I've been thinking about interaction checkers for a while now. Not in the abstract, but because I've watched people actually use them. Somebody at a festival pulls out their phone at 11pm, types in two substances, and stares at a colored square on a chart trying to figure out if they're about to make a dangerous decision. That moment is where harm reduction either works or it doesn't.
The tools we have right now are good. Some of them are really good. But the gap between what exists and what's possible is wider than most people realize, and I think it's worth talking about openly, even as somebody who partners with the organizations building these tools.
The current landscape (and why it matters)
There are two main players doing this work in the harm reduction space right now. TripSit's combination chart has been the go-to reference for years. It covers around 24 substances with pairwise ratings, it's open-source, and it's been used by more people than probably anyone can count. It's the foundation.
Then there's Substancy's interaction checker, which is a newer entry from our partners at Drugs and Me. They're taking a different approach. Instead of a single color square per pair, they're building a dose-response matrix that accounts for effect level, route of administration, and specific risk factors like body temperature, heart rate, and neuronal toxicity. It's a step toward the kind of detail that clinicians work with, but presented for people who actually use these substances.
Both of these tools are free, which matters. Both are built by people who care about this work. And both are making real decisions safer for real people. That's not nothing. That's actually a lot.
"The moment someone opens an interaction checker is the moment harm reduction either works or it doesn't. Everything about the tool should serve that moment."
What Substancy is getting right
I want to call out some specific things Substancy is doing that push the space forward, because I think they're easy to miss if you're just glancing at the interface.
The effect level selection is a smart move. Somebody who's had three beers is having a fundamentally different interaction with another substance than somebody who's had one. Most charts don't account for that. Substancy does, and it changes the output. That's closer to how clinicians think about drug interactions, where dose is always part of the equation, never just the substance name.
The route of administration dropdown is another one. Smoking cannabis is a different drug experience than eating an edible, and those two routes interact differently with other substances because of onset timing, bioavailability, and duration. Including ROA in the interaction check is the kind of detail that makes the output actually useful instead of generically cautionary.
And the per-system risk breakdown, showing separate ratings for body temperature, heart rate, respiratory rate, liver toxicity, and so on, gives people a more specific picture of what's actually at stake. Not just "danger" but "danger to what, specifically." That's the kind of information that changes behavior.
The partnership with Drug Science and the involvement of organizations like Reagent Tests UK also signals credibility in the right direction. This isn't a random website making claims. There are named organizations behind it.
Where the whole space needs to go
Here's where I start thinking out loud. This isn't a criticism of any specific tool. It's a vision for what a 10/10 interaction checker could look like if we built it from the person outward instead of from the pharmacology inward.
I keep coming back to a simple question: who is the person using this tool right now, and what do they actually need?
Key Insight
People use interaction checkers in three or four very specific situations. Each one needs a different experience. Somebody panicking at a festival is not the same user as somebody planning a session next weekend, and neither of them is the same as someone trying to help a friend who just told them they took two things.
Person-centered triage. The first thing an interaction checker should ask isn't "what substances" but "who are you right now?" Someone needing fast answers doesn't need to scroll past pharmacology. Someone planning ahead wants all the depth. Someone worried about a friend needs emergency action steps, not enzyme pathway explanations. The tool should reshape itself based on that initial answer.
Trauma-informed design. A lot of the people using these tools have complicated relationships with substances, medical systems, or both. The language matters. "Substance use" not "abuse." "Person who uses drugs" not "addict." No moralizing. No unsolicited recovery resources bolted onto every page. The tone should be a knowledgeable friend, not a disappointed doctor.
Source transparency with tiers. This is the one I feel strongest about. Some interaction data comes from systematic reviews of controlled clinical studies. Some comes from a single case report. Some comes from PsychonautWiki trip reports. All of that information has value, but the user needs to know which is which. A tiered source system, where every claim shows its evidence level, builds the kind of trust that keeps people coming back. When the evidence is thin, say so. The harm reduction audience has been lied to by enough institutions. Honesty about uncertainty is what earns credibility.
Timing and order matter. Taking MDMA first and then psilocybin two hours later is a fundamentally different experience than the reverse order. Most interaction charts treat these as the same combination. They're not. The classic "hippie flip" protocol exists because the order changes the pharmacology. A next-generation checker should show this.
Medication flags that persist. The SSRI and MDMA interaction is probably the most common question in harm reduction, and it's also one of the most dangerous. A tool that lets you flag your daily medications once and then shows relevant warnings on every interaction check would catch people who don't think to look up their antidepressant as a "substance." Same for MAOIs, lithium, and benzodiazepines.
The content gap nobody talks about
Every interaction checker tells you the risk. Almost none of them tell you what to do about it.
If I look up alcohol plus MDMA and the tool says "unsafe," what am I supposed to do with that? I need to know: if I've already taken both, what do I watch for? At what point do I call for help? What do I actually say to the 911 dispatcher? How do I put someone in the recovery position? What does the next 72 hours look like?
This is the content that saves lives, and it's the content that's hardest to produce because it requires clinical knowledge, not just pharmacological data. But it's what separates a reference chart from a tool that actually serves people in the moment they need it.
Safety Note
If you're ever unsure about a combination you or someone near you has taken, the Fireside Project (62-FIRESIDE) offers free, confidential peer support. For medical emergencies, call 911. Good Samaritan laws protect you in most US states when calling for help.
Connecting the silos
One thing I notice across all existing tools is that they operate independently. You check an interaction, get a result, and that's it. But the actual user journey is longer than that. Before you take anything, you should be preparing your set and setting. You should be testing your substances with a reagent kit (our resources page links to DanceSafe, Bunk Police, and Miraculix for testing kits). And afterward, you should be integrating the experience.
The interaction checker sits in the middle of that arc. A truly connected tool would link outward to preparation resources, reagent testing, crisis support lines like Fireside, and aftercare guidance. Not as an afterthought but as part of the core flow. You look up a combination, and the tool helps you prepare for it, test it, stay safe during it, and process it afterward.
This is what Substancy is starting to build toward, and it's what organizations like TripSit have been laying the groundwork for. The data they've been gathering, the interaction reports they've been cataloging, the community they've built around open-source harm reduction data, all of that is infrastructure that the next generation of tools will build on.
The privacy question
I want to flag something that doesn't get talked about enough. The people using interaction checkers have real legal exposure. They're looking up information about controlled substances, often on their personal devices, sometimes while in possession of those substances. A tool that's serious about serving this population needs to be serious about privacy.
No accounts. No tracking. No third-party analytics scripts. Client-side processing. The ability to work offline. A plain-language privacy policy that says what isn't collected. These aren't nice-to-haves. For this specific audience, they're trust requirements.
"Honesty about uncertainty is what earns credibility. When the evidence is thin, say so. The harm reduction audience has been lied to by enough institutions."
What I'd love to see next
I'm not building an interaction checker. That's not Pada's lane. But I am somebody who cares a lot about the space and partners with the people who are building these tools. So here's my wish list for whoever picks up the next version, whether that's Substancy iterating on what they have, TripSit evolving their chart, or someone entirely new.
Build it with the people it's meant to serve. Not a focus group at a university. Sit at festival medical tents, at needle exchanges, at integration circles, and watch people try to use the current tools. Watch where they get lost. Watch where they close the tab. The best harm reduction tech comes from people who understand the context because they've been in it.
Make it work on a bad phone with one bar of service. Server-side render the content. No JavaScript required for the core safety information. This is a functional requirement for the actual use case, and most current tools fail it.
And keep it open. Open data, open source, openly governed. The interaction chart is a public good. It should be built like one.
Frequently Asked Questions
What is a drug interaction checker?
A drug interaction checker is a tool that shows you the risks of combining two or more substances. It typically rates combinations from low risk to dangerous based on pharmacological data and community reports. The most well-known harm reduction versions are TripSit's combination chart and Substancy's interaction checker on the Drugs and Me platform.
Are interaction checkers reliable?
They're a useful starting point, not a final answer. The best ones cite their sources so you can see the evidence quality. Some combinations are well-studied clinically, while others rely primarily on community reports. Always do additional research, and when in doubt, assume higher risk.
What's the most dangerous drug combination?
Combinations that cause respiratory depression are the most immediately life-threatening: opioids with benzodiazepines, opioids with alcohol, GHB with alcohol, and GHB with benzodiazepines. MDMA with MAOIs can cause fatal serotonin syndrome. These combinations should always be avoided.
Can I take MDMA if I'm on antidepressants?
SSRIs block the serotonin transporter that MDMA relies on. At best, MDMA won't work. At worst, the combination can cause serotonin syndrome, which is a medical emergency. Do not stop taking your SSRI to use MDMA without consulting your prescriber, as abrupt discontinuation has its own serious risks.
Where can I test my substances before taking them?
Reagent testing kits are available from DanceSafe, Bunk Police (Pada community discount: 10% off with code PADA10), and Miraculix for quantitative testing. Fentanyl test strips should be used on every substance, every time. See our full testing resources.
Pada is a harm reduction company. We don't sell or endorse the use of any controlled substances. This content is for educational purposes only.