Social networks and infectious disease: The Colorado Springs study

Klovdahl, Alden S.; Potterat, John J.; Woodhouse, Donald E.; Muth, John; Muth, Stephen Q.; Darrow, William W.

doi:10.1016/0277-9536(94)90302-6

Cited by 323 publications

(198 citation statements)

References 13 publications

Supporting

Mentioning

184

Contrasting

Unclassified

Order By: Relevance

“…As such, researchers attempted to construct a network census of high-risk heterosexuals in Colorado Springs, focusing particularly on sex workers and drug injectors and their sexual and drug partners (23)(24)(25)(26). We restrict attention to the giant component of the network, comprising 4,430 individuals and 18,407 edges, representing social, sexual, and drug affiliation.…”

Section: Methodsmentioning

confidence: 99%

Assessing respondent-driven sampling

Goel

Salganik

2010

Proc. Natl. Acad. Sci. U.S.A.

328

351

View full text Add to dashboard Cite

Respondent-driven sampling (RDS) is a network-based technique for estimating traits in hard-to-reach populations, for example, the prevalence of HIV among drug injectors. In recent years RDS has been used in more than 120 studies in more than 20 countries and by leading public health organizations, including the Centers for Disease Control and Prevention in the United States. Despite the widespread use and growing popularity of RDS, there has been little empirical validation of the methodology. Here we investigate the performance of RDS by simulating sampling from 85 known, network populations. Across a variety of traits we find that RDS is substantially less accurate than generally acknowledged and that reported RDS confidence intervals are misleadingly narrow. Moreover, because we model a best-case scenario in which the theoretical RDS sampling assumptions hold exactly, it is unlikely that RDS performs any better in practice than in our simulations. Notably, the poor performance of RDS is driven not by the bias but by the high variance of estimates, a possibility that had been largely overlooked in the RDS literature. Given the consistency of our results across networks and our generous sampling conditions, we conclude that RDS as currently practiced may not be suitable for key aspects of public health surveillance where it is now extensively applied.disease surveillance | snowball sampling | social networks T he development and evaluation of public health policies often require detailed information about so-called hard-to-reach or hidden populations. For example, HIV researchers are especially interested in monitoring risk behavior and disease prevalence among injection drug users, men who have sex with men, and commercial sex workers-the groups at highest risk for HIV in most countries. Unfortunately, however, these high-risk groups are not easily studied with standard sampling methods, including institutional sampling, targeted sampling, and time-location sampling (1).Respondent-driven sampling (RDS) (2-4) facilitates examination of such hidden populations via a chain-referral procedure in which participants recruit one another, akin to snowball sampling. RDS is now widely used in the public health community and has been recently applied in more than 120 studies in more than 20 countries, involving a total of more than 32,000 participants (5). In particular, in helping to track the HIV epidemic, RDS is used by the Centers for Disease Control and Prevention (CDC) (6, 7) and by the United States President's Emergency Plan for AIDS Relief.RDS is a method both for data collection and for statistical inference. To generate an RDS sample, one begins by selecting a small number of initial participants ("seeds") from the target population who are asked-and typically provided financial incentive-to recruit their contacts in the population (2). The sampling proceeds with current sample members recruiting the next wave of sample members, continuing until the desired sample size is reached. Participants are usually all...

show abstract

Section: Methodsmentioning

confidence: 99%

Assessing respondent-driven sampling

Goel

Salganik

2010

Proc. Natl. Acad. Sci. U.S.A.

328

351

View full text Add to dashboard Cite

show abstract

“…Potterat et al (15) Prostitutes' Number of Partners. We estimated prostitutes' number of partners from our prospective study of the sexual, drug-using, and social networks of persons (including prostitutes) presumed to be at high risk for HIV infection in Colorado Springs (16). We recruited respondents from the county sexually transmitted diseases clinic, HIV testing site, drug treatment program, and outreach activities, and then enrolled some of respondents' sexual, drug-using, and social contacts through a link-tracing design.…”

Section: Adjusting the Discrepancy For Prostitutionmentioning

confidence: 99%

Prostitution and the sex discrepancy in reported number of sexual partners

Brewer

Potterat²,

Garrett³

et al. 2000

Proc. Natl. Acad. Sci. U.S.A.

View full text Add to dashboard Cite

One of the most reliable and perplexing findings from surveys of sexual behavior is that men report substantially more sexual partners than women do. We use data from national sex surveys and studies of prostitutes and their clients in the United States to examine sampling bias as an explanation for this disparity. We find that prostitute women are underrepresented in the national surveys. Once their undersampling and very high numbers of sexual partners are factored in, the discrepancy disappears. Prostitution's role in the discrepancy is not readily apparent because men are reluctant to acknowledge that their reported partners include prostitutes.A cross the world, probability sample household surveys of adult sexual behavior show that men report substantially more sexual partners than women do (1-13). This finding is puzzling, because in a closed population of heterosexuals, men and women actually have the same number of sexual partners in the aggregate. Explanations for this discrepancy pertain to either sex-linked reporting bias or sampling bias. Sex-linked reporting bias means that, for whatever reason, men overreport and͞or women underreport the number of their partners. Sampling bias refers to the undersampling of women who have had many partners and͞or women with whom sampled men have had sex but are outside the sampling frame. In our analysis, we evaluated sampling bias related to prostitution as an explanation for the disparity.Measuring the Discrepancy Data. We used data from the 1988-1991 General Social Surveys (GSS) (14) and the 1992 National Health and Social Life Survey (NHSLS) (8). These cross-sectional surveys involved multistage area probability samples of adults living in United States households (the NHSLS included adults under age 60 only). Although the population of the United States is not closed, we treated it as if it were, given the consistency of the measured sex discrepancy in countries from each major region of the world and at different levels of development.We used data based on responses to questions in paper self-administered questionnaires (SAQs), except as noted. In both surveys, two questions ask about the number of sexual partners in the last 12 months and 5 years (asked in the GSS since 1991), respectively. Response options include 1, 2, 3, 4, 5-10, 11-20, 21-100, and more than 100. In the 1991 GSS, some respondents answered these questions in an open-ended fashion. For each recall period, we recoded open-ended responses to the closed-ended response categories. Following T. W. Smith's procedures (personal communication), we recoded response categories involving ranges to the midpoints and recoded ''more than 100'' partners to ''101.'' For the NHSLS 12-month recall period, we followed Laumann et al. (8) by using the greater of the SAQ response and the face-to-face response to a similar question. For both surveys, we defined heterosexuals as those who reported only opposite-sex partners for a given recall period.Our analyses of the GSS are based on data from 1988 to 1991 com...

show abstract

“…The remaining exogenous cases are (44) 77 (24) 170 (41) 52 (29) Hispanic 18 (13) 65 (20) 68 (16) 32 (18) Indian 24 (19) 81 (18) 21 (4) 3 (2) 2 (1) 5 (2) 6 (1) 2 (1) Other 3 (2) 18 (6) 4 (1) 8 (4) Unknown 1 (1) 4 (1) 79 (16) 27 (13) Residence in city 118 (92) 430 (97) 417 (84) 101 (94) 137 (99) 321 (97) 352 (79) 155 (78) Residence outside city 10 (8) 13 (3) 80 (16) 6 (6) 1 (1) 9 (3) 92 (21) 44 (22) Residence unknown 0 (0) 0 (0) 45 (8) 14 ( the sex partners named by cases in the two cities who subsequently became laboratory-confirmed cases. In Winnipeg, 16% of male contacts and 6% of female contacts were from outside the city, whereas in Colorado Springs, exogenous contacts were more evenly distributed between the genders, 21% and 22%, respectively.…”

Section: Resultsmentioning

confidence: 99%

“…19 Both social and sexual relationships can be illustrated and analyzed simultaneously, thus augmenting the context of STI transmission and indicating possible sex partners respondents may omit. However, groups of people directly and indirectly linked through sexual partnerships (sexual networks) are more commonly the focus of STI prevention and transmission research.…”

Section: Concepts and Methodsmentioning

confidence: 99%