The US overdose story includes real progress. Deaths have fallen in the periods covered by recent national releases. That does not settle a family’s treatment decision. A medication-free discharge can look like a clear endpoint, but opioid-treatment evidence asks a different question: does the person remain safer, engaged in care, and able to sustain recovery afterward?
Key findings
69,973: predicted provisional US overdose deaths in 2025 in CDC’s May 2026 release, compared with 81,313 in 2024.
44,564: opioid-involved deaths in 2025 in the same release, down from 55,296 in 2024.
15.7%: past-year medication receipt among people aged 12 or older with past-year opioid use disorder in SAMHSA’s separate 2025 survey of the US civilian, noninstitutionalized population.
67,798: predicted deaths in the rolling 12 months ending March 2026, reported by CDC in August 2026. This is not a first-quarter total.
These measures describe different populations and periods. They cannot be combined to calculate what caused the mortality decline. The survey percentage is not the proportion of all medication recipients: some people receiving medication did not meet past-year disorder criteria. [1] [2] [3]
The national decline deserves accurate reporting
Source: CDC/NCHS, May 13, 2026. Both values use the same release vintage. The decline is almost 14%; the numbers are not final registrations. [1]
The later August update reported a rolling-year total 12.3% below the corresponding period a year earlier. A rolling year overlaps the calendar-year series and comes from a newer release. Treating it as a separate, nonoverlapping annual observation would misrepresent the trend. [3]
National improvement also does not mean every jurisdiction improved. CDC’s May release reported increases in some states. Population surveillance informs planning, but neither a reassuring average nor a worsening national figure can substitute for an individual’s clinical assessment.
| Measure | Value | Scope |
|---|---|---|
| All-drug overdose deaths, 2025 | 69,973 | Predicted provisional count, May 2026 release |
| Opioid-involved deaths, 2025 | 44,564 | Subset of overdose deaths; drug categories can overlap |
| Medication receipt among people with past-year OUD | 15.7% | 2025 survey estimate |
| All-drug deaths, year ending March 2026 | 67,798 | Rolling-year predicted count, August release |
Sources: CDC and SAMHSA. The rows are not one patient cohort and must not be added together. [1] [2] [3]
Medication-free is not a universal definition of recovery
Stopping harmful or uncontrolled opioid use is not the same as stopping every medication used in treatment. CDC recommends evidence-based medication treatment for opioid use disorder. It states that detoxification alone, without that treatment, is not recommended because of the risks of resumed use, overdose, and death. [4]
That creates an important question for a program advertised as a rapid reset. Does its endpoint describe completed withdrawal management or a supported pathway toward sustained recovery? A completed stay does not answer the question unless the plan after discharge is clear.
The evidence does not require every person to receive an identical medication or remain on an unchanging plan. Preferences and clinical circumstances matter. There is no recommended fixed duration limit for methadone or buprenorphine treatment of opioid use disorder; any reduction or discontinuation should be planned with the treating clinician. It does challenge the idea that the absence of treatment medication is always proof of success.
Retention and mortality are linked in cohort research
A 2017 review combined 19 observational cohorts. In methadone cohorts, pooled all-cause mortality was 11.3 deaths per 1,000 person-years during treatment and 36.1 outside treatment. In buprenorphine cohorts, the corresponding rates were 4.3 and 9.5. These were periods in and out of treatment, not a randomized comparison between medications. [5]
Source: Sordo and colleagues, BMJ, 2017. The pooled unadjusted out-to-in rate ratio was 3.20, with a 95% confidence interval of 2.65 to 3.86. Different cohorts cannot support a head-to-head medication ranking. [5]
Person-years combine people and observation time. These rates are not percentages dying during a single fixed program, and they are not 2026 rates for newly treated patients. Their contribution is evidence about retention, interpreted alongside current guidance and individual assessment.
For affluent families, continuity is a purchasing question
Resources may make private assessment or travel for care possible. They do not automatically establish access to every appropriate treatment or agreement between clinicians in different jurisdictions. The relevant question is who remains responsible after the initial setting.
A proposed pathway should identify the coordinating clinician, follow-up arrangements, and response to changing needs. Records should be shared with permission, and the receiving service should be involved. A recommendation that another professional take over is not the same as an agreed transfer.
No cited dataset measures how often affluent families experience fragmented care. The client-specific value here is a set of questions to make explicit before a purchase, not an invented percentage describing wealthy patients.
Survival matters, and so does the rest of life
Keeping people alive is central. Broader recovery reporting should also describe substance use, disorder symptoms, physical and mental health, relationships, and meaningful daily activities. These measures complement safety rather than compete with it.
For someone returning to leadership, attending meetings is not sufficient evidence of recovery. Sustainable participation and the difficulties that prompted care may be better goals. The patient’s priorities should shape the evaluation rather than a provider’s preference for a striking success claim.
Every follow-up needs an eligible count and an observed count. People who cannot be reached should not silently disappear or be assumed to match respondents. Their outcomes are unknown, and the effect of missing information belongs in the interpretation.
The treatment gap is not only a financial question
SAMHSA’s July 27, 2026 release of the 2025 survey results describes a gap between substance-use treatment need and receipt. Reasons reported by specified subgroups include wanting to manage alone, insufficient time, and concern about other people’s reactions. Those responses are not a survey of wealthy executives. [2]
The affluent-client research question is whether similar barriers remain when cost is less restrictive. Professional standing, family expectations, and public visibility are plausible subjects to study, but their effects need measurement rather than assumption.
A confidential first discussion should clarify options, responsibilities, and risks without promising an outcome. People should not have to accept a simplistic identity or a single treatment philosophy before their circumstances are assessed.
What a private program should explain
| Area | Explanation to request |
|---|---|
| Clinical plan | Why the approach is appropriate, including medication’s role |
| Safety | How overdose risk, other substances, and medical needs are assessed |
| Continuity | Who provides ongoing care after the initial program |
| Outcomes | What is measured, when, and among whom |
| Adaptation | What changes if improvement is insufficient |
These questions concern the completeness of the care plan, not a requirement that every service guarantee permanent recovery. A useful explanation makes uncertainties and responsibilities visible so that the patient can understand the next step.
The plan should also address safety beyond the treatment episode. Opioid tolerance can fall after stopping use, so returning to a previously used dose can cause a fatal overdose. CDC recommends overdose-reversal education and access to naloxone for people with opioid use disorder and relevant household members. During a suspected overdose, call local emergency services immediately and use available reversal medication according to its instructions and emergency guidance. Naloxone is temporary, and additional doses may be needed according to its instructions. Stay with the person until emergency help arrives, even if they appear to recover. [4]
For international care, local availability and prescribing arrangements need to be established with the clinicians involved. A favorable study result cannot guarantee that a receiving service will provide a particular medication or accept an incomplete handover.
What the numbers cannot tell us
The mortality estimates remain provisional. Calendar-year and rolling-year totals overlap and come from different releases. Survey medication receipt is not a mortality measure. The survey’s opioid use disorder measure does not capture symptoms arising solely from illegally manufactured fentanyl use. Historical cohort rates cannot establish a causal contribution to the recent national decline.
The evidence does not rank private clinics, estimate outcomes among HNW individuals, or report THE BALANCE patient results. It supports careful questions about treatment and retention without replacing an individualized clinical decision.
The outcome should describe the whole pathway
A family’s comparison should begin before admission and continue beyond discharge. What clinical need was identified, what treatment was delivered, and what support followed? A short stay can be one part of care without being the entire intervention. Reporting only the visible episode can obscure the ongoing treatment associated with later improvement.
Programs should make clear whether their figures describe symptoms, engagement, safety, or functioning. Those outcomes are complementary, not interchangeable. A favorable response in one area should not be used to imply that every other outcome improved, or that the same result applies to patients who were not assessed later.
The bottom line
The falling death count deserves to be reported accurately. It does not make treatment unnecessary or every approach equally effective. The stronger test of care is sustained engagement, safety, and meaningful recovery after the visible treatment episode ends, not whether discharge produces the shortest medication list.
For journalists
Key contrast: falling predicted overdose deaths alongside limited medication receipt among surveyed people with past-year opioid use disorder.
Important caveat: the surveillance, survey, and cohort figures are different evidence sources and must not be combined into one causal calculation.
Suggested attribution: THE BALANCE analysis of CDC, SAMHSA, and published opioid-treatment evidence.
Methodology and sources
The mortality chart uses the May 2026 release consistently. The August rolling-year update is labeled separately. Cohort mortality retains its original person-year units and publication date. No new pooled analysis or provider-effect estimate was produced.
- Centers for Disease Control and Prevention, National Center for Health Statistics. (2026, May 13). U.S. Overdose Deaths Decrease for Third Consecutive Year in 2025. Provisional estimates.
- SAMHSA. 2025 NSDUH treatment-gap findings, July 2026.
- CDC. Rolling-year update, August 31, 2026.
- CDC. Treating opioid use disorder.
- Sordo and colleagues. Mortality risk during and after opioid substitution treatment: systematic review and meta-analysis of cohort studies. BMJ, 2017;357:j1550.
