Article Text

P4.84 How many account for how much? lorenz curves and the concentration of condomless sex partnerships
  1. Peter Saxton1,
  2. Andrew Pasley1,
  3. Anthony Hughes2,
  4. Nigel Dickson3,
  5. Adrian Ludlam2
  1. 1University of Auckland, Auckland, New Zealand
  2. 2New Zealand AIDS Foundation, Auckland, New Zealand
  3. 3University of Otago, Dunedin, New Zealand


Introduction Lorenz curves, while developed to measure income inequality, can describe the unequal distribution of sexual contacts in a population. Sexual partnerships are typically skewed with most individuals reporting few and some reporting many, but the degree of concentration is poorly understand. A better understanding has important implications for HIV and STI prevention, care and research, notably targeted PrEP and test-and-treat.

Methods We analysed data collected from an ongoing national HIV behavioural surveillance programme among gay and bisexual men (GBM) conducted in offline and online community settings in 2014. Participation was voluntary, anonymous and self-reported. Data were collected on the number of condomless anal intercourse partnerships (CAIPs) in the previous six months. We aimed to (i) describe the distribution of CAIPs; (ii) measure the concentration of CAIPs using Lorenz curves; (iii) identify potential “core” groups and compare their characteristics to other GBM.

Results Of 3027 GBM responding, 1575 (52%) reported at least one CAIP. Of these 1575, the median, mean and range was 1, 3.5 and 1–250 CAIPs, and the distribution was skewed with 90%, 95% and 99% percentiles being ≤6,≤10 and≤39 CAIPS respectively. An aggregate 5525 CAIPs were reported, with 9% of GBM accounting for 52% of all CAIPs (those reporting >6 CAIPs); 5% accounting for 40% of all CAIPs (those reporting >10 CAIPs); and 1% accounting for 19% of all CAIPs (those reporting 40+ CAIPs). The Gini index was 0.59 indicating high inequality. GBM in these three “cores” were more likely to be older, diagnosed HIV positive, have recent STI diagnoses and negative attitudes towards condoms and safe sex obligations.

Conclusion A large volume of condomless sex partnerships in this community sample was generated by a small proportion of GBM representing highly sexually connected nodes. Clinically, interventions with these GBM such as PrEP and early diagnosis and treatment can disrupt transmission of HIV and STIs across sexual networks and must be attractive and accessible to them.

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.