Outbreaks in healthcare settings affect vulnerable populations, disrupt normal routines and may spread to other healthcare institutions (HCIs). Outbreaks can be limited in extent by good routines for detection, management of cases and other infection-control measures [1]. Norovirus infection is most often seen in the winter months and is a common cause of outbreaks in HCIs [2] as it has a low infectious dose, short incubation period, and symptoms such as diarrhoea and vomiting which facilitate spread. Symptoms normally lasts around one to three days, but can be longer in hospital patients [3]; and in this type of setting, infection can lead to slower recovery from other illness and even death [4]. Norovirus can be divided into several genogroups and genotypes [5]. Genogroup II genotype 4 is the most prevalent genotype globally [6] as well as in the Nordic countries [6]. There is no vaccine and immunity is not well understood; at best it is strain-specific but probably only partial and shortlived as the virus readily undergoes mutation [7, 8]. Humans are the only reservoir of the virus and spread of the infection in outbreaks is particularly difficult to control because of the low infectious dose, its stability in the environment and efficient transmission by person-to-person contact and exposure through contaminated surfaces [9]. Norway has national recommendations on norovirus infection in long-term-care facilities (LTCFs) in which the most important measure is isolation or cohort nursing of sick residents. Exclusion of sick staff until 48 h after they are symptom free is also recommended [10]. In a hospital setting, the infection-prevention-control unit will have local procedures. There are around 60 hospitals and 950 LTCFs in Norway [11]. The responsibility for management of local outbreaks lies within the hospital or with the community medical officer (one in each of the 422 municipalities) for outbreaks in LTCFs. All suspected outbreaks in Norwegian HCIs, regardless of the causative pathogen, should be alerted by law to relevant actors, including the Norwegian Insititute of Public Health (NIPH), to facilitate communication and response [12, 13]. The aim of this study was to describe, for the first time, the epidemiology and impact of these outbreaks in order to identify areas which may improve outbreak response.

This study shows that norovirus outbreaks pose an important burden for HCIs all over Norway, especially in the winter months. In addition to affecting an already vulnerable population, this study shows that these outbreaks indeed also impact on the internal workflow and resources, with a conservative estimate of around 1800 days of absenteeism per year due to these outbreaks, during which staff would have to be covered for by other internal or external healthcare staff.

Surveillance of norovirus outbreaks exists in Germany and Scotland. In Germany, reporting of norovirus outbreaks in HCIs has been mandatory since 2001. In contrast to what is seen in Norway, outbreaks were smaller (median nine cases vs 15 in this study) and around 80% of norovirus outbreaks were reported from hospitals (vs 23% in this study) during the first 12 months after introduction of the system [15]. Varying ways of counting interdepartmental outbreaks, better collaboration with the local level or under-reporting from hospitals may explain this. In Scotland, surveillance of ward closures due to norovirus infection has been in place since October 2017. From then until week 26, 2018, 219 wards or bays have been closed due to confirmed or suspected norovirus [16]. This is markedly more than the 16 reported outbreaks in hospitals in Norway 2017/18, in a population of similar size. The occurrence of norovirus outbreaks has also been studied prospectively; Curran et al. [17] aimed to identify the index cases of norovirus outbreaks in the UK and Ireland in 54 acute and non-acute healthcare centres; only five out of the 54 included centres did not experience any outbreak during one winter. Also, Lopman et al. found that 171 inpatients units, had on average 1.3 gastroenteritis outbreaks in the 1-year follow-up period. Of these, 63%were caused by norovirus [2].

It was seen that a small proportion of residents at LTCFs were admitted to hospital during norovirus outbreaks. This may be necessary in severe cases despite the risk of spread from one institution to the next. Our results suggest that hospitals are affected by norovirus outbreaks earlier in the epidemiological year than LTCFs. Potentially because there is a greater influx of patients from the community, where norovirus circulates, to and from hospital than between the community and LTCFs as also suggested by Sadique et al. [18]. This finding, however, could only be evaluated on the national level, as the number of reported outbreaks is low. That the start of the outbreak season seemed to start earlier in hospitals than in LTCFs, at least at the national level, suggests an opportunity that with improved communication, hospitals could alert LTCFs within the same area in order to prepare for the outbreak season and limit the extent of further outbreaks.

Slightly more cases were seen amongst healthcare staff in hospitals compared to LTCFs, though no information about the number of healthcare staff at risk during the outbreaks is available. The patient or resident:healthcare-staff ratio varies with the level of care needed and type of department and will most often be higher in hospitals. Whether this explains the slightly higher proportion of staff affected in hospital outbreaks is unknown. Nevertheless, healthcare staff do represent a big proportion of cases in the reported outbreaks, indicating a need for improved compliance with infection prevention and control measures. Outbreaks are an economic burden for HCIs, both as infected staff need to be covered for during illness and ‘quarantine’ and cohort nursing may require extra staff.

The relatively high number of people infected during an outbreak underscores the infectiousness of norovirus and norovirus can serve as a worst-case scenario for introduction of other, more virulent, person-to-person transmitted pathogens into HCIs. With the current information captured in the alert system, it was not possible to assess the extent to which national recommendations were followed and/or which infection prevention control procedures are in place locally. But the high number of people infected do suggest a potential for limiting spread, for example by having systems and routines in place before outbreaks happen, as advised in the national recommendations.

Even though NIPH routinely promotes the web-based outbreak alert system and teaches outbreak management, both at the regional and national level, in order to strengthen local capacity and encourage the use of the alert system, under-reporting is still apparent. If the under-reporting of outbreaks reflects a lack of awareness concerning outbreak management, or a lack of communication between the LTCF and the municipal doctors about ongoing outbreaks, it is worrying. The alert system serves to alert relevant stakeholders so that outbreak support and advice can be given in an early phase. The alert system can also be used for statistical purposes to get a national overview of outbreaks which will facilitate targeted capacity building, guideline development and communication messages in order to increase awareness and investigate whether there are any changes in trends.

Limitations
This study has three main limitations: the sensitivity of the norovirus outbreak definition and under-reporting of number of outbreaks and number of cases in each outbreak. Classification as a norovirus outbreak is dependent on local definitions. The infection prevention measures for diarrhoea and vomiting are the same for all the common pathogens in this setting. Samples were submitted for testing in two thirds of the outbreaks and most were confirmed as norovirus at the time of reporting or updating. Information about the genotypes of the isolated strains from each outbreak or of dominant strain of the season was not available. For this reason, it was not possible to evaluate the effect of the genotype.

Concerning under-reporting, the number of outbreaks notified through the outbreak alert system and reported here, most likely represent only a proportion of all norovirus outbreaks occurring in Norwegian HCIs. Although outbreaks were reported from all parts of Norway, some areas had not reported any outbreaks of any kind during the 13-year study-period.

The alert system is used for the mandatory alerting of suspected outbreaks. Reporting should happen as soon as the outbreak is suspected and before the full extent of the outbreak is known. Even though the system sends a reminder to update the details about the outbreak, including the case numbers, three weeks after the initial alert, some under-reporting of the extent of each outbreak is expected.

This is the first comprehensive description of norovirus outbreaks in HCIs in Norway. Even though the analyses revealed under-reporting that is unlikely to reflect the real epidemiology, this study clearly shows that these outbreaks affect both hospital and LTCFs all over Norway. Norovirus infection may delay medically important procedures and recovery, but also presents a major challenge to the functional ability of an HCI and its resources as up to one-half of cases were healthcare personnel.

It is recommended that NIPH promotes the outbreak alert system to increase reporting and improve the quality of the data and strengthen local capacity for outbreak management and general infection control. It is also recommended to investigate possibilities for improving communication between hospitals and LTCFs regarding when the norovirus season starts and progresses, for hospitals and LTCFs to be prepared and to take early action to prevent and limit further spread.

Read full article: https://www.journalofhospitalinfection.com/article/S0195-6701(19)30268-3/fulltext?dgcid=raven_jbs_etoc_email

© 2019 The Authors. Published by Elsevier Ltd on behalf of The Healthcare Infection Society.