key: cord-0294759-pdeotba5
authors: Clarke, E.; Horner, P.; Muir, P.; Turner, K. M. E.; Harding-Esch, E.
title: Online tests for sexually transmitted infections - Friend or Foe? An analysis of providers in the United Kingdom.
date: 2021-07-05
journal: nan
DOI: 10.1101/2021.07.01.21259784
sha: b3034fc7252fc38dd06150c243270259f05fbff6
doc_id: 294759
cord_uid: pdeotba5

Objectives: Online testing for sexually transmitted infections (STIs) may contribute to overcoming barriers to traditional testing such as stigma and inconvenience. However, regulation of these tests is lacking, and the quality of services is variable, with potential short- and long-term personal, clinical and public health implications. This study aimed to evaluate online tests available in the UK against national standards. Methods: Providers of online STI tests (self-sampling and self-testing) in the UK were identified by an internet search of Google and Amazon (June 2020). Website information on tests and care was collected, and further information requested from providers via an online survey, sent twice (July 2020, April 2021). The information obtained was compared to British Association for Sexual Health and HIV (BASHH) guidelines for diagnostics and standards of STI management. Results: 31 providers were identified: 13 self-test, 18-self-sample, and two laboratories that serviced multiple providers. Seven responded to the online survey. Many conflicts with national guidelines were identified, including: lack of health promotion information, lack of sexual history taking, use of tests licensed for professional use only marketed for self-testing, inappropriate infections tested for, incorrect specimen type used, and lack of advice for post-diagnosis management. Conclusions: Very few online providers met the BASHH national STI management guidelines standards that were assessed, and there is concern that this will also be the case in areas that were not covered by this study. For-profit providers were the least compliant, with concerning implications for patient care and public health. Regulatory change is urgently needed to ensure that online providers are compliant with national guidelines to ensure high-quality patient care, and providers are held to account if non-compliant.

• Online providers help overcome many barriers to STI testing and are increasingly 57 popular, but quality of services is not assured 58

• Many online testing services, particularly for-profit providers, did not comply with 59 national guidelines 60

• Substandard services can lead to serious personal, clinical and public health 61 implications, such as inappropriate testing, inappropriate antimicrobial prescribing, 62 unnecessary emotional distress and missed diagnoses 63

• Regulatory change is required to ensure online providers comply with national 64 guidelines and are held to account when they do not 65 66 . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted July 5, 2021. ; https://doi.org/10.1101 https://doi.org/10. /2021 

Sexually transmitted infections (STIs) are an increasing public health problem in the world, [1] 68 including the United Kingdom (UK). [2] Early diagnosis is a core intervention for guiding 69

A key concern relating to online testing is the quality of the tests themselves. Although 94 product regulatory standards such as CE-marking are often used to assess product 95 quality, [11] this may not always indicate good performance. [12] Poor diagnostic accuracy 96 can lead to false-positives resulting in unnecessary treatment, with resulting AMR risk and 97 relationship implications, and false-negatives enabling further transmission and RSH 98 sequalae. [13] Furthermore, some testing panels include infections that STI guidelines 99

recommend should not be routinely tested for, [14, 15] According to click through data, [17] the majority of Google users do not go past the first 116 page of search results, however for thoroughness the first five pages of results were 117 . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted July 5, 2021. ; https://doi.org/10.1101 https://doi.org/10. /2021 screened by title and description. All Amazon results were screened. Inclusion criteria for a 118 provider were: 119 1) The test had to be available in the UK 120

2) The test had to be either for self-sampling or self-testing 121

3) The test was not provided by an individual borough (administrative unit), as these are 122 geographically limited and often use National Health System (NHS) services. 123

Data on the tests available, price, specimen type and advice/information were extracted 124 iteratively from eligible websites. For products identified through third-party sellers, the 125 original provider was identified and any other tests they provided also recorded. 126

Further information was requested from providers and associated laboratories through an 128 online questionnaire, guided by categories identified during data extraction and BASHH 129 guidelines. [7] Questionnaires were tailored for each provider, depending on the tests they 130 provided and the information available on their website. The full set of questions is available 131

in Supplementary Table 2 . The questionnaire was first sent to providers in July 2020. In 132

March 2021, BASHH published a position statement regarding online services. This 133 highlighted inappropriate use of multiplex testing platforms, and suboptimal antibiotic 134 treatment regimens for bacterial STIs, [12] emphasising the presence of inappropriate 135 testing, incorrect specimens and poor treatment options that may have implications for AMR. 136

The statement called for increased regulation of these services, as some providers' practices 137 are inconsistent with national guidelines. The questionnaire was sent again in April 2021 to 138 the providers who did not respond in the first round, in the hope that the position statement 139 publication would increase the response rate. 140

Comparison with national and regulatory diagnostic guidelines 141 . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted July 5, 2021. ; https://doi.org/10.1101 https://doi.org/10. /2021 Data obtained from provider services were categorised into: test type, specimen type, 142 diagnostic accuracy, health information, follow-up/treatment, and accreditation. Comparison 143 of services offered with BASHH guidelines was then conducted. [7, 16] A full list of guidelines 144 used for specific pathogens is available in Supplementary Table 3 . 145

Overview of provider responses 147

The Google and Amazon search returned 13 self-test and 18 self-sample providers, as well 148 as two laboratories that serviced multiple providers. In the first round of surveys, two 149 providers completed the questionnaire, and one requested a phone call. The second round 150 of surveys prompted four more replies. Therefore, most information was collected from 151 provider websites. Provider names have been anonymised, in accordance with the survey 152 terms of consent (Supplementary Table 2 ). The main guidelines providers were compared 153 to are summarised in Table 1 . Tests available and specimen type requested are shown in 154 Table 2 , compared to national guidelines. Further details of tests identified can be found in 155

Supplementary Tables 3a and 3b. In general, the providers that were closest to the 156 guidelines were free services, providing an appropriate range of tests, correct sample types, 157

and up-to-date, comprehensive information. 158 . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted July 5, 2021. [7] These were used to assess whether online providers were providing an adequate standard of care to patients Standard BASHH Standards for STI Management 1: Test audience and pre-test process

• Online services are not appropriate for people with genital symptoms or symptoms of sexual infection

• There should be the recording of a medical and sexual history, including use of contraception and drug and alcohol risk assessment, whether taken by a healthcare professional or self-completed.

• Specimens for STI testing including as a minimum those for chlamydia, gonorrhoea, syphilis and HIV from all relevant exposed sites.

• Specimens for microbiological testing obtained during the examination should be in line with national guidance. This should include samples from extra-genital sites if indicated in the sexual history.

• If home sampling or testing kits are provided, it is the providers' responsibility to ensure there are clear and comprehensive written instructions on how to use the kit and how to access follow up and support if necessary 3: Health information and signposting

• The STI tests offered to individuals should be explained clearly, including what they are for, how samples are taken, the limits of each test i.e. window periods

• Health promotion and prevention interventions including encouragement of safer sex behaviour and condom usage.

• Providers should ensure that appropriate advice is given about effective preventative interventions such as post exposure prophylaxis

• Have clear agreed care pathways in place to ensure people have access to appropriate care based on their need.

• Laboratories should be United Kingdom Accreditation Services (UKAS) accredited and have evidence of External Quality Assessment (EQA), Internal Quality Control (IQC) and Internal Quality Assurance (IQA). Self-Sample

B = Blood; O=Oral transudate; S = Swab; Sp? = species unclear; U = Urine; US = urethral swab; V = Vaginal swab; ? = specimen unclear. This reflects where a provider explicitly states what sample is used. This is often not included on package tests, so may not reflect all tests offered; * = most tests requested urine, but a vaginal swab was available separately. † Indicates a free service

Standard 1 was often not met. BASHH advice is that symptomatic users are not suitable for 160 online services as they need to be examined, [7] yet providers specifically targeted this group 161 and recommended a large panel of tests. Additionally, five self-test providers offered tests 162 that were marked as being for professional use only. With regards to sexual history taking 163 pre-testing, some providers used an online questionnaire to recommend the most 164 appropriate options, but most providers did not have this feature. 165

Whilst providers did offer tests for the minimum requirement in standard 2, these were often 167 available individually and in various packages, leaving users able to pick and choose. For All 18 self-sample providers offered tests for chlamydia and gonorrhoea, but availability 174 varied for other tests (Table 1) . Self-sample tests were available in various combinations, 175 with as many as 12 tests in one bundle. Free services provided a smaller range of tests than 176 paid services. Ten self-sample services offered tests individually or within bundles for 177 organisms generally regarded as commensal, such as Ureaplasmas or Mycoplasmas, 178 however, it is recommended that these should not be routinely tested for. [14, 15] It was also 179 often unclear precisely which species was being referred to. Gardnerella infection was often 180 used as a proxy for bacterial vaginosis, contrary to recommendations. [19] Private providers 181 often exaggerated the pathogenicity and importance of testing for these commensal 182 organisms. For example, two paid services claimed an advantage over the NHS by testing 183 . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted July 5, 2021. Eleven self-test providers reported sensitivity and specificity estimates, all of which were 197 >85%, however information about reference tests used or sample sizes was often 198 unavailable. Due to lack of website information and low survey response rate, it was not 199 possible to obtain information on diagnostic accuracy from most self-sample providers. 200

Those who reported these results (n=5) gave values >95% for sensitivity or specificity, but in 201 three of these, only the HIV test performance was stated. 202

To assess standard 3, we extracted information on whether sites gave information on STI 204 symptoms, window periods, transmission details and health promotion (e.g. condom use). 205

Health information was often lacking or inconsistent. For self-tests, as the package insert 206 was only available for four providers, it was difficult to ascertain health promotion materials 207 that may be provided to those purchasing the tests. As self-tests were often marked as 208 professional use only, there could be lack of appropriate health information or signposting, 209 . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted July 5, 2021. ; https://doi.org/10.1101/2021.07.01.21259784 doi: medRxiv preprint as information would be targeted at a clinician. Using information from provider websites, 210 only five self-sample providers provided information on window periods, transmission details 211 and infection prevention. For other providers, information was not on the same page as the 212 test or was inconsistently mentioned across different tests. One provider gave links to 213

Wikipedia. Advice on accessing HIV post-exposure prophylaxis (PEP) was also not 214 mentioned by eight of the self-sample providers; guidelines stipulate that users in need of 215 this service should be directed to a clinic. [7] 216

It was difficult to assess whether standard 4 was met as often post-diagnosis processes 218

were not shared. Where post-positive diagnosis information was provided, options involved a 219 private consultation, treatment ordered online (mainly for chlamydia), or advice to visit a 220

General Practitioner (GP). Partner notification was often mentioned non-specifically and may 221 instead have been discussed post-diagnosis. Exact treatment options were unclear, however 222 at least one provider offered an oral course of azithromycin and cefixime for gonorrhoea 223 treatment which was easy to purchase online, instead of the recommended intramuscular 224 ceftriaxone as per BASHH recommendations, which requires a visit to a healthcare 225

Accreditation 227

Although the BASHH standards do not refer to accreditation for self-tests, it is recommended 228 that they hold the CE-mark.

[18] Eleven self-test providers had at least one of their tests CE-229

Drug Administration (FDA) accreditation. One self-test provider marked their chlamydia and 231 gonorrhoea self-tests with an NHS logo, describing themselves as an NHS provider, but 232 whether that product had received NHS endorsement was unclear. For self-sample 233 providers, United Kingdom Accreditation Service (UKAS) accreditation was claimed as 234 required in Standard 5, however, it was often used as a blanket term for the laboratory 235 . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted July 5, 2021. ; https://doi.org/10.1101 https://doi.org/10. /2021 without details of the specific laboratory service that had received accreditation.

[23] Of the 236 two laboratory providers we were able to identify, one had UKAS accreditation for certain 237 STI tests and the other had UKAS accreditation for tests other than STIs. Care Quality 238

Commission (CQC) accreditation was reported for 12 self-sample providers, although mostly 239 only for laboratories used for sample testing, as opposed to the providers themselves. 240

This study identified and analysed 31 providers of online tests in the UK. We found 243 significant areas of suboptimal service that often conflicted with national guidelines on STI 244 diagnostics and management. These included a lack of health promotion information, lack of 245 sexual history taking, use of tests licensed for professional use only marketed for self-246 testing, inappropriate infections tested for, incorrect specimen type used, and lack of advice 247 for post-diagnosis management. As a result, users are at risk of taking unnecessary tests, 248

with poor performance, that could lead to incorrect results, inappropriate management and 249 receiving inadequate clinical information and support. 250

The first limitation of this study was low questionnaire response rate from providers, despite 251 a follow-up in 2021 following the BASHH position statement publication,[12] meaning that 252 not all aspects of care could be evaluated. Data considered missing in our analysis may 253 have been available once the user had bought the test. Furthermore, data were extracted 254 from websites in July 2020 but providers may have subsequently updated their websites. 255

Although our internet search was comprehensive, it is not possible to identify all online STI 256 test providers. The sample analysed here may therefore not be fully representative of all 257

providers. This lack of representativeness may be further compounded by the small number 258 of providers who responded to the survey and for whom we therefore have more extensive 259 data. However, the low response rate we observed has been seen in similar studies where 260 providers were contacted for information, and is not unique to our study. [24, 25] In addition, 261 . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted July 5, 2021. as our study was unfunded, tests could not be purchased to identify whether information was 262 available post-purchase. This also meant we were unable to test the services independently, 263 either from a user perspective through a "mystery shopper" exercise, [26] or from a 264 diagnostic accuracy perspective by independently assessing test performance claims. These 265 are obvious next steps for future work. Despite this, we were able to collect large amounts of 266 information from provider websites, giving an accurate perspective of what a consumer 267 would experience when choosing to order an online test. 268

Whilst it was difficult to assess test performance in the identified providers due to lack of 269 available information and inability to perform independent evaluations, it is expected that test 270 performance was sub-optimal in at least some instances. The lack of appropriate health information given by providers poses a risk to users on 278 multiple levels. Access to healthcare professionals as part of online STI services is 279 recognised as important for offering information, technical assistance, and support such as 280 reassurance and navigating relationship implications. [28] Receiving accurate information 281 regarding appropriate services and tests is critical to providing appropriate patient care, 282 ensuring that patients receive the correct tests relevant to their situation. In contradiction to 283 this, we found that several online providers specifically targeted symptomatic patients, 284

whereas standard BASHH guidelines are clear that online services are appropriate for 285 asymptomatic individuals only, [7] as well as not being signposted to vital services such as 286 PEP for HIV. [7] Patients were also frequently offered testing for commensal Mycoplasmas 287 and Ureaplasmas, [14] which could lead to unnecessary costs, treatments, and results of 288 . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted July 5, 2021. ; https://doi.org/10.1101 https://doi.org/10. /2021 uncertain significance, [15, 29] resulting in emotional distress and poor antimicrobial 289 stewardship. These additional tests were only found in paying providers' services, 290 suggesting that they may be more motivated by profit than by high quality healthcare 291 provision.

[30] It would be important to understand why individuals choose to pay for testing 292 rather than opting for the free services, to help ensure patients are offered the best possible 293

Whilst this is the first time to our knowledge that an evaluation of UK online testing providers 295 has been conducted, studies in other countries externally assessing online test providers 296 have reported similar results. A 2010 study of online tests in America was able to perform 297 independent assessments of online STI test providers, finding that they were hard to contact, 298 and although self-tests had poor performance, self-sample tests had high accuracy.

[27] 299

Providers of chlamydia online tests in The Netherlands were found to often not meet quality 300 indicators regarding health promotion or follow-up (especially for self-tests), but the process 301 of an evaluation taking place did provoke providers to improve their service. [25] Similarly, an 302

Australian study of HIV self-tests showed that none conformed to national product 303 guidelines, and often had inadequate pre-test information and linkage to care. [24] These 304 studies demonstrate that sub-optimal online testing service provision is a problem across the 305 world. Actions such as publications highlighting short-falls and position statements with 306 recommendations may create short-term impacts. However, if there are no mechanisms to 307 maintain improved practice and prevent providers from, for example, appearing under a 308 different name,[31] these efforts are of little long-term benefit. For there to be sustained 309 improvements in patient care, regulatory change is needed so that providers are regularly 310 monitored and can be held to account. Part of this involves services frequently being 311 evaluated against national guidelines, which must also be continually updated to adapt to 312 evolving service provision. [8, 12 ] 313 314 . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted July 5, 2021. ; https://doi.org/10.1101 https://doi.org/10. /2021 

Online testing is a welcome addition to STI diagnostics, offering a convenient and flexible 316 option for users. However, the proliferation of providers that do not follow guidelines, in 317 particular for-profit sites, jeopardises these advantages and puts users at risk. If current 318 trends continue, online testing usage will increase, resulting in more online providers as 319 demand increases. Regulatory change is required to ensure that the standard of care 320 received online meets national guidelines to protect patients and the wider population from 321 the repercussions of underperforming or inappropriate tests. If we do not act now, patients 322 will continue to receive sub-optimal care with potentially significant adverse personal, 323 clinical, and public health implications. 324 . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted July 5, 2021. ; https://doi.org/10.1101 https://doi.org/10. /2021 Funding and all other required statements 335 KMET acknowledges support from HDRUK CFC 0129 and PH and KMET acknowledge 336 support from the Health Protection Research Unit in Behavioural Science and Evaluation, 337 NIHR 200877, at University of Bristol. 338 . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted July 5, 2021. ; https://doi.org/10.1101 https://doi.org/10. /2021 

Sexually Transmitted Infections and Screening for Chlamydia in England

The impact of the COVID-19 pandemic on prevention, testing, diagnosis and care for sexually transmitted infections, HIV and viral hepatitis in England

To be on the safe side': a qualitative study regarding users' beliefs and experiences of internet-based self-sampling for Chlamydia trachomatis and Neisseria gonorrhoeae testing

Self-sampling and self-testing for STIs and HIV: the case for consistent nomenclature

STI Testing Using Online Companies: Benefits, Drawbacks and Call for Official Guidance

British Association for Sexual Health and HIV. BASHH Standards for the Management of Sexually Transmitted Infections (STIs)

British Association for Sexual Health and HIV/The Faculty of Sexual and Reproductive Healthcare. Standards for Online and Remote Providers of Sexual and Reproductive Health Services

National Chlamydia Screening Programme Audit Report: internet-based chlamydia testing

Can remote STI/HIV testing and eClinical Care be compatible with robust public health surveillance?

Guidance: CE Marking

British Association for Sexual Health and HIV. BASHH Position Statement on the Innapropriate Use of Multiplex Testing Platforms, and Suboptimal Antibiotic Treatment Regimens for Bacterial Sexually Transmitted Infections

The natural history of Chlamydia trachomatis infection in women: a multi-parameter evidence synthesis

Should we be testing for urogenital Mycoplasma hominis, Ureaplasma parvum and Ureaplasma urealyticum in men and women? -a position statement from the European STI Guidelines Editorial Board

Diagnosis of some genital-tract infections: part 2. Molecular tests and the new challenges

British Association for Sexual Health

Online Reputation Statistics That Every Business Owner Needs To Know

HIV Testing and Self-Testing: Information update

British Association for Sexual Health and HIV. UK National Guideline for the Management of Bacterial Vaginosis

UK national guideline for the management of infection with Chlamydia trachomatis

UK national guideline for the management of infection with Neisseria gonorrhoeae

UK national guideline for the management of anogenital herpes

Evaluation and enumeration of online test providers for sexually transmitted infections, specifically chlamydia, in the Netherlands

O20 Online prescribing for sexually transmitted infections -what's on offer!

Utilising the internet to test for sexually transmitted infections: results of a survey and accuracy testing

Using the eSexual Health Clinic to access chlamydia treatment and care via the internet: a qualitative interview study

Are Mycoplasma hominis, Ureaplasma urealyticum and Ureaplasma parvum associated with specific genital symptoms and clinical signs in non-pregnant women?

E-health: potential benefits and challenges in providing and accessing sexual health services

Point-of-care tests using enzyme detection to diagnose Chlamydia trachomatis infection do not work

We thank all providers who responded to the questionnaire. We would also like to thank the 326 BASHH Bacterial Special Interest Group (BSIG) members for reviewing the manuscript. 327 EMHE would like to thank Julius Schachter who set the challenge to address this issue at 328 the 8 th Meeting of the European Society for Chlamydia Research, September 2016, Oxford, 329 UK. It is not yet the outcome you hoped for, but progress is slowly being made. 330

The authors declare no competing interests. 333