RTI Measure Tracks Hospital Care After Opioid Treatment Begins

At the median opioid treatment program in the United States, one in twenty Medicaid patients who started care in 2024 had a substance use disorder-related emergency department visit or hospital stay within 90 days, according to a study published October 25 in the journal Addiction. The measure is the first designed to score opioid treatment programs on what happens to patients after they walk in, rather than on whether they walked in at all.

Researchers at RTI International calculated the rate for 1,226 programs using 412,411 new treatment episodes involving 257,986 Medicaid enrollees diagnosed with opioid use disorder. The median program rate was 5.0 percent, but the middle half of programs ranged from 2.8 percent to 8.0 percent — a spread wide enough that a patient's odds of landing in an emergency department say something about the clinic, not only about the patient.

What the measure counts, and what it leaves out

The definition is deliberately narrow. An episode counts against a program if, within 90 days of starting treatment, the patient has an emergency department visit or an inpatient stay carrying a substance use disorder diagnosis or a drug poisoning code. Routine counseling visits, missed appointments and positive drug tests do not count. Neither does anything that happens after the 90-day window closes.

That construction means the measure runs on claims data the Medicaid program already holds. States, health plans and federal agencies could compute it without asking treatment programs to file a single additional report — an unusual property in a field where quality reporting has historically meant paperwork for clinics with thin administrative staff.

The 2024 measurement year produced more than 103,300 qualifying acute care events. Of those, 80.2 percent were emergency department visits and 19.8 percent were inpatient hospitalizations. Opioid-related disorders were the most common principal diagnosis, at 47.4 percent of events, which means the majority of the acute care being counted was for something other than opioids — alcohol, stimulants, or a combination.

Variation that survives casemix adjustment

The obvious objection to any clinic-level rate is that sicker patients produce worse numbers. The authors tested that directly. After adjusting for patient mix, 46 of the 1,226 programs moved from the 25th percentile of the distribution to the median, and 20 programs moved from the median to the 75th percentile. Casemix explained some of the spread at the tails, but most of the gap between the best and worst programs held.

"While many existing measures focus on whether patients begin treatment or remain in care, this new measure examines whether patients experience serious substance use-related acute care events after treatment starts," said Barrett Montgomery, a drug use epidemiologist at RTI and the study's first author. He described the measure as a complement to existing metrics rather than a replacement.

That framing matters because opioid treatment programs already report to several masters. They are certified by the Substance Abuse and Mental Health Services Administration and inspected under federal narcotic treatment regulations, licensed by states, and often accredited by bodies such as CARF or the Joint Commission. Retention has been the outcome measure of choice: this blog covered an RTI analysis in September that found programs keeping patients past 30 days recorded fewer hospitalizations, emergency visits and deaths. A retention number and an acute care number describe different failures. One catches the patient who leaves; the other catches the patient who stays and still ends up in a hospital.

The blind spot the measure is meant to fix

Program staff usually learn about a hospitalization only if the patient mentions it. "Opioid treatment programs often do not know when their patients are hospitalized or visit an emergency department after beginning care," said Tami Mark, a distinguished fellow in behavioral health financing and quality measurement at RTI. "This measure provides a way to identify those events and compare outcomes across programs, creating opportunities to better understand where additional support may be needed."

The research was supported by the National Institutes of Health through the HEAL Initiative under award 1RM1DA059375, part of a broader federal push to build quality measurement into addiction treatment the way it was built into surgery and cardiology decades ago.

What a 5 percent rate cannot tell you

A claims-based rate cannot distinguish a program that discharges people too early from a program that serves a county with no psychiatric beds, no sober housing and a two-hour wait for a crisis line. A hospitalization for alcohol withdrawal may signal a clinical gap, or it may signal that the patient had nowhere else to sleep. The measure also excludes Medicare-Medicaid dual eligibles and anyone receiving buprenorphine in a primary care office rather than a certified program, so it describes a slice of the treatment system, not the whole of it.

For patients, the practical value is prospective rather than immediate: it gives state Medicaid agencies and county commissioners a number they can put next to the retention figure when they decide which programs to fund and which to question. People looking for treatment now can reach the SAMHSA National Helpline at 1-800-662-4357 or search medication-assisted treatment programs and opioid treatment programs by location. The measure published this week is a tool for the agencies that pay for that care; whether any of them adopt it is a decision still to be made.