Background
Data from the 2021 WHO global progress report on HIV, viral hepatitis, and sexually transmitted infections indicate a global incidence of 128 million new chlamydia and 82 million new gonorrhoea cases in 2020 [
1]. Representing neglected pandemics, these infections cause a significant global disease burden. There are population groups who are disproportionately affected by STIs, including men who have sex with men (MSM), sex workers (SW) and their clients, transgender people (TG), adolescent girls and young women (AGWY), and pregnant women [
1]. There is also a high prevalence and incidence of STIs among people taking pre-exposure prophylaxis for HIV (PrEP) [
2] and young women attending contraceptive services in East and Southern Africa [
3], many of which would have been missed if syndromic STI management had been used. This has led to a push for greater access to aetiological testing in PrEP programmes [
4]. Major gaps persist in the availability of diagnosis and treatment for CT/NG, with STI programmes and services generally underfunded despite high levels of morbidity and mortality.
To control STIs, earlier detection and treatment are needed. Yet, a significant challenge is that most STIs are asymptomatic and require testing to identify infection. Undiagnosed and untreated STIs can lead to onward transmission and morbidity such as reproductive organs inflammation, reproductive morbidity and infertility, and vertical transmission to neonates. Viral and bacterial STIs can increase the risk of acquiring HIV, as increased viral loads of HIV can be found in genital tracts during STI coinfection [
5]. Furthermore, inappropriate management of gonococcal infection may accelerate the emergence of multidrug-resistant NG [
6]. This underscores the need for aetiological diagnosis to optimise STI management.
An aetiological diagnosis that tests all appropriate anatomic sites is needed. Evaluation for CT/NG at extragenital sites is critical for some population groups (e.g., MSM, SW, TG), as a significant proportion of infections would be missed if only genital testing were undertaken [
7]. Studies have demonstrated that up to two thirds of NG cases would be missed if only urethral or urine samples were tested in MSM [
8]. CT and NG are highly transmissible and often asymptomatic, and early detection relies on regular and comprehensive testing of multiple anatomic sites for those at higher risk [
8,
9]. Since 2010, the US Centres for Disease Control (CDC) have recommended using NAAT to test for extragenital CT and NG, as molecular testing improved sensitivity compared to culture [
10]. The current Australian STI management guidelines recommend that pharyngeal, anorectal, and urethral testing is undertaken in asymptomatic MSM [
9]. In women, a pharyngeal swab and anorectal swab are also recommended depending on reported sexual practices [
9].
Whilst testing from three anatomic sites individually would be ideal, the increased cost of NAAT over culture is a major limitation [
11], especially in low- and middle-income countries and other resource-limited settings. Furthermore, testing multiple anatomic sites separately can increase costs and workload, especially when implementing testing at or near the point of care. Several studies have investigated the pooling of specimens from triple anatomic sites from a single individual, but its accuracy varied across studies [
12‐
14]. Currently, there is no clear consensus whether pooling has adequate accuracy for populations at higher risk. Furthermore, a 2018 UK study reported that most clinicians regarded the existing evidence of pooling as insufficient to justify implementation in clinical practice [
15]. Since 2018, many more studies have been published, and a critical appraisal of all available evidence is helpful to guide future guidelines and practices.
If pooling of samples from the pharynx, urethra/endocervix, and anorectum within a single individual is demonstrated to be both highly sensitive and specific for Chlamydia trachomatis and Neisseria gonorrhoea detection, this could provide significant cost savings and influence national guidelines and clinical practice. The primary aim of this systematic review was to review and critically appraise the existing evidence regarding the diagnostic accuracy of pooled samples from triple anatomic sites of one individual for the testing of CT/NG using a single sample from one anatomic site as the reference. The secondary aims were to assess the cost impact of using pooled specimens, the patient and provider acceptability of the pooled sample approach, and the effects of pooled testing on health equity.
Discussion
Our systematic review appraises the current evidence of the diagnostic accuracy of pooling from triple anatomic sites for CT/NG testing. The combined sensitivity for CT and NG were 93.1% and 94.1%, respectively. The combined specificity for CT and NG was 99.4% and 99.6%, respectively. Programmes and services will need to assess whether the small decrease in sensitivity associated with multisite pooled testing warrants the substantial cost savings and potential improvements in health inequity.
The benefit of pooling will be restricted to testing for those reporting extragenital sexual practices. In a UK study involving MSM and women from the general population, Wilson et al. showed that pooling could save up to £18.22 per individual tested, a significant cost saving when multiplied by the number of people tested at the population level [
22]. Viewed a different way, if there is a fixed budget, pooling could increase the numbers of people tested and the frequency of testing for those at higher risk [
21,
22]. In particular, this could enable greater access to testing, including more regular testing for those with a higher risk of STIs. This includes individuals taking PrEP, where most national guidelines recommend routine triple-site testing for users [
29]. Pooled testing encourages multisite STI testing at a lower expense, which is more effective in detecting CT/NG infections than the single-site testing in which many extragenital infections would be missed [
20,
27]. The prevalence at the extragenital site is usually higher than the genital site, and they are usually asymptomatic; therefore, both anorectal and pharyngeal specimens are recommended to be included in testing [
30]. Verougstraete et al. [
20] reported that 40% of CT and 60% of NG infections would have been missed if only genital samples were tested. Similarly, if only urine samples were tested among 76 participants enrolled in Badman et al. [
27], 82% of CT and 85% of NG infections would be missed. Furthermore, the background prevalence of CT/NG influences the cost-effectiveness of pooled testing. Pooled testing would demonstrate higher cost savings in low prevalence settings/populations or high prevalence settings/populations if retesting was not required when a uniform treatment protocol is used as recommended by WHO as the first line for both CT/NG [
31]. In addition to the potential for cost savings, participants reported high acceptability to self-collecting samples, and we found no significant difference in the diagnostic accuracy compared to clinician collected samples. Therefore, self-collected pooled specimens could further reduce barriers to STI testing [
20]. By ensuring multisite testing among relevant populations, pooling can expand the testing coverage and increase the frequency of testing among those at higher risk of infection. This could potentially limit the CT/NG pandemic and reduce HIV transmission.
There are several potential limitations to multisite pooled testing. The lack of anatomic-site specific results was a common concern in the studies reviewed and suggests the need to retest those with positive results. Although some guidelines have some differences in the treatment recommended for CT and NG depending on the site of infection, in the current WHO guidelines, the first therapeutical line recommended can be the same for infections regardless of anatomic site [
31,
32]. Using pooled testing may impair epidemiological data collection for reporting CT/NG infections, as the site of infection would not be known without testing individual sites. Whilst demonstrated to reduce the total costs of testing, pooled testing may still not be affordable in resource-constrained settings without established STI testing infrastructure or not adopting molecular point-of-care testing for STIs or other infections such as tuberculosis or to measure HIV viral load. Not only does pooled testing require access to NAAT, but it also requires laboratory processes for combining samples before testing. Considerations include the transport and handling of samples, mixing of samples, the amount of diluent used, potential contamination of samples, laboratory staff training, and the storage of individual samples for re-testing if required [
15,
20]. Finally, with lower bacterial load in the oropharynx compared to the genital and anorectal sites [
33], there is a potential for pooled testing to miss oropharyngeal infections [
13,
26,
27,
34]. Future studies should investigate how to further optimize detection of oropharyngeal infections.
Limitations of this review include the high variation in the pooling method and the assays used for each study. We are therefore unable to comment on an optimal method of pooling. This review is also limited by our exclusion of studies in languages other than English. Additionally, our data was mainly gathered from high-income countries (82.6%, 14/17) and used MSM as a study population (70.6%, 12/17), limiting the generalisability of this review to other populations and low- and middle-income countries. Studies were conducted in the community outpatient setting (47.1%, 8/17), where symptomatic people are more likely to attend and therefore increase the likelihood of returning a positive result.
Our systematic review has highlighted areas for further research, including the feasibility and patient acceptability of self-collected sampling combined with self-pooling and feasibility from a laboratory staff perspective. Appropriately powered studies are required for the evaluation of pharyngeal NG and CT sensitivities. Further studies should also be undertaken to assess the acceptability of pooling by providers considering the synthesis of evidence and approvals from peak bodies such as the WHO. Future studies should be conducted to test the impact of urine volume and order of swabbing on the diagnostic accuracy of pooled samples. In addition, there is a need for more implementation studies to assess any treatment delays or additional costs associated with retesting individual samples if anatomic site-specific information is needed to guide treatment.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.