ventilation. We stopped using nebulizers in the operating complex. We used and changed a heat–moisture exchanger filter, a second microbial filter at the end of the expiratory tubing, circle system tubing, and carbon dioxide absorbent after every patient. To reduce wastage, only one absorbent canister was filled to a quarter. It was not practical to change the fixed tubings and ventilator bellows of the anesthetic machine.

When it was safe to do so, tracheal intubation was performed under full paralysis, and patients were extubated “deep” to minimize coughing and gagging. When disconnection of the patient and the circuit was required to facilitate positioning, the disconnection was made, leaving the heat–moisture exchanger filter attached to the endotracheal tube or laryngeal mask. Low-risk patients recovered in the PACU, whereas medium- and high-risk patients undergoing surgery in the isolation OR recovered fully there and were then transferred to the ward, avoiding the PACU.

Because SARS is a new disease, these changes were not evidence based. Although no patients or staff were infected in the OR, we are unable to determine which interventions were necessary, because everything was done at once. In particular, grouping of staff into cohorts may not have been useful, and these changes have greatly affected operating efficiency and cost.

Guidelines for preventing the transmission of mycobacterium tuberculosis in health-care facilities. Mortality and Morbidity Weekly Report 1994; 994/43:1–132