Article Text

PDF

Utilising the internet to test for sexually transmitted infections: results of a survey and accuracy testing
  1. Sherria L Owens1,
  2. Nick Arora2,
  3. Nicole Quinn3,
  4. Rosanna W Peeling4,
  5. King K Holmes5,
  6. Charlotte A Gaydos3
  1. 1Morgan State University, School of Community Health and Policy, Baltimore, Maryland, USA
  2. 2Johns Hopkins University, Homewood Campus, Baltimore, Maryland, USA
  3. 3Johns Hopkins University, Division of Infectious Diseases, Baltimore, Maryland, USA
  4. 4The Sexually Transmitted Diseases Diagnostics Initiative, World Health Organization, Geneva, Switzerland
  5. 5University of Washington, Departments of Global Health and Medicine, Seattle, Washington, USA
  1. Correspondence to Dr Charlotte A Gaydos, Johns Hopkins University, Division of Infectious Diseases, International STD Laboratory, 530 Rangos Building, 855 North Wolfe Street, Baltimore, MD 21205, USA; cgaydos{at}jhmi.edu

Abstract

Background Searching the internet for information about sexually transmitted infections (STI) is common. The goal of the study was to discover which internet sites offer STI tests and obtain information about the services and their validity.

Methods Using internet web-based search engines, information was collected from the sites about STI testing services, costs and types of tests offered, and tests were evaluated for accuracy. ‘Business’ functions regarding consent and return of results were investigated. Contact attempts were made by phone, e-mail or ‘contact us’ links and by mail. Test kits were ordered from six commercial internet sites and one public health site. Their accuracy was evaluated for chlamydia and gonorrhoea.

Results The study identified 27 national/international internet sites offering STI self-collection kits and services. Tests were available for gonorrhoea, syphilis, chlamydia, HIV, herpes, hepatitis viruses, trichomonas, mycoplasma and ‘gardnerella’. All attempts to administer the survey yielded unsatisfactory results. After sending the survey by mail/e-mail to all the sites, four responded, two with the survey. Six websites appeared invalid based on non-deliverable e-mails and returned letters. The remainder did not respond. Test results were obtained from five of seven ordered kits. Two websites who were sent mocked urine specimens never provided results. The two ‘perform-it-yourself’ kits yielded false-negative results. Two mail-in urine specimens yielded correct positive results. The public health site kit yielded correct positive results.

Conclusions The internet STI testing sites were difficult to contact and demonstrated unwillingness to answer consumer-specific questions. Test accuracy varied, with home tests having poor accuracy and mail-in specimens demonstrating high accuracy.

  • Chlamydia trachomatis
  • diagnosis
  • gonorrhoea
  • internet
  • internet testing self-sampling for STI testing
  • outreach services
  • public health
  • sexually transmitted infections
  • STI testing services
View Full Text

Statistics from Altmetric.com

Home testing may increase access to sexually transmitted infection (STI) testing, especially for adolescents.1 2 Often seen as an inexpensive and convenient alternative to a doctor visit, home-based, self-testing diagnostic tests may be the method of choice for some individuals. Home testing for those who are embarrassed or unwilling to go to a clinic could lead more people to get diagnosed and seek treatment.3–6 This may be appealing for adolescents, who are often at high risk of STI. The National Initiative to Improve Adolescent Health by 2010 has emphasised the need to reduce the proportion of adolescents and young adults with Chlamydia trachomatis infections and ‘challenges the Nation to create new ideas, methods, and strategies, to move forward promoting adolescent health’.7

The majority of sexually active teens do not routinely seek screening or diagnostic testing for STI.4 Teen-friendly STI services are not often available, and satisfaction with healthcare providers and provider style is particularly important to teens.4 Private internet services thus offer alternatives to clinics, which require parental support and payment.5 Adults and adolescents are increasingly using the internet to search for information about health care and STI.8 People seeking sex partners also use the internet, and may appear to be at greater risk of STI.9 Internet-use statistics are staggering. According to the Pew American Life Project, 168 000 000 Americans use the internet per day and the number is growing at 63 000/day.10

Nearly all adolescents in the USA have gone online, 75% have a home computer,11 and internet communicating may lead to high-risk behaviour,12 especially for those seeking sex partners.9 A disproportionate number of adolescents, who were frequent ‘chat-room-users’ have run away from home and experimented with drugs and alcohol.13 On the other hand, many adolescents have searched for accurate information about sex online.14 15 Youth Risk Behavior Surveillance Reports indicated that 47.8% of high school students have had sex.16 One study reported the average frequency of coital events was 0.94 per week17; young internet users seeking sex partners may be at high risk of STI.13 18 However, because the internet has the ability to provide accurate educational information, using it also to provide STI services has much appeal.

New online offer STI testing services and clinics are available, where individuals can drop off self-collected samples. Also accessible are internet sites that allow individuals to collect samples at home and mail them directly to a testing laboratory.19 Although some publicly funded services offer free testing, other home testing services may be expensive and are not regulated.20 We therefore sought to survey internet sites that offered self-collection/testing to obtain information about their services. We subsequently investigated the reliability of such services by ordering and evaluating the accuracy of such test kits.

Methods

We used several web-search engines (ie, Google, Yahoo, Ask, Dogpile and MSN), to identify internet sites offering tests for STI, and to obtain information about services, costs, accuracy, types of tests offered, privacy policies and return of test results. Internet business sites identified were contacted for an in-depth interview about their programme, using verbal consent for the interview. Contact methods included phone, e-mail and mailed correspondence in that order. A standardised survey instrument was used. Information requested on the survey included turn-around time for testing, participation in proficiency testing programmes, efforts to maintain specimen integrity during shipping, types of STI testing offered, type of tests used, quality control and quality assurance issues. Of special interest was whether the site provided treatment or referral for treatment of individuals with a positive test result. If the interview could not be conducted on-line, by e-mail, or by phone, a written survey was sent via US mail. The individual sites were informed in the survey that they might be sent a ‘mocked’ proficiency sample for testing in order to evaluate accuracy. Sites were assured that for publication and analysis purposes the websites would not be identified by name or URL.

The second part of this study involved ordering test kits from six commercial internet sites and one public health site (http://www.iwantthekit.org)19 (which is the authors' own site) and evaluating their performance using known quantities of C trachomatis and Neisseria gonorrhoeae. Instructions provided by the supplier were followed except that the kit collection/transport tubes or testing ‘wands’ and ‘cartridges’ were inoculated with 106 organisms per test. For the four kits using urine specimens, ‘mocked’ transport media with 106 organisms per test were mailed back to the testing facility. For the two ‘self-test’ kits for vaginal and urine samples, testing instructions were followed by laboratory personnel. Costs for our purchased kits ranged from US$74 (∼£40) for a single ‘self-test’ kit for chlamydia to US$99 (US$115 with shipping) (∼£62) for a combination chlamydia/gonorrhoea urine test. Both phases of the study were rated as ‘exempt’ for consent by the Johns Hopkins University Institutional Review Board.

Results

Information about STI testing obtained from internet searches

We identified 27 national or international internet sites offering a range of STI self-testing kits. Sites were identified as international (13), US-based (10) sites using URL extensions and contact information, or ‘distributor sites’ (four) (table 1). These commercial ‘distributor sites’ were designated as such if they simply sold STI test sample collection kits in bulk, but did not offer any testing services. The tests available, as listed, varied by site, and included one or more tests: N gonorrhoeae, syphilis, C trachomatis, HIV, herpes simplex virus (HSV), hepatitis A, B and C viruses (HAV, HBV, HCV), Trichomonas vaginalis, Mycoplasma genitalium and gardnerella/yeast.

Table 1

Information obtained from survey of internet sites

For the 23 US and international sites (other than the distributors of test kits), the most common test offered was for chlamydial infection (20), followed by gonorrhoea (13), HIV (nine), HSV and HCV (eight each), HBV (seven), syphilis (six), HAV (four), trichomonas (three), and M genitalium, gardnerella/yeast (one each) (table 1). Of the four distribution websites identified, three sold collection kits for chlamydia, four offered collection kits for gonorrhoea and one offered a collection kit for trichomonas (table 1).

Available contact information for sites

All of the international sites identified had contact information listed on their website; however, six of the websites appeared invalid based upon both non-deliverable e-mails and returned letters sent via the US mail. Of the US sites identified, all 10 had contact information listed on their website, but two of the contact sources appeared invalid. All four of the distribution sites recognised had contact information listed on their website; however one had invalid information, which did not allow contact.

It was noted from information available on some of the websites that services other than sexually transmitted disease (STD) testing were offered. One international site offered telephone consultation after testing for an additional fee, whereas one US site offered consults free of charge in the form a counselling referral. Of the total 27 sites identified, one US site (which is the authors' own site) also offered free testing and referred infected patients for treatment. There were no distribution or international sites that offered treatment services, treatment referrals, or any type of follow-up. Nine international, four US and all four of the distribution sites offered additional testing services, including drug, fertility and glucose tests.

One of the international sites sent their clients' results to external laboratories, while two of the sites offered ‘results in seconds’ using simple point of care rapid tests. One US site indicated that they used an external laboratory for mailed samples. Sites using external laboratories did not process the samples; they sent the specimens to a third party for testing and then reported the results to the participant.

Costs associated with STI testing/vaccination obtained from the websites

One US site (a public health site) offered free testing for chlamydia, gonorrhoea and trichomonas infections. The lowest costs associated with a US site were US$19 for a HIV 1/2+0 Plus* rapid test kit and US$23 for a syphilis rapid test kit. The lowest cost was US$19 (∼£10) for a single chlamydia test. The highest costs were US$799 (∼£429) for HIV proviral DNA by PCR and US$645 (∼£346) for an HIV antibodies test. For an STI panel that includes testing for gonorrhoea, chlamydia, syphilis, HSV types 1 and 2, the cost was US$399 (∼£214) (table 1).

Return of test results as noted by information available on the websites

For the 10 US sites and 13 international sites that provided testing, results could be sent via e-mail, US mail and by phone. This excluded HIV testing because those tests were anonymous (at the consumer's request) and confidential. The clients had to call for the results by identifying themselves by a series of numbers or a code. Listed on each site was information regarding their ‘confidential testing’ policies. ‘Confidential testing’ meant the clients' results and any other information would be a part of their medical record and would only be used for discussion with their medical professionals.

Survey responses

Of all 27 sites, only two, both US websites provided completed surveys. This refusal to participate was despite attempts to contact sites via the phone, by the website's listed contact information and by US and international postal systems. The two sites that responded used nucleic acid amplified tests (NAAT) for chlamydia and gonorrhoea. Collection devices/kits were supplied by the laboratories. Turnaround was 1–7 days for these sites, both of which were accredited by professional organisations and participated in laboratory proficiency programmes. Both laboratories reported having quality assessment/continuous improvement plans, biosafety programmes and laboratory information systems. Both also reported the use of US Food and Drug Administration (FDA)-cleared assays and electronic requests for tests.

Treatment or follow-up

Results for the provision of treatment obtained from the website varied by site. Most sites offered no treatment options. One site reported ‘see a healthcare professional for treatment as soon as possible’ when a positive gonorrhoea test was received. A couple offered more comprehensive follow-up, such as assistance in setting up a treatment appointment at a free clinic.

Results of test kits ordered from the internet

Of the six websites from which kits were ordered with payment by credit card, results were obtained from four. The seventh site was the authors' site. Two sites never returned results despite repeated requests. The two cartridges from one site for a ‘perform-it-yourself’ urine and vaginal swab combo CT/GC gave false-negative results when 106 organisms were applied (performed with four separate tests). Another ‘perform-it-yourself’ chlamydia kit for a vaginal swab or ‘wand’ was also negative when 106 chlamydia organisms were applied. One mail-back urine kit for gonorrhoea was correctly identified as positive, and another mail-back urine kit for gonorrhoea and chlamydia was correctly identified as positive for both. The kits from the author-affiliated site provided correct chlamydia and gonorrhoea results from mock-up specimens. The two UK websites could not locate results after mocked urine specimens were sent.

Discussion

Phase one of this two-phase study demonstrated that most internet sites that provide testing for STI were hard to approach and contact directly. We have focused on understanding the entire process involved in using an internet site as a mode for STI testing. Our findings should be useful for anyone considering accessing the internet for the diagnosis of STI. Earlier publications have demonstrated that urine can be used to test for several STI using NAAT; self-collected vaginal specimens have also been shown to be effective for the diagnosis of STI by NAAT; finger-stick whole blood can be used to screen for syphilis and oral fluid can be used to screen for HIV infection.19 21–24 However, phase two of our study, for which we purchased test kits from the internet, demonstrated that the quality of internet STI testing services and of STI tests offered through the internet were often poor. There should be a mechanism in place to monitor and document the quality of tests being offered by these sites, such as state, federal, or international regulation, as well as required participation in proficiency testing programmes. There is currently no mechanism worldwide for the regulation of products and services offered through the internet.20

From an STI prevention perspective, use of the internet to seek sexual partners has been shown to be associated with risks.9 15 18 However, internet use for STI information and testing may have potential benefits. In order for internet sites to offer innovative, comprehensive and flexible testing, internet website medical and laboratory directors and administrators need to be a part of the public health system for case reporting and disease control. This strategy could be enhanced by a social marketing campaign. The public currently has little means of measuring the success of internet-based disease control efforts. Enhanced oversight and surveillance is needed for those who use the internet for testing options and could improve the business aspect of the internet sites while helping with public health prevention efforts.

The findings in this report are subject to limitations. First, research results for phase one may not be generalisable due to low responder rates. Research on internet sites allows researchers to collect data in new ways, as it may provide anonymity to the participants. However, McFarlane et al9 demonstrated that initial face-to-face recruitment offered an opportunity to obtain participant buy-in, establish credibility and offer research information that may be difficult to accomplish online. However, face-to-face recruitment or any form of contact of internet site medical directors/administrators was not possible, as evidenced by high rates of non-response in our survey. Second, sites may not have been forthcoming about their testing and reporting practices if they perceived this survey as a business threat. This may be an indication of their unreliability, especially if their laboratory is not accredited and does not participate in proficiency programmes. Some of the sites could only be reached by e-mail; other sites listed mail, e-mail and fax contact methods. We had anticipated that this would improve our chances of receiving responses to the survey, but unfortunately it did not. Third, the survey instrument used involved extended questioning. For example, ‘Do you have a written lab procedures manual for the tests you perform?’ and ‘Is your lab enrolled in a proficiency program?’ These questions were meant to be answered by a laboratory director, but it was mostly impossible to reach this or any person. It was impossible to ascertain whether sites even had a medical/laboratory director.

Results of the phase two part of the study, which evaluated the accuracy of the test kits sold through the internet, were frequently inaccurate. Only two of the six kits purchased yielded correct test results and both were for mail-in urine specimens; two of the six provided no results at all. These two sites, both in the UK, were for urine, to which we added chlamydia and gonorrhoea, and neither ever reported results, despite many contact attempts. Of the two kits with correct results, the first offered testing for gonorrhoea; this was inferred to be a NAAT assay from the fact that urine is only approved as a test sample for NAAT assays. The second tested for both chlamydia and gonorrhoea and was also inferred to be a NAAT assay by the urine collection device. Neither site reported the kind of test they performed. Both of the remaining kits were ‘perform-it-yourself kits’ and performed inadequately. One was a cartridge point of care test, which tested for chlamydia and gonorrhoea from both urine and a vaginal swab. Although tested with 106 organisms in each case, the results were negative. Another test was a vaginal swab ‘wand’ for chlamydia only, which also tested negative when 106 organisms were added. These results are notable in that an infected individual may receive a false-negative result. Such individuals could face serious health consequences.

Self-obtained genital specimens are not yet cleared by the FDA for home collection, even self-obtained vaginal swabs, although they are highly acceptable to women.25 26 Looking to the future, this would be an important step in facilitating home collection for internet-recruited self-obtained specimens, if regulation becomes possible and the accuracy of samples collected via internet recruitment improves. The use of self-obtained vaginal swabs for the detection of chlamydia and gonorrhoea has been judged to be highly accurate.27 28 Our group agrees with the National Institute of Health's recommendation that the STI research community and the public health community should work together with the diagnostics industry to develop validation protocols for these alternative home-collected specimens, such as self-obtained vaginal swabs or penile swabs in order to obtain the data needed for FDA clearance.22

Consumers cannot know with any degree of certainty that test results are correct if the FDA has not approved home collection kits and sites are not regulated. Many of these sites are not accountable for either the provision or the accuracy of their products. In addition, the internet sites that offer STI testing should be encouraged to facilitate referral for treatment and enhance post-diagnosis counselling and reporting mechanisms. We could not discover whether any sites offered assurance of treatment. Face-to-face counselling provides lasting networks of support and links individuals to services, and thus these counselling programmes should be made available to people after receiving test results through at-home collection methods, including internet recruitment, in order to serve public efforts at disease control.

In summary, we found that STI testing services are available from the internet, and several sites offer STI kits for at-home testing; however, it maybe difficult for consumers to contact these sites and often the results are not accurate. The internet is becoming a portal that will be used more frequently for testing of STI. FDA clearance of self-collected urogenital swab specimens at home is badly needed to support better quality. We also recommend that regulatory controls be put in place to protect internet consumers from receiving inaccurate results, and to ensure that results are returned for purchased kits. With proper regulation, internet services could provide accurate, confidential results for individuals who are reluctant to use clinical services or may not attend clinics for routine testing.

Short summary

National and international internet sites offering STI self-testing kits and testing services were surveyed and evaluated for test accuracy. Most sites were difficult to contact. Some sites offered ‘perform-it-yourself’ kits that gave inaccurate results.

Key messages

  • The internet is becoming a portal that will be used more frequently for testing of STI.

  • STI testing services are available from the internet and several sites offer kits for at-home testing; but it is difficult for consumers to contact these sites.

  • Often internet-solicited STI results are not accurate.

  • FDA clearance of self-collected urogenital swab specimens at home is required. Regulatory controls are needed to protect internet consumers from receiving inaccurate results.

  • With regulation, internet services could provide accurate results in a confidential manner for those who are reluctant to use clinics for routine testing.

References

View Abstract

Footnotes

  • Competing interests CAG is the founder of one of the web sites reviewed: http://www.iwantthekit.org.

  • Ethics approval This study was conducted with the approval of the rated as exempt for consent by Johns Hopkins Institutional Board.

  • Provenance and peer review Not commissioned; externally peer reviewed.

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.

Linked Articles

  • Whistlestop tour
    Jackie A Cassell