The Encounters Happened. The Prevention Didn't.

On July 17, 2026, the Centers for Disease Control and Prevention (CDC) released findings from SHIELD, the Surveillance of HIV-related Service Barriers Among Individuals with Early or Late HIV Diagnoses. The project surveyed 462 people in Florida, Louisiana, Michigan, and the City of Houston who had received an HIV diagnosis at either the earliest or the latest stage. One finding organizes everything else: 78% of participants diagnosed at stage 0 and 63% of those diagnosed at stage 3 had seen a healthcare worker during the 12 months before their diagnosis. Only 38% and 20% discussed HIV testing at that visit. For PrEP, the numbers were 23% and 6%. More than half of participants diagnosed early, and 71% of those diagnosed late, reported never being offered an HIV test before the test that diagnosed them.

The people were in the room. The conversation did not happen.

What SHIELD Was Built to Ask

Case surveillance tells us who received a diagnosis and when. It cannot tell us what preceded it. SHIELD closes that gap, drawing participants from enhanced HIV/AIDS Reporting System databases and surveying them a median of 8 months after diagnosis. Of 3,360 people sampled, 496 participated and 462 entered the analytic sample.

The limits matter and CDC states them: an unweighted convenience sample, descriptive only, cross-sectional, with no generalizability beyond the 4 project areas. Dr. Pollyanna Chavez, a former CDC epidemiologist who worked on the study’s design, described its purpose on LinkedIn as answering what case surveillance cannot. What SHIELD offers is not a national estimate. It is a map of where the system had contact and did nothing with it.

The Missed Encounters Cluster Where Suspicion Should Be Highest

Among participants who tested for other sexually transmitted infections (STIs) in the year before their HIV diagnosis, 33% of those diagnosed at stage 0 and 52% of those diagnosed at stage 3 were not offered an HIV test at that same visit. Feeling sick was the most common reason for the test that eventually produced the diagnosis, cited by 56% and 53%. For half of participants diagnosed late, that test happened in a hospital, emergency room, or other inpatient setting. Nearly half of people diagnosed at stage 3 had never tested for HIV at all.

None of this should be particularly surprising. CDC's analysis of ambulatory care found HIV testing occurred at 0.63% of physician office visits, 2.65% of community health center visits, and 0.55% of emergency department visits between 2009 and 2017. The regional trend has since moved backward: across 9 Deep South states, the share of adults who had ever tested fell from 47% in 2019 to 41% in 2023, with significant declines in past-year testing in Florida and Louisiana. The pandemic accounts for part of that drop, but not for its persistence. National testing and diagnoses returned to pre-pandemic levels by 2021, while testing across the Deep South remained below its 2017 rate through 2023. 2 SHIELD jurisdictions were losing ground on testing during the years SHIELD was collecting its data, and they did not recover when the rest of the country did.

Why Nobody Raises It

Both sides of the encounter fail, for reasons that are structural rather than personal.

People did not test because they did not believe they needed to. Self-assessed lack of perceived risk was the leading reason for never asking for a test, cited by 47% of participants diagnosed early and 51% of those diagnosed late. Awareness was thinner still: 42% of people diagnosed at stage 3 had never heard of PrEP, and 62% had never heard of an HIV self-test.

Disclosure ran the same direction. Among men who have sex with men diagnosed at stage 3, 47% had not told a healthcare worker they were attracted to or had sex with men. Roughly 1 in 5 participants reported experiencing discrimination from a healthcare worker before their diagnosis, most often attributed to sexual orientation. Nondisclosure and discrimination are the same phenomenon viewed from opposite ends of the exam table. Any screening model that depends on a person volunteering sensitive information to a provider or system they have reason to distrust will systematically miss the people it most needs to reach.

Any screening model that depends on a person volunteering sensitive information to a provider they have reason to distrust will systematically miss the people it most needs to reach.

On the provider side, a systematic review of 28 studies identified 6 recurring barriers to PrEP prescribing, among them limited provider knowledge, cost assumption concerns, provider-driven interpersonal stigma, and provider discomfort discussing sexual behavior.

Underneath all of it sits the payment structure. Under bundled arrangements, providers receive the same reimbursement regardless of which procedures they perform, so no additional payment attaches to an HIV test. Providers also describe what follows a reactive result, the state notification and the linkage-to-care work, as effort they cannot bill.

A 2022–2023 survey of emergency department directors found routine HIV screening was the least offered preventive service of any measured, at 19%, with 58% citing increased costs to their department. We are asking healthcare staff to perform unpaid work in systems that measure them on throughput, then treating the resulting gap as a matter of individual attitude.

Coverage on Paper, Access in Practice

The legal requirement is settled. In Kennedy v. Braidwood Management, Inc., decided June 27, 2025, the Supreme Court preserved the Affordable Care Act's requirement that qualified health plans cover U.S. Preventive Services Task Force grade A and B services without patient cost-sharing. Routine HIV screening and PrEP both carry grade A ratings.

What that produces in practice is a separate question. The AIDS Institute reviewed 310 marketplace plans across all 50 states and the District of Columbia for the 2026 plan year. More than half, 52%, do not list any form of long-acting injectable PrEP as covered without cost sharing. Fully 84% fail to list all forms of oral and injectable PrEP. 25% do not include PrEP on a preventive services list at all. 60% of plans earned a failing grade, and the average plan failed in 72% of states. CDC research cited in the same report found 20-30% of commercially insured PrEP users were improperly charged cost-sharing in 2021 and 2022.

The enforcement design explains the result. Regulators assume plans are complying unless enrollees report a problem, placing the burden on people to recognize an improper charge, understand the requirement being violated, and appeal. State Insurance appeals processes are burdensome on patients, often difficult to identify or navigate, and resolution typically takes significantly longer than even appeals to insurance carriers themselves. 9 states have commissioners who issued PrEP coverage guidance, and those states graded above the national average. That is 18% of regulators. Michigan is among them. Florida, Louisiana, and Texas are not, which means 3 of the 4 SHIELD jurisdictions have no state enforcement posture on the very coverage SHIELD participants would have needed.

Mississippi shows what enforcement looks like when advocates supply it. In June 2026, the HIV+Hepatitis Policy Institute filed a complaint with the state insurance commissioner alleging Blue Cross and Blue Shield of Mississippi unlawfully restricted PrEP access. Within weeks, the insurer added Descovy and Apretude to its formulary. The commissioner told Mississippi Today it is not his responsibility to regulate which drugs a private plan covers.Mississippi has the highest HIV death rate in the country and the lowest rate of PrEP use relative to new diagnoses. As HIV + Hep Institute's Carl Schmid put it, it is the commissioner's "job to enforce the law, not to decide who deserves" protection.

Schmid is correct, as found in Centers for Medicare and Medicaid Services' (CMS) Marketplace & Private Insurance Compliance and Enforcement guidance indicate. Waivers for state-driven primary enforcement may be issued when a state notifies CMS of an inability to enforce and submits to a Collaborative Agreement. As of the above described actions and the time of this writing, Mississippi is not among the 4 states that have sought and been granted such a waiver.

Nationally, 26.3% of people who could benefit from PrEP are taking it, including 15% of Black people, 18% of Hispanic and Latinx people, and 9% of women.

Default Screening Works Where It Is Required

Between April 2022 and January 2025, England's National Health Service ran opt-out bloodborne virus testing across 34 emergency departments in high-prevalence areas, delivering more than 7 million tests. The program produced 719 new HIV diagnoses and re-engaged 291 people previously diagnosed but out of care. Uptake reached 70% of eligible patients; 60% had no prior testing record.

The finding that should stop us: 80% of people newly diagnosed through the program had attended an emergency department in the year before it launched. That is the SHIELD pattern, in a health system that responded by making the test automatic.

Domestically, when Florida's only legal syringe services program moved to bundled opt-out testing, uptake rose 42.4% immediately and held across 22 months, with acceptance climbing from 33.1% to 91.3%. The people did not change. The default did.

What We Should Push For

CDC should finalize its draft screening recommendations. In December 2024, CDC published draft Recommendations for HIV Screening in Clinical Settings proposing at least 1 HIV test in a lifetime for everyone 15 and older, eliminating the upper age limit, encouraging automated laboratory orders, and treating general consent as sufficient. Comment closed January 2, 2025. 19 months later, the 2006 recommendation remains operative. Finalization establishes the standard of care against which practice is measured, converting screening from a discretionary act into an accountable one.

Health systems should automate the order and code it correctly. Auto-ordering an HIV antigen/antibody testalongside a complete blood count for patients aged 15 to 65 with no prior test in 12 months removes the requirement that a clinician remember to ask. Bundling HIV into STI panels closes the gap half of SHIELD's late-diagnosed participants fell through. Correct procedural coding ensures the screening is reimbursed and the patient carries no cost sharing.

State insurance commissioners should issue PrEP coverage bulletins and audit compliance. 9 have. Advocates in the remaining 41 states have a specific ask with a proven template.

States should move routine HIV screening from optional to required. Essential Health Benefits benchmark plans are state-determined, a lever available now without federal action. The same applies to traditional Medicaid, where routine screening remains optional. Texas offers the cautionary case: HB 50 would have made HIV testing part of routine STI screening on an opt-out basis, passed the House 116 to 21, and died in Senate committee for the second consecutive session. Houston is a SHIELD jurisdiction.

Conclusion

In 2022, an estimated 13% of the roughly 1.2 million people living with HIV in the United States were undiagnosed, unaware of their status and outside of care. That share has moved 1 percentage point since 2018. Ending the HIV Epidemic targets a 90% reduction in new transmissions by 2030, a goal that depends entirely on diagnosing people earlier than we do now.

The funding picture complicates that. The FY2027 House Labor-HHS bill advanced June 9, 2026 holds the EHE line at $220 million while cutting HIV, viral hepatitis, STI, and tuberculosis prevention by $818.1 million and eliminating the Ryan White EHE component. The initiative's name survives; the infrastructure it runs on does not. Senate action remains pending.

SHIELD gives us something more useful than another measure of distance from the goal. It identifies the specific moments when the healthcare system had a person in front of it and let the opportunity pass. Those moments are addressable through order sets, coding rules, benchmark plans, and enforcement bulletins. We have the tools, the evidence that they work, and now the data showing precisely where they are not being used.