Shorter outpatient wait-times for buprenorphine are associated with linkage to care post-hospital discharge

Abstract

Background

Inpatient addiction consult services (ACS) lower barriers to accessing medications for opioid use disorder (MOUD), however not every patient recommended for MOUD links to outpatient care. We hypothesized that fewer days between discharge date and outpatient appointment date was associated with improved linkage to buprenorphine treatment among patients evaluated by an ACS.

Methods

We extracted appointment and demographic data from electronic medical records and conducted retrospective chart review of adults diagnosed with opioid use disorder (OUD) evaluated by an ACS in Boston, MA between July 2015 and August 2017. These patients were initiated on or recommended buprenorphine treatment on discharge and provided follow-up appointment at our hospital post-discharge. Multivariable logistic regression assessed whether arrival to the appointment post-discharge was associated with shorter wait-times (0–1 vs. 2+ days).

Results

In total, 142 patients were included. Among patients who had wait-times of 0–1 day, 63% arrived to their appointment compared to wait-times of 2 or more days (42%). There were no significant differences between groups based on age, gender, distance of residence from the hospital, insurance status, co-occurring alcohol use disorder diagnosis, or discharge with buprenorphine prescription. After adjusting for covariates, patients with 0–1 day of wait-time had 2.6 times the odds of arriving to their appointment [95% CI 1.3–5.5] compared to patients who had 2+ days of wait-time.

Conclusion

For hospitalized patients with OUD evaluated for initiating MOUD, same- and next-day appointments are associated with increased odds of linkage to outpatient MOUD care post-discharge compared to waiting two or more days.

1. Introduction:

Medication for opioid use disorder (MOUD) constitutes first-line, evidence-based therapy for patients with opioid use disorder (OUD) (Leshner and Mancher, 2019). However, there have traditionally been significant barriers to accessing MOUD, including long waitlists, difficult admissions processes, involuntary discharge, and stigma (Richert and Johnson, 2015).

Inpatient addiction consult services reach patients who may not otherwise seek medical treatment and have been demonstrated to lower the barriers to MOUD (Trowbridge et al., 2017; Wakeman et al., 2017). In one study, patients seen by an addiction consult service (ACS) had 2.6 times the odds of engaging in substance use disorder treatment post-discharge compared to controls (Englander et al., 2019a). Notably, however, not every patient initiated on MOUD in the hospital successfully links to outpatient care.

In a previous study, we found the linkage rate to the first appointment post-hospital discharge for methadone to be considerably higher (76%) than buprenorphine (49%) (Trowbridge et al., 2017). This may have been due to the fact that methadone follow-up is next-day, whereas buprenorphine follow-up could take up to 7 days. Another study found a 72% linkage rate for buprenorphine post-hospital discharge within a 6-month period (Liebschutz et al., 2014).

Determining factors associated with improved linkage to care after inpatient discharge could help maintain patients’ connection to evidence-based treatment for their OUD. We describe the process for linking patients from our inpatient ACS to an outpatient low-barrier access addiction clinic for buprenorphine treatment and evaluate the association between arrival to follow-up appointment and wait-times.

We hypothesized that among patients with OUD recommended buprenorphine by an ACS, a reduced number of days between hospital discharge date and outpatient follow-up appointment would lead to improved odds of arrival to an outpatient appointment and thus linkage to buprenorphine treatment.

2. Methods:

2.1. Study Setting and Design

We conducted a retrospective cohort study at Boston Medical Center (BMC), an urban, academic hospital with an ambulatory care center located in Boston, Massachusetts. Approximately 17% of BMC inpatients had a substance use disorder diagnosis on discharge (Walley et al., 2012). The inpatient ACS was formed in July 2015 to provide consultation services for primary medical and surgical teams for patients with opioid and other substance use disorders; patients with OUD were offered MOUD in the hospital and linked to treatment post-discharge (Trowbridge et al., 2017).

There were two low-barrier access addiction clinics providing MOUD (specifically, buprenorphine and naltrexone) to which patients were referred over the course of this study. The first was a once-weekly clinic that facilitated access to substance use disorder treatment for those recently discharged from the hospital (Walley et al., 2015). Patients were referred here until 7/31/2016, after which point they were referred to another low-barrier access addiction clinic, described below.

The second was an opioid urgent care clinic open Monday-Friday. This was one of three urgent care clinics grant-funded by the Massachusetts Department of Public Health to focus on addiction treatment in response to the opioid crisis. This clinic opened on 8/1/2016 and was created to maximize accessibility and lower barriers to care through its presence on the ground floor of BMC, a dedicated addiction nurse specialist, and short-term stabilization. It is further described elsewhere (Roy et al., 2020).

2.2. Population and Timing

The study sample included 1) BMC inpatients over the age of 18 with OUD evaluated by the ACS from July 17, 2015 to August 31, 2017; 2) patients for whom ACS either recommended only or recommended and initiated buprenorphine; and 3) patients who were given an appointment to follow-up in a low-barrier access addiction clinic at BMC on hospital discharge. Only a patient’s first encounter with the BMC ACS was included for analysis. Exclusion criteria included: 1) discharge location to another facility, 2) inability to verify the appointment date, 3) appointment for a reason other than OUD, or 4) patient had a premature discharge prior to appointment creation.

We extracted data from the BMC electronic medical record and performed medical record reviews to verify appropriate variables. All patient records were de-identified.

2.3. Measures: Independent, Dependent, and Covariates

The main outcome was whether the patient arrived or did not arrive to the outpatient appointment at a low-barrier access addition clinic post-hospital discharge. We dichotomized this variable as “arrived” or “not arrived.” “Not arrived” included both no-shows and cancellations, regardless of whether the patient rescheduled.

The primary exposure variable was the number of nights that patients waited between the date of their discharge from the hospital and the date of their follow-up appointment post-discharge. This independent variable was “wait-time,” which we dichotomized into 0–1 day (same- or next-day appointment) and 2+ days based on the distribution of the data and clinical relevance. These categories referred to actual days of the week rather than business days.

We included the following covariates from the electronic medical record: age (in years), gender (male/female), distance of residence from the hospital (in miles), insurance type (Medicaid, Medicare, Commercial, Other), co-occurring alcohol use disorder diagnosis (yes/no), and whether a buprenorphine product was prescribed on discharge (yes/no). Patients prescribed buprenorphine product on discharge were given enough supply to last from day of discharge to day of follow-up appointment. A prescription was provided on discharge in order to maintain patients on treatment, reduce harm, and hopefully improve outpatient engagement.

2.4. Data Analysis

We described frequencies and counts to highlight how demographics and insurance characteristics differed by wait-time. We performed multivariable logistic regression to evaluate the association between wait-time and arriving to the post-discharge follow-up appointment. Odds ratios were calculated to compare the odds of arrival for 0–1 days vs. 2+ days of wait-time. We did not adjust for the day of the week a patient was discharged nor for the clinic to which they were referred, as the day of the week and clinic type were collinear with the primary exposure variable, wait-time. We conducted analyses using SAS software, Version 9.4.

This study was approved by the Boston University Medical Campus Institutional Review Board.

3. Results:

Of 142 patients, 77 (54%) arrived to their outpatient appointment post-discharge. A total of 56 (39%) patients had a wait-time of 0–1 day, while 86 (61%) patients had a wait-time of 2+ days. A majority of patients were insured by Medicaid and were male (see Table 1). The patient groups did not differ significantly in terms of age, distance from the hospital, insurance type, alcohol use disorder diagnosis, or discharge with buprenorphine.

Table 1:

Sample Demographics to a Low Barrier Access Addiction Clinic By Wait-Time

1a. Sample Demographics	0–1 day	2+ days	p-values
	N=56	N=86
Arrived % (n)	63% (35)	42% (36)	0.02
Male % (n)	73% (41)	76% (65)	0.91
Age mean (SD)	43 (11.8)	46 (11.0)	0.09
min, max	20, 64	22, 67
Distance from the hospital, in miles mean (SD)	6.0 (9.9)	5.7 (10.9)	0.66
Insurance % (n)
Medicaid	68% (38)	67% (58)	0.13
Medicare	13% (7)	4% (12)
Commercial	11% (6)	13% (11)
Other (free care)/Unspecified	9% (5)	6% (5)
Alcohol use disorder % (n)	13% (7)	8% (7)	0.39
Discharged with buprenorphine prescription % (n)	75% (42)	79% (68)	0.23

Regarding the proportion of arrivals, 63% of patients with a wait-time of 0–1 days arrived to their appointment compared to 42% of patients with a wait-time of 2 or more days (p=0.02). After adjusting for gender, age, distance of residence from the hospital, insurance type, co-occurring alcohol use disorder diagnosis, and discharge with buprenorphine, patients with a wait-time of 0–1 days had 2.6 times the odds (95% CI 1.3–5.5) of arriving to their outpatient appointment post-discharge from the hospital compared to patients with a wait-time of 2+ days (see Table 2).

Table 2:

Odds of Arrival to a Low Barrier Access Addiction Clinic By Wait-Time

	Arrived vs. Not Arrived	Arrived vs. Not Arrived
	Odds Ratio (95% CI)	Adjusted Odds Ratio (95% CI)
Wait-times, in days
0–1	2.3 (1.2, 4.6)	2.6 (1.3– 5.5)
2+	Ref.	Ref.
Male	1.4 (0.6, 2.9)	1.5 (0.7–3.5)
Age	1.0 (1.0, 1.0)	1.0 (1.0–1.1)
Distance from the hospital, in miles	1.0 (1.0, 1.0)	1.0 (1.0–1.1)
Insurance
Medicaid	0.4 (0.1, 1.2)	0.4 (0.1–1.2)
Medicare	0.3 (0.1, 1.2)	0.3 (0.1–1.0)
Commercial	Ref.	Ref.
Other (free care)/Unspecified	0.3 (0.1, 1.4)	0.2 (0.0–1.3)
Alcohol Use Disorder	1.9 (0.6, 6.0)	1.9 (0.6–6.5)
Discharged with buprenorphine prescription	1.6 (0.7,3.6)	2.2 (0.9–5.7)

4. Discussion:

This retrospective study found that patients seen by an inpatient addiction consult service who received a same-or next-day follow-up appointment post-discharge for buprenorphine treatment were more likely to attend their appointment, compared to patients waiting two or more days.

Buprenorphine treatment is known to reduce overdose risk and overall mortality among people with OUD, as seen in a large, observational study of people who had previously suffered a nonfatal opioid overdose (Larochelle et al., 2018). The population included in this study were individuals with OUD who were out-of-treatment and thus remained at risk of opioid overdose. Consequently, ensuring linkage to treatment with evidence-based medications such as buprenorphine may reduce opioid-related morbidity and mortality. Linkage to treatment is the first step towards retention in care. Among a LatinX population, low-barrier access to buprenorphine was associated with greater odds of retention in treatment at three months (Lee et al., 2019). Retention of patients in MOUD treatment is associated with lower overdose and all-cause mortality (O’Connor et al., 2020).

Ensuring initiation or continuation of buprenorphine post-discharge through better linkage programs may improve acute care utilization as well. Medication for addiction treatment (MAT) has been shown to reduce the hazard of inpatient behavioral health readmission (Reif et al., 2017). However, another study found that while in-hospital initiation of MOUD was a promising method for facilitating treatment linkage post-discharge, MOUD initiation did not reduce medical inpatient readmission rates (Nordeck et al., 2018).

This study adds to others that have explored factors that contribute to initiation of and linkage to substance use disorder treatment post-hospital discharge. A study in Portland, OR found that homelessness, past treatment with methadone maintenance, and partner substance use were associated with initiating treatment on hospital discharge (Englander et al., 2019b). Another study in Boston, MA looked at the role of post-discharge navigation in reducing barriers to engagement post-hospitalization but found no effect (Wakeman et al., 2020). A group in Seattle, WA compared patients admitted with trauma injuries to non-trauma patients for engagement in care for buprenorphine 30 days post-discharge (63.2% v 48.2%, p = 0.16) (Bhatraju et al., 2020).

Our study is consistent with previous reports concerning wait-time within and outside addiction treatment settings. In the same clinic, we studied outpatients seeking addiction care for opioid and alcohol use disorders and found arrival rates of 82% for same-day, 53% for next-day, and 39% for 2 or more days (Roy et al., 2020). Increased wait-time has also been shown to impact the probability of arrival concerning outpatient gastroenterology (Shrestha et al., 2017) and ophthalmology clinics (McMullen and Netland, 2015); wait-times of less than 5 days and 2 weeks, respectively, appeared to have the most benefit. The negative effects of increased wait-time were additionally present concerning arrival for outpatient MRIs, and were most pronounced in vulnerable populations such as those with non-commercial insurance (Daye et al., 2018), similar to the demographics within our study.

4.1. Limitations

There are several limitations to our study. First, this was a retrospective study and we cannot make causal determinations based on our findings. We also had a small sample size and the majority of our population had state-based insurance, thus the results may not be generalizable to larger populations or those with private insurance. Due to limitations of the datasets, we were unable to account for race, ethnicity, and housing status as well, which are known to be associated with addiction treatment utilization (Chatterjee et al., 2017; Wu et al., 2016). We were unable to adjust for the specific day of discharge because the day of the week was collinear with the primary exposure variable (wait-time). Finally, our data set was limited to buprenorphine and did not extend to other MOUD (methadone and naltrexone).

4.2. Future Directions

Based on these findings, increasing clinic availability to provide MOUD could improve continuity of care between the inpatient and outpatient setting by decreasing wait-times for outpatient linkage. Other factors that could be explored include access to transportation and child care and criminal justice involvement. Future areas of evaluation include examining how wait-times are associated with retention in care, hospital readmission rates, and the cost-effectiveness of improving linkage and retention in addiction treatment of these interventions.

5. Conclusion

Patients connected to MOUD after being evaluated by an inpatient ACS were more likely to arrive if the wait-time to the outpatient follow-up appointment was the same- or next-day post-discharge compared to waiting two or more days. Clinics providing MOUD should aim to increase the availability of their services to maximize the continued engagement of patients to evidence-based therapy initiated during inpatient care.

Highlights.

• Addiction consult can improve buprenorphine treatment for hospitalized patients.

• Not every patient started on buprenorphine inpatient links to outpatient treatment.

• Clinics should engage with recently discharged patients with opioid use disorder.