Three Drugs, Two Pills, One Algorithm: A Cross-Sectional Pilot Study of Consolidated Guidelines for HIV PrEP, PEP, and Treatment Initiation for Resource-Limited and Non-specialty Settings in the United States

Abstract

Background

Accessing HIV prevention and treatment remains difficult throughout the United States (US). With the shortage of infectious disease physicians, increasing the number of clinicians offering HIV Pre-Exposure Prophylaxis (PrEP), HIV Post-Exposure Prophylaxis (PEP), and Test & Treat (T&T) is imperative. This exploratory project created a consolidated algorithm for HIV treatment and prevention that simplifies care for US providers with limited resources or experiences.

Methods

The algorithm was designed for PrEP, PEP, and treatment initiation using published guidelines from multiple clinical bodies. Three medications were included in the algorithm: tenofovir disoproxil fumarate, emtricitabine, and dolutegravir. US medical students completed an eight-question pretest unassisted then repeated questions with the algorithm to test the algorithm’s functionality. Questions included clinical scenarios on HIV testing, PrEP, PEP, and T&T. Pearson’s chi-squared compared pre vs post-test scores; total scores (number of items correct) were analyzed using Paired Wilcoxon signed rank test.

Results

97 students from 13 medical schools completed pre and post-test questions. Correctness increased from 3 of 8 (pretest) to 7 of 8 (post-test) (p< 0.001). All questions saw statistically significant improvement in correctness when assisted by the algorithm including correctly prescribing PEP, selecting appropriate labs for PrEP initiation, and screening patients with prior cabotegravir-based PrEP using HIV RNA quantitative (p<0.001 for all).

Conclusions

This algorithm improved medical students’ abilities to accurately answer HIV clinical care questions about HIV prevention and treatment. Further studies are needed to understand the algorithm’s effect in clinical settings and how to implement it in practice.

Introduction

The United States (US) has never had more medications at a clinician’s disposal to prevent or treat HIV. In the past two decades, the approval of multiple medications within new drug classes has reduced pill burden and side effects while increasing treatment options. These medicines have increased the number of people with HIV (PWH) achieving undetectable viral loads and prevent PWH from transmitting the virus to other people. ¹ While there is no vaccine for HIV, pre-exposure prophylaxis (PrEP) regimens are medications taken by people without HIV to prevent HIV. There are four FDA approved options for PrEP currently on the market, two oral and two injectable. ² Post-Exposure Prophylaxis (PEP) regimens can be prescribed to prevent HIV in those with recent exposures by preventing seroconversion. ³ With these technologies, once daily medications can treat HIV in a pregnant person to prevent vertical transmission, prevent HIV transmission in serodiscordant couples, and consolidate therapy among PWH. Despite these advancements, new HIV infections in the US decreased by only 33% between 2012 and 2022, far from the National Ending the HIV Epidemic goal of reducing new HIV diagnoses by 90% by 2030.^4-6

There are increasing headwinds facing those who aim to make HIV care the mainstay of their career. Congressional considerations to defund federal programs like 340B funding (which provides drugs at lower costs), and the Ryan White Act (which provides medical and dental service coverage for PWH) pose challenges. ⁷ Many states have also been passing legislation threatening HIV care including refusal of Centers for Disease Control and Prevention (CDC) funding for HIV prevention organizations and had court challenges to guaranteed PrEP care coverage by insurers.^8,9 These laws and proposals could significantly alter the structure, funding, and viability of achieving the national HIV reduction goals.

Compounding this problem is a struggling pipeline for new HIV providers. Infectious disease (ID) fellowships have seen a decreasing match rate in recent years. ¹⁰ Although the factors affecting fellowship match rates are outside the scope of this paper, greater than 80% of US counties lack an ID physician. ¹¹ Having an ID provider does not ensure access to HIV care, however. A survey of ID physicians found that only 67% provided care for PWH. ¹² Although HIV prevention and treatment care can be provided by primary care providers (PCPs), evidence has shown that exceedingly few general practitioners have ever written for antiretroviral drugs for PrEP, PEP, or treatment. ¹³ This has led to the “purview paradox” where ID providers believe PrEP should be prescribed by PCPs, while PCPs believe PrEP should be provided by ID specialists. ¹⁴ Many states have passed laws allowing pharmacists to write for PrEP and PEP to expand access to care. ¹⁵ It is imperative that providers in general fields have easy access to clear guidance on PrEP, PEP, HIV treatment, and linkage to specialists. To the best of the authors’ knowledge, no single algorithm exists that includes HIV testing, PrEP, PEP, and treatment initiation for the US healthcare system. We developed an algorithm to provide these services to patients cared for in non-specialty and resource-limited settings in the US. The algorithm was designed to reduce erroneous laboratory testing and offer consolidated medications. To test this algorithm, we chose to conduct an exploratory pilot with medical students. We hypothesize that this algorithm would improve medical students’ ability to correctly select care plans for HIV-related clinical scenarios.

Methods

Algorithm Development

The foundation of the algorithm (Figure 1) was made by referencing the New York State Department of Health AIDS Institute, the International Antiretroviral Society, the World Health Organization, and the Department of Health, Human Services Office of AIDS Research, and consulting with a practicing physician and medical director of a federally funded HIV clinic. The algorithm can be used in the care of all patients needing HIV testing who have not previously tested positive for HIV and weigh at least 35 kg. Patients with signs of AIDS/advanced HIV or opportunistic infections are immediately deferred for an ID referral.^16,17 These patients also would likely benefit from rapid initiation of antiretroviral therapy (ART) which should be done under the guidance of an ID physician given the heightened risk for immune reconstitution inflammatory syndrome.^17,18

Figure 1.

The novel algorithm developed containing care plans for pre-exposure prophylaxis, post-exposure prophylaxis, and deferral guidance. Ab, antibody; AIDS, acquired immunodeficiency syndrome; Ag, antigen; CMP, comprehensive metabolic panel; CrCl, creatinine clearance; DTG, dolutegravir; FTC, emtricitabine; HBsAg, hepatitis B surface antigen; HBsab, hepatitis B surface antibody; HCV, hepatitis C virus; HIV, human immunodeficiency virus; PEP, post-exposure prophylaxis; PO, by mouth (per os); PrEP, pre-exposure prophylaxis; PREP, pre-exposure prophylaxis; QD, once daily (quaque die); RNA, ribonucleic acid; RPR, rapid plasma reagin; STI, sexually transmitted infection; TDF, tenofovir disoproxil fumarate; VDRL, Venereal Disease Research Laboratory (test)

HIV Testing

First, we must determine the last possible HIV exposure (via sexual contact or intravenous drug usage (IVDU)) for the patient. If the last exposure has been within 72 hours, these patients are candidates for PEP barring ongoing medical issues such as liver or renal disease that may lead to acute decompensation if started on full ART regimen.^17,18 Patients are to begin a PEP regimen and be referred to an ID specialist for guidance on retesting and de-escalation of therapy. If a patient’s last exposure was over 72 hours ago, additional factors must be stratified to identify the appropriate HIV test. While HIV RNA quantitative testing narrows an individual’s window period, the cost of the test and lack of in-house availability for many resource-limited settings may make it cost prohibitive or irrelevant due to delayed result times. ¹⁹ Key scenarios where a HIV RNA quantitative test is required as part of their initial evaluation include individuals with symptoms of acute HIV infection or with injectable cabotegravir PrEP use in the past 12 months. ²⁰ For these patients, HIV antibody/antigen (HIV Ab/Ag) testing may be negative due to possible seroconversion in the window period or the lagging effects of cabotegravir in the system, respectively. ²⁰ Cabotegravir as PrEP accounts for less than three percent of PrEP prescriptions in the US as of June 2025, but providers should still be prepared to encounter patients with recent cabotegravir use. ²¹ The same guidance applies to lenacapavir for PrEP, but lenacapavir was approved for PrEP after this study had concluded. Patients without symptoms of acute HIV infection or use of injectable PrEP in the past 12 months are adequately screened using HIV antibody/antigen testing. ¹⁹ Patients with reactive HIV RNA or HIV Ab/Ag are candidates for rapid ART initiation. ²² Patients undergoing HIV testing should be tested for site-specific sexually transmitted infections, syphilis, Hepatitis C (HCV) RNA and Hepatitis B (HBV) (as ART regimens can partially treat HBV) and be referred to an ID provider if assistance is needed treating these conditions. ¹⁷ A patient with negative serologies for HIV, HBV, and HCV should discuss future risk and consider PrEP with their clinician.

Following the aforementioned screening and treatment protocol, medications were then selected to implement. The medications offered by the algorithm were selected to reduce provider confusion and patient costs to implement this protocol while also broadly covering most patient demographics and conditions. Ultimately, this led researchers to select three medications that are formulated in two tablets: 1) Tenofovir Disoproxil Fumarate (TDF)/Emtricitabine (FTC) and 2) Dolutegravir (DTG).

PrEP

There are several reasons TDF/FTC was chosen for PrEP in this algorithm. Tenofovir alafenamide (TAF)/emtricitabine (FTC) is only approved for those assigned male at birth (AMAB) and at-risk of HIV through sexual exposure. ²³ This excludes use of TAF/FTC for PrEP in those assigned female at birth (AFAB) and those with IVDU. Given this, a similar algorithm using TAF/FTC for PrEP would still require TDF/FTC on formulary to provide PrEP to these demographics. Similarly, TDF/FTC will provide a preferred backbone to PEP and treatment initiation as will be discussed in their respective sections. While TDF regimens used in the algorithm restrict treatment initiation to those with creatinine clearance of 60mL/min, patients with stage 3a kidney disease or worse renal impairment benefit from specialist management of HIV prevention given the multifaceted considerations in these patients. ²⁴ Lastly, TDF/FTC being generic greatly reduces its cost. A patient’s ability to afford medication, as well as costs for health entities to stock the medication, must be considered when designing infrastructure for new PrEP programs. ²⁵ Medicaid data shows a national average drug acquisition cost (NADAC) of TDF/FTC 300/200mg tablets as low as $15.69 for thirty tablets while an equivalent supply of TAF/FTC 25mg/200mg has a NADAC of $2,137. ²⁶ TDF/FTC ensures a safe and affordable PrEP regimen regardless of patient demographics, insurance coverage, or health-entity 340B program.

HIV Treatment

Patients who test positive for HIV and need rapid ART initiation require several considerations. In rapid treatment initiation, many HIV laboratory tests related regimen selection may not yet be available (e.g. genotyping, viral load, CD4 count). However, delaying treatment to perform these labs has a measurable worsening of patient outcomes. ²⁷ For rapid initiation in the absence of these labs, therapy with two nucleotide reverse transcript inhibitors (NRTIs) and an integrase inhibitor (INSTI) is preferred. ²⁸ Similar considerations have been applied in Colombia’s International Center for AIDS Care and Treatment Program’s recommendation of Lamivudine/TDF/DTG to expand INSTI use in Africa via the President’s Emergency Plan for AIDS Relief formulary. ²⁹ This regimen, however, is not available on the US market as a single tablet. Of these NRTI’s, TDF/FTC offers low side effect profile and HBV coverage. ¹⁸ DTG is a once daily INSTI that provides a high resistance barrier and reduced metabolic side effects compared to other ART drug classes. ²⁸ This high barrier to resistance regimen is approved as initial treatment in patients with unknown viral loads. ²⁰ DTG is also approved in those who are pregnant or may become pregnant. ³⁰ Should a health entity implement the algorithm, generic TDF/FTC would already be available and DTG would offer a preferred treatment initiation regimen at roughly half the NADAC of having Bictegravir (BIC)/TAF/FTC) on formulary.^22,26

While this regimen provides non-inferior coverage in these clinical scenarios, it may not be the first choice of those with wide drug selections available. ²² BIC/TAF/FTC is often selected as many clinicians’ first choice for treatment naïve PWH as it is a single tablet. While DTG was once the preferred daily INSTI for PWH who are or may become pregnant, BIC was recently approved for use in pregnancy. ³¹ BIC is not available in a single drug tablet and is only on the US market combined with TAF/FTC. TAF/FTC is often widely used as the preferred oral regimen for PrEP in individuals assigned male at birth or backbone for ART given the renal protective profile compared to TDF/FTC. A recent meta-analysis found no significant differences in adverse events related to bone mineral density or renal tubular injury when comparing TAF to TDF. ³² Other studies have shown that TAF had fewer renal adverse events and favorable renal biomarkers at 96 weeks of therapy when compared to TDF. ³³ The role of the algorithm, however, is to rapidly engage patients who test positive in effective treatment which can later be tailored by a specialist. Some clinicians may be concerned about pill fatigue when using DTG+TDF/FTC in the age of single tablet regimens. While this regimen is two tablets, both medications can be taken at the same time, once daily. ²² Much of the existing data on HIV adherence improvements with pill burden has been done comparing twice daily to once daily regimens. ³⁴

PEP

DTG +TDF/FTC is a recommended regimen for PEP in all individuals including those who are pregnant. ³⁰ Discrepancies exist between whether TAF or TDF based PEP regimens should be recommended. International Antiviral Society (IAS) guidelines recommend either TAF or TDF. CDC guidance previously recommended only TDF-based PEP regimens, but has recently approved TAF for use in PEP.^18,35 The existing debate centers around TDF’s faster time to elevated drug levels in the body compared to TAF. ³³ While it was believed that this bioavailability difference would provide superior protection in TDF regimens, there is little data to support this. Rather, more data exists showing that rapid initiation of a PEP regimen is more important than the NRTI backbone. ³⁶

The complete algorithm is displayed in Figure 1.

Algorithm Testing

To initially test the algorithm’s efficacy in helping clinicians in under-resourced settings, the decision was made to do preliminary research with medical students. While medical students are not practicing independently, we chose medical students because they provided a controlled and accessible population with foundational clinical knowledge, allowing for early assessment of algorithm usability, comprehension, and decision-making without the confounding effects of prior clinical habits. Furthermore, using a convenience sample allowed detection of any shortcomings or misinterpretation of the algorithm in a setting where it would not negatively influence those providing care to patients. Researchers plan to proceed with future studies using the algorithm with clinicians pending the data of this initial study. To evaluate the algorithm’s feasibility and function with medical students, an eight-question test was developed by the research team. The test contained board style questions including a patient scenario related to HIV testing, prevention and treatment initiation (Figure 2). Topics included:

Figure 2.

A list of the topics assessed in pre and post-testing. Ab/Ag, antibody/antigen; AIDS, acquired immunodeficiency syndrome; HIV, human immunodeficiency virus; PEP, post-exposure prophylaxis; PrEP, pre-exposure prophylaxis

Students completed the unrandomized test twice, first unassisted, then using the algorithm as a reference. External resources were restricted for pre and post-test. Between the pre and post-test, a slide appeared showing the algorithm for review prior to post-test. The algorithm was attached for reference with each post-test question. Students were emailed a link to the Qualtrics survey containing a written consent form and the test to complete virtually. Students were given an estimated time of completion of twenty-five minutes, but there was no time limit for pre or post-test. Survey data collection occurred between November of 2023 and June of 2024. Study inclusion criteria were current enrollment at a US allopathic or osteopathic college of medicine; completion of their institutions’ HIV coursework which ensured participants had a frame of reference for the information. Exclusion criteria were participants who had not completed their institutional HIV coursework and participants from other health sciences fields, such as nurse practitioners, pharmacists and physician assistants to avoid inherent bias that may exist in the HIV training in these fields. This study and its reporting adhere to the standards for Intervention Description and Replication (see supplemental materials for checklist). ³⁷ We hypothesized that the use of the algorithm would significantly improve students’ abilities to accurately answer questions about PrEP, PEP, and treatment initiation. Independent variable examined was the algorithm as an intervention with the dependent variable being the change in post-test correctness compared to pretest.

This study was approved by the Institutional Review Board at the Alabama College of Osteopathic Medicine.

Statistical Analysis

A total of 210 medical students at 13 medical schools were contacted about the survey. 153 students participated with ninety-seven(N=97) of those completing both pre- and post-test questions (46% complete response rate). Descriptive statistics were obtained. Categorical and continuous variables were provided with counts and percentages and medians with interquartile ranges, respectively. Further statistical analysis was performed to detect the changes in correct responses during the post-test when provided the algorithm. Pre vs. Post comparison of individual question items utilized Pearson’s chi-squared test while the raw total score (number of items correct) used the Paired Wilcoxon signed-rank test. All statistical analyses were performed in R Statistical Software (v4.2.1; R Core Team 2022).

Results

Analysis of pre versus post comparison of individual questions and total score of participants is provided in Figure 3. On average, students correctly answered three out of eight questions on the pretest and seven of eight questions correctly on the post-test with algorithm assistance (p<0.001). All individual questions saw statistically significant increases in correctness in posttest (p<0.001, 0.040, and 0.004 for respective questions.) (Figure 4).

Figure 3.

A comparison of the total score correctness of participants both pre and post testing with the algorithm including correctness scores to individual questions and statistical significance. AFAB, assigned female at birth; Ab/Ag, antibody/antigen; AIDS, acquired immunodeficiency syndrome; ART, antiretroviral therapy; CMP, comprehensive metabolic panel; HIV, human immunodeficiency virus; IVDU, intravenous drug use; OI’s, opportunistic infections; PEP, post-exposure prophylaxis; PrEP, pre-exposure prophylaxis; RNA Quant, ribonucleic acid quantitative (Test)

Figure 4.

A comparison of the number of correct responses per question by students in pretest vs post-test. Post-test correct percentages are displayed in light green while pretest correctness is displayed in dark blue. The dashed red line vertical line correlates to fifty percent correct. AFAB, assigned female at birth; Ab/Ag, antibody/antigen; AIDS, acquired immunodeficiency syndrome; ART, antiretroviral therapy; CMP, comprehensive metabolic panel; HIV, human immunodeficiency virus; IVDU, intravenous drug use; OI’s, opportunistic infections; PEP, post-exposure prophylaxis; PrEP, pre-exposure prophylaxis; RNA, ribonucleic acid quantitative (Test)

Discussion

This study was intended to be an exploratory pilot of the newly developed algorithm using a convenience sample of medical students. Thus, this study was not powered to detect differences between different groups of medical students nor were demographics data collected. Furthermore, this study used a sample which may have introduced a sampling bias. Thirteen percent of students in the post-test selected an HIV regimen for treatment with medications not mentioned in the algorithm. It is possible that a time constraint or the virtual format may have affected students’ ability to use or interpret the algorithm. While the survey was not timed, the attrition of 153 students who began the survey to ninety-seven who completed the survey may demonstrate students had competing priorities with coursework or board exams as compared to voluntary research participation. This may also have limited the sample size. Conducting a study with a larger sample size may help to better identify if these trends continue. Likewise, conducting this pilot among medical students rather than practicing clinicians limits external validity. The improvements in student performance suggest the algorithm is interpretable and usable by individuals with foundational clinical knowledge, but real-world effectiveness in non-specialty or resource-limited settings may differ due to workflow, time pressures, and provider experience. Future studies will evaluate the algorithm with practicing clinicians to assess feasibility, effectiveness, cost of implementation, and generalizability to intended clinical contexts. Similar testing would be needed in other career fields (pharmacist, nurses, etc.) and among practicing providers to see if the associations in this study could be replicated in other healthcare roles. Future studies using this algorithm may also require post-test qualitative interviews with users to assess thoughts and confidence of the algorithm.

The results of the current study indicate medical students’ abilities to answer HIV-related care questions correctly were significantly improved when using the algorithm compared to unassisted. For all of the topics tested as well as overall scores, there was a statistically significant improvement in post-test correctness when using the algorithm. This demonstrated that the algorithm was able to accurately guide medical students in selecting appropriate care plans. However, we acknowledge potential threats to internal validity given testing effect bias (participants saw the same questions twice during pre- and post-test), learning/memory bias, and exposure bias from reviewing the algorithm before the post-test. Nevertheless, the algorithm shows promise for additional preclinical study with providers without significant additional HIV-training to evaluate if trends persist.

Studies prior to this have been done in attempts to increase providers’ understanding of HIV care. However, many have been limited to improving resident physicians’ awareness of PrEP or HIV without improving the clinical metrics necessary to provide this care in the future such as understanding regimen selection or laboratory requirements for implementation. One previous study of allied health professions was able to significantly improve medical residents’ awareness of PrEP and patient candidacy for PrEP. ³⁸ This study, however, did not evaluate residents’ ability to order appropriate testing required for PrEP initiation, regimen selection, or maintenance.

Some topics tested improved much more comparatively. HIV RNA quantitative testing for those with prior CAB use would not be expected knowledge for a medical student given the drug’s recent introduction which may explain the large improvement on this question. The smallest improvement was seen in HIV screenings using Ab/Ag. This may be related to student’s lack of perception about the cost of laboratory testing. Students selected answers that would detect HIV, but selected the most expensive way to do so (HIV RNA quantitative w Reflex Resistance + HIV Ab/Ag). This test would be unnecessary given the lack of a window period for the patient in the question. Rewording the question to discuss cost effectiveness may have yielded different results. Future versions of this question may need to include text about consideration of cost to the patient or medical infrastructure. Similarly, when asked when to defer treatment initiation to an ID specialist, many students selected patients with positive HBV or HCV serology. There was statistically significant improvement this question, but students may have misinterpreted the algorithm’s deferral of PrEP in patients with viral hepatitis to an ID specialist.

Implementation Considerations

Implementation of the algorithm in real-world clinical settings will require careful consideration of several factors. Workflow integration must be optimized to minimize disruption to routine clinical tasks, and providers will need adequate training to ensure accurate and consistent use. Legal scope-of-practice regulations vary across states, which may affect which providers can independently use the tool for initiating PrEP, PEP, or HIV treatment. Future studies should evaluate these implementation factors alongside algorithm effectiveness to guide successful adoption in non-specialty and resource-limited settings. To pursue those studies, adjustments to the algorithm would be required. Firstly, the algorithm would need to include a protease-based regimen as initial therapy for patients who test positive with recent injectable CAB PrEP to address concerns of INSTI resistance. ³⁹ Additionally, if the implementation site can reduce barriers to TAF-based PrEP and treatment, this would increase access for those with further reduced renal function. Similarly, it may be beneficial in a clinical setting to provide a list of AIDS-defining criteria to better identify patients who need urgent ID referral.

Conclusions

With assistance from a streamlined algorithm, students were able to answer more questions correctly about HIV care. The basis of the study showing improved decision-making in those with a baseline understanding of HIV care provides a proof of concept for future research on a modified, consolidated algorithm with practicing clinicians. Future directions should focus on evaluating the algorithm’s efficacy when used by providers in educational settings to ensure it translates similarly into improved decision making in its target providers as well as the cost to implement the algorithm as a care pathway.

ORCID iD

Colton Boney https://orcid.org/0009-0006-3647-0321

Ethical Considerations

This study was reviewed and determined to be exempt from Institutional Review Board (IRB) oversight by Alabama College of Osteopathic Medicine in accordance with the U.S. Department of Health and Human Services regulations at 45 CFR §46.104(d) (Common Rule), which outlines categories of research exempt from IRB review (Approval No. 23-09-12-001).

Consent to Participate

All study participants provided written informed consent prior to participation. No identifiable private information was collected, and disclosure of participant responses would not reasonably place subjects at risk.

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.*