Sexually Transmitted Infection Symptom Reporting to AI-Based Virtual Triage and Care Referral

George A. Gellert, MD, MPH, MPA ^1, Maria Marecka, MD ^2, Gabriel L. Gellert, BS ^1, Tim Price, MS ^3, Katarzyna Trybucka, MD ^{4, Mateusz} Palczewski, MD, PhD ⁴

1. George A. Gellert, MD, MPH, MPA Infermedica, San Antonio, USA
2. Maria Marecka, MD Infermedica, New York City, USA
3. Gabriel L. Gellert, BS Infermedica, San Antonio, USA
4. Tim x Tim Price, MS Infermedica, London, United Kingdom
5. Katarzyna Trybucka, MD Infermedica, Wroclaw, Poland
6. Mateusz Palczewski, MD, PhD Infermedica, Wroclaw, Poland

OPEN ACCESS

PUBLISHED: 31 May 2025

CITATION: Gellert, GA., Marecka, M., et al., 2025. Sexually Transmitted Infection Symptom Reporting to AI-Based Virtual Triage and Care Referral. Medical Research Archives, [online] 13(5). https://doi.org/10.18103/mra.v13i5.6475

DOI https://doi.org/10.18103/mra.v13i5.6475

COPYRIGHT: © 2025 European Society of Medicine. This is an   open-access    article distributed    under the terms of the Creative Commons Attribution License, which    permits    unrestricted    use, distribution, and reproduction in any medium, provided the original author and source are credited.

ISSN 2375-1924

ABSTRACT

Objective: Evaluate the effectiveness of artificial intelligence-based virtual triage and care referral in describing, engaging and influencing care seeking behavior among patients with symptoms indicative of sexually transmitted infections.

Methods: Evaluated 4,487,191 virtual triage encounters among patients aged 12+ years over 36 months for sexually transmitted infections ranked as top three conditions. Demographics and type of care seeking before and following virtual triage were quantified and relationships assessed for statistical significance.

Results: A total 115,109 (2.6%) virtual triage encounters were consistent with a possible sexually transmitted infection. Most patients were female (64.5%) and 89.2% were aged 18–44. Over half of patients completed the virtual triage encounters in English or Spanish. Frequency of symptoms consistent with sexually transmitted infections reflected low detection rates due to the largely asymptomatic nature of these conditions. Of 5,717 patients completing pre- and post-triage care intent surveys, care seeking increased 6.1% post-triage (p<0.05); those uncertain about their care intent or planning no action decreased 20.0% (p<0.05). However, 46.3% of patients opted for contraindicated self-care measures post-triage, a 13.8% increase (p<0.05).

Conclusions: Virtual triage substantially influenced care seeking among individuals reporting symptoms suggestive of a sexually transmitted infection. Patients unsure how to address their symptoms or who planned inaction decreased significantly, while users intending to seek care increased. Virtual triage reduced indecision among patients by converting passive symptom acknowledgement into proactive intent to seek care. Increase in patients opting for self-care measures after virtual triage indicated a possible sexually transmitted infection, and despite a recommendation to seek care, suggests that certain barriers to care seeking for these infections may be too deeply rooted to be addressed by virtual triage technology alone.

Keywords

virtual triage and care referral; artificial intelligence; digital triage; sexually transmitted infections; telemedicine; STI disease control

Introduction

Sexually transmitted infection (STI) rates continue to rise globally and remain a significant public health challenge with low detection and treatment rates due to multiple barriers to care. In 2024, the World Health Organization (WHO) reported a large global increase of over 1 million STIs acquired daily. In the USA, chlamydia, gonorrhea, and syphilis reached record highs in 2019, the sixth consecutive year of STI rate increases, with only a small decrease in 2023. Sexually transmitted infections affect 1.2 billion people globally, mostly young in low and middle income countries. Prompt diagnosis and treatment of STIs is imperative to reducing community transmission. Social factors influence STIs, including class, race, gender, and sexual orientation. While diagnosis and treatment of STIs have improved, social stigma remains a major barrier to testing and treatment, causing embarrassment and shame, fear of unfair judgment and discrimination, and results in delayed diagnosis and treatment. Timely access to appropriate care is critical for preventing long-term complications and controlling community spread, but this access is often hindered by the asymptomatic nature of most STIs until late in the course of illness. It is estimated that 20% of the US population had an STI on any given day in 2018, yet only 2.2 million cases of chlamydia and gonorrhea were diagnosed, just 0.7% of the population. This is partly due to lack of symptoms, with only 6% of females and 11% of males with chlamydia and 45% of males and 14% of females with gonorrhea ever becoming symptomatic. Detection of STIs is further undermined by a majority of patients underestimating their STI risk and delaying care seeking. Over 85% of women and 70% of men perceived themselves as not at all or not very much at risk of STIs despite reporting high risk sexual behaviors. Among individuals experiencing STI symptoms, over 38% reported delaying care seeking. Care barriers include limited access to healthcare services, particularly in rural or underserved areas, socioeconomic inequities, social stigma, low sexual health literacy, financial impediments, and concerns about confidentiality. Innovative applications of health information technology have been explored to manage STIs and to enhance patient knowledge and symptom recognition. User reviews of an online sexual health symptom checker showed that the tool was perceived as providing helpful individualized, evidence-based health information to improve healthcare seeking. One study found that online STI evaluation ameliorated some patients’ feelings of shame and stigma. Public use of symptom checkers and in particular artificial intelligence (AI)-based virtual triage and care referral (VTCR) has grown, as they offer an easily accessible form of remote healthcare, providing evidence-based clinical triage on the go. Due to their anonymity, online symptom checkers are a platform where patients can comfortably disclose STI symptoms, mitigating social stigma. By providing automated, anonymous health assessments on demand, virtual triage and care referral may reduce barriers related to accessibility and lack of health literacy. Very little research has evaluated whether VTCR can improve patient ease and experience when reporting STI related symptoms, reduce stigma and care delays, and expedite appropriate care referral. AI-based VTCR has demonstrably improved post-triage clinical care acuity alignment, with 35% of patients altering care plans to match the recommendation of virtual triage. Virtual triage and automated care referral can enhance early detection of high incidence and life-threatening conditions, potentially reducing diagnostic and care delays which negatively impact clinical outcomes. It may potentially improve patient experience and improve STI detection and care referral. While studies have assessed the accuracy of virtual detection of STIs, few describe the patient demographics or impact of VTCR on patient care seeking behavior. The present study seeks to explore and reduce this knowledge gap.

Methods

STUDY OBJECTIVE:

To evaluate the utility of an AI-based VTCR engine in engaging patients with symptoms suggestive of STIs, and assess its effectiveness in helping them seek clinically appropriate healthcare. The study aimed to gather data on patient demographics and clinical characteristics in order to understand the potential role of VTCR in improving STI surveillance, detection and care delivery.

STUDY DESIGN:

A retrospective cohort design collected patient self-reported data and demographics through a free online VTCR engine, Symptomate, over a three-year period from October 2021 to October 2024. Patient reported pre- and post-triage healthcare seeking intentions were collected from April 1, 2022 to October 1, 2024, when the intent survey was implemented in the VTCR online platform.

SETTING AND DESCRIPTION OF INTERVENTION/VIRTUAL TRIAGE ENGINE UTILIZED:

The Infermedica Symptomate VTCR engine is designed for general public use and completes evidence-driven analyses of 800 diseases, 1,500 symptoms, and 200 risk factors. Utilizing AI, VTCR assesses symptoms shared by patients and identifies probable conditions which correlate with the clinical presentation and history. Virtual triage then refers the patient to the safest and most acuity-level appropriate care by identifying potential somatic or mental health symptoms that warrant further professional evaluation. The VTCR encounter concludes by providing a summary and analysis of the patient’s reported symptoms, along with a recommendation to engage self-care, visit a primary care or specialist physician on an outpatient basis routinely or urgently, or proceed to an emergency department (ED) via self-transport or ambulance. Data was extracted from the Symptomate application, which is available in 24 languages, and has over 19 million completed evaluations.

SAMPLE SELECTION AND INCLUSION/ELIGIBILITY CRITERIA:

During the 36-month study, a total of 4,487,191 VTCR encounters were completed by patients aged 12 or older. Study eligibility criteria included completed encounters in which the VTCR engine identified a sexually transmitted or related illness as one of the top three most likely conditions, yielding a study sample size of N=115,109. Because data for these analyses were drawn exclusively from completed encounters, where all demographic variables were fully recorded, there was no need to employ data imputation methods, nor were any cases excluded due to missing data.

DATA CAPTURE AND ANALYSES:

Prior to and after completing a VTCR encounter, patients could opt in to report their intended care seeking action, enabling a comparison of pre- and post-VTCR care seeking intent. This enables an assessment of the impact of VTCR in educating patients about and improving care acuity alignment, potentially reducing care seeking delays. Virtual triage and care referral recommendations were divided into four levels of care acuity: self-care, routine outpatient consultation when possible, urgent outpatient consultation (within 24h), and ED care. Cross-tabulations were completed and tables generated comparing male and female patient segments for symptom and care intent reporting.

ETHICS STATEMENT:

All patients in this study provided explicit consent prior to all VTCR encounters for their data to be used for research purposes, with analyses reported in a fully de-identified, anonymous manner in the aggregate. Ethics review board approval was waived as a result.

Results

PATIENT DEMOGRAPHICS AND MOST COMMON SEXUALLY TRANSMITTED INFECTIONS:

A dataset of 4,487,191 Symptomate encounters completed over the three-year study period was examined. Of these, 115,109 (2.6%) were consistent with a possible STI or STI-related condition and ranked as one of the top three likely conditions by VTCR. Encounters may have been identified as having more than one STI or STI-related condition within the top 1-3 rankings. As a result, the total number of encounters in Table 1 exceeds 115,109; however, analyses were conducted on unique encounters to avoid duplication. Regarding patient demographics, 64.5% were female. Patients aged 18 to 44 years old comprised 89.2% of the sample. The encounter languages selected by patients were analyzed, with English most widely used (37.0%) followed by Spanish (20.0%) and French (9.6%). Less frequently used languages included Russian (7.6%), German (5.4%), Polish (5.1%), Portuguese (4.5%), Arabic (3.7%) and Italian (3.2%). Among all users, 3.4% of all encounters identified an STI as one of the top three conditions. The most common STI or STI-related condition was trichomoniasis (26,748 or 0.4% of all encounters and 23.2% of all unique STI encounters), gonorrhea (25,596 or 0.4% of all encounters and 22.2% of all unique STI encounters), and pelvic inflammatory disease (17,434 or 0.4% of all encounters and 15.1% of all unique STI encounters).

PRE- AND POST-VTCR CARE SEEKING INTENT:

Findings among 5,717 patients who completed the pre- and post-triage care seeking intent surveys between April 1st, 2022, and October 1st, 2024 are shown in Table 4. Pre-VTCR, most intended to engage self-care or did not know/were uncertain about what care to seek (67.2%). Post-VTCR, intent to seek professional care increased by 6.1% to 39.0% (p<0.05). The proportion of patients uncertain of what care to pursue, or who planned to do nothing, decreased by 20.0% from 34.7% pre-triage to 14.7% post-triage (p<0.05). The percentage of patients who planned to manage their symptoms with self-care measures, clinically contraindicated for STIs, increased by 13.8% post-triage to 46.3% from 32.5% pre-triage (p<0.05). This was also contrary to VTCR output, where only 0.7% of triage recommendations were to self-care. The findings illustrate the difficulty in favorably influencing patient STI care intent, with almost half still planning self-care post-triage contrary to VTCR output, and a substantial proportion still not planning to consult a healthcare professional.

Discussion

Symptoms consistent with STIs were observed at a rate of 2.6% in our study, which is in keeping with CDC data indicating that STIs like chlamydia and gonorrhea were diagnosed in only 0.7% of the US population in 2023. This low detection rate reflects the predominantly asymptomatic nature of many STIs. Despite estimates suggesting that roughly 20% of the population have an STI in any given year, the vast majority experience little to no noticeable symptoms. The demographic distribution of elevated STI prevalence among the patients in this study is also consistent with global epidemiological data, where STIs more commonly affect younger adults. Patients aged 18 to 44 years old accounted for 89.2% of those with STIs. Higher representation of females (64.5%) in our dataset also may have contributed to an overall STI rate of 2.6%, as there is evidence that in the US males have higher STI incidence and higher likelihood of a symptomatic course relative to asymptomatic presentations among females. The skew to higher general female use of VTCR may reflect greater female engagement of digital health tools and healthcare seeking among females, and aligns with other reports of public VTCR use. The ranking of different STI etiologies in this population aligns with statistical data on symptomatic STI presentations. In the US chlamydia, trichomoniasis, genital herpes and human papillomavirus account for 97.6% of all prevalent and 93.1% of all incident STIs. This is consistent with STI rates presented in Table 1, except for gonorrhea, which ranks higher. This discrepancy may be due to rarer conditions like gonorrhea presenting with symptoms more often than common asymptomatic infections like chlamydia, resulting in VTCR technology detecting them at higher rates. The reported changes in patient care intent present a complex picture in terms of improving care delivery to individuals with STIs. On the one hand, VTCR substantially (20.0%) reduced the number of patients uncertain of their care intent, or planning not to seek care, a positive impact. In these instances, VTCR successfully reduced patient indecision, inaction and care delay. There was also a statistically significant increase of 6.1% in the percentage of users intending to seek care following VTCR, observed most frequently in non-urgent outpatient consultations, which is often the most appropriate referral for many STIs. Thus, VTCR tools can transform passive STI symptom acknowledgment into proactive healthcare engagement and potential care seeking. On the other hand, a 13.8% increase in users opting for self-care measures post-VTCR is concerning since STIs are not treated with self-care measures. We suspect this includes patients planning to obtain telemedical care with at-home diagnostic assays and treatment. Further details on the specifics of self-care intent were not captured, and because “self-care” may be understood by some individuals as care involving at-home testing and virtual treatment, future studies should capture granular data on what is perceived as self-care intent. Overall, the results suggest VTCR can address certain barriers to STI care seeking by improving health awareness and education, demonstrated by the large reduction of patients post-VTCR who remain uncertain of how or whether to proceed.

Conclusions

Artificial intelligence-based VTCR substantially influenced care seeking among individuals reporting symptoms suggestive of STI. Prior to VTCR, over one-third of users were unsure how to address their symptoms or planned inaction. Following virtual triage, this proportion decreased significantly, while users intending to seek care increased. Virtual triage and automated care referral reduced indecision among patients by converting passive symptom acknowledgement into proactive intent to seek care. Virtual triage improved health awareness and can potentially expedite care access among patients reporting likely STIs, thereby achieving a positive public health impact. A concerning increase in users opting for self-care measures after VTCR indicated a possible STI, and despite a recommendation to seek care, suggests that certain barriers to STI care seeking such as stigma, financial concerns, trust or healthcare accessibility may be too deeply rooted to be addressed by VTCR technology alone. Nonetheless, the data demonstrated that VTCR can potentially support public health efforts to improve STI care seeking behavior.

Declarations

Clinical Trial: Not applicable.

Consent for Publication: All co-authors have reviewed and approved this article for publication.

Availability of Data and Material: Study data may be made available upon reasonable request.

Competing Interests: All authors are either employees of or medical advisors to Infermedica.

Funding: No external funding supported this work.

Authors contributions: GAG, MM, GLG, TP, KT and MP designed the study methodology, completed analyses and interpreted the data; GAG, MM and GLG wrote and edited all drafts of the manuscript; GAG, MM, GLG, TP, KT and MP reviewed and organized the data presentation and validated the data analyses; GAG, MM and GLG wrote the initial draft and edited subsequent drafts of the manuscript; GLG assisted with project management, literature search, reference integration, and journal submission.

References

1. World Health Organization, New report flags major increase in sexually transmitted infections, amidst challenges in HIV and hepatitis, May 21, 2024. https://www.who.int/news/item/21-05-2024-new-report-flags-major-increase-in-sexually-transmitted-infections—amidst-challenges-in-hiv-and-hepatitis. Accessed 23 March 2025.
2. Centers for Disease Control and Prevention, Sexually transmitted disease surveillance 2019: National overview – Sexually Transmitted Disease Surveillance, 2019. April 13, 2021. Accessed 23 March 2025. https://stacks.cdc.gov/view/cdc/105137. Accessed 23 March 2025.
3. Centers for Disease Control and Prevention, Reported STDs reach all-time high for 6th consecutive year. https://archive.cdc.gov/www_cdc_gov/media/releases/2021/p0413-stds.html#:~:text=The%20newly%20released%202019%20STD,STDs%20between%202015%20and%202019. Accessed 23 March 2025.
4. Centers for Disease Control and Prevention. Sexually transmitted infections (STIs): National overview of STIs in 2023, https://www.cdc.gov/sti-statistics/annual/summary.html. Accessed 23 March 2025.
5. de Wit JBF, Adam PCG, den Daas C, Jonas K. Sexually transmitted infection prevention behaviours: Health impact, prevalence, correlates, and interventions. Psychol Hlth 2023;38(6):675-700. DOI: 10.1080/08870446.2022.2090560
6. Dalby J, Stoner BP. Sexually transmitted infections: Updates from the 2021 CDC guidelines. Am Fam Phys 2022;105(5):514-520.
7. O’Byrne P, Orser L, Kroch A. Rates of sexually transmitted infections are rising. BMJ 2023:30;381:1492. DOI: 10.1136/bmj.p1492
8. Garcia PJ, Miranda AE, Gupta S, Garland SM, Escobar ME, Fortenberry JD. The role of sexually transmitted infections (STI) prevention and control programs in reducing gender, sexual and STI-related stigma. Eclin Med 2021 Feb 24;33:100764. DOI: 10.1016/j.eclinm.2021.100764
9. Lee ASD, Cody SL. The stigma of sexually transmitted infections. Nurs Clin North Amer 2020;55(3):295-305. DOI: 10.1016/j.cnur.2020.05.002
10. King AJ, Bilardi JE, Towns JM, Maddaford K, Fairley CK, Chow EPF, Phillips TR. User views on an online sexual health symptom checker tool: Qualitative research. JMIR Form Res 2024;4;8:e54565. DOI: 10.2196/54565
11. Karamouzian N, Knight R, Davis WM, Gilbert M, Shoveller J. Stigma associated with sexually transmissible infection testing in an online testing environment: Examining the perspectives of youth in Vancouver, Canada. Sex Hlth 2027;15:46-53. https://doi.org/10.1071/SH17089.
12. Low N, Broutet N, Adu-Sarkodie Y, Barton P, Hossain M, Hawkes S. Global control of sexually transmitted infections. Lancet 2006;2;368(9551):2001-16. DOI: 10.1016/S0140-6736(06)69482-8
13. Malek AM, Chang CC, Clark DB, Cook RL. Delay in seeking care for sexually transmitted diseases in young men and women attending a public STD clinic. Open AIDS J 2013;14;7:7-13. DOI: 10.2174/1874613620130614002
14. Tilson EC, Sanchez V, Ford CL, et al. Barriers to asymptomatic screening and other STD services for adolescents and young adults: Focus group discussions. BMC Publ Hlth 2004;4:21. DOI: 10.1186/1471-2458-4-21
15. Leichliter JS, Copen C, Dittus PJ. Confidentiality issues and use of sexually transmitted disease services among sexually experienced persons aged 15-25 Years – United States, 2013-2015. MMWR 2017;66(9):237-241. DOI: 10.15585/mmwr.mm6609a1
16. Kenyon C, Herrmann B, Hughes G, de Vries HJC. Management of asymptomatic sexually transmitted infections in Europe: Towards a differentiated, evidence-based approach. Lancet Reg Hlth Eur 2023;34:100743. DOI: 10.1016/j.lanepe.2023.100743
17. Centers for Disease Control and Prevention. Sexually Transmitted infections prevalence, incidence, and cost estimates in the United States, https://www.cdc.gov/sti/php/communicationresources/prevalence-incidence-and-cost-estimates.html#:~:text=CDC%20estimates%20indicate%20about%2020,billion%20in%20health%20care%20costs%20alone. Accessed 23 March 2025.
18. Korenromp EL, Sudaryo MK, de Vlas SJ, Gray RH, Sewankambo NK, Serwadda D, Wawer MJ, Habbema JD. What proportion of episodes of gonorrhoea and chlamydia becomes symptomatic? Int J STD AIDS 2002;13(2):91-101. DOI: 10.1258/0956462021924712
19. Clifton S, Mercer CH, Sonnenberg P, Tanton C, Field N, Gravningen K, Hughes G, Mapp F, Johnson AM. STI Risk Perception in the British population and how it relates to sexual behaviour and STI healthcare use: Findings from a cross-sectional survey (Natsal-3). Eclin Med 2018;2-3:29-36 DOI: 10.1016/j.eclinm.2018.08.001
20. Gellert GA Garber L, Kabat-Karabon A et al. Using AI-based virtual triage to improve acuity-level alignment of patient care seeking in an ambulatory care setting. Intl J Hlthcr 2024;10. 41.10.5430/ ijh.v10n1p41. DOI: https://doi.org/10.5430/ijh.v10n1p41
21. Gellert GA, Kabat-Karabon A, Gellert GL et al. The potential of virtual triage AI to improve early detection, care acuity alignment, and emergent care referral of life-threatening conditions. Front Publ Hlth 2024;12:1362246. DOI: 10.3389/fpubh.2024.1362246
22. World Health Organization, Sexually transmitted infections (STIs), May 21, 2024. https://www.who.int/news-room/fact-sheets/detail/sexually-transmitted-infections-(stis).
23. Kreisel KM, Spicknall IH, Gargano JW et al. Sexually transmitted infections among US women and men: Prevalence and incidence estimates, 2018. STDs 2021;48(4):208-214. DOI: 10.1097/OLQ.0000000000001355
24. Gellert GA, Orzechowski PM, Price T et al. A multinational survey of patient utilization of and value conveyed through virtual symptom triage and healthcare referral. Front Publ Hlth 2023. DOI: https://doi.org/10.3389/fpubh.2022.1047291