JMIR Publications

ABSTRACT

Background:

HIV and sexually transmitted infections (STI) are major global public health concerns. Over one million curable STIs occur every day amongst people aged 15–49 years worldwide. Insufficient testing or screening substantially impedes the elimination of HIV/STI transmission.

Objective:

The aim of our study is to develop an HIV/STI risk prediction tool using machine learning algorithms.

Methods:

We used clinic consultations tested for HIV/STI at the Melbourne Sexual Health Centre between March 2, 2015, to December 31, 2018, as the development dataset (training and testing dataset). We also used two external validation datasets including data in 2019 as the external 'validation data 1' and data during 01/2020 and 01/2021 as the external 'validation data 2'. We developed 34 machine learning models to assess the risk of acquiring HIV, syphilis, gonorrhoea, and chlamydia. We created an online tool to generate an individual's risk of HIV/STI.

Results:

The important predictors for HIV/STIs risk were gender, age, men who reported having sex with men, number of casual sexual partners, and condom use. Our ML-based risk prediction tool named MySTIRisk performed at an acceptable or excellent level on testing datasets (area under the curve (AUC) for HIV= 0.78; syphilis = 0.84; gonorrhoea =0.78; chlamydia =0.70) which had stable performance on both external validation data in 2019 (AUC for HIV= 0.79; syphilis = 0.85; gonorrhoea = 0.81; chlamydia = 0.69), and data in 2020-2021(AUC for HIV= 0.71; syphilis= 0.84; gonorrhoea =0.79; chlamydia =0.69).

Conclusions:

Our web-based risk prediction tool could accurately predict the risk of HIV/STI in clinic attendees with simple self-reported questions. MySTIRisk could serve as an HIV/STI screening tool in clinic websites or digital health platforms to encourage individuals at risk of HIV/STI to have testing or start HIV pre-exposure prophylaxis. The public can use this tool to assess risk and then decide if they would attend a clinic for testing. Clinicians or public health workers can use this tool to identify high-risk individuals for risk management or further interventions.