Global Alert and Response (GAR)

WHO SARS Scientific Research Advisory Committee concludes its first meeting

22 October 2003

The WHO SARS Scientific Research Advisory Committee concluded its first two-day meeting in Geneva on 21 October. The meeting, which brought together more than 30 leading SARS researchers, aimed to identify the specific research most urgently needed to understand the disease better and prepare for its possible recurrence.

Participants agreed on the priority research questions to be addressed in the coming months. Questions fall into the main areas of epidemiology, laboratory diagnostics, outbreak management, case management, including treatment outcomes and infection control, and social impact. The research agenda, which will undergo further review by participants, is expected to be finalized for public release within a week.

Preparation of the research agenda was based on a discussion of several problems that emerged during the outbreak and an analysis of subsequent data, including areas where additional scientific evidence was needed to identify the most cost-effective control strategies. Participants expressed a great need to ensure that the world can recognize and respond to a recurrence of SARS in ways that work faster to achieve control and are less costly and socially disruptive.

Some of the main issues discussed

The global SARS alert system

Concerning the level of operational preparedness to detect new cases, participants noted that the system of alert now in place will probably work well to pick up clusters of severe disease. False-alarms are fully expected and have already occurred. They provide reassurance that the alert system is working well. However, participants agreed that sporadic cases of SARS may not be detected early.

Mild SARS cases and cases with atypical symptoms could be initially missed, hence the importance for several lines of defence in which the intensity of surveillance is determined by the level of risk (zone of re-emergence, nodal areas, and low risk areas). Moreover, the system for global alert is weakened by the fact that at least half of the world’s population has no access to services that can perform simple chest X-rays.

Some countries are continuing a policy of heightened alert for persons with a recent travel history to one of the initial outbreak sites. Failure to elicit such a history should not automatically exclude SARS as a diagnosis if other clinical, laboratory, and epidemiological features are consistent with the diagnosis. The Hotel Metropole incident, in which contact with a single infected person seeded the outbreak’s international spread, was a random event illustrating the limitations that travel history can have in the post-outbreak situation. Any major transportation hub near an international airport could be the scene for a similar seeding of international spread.

Preparedness in resource-poor settings

Following the global alerts of 12 and 15 March, most countries experiencing imported cases were able to prevent any further transmission or hold the number of cases to very small numbers, indicating the effectiveness of a high level of alert. It is hoped that, should SARS re-emerge, the disease would occur in small pockets of cases that could, with the present level of vigilance, be rapidly detected and contained.

A major concern is the occurrence of cases in resource-poor settings, where health infrastructures might not be able to cope with the demands of case detection, isolation, intensive care, and contact tracing. Outbreak dynamics in such settings could take on different dimensions than seen during the initial outbreak. In the absence of sophisticated hospital facilities and treatment procedures, which are known to amplify transmission, developing countries might not experience explosive increases in cases for the same reasons as seen elsewhere. Many other factors, such as crowding on wards and lack of adequate isolation facilities, could contribute to so-called “super-spreading” events.

Some discussion centered on ways to identify areas that might be at particular risk. One approach would be to assess air passenger flows to determine which cities are most vulnerable to international spread. Support for enhanced surveillance and preparedness plans could then be targeted to areas considered at greatest risk.


Diagnosis of SARS is difficult, and all currently available laboratory tests have limitations. At present, there is no single “gold standard” that can be recommended for laboratory testing. Different tests are appropriate for different stages of the disease. Another problem affecting the reliability of available serological tests is cross-reaction with other human coronaviruses, which are widespread.

The laboratory confirmation of two or more cases in the same health care unit during a ten-day period would have great significance in the post-outbreak setting, perhaps signalling the start of another global emergency. For this reason, and in view of the inadequacy of diagnostic tests, the experts recommended a policy of rigorous verification, whereby positive tests in one laboratory would need to be confirmed by an external, internationally recognized reference laboratory. A system of quality assurance of tests was also judged essential, as were internationally-agreed algorithms that could help clinicians make difficult diagnostic choices.

These issues are being discussed further at the SARS Laboratory Network Meeting on 22 October.

Diagnosis would also benefit from some method of identifying, perhaps through immunological or genetic markers, which patients might develop severe illness and which patients might contribute to super-spreading events. An additional concern is the development of diagnostic tests suitable for use in resource-poor settings.

Possible evolution of the SARS coronavirus

Studies of coronaviruses in a number of animal species indicate that some of these viruses readily cross from one species to others, lending weight to theories that the SARS virus originally jumped from some natural animal host to humans. Recent experimental studies have easily infected two animal species with the SARS virus, suggesting that the range of animal hosts where the SARS virus could “hide” between outbreaks is much wider than initially suspected.

In animal diseases caused by coronaviruses, the severity of disease can be exacerbated by such factors as co-infection with other respiratory viruses, including influenza viruses, some forms of treatment, and stress, especially during transport. Virus shedding is longer, and respiratory symptoms are more severe following exposure to aerosolized virus. These features may shed some light on super-spreading events, which remain poorly understood. In Hong Kong, for example, transmission was amplified in hospitals where certain aerosol-producing procedures, including the use of nebulizers, were performed during the earliest days of the outbreak.

Animal studies indicate that coronaviruses can be spread mechanically. This mode of spread may have contributed to events, such as the Amoy Gardens outbreak, in which some environmental contact with the virus clearly played a role.

Some coronaviruses are believed to survive in their animal hosts despite seasonal changes in temperature. Others have mutated into strains causing much milder disease. While these features could hold true for the SARS virus, much more research is needed before any conclusions about the future evolution of the SARS virus can be reached.

Amplification in hospitals: the role of infection control

The amplification of transmission within well-equipped hospitals was a striking feature of SARS. In some cases, staff became infected despite wearing full protective equipment. Participants presented further compelling evidence that certain sophisticated procedures, especially those that generate aerosols, greatly amplify the risks of infection; less sophisticated hospitals fared better. Some cases of hospital-based infection could be attributed to clear lapses in infection control. In this and several other cases, research in the social sciences could assist in improving behaviour while also helping staff cope with the considerable stress of dealing with a new disease, a new type of patient, and new requirements for care.

A disproportionately large amount of resources is needed to prevent transmission of airborne infections. The required infection control procedures are also extremely demanding for staff. As SARS is not thought to be an airborne infection, research should clarify the need for these measures compared with measures needed to prevent transmission via droplets and contact with body fluids. It was also agreed that the need for protective equipment is related to procedures, and that, in advance of another emergency, consideration needs to be given to equipment that provides a good standard of care and adequate protection while also being affordable and sustainable.

Laboratory biosafety

Better measures are needed internationally to prevent the recurrence of SARS following a laboratory biosafety accident. Health authorities should be encouraged to produce inventories of all facilities conducting research on the virus or holding samples. This information should then be communicated internationally to WHO.

Control interventions

Rapid case detection and isolation, along with contact tracing and follow up, were generally regarded as an effective way to deal with the emergence of a new disease. It reduces the pool of infectious individuals circulating in the community and, in the case of contacts, heightens the level of clinical suspicion, especially in the absence of reliable diagnostic tests. However, good support from information technologies is needed to manage the data required for contact tracing more efficiently. Research on the most effective approaches to contact tracing is also needed.

Another important need is for the real-time compilation of data from epidemiological, clinical, and laboratory sources, and their integration into a single dataset available to the international community. Access to such a database would contribute greatly to assessment of the effectiveness of all interventions, including treatment, in real time as a SARS outbreak – or outbreak of any other emerging disease – unfolds.