Predictive Analytics + Clinical Decision Support for HIV Prevention

(PrEDICT Trial)

Recruiting at 15 trial locations

Overseen By Julia Marcus, PhD

Age: 18+

Sex: Any

Trial Phase: Academic

Sponsor: Harvard Pilgrim Health Care

Disqualifiers: HIV/STI care, Correctional, Mobile, School-based

No Placebo Group All trial participants will receive the active study treatment (no placebo)

Compensation: Varies

Number of Visits: Unknown

16 Participants Needed

Trial Summary

Will I have to stop taking my current medications?

The trial information does not specify whether you need to stop taking your current medications.

What data supports the effectiveness of the treatment Predictive Analytics and Clinical Decision Support for HIV Prevention?

Research shows that predictive analytics can help identify HIV patients at risk of dropping out of care, allowing for timely interventions. A machine learning model improved prediction accuracy, flagging high-risk patients more effectively than previous methods, which can help keep patients in care and reduce HIV transmission. ^{1 2 3 4 5}

Is Predictive Analytics + Clinical Decision Support generally safe for humans?

The research articles focus on improving the identification and prediction of adverse events (unwanted side effects) in medical care, but they do not provide specific safety data for Predictive Analytics + Clinical Decision Support as a treatment. They highlight the importance of accurate adverse event detection to enhance patient safety. ^{6 7 8 9 10}

How does the Predictive Analytics and Clinical Decision Support treatment for HIV prevention differ from other treatments?

This treatment is unique because it uses machine learning and predictive analytics to identify individuals at high risk of HIV infection, allowing for targeted prevention strategies like pre-exposure prophylaxis (PrEP). Unlike traditional methods, it leverages electronic health records and real-time data to make informed decisions, potentially improving early intervention and reducing HIV transmission. ^{11 12 13 14 15}

What is the purpose of this trial?

Scale-up of HIV preexposure prophylaxis (PrEP) is a key strategy of the U.S. initiative to end the HIV epidemic, but healthcare providers lack tools to support PrEP discussions and prescribing for patients likely to benefit. This research will evaluate whether integrating automated tools into electronic health records to help providers efficiently and equitably identify potential candidates for PrEP, discuss PrEP, and prescribe PrEP can improve PrEP initiation and persistence in safety-net community health centers. It will achieve this by conducting a stepped-wedge trial of a decision support tool with an embedded HIV prediction model to identify patients likely to benefit from PrEP. The intervention will be delivered to healthcare providers in 16 community health centers within the national OCHIN network.

Eligibility Criteria

This trial is for healthcare providers at community health centers that saw over 500 patients in 2024, had at least 10 new HIV diagnoses in 2023, and offer primary care services (excluding pediatrics) active on OCHIN Epic since January 1, 2023.

Inclusion Criteria

My healthcare facility reported at least 10 new HIV cases in 2023.

My primary care provider is active on OCHIN Epic since 1/1/2023 and does not specialize in pediatrics.

My healthcare provider had over 500 patients in 2024.

Timeline

Screening

Participants are screened for eligibility to participate in the trial

2-4 weeks

Intervention

Providers receive point-of-care notifications via the EHR-embedded decision support tool about patients at increased predicted HIV risk

30 months

Follow-up

Participants are monitored for PrEP initiation and persistence

30 months

Treatment Details

Interventions

• Predictive Analytics and Clinical Decision Support

Trial Overview The study tests a decision support tool integrated into electronic health records. It uses predictive analytics to help providers identify who might benefit from HIV PrEP, discuss it with them, and prescribe it more effectively and equitably.

Participant Groups

Group I: Standard of Care Active Control 1 Intervention

Data will be extracted from EHRs of control clinics as a comparison, and those clinics will not actively participate in study activities

Group II: Predictive analytics and clinical decision support Active Control 1 Intervention

Providers in enrolled intervention clinics will receive point-of-care notifications via the EHR-embedded decision support tool about patients who are at increased predicted HIV risk and therefore likely to benefit from PrEP

Who Is Running the Clinical Trial?

Harvard Pilgrim Health Care

Lead Sponsor

Trials

61

Recruited

27,990,000+

Oregon Health and Science University

Collaborator

Trials

1,024

Recruited

7,420,000+

National Institute of Mental Health (NIMH)

Collaborator

Trials

3,007

Recruited

2,852,000+

Oregon Health & Science University (OHSU)

Collaborator

Trials

1

Recruited

20+

OCHIN, Inc.

Collaborator

Trials

24

Recruited

9,964,000+

Findings from Research

A predictive model for preventable adverse events (AEs) in hospitalized older patients was developed using data from 6096 patients across two studies, but it was found to be unsatisfactory in accurately predicting these events.

Key risk factors identified included increased age, elective admissions, and admissions to surgical departments, but common factors like comorbidity did not effectively predict preventable AEs.

NCBI (Pubmed)

pubmed.ncbi.nlm.nih.gov

Can preventable adverse events be predicted among hospitalized older patients? The development and validation of a predictive model. Van De Steeg, L., Langelaan, M., Wagner, C.[2022]

In a study of 1047 patients at a large teaching hospital, 17.7% experienced serious adverse events, with longer hospital stays increasing the likelihood of such events by about 6% for each additional day.

The majority of adverse events were linked to individual errors (37.8%) or interactive causes (15.6%), highlighting the need for healthcare providers to focus on these areas for proactive error prevention.

NCBI (Pubmed)

pubmed.ncbi.nlm.nih.gov

An alternative strategy for studying adverse events in medical care. Andrews, LB., Stocking, C., Krizek, T., et al.[2022]

References

1. United States pubmed.ncbi.nlm.nih.gov

Multicenter Development and Validation of a Model for Predicting Retention in Care Among People with HIV. [2023]

2. United States pubmed.ncbi.nlm.nih.gov

Efficacy of a clinical decision-support system in an HIV practice: a randomized trial. [2021]

3. England pubmed.ncbi.nlm.nih.gov

Predictive Analytics for Retention in Care in an Urban HIV Clinic. [2021]

4. United States pubmed.ncbi.nlm.nih.gov

Clinical decision tools are needed to identify HIV-positive patients at high risk for poor outcomes after initiation of antiretroviral therapy. [2018]

5. China pubmed.ncbi.nlm.nih.gov

Developing prediction models for clinical use using logistic regression: an overview. [2021]

6. England pubmed.ncbi.nlm.nih.gov

Can preventable adverse events be predicted among hospitalized older patients? The development and validation of a predictive model. [2022]

7. England pubmed.ncbi.nlm.nih.gov

Improving patient safety via automated laboratory-based adverse event grading. [2021]

8. England pubmed.ncbi.nlm.nih.gov

An alternative strategy for studying adverse events in medical care. [2022]

9. United States pubmed.ncbi.nlm.nih.gov

Novel Approaches for Dynamic Visualization of Adverse Event Data in Oncology Clinical Trials: A Case Study Using Immunotherapy Trial S1400-I (SWOG). [2023]

10. England pubmed.ncbi.nlm.nih.gov

Leveraging electronic health records for predictive modeling of post-surgical complications. [2019]

11. Netherlands pubmed.ncbi.nlm.nih.gov

How to Identify Potential Candidates for HIV Pre-Exposure Prophylaxis: An AI Algorithm Reusing Real-World Hospital Data. [2021]

12. England pubmed.ncbi.nlm.nih.gov

Generalizable pipeline for constructing HIV risk prediction models across electronic health record systems. [2023]

13. United States pubmed.ncbi.nlm.nih.gov

Using Smartphone Survey Data and Machine Learning to Identify Situational and Contextual Risk Factors for HIV Risk Behavior Among Men Who Have Sex with Men Who Are Not on PrEP. [2023]

14. Switzerland pubmed.ncbi.nlm.nih.gov

Machine Learning-Based HIV Risk Estimation Using Incidence Rate Ratios. [2022]

15. England pubmed.ncbi.nlm.nih.gov

Using machine learning approaches to predict timely clinic attendance and the uptake of HIV/STI testing post clinic reminder messages. [2022]