Journal of the New Zealand Medical Association, 20-February-2004, Vol 117 No 1189
Norovirus outbreaks in a hospital setting: the role of infection control
Sarah Lynn, Julianne Toop, Carl Hanger and Nigel Millar
Noroviruses (Norwalk-like viruses), previously known as small round structured viruses, are a common cause of outbreaks of viral gastroenteritis. The illness is characterised by rapid onset of nausea and vomiting, often projectile, along with diarrhoea. The symptoms are usually mild and self-limiting, with a mean duration of illness of 12–60 hours. Relapse is uncommon but recognised and mortality rates are low, even in hospital outbreaks. The attack rate is around 50% and the infective dose is as small as 1–10 virus particles. The incubation period is 24–48 hours and excretion of the virus in faeces can continue for up to 7–10 days. Transmission is by person-to-person contact through aerosol or direct contact, or through consumption of food handled by an infected food worker.
Outbreaks have been described in a number of settings with most exposures in New Zealand occurring in restaurants through contaminated food.1 Hospitals and rest homes are also affected, resulting in morbidity for patients and staff, staffing shortages, hospital disruption and ward closure. Prompt recognition and infection control measures can minimise the consequences. We report our experience of two NV gastroenteritis outbreaks that occurred in separate geriatric rehabilitation wards within an 18-month period, and the role of infection control in limiting their spread.
Setting The two outbreaks occurred in separate rehabilitation wards for older people. The hospital described has five wards (131 beds) for geriatric rehabilitation and two wards (44 beds) for psychogeriatric patients. The first outbreak occurred during winter in a 25-bed ward occupied by 28 patients at the time of the outbreak. This ward provides rehabilitation for mainly post-operative patients. The second outbreak occurred during autumn in a 23-bed stroke rehabilitation ward. Both wards have a mixture of rooms containing one to four beds and both have shared bathroom facilities.
Case definition A case was defined by the sudden onset of vomiting, with or without diarrhoea. Other symptoms could include nausea, abdominal cramps, myalgia, headache, chills and fever. The person presenting with signs and symptoms of illness had to have been in contact with cases or in the environment/geographic area in which the outbreak was occurring.
Laboratory methods Faecal specimens were collected and sent to the Institute of Environmental Science and Research Limited for identification of NV by reverse transcriptase polymerase chain reaction. This technology provides more rapid identification of a virus in faecal specimens than previous methods. Specimens were also examined for Clostridium difficile enterotoxin, Rotavirus, Shigella, Salmonella and Campylobacter species.
Faecal specimens were collected if a patient had symptoms that met the case definition. Once NV had been identified in an outbreak, further faecal specimens were not required and patients were diagnosed on clinical grounds, providing they met the case definition.
A register was kept in each ward and filled in by staff or the infection control nurse.
Outbreak 1 Outbreak 1 lasted 14 days with the first case presenting on 25 July 2000 and the last case developing symptoms on 7 August 2000. The infection control nurse and the public health service were notified on 27 July after five cases had developed in the preceding 12 hours during the late afternoon and night. On investigation, one patient and two staff members had also succumbed over the previous 48 hours. No precautions had been put in place and by the end of 27 July a total of 14 cases had presented. Contact precautions were commenced, staffing guidelines (Table 1) instituted and the ward was closed for 11 days.
Table 1. Staffing guidelines for use during outbreak of acute gastroenteritis
Forty one cases met the case definition, with 16 patients and 25 staff affected (Figure 1).
Figure 1. Timeline of symptom duration in Norovirus outbreak 1
The attack rate was 57.1% for patients and 41% for staff. The average age of the patients was 79.5 years (range 65–93). The mean duration of sickness was 2.2 days for patients and 1.2 days for staff. Three cases relapsed. One patient died, with gastroenteritis the precipitating event of his final illness. Seven of nine faeces samples collected from patients were positive for NV and the other two were negative. Three of the positive NV cases were also positive for Clostridium difficile enterotoxin. Once the virus was isolated, no further samples were sent for virus identification and all following cases were diagnosed on the case definition and clinical history. As there were no cases identified outside of the affected ward, a food-borne source from the hospital kitchen was considered unlikely.
Modified staffing restrictions were commenced immediately for staff meeting the case definition. A strict staffing regime as suggested by Chadwick,2 as per Table 1, was developed and enforced from Day 5 of the outbreak. During the early days of the outbreak there was not full awareness of both the highly infectious nature of the virus and the importance of rigorous cleaning followed by disinfection. For example, a commode containing diarrhoea from a patient who met the case definition was knocked over in a four-bed room and the area was not disinfected nor the carpet steam cleaned. It was 72 hours before definitive cleaning action was taken.
After the outbreak ended, an NV education series was conducted throughout the hospital discussing the case definition, control of spread, staffing issues, patient and staff outcomes, and the lessons learned.
Outbreak 2 Outbreak 2 lasted for 16 days from 6 March 2002 to 20 March 2002. Ward staff instituted infection control precautions immediately on suspicion of gastroenteritis as soon as a third case was noted early on 8 March. The infection control nurse was contacted immediately. The ward was closed on 8 March as more cases developed over the day and the public health service was notified. The ward was closed from that date for a total of six days. On the date of reopening there had been no new cases for 24 hours. Three further cases developed after reopening but cross infection could be contained by contact precautions. Twenty four cases met the case definition, with 13 patients and 11 staff affected (Figure 2). The attack rate was 56.5% for patients and 18% for staff. The average age of the patients was 72.3 years (range 65–86). The mean duration of sickness was 2.5 days for patients and 3.5 days for staff. There were no deaths on the ward during this time period. Four of five faecal specimens from symptomatic patients were positive for NV. No other causative organisms were identified.
Several amended measures were used in this outbreak to contain the infection (Table 2). Staffing restrictions as per Table 1 were put in place immediately with strong support from management. Paid sick leave over and above normal sick leave was made available for permanent staff to encourage compliance with the 48-hour recommendation. The cleaning service was involved immediately and a rigorous cleaning regime put in place. Strict cleaning and disinfection were undertaken for any room contaminated by patient vomit or diarrhoea. Patients remained on the ward for their rehabilitation therapy rather than attending other therapy areas. Intensive promotion of the importance of hand hygiene was performed on the ward to educate staff, patients and visitors. This was backed up by the ready availability of alcohol-based hand gels for all to use.
Figure 2. Timeline of symptom duration in Norovirus outbreak 2
Table 2. Precautions for any outbreak of vomiting and diarrhoea
NV outbreaks are characterised by a high attack rate of the virus and rapid spread to both patients and staff. We have shown in this report that the impact of hospital-acquired NV gastroenteritis can be limited through a strict programme of infection control measures and ward closure (Table 2). By using the case definition, staffing guidelines and infection control measures that were developed during the first outbreak, we were able to implement this programme sooner and more effectively in the second outbreak. This resulted in a shorter duration of ward closure and lower staff attack rate. These differences were seen despite the outbreak duration and patient attack rate being similar. The fact that the staff reported a longer duration of illness suggests that the staffing restrictions on returning to work were more strictly adhered to during this episode as opposed to the first outbreak. It is difficult to enforce sick leave for nursing staff, particularly casual or bureau staff who do not receive sick pay.3 This was acknowledged in the second outbreak with the introduction of special paid sick leave for permanent staff, which encouraged staff to comply with the guidelines.
Rapid recognition and notification of a gastroenteritis outbreak is important. While it can be difficult on a geriatric ward to distinguish an outbreak of gastroenteritis from other causes of diarrhoea and vomiting it is important to maintain vigilance when several patients present at once. Delay in instituting control measures can significantly extend the duration of an outbreak.4 In both outbreaks there was a delay of 48 hours before the infection control nurse was notified; however, the index of suspicion was much higher in the second outbreak and infection control measures were put in place by the ward staff immediately.
Close liaison between management and staff allowed us to close the affected wards to new admissions and restrict staff movements. Requiring staff to stay away until 48 hours after their last symptom was a contentious issue. The infection control nurse worked closely between nursing coordinators, unit managers and individual staff members to ensure adherence to required staff restrictions as per Table 1. Ward closure can be difficult to achieve if the need for beds is high, such as in acute hospitals or in winter when beds are in short supply. We believe both ward closure and staffing restrictions were essential aspects of managing the outbreaks and allowed us to prevent the spread of infection to other areas, thereby avoiding hospital-wide disruption.
Recent reports show just how severely hospitals can be affected if an outbreak is not contained. In 2002 the Victoria Infirmary in Glasgow was forced to close to admissions for a week after almost 300 patients and staff were affected.5 Nine other Scottish hospitals had temporary ward closures. An outbreak in England in 1994 closed a 220-bed hospital due to staff shortages.6 The costs of staff sickness, bed closures and control measures can be significant. An outbreak in 1995 in Australia caused closure of several wards for up to 22 days and resulted in estimated costs of AU$7600 for sick leave and AU$10 600 for bed closures.7 Early closure is likely to result in an overall net gain, with the initial bed shortages being compensated for by lower overall disruption.
Evidence-based, clear and workable clinical infection control systems can prevent the spread of an outbreak. Prompt identification of a cluster of patients with unexplained vomiting and diarrhoea and immediate implementation of contact precautions are essential. Involvement of the cleaning service from the outset is also important, as is educating nursing staff to clean and disinfect after any environmental contamination. Limiting the movement of patients from the outbreak ward and tight controls on where the symptomatic patients are placed, or moved to, within the ward are also vitally important. In both outbreaks, if a patient met the case definition and was in a four-bed room, the entire room was placed under contact precautions. Only in the later stages of the outbreak were infected patients moved to either single rooms or grouped together. This was to facilitate the re-opening of wards and was decided in close liaison with the infection control nurse and the clinical staff. Our team also found that constant feedback, up-to-date information and sharing of the attached timelines (Figures 1 and 2) assisted staff to comply with precautions put in place.
A significant proportion of reported outbreaks of NV in New Zealand occur in healthcare settings,1 and the greatest burden of illness is seen in the institutionalised elderly.8 The key points to managing and limiting an outbreak of this kind include early recognition and notification, prompt implementation of infection control measures, education of staff, patients and visitors, staffing restrictions and ward closures.
Author information: Sarah J Lynn, Senior Registrar, Older Persons’ Health, The Princess Margaret Hospital, Christchurch; Julianne M Toop, Infection Control Practitioner, Southern Community Laboratories, Christchurch; H Carl Hanger, Physician/Geriatrician; Nigel D Millar, Clinical Director, Older Persons’ Health, The Princess Margaret Hospital, Christchurch
Correspondence: Dr Carl Hanger, Older Persons’ Health, The Princess Margaret Hospital, PO Box 800, Christchurch. Fax: (03) 337 7823; email: firstname.lastname@example.org
issue | Search journal |
Archived issues | Classifieds
| Hotline (free ads)
Subscribe | Contribute | Advertise | Contact Us | Copyright | Other Journals