SARS with the Benefit of Hindsight
Monthly Topics in Clinical Biosecurity Series
By Eric Toner, M.D., June 2, 2005
Severe Acute Respiratory Syndrome (SARS) first appeared on November 16, 2002 in the Guangdong Province of southern China. The disease was found to be caused by a novel virus, the SARS-related coronavirus (SARS CoV). Genetic analysis suggests that there were several distinct initial genotypes which probably arose as separate zoonotic events . It appears that these outbreaks originated in the live animal markets of the Pearl River Delta [MAP]. Although several animals in the markets were found to harbor viruses very similar to the human SARS, the virus was found most frequently in the Himalayan Palm Civet. It is presumed that the infection was imported into the animal markets from a wild animal reservoir, but, as yet, this reservoir has not been found.
Although it appears that at some time around November 2002, a mutation occurred in the SARS CoV precursor virus that enabled human infection, it was not until the beginning of January 2003 that efficient human-to-human transmission began. The result was a dominant genotype that led to the first pandemic of the 21st century . Over the course of 3 months, SARS spread to 29 countries around the world, with approximately 8,000 cases and 774 deaths. The vast majority of cases and all of the deaths were in adults, 50% of victims were infected in hospitals, and 21% of victims were healthcare workers.
One of the most remarkable features of the SARS outbreak was the phenomenon of superspreading events (SSE). Most victims of SARS passed it on to an average of <3 other people . However, 12 individuals spread the disease to an average of 60 people each. In the case of one SSE (Amoy Gardens), a single individual was responsible for >300 cases .On several occasions, including at Prince of Wales Hospital, a single patient caused nosocomial outbreaks of more than 100 additional cases. One symptomatic SARS patient transmitted the disease to >20 other passengers on a China Air flight . Some SSEs were clearly related to environmental factors such as medical procedures and peculiarities of high rise apartment building plumbing. Others may have been related to host factors such as age, the presence of diabetes, or renal insufficiency, while others appeared to be due to differences in the virulence of particular strains which caused chains of SSEs. Some events may have been the result of several of these factors. Although other diseases sometimes feature variable degrees of transmission, the degree of variation seen with SARS appears to be unique. It may be the case that as a single-stranded RNA virus, SARS CoV is highly mutable, resulting in multiple simultaneous strains which may vary in virulence. Although this occurs with other viruses, such as Influenza A, the high morbidity and mortality of SARS would make these variations more apparent. Another factor contributing to variability of transmission may be its association with age and co-morbid illness. Because the disease is more contagious in the elderly and infirm and the degree of infectivity is directly related to the severity of the symptoms, the virus had the opportunity to spread in places where elderly, acutely sick people were congregated, namely hospitals.
While it was immediately apparent that SARS was spread by a respiratory route, it has not been clear whether this was due to large droplets, which travel 3-6 feet and can be blocked by a simple mask, or small droplet aerosols that can spread over long distances and require a higher degree of protection such as an N95 mask and negative pressure isolation. There has been increasing evidence that SARS was at least partially spread by small droplet aerosols. This mechanism is the most plausible explanation of the SSEs involving the China Air flight and Amoy Gardens (above) [4, 6]. Studies of the Prince of Wales Hospital SSE strongly suggest aerosol (airborne) transmission of SARS to both medical students and other patients [5, 8]. A recent study from Toronto provides direct evidence of the presence of suspended aerosol droplets in the room of a SARS patient. The Toronto study also found SARS CoV on surfaces both in and outside the patient?s room . Clearly, airborne respiratory precautions are essential for SARS patients.
The association of the spread of SARS with modern hospitals is clear. Not only did half of the victims become infected in hospitals, but the spread within hospitals was clearly related to certain high risk medical procedures such as endotracheal intubation, bag ventilation, and non-invasive mechanical ventilation. It is also clear that transmission within hospitals could be limited by appropriate use of isolation, personal protective equipment, and training. This phenomenon of hospital-related amplification of an epidemic has occurred with other diseases as well, such as measles, chicken pox, tuberculosis, and smallpox, but the striking relationship between hospitals and the spread of SARS may be related to the close correlation between degree of infectivity and illness in SARS.
One has to wonder, however, how much these medical procedures contribute to nosocomial infections with more common diseases on a day-to-day basis. For instance, if respiratory precautions were used everyday, as they eventually were with SARS, would there be a significant reduction in nosocomial disease? What about diseases such as influenza or Respiratory Syncytial Virus (RSV) that are not typically considered nosocomial?might those infections be reduced through more vigilant precautions? These questions seem worthy of study.
Rapid action by regional, national and international public health agencies and laboratories led to the rapid discovery of the etiologic agent and containment of SARS. Had more open information been available in the first few weeks of 2003, it might have been controlled even earlier. The tools employed were the basic tools of public health: isolation, contact tracing, and quarantine. The national health agencies, doctors, and nurses in the affected hospitals, and the WHO should be congratulated for their success in containing SARS; however, this was a small "pandemic," and the affected nations cooperated with the WHO and allowed investigators in. It is not clear how much the basic biology of the SARS CoV contributed to the rapidity of its control. Furthermore, nosocomial rather than community transmission was the main cause of spread, and once that was realized and control measures were taken, interruption of the chain of transmission was more straightforward.
Novel human pathogens will certainly continue to emerge as interactions between humans and rare animals in remote locations increase, as evidenced by the emergence of HIV, Ebola, Lassa, Marburg and Monkeypox in the past 30 years. In addition, new and old diseases will continue to spread rapidly to new locations, as with West Nile virus. The challenge is to create a medical and public health infrastructure capable of meeting these threats. On the public health side, sensitive surveillance systems, rapid outbreak investigation teams, rapid laboratory investigation, and coordinated response regimes are needed, as is an underlying framework of agreements. On the medicine side, isolation capacity and procedures clearly need to be improved and utilized day-to-day. Patients and visitors with febrile respiratory illnesses should wear masks. Surge capacity needs to be increased and tools for early outbreak detection created. Mechanisms to share clinical information in real time need to be developed and/or improved.
Many of the issues raised by SARS are the same as those that would be faced following an intentional release of a contagious bioweapon or other emerging respiratory diseases such as pandemic influenza. It is important to implement what we have learned from SARS before we have another opportunity to learn the same lessons again.
"Molecular Evolution of SARS", Science 2004;303: 1666-1669
Guan, Peiris, Lancet 2004; 363: 99-104
Donnelly et al, Lancet ID 2004; 4 : 672-83
Ignatius et al, NEJM 2004; 350:1731-9
Ignatius et al, CID 2005:40 (1 may)
Olson et al, NEJM 349:2416-22
Booth et al, JID 2005:191 (1 may)
Wong et al, EID 2004;10: 269-276