A Previously Uncharacterized Infectious-like Event Leading to Spatial Spread of Deaths Across England and Wales: Characteristics of the most Recent Event and a Time Series for Past Events

Abstract

Aims: In 2003, 2008 and 2012, deaths in the UK showed unexpected and unexplained large increases similar in magnitude to the larger influenza epidemics seen before 2000. However there were no unusual levels of influenza to explain these spikes. This study aims to investigate the spread of a presumed infectious agent across England and Wales during the 2012 event, and to establish a longer time-series for these outbreaks. Study Design: Longitudinal study of deaths. Place and Duration of Study: Deaths for residents of Local Authority (LA) and regional areas in England and Wales from January 2009 to July 2014. Analysis of monthly deaths from January 1951 to December 2012 for the whole of England & Wales, to detect events prior to 2012. Analysis of calendar year deaths between 1963 and 2013. Methodology: Running twelve month totals are used to detect the onset of a step-like increase in deaths which endures for twelve to eighteen months before abating. Results: These events can be traced back to the early 1950’s where they were intertwined with influenza epidemics. Moderately slow infectious-like spread across the UK occurs over a two year period. The last event which peaked in 2012 and 2013 led to a minimum estimate of 42,000 excess deaths, although 60,000 deaths are a more likely estimate. An additional event in 2010, which affects around 30% of LAs, appears linked with the swine flu epidemic and leads to underestimation of deaths in these locationsfor the 2012 event. The magnitude of the increase associated with the 2012 event decreased with increasing LA size, an effect which is due to the modifiable areal unit problem (MAUP). These events are always linked to large increases in emergency medical activity and emergency department attendances. The increase in deaths at local authority level is highly variable ranging from +5% to +30%, and this variability leads to large differentials in the cost pressures experienced by the local health services. In particular, 90% of LAs in London experienced the lowest percentage increase from the 2012 event.As a general rule it was observed that local authorities experiencing a low percentage increase in death for the 2012 outbreak, had experienced a high percentage increase following the 2008 outbreak and vice versa. This implies that the events are linked to the same agent. Conclusion: The existence of a new type of infectious outbreak, with relatively slow spread, has been confirmed and is part of a longer time-series of outbreaks. Large numbers of deaths appear to be associated with each event, although the 2012/2013 event appears to have led to the highest number of deaths in any of these outbreaks since 1950. Urgent action is required since the next outbreak is due around 2016 to 2018.