In addiction medicine a falling opioid dose is usually read one way: somebody tapered the patient. The dose is something done to a person, and the clinical question is how fast and how well it was done.
There is a second route to the same number, and it produces a different patient at the end of it. The requirement itself falls, because the thing generating the pain was treated. That distinction is worth holding onto, because the two look identical on a prescription record and feel nothing alike to the person living through them.
Why the distinction is not semantic
The harms of getting a taper wrong are well quantified. In a cohort of 113,618 patients on stable higher-dose opioid therapy, tapering was associated with 9.3 overdose events per 100 person-years against 5.5 in non-tapered periods, and 7.6 mental health crisis events against 3.3 — and the faster the dose came down, the worse both got. A follow-up analysis found the elevated risk persisting up to two years after the taper began. We covered that work when it published: opioid tapering and the risk of overdose or mental health crisis.
So a dose reduction is not self-evidently a good outcome. It is an intervention with its own risk profile, and whether it helps depends almost entirely on what else changed.
The mechanism that lets a dose fall on its own
Sustained opioid exposure can produce two separate problems that both look like undertreatment. Tolerance means the drug still works but more is needed. Opioid-induced hyperalgesia means the nervous system has been recalibrated toward pain, so more drug feeds the problem rather than fixing it. The proposed mechanisms include the central glutamatergic system and NMDA receptor activation, spinal dynorphin, and descending facilitation.
The clinical tell is what a dose increase actually buys. With tolerance it helps, at least for a while. With hyperalgesia it buys a brief improvement and then a worse baseline, often with pain spreading beyond the original site. One review puts the consequence bluntly: where hyperalgesia is operating, further opioid prescribing is largely futile.
This is where the second route comes in. A 2019 review of opioid tolerance and hyperalgesia notes that interventional techniques which reduce the pain input can allow the opioid dose to come down, and in doing so revert the mechanisms producing tolerance and hyperalgesia in the first place. The dose falls as a consequence of treating something, not as a target imposed on the patient.
What that looks like in a treated population
On International Overdose Awareness Day in August 2026, Dr. Gurpreet Singh Padda, MD, MBA, MHP published the opioid figures from his interventional pain practice rather than a position statement.
- New patients arrive on an average of more than 90 morphine milligram equivalents per day, after more than two and a half years in pain.
- Within 90 days of active interventional treatment, 21% are completely off opioid pain medication.
- Within one year, 34% are completely off all opioid pain medication.
- Of those who cannot be fully weaned, the large majority are brought below 30 MME per day.
Those are practice-reported figures from that population, not trial outcomes, and individual results vary. The release was carried by AP News; the full announcement is here.
Where this matters for addiction treatment
Roughly one in twenty patients in that practice is co-managed for addiction alongside severe pain — the population that tends to fall between two services, because pain clinics are wary of the addiction history and addiction programs are not equipped to treat the pain. Both problems are real at the same time, and treating only one of them reliably makes the other worse.
The practical implication is not that everyone should be weaned. It is that why a dose is coming down predicts whether the reduction holds. A taper against a pain generator nobody has addressed is arithmetic, and the cohort data above describe what tends to happen next. A reduction that follows treatment of the generator is a different event.
It is also worth being honest about the evidence boundary. A systematic review of opioid-induced hyperalgesia outside the surgical setting found the signal real in experimental models but inconsistent enough that its clinical impact could not be quantified. Hyperalgesia is a mechanism to suspect and test for, not a diagnosis to assume.
If this is your situation
If you or someone you are treating is on a dose that keeps climbing without a corresponding gain in function, the question worth asking is not only how to come down but what has been done to find and treat the pain generator. Our approach to opioid addiction treatment, and about Dr. Padda.
Sources
- Agnoli A, Xing G, Tancredi DJ, Magnan E, Jerant A, Fenton JJ. Association of dose tapering with overdose or mental health crisis among patients prescribed long-term opioids. JAMA. 2021;326(5):411–419. PMID 34342618 · DOI 10.1001/jama.2021.11013.
- Fenton JJ, Magnan E, Tseregounis IE, Xing G, Agnoli AL, Tancredi DJ. Long-term risk of overdose or mental health crisis after opioid dose tapering. JAMA Network Open. 2022;5(6):e2216726. PMID 35696163 · DOI 10.1001/jamanetworkopen.2022.16726.
- Mercadante S, Arcuri E, Santoni A. Opioid-induced tolerance and hyperalgesia. CNS Drugs. 2019;33(10):943–955. PMID 31578704 · DOI 10.1007/s40263-019-00660-0.
- Colvin LA, Bull F, Hales TG. Perioperative opioid analgesia — when is enough too much? A review of opioid-induced tolerance and hyperalgesia. The Lancet. 2019;393(10180):1558–1568. PMID 30983591 · DOI 10.1016/S0140-6736(19)30430-1.
- Lee M, Silverman SM, Hansen H, Patel VB, Manchikanti L. A comprehensive review of opioid-induced hyperalgesia. Pain Physician. 2011;14(2):145–161. PMID 21412369.
- Yang DZ, Sin B, Beckhusen J, Xia D, Khaimova R, Iliev I. Opioid-induced hyperalgesia in the nonsurgical setting: a systematic review. American Journal of Therapeutics. 2019;26(3):e397–e405. PMID 29726847 · DOI 10.1097/MJT.0000000000000734.
Bibliographic records in this article were retrieved from PubMed. Every reference links to its PubMed record and DOI.