|dc.description.abstract||With respect to vaccine-preventable diseases (VPDs), serological surveys are a special surveillance tool using the prevalence of disease-specific antibody to measure exposure to vaccination or infection in the population. A serological survey is based on the collection and testing of blood from a sample of the population, selected in some appropriate manner. Its application to individual vaccine-preventable diseases depends on a number of factors related to both laboratory and sampling methods. With respect to laboratory methods, relevant factors include whether the assay, developed and validated for serological diagnosis in individuals, is appropriate for use in a population-level survey, whether a threshold level of antibody has been validated as a correlate of protection against a particular vaccine-preventable disease, and whether the relationship between the immune status of the population and herd immunity can be determined. For sampling methods, the major consideration is the relative merit of purposive and opportunistic designs. This thesis examines methodological issues relevant to the validity of serological surveys of vaccine-preventable diseases and systematically reviews their application to date internationally, and with respect to illustrative diseases in Australia, to formulate recommendations for their future use.
The first chapter reviewed the literature relevant to key methodological issues including the validity and suitability for application at the population level of various serological tests originally developed for diagnosis in individuals; the availability and validity a correlate of protection against particular vaccine-preventable diseases; the interpretation of vaccine-induced immunity at population level in the context of herd protection; and finally the strengths and limitations of different sampling approaches. It concluded that the use of current serological tests varies greatly depending on type of pathogens and there are major methodological gaps or logistic barriers in some areas for some infections and tests. The step from description of immunological status in individuals to their implications at the population level is also complex. Based on synthesis of these findings, vaccine- preventable diseases were divided into two broad categories with respect to the validity and applicability of use of current serological methods in measuring and interpreting immunity at a population level: Group 1: suitable for serological surveys with minor caveats – measles, rubella, varicella, hepatitis B, hepatitis A, diphtheria, tetanus, poliomyelitis, influenza, Haemophilus influenzae serotype b (Hib), and meningococcal disease; Group 2: suitable for use only in restricted circumstances – pertussis, human papillomavirus (HPV), pneumococcal disease, and mumps.
In the second chapter, these methodological considerations were then placed in the context of a systematic review of the international literature on population-based serologic surveys for vaccine-preventable diseases, in both high income and middle-lower income countries, to formulate evidence on the practical application of serological surveys as part of programs in surveillance and vaccination internationally. The results of this work confirmed that the interpretation of serosurvey data and its contribution to developing and maintaining immunisation programs varied by disease and setting. In middle-lower income countries, the use of serological surveys has been limited, contributing most to control and vaccine impact evaluation for hepatitis B, poliomyelitis, measles and rubella, including congenital rubella syndrome (CRS). In high income settings, serological surveys for all major vaccine-preventable diseases have been extensively undertaken, serving either to inform vaccine strategy adjustment or vaccination campaigns for measles, rubella, poliomyelitis and diphtheria, or to provide baseline prevalence data at country level for newly introduced vaccines including human papillomavirus, varicella, meningococcal C conjugate (MCC) and pneumococcal conjugate vaccines (PCV). Serosurveys have also been used to study trends in the epidemiology of pertussis, and influenza and to parameterize disease transmission dynamic models to predict the impact of changes in vaccination strategy for varicella or HPV. In both high and middle-lower income country settings, this review provided evidence that serological surveillance can be a powerful tool complementing other surveillance methods in monitoring and evaluating immunisation programs, depending on both the validity and applicability of the available serosurvey methods for disease under study and the quality and availability of other surveillance in specific settings.
In the next three chapters, original analyses of data collected through the Australian serosurveillance program were conducted for three vaccine-preventable diseases—hepatitis B, diphtheria and tetanus. These three diseases were chosen because of their difference and the opportunities these afforded to illustrate a range of issues pertaining to test validity and the supplementary contribution of serological data to program monitoring at population level. Specifically, for hepatitis B, serological survey methods across three serosurveys had changed, and the implications of these changes in the context of Australia, a country with very low prevalence of hepatitis B virus (HBV) infection in general population, were unclear. Diphtheria was interest because elimination of disease had been maintained despite presumptive vaccine-induced protection, based on well-established antibody correlates of protection being absent in a significant proportion of the population. For tetanus, unlike almost all other vaccine-preventable diseases, clinical disease does not result in production of protective levels of circulating antibody (in this case antitoxin), so seroprevalence data provide a uniquely accurate picture of the distribution of vaccine-induced antibody in the population, unaffected by consideration about disease exposure, for correlation with data on disease incidence.
For hepatitis B, in 2007, compared with 1997-98, the prevalence of antibody to hepatitis B surface antigen (anti-HBs) increased significantly in all age groups below 24 years, more than doubling in children one to six years of age. Prevalence of past hepatitis B infection (anti-HBc) in adolescents aged 12 to 19 years reduced by 77%. The prevalence of anti-HBs was measured in vaccine-eligible age-groups to assess vaccine-acquired immunity; but as vaccine-induced antibody waned 38% in five years, and there is evidence that an anti-HBs level below 10mIU/mL does not necessarily indicate the loss of protection, anti-HBs alone is limited as a measure of vaccine impact. As testing for anti-HBc was not performed in anti-HBs negative individuals, prevalence of past HBV infection may have been underestimated, because a small proportion of anti-HBs negative sera can test positive for anti-HBc. It was concluded that testing sera for all three HBV serological markers (HBsAg, anti-HBs and anti-HBc) with appropriate testing algorithms was necessary for more complete evaluation of the impact of hepatitis B vaccine introduction, even in a low prevalence country like Australia. For diphtheria, only two-thirds of the Australian population aged three years or older had protective levels of diphtheria antitoxin considered indicative of protection (≥0.1IU/mL), but cases of diphtheria have been rare for decades. This discrepancy is consistent with diphtheria having a low basic reproduction number (R0) and herd immunity threshold, but also highlights uncertainty about the interpretation of the presumptive protective antitoxin level of >0.1 IU/mL in low incidence settings with respect to local or travel-associated exposure. For tetanus, the proportion of persons aged up to 49 years having protective antibody levels was very high, but older adults, particularly older women, often lacked protective titres. This finding is consistent with the pattern of age-specific incidence of tetanus over the past two decades in Australia, with the elderly accounting for the highest number of tetanus cases and almost all deaths due to tetanus. Serological results were congruent with available national tetanus morbidity and mortality data, suggesting the methodological approach for tetanus was reliable. Results of these three studies confirm that Australian serological surveillance program is a useful complement to routine surveillance data in evaluating immunisation programs, and that the use of opportunistically collected residual specimens is an appropriate and efficient way to conduct serological surveys. The consistency of its methodology is one of the major advantages of the Australian program, which ensures results are comparable with previous serosurveys. If any changes to methodology are made, potential implications of these changes on data validity and utility should be evaluated.
This thesis represents the most extensive international systematic review of methodological considerations for, and application of, serological surveys and also examines issues relating to three diseases in the Australian context, with reference to international data. It is concluded that measurement and interpretation of immunity in populations is complex, particularly the extrapolation of immunologic status at the individual level to population level. Despite significant caveats, for most vaccine-preventable diseases, serological surveys are a valid tool to evaluate and interpret population immunity, and can substantially enhance other data sources. For some vaccine-preventable diseases, there are significant limitations to the application of serological surveys, and other lines of evidence are needed to interpret their results in order to reach valid conclusions about population-level immunity.||en_AU|