The diagnosis and epidemiology of pertussis in Australia

Abstract

Pertussis, more commonly known as whooping cough, is one of the most common vaccine-preventable infections reported in Australia. In the last decade, pertussis incidence has increased in Australia and overseas, with large sustained epidemics occurring most notably in developed countries. Australia had a pertussis epidemic between 2009 and 2012, which peaked during 2011 with 38,602 notified cases. The role of improved diagnostic methods, particularly the use of polymerase chain reaction (PCR) testing, has been hypothesised to have led to improved case ascertainment and superior detection of disease activity. This Thesis is comprised of five distinct studies contributing to the investigation of this hypothesis. In Australia, the incidence of pertussis is predominantly monitored using notification data, with the case definition for notifications heavily reliant on laboratory confirmation. By analysing pertussis notifications between 1991 and 2013 by diagnostic test method, I was able to demonstrate that PCR quickly replaced all other laboratory methods for pertussis diagnosis from 2005 onwards, with the exception of persisting serology use in a proportion of adolescents and adults. In addition, a change in the age distribution of notifications was observed over time, with increasing pertussis incidence among older children and adolescents. As notification data can be biased by changes in testing practices, awareness, and definitions, my next study investigated pertussis testing trends in the general practice (GP, primary care) setting between 2000 and 2011 using data from the Bettering the Evaluation And Care of Health (BEACH) program. Among a stable set of GP encounters for respiratory illness, it was observed that the likelihood of individuals being tested for pertussis in 2010 2011 was seven times higher compared to ten years earlier. These findings, at least in part, suggest that notifications have been amplified due to increased testing. To gather more information about the contribution of increased testing to the pertussis epidemic, the incidence of severe pertussis cases admitted to intensive care units (ICUs) was reviewed, as the ICU setting is less likely to be biased by changes in diagnostic testing. Data on Australian paediatric pertussis-related ICU admissions between 1997 and 2013 were obtained from the Australian and New Zealand Paediatric Intensive Care (ANZPIC) Registry. I found that there had been an increase in pertussis-related ICU admissions during the epidemic period (2009-2012), likely indicating a true increase in pertussis activity in Australia. Focussing then on a subset of the ANZPIC Registry data, the diagnostic testing of pertussis-related paediatric ICU admissions in Queensland was reviewed. This analysis identified that PCR largely replaced other diagnostic methods over time and became the predominant diagnostic test, mirroring what had previously been observed among national pertussis notifications. With my final study, I investigated the impact of diagnostic testing changes on the understanding of pertussis epidemiology, focussing specifically on pertussis seasonality, which although a previously accepted phenomenon, can no longer be observed in notification data. By calculating the proportion of positive pertussis tests conducted, this study demonstrated that pertussis remains a seasonal illness, with peaks in summer each year, highlighting the need to monitor underlying testing patterns to truly understand disease epidemiology. When considered in combination, the findings of my Thesis support the hypothesis that while there was a real increase in pertussis activity in Australia between 2009 and 2012, the increased likelihood of testing and shift to more sensitive PCR-based diagnosis are likely to have magnified the recorded incidence of pertussis. The practical effect of these findings together with rapidly waning immunity following the introduction of acellular vaccines, and the removal of the vaccine booster dose for 18 month olds in 2003 in Australia, all contributed to the recent resurgence of pertussis. This Thesis makes an important and original contribution to understanding the recent resurgence of pertussis in Australia and overseas by addressing gaps in knowledge about changes in diagnostic testing. Although a true increase in pertussis incidence occurred in Australia, increasing PCR diagnosis allowed better case detection across all age groups and healthcare settings. Greater understanding of pertussis epidemiology as a result of PCR diagnosis should ultimately improve pertussis immunisation and control programs, through better identification of at-risk groups.