key: cord-272135-a09bf50o
authors: Brouqui, Philippe; Puro, Vincenzo; Fusco, Francesco M; Bannister, Barbara; Schilling, Stephan; Follin, Per; Gottschalk, René; Hemmer, Robert; Maltezou, Helena C; Ott, Kristi; Peleman, Renaat; Perronne, Christian; Sheehan, Gerard; Siikamäki, Heli; Skinhoj, Peter; Ippolito, Giuseppe
title: Infection control in the management of highly pathogenic infectious diseases: consensus of the European Network of Infectious Disease
date: 2009-04-22
journal: Lancet Infect Dis
DOI: 10.1016/s1473-3099(09)70070-2
sha: 
doc_id: 272135
cord_uid: a09bf50o

The European Network for Infectious Diseases (EUNID) is a network of clinicians, public health epidemiologists, microbiologists, infection control, and critical-care doctors from the European member states, who are experienced in the management of patients with highly infectious diseases. We aim to develop a consensus recommendation for infection control during clinical management and invasive procedures in such patients. After an extensive literature review, draft recommendations were amended jointly by 27 partners from 15 European countries. Recommendations include repetitive training of staff to ascertain infection control, systematic use of cough and respiratory etiquette at admission to the emergency department, fluid sampling in the isolation room, and analyses in biosafety level 3/4 laboratories, and preference for point-of-care bedside laboratory tests. Children should be cared for by paediatricians and intensive-care patients should be cared for by critical-care doctors in high-level isolation units (HLIU). Invasive procedures should be avoided if unnecessary or done in the HLIU, as should chest radiography, ultrasonography, and renal dialysis. Procedures that require transport of patients out of the HLIU should be done during designated sessions or hours in secure transport. Picture archiving and communication systems should be used. Post-mortem examination should be avoided; biopsy or blood collection is preferred.

Over the past two decades, many new and re-emerging diseases have posed threats to public health and have provided new challenges for infectious disease researchers worldwide. Expansion of human populations has caused both a greater proximity to wildlife habitats, resulting in the emergence of new zoonoses, and a massive urbanisation process, which facilitates the rapid spread of communicable diseases in human beings. Travel across the world has become increasingly frequent, resulting in the ever-increasing risk of worldwide contagion spread. These emerging problems were highlighted during the severe acute respiratory syndrome (SARS) epidemic. 1 Imported highly infectious diseases (HID), such as Lassa fever and viral haemorrhagic fever, have been reported many times in the literature but have seldom been at the origin of an outbreak. 2,3 Terrorist attacks with biological agents pose a substantial threat to the safety, health, and security of the citizens of every country.

As defi ned in the 2007 issue of the Biosafety in Microbiological and Biomedical Laboratories manual, 4 "group 3 agents are pathogens that usually cause serious human or animal disease but do not ordinarily spread from one infected individual to another and eff ective treatment and preventive measures are available," whereas "group 4 [agents] are pathogens that usually cause serious human or animal disease and that can be readily transmitted from one individual to another, directly or indirectly. Eff ective treatment and preventive measures are not usually available." Several laboratoryassociated infections with group 3 and 4 pathogenic agents have already been reported. [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] Although laboratories that handle group 3 and 4 agents should comply with biosafety regulations, laboratory leakage can happen at any time, for example, when working with a known agent or when attempting to isolate an unknown infectious agent, such as occurred with mimivirus. 21 Experience shows that the recognition and isolation of a new infectious agent is often followed by a reported laboratory-acquired infection caused by the new isolate, as was reported for SARS. 22, 23 The infection of a single laboratory worker with a highly infectious agent could be the origin of an outbreak, particularly if the agent has the capability of human-to-human transmission (ie, SARS-associated coronavirus). 23 In some situations, such as a cough in so-called "superspreader" patients with extremely drug-resistant tuberculosis, smallpox, or SARS-associated coronavirus, or exposure to infected blood of a patient with late-phase haemorrhagic fever, the inoculums to which health-care workers are exposed are likely to be equivalent to those received by a laboratory worker during specimen handling. The care of such patients should consequently be administered in a way to ensure the same level of protection and safety to health-care workers as to laboratory workers exposed to the same agent.

The European Network for Infectious Diseases (EUNID) comprises 30 national representatives and experts from 16 European member states and is funded by the European commission, within its Public Health and Risk Assessment Programme. 24 It was created to exchange information, share best practices, develop training, and improve the connection between national and regional Review infectious disease experts. This network of clinicians, public health epidemiologists, microbiologists, infectioncontrol and critical-care clinicians, who are experienced in the management of HID, represent national or regional infectious diseases units designated to care for patients with HID. One of EUNID's agreed tasks was to develop a consensus statement on the design and operation of high-level isolation units (HLIU) in Europe. 25 An HID is transmissible from person to person, causes life-threatening illness, presents a serious hazard in health-care settings and the community, and therefore requires specifi c control measures. 25 An HLIU is defi ned as a health-care facility that is specifi cally designed to provide safe, secure, high-quality and appropriate care, with optimum infection containment, infection prevention, and control procedures for a single patient or a small number of patients who have, or who may have, an HID. These units have also been described as biocontainment units by US experts. 26 We defi ne an isolation room as a single-bed room with negative pressure.

Our task was to develop a consensus statement for infection control during clinical management and invasive procedures in HID patients admitted to an HLIU. To assess this task, an expert (PB) was co-opted and charged by EUNID to do an extensive review and draft recommendations on infection control in the situations mentioned above. The draft recommendations were amended by the coordination team, and further refi ned jointly by the 27 partners from 15 European member states and the coordination team. The fi nal consensus was reached and validated during two consecutive meetings in Rome in May and October, 2007. The draft manuscript was then shared within the network by e-mail and revised to incorporate comments and additional evidence provided by participants until a fi nal version was reached.

Among the several possible scenarios that may be considered, epidemics outside Europe of a yet unknown contagious agent or a known group 3 or 4 agent, such as SARS-associated coronavirus or viral haemorrhagic fever, are the most likely to occur. Alternatively, a laboratory worker may become sick after being exposed to a known agent in a registered biosafety level 3 or 4 laboratory while doing his or her duty, such as in the last SARS outbreak in China and Singapore. 23, 27 A third possible situation is intentional release of a bioterrorism agent. If an outbreak of a human-to-human transmissible disease begins in one country, the fi rst case it likely to be missed. Thus, other implementations, such as routine respiratory and hand hygiene in health-care settings, healthcare personnel surveillance, and prompt reporting of such patients to national public-health authorities are mandatory. 28 The decision to care for patients in an HLIU is based on the capability of the agent to have human-to-human transmission, its transmission rate, and the availability of primary or secondary prophylaxis, such as vaccines or eff ective antimicrobial therapy. Although the role of 

Hospital-acquired pneumonia 21 

HLIU=High-level isolation unit. SARS=severe acute respiratory syndrome. *As defi ned by the Centers for Disease Control and Prevention, WHO, and the European Community Directive. 38 †Note that most of the guidelines are based on a very small number of clinical cases and that level of evidence and grading are not necessarily accurate. ‡In vaccinated personnel. §Not an offi cial classifi cation, but we recommend this level because laboratory-acquired pneumonia has previously occurred. 

Emergency departments of general hospitals are the primary units where interaction with an unknown HID is most likely, and because many such units were not prepared for this kind of medical situation, they paid a heavy price during the SARS outbreak. [46] [47] [48] [49] Even now, most of our hospitals are still not prepared to face these kinds of situations. 50 Until a suitable network of care for such patients, with communication between health-care workers prior to referral, becomes eff ective in each country, patients suspected of being infected with a highly contagious agent are more likely to be referred to the emergency department of a general hospital by their general practitioner. As a consequence, the emergency department of any hospital should be prepared for such events, and both training and structural features should be implemented. 28 Based on studies of SARS transmission, measures designed to control respiratory droplets and secretions, along with hand hygiene, would seem to off er signifi cant protection to other patients and health-care workers who have close contact with source patients. 51, 52 Given the challenge of recognising early HID cases, and considering the potential for the spread of respiratory infections in health-care settings, the US Centers for Diseases Control and Prevention (CDC) recommended a broader strategy to prevent health-care-associated transmission of respiratory illnesses in response to SARS. 53 In addition to standard precautions, the CDC suggest that an effi cient measure to reduce the risk of infection transmission needs to be systematically implemented by health-care workers. The CDC described the new standard approach to manage patients with febrile respiratory illness as "respiratory hygiene" or "cough etiquette". 54 Patients with cough and fever should be encouraged to report symptoms at admission and should be encouraged or asked to wear a surgical mask and wash their hands after contact with respiratory secretions. These patients should be separated from other patients in the waiting area, and should be examined and assessed as soon as possible by the emergency staff in a single room. 54 Signs should be displayed in the waiting areas to promote these measures and to educate patients and health-care workers, and emergency staff should comply with droplet precautions.

However, because the modes of infectious agent transmission are often underestimated, as was recently reported for infl uenza and SARS, 55 and because tuberculosis cannot be identifi ed without biological testing, EUNID recommends that droplet precaution should be upgraded to airborne precaution each time Situations in which a patient would need to be admitted to an HLIU • Patients with an unknown human-to-human transmittable or a potentially transmittable epidemic febrile illness that is native or imported from abroad • Patients with a known infectious disease caused by a group 3 or 4 agent* At admission of patients with HID to an emergency department • Systematically apply standard precautions and cough and respiratory etiquette • Set up at least one single room with a dedicated route and direct access, or an isolation room as recommended by EUNID for a referral hospital, † if HLIU cannot be used for ruling out HID diagnoses • Off er special training to the emergency department team • Retain close relationships with the HLIU team of the referral hospital 

• For infection-control reasons, all children suspected to be infected with an HID should be admitted to an HLIU • Family participation should be minimised • The HLIU responsible manager should make all eff orts to be prepared and be able to provide nursing care compatible with children's requirements

• If possible, perform intensive-care therapy in the HLIU in collaboration with the infectious disease team; the HLIU should be pre-equipped for critical care • If a patient with HID is cared for in the intensive-care unit, the unit should be subjected to negative pressure • Manual ventilation duration during resuscitation procedures should be reduced to a minimum • Use NPPV instead of facial mask aerosol therapy when possible and intubation/ mechanical ventilation instead of NPPV when safely achievable with maximum precautions • Endotracheal intubation should be done with rapid sequence induction by the most skilled person available, who should wear personal protective equipment • Meticulous infection-control measures must be followed in case of ventilated patients with HID, particularly during suctioning, tracheotomy, and bronchoscopy with or without bronchoalveolar lavage HLIU=High-level isolation unit; NPPV=non-invasive positive pressure ventilation. *As summarised in tables 1 and 2. †A hospital to which the patients with HID are usually referred to either by national authority directives or by their own expertise and reference. ‡This remains a matter of debate in our group.

that involvement of one of these agents is suspected. Consequently, chest radiography should be done separately from other patients, a systematic examination of sputum for acid-fast bacteria should be prescribed when pneumonia is diagnosed (to rule out tuberculosis), and the patient should remain in isolation until the threat of tuberculosis is eliminated. 56 Patients with suspected HID should be placed directly in an emergency department isolation room (EDIR), if available. During admission, such patients should avoid any contact with other patients and unprotected healthcare workers, and therefore direct access to the EDIR from outside the unit is required. 57 EDIRs should comply as much as possible with the design and operational management recommendations made by EUNID for HLIU. 25 Whereas general respiratory hygiene rules and cough etiquette should apply to every emergency department of every general hospital, isolation rooms could be available in referral hospitals only. A patient with a possible or confi rmed HID, if not admitted directly to an HLIU, should be transferred from the EDIR to the HLIU in a secure manner, by use of a safe isolation transportation system, 58 or at least by staff wearing personal protective equipment and using a clear and secure route. Although these general recommendations may be diffi cult to apply in the usual understaff ed, overworked, and busy environments, implementation of these measures is important. These recommendations would also help prevent the transmission of many other important pathogens that are spread by the droplet route, such as infl uenza and Mycoplasma pneumoniae. 53

Clinicians should remember that the most likely cause of fever in a tropical traveller is malaria, which is far more common than a new or emerging HID. Consequently, EUNID fi rst advocates that everything possible should be done to rule out the most frequent diff erential diagnoses without delay. To reduce the risk of transmission to health-care workers, samples from patients should be taken in the EDIR or HLIU, depending on availability. Severely ill patients may necessitate frequent blood sampling and intravenous line placement for reliable antimicrobial or antiviral drug administration. To reduce exposure due to accidental needle-stick incidents, we recommend the use of a routinely secured arterial line and central venous line access by an HID-trained physician, thus allowing safe serial blood sampling and drug administration without further needle-based procedures.

If possible, all routine diagnostic tests should be processed in a biosafety level 3 or 4 laboratory that is located close to the HLIU or on the same campus to avoid unnecessary transportation of contaminated samples. 4, 25, 59 Once inactivated, the sample can be processed in a routine clinical laboratory with PCR or blood fi lm analysis. Under certain handling precautions, the use of a certifi ed autoanalyser in routine tests has been suggested as being safe. However, some experts believe that samples that are likely to be highly contagious, such as blood contaminated by Ebola virus, cannot be handled safely in a routine laboratory. The development of a pointof-care test, such as for arterial blood gas, blood electrolyte, or haemoglobin content, and more recently for microbiology (ie, malaria), or for early diagnosis of lower respiratory-tract infections, off ers an alternative to routine tests because they can be done at patients' bedsides. 60

Hospital-acquired infections can cause major problems in paediatric wards, and compliance with isolation procedures needs to be ensured. 61, 62 During the SARS epidemic, infection-control measures overshadowed family-centred nursing practices in the management of paediatric patients, 63 and created inevitable confl ict. 64 To eff ectively control infection, family participation should be minimised, and all children with suspected HID should be admitted to the HLIU. Consequently, everything should be organised in the HLIU to provide nursing care that respects the privacy of the parents and children.

The risk of being infected with SARS-associated coronavirus was reported to be about 13-times higher among physicians and nurses who performed or assisted in endotracheal intubations in intensive care units (ICUs) than in those who did not. 65 Nurses who became ill were often exposed to SARS-associated coronavirus within 48 h of a patient's admission, during which time the patient usually deteriorated with symptoms, increasing the spread of droplets or aerosols (eg, dyspnoea, cough, etc). 66, 67 Patients admitted to the ICU are usually severely ill, and are likely to have a high viral load and to be at a point of maximum infectiousness. 68 Non-invasive positive pressure ventilation (NPPV) is a mode of ventilation assistance used in early acute respiratory failure and acute respiratory distress syndrome. NPPV reduces the intubation rate and is eff ective in the treatment of SARS-related acute respiratory failure without posing infection risks to health-care workers. [69] [70] [71] [72] In this setting, intubation has been avoided in up to two-thirds of cases in some studies. 73, 74 If aerosolisation and airborne transmission are discounted, respiratory secretions or fl uids, or both, are the main route of SARS transmission, ultimately leading to an overwhelming risk of infection via intubation. However, this mode of transmission does not occur with other airborne pathogens, most notably avian infl uenza, which can aerosolise or disperse up to 0·5 m around a patient undergoing NPPV. 75 The main problem with dispersion through NPPV is not the exhalation portion of the respiratory circuit but the inevitable mask Review leaks. Consequently, intubation or mechanical ventilation should be preferred to NPPV if safely achievable to control droplet aerosolisation and dispersion. Consequently, endotracheal intubation should be done by the most skilled person available. In addition, that person should wear personal protective equipment, including eye protection, and use rapid sequence induction. 43 ICU rooms should be subjected to negative pressure and a minimum of 15 air changes per hour. 76, 77 Some researchers advocate the use of a powered air-purifying respirator, particularly during high-risk manipulations such as endotracheal intubation, because it off ers supplementary protection with a better fi t to the healthcare worker's face. 78, 79 However, its use is a matter of debate because the risk of dysfunction seems important, which may increase the risk of exposure. Problems in cleaning, disinfection, and storage of the respirator may also increase the risk of exposure.

Endotracheal intubation is not the only high-risk procedure in ICU patients; cardiopulmonary resuscitation was also reported to be a very high-risk procedure during the SARS epidemic. 80 Patients with other HIDs, such as viral haemorrhagic fever, should also be treated in an ICU. Although the main route of transmission of viral haemorrhagic fever is contact with body fl uids, airborne transmission has been suggested with the Reston and Zaire strains of Ebola virus in monkeys. 81 In an outbreak of Ebola in a Johannesburg hospital, no subsequent cases of the disease occurred despite the staff being involved in numerous hazardous procedures. 82 Despite the fact that this hospital opted for high-level barrier nursing, which entailed the isolation of the patient in a cubicle and the use of protective clothing plus high-effi ciency particulate air-fi ltered respirators to minimise exposure to aerosols, universal blood and body-fl uid precautions may have been suffi cient for protection. 82

The management of invasive diagnostic or therapeutic procedures in a patient with HID is a challenge. However, there were few reports on hospital-acquired HIDs during invasive procedures before the SARS era. In fact, until now, HID outbreaks had only occurred in countries or at times when such techniques were not available, and the recent SARS epidemic has revealed the risks in such situations. 83 Of note, available evidence on risk factors is weak and somewhat indirect, according to the commonly accepted hierarchy of evidence. Much work needs to be done to separate the essential risk factors from the superfl uous ones. High-risk aerosolgenerating procedures were well summarised in the last WHO interim guidelines. 43 Panel 2 shows the EUNID recommendations for infection control in the management of patients with HID during these special procedures.

Although diagnostic bronchoscopy or fl exible lung endoscopy is not necessary in some scenarios (ie, an ongoing outbreak of a known disease), some situations need such invasive procedures to rule out diff erential diagnoses or to collect samples for laboratory • The examination should be kept as short as possible to answer the clinical questions • For HLIU-admitted patients, bedside radiography should be provided to avoid transport of patients; radiographic equipment should then be kept in the HLIU • Radiographs should be interpreted only by a designated radiologist who is aware of infection-control procedures, and by use of a picture-archiving and communication system, if available 84 • For ultrasound scanning, a sonographic scanner should be designated as a portable radiograph to be used only for HID patients • For CT or MRI, we strongly recommend that the department appoints a staff member to monitor and ensure that all department staff fully comply with the infection-control measures according to the guidelines • Designated sessions or hours, either out of offi ce hours or at the end of a session, should be assigned for such patients

• Treat HID patients who require dialysis at their bedside with either peritoneal dialysis or haemodialysis • Designate dedicated haemodialysis machines • Decontaminate dialysate as infectious waste Post-mortem examination • Risks and benefi ts must be carefully considered • Limited autopsy or post-mortem collection of blood and percutaneous biopsy are preferred • The biosafety precautions recommended for clinicians and laboratory staff working with infected patients and specimens must also be followed during post-mortem examination • Perform the autopsy only if necessary and in a biosafety level 3/4 isolation room, which can serve as the HLIU HLIU=High-level isolation unit.

investigation. SARS transmission has been reported or suggested after the intubation of patients, 84 and use of a nebuliser by health-care workers in patients with SARS resulted in a major outbreak of the disease. 85 In a retrospective study among critical-care nurses in Toronto, 86 the probability of a SARS infection was 6% in nurses who assisted during intubation, suctioning, and manipulating the oxygen mask. In the same study, wearing a mask, especially an N95 mask, was deemed protective. 86 A high-fl ow-rate oxygen mask may also result in health-care worker infection, 83 and we have thus suggested that NPPV is preferred to facial mask aerosol therapy if available. Bronchoscopy has also been suggested to increase SARS-associated coronavirus transmission in health-care workers. 71 The aerosolisation of lung pathogens during fl exible endoscopy and hospital-acquired infections during these procedures are both well documented and have led to standard guidelines for fl exible endoscopy. 87 Similar transmissions would probably occur with other respiratory agents, such as avian infl uenza and hantavirus pulmonary syndrome. Bronchoscopy, airway suctioning, and other procedure that may induce coughing and expose health-care workers to potentially infected aerosolised respiratory droplets pose an increased risk of transmission of those agents. In most hospitals, rooms dedicated for bronchoscopy are under negative pressure, but these rooms are not necessarily air fi ltered. 88 

In addition to airborne transmission, SARS-associated coronavirus may also be transmitted by direct contact with infected respiratory secretions and other body fl uids, similar to viral haemorrhagic fever viruses. 49, [89] [90] [91] Contact with contaminated environmental surfaces and inanimate objects is suspected to have resulted in the transmission of SARS, as suggested by reports that some health-care workers became infected even though they had no direct contact with SARS-infected patients. 47, 89, [92] [93] [94] Data also suggest that SARS-associated coronavirus and orthopoxvirus can survive on hard surfaces such as plastic and stainless steel for several hours, if not days. 40, 47, 94 Moreover, many group 3 and 4 viruses, including SARS-associated coronavirus, have been identifi ed in human faeces. 47, 90, 93, 94 Although there is no published report of transmission of SARS to health-care workers and other patients during gastrointestinal endoscopy, the potential for such transmission exists. 95

Most of our knowledge on managing infection control in radiology departments comes from experience of tuberculosis and SARS. Radiology technicians have a relative risk for a positive tuberculin skin test of 1·7 compared with other health-care workers, and those working for less than 1 year have a lower risk of infection, indicating that radiology technicians are more exposed to tuberculosis during their practice. 96 At the Prince of Wales Hospital in Hong Kong in March, 2003, at least 50 health-care workers, including radiographers, were infected by SARS. 97 Because imaging plays an important part in the diagnosis and management of HID, the role of the radiology department is to provide an immediate and effi cient radiological service for patients with suspected or confi rmed HIDs. Chest radiography is mandatory in such a situation.

To minimise the risk of cross-infection, transportation of patients with HID should be as limited as possible. For ambulatory patients with suspected HIDs, a satellite radiography centre should be set up with portable radiography equipment in the vicinity of the EDIR dedicated to HID patients in order to confi rm or to reject the diagnosis. 98 For patients in the HLIU, bedside radiography should be provided to avoid transportation of patients. 97, 99 To avoid transmission of fomites, the radiograph fi lm should be handled with care and should only be interpreted by a designated radiologist who is aware of infection control. The fi lm-processing area, where cassettes are brought back to the department after bedside radiography in the HLIU, should be considered as high risk, unless the cassettes were processed in the HLIU by following a protocol of double bag sealing, and should be disinfected. 98, 99 Images should be interpreted through a picture-archiving and communication system, if available. 84, 97, 100 For ultrasound scanning, a sonographic scanner should be designated as a portable radiograph that is only used for HID patients. One machine should be dedicated for a specifi c area such as the HLIU. The examination should be kept as short as possible to answer the clinical questions. The transducer should be covered with disposable covers that are discarded between patients.

The value of a CT scan in assessing the diagnosis of HIDs, such as SARS, has been established, and a CT scan is sometimes mandatory for a patient's assessment. 101 Because this examination can be done only in the radiology department, stringent infection-control measures need to be followed, and the examination should be done only if absolutely necessary for the patient's recovery.

We strongly recommend that the radiology department appoint a staff member to monitor and ensure that all staff fully comply with the infection-control measures according to the guidelines. Designated sessions or hours, either outside offi ce hours or at the end of a session, should be assigned for such patients. Patients should be transported in a special isolation carrier or in a defi ned way to avoid any contact with other patients or unprotected personnel. 58, 100 The department should be divided into low-risk and high-risk areas. After a CT scan, the gantry table and fl oor should be cleaned, and any bed linen should be changed. Film cassettes should be decontaminated properly before fi lm processing. In all Review cases, radiology technicians, radiologists, and other radiology personnel should comply with universal precautions, including the wearing of a mask, cap, gown, and gloves during direct contact with patients.

The main reported dialysis-associated infection is viral hepatitis. As a consequence, guidelines have been edited to prevent nosocomial transmission of this agent to personnel and patients. 102 When the guidelines were followed in a European study, 103 no reported hospitalacquired cases of hantavirus haemorrhagic fever with renal syndrome were reported, despite 30-50% of patients being in need of haemodialysis.

Most of our knowledge in the management of HID with renal failure has been acquired from SARS outbreaks. By comparison with other patients, the care of patients undergoing renal dialysis poses several additional infection-control issues in the disposal of spent dialysate (both haemodialysis and peritoneal dialysis) and in the prevention of cross-contamination within the dialysis unit. 104 During the SARS episode, patients receiving dialysis were kept in the SARS isolation ward with the other patients with SARS. All patients with peritoneal dialysis were treated with intermittent peritoneal dialysis during their hospital stay. The dialysis exchange was done by the ward staff , who wore full protective gear, as recommended by WHO, including waterproof disposable gowns, caps, gloves, face shields, and N95 face masks. 104 The spent peritoneal dialysis effl uent was decontaminated with a 2% sodium hypochlorite solution.

Haemodialysis was also done by the ward staff , who wore full protective gear, in a room especially equipped for SARS patients in the isolation ward. 104 Dedicated haemodialysis machines were used with an ordinary tap water supply that passed through a fi lter without reverse osmosis or any other water treatment. Spent dialysate was decontaminated as described above, and all of the blood tubing was discarded as infectious waste. Because they were potentially contaminated, unspent dialysate concentrate and the sodium bicarbonate cartridge were also discarded as infectious waste. The dialysis machine was disinfected after each haemodialysis session with a sodium hypochlorite solution according to the manufacturer's instructions. Particular attention should be paid to the infection control of dialysate effl uents and decontamination of the machines.

Although autopsies have been done safely on patients with HID in some circumstances, without prior knowledge of diagnoses such as Ebola haemorrhagic fever, HID agents are transmissible at autopsy, which raises concerns about the protection of pathologists and autopsy personnel. 105 Tuberculosis was the fi rst autopsytransmitted disease to be reported in the literature, 106 and transmission is also thought possible with multidrugresistant or extremely drug-resistant tuberculosis. Aerosol production during autopsy, particularly from ruptured organs, had been recognised early in this situation and has led to some precautions. 107 During the fi rst reported episode of hantavirus pulmonary syndrome, the fi rst fi ve suspected patients were autopsied with standard precautions only, even though the agent was isolated and classifi ed as a group 3 agent. 108 Fortunately, no transmission to autopsy personnel occurred. During the SARS episode, many autopsies were done, and although there was no case of transmission, several investigators have raised concerns over biosafety in autopsy rooms. 105, 108, 109 Recent guidelines have been published to prevent infection during autopsy. 110 Before an autopsy is done on a patient suspected to have died from an HID, the possible risks and benefi ts must be carefully considered. 58 Limited autopsy or post-mortem collection of blood and percutaneous liver biopsy material may be appropriate. Several pathologists suggest that safety measures applied to laboratory workers should also be applied during and after autopsy. Furthermore, patients who have died from an unknown HID or from a known group 3 or 4 agent should be autopsied only if necessary and in a biosafety level 3 or 4 isolation room. 58, 108, 109 

The literature on HIDs, particularly SARS, indicates that there is a need for hospitals to be prepared for these events and that HLIUs urgently need to be built in European member state hospitals. Research and development of universal, bedside, reproducible, and transferable diagnostic tools are mandatory. Prompt reporting to the authorities is needed so that a rapid response can be organised. These measures should be accompanied by harmonised recommendations for the

The literature review was done by use of Medline and the Web of Science with the following key words: "SARS [all]", "laboratory-acquired infection", "laboratory-associated infection", "imported [each agent's name]", and "[each group 3 and 4 agent's name]". More than 1400 references were obtained. Selection criteria within the topics were the impact factor and half-life of the journal, followed by journal availability. The topics selected were as follows: patient's criteria for admission to the HLIU, admission to the emergency department, safe sampling for laboratory investigation, isolation of suspected HID patient in HLIU, intensive-care management of HID patients, HID in paediatric patients, practice of invasive procedures such as bronchoscopy and gastroscopy, radio imaging, renal dialysis, and post-mortem examination of HID patients. Only English-language articles were reviewed. Websites such as WHO, CDC, and other scientifi c and international research society's guidelines were also used.

Review safe care of these unusual patients. The recommendations reported here by our group will hopefully help establish consensual protocols. Networking for the standardisation of procedures and the management of these patients is mandatory.

We declare that we have no confl icts of interest.

Laboratory-acquired infections with Hantaan virus, the etiologic agent of Korean hemorrhagic fever

Laboratory acquired tick-borne meningoencephalitis: characterisation of virus strains

Brief report: treatment of a laboratory-acquired Sabia virus infection

Laboratory-acquired Sabia virus infection

Laboratory-acquired vaccinia infection

Laboratory infection of a technician by mimivirus

Biological safety cabinetry

Laboratory-acquired severe acute respiratory syndrome

European Network of Infectious Diseases

Framework for the design and operation of high-level isolation units: consensus of the European Network of Infectious Diseases

Designing a biocontainment unit to care for patients with serious communicable diseases: a consensus statement

Recent Singapore SARS case a laboratory accident

Severe acute respiratory syndrome: preparedness, management, and impact

Using models to identify routes of nosocomial infection: a large hospital outbreak of SARS in Hong Kong

Risk group classifi cation for infectious agents

Nosocomial transmission of Q fever

Q fever in pregnancy

Q fever during pregnancy-a risk for women, fetuses, and obstetricians

Risks and prevention of nosocomial transmission of rare zoonotic diseases

Nosocomial infection due to Pseudomonas pseudomallei: two cases and an epidemiologic study

Guidelines for preventing the transmission of Mycobacterium tuberculosis in health-care settings

Extensively drug-resistant tuberculosis: current challenges and threats

on the protection of workers from risk related to exposure to biological agents at work. Offi cial Journal of the European Communities

Smallpox as a biological weapon: medical and public health management. Working Group on Civilian Biodefense

Bichat guidelines for the clinical management of haemorrhagic fever viruses and bioterrorismrelated haemorrhagic fever viruses

Novel chikungunya virus variant in travelers returning from Indian Ocean islands

Human-to-human transmission of hantaviruses

Infection prevention and control of epidemic-and pandemic-prone acute respiratory diseases in health care: WHO interim guidelines

WHO issues guidelines to manage any future SARS outbreak

B virus (Herpesvirus simiae) infection in humans: epidemiologic investigation of a cluster

SARS exposure and emergency department workers

SARS outbreak in the Greater Toronto Area: the emergency department experience

Responding to the severe acute respiratory syndrome (SARS) outbreak: lessons learned in a Toronto emergency department

Emergency departments (EDs) in the United Kingdom (UK) are not prepared for emerging biological threats and bioterrorism

Eff ectiveness of precautions against droplets and contact in prevention of nosocomial transmission of severe acute respiratory syndrome (SARS)

Guideline for isolation precaution: preventing transmission of infectious agents in healthcare settings

Respiratory hygiene/ cough etiquette in healthcare settings

Transmission of infl uenza: implications for control in health care settings

Foundations of the severe acute respiratory syndrome preparedness and response plan for healthcare facilities

Hospital preparedness for severe acute respiratory syndrome in the United States: views from a national survey of infectious diseases consultants

Hospital preparedness to bioterrorism and other infectious disease emergencies

Early diagnosis of lower respiratory tract infections (point-of-care tests)

Preventing nosocomial infection in paediatric wards

Pediatric nosocomial infections: children are not little adults

Balancing infection control practices and family-centred care in a cohort of paediatric suspected severe acute respiratory syndrome patients in Hong Kong

Bowlby and Robertson revisited: the impact of isolation on hospitalized children during SARS

Transmission of severe acute respiratory syndrome in critical care: do we need a change?

Transmission of severe acute respiratory syndrome during intubation and mechanical ventilation

Illness in intensive care staff after brief exposure to severe acute respiratory syndrome

Clinical progression and viral load in a community outbreak of coronavirus-associated SARS pneumonia: a prospective study

Development of a standard treatment protocol for severe acute respiratory syndrome

Clinical analysis of 45 patients with severe acute respiratory syndrome

Description and clinical treatment of an early outbreak of severe acute respiratory syndrome (SARS) in Guangzhou, PR China

A hospital outbreak of severe acute respiratory syndrome in Guangzhou, China

A retrospective study of 78 patients with severe acute respiratory syndrome

Non-invasive versus invasive mechanical ventilation for respiratory failure in severe acute respiratory syndrome

Noninvasive positive-pressure ventilation: an experimental model to assess air and particle dispersion

ICU management of severe acute respiratory syndrome

Practical guidelines for infection control in health care facilities

SARS: ventilatory and intensive care

Severe acute respiratory syndrome-Singapore

Possible SARS coronavirus transmission during cardiopulmonary resuscitation

Kass handbook of infectious diseases: exotic viral infections

Unexpected Ebola virus in a tertiary setting: clinical and epidemiologic aspects

The epidemiology of the outbreak of severe acute respiratory syndrome (SARS) in Hong Kong-what we do know and what we don't

Application of Ris/ PACS in the prevention and treatment of SARS at the hospital radiology

SARS: experience at Prince of Wales Hospital, Hong Kong

SARS among critical care nurses

American College of Chest Physicians and American Association for Bronchology [corrected] consensus statement: prevention of fl exible bronchoscopy-associated infection

The American Institute of Architects. Guidelines for design and construction of hospital and health care facilities

SARS transmission, risk factors, and prevention in Hong Kong

SARS: epidemiology, clinical presentation, management, and infection control measures

Listening to SARS: lessons for infection control

An outbreak of severe acute respiratory syndrome among hospital workers in a community hospital in Hong Kong

Managing SARS amidst uncertainty

Management of hospital-acquired severe acute respiratory syndrome with diff erent disease spectrum

Recommendations for the prevention of transmission of SARS during GI endoscopy

Risk of tuberculous infection among healthcare workers in a tertiary-care hospital in Ankara

Eye of the storm: the roles of a radiology department in the outbreak of severe acute respiratory syndrome

Severe acute respiratory syndrome: avoiding the spread of infection in a radiology department

Rapid creation of a temporary isolation ward for patients with severe acute respiratory syndrome in Taiwan

Severe acute respiratory syndrome: management and reconfi guration of a radiology department in an infectious disease situation

Value of CT in assessing probable severe acute respiratory syndrome

European Best Practice Guidelines Expert Group on Hemodialysis (ERA)

Hantavirus infections in Europe

Severe acute respiratory syndrome in dialysis patients

The modern autopsy: what to do if infection is suspected

Tuberculosis acquired in laboratories and necropsy rooms

Aerosols in the mortuary

Hantaviral biosafety issues in the autopsy room and laboratory: concerns and recommendations

Biosafety level 3 laboratory for autopsies of patients with severe acute respiratory syndrome: principles, practices, and prospects

Health and safety at necropsy