Ask AI
Back Back
HIV ASSIST Guide

CME

How to Use the HIV-ASSIST Tool: A Step-by-Step Guide for HCPs in Asia

Physicians: maximum of 1.00 AMA PRA Category 1 Credit

Released: June 11, 2025

Expiration: June 10, 2026

Maunank Shah
Maunank Shah, MD, PhD
CME/CE Information

Activity

Progress
1 2
Course Completed

HIV-ASSIST: A Decision Support Tool for ARV Selection

I would like to begin by giving an overview of how to access the HIV-ASSIST tool. HIV-ASSIST is available at www.HIVASSIST.com.

HIV-ASSIST is free to use and publicly available. It is also available as a smartphone application, for both Android and iPhone.1

Download

HIV-ASSIST: Input Key Viral and Patient Characteristics With a User-Friendly Web-Based Tool

After going to the website and clicking on “start,” you will be taken to an input screen, which is a easy-to-use, web-based interface where you can enter key patient characteristics that the system will use to evaluate optimal antiretroviral (ARV) choices.1 

Download

HIV-ASSIST: A Decision Support Tool for ARV Selection Available as a Mobile App

Alternatively, if you were to use the smartphone application, you would open the app and see a screen like what is shown here.1

Tapping on “start” will lead you through a series of screens into which you enter patient characteristics.

Download

HIV-ASSIST: Input Is Individualized for Each Patient’s Viral Attributes

Next, I will share some insights into how HIV-ASSIST works to give you a better understanding of how to interpret the results.

HIV-ASSIST is based on a branch of decision theory called multiple-criteria decision analysis. Here, you see a schematic of how the algorithms work to prioritize ARV selections for your individual patient scenarios.1

Sometimes, people ask if HIV-ASSIST is using artificial intelligence or machine learning, and the answer is no. HIV-ASSIST is based on a series of rules and logic-based algorithms that seek to apply the available evidence and clinical guidelines in order to make appropriate ARV selections for individual case scenarios.1

It does so in the following manner: First, the system considers the different viral attributes that you entered. This includes things like HIV-1 RNA and resistance mutations.1

Download

HIV-ASSIST: Input Is Individualized for Each Patient’s Comorbidities, Comedications, and Characteristics

Next, the system considers the various patient characteristics that you entered, such as comorbidities, comedications, treatment history, and CD4 count.1

Download

HIV-ASSIST: Provides Individual ARV Utility Scores

The system then evaluates all of those different factors and prioritizes or penalizes individual ARVs through an internal multiple-attribute utility function to determine the optimal regimen.1

To say this in a simpler way, for example, if there is a drug resistance mutation, an ARV that is compromised by that mutation would get a mathematical penalty, causing that ARV to be deprioritized.1

Similarly, the system automatically links to the drug interaction checker from the University of Liverpool. If you entered a comedication that has a significant drug interaction with a certain ARV, that specific ARV will be penalized and therefore deprioritized in the HIV-ASSIST algorithms.1

Download

HIV-ASSIST: Multidrug Aggregate Score Informed by Guidelines and Scientific Advisory Panel

After synthesizing those particular characteristics to determine utility scores for individual ARVs, the HIV-ASSIST algorithms evaluate each multidrug regimen on the basis of 2 primary factors:

First, the tool evaluates the available scientific evidence that a particular ARV regimen will be likely to lead to sustained viral suppression.1

Second, the algorithms consider the domain of tolerability for each regimen and factors in other patient preferences. This includes aspects like pill burden and the expected adverse effect profile.1

Download

HIV-ASSIST: Includes User Options for Additional Prioritization

Additional prioritization options include aversion to large pills or needles, preference for once-daily dosing, adherence challenges, and number of active drugs.1

HIV-ASSIST is unique in allowing users to enter all of these various considerations to fully individualize ART for each patient.1

Download

HIV-ASSIST: Score Is the Composite of Different Considerations, Not an Absolute Ranking

Finally, the system considers each factor simultaneously to generate a final HIV-ASSIST score for each regimen evaluated, which synthesizes the domains of efficacy, tolerability, and patient preference.1

Lower scores, color-coded in green, signify strong evidence that the regimen is likely to be virologically efficacious and tolerable for sustained usage.1

Intermediate scores, color-coded as yellow or orange, represent some form of trade-off. For example, there may be instances where a particular regimen possesses strong evidence for efficacy, but requires multiple pills multiple times a day. This regimen therefore would get penalized on the basis of tolerability and would show as a lower score on the ranked list.1

High scores, color-coded as red, indicate that these regimens are associated with poor efficacy, poor tolerability, or limited evidence for use and would not be recommended for the patient in question.1

Download

HIV-ASSIST Output: Potential Regimens With Color-Coding and Weighted Scores

The HIV-ASSIST system is completely thorough in that it actually evaluates every possible combination of 2, 3, 4, 5, and 6 drugs from the available ARVs and ranks them based on the composite internal HIV-ASSIST weighted score that I previously described.1

Shown here is what a common output might look like. At the top of the screen, the user inputs are summarized, and the ranked list of ART regimens appears below.1

The different ARVs are listed based on the HIV-ASSIST weighted score in the second column. On this dashboard, you can also see the number of active drugs in the regimen, based on the number of included ARVs and any drug resistance.1

This table-based view will also show a snapshot of the total pill burden, as well as the dosing frequency.1

There is also quite a lot more information available to view when you click on each of these different regimens, which will be demonstrated through a case example.1

Download

HIV-ASSIST Output: Expand Instructions Link to Access Explanation of Weighted Scoring and Color Coding

If you ever forget what the color coding means, you can click on the blue bar that says, “Instructions.”1

Download

HIV-ASSIST Output: Lowest Scores Represent Preferred Regimens Backed by Strong Clinical Evidence

When you do so, the system will expand to show you an explanation of the HIV-ASSIST scoring system.1

To recap, lower scores ≤1.5 are represented in dark green, indicating strong evidence for both likelihood of viral suppression and tolerability. Regimens with good likelihood of achieving viral suppression, but some disadvantages for tolerability, such as higher pill burdens or more adverse effects, will have a higher HIV-ASSIST score and may show up as light green or yellow. This does not mean that those regimens are inherently worse, but rather shows that they are ones for which there are some trade-offs in one of these domains, for the specific patient being considered.1

For ease of use, if there is evidence against use of a regimen, it will not appear on the list of individualized ART options.1

Download

HIV-ASSIST Outputs Rational Choices: HCP Weighs the Pros and Cons

Overall, HIV-ASSIST develops a single independent composite score for each ARV option, synthesizing various different domains including drug resistance, drug interactions, and comorbidities into a single weighted score by which regimens are ranked.1

 

Lower scores are better and are more likely to achieve viral suppression and be well-tolerated, whereas higher scores represent regimens with trade-offs. Trade-offs can include less evidence to support usage and/or lower tolerability.1

Clicking on a particular regimen pulls up additional rationale that can be used to adjudicate the prioritizations and penalties applied by the HIV-ASSIST system.1

Ultimately, HIV-ASSIST seeks to be a decision aid and an education tool to support, but not replace, clinical decision-making. The hope is that it will provide detailed rationale for decision-making in a transparent manner in a simple, easy-to-use tool.1

Download
BACK | NEXT