Good news re opioid deaths

[jon-nyc]

Down almost half from peak.

From WaPo

[xenon]

Any theories on why?

[jon-nyc]

Supply shock I think.

[Mik]

Venezuela!!

[jon-nyc]

Short answer: there is no single cause. The sharp decline starting in summer 2023 is real, but it reflects several forces hitting at once, some intentional and some accidental.

Here’s the cleanest way to understand it.

⸻

1. Naloxone saturation finally crossed a threshold

This is probably the biggest contributor.

By mid-2023:
• Naloxone was OTC, cheap, and everywhere
• Police, EMTs, shelters, libraries, bars, and users themselves had it
• Multiple doses per event became common practice

What changed wasn’t naloxone existing — it was ubiquity + speed.
Enough overdoses were reversed before hypoxia became fatal that deaths dropped rapidly.

Key point: Naloxone doesn’t reduce overdoses — it reduces fatal overdoses. The curve you’re seeing is deaths, not use.

⸻

2. The fentanyl market changed (quietly but decisively)

This is underappreciated.

Evidence from toxicology and drug-checking shows:
• Lower fentanyl purity and variability
• Fewer ultra-potent analogs (e.g., carfentanil spikes)
• More xylazine-adulterated but less acutely lethal mixes

Why?
• Supply chain disruptions
• Shifts in Mexican production and distribution
• Dealers responding to customer death backlash (dead customers don’t buy)

The drug supply became more predictable, which paradoxically saves lives.

⸻

3. People adapted their behavior

After years of carnage, user behavior changed:
• Test doses
• Never using alone
• Staggered dosing
• Widespread fentanyl test strip use

This is Darwinian harm reduction — brutal, but real.

Once a critical mass of users changes behavior, death rates can fall fast even if addiction rates don’t.

⸻

4. Medication-assisted treatment finally scaled

By 2023:
• Buprenorphine prescribing barriers were relaxed
• Methadone access expanded in many states
• Jail/prison initiation programs increased

MAT doesn’t eliminate use — it reduces lethality, especially fentanyl lethality.

Again: fewer deaths, not necessarily fewer users.

⸻

5. There is some statistical compression — but not enough to explain this

Yes, there are:
• Reporting lags
• Provisional data effects
• Cause-of-death coding delays

But:
• The decline is too large and sustained to be an artifact
• Independent datasets show the same inflection

This isn’t just bookkeeping.

⸻

What this is not
• Not a sudden success of the “war on drugs”
• Not primarily incarceration
• Not abstinence
• Not moral reform

It’s engineering, distribution, and adaptation, not virtue.

⸻

The uncomfortable but accurate summary

Overdose deaths fell not because fewer people use opioids, but because we finally made opioid use less fatal.

That’s why the drop is steep and sudden — once survival probability crosses a threshold, curves bend fast.

If you want, I can:
• Break down which states drove most of the decline
• Explain why deaths from xylazine didn’t rise in parallel
• Or compare this to historical declines in HIV deaths once ART scaled (the pattern is eerily similar)