Since the year 2000, the opioid-related mortality rate in Oregon has tripled. The increase of opiate and heroin use has brought with it the increase of Hepatitis C and other medical burdens. Substance use disorders (SUDs) are associated with higher rates of hospitalization and can contribute to poor medical and behavioral health outcomes. Historically, people with severe substance use disorders have not responded well to medical interventions, don’t access programs, or drop out of the programs.
“But once I was determined enough to tell myself, ‘I’m going to do it,’ it wasn’t hard anymore" -Program Participant
The team helped a large percentage of people with substance use disorder to participate in treatment in the hospital. They observed that many people who began treatment were still participating 90 days after they were discharged.
Nearly half (46.4%) of the hospital patients who engaged in treatment services were still engaged in those services 90 days post-discharge. This is a key outcome metric, well correlated with other, more difficult to obtain measures such as drinking/using days, employment status, and health status. While 46.4% is lower than our desired metric of a 50% retention rate, it is a surprisingly high rate of retention given the severity of the substance use disorder and the medical complications in this population.
The program was able to reduce the number of emergency department admissions related to SUD by greater than 100 admissions. This greatly exceeded their original goal of reducing emergency department admission down to 2.23% of the identified group.
269 patients were referred to participate in the intervention. In the year prior to the intervention, this group had 702 ED admissions (2.61 admits/patient) and 458 hospital admissions (1.70 admits/patient). The range of admissions was substantial, with most patients having had one or zero admissions the previous year, while a minority had many admissions. The highest readmission rate was a patient with 29 ED admissions and eight hospital admissions in the past year.
The two intervention groups, Consult/Yes-Treatment, and Consult/No-Treatment, were roughly comparable groups. The Yes-to-Treatment group had a 35% higher rate of ED admissions than the No-to-Treatment group, potentially showing a higher rate of distress in that group, or potentially a higher rate of help-seeking behavior that was able to be redirected. The No-to-Treatment group had the majority of the Cannabis Use Disorder patients, reflecting a common pattern seen in community-based SUD treatment, where people with Cannabis Use Disorders are more resistant to addressing their substance use disorder than other groups.
Both interventions, the Consult/Yes-Treatment and the Consult/No-Treatment have shown substantial reductions in readmissions rates. The Consult/Yes-Treatment cohort showed a slightly greater rate of decline in the readmission rate than the Consult/No-Treatment cohort. Our preliminary results are that providing SUD consultations to patients on medical floors appears to be strongly correlated to reduced readmission rates. It is more difficult to interpret the results dependent on whether a patient admitted directly into SUD treatment or not. We do not know the many factors which may lead one of these patients to be ready for treatment at the moment they were approached or not. We may want to get a picture of the total disease burden these patients have and how that may have shaped the decisions they made. The hospital may want to explore using LACE+ scores or other risk adjustment scores to gain insight into the decision-making process these patients are making. We may also want to look at using a Readiness-to-change scale in the interview to better understand the choices the patients are making.
87% of all identified patients were willing to talk with the SUD staff. Patient acceptance of this level of engagement was quite high and well appreciated. 53% of all identified patients got two follow-up contacts from a recovery mentor after discharge from the hospital.
In the first year of operation, hospital staff identified 269 patients that met the criteria for a consult. For analysis purposes, these patients were divided into four categories:
(a) Accepted a consult and subsequently entered treatment
(b) Accepted a consult and subsequently refused treatment
(c) Refused an initial consult
(d) Died.
(a) Accepted a consult and subsequently entered treatment
(b) Accepted a consult and subsequently refused treatment
(c) Refused an initial consult
(d) Died.
There were five patients (2%) who died. All five patients had accepted a consult, but refused to enter treatment. Four of the five had a primary Alcohol Use Disorder and one had a primary Opioid Use Disorder. Given the referred patients’ severe medical complications and their poor health status upon admission, this is a very low number.
35 patients (13%) refused a consult with the SUD team and we were not able to collect any further information on them related to their substance use.
132 patients (49%) accepted a consult and many talked at length with the SUD staff but did not subsequently enter SUD treatment. These patients can be viewed as having received a brief intervention related to their substance use. Notably, of the nine patients that were identified as having a Cannabis Use Disorder, only one accepted referral into treatment, a far lower rate than for people with alcohol, methamphetamine, or opioid use disorders. This correlates with the experiences of community-based SUD treatment providers, who find that cannabis users are the most resistant to examining the consequences of their drug use on their lives.
97 patients (36%) accepted a consult and subsequently entered treatment services upon discharge from the hospital. Of these 97 patients, 69 (71.1%) entered outpatient services upon discharge, 20 (20.6%) entered residential treatment services, and 5 (5.2%) entered detox services. A key finding is that the project facilitated rapid access into treatment services upon discharge, where previously there had been almost zero treatment access.
Nearly half (46.4%) of the hospital patients who engaged in treatment services were still engaged in those services 90 days post-discharge. This is a key outcome metric, well correlated with other, more difficult to obtain measures such as drinking/using days, employment status, and health status. While 46.4% is lower than our desired metric of a 50% retention rate, it is a surprisingly high rate of retention given the severity of the substance use disorder and the medical complications in this population.
Embedding SUD staff on the medical floors of the hospital is a novel activity and to show a return-on-investment, the intervention must demonstrate a robust correlation with subsequently reduced hospital and ED utilization On the other hand, we should not expect an academically pure outcome in a real-world application. Other cost management decisions organizations routinely make are based on strong correlation and common-sense analysis of causative links. This should be our standard for the analysis of the cost-benefit of this pilot.
As the pilot has progressed, it has become clear that we do not have a useful comparison group. Only 35 patients over the past year have refused to talk to the SUD staff. This group’s readmission rate was lower than the other two groups, but we have little information on those people and do not have permission to track them. We do not know if their substance use pattern was different than the other two groups, or if their life circumstances were different. We can make no inference on the readmission pattern given the paucity of information.
We functionally have two interventions; a brief intervention by a SUD professional without follow-up services and a brief intervention followed by engagement in community-based SUD services. Both interventions appeared to produce clinically significant results, with treatment engagement appearing to be slightly more robust than non-treatment engagement.