The transmission dynamics of gonorrhoea: modelling the reported behaviour of infected patients from Newark, New Jersey

Abstract

A survey of the sexual behaviour of gonorrhoea patients in Newark was undertaken to evaluate parameters within a model of gonorrhoea transmission. Modelling work aimed to explain observed epidemiological patterns and to explore the potential impact of interventions. Reported behaviours, along with values derived from the literature, were used within a standard deterministic model of gonorrhoea transmission, where the population was stratified according to sex and rates of sex-partner change. The behaviours reported, particularly among women, are insufficient by themselves to explain the continued existence of gonorrhoea within the population. The majority of symptomatic patients seek treatment within a few days, and report that they do not have unprotected sex while symptomatic. The proportion of patients with low numbers of sex partners suggests that sexual mixing between people categorized according to sexual behaviour is near random. To explain the persistence of gonorrhoea, there must be some patients who, when infected, do not seek care in public clinics. In addition, gonorrhoea incidence in the model is sensitive to change, such that very small reductions in risk behaviour could lead to its elimination. This does not accord with the observed failure of many interventions to eliminate infection, suggesting that the modelled infection is too sensitive to change. The model, which has been influential in gonorrhoea epidemiology, is not consistent with the observed epidemiology of gonorrhoea in populations. Alternative models need to explore the observed stability of gonorrhoea before robust modelling conclusions can be drawn on how best to control infection. However, the current results do highlight the potential importance of asymptomatic infections and infections in those who are diseased and do not attend public health services. Screening and contact-tracing to identify asymptomatic infections in both men and women will be more effective in reaching those who maintain the infection within the community rather than simply treating symptomatic cases.

Frequency–dependent incidence in models of sexually transmitted diseases: portrayal of pair–based transmission and effects of illness on contact behaviour

“…We simulated epidemics for these two main classes of STD. To represent bacterial STDs, we drew parameter values from the literature on gonorrhoea, chlamydia and trichomoniasis Garnett et al 1999;Bowden & Garnett 2000): transmission rates are high, and infected individuals recover without immunity in roughly one month. To represent viral STDs, we chose parameters in the range appropriate for HIV and HSV-2 (Anderson et al 1989;Castillo-Chavez 1989;Mertz et al 1992;Blower et al 1998Blower et al , 2000: transmission is relatively slower, but there is no recovery.…”

“…The infection status of individuals (both human and animal) has been shown to affect their contact behaviour in STDs and other diseases (Kennedy et al 1987;Loehle 1995;Able 1996;Beckage 1997;Garnett et al 1999;Kiesecker et al 1999;Webster et al 2003), and evolutionary biologists have long speculated that parasites play a role in mating behaviour (Hamilton & Zuk 1982;Boots & Knell 2002). Proposed or observed mechanisms for lowered contact rates include debilitation and reduced vigour (e.g.…”

“…Thrall & Antonovics 1997;Blower et al 1998Blower et al , 2000Bowden & Garnett 2000;Sullivan et al 2001;Boots & Knell 2002). In their review of STD modelling, Anderson & Garnett (2000) point out that simpler models are complementary to complex simulations and help to extract general principles and reach robust conclusions, which aid in policy development (though see Garnett et al (1999) for a discussion of potential shortcomings). McCallum et al (2002) observe that simple treatments of transmission will continue to be necessary, and that correspondence with more complex models is an important area for investigation.…”

A metapopulation modelling framework for gonorrhoea and other sexually transmitted infections in heterosexual populations

“…Moreover, individuals in such locales formed sexual partnerships with each other more frequently than would be expected by chance (Potterat et al 1985;Zenilman et al 1999), suggesting a pattern of 'assortative mixing' between members of core populations. However, when Garnett and colleagues applied a modified version of the model by Hethcote and Yorke to empirical data (Garnett et al 1999), the model suggested that, contrary to the assortative mixing observed in empirical studies, 'near-random mixing' was required to reproduce the observed distribution of gonorrhoea. In addition, Garnett's work suggested that their modelling approach had other limitations in that it was highly sensitive to assumptions on the parameters that describe gonococcal transmission, and as such was inadequate for testing interventions aimed at reducing gonococcal transmission.…”

Antibiotic-resistant Neisseria gonorrhoeae spread faster with more treatment, not more sexual partners

“…Information about parameter estimates for N. gonorrhoeae is scarce, so we chose to use non-informative priors for all parameters except the duration of infection which was informed by Garnett et al [25]. The ranges for the expected prevalence and incidence of diagnosed infections in HMW were based on the National Health and Nutrition Examination Survey [26] and surveillance data [27], both from CDC.…”