← sti.care labs

Open questions

The honest calls I don't want to make alone. The first part is a quick description of what the thing is, so we're working from the same picture; the rest is the set of questions I most want outside eyes on. Take them as "help me see what I can't," not "approve my work." If any of this is your lived experience or your field, I'd genuinely value your read. There's a feedback link at the bottom.

What it is (the short version)

An app that fights STI/HIV stigma in product form: make testing and disclosure ordinary, and help people make science-based decisions, without the clean/dirty framing. It is harm reduction, never a filter that sorts people or decides who may have sex. It is live at sti.care, which makes these questions more urgent, not less: some of the calls below are built and running, and I would still rather hear I got one wrong now than keep shipping on top of it.

How it works:

The honest tension I'm sitting with: rewarding active HIV protection is a mildly "healthist" choice. I think mid-epidemic it's worth it, and I keep it from becoming a serostatus gate by making the cheapest route (condoms) count and putting free PrEP/condoms/testing one tap away. But whether that line sits right is exactly what I want feedback on.

The questions

1. Does the whole thing help or harm poz folks: gut check

This is the meta-question, the one the others feed into. Intent is anti-stigma: U=U in the core, no clean/dirty framing, harm-reduction not moralizing. But intent and impact differ.

Question: sitting with the whole concept, does it reduce stigma, or does a status- sharing app, however carefully built, inevitably make poz people feel sorted and surveilled? Is this a tool you'd want to exist?

2. Who can't reach "blue," and whether that redraws a clean/dirty line

Blue requires being actively protected against HIV (PrEP, undetectable, or public condom commitment), so some people can't reach it even if diligent:

Question: is rewarding active HIV protection the right call mid-epidemic, or does it throw the newly-diagnosed and the PrEP-decliners under the bus?

The sharpest version of this is detectable HIV in care. Today detectable HIV is an active infection on the clearance axis, so no route reaches blue until suppressed. But the app already treats condoms as equal-tier HIV prevention (the next question), and blue is defined as "actively protecting," not "no transmissible virus" and explicitly not "clean." By that logic a detectable person who publicly commits to condoms is actively protecting, and the same condom route is already the only road to blue for the HIV-negative PrEP-decliner above. The condom preference is worn by a broad mixed population, so requiring it for this subgroup wouldn't out them; it's the same camouflage the PrEP-decliner already relies on. Allowing it would also erase the "tests yet never blue" inference above. So refusing it arguably puts the clean/dirty line back on the clearance axis, the one place blue can't be reached by behavior at all.

The reason I lean no: U=U neutralizes the source biologically and continuously, while condoms over a detectable person leave live, uncontrolled virus behind a per-act barrier and a self-reported commitment. That's the largest difference of degree anywhere on the route axis, and it means blue would be sitting on top of uncontrolled live virus. There's also a knock-on cost: saying yes would force partner-notification (the receiving-end question below) to stop being uniformly gentle. A condom failure with a detectable partner is a real possible-HIV- exposure, which argues for a sharper, faster, less content-free "get tested now, consider PEP" alert, in direct tension with the calm, batched, anonymous design I chose there. So the badge change wouldn't stay contained to the badge.

Question: is treating detectable HIV as an absolute clearance blocker the right call, or is it the clean/dirty line sneaking back in? If condoms genuinely count as prevention, can the app justify accepting them for negatives but not for a detectable person in care? The blocker is what shipped (it is locked in the decisions log), but I want outside eyes on whether keeping it is principled or just fear; this is the decision I would most readily reopen for a strong argument.

One more residual on the same axis: wearing "condoms always" without the HIV-prevention label weakly implies "HIV-negative and not on PrEP." I've accepted that as the lesser harm (the three-state condom preference softens it compared to a binary), but I'd value a check on that acceptance.

3. Grouping condoms with PrEP and U=U

The three routes share the same positive state: no ranking. Condoms are real, evidence- based prevention, but typical-use efficacy is lower than PrEP/U=U. I grouped them anyway, because a lower tier for condom users re-creates the sorting I'm trying to kill, and I'm committing the app to honest info pages on the differences instead of encoding a hierarchy.

Question: does "publicly commits to condoms" genuinely belong in the same tier as PrEP/U=U, or am I flattening a difference that matters? Does leaning on education-instead-of- ranking read as honest or as a dodge?

4. Does a shareable badge become a gate?

The product refuses to be a filter: it never sorts people, never decides who may have sex, and gray is a wide bucket precisely so a look can't decode to a diagnosis. But a thing I can share is a thing someone can ask me to show. I can build it so the app is never a gate and still watch "can I see your passport?" harden into an expected step before sex, blue as the price of admission. The product refuses to be a gate; it can still be used as one.

Question: is that social pressure a real risk, disclosure becoming demanded rather than offered? And is there anything in the design that would blunt it, or does a shareable blue badge inevitably invite being used as a checkpoint?

5. Rewarding access

Rewarding PrEP and testing rewards people who can get to PrEP and testing, which tilts toward the insured and the well-resourced. I answer this inside the product: condoms count toward blue as a route that needs no clinic, and free testing, free condoms, and free/low-cost PrEP finders are one tap away. But that's my answer from where I sit.

Question: is counting condoms and putting free options one tap away actually enough to keep blue from privileging people with healthcare access, or does the reward still land hardest on those who already have the most? I'd especially value this from someone who lives closer to the access barriers than I do.

6. The partner-notification feel

If someone tests positive, their recent contacts get an anonymous, content-free nudge ("a contact suggests getting tested"), never who, when, or what. It's batched and delayed so timing can't out anyone, and it routes to free testing + PEP. The person controls who's on the list and can remove anyone easily before it sends. I kept removal frictionless, because making it hard just pushes people toward telling no one.

Question: does that feel supportive or accusatory on the receiving end? Could a content- free "go get tested" from an unknown source land as more alarming than a direct conversation?

7. What counts as a "known active infection"

This line is now decided and built: untreated gonorrhea/chlamydia/syphilis gray someone until treated; detectable HIV until suppressed; chronic lifelong conditions like HSV and HPV never gray anyone (almost every sexually active adult has been exposed to HPV, and graying someone forever over herpes would be cruel). Prior-treated syphilis serology doesn't break "clear" either; only a genuine reinfection does. For a transient high-transmission moment like a visible HSV outbreak, the answer is the ordinary pause: voluntarily gray, identical to every other gray, no reason shown, and the education teaches that pausing during an outbreak is the considerate move.

Question: does that built line feel right from the inside? Anything about how the app treats chronic vs. acute conditions, or about teaching "pause during an outbreak" as the considerate move, that would feel stigmatizing to people living with HSV/HPV?

8. Groups, and the "clean club" risk

Groups are built and live: people share status within a trusted set and get anonymous "go test" nudges if someone tests positive. The shape that shipped is deliberately mutual-care, not exclusionary: no rankings, no counts, joining/leaving looks ordinary, being in a group is itself the sharing, you join under a handle you choose, and I rejected letting groups set a testing-cadence bar ("tested within 30 days to belong" implies a verifiable standard the system can't provide, and it's the door I don't want to exist).

Question: with the construct real, does it still drift toward "you have to be blue to belong," a serosorting club, no matter how it's framed? If you've seen group dynamics like this elsewhere (WhatsApp circles, party lists, PnP communities), what drift should I be watching for that the design doesn't already block?

9. For a clinical reviewer: is a fresh "all clear" too frictionless? (the testing window)

This one is a clinical question more than a lived-experience one, so it's really aimed at an epidemiologist or clinician. Blue wants a core-panel test (HIV, syphilis, gonorrhea, chlamydia) within 90 days, all clear, but the panel isn't one moment. Results land on different clocks, and every test has a window period where a recent infection won't show yet. Syphilis is the sharp case: you can truly know you're negative for HIV/gonorrhea/chlamydia today while the syphilis result is still pending, or test so soon after an exposure that "clear" is false reassurance. An incomplete panel already stays gray (you can log what's back and add syphilis when it arrives), but a complete, just-tested panel flips blue even when the slowest result hasn't had time to mean anything. Right now that residual is left to explainer copy, not the badge. Any sharper handling has to stay on-device and never show. A visible "syphilis pending" would be exactly the tell the two-state badge exists to prevent.

Question: is a frictionless "all clear" too generous? Should the testing date (and which result came back, and when) factor into blue more honestly (capture per-result dates or a "not back yet" input, and hold blue until the slowest-seroconverting piece is both in and past its window), or does that pile on friction and a leak surface for a nuance most people won't model correctly anyway? Today the residual is explainer-only and the decisions log carries this exact item as "for outside review"; any sharper handling would stay on-device and never show.

10. Anything I'm not seeing

If you live closer to this than I do, what's the question I haven't thought to ask?

Have a perspective on any of these? I'd genuinely value your read.