Introducing a novel “real-time” outbreak alert and notification system to monitor SARS-CoV-2 outbreaks and case fatality in elderly care facilities,the Netherlands,2020–2022

Abstract

Surveillance systems collating individual-level data may limit timely information sharing during rapidly evolving, infectious disease outbreaks. We present a digital outbreak alert and notification system (MUIZ) in which institutional-level data are reported, allowing real-time outbreak monitoring in elderly care facilities (ECF). We describe trends in the number of outbreaks, mean case number per outbreak, and case-fatality rate (deaths/recovered + deaths) of SARS-CoV-2 in ECF notified through MUIZ in the Rotterdam area (April 2020-March 2022). Overall, 369 outbreaks were reported from 128 ECF that registered with MUIZ (approximately 85% of all ECF), and 114 (89%) notified at least one SARS-CoV-2 outbreak. Trends were consistent with the concurrent national epidemiology and societal control measures in place. MUIZ is a simple outbreak surveillance tool that was highly adopted and acceptable to users. Dutch PHS regions are increasingly adopting the system and it has potential for adaptation and further development in similar institutional outbreak settings.

Introduction

In the Netherlands, outbreaks of infectious diseases are notifiable by law. ¹ Each regional public health service (PHS) is responsible for infectious disease control in its area. Traditionally, health care facilities reported outbreaks directly to the PHS, mainly by phone. There was no legal requirement or established process by which healthcare facilities could directly inform other institutions. Following a number of major, multi-institutional outbreaks in hospitals and elderly care facilities (ECF) in the Rotterdam-Rijnmond area^2,3 (total population 1.3 million), representatives from the PHS, nursing-home governance bodies, ECF, elderly care physicians, experts in infection prevention, microbiologists, physicians, and managers collaborated with a software developer to develop a rapid outbreak alert system known as MUIZ ⁴ (“Meldpunt voor Uitbraken van Infectieziekten en Bijzonder Resistente Micro-organismen,” translated as “Hotline for notification of infectious disease outbreaks and highly resistant microorganisms”). MUIZ was launched in February 2017.^5,6 The initial focus was on outbreaks of common organisms such as norovirus, influenza, and scabies, and also multi-drug resistant organisms (MDRO). The system was later extended to include a SARS-CoV-2 module, developed over a 3-week period in March 2020.

MUIZ is a secure web-based application to notify outbreaks and to exchange outbreak information between healthcare facilities and the PHS in real-time. It is an ecological system only and does not collect individual-level, personal data. To optimize expediency in notification, reporting through MUIZ is kept deliberately simple by limiting outbreak registration requirements to a minimum. In the SARS-CoV-2 module, notifiers are asked to provide aggregate numbers of cases reported, recoveries, deceased patients, and patient transfers in and out to account for cohort nursing. For more detail on the technicalities of MUIZ and the SARS-CoV-2 module, please refer to the Supplemental Material. The module went live on April 4, 2020. In this study, we report retrospective trends in the number, size, and case-fatality rate of SARS-CoV-2 outbreaks in ECF as notified through MUIZ, and consider the feasibility of using MUIZ to inform timely public health response and healthcare planning.

Design and methods

The MUIZ notification system

MUIZ is accessed via an online notification interface (Figure 1), to which users have differing levels of access (see Supplemental Material for technical detail). Some users can directly enter data (e.g. healthcare providers on site at the elderly care facility who suspect an outbreak) and others have reader-access only (e.g. authorized PHS staff). Once the criteria for outbreak notification are met by the notifier, an outbreak alert is opened within MUIZ and an email notification is issued to users that wish to receive these notifications. As the outbreak evolves, the notifier can update the outbreak status (e.g. number sick, number of fatalities, recoveries, etc.). This is automatically reported to users by email. Users can select a specific outbreak and retrospectively view all details of the outbreak, as entered by the notifier, including the full update history and the outbreak location.

Figure 1.

MUIZ outbreak notification tool, illustration of the information available to the user directly on login.

The SARS-CoV-2 module

Within MUIZ (Figure 1), a SARS-CoV-2 outbreak was defined as one or more suspected or confirmed SARS-CoV-2 cases in a resident or staff member of an elderly care facility. Once these criteria were met, notifiers were requested to provide: the date of onset of the outbreak (date first case identified at initial notification); the number of suspected and confirmed cases at the facility at the time of reporting (identified as residents and/or staff); number of transfers to and from other ECF and to hospitals, and related recoveries and fatalities among the elderly care facility residents. Notifiers received a reminder email every 3 days with the request to update the outbreak, which could be done at any time. In this way, the notifiers record the history and evolution of the outbreak. Once the outbreak is over, notifiers were required to specify the date of the “first day without disease” and the outbreak’s cumulative figures. Since new cases could still arise, notifiers were asked to close their notification after twice the maximum incubation period of SARS-CoV-2 (initially 28 days, reduced to 14 days after the study period). They also received reminder emails for this.

Data analysis

Outbreaks included in this study extended from April 2020 to March 2022, inclusive. The data analysis was confined to cases among residents only (cases among staff were notified less consistently and the primary concern was cases and outcomes among vulnerable elderly residents). It was descriptive and included estimates of: the total number of new outbreaks notified each calendar month (defined as outbreak alerts opened for the first time), total number of outbreaks ongoing on the first of this calendar month (already open, but ongoing); the absolute number of recoveries, deaths, and cases still sick from all monthly outbreaks by the end of the subsequent month (e.g. for new outbreaks notified in October 2020, the outcome reference date was December 1st, 2020). The mean size of outbreaks notified each calendar month was estimated using the sum of these recoveries, deaths and cases still sick. To reflect severity among detected cases as the pandemic evolved, a crude case-fatality rate among residents (CFR: no. of deaths/no. recovered + deaths ⁷ ) was also estimated at the outcome reference date. Total numbers (new outbreaks, ongoing outbreaks, size of outbreaks) and CFR were mapped over time, described at monthly intervals (Figures 2 and 3).

Figure 2.

Trend in number and size of outbreaks notified in the SARS-CoV-2 module of MUIZ, April 2020 to March 2022.

Figure 3.

Trend in case fatality rate after 1 month with 95% confidence interval for cases in outbreaks notified in the SARS-CoV-2 module of MUIZ, April 2020 to March 2022.

Additional information

To set the data in the concurrent national epidemiological context, information on the course of the SARS-CoV-2 epidemic nationally, the introduction of new COVID variants and the timing of imposed control measures was retrieved from three Dutch government sources.^8–10 These are noted in the figure along with the outbreak trends.

Results

In total, 128 ECF were registered with MUIZ, of which 114 (89%) notified at least one SARS-CoV-2 outbreak during the study period. In total, 369 outbreaks were notified.

Notifications of new and ongoing outbreaks, and size of outbreaks

The highest number of new outbreaks notified was in April 2020 and the mean size of outbreaks notified peaked in October of 2020 (Figure 2). The number and size of new outbreaks notified decreased from October 2020 until April of 2021, although the size of notified outbreaks increased again in March, when the alpha variant became dominant. From January to June 2021 the number of new outbreaks notified remained low (<10). In July and August, the number and size of outbreaks notified increased again as the delta variant became dominant in the Netherlands in July 2021. This increasing trend continued, peaking in November 2021 when 47 new outbreaks and 28 ongoing outbreaks were notified. Trends in the number and size of new and ongoing outbreaks are set in the context of concurrent national control measures in Figure 2. From January 2022 onward, control measures were gradually relaxed whilst the milder but highly transmissible omicron had become dominant in the Netherlands. Where outbreaks were notified, the mean size increased initially, and the number of outbreaks categorized as “ongoing” remained high.

Case-fatality rates

The mean CFR (Figure 3) fluctuated widely over time with an initial maximum of 41% (95% CI: 36%−45%). It decreased from December 2020 to zero case fatalities in June 2021. Subsequently, the CFR rose again to peak in September 2021, declining thereafter. Overall, the confidence intervals are wide (notably for June 2021 and September 2021, when only three and two new outbreaks, respectively, were notified). Concurrent control measures are presented in Figure 3.

Discussion

The “real-time” outbreak alert and notification system, MUIZ, was a collaborative response to an urgent need. It allowed rapid communication of information on infectious disease events in ECF to other healthcare providers and the PHS, to protect the health of both residents and staff. Whilst there is not a single point of registration for ECF, the PHS estimates that about 85% of care homes in the Rotterdam-Rijnmond PHS region registered with MUIZ. Many facilities had registered before the start of the SARS-CoV-2 epidemic, and they all maintained their registration throughout the pandemic. MUIZ has subsequently been adopted by 13 of the 25 PHS regions in the Netherlands. Combined with the overall number of outbreaks notified during the study period, this suggests substantial “buy-in” and utilization of the system by stakeholders.

Access to real-time data on the numbers of infections and deaths and CFR was a key indicator for public health action early in the pandemic, when high attack rates and deadliness of SARS-CoV-2 in vulnerable elderly populations became apparent. It is likely that the stability of the estimates changed over time as polymerase chain reaction and antigen testing became readily available and elderly residents survived the infection more often. In addition, the SARS-CoV-2 outbreaks were notified in a complex, highly dynamic epidemiological context related to rapid viral evolution and seasonal effects resulting in different waves, changes in diagnostics, introduction and withdrawal of control measures within ECF and at societal level, availability of vaccines, waning immunity, and later boostering. Ecological level data, such as reported here, is also notoriously prone to bias with under-diagnosis and differences in notification practices within and between ECF a practical reality. Despite the crude nature of the estimates, however, trends in the number and size of the outbreaks reported and CFR were credible and generally reflected the wider epidemiological context as reported by the Dutch government “Corona dashboard” for elderly care facilities. ¹¹

When the highly transmissible omicron variant emerged, the utility of the aggregate data was more limited. Notifiers were requested to close an outbreak notification after twice the maximum incubation period of SARS-CoV-2 (28 days). Omicron’s shorter incubation period (3.42 vs 6.65 days for wildtype ¹² ), fewer societal control measures and widespread viral reintroductions, likely led to concurrent, ongoing outbreaks that were then indistinguishable. This is reflected in the data as the number “ongoing outbreaks” increased in early 2022, while the size and number of new outbreaks decreased. This could reflect a lack of notification of outbreak closure on behalf of the notifier for myriad reasons, including increased workload, but it is also possible that not all new cases were notified due to less serious morbidity and a lower priority for notification for notifiers.

Despite these limitations, when expediency was key in practice, rapid access to aggregate data from MUIZ was vital for regional PHS and other stakeholder organizations involved in health care planning and delivery at regional level, including the elderly care facilities themselves. It allowed for an overview of the seriousness and impact of the infection as outbreaks evolved in real time; it provided an estimate of the burden of disease within elderly care facilities and helped to identify the needs for healthcare continuity; it allowed for a high-level comparison between outbreak locations, facilitating discussion on differences in characteristics between locations that might explain differential outcomes in morbidity and mortality. In practice, this high-level perspective was essential to allow for planning between the PHS and the institutions themselves, facilitating discussion on outbreak control measures, responsive management of staffing levels, and maintaining continuity of care.

Reflections on the data notified over the 2-year period highlight the challenge to maintain flexibility in the system and retain its adaptability for unforeseen events (e.g. ability to change criteria for opening/closing outbreaks as new information emerges), particularly as PHS regions increasingly adopt the system, rendering decision making more complex. Since its inception, MUIZ has become a more standard means of reporting institutional outbreaks to the PHS. The next steps for MUIZ include evaluation of the system to explore user profiles, their needs and barriers and motivators for use, and the potential for MUIZ as part of the pandemic preparedness process. Additional functionality will be added, allowing linkage between MUIZ and existing multidisciplinary electronic patient dossiers for older people with chronic conditions (Ysis ¹³ ). It is also intended to ensure compatibility with a new digital outbreak manager for use within healthcare institutions, that is currently under development. ¹⁴ In conclusion, MUIZ is a simple tool to notify outbreaks and exchange information that was highly acceptable to users during the first 2 years of SARS-CoV-2. It has high potential for adaptation and further development in similar institutional outbreak settings.

Supplemental Material

sj-docx-1-phj-10.1177_22799036231160634 – Supplemental material for Introducing a novel “real-time” outbreak alert and notification system to monitor SARS-CoV-2 outbreaks and case fatality in elderly care facilities, the Netherlands, 2020–2022

Supplemental material, sj-docx-1-phj-10.1177_22799036231160634 for Introducing a novel “real-time” outbreak alert and notification system to monitor SARS-CoV-2 outbreaks and case fatality in elderly care facilities, the Netherlands, 2020–2022 by Abraham Meima, Jane Whelan, Jan Dijks, Nicoline van der Hagen, Marco van Duuren and Aimée Tjon-A-Tsien in Journal of Public Health Research