Description of Implementing a Mail-based Overdose Education and Naloxone Distribution Program in Community Supervision Settings during COVID-19

Abstract

Introduction:

This study uses the Exploration, Preparation, Implementation, and Sustainment (EPIS) model to retrospectively describe the mail-based overdose education and naloxone distribution (OEND) program developed in collaboration with the Kentucky Department of Corrections (DOC) for use in the HEALing Communities Study in Kentucky (HCS-KY) and details the reach of this innovative delivery model.

Methods:

HCS-KY is a community-engaged cluster-randomized trial assessing the effects of implementing evidence-based practices, including OEND, on overdose death reduction across 16 communities highly impacted by the opioid epidemic in Kentucky.⁶ The study launch coincided with the COVID-19 pandemic. All coalitions in the 16 HCS-KY counties selected OEND implementation in community supervision offices; however, pandemic limitations on in-person reporting made face-to-face OEND unfeasible. This study uses the EPIS phases to understand how the unique inner and outer contextual factors of the pandemic drove innovation, including five implementation strategies to promote the mail-based OEND program. Internal study management trackers data measured implementation reach.

Results:

Implementation occurred in all 16 counties. All promotional strategies used in the first 8 counties (Wave 1) were carried over to the second 8 counties (Wave 2), except letters were not sent to community supervision clients in Wave 2 counties. Across both waves, 1,759 people accessed the Typeform^™ website to receive overdose education, complete a brief demographic survey, and 1,696 had naloxone shipped to their homes. Greater reach occurred in Wave 1 and in rural counties. Of the participants, 81.13% were white, 61.17% were female, 51.79% were between the ages of 35-54, 18.82% had previously experienced an overdose, and 69.07% had witnessed an overdose. Sites sustained three of the five implementation strategies for publicizing the OEND website at the study’s end but not letters and texting.

Conclusions:

Mail-based OEND programs are an appropriate delivery method for ensuring access to life-saving medication for people on community supervision and may encourage treatment. Strategies to promote the OEND program that were high-effort for agency and study staff, such as letters, or high-cost, such as texting, were not sustainable. Implications for OEND best practices, including innovative technology use within community supervision settings are addressed.

1. Introduction

Drug overdose is the leading cause of death among people who were formerly incarcerated (Binswanger et al., 2013; Fiscella et al., 2020). Opioids play a prominent role in post-release drug overdose deaths – prior research has shown that opioid overdose constituted the majority of post-release drug overdose deaths, with some studies showing opioids as responsible for more than half of overdose deaths among individuals involved in the criminal legal system (Alex et al., 2017; Binswanger et al., 2013, 2020). Furthermore, evidence has shown that individuals on parole had the highest overdose mortality rates compared to those in prison and in the community, and individuals were at higher risk for opioid overdose death during the periods of transition from prison and from community supervision (Binswanger et al., 2020). These risk factors highlight the need for overdose education and naloxone distribution (OEND) within community supervision settings, which includes probation (i.e., an alternative to incarceration) and parole (i.e., early release from incarceration), in which a person must comply with court-ordered or Parole Board-ordered stipulations of supervision under a probation or parole (P&P) officer. The COVID-19 pandemic increased the number of people on community supervision (Aslim & Mungan, 2020), and the onset of COVID-19 coincided with an increase in overdose deaths (Imtiaz et al., 2021). OEND is an evidence-based practice (EBP) that reduces opioid-related mortality (Razaghizad et al., 2021) but is not widely available to people within community supervision settings (Grella et al., 2021). Emerging evidence also suggests that intervention during overdose events may increase interest in treatment (Jones et al., 2021; Pollini et al., 2006), suggesting that OEND programs present an opportunity for linkage and referral to treatment. Research on improving the implementation of EBPs in real-world environments, especially criminal legal system (CLS) settings, is lacking (Proctor et al., 2023). The purpose of this paper is to contribute evidence to this understudied area by describing Kentucky’s HEALing (Helping to End Addiction Long Term^SM) Communities Study implementation process along the EPIS (Exploration, Preparation, Implementation, and Sustainment) framework (Aarons et al., 2011; Moullin et al., 2019) for implementing a mail-based OEND program in community supervision settings beginning during COVID-19.

1.1. Lack of OEND programs in criminal legal settings

Prior research has evaluated programming in prisons and jails that provide OEND to individuals during the release process. For example, a study of the San Francisco County Jail OEND program found that almost a third of individuals who received naloxone upon release reported using it to reverse an overdose, and almost half of the participants obtained a refill in the community (Wenger et al., 2019). The volume of naloxone used and refilled by participants speaks to the urgent need and acceptability of providing OEND resources among the reentry population. An evaluation of Scotland’s National Naloxone Programme found that distributing naloxone to incarcerated persons upon release reduced the proportion of opioid-related deaths by 22% when compared to the years before the program was initiated (Bird et al., 2016).

Despite the evidence for OEND in CLS settings and more broadly (Razaghizad et al., 2021), it typically is not offered to clients of United States (U.S.) CLS agencies (Grella et al., 2021; Scott et al., 2022). The largest CLS setting, community supervision offices, is ideal for OEND programs considering its breadth of reach and the heightened risk of overdose faced by individuals under community supervision (Binswanger et al., 2007, 2020; Bureau of Justice Statistics, 2018). While previous research has examined the implementation of medication for opioid use disorder and contingency management programs in community supervision settings (Drazdowski et al., 2024; Kang et al., 2024; Martin et al., 2021), there is limited research on OEND programs despite the potential reach of this venue (Brinkley-Rubinstein et al., 2018). In a scoping review by Grella and colleagues (2021) of factors that influence opioid overdose prevention for CLS-involved populations, only one paper on community supervision settings was included which focused on individual-level characteristics associated with willingness to receive OEND (see Cropsey et al., 2013). This highlights the OEND gap in community supervision settings. The present study addresses this gap in the literature by describing efforts to implement mail-based OEND programs in Kentucky community supervision offices. Notably, Kentucky ranks 9^th among U.S. states with the highest number of individuals under community supervision per 100,000 adults in 2021 (Kaeble, 2023), ranks fourth-highest in fatal drug overdoses (Centers for Disease Control and Prevention, 2023), and was one of the states with the largest increase in opioid-related fatalities during COVID-19 (Friedman & Akre, 2021).

1.2. Mail-based OEND programs are an effective, confidential modality to reach high-risk populations such as people on community supervision, especially during COVID-19

Mail-based OEND programs provide naloxone to participants upon completion of an online course (Sisson et al., 2023; Yang et al., 2021). These programs are designed to overcome challenges to distributing harm reduction materials and may be ideal for implementation in CLS settings. Further, even though mail-based OEND programs existed prior to COVID-19, they may address barriers during COVID-19, when prisons and jails released individuals deemed lower-risk to community supervision to reduce disease spread (Aslim & Mungan, 2020; Dir et al., 2022; Marcum, 2020; Schwalbe & Koetzle, 2021). Community supervision offices had to quickly adapt to increased caseloads and pandemic constraints, resulting in greater use of telehealth and virtual services (Marcum, 2020; Oser, Rockett, et al., 2024; Schwalbe & Koetzle, 2021). However, limiting in-person visits to community supervision offices impeded the ability to deliver services (Marcum, 2020).

Mail-based OEND programs address challenges arising from the COVID-19 pandemic but also circumvent other delivery challenges. For example, mail-based OEND programs overcome geographic barriers to naloxone distribution. Programs such as the Needle EXchange Technology (NEXT) harm reduction program provide naloxone to individuals at risk of overdose across the U.S. by mailing naloxone upon completion of a training video (Sisson et al., 2023; Yang et al., 2021). Increasing access to mail-based OEND is critical in rural areas as they have limited access to emergency medical services that respond to overdoses (Hanson et al., 2020; Mell et al., 2017). Furthermore, mail-based OEND combats stigma from association with the use of harm reduction services, because recipients are able to obtain OEND at their private residence, rather than seen (and potentially stigmatized by others) at more public venues (Barnett et al., 2021; Hayes et al., 2021). These programs bypass misguided community concerns about brick-and-mortar harm reduction programs and feared negative community impacts of expanded harm reduction (Hayes et al., 2021). Mail-based OEND programs in community supervision settings offer anonymity as the service recipient is not recognized while receiving OEND. Centralizing OEND programs via mail-based delivery has the additional benefit of reducing CLS staff burden by eliminating the need to provide OEND on an individual basis. These benefits were the impetus for designing a mail-based OEND program with implementation strategies as part of the HEALing Communities Study in Kentucky (HCS-KY).

1.3. EPIS model

Implementation science frameworks are useful for planning and characterizing the process of implementation to ensure success, including sustainment of an EBP (Aarons et al., 2011; Glisson & Schoenwald, 2005; Metz & Albers, 2014). While there are myriad implementation science models (Tabak et al., 2012), the EPIS framework (Aarons et al., 2011; Moullin et al., 2019) is a holistic, multilevel approach to the strategic implementation of an EBP. EPIS entails a four-phase implementation process, while also considering the inner and outer contexts and bridging and innovation factors that influence implementation. The Exploration phase includes the identification of an EBP that will address an unmet need among clients served by an organization. The Preparation phase involves identifying barriers and factors of implementation at both the inner and outer contextual levels, preparing staff and resources, and devising a plan for implementation that maximizes facilitators and addresses potential barriers. The third phase, Implementation, is the initial use of the EBP within a specific organization and monitoring the implementation process. The Sustainment phase focuses on the continuation of the EBP, which is a dynamic ongoing process that is often overlooked in implementation research despite being key to sustaining positive public health impacts.

This study draws upon the EPIS model (Aarons et al., 2011) post hoc to retrospectively describe the development, implementation process, reach, and sustainment of an innovative mail-based OEND program in collaboration with the Kentucky Department of Corrections (DOC) for use in HCS-KY. It should be noted that the EPIS model was the implementation science framework used in the HCS-KY application, but more details about the model and approach for assessing implementation in the HEALing Communities Study can be found in an article by Knudsen and colleagues (2020). Two research questions are examined. First, what efforts are needed to successfully implement a mail-based OEND program in community supervision settings during COVID-19? Second, what is the reach (defined as units of naloxone distributed per year per 100 people on community supervision) of this mailed-based OEND program in community supervision settings?

2. Methods

2.1. Study context

The HEALing Communities Study is a parallel-group, cluster randomized waitlist-controlled trial that tests the effect of the community-level Communities That HEAL intervention on reducing opioid overdose deaths, among other outcomes, across 67 communities in four states (HCS Consortium, 2020). The Communities That HEAL intervention includes opioid-focused community coalitions working through a multi-phased process to select strategies to promote the uptake of EBPs within the Opioid-overdose Reduction Continuum of Care Approach for implementation in organizations within their communities (Chandler et al., 2023; Sprague Martinez et al., 2020; Winhusen et al., 2020; Young et al., 2022). OEND is one of the primary EBPs within the Opioid-overdose Reduction Continuum of Care Approach, and CLS settings were a priority. The methodology of the HEALing Communities Study is described in detail elsewhere (HCS Consortium, 2020).

This paper focuses on the 16 communities (operationalized as counties) in Kentucky, which comprise more than 40% of the state’s population. The 16 communities were randomized to receive the Communities That HEAL intervention as part of “Wave 1” (i.e., randomized to receive the intervention between January 2020 and June 2022; n=8) or “Wave 2” (i.e., delayed intervention occurring between July 2022 and December 2023, n=8). OEND in community supervision settings was chosen as an EBP from the Opioid-overdose Reduction Continuum of Care Approach for implementation by all 16 community coalitions in the 16 HCS-KY counties.

The HEALing Communities Study launch in Wave 1 communities coincided with the beginning of the COVID-19 pandemic in early 2020. The HCS-KY site used a “hub with many spokes” strategy, in which the University of Kentucky acted as a central “hub” for naloxone training and distribution to partner agencies (Knudsen et al., 2023; Oyler et al., 2024). For OEND implementation where in-person delivery was feasible, implementation facilitators (trained in implementation science methods) worked to recruit partner organizations (“many spokes”) to deliver OEND to their clients. Evaluation of HCS-KY Wave 1 data found that 69% of organizations contacted for OEND partnership accepted at least one shipment of naloxone (Knudsen et al., 2023), and CLS agencies (e.g., jails, drug courts) made up 10% of community partners – second only to behavioral health agencies (Oyler et al., 2024). Given the constraints of the COVID-19 pandemic, an alternative model was needed for agency partners like the Kentucky DOC Division of Probation and Parole (P&P) where in-person OEND was not feasible because these agencies had greatly reduced in-person meetings. Thus, in partnership with the Kentucky DOC, the HCS-KY Naloxone Hub created and managed a mail-based OEND program for individuals on community supervision.

The Kentucky DOC administers the Division of P&P, which provides community supervision, investigative, and reentry services. HCS-KY partnered with 21 community supervision offices that served residents of HCS-KY’s 16 target counties. One large urban county had five community supervision offices, while other community supervision offices were regional and served clients in multiple counties. These offices’ reporting hours varied widely, with some offices seeing clients five days per week and others only two mornings each month. Beginning in March 2020 throughout the height of the COVID-19 pandemic, clients deemed lower-risk reported to their P&P officer via phone to reduce COVID-19 transmission (7% of the supervision cases were assessed as high, 62% moderate, 26% low, and 5% administrative in 2020) (Kentucky Department of Corrections, 2021). DOC approved the resumption of in-person research activities in February 2022. To be eligible for the mail-based OEND program, individuals had to reside or report in one of the 16 HCS-KY target counties.

2.2. Procedures: Buy-in, development, and operations of mail-based OEND program for people on community supervision

Figure 1 illustrates the timeline of implementation activities by EPIS phase. The Exploration phase entailed a meeting in November 2020 with DOC leadership including the Director of P&P, Director of the Division of Addiction Services (which provides clinical and administration oversight for all substance use disorder (SUD) services to people on community supervision and/or incarcerated), and staff from the Office of Research and Legislative Services (which oversees external research activities, data requests, DOC’s state/federal research portfolio, and DOC input on legislative processes). DOC provided preliminary support for the HEALing Communities Study during the grant application/award phases, which included a meeting to overview the study, a letter of support for the grant application, and DOC participation in the NIDA site visit to determine if Kentucky would receive an award. The November 2020 meeting sought approval specifically for implementing OEND in community supervision settings. HCS-KY faculty and staff provided a brief overview of the study, discussed the importance of OEND including the evidence base, and shared that all Kentucky community coalitions selected community supervision offices as a venue for OEND. Discussions focused on the unmet need for OEND and internal and external factors, primarily COVID-19, that would influence the implementation process. At the end of this meeting, the DOC agreed to partner with HCS-KY to provide OEND to people on community supervision but shared that in-person OEND was not feasible at that time due to the external factor of COVID-19 pandemic limitations.

Figure 1.

Timeline of Implementation Activities of the Mail-Based OEND Program in Community Supervision Settings by EPIS Phase

The Preparation phase occurred in partnership with the DOC to co-design an approach for virtually providing OEND to people via community supervision offices. Four additional meetings led by implementation facilitators using a structured meeting guide were held with DOC leadership members during November and December 2020 to elicit barriers and facilitators to virtual mail-based OEND implementation. The research team reviewed meeting minutes and identified the following as facilitators: willingness to collaborate, understanding of the benefit of OEND for people on community supervision, and the existence of virtual strategies to promote the OEND program (e.g., DOC website and a new supervision application). Barriers included: limited in-person reporting, the need for strategies to ensure all people on community supervision were aware of the OEND program, and limited community supervision staff resources to implement an OEND program. Based on these factors, the HCS-KY research team and DOC leadership collaboratively identified five implementation strategies, defined as systematic processes to adopt and integrate EBPs into usual care (Powell et al., 2011). These implementation strategies were co-designed to address barriers in promoting the mail-based OEND program by driving individuals to the Typeform^™ website. These included: 1) letters sent to community supervision clients, 2) print materials (e.g., flyers distributed by P&P officers), 3) installation of televisions in community supervision offices playing promotional videos, 4) informational postings on the Kentucky DOC website and social media, and 5) text messages and information sent via the Kentucky DOC Supervision Application. Together, the strategies ensured information was available at every point of contact an individual may have with a Kentucky community supervision office or P&P officer during the pandemic.

The letter-based strategy informed community supervision clients about the OEND program. Prepared stamped, sealed letters with OEND information and a corresponding number of blank address labels were sent to each county office, with the volume determined by current client numbers. Letters included a testimonial shared by a certified Peer Support Specialist working with community supervision clients about the benefits of carrying naloxone. Community supervision office staff then printed and placed address labels before delivering the addressed letters to the U.S. Postal Service. Office staff mailed these materials to all community supervision clients who resided in a Wave 1 county in January 2021, late February/early March 2021, and May 2021.

Print materials included county-specific posters provided to each office to display in the waiting room or adjacent area (see Appendix A, Fig. A.2). Attached to each poster were sticky note ‘coupons’ displaying the OEND website address, QR code, and study information that clients could tear off and take with them. P&P officers were provided with informational OEND palm cards to give to clients during in-office visits. These materials were available to any individual who entered a community supervision office. If an office served clients from multiple counties, these individuals also had access to promotional materials for the OEND program.

To expand the visibility of the OEND program, another implementation strategy was for HCS-KY to use community supervision office televisions (see Appendix A, Fig. A.3). HCS-KY purchased and installed televisions for community supervision waiting rooms that did not already have a television to play a continuous video loop of the HCS-KY overdose education video and communication campaign materials promoting all individuals’ ability to carry life-saving naloxone and challenge stigmatizing beliefs (Stein et al., 2023). A QR code on the TV directed individuals to the Typeform^™ website for the mail-based OEND program. Similar to the print materials strategy, individuals who entered a community supervision office were exposed to these promotions, independent of the county of residence. This included individuals at higher risk required for in-person reporting during the height of COVID-19 but was also used for everyone after the resumption of all in-person reporting. Televisions in eight Wave 1 offices were installed in January 2021 and televisions in four Wave 2 offices were installed between January-April 2023.

In addition, the Kentucky DOC posted language on its website and social media accounts starting in October 2021 to advertise the availability of free naloxone and instruct clients on how to request naloxone to be confidentially shipped to their home by the HCS-KY Naloxone Hub (see Appendix A, Fig. A.4). Individuals accessed the Typeform^™ website, which was linked to DOC’s website, by selecting the county in which they reside or in which their P&P office is located. The website noted that information on who participated in the training and/or received naloxone would not be shared with DOC or any other individuals/agencies.

Finally, a text message strategy was implemented via the Kentucky DOC Supervision application (see Appendix A, Fig. A.5). At the beginning of each month for the duration of the intervention, a text message with a brief description and the OEND website link was sent to all individuals residing in a HCS-KY county who had a cell phone number on file within the Kentucky Offender Management System. The Supervision application was an opt-in strategy that allowed people on community supervision to have access to resources and connect virtually with their P&P officer in lieu of in-person reporting if they were a low-risk or administrative-level client. The monthly texting began in April 2021 and continued through December 2022 for Wave 1 counties (even though the Wave 1 intervention period ended June 2022). For Wave 2 counties, in preparation for implementation, the program sent texts in mid-December 2022, then bi-monthly for January and February 2023, and once per month thereafter between March and December 2023.

DOC leadership suggested a series of preparatory meetings with district supervisors and local community supervision staff for on-boarding offices to the mail-based OEND program. HCS-KY faculty created standard operating procedures, including a PowerPoint presentation, for the mail-based OEND program in community supervision offices. An HCS-KY implementation facilitator used these resources to lead implementation preparation meetings with local county community supervision leadership to garner additional “buy-in” at the district/county level, as P&P officers are in a position to encourage clients to take advantage of the OEND program. The HCS-KY implementation facilitator served as the point-of-contact for the P&P district supervisors and assisted with coordinating the implementation process, trouble-shooting, and sharing resources.

This phase was supported by the HCS-KY Naloxone Hub (Knudsen et al., 2023; Oyler et al., 2024) in creating a website for overdose training as well as collaborating with the DOC in formulating strategies to direct people to the website. Naloxone distribution relied upon the HCS-KY Naloxone Hub (Knudsen et al., 2023; Oyler et al., 2024), which was responsible for ordering, labeling, and shipping naloxone to people who completed the online overdose training.

The program created a survey-based Typeform^™ website to provide an overdose education video produced by the HCS-KY team and to collect study-related information. Respondents watched a 9-minute video covering overdose recognition, prevention, and response, including intranasal naloxone administration, and answered three knowledge-based questions regarding the content of the video. If respondents selected incorrect responses to the knowledge-based questions, the website displayed the correct answer. See Appendix A, Fig. A.1 for an example of a knowledge question and the overdose education video. Respondents provided basic demographic information, affiliated community supervision office/county of residence, and overdose history. “Prefer not to say” was a response option for all demographic questions. Respondents also provided their name and mailing address to allow naloxone, along with an educational brochure and county-specific SUD treatment and harm reduction resource guide, to be shipped directly to their home. An honest broker managed the Typeform^™ account and all related data. The broker extracted identifying information and sent it separately to the HCS-KY Naloxone Hub, where staff mailed shipments to respondents weekly. The honest broker imported study-related information (e.g., units mailed, demographics) directly into the project database (Harris et al., 2009, 2019). No study members had access to the Typeform^™ site after its initial building, and the program shared no data with DOC.

2.3. Data collection, measures, and analysis

After watching the overdose education video, the program asked participants to complete a brief online survey. Variables were modeled after an online OEND training resource (NDRI USA, 2019) and included sex, age category, race, and history of overdose (both personal and witnessed). Individuals could skip any questions and the Research Electronic Data Capture platform collected the data (Harris et al., 2009, 2019). Advarra Inc., the HEALing Communities Study single Institutional Review Board, approved this data collection as a site-specific modification (Pro00038088).

Descriptive statistics characterize the demographic characteristics of individuals on community supervision who received OEND. The Kentucky DOC provided county-level supervision numbers at the start of OEND program implementation to calculate reach, meaning the rate of OEND among 100 clients on community supervision per county per year. The Implementation phase of the mail-based OEND program occurred across 24 months for the 8 Wave 1 counties and across 12 months for the 8 Wave 2 counties. To calculate the reach rate, the study divided the number of naloxone units distributed by the number of people on supervision per county. For Wave 1 counties, this number was divided by two since the implementation phase was twice as long as in Wave 1 counties.

3. Results

Figure 1 displays the timeline of implementation activities of the mail-based OEND program by EPIS phase. Four of the five implementation strategies used in Wave 1 were carried over to Wave 2. Specifically, letters were not sent to community supervision clients in Wave 2 counties, as this strategy entailed a time-intensive process for community supervision staff, and there was less need for this strategy as in-person reporting at community supervision offices had resumed.

Table 1 displays the number of units of naloxone distributed by the mail-based OEND program, the number of people on supervision, and the units distributed per year per 100 people on community supervision in each county. Across both waves, 1,696 people had naloxone shipped to their home address. This reach equates to approximately 7.38% of the people on community supervision in the 16 counties. There were differences between Wave 1 and 2 naloxone distribution rates, with a higher rate in Wave 1 counties. Approximately 6.63 naloxone units per year per 100 people on community supervision were distributed in Wave 1 counties as compared to 6.02 naloxone units in Wave 2 counties. Moreover, across the 16 counties, naloxone distribution was higher in rural counties compared with urban counties (8.05 units in rural counties, as compared to 4.98 units per year per 100 people on community supervision in urban counties).

Table 1.

Naloxone Distribution by County for the Mail-based OEND Program in Community Supervision Settings

County	Wave	Rural or Urban Status	Total Units Distributed	# of People on Community Supervision	Units Distributed Per Year Per 100 People on Community Supervision
County A	1	Rural	90	856	5.26
County B	1	Urban	28	322	4.35
County C	1	Urban	116	631	9.19
County D	1	Urban	379	3,015	6.29
County E	1	Urban	233	2,098	5.55
County F	1	Rural	43	235	9.15
County G	1	Rural	50	377	6.63
County H	1	Rural	101	759	6.65
County I	2	Urban	34	2,525	1.35
County J	2	Urban	12	227	5.29
County K	2	Rural	17	194	8.76
County L	2	Rural	29	364	7.97
County M	2	Urban	5	238	2.10
County N	2	Urban	498	10,442	4.77
County O	2	Urban	22	370	5.95
County P	2	Rural	39	326	11.96
TOTAL			1,696	22,979	7.38

Table 2 displays the demographic characteristics of 1,759 people who accessed the mail-based OEND program through community supervision offices. There were 63 people who did not receive naloxone as they did not enter an address or entered an invalid address. Respondents were primarily female (61.17%) and white (81.13%), and more than half were in the 35-54 age range (51.79%). In comparison, people in the U.S. on community supervision are less likely to be female (25% of the probation and 12% of the parole populations) and white (54% of the probation and 44% of the parole populations) (Kaeble, 2023). Almost one in five individuals reported having experienced an overdose (18.82%), and more than two-thirds (69.07%) had witnessed an overdose in their lifetime. It is worth noting that more people accessing the mail-based OEND program in Wave 1 reported experiencing a prior overdose (22.78%) compared with people in Wave 2 (12.10%).

Table 2.

Self-Reported Demographics of Individuals who Participated in the Mail-based OEND Program in Community Supervision Settings by HEALing Communities Study Wave

	Wave 1 (n=1,106)	Wave 2 (n=653)	Total (N=1,759)
Sex
Female	637 (57.59%)	439 (67.23%)	1076 (61.17%)
Male	435 (39.33%)	184 (28.18%)	619 (35.19%)
Other	0	3 (0.46%)	3 (0.17%)
Prefer not to say/missing	34 (3.07%)	27 (4.13%)	61 (3.47%)
Age Category
34 and younger	364 (32.91%)	209 (32.01%)	573 (32.58%)
35-54	580 (52.44%)	331 (50.69%)	911 (51.79%)
55 and older	135 (12.21%)	94 (14.40%)	229 (13.02%)
Prefer not to say/missing	27 (2.44%)	19 (2.91%)	46 (2.62%)
Race
African American/Black	81 (7.32%)	76 (11.64%)	157 (8.93%)
White	936 (84.63%)	491 (75.19%)	1,427 (81.13%)
Asian	9 (0.81%)	5 (0.77%)	14 (0.80%)
American Indian/Alaska Native	3 (0.27%)	2 (0.31%)	5 (0.28%)
Native Hawaiian or Other Pacific Islander	6 (0.54%)	2 (0.31%)	8 (0.45%)
Other†	31 (2.80%)	30 (4.59%)	61 (3.47%)
Prefer not to say/missing	40 (3.62%)	47 (7.20%)	87 (4.95%)
Prior Overdose
Yes	252 (22.78%)	79 (12.10%)	331 (18.82%)
No	816 (73.69%)	542 (83.00%)	1,358 (77.14%)
Prefer not to say/not sure/missing	38 (3.44%)	32 (4.90%)	70 (3.98%)
Witnessed Overdose
Yes	768 (69.44%)	447 (68.45%)	1215 (69.07%)
No	282 (25.41%)	163 (25.00%)	445 (25.24%)
Prefer not to say/not sure/missing	56 (5.06%)	43 (6.59%)	99 (5.63%)

^†Includes multi-race in which respondents selected more than one race, multi-race-ethnicity, and/or other

In the Sustainability phase following the intervention period, which ended in December 2022 for the 8 Wave 1 counties and December 2023 for the 8 Wave 2 counites, management of the OEND website transitioned to the Kentucky Pharmacists Association, with the naloxone supplied through state funding (see Knudsen et al., 2023 for details). As part of the sustainment of the mail-based OEND program, three of the five implementation strategies continued: print materials in community supervision offices, promotional videos on televisions in community supervision offices, and information on how to access the mail-based program posted on the Kentucky DOC website. The letter strategy was discontinued due to the burden on DOC staff time and the text message strategy was not sustained due to cost. During the Sustainability phase, between January 2023 and April 2024 for Wave 1 counties and between January 2024 and April 2024 for Wave 2 counties, the Kentucky Pharmacists Association has distributed 610 units of naloxone to people on community supervision within the 16 counties (detailed distribution data by county per month is not available).

4. Discussion

People on community supervision, especially those recently released from prison and jail, are at elevated risk for return to drug use and fatal overdoses (Binswanger et al., 2013; Fiscella et al., 2020). During the height of the COVID-19 pandemic, rapid early releases occurred nationwide for people incarcerated for non-violent offenses to reduce COVID-19 transmission, which may have left little time for SUD treatment linkage and service delivery and harm reduction efforts (Dir et al., 2022; Marcum, 2020). In Kentucky, the SAMHSA-funded state opioid response effort provided 1,931 naloxone units with written training materials to people in jails or prisons who were receiving sentence commutations or were on home incarceration between April and August 2020 (K. Marks, personal communication, March 29, 2024). People who were incarcerated were also provided a treatment/re-entry resource packet including information for dedicated staff to help link people to resources and presumptive eligibility with Medicaid cards mailed to the individual’s home placement address. However, naloxone distribution occurred without active overdose education and did not reach the largest segment of people in the CLS – people on community supervision. EBPs such as mail-based OEND help mitigate the risk of fatal overdose; this was particularly important during the study as overdoses were rising due to the COVID-19 pandemic and fentanyl was expanding in the illicit market (Friedman & Akre, 2021). Moreover, non-fatal overdose experiences may serve as an impetus for seeking SUD treatment services (Jones et al., 2021; Pollini et al., 2006), and the mail-based OEND strategy included the provision of county-specific resource guides with information on where to access medication for opioid use disorder. Despite the challenges of the pandemic, a model of mail-based OEND was successfully implemented in 16 Kentucky counties. These findings may inform mail-based OEND program implementation in community supervision settings in other states and/or other CLS settings.

Overall, the Exploration phase revealed DOC leadership’s commitment to the provision of OEND for individuals on community supervision. There was a high degree of readiness to change and acceptability of a mail-based OEND program, which is key for implementation (Grella et al., 2021). HCS-KY and DOC leadership discussions focused on the need for OEND for people on community supervision when overdose rates were increasing and access to venues that usually provide OEND were limited due to pandemic-related shutdowns (e.g., syringe service programs, recovery community centers). HCS-KY was able to supply financial resources for the OEND program, which addresses a common barrier and can promote readiness to change within an agency. DOC leadership shared considerations surrounding the external factor of COVID-19 that may be a barrier to the OEND program implementation process.

HCS-KY and DOC leadership discussions of internal and external factors influencing the implementation process continued into the Preparation phase. Limited community supervision staff time, public health safety precautions, and a reduction of in-person reporting during COVID-19 led to the design of the mail-based OEND program. Although take-home naloxone programs in which overdose education occurs face-to-face and naloxone is immediately distributed have been implemented in correctional facilities prior to release pre-COVID-19 (Parmar et al., 2017; Pearce et al., 2019; Sondhi et al., 2016; Wenger et al., 2019; Zucker et al., 2015) as well as in local jails in partnership with HCS-KY during COVID-19 (Oser, McGladrey, et al., 2024), it was not possible in community supervision offices during the pandemic. Thus, a passive, opt-in strategy of a mail-based OEND program allowed people on community supervision to access this life-saving medication in their homes during the height of social distancing.

The EPIS framework states that activities occurring during the Exploration and Preparation phases influence the effectiveness of implementation strategies (Aarons et al., 2011; Moullin et al., 2019). Due to the partnership with DOC, the mail-based OEND program reached almost one in ten people reporting to community supervision offices during the Implementation phase without any active in-person promotion. This suggests that mail-based OEND programs are a helpful option for people on community supervision, especially during COVID-19, but reach could be increased if supplemented with an in-person OEND program.

It is noteworthy that people reached by the mail-based OEND program were more likely to be female and white as compared to the U.S. community supervision population; and, these trends align with other research in a population of people who use drugs (Khan et al., 2023). The HCS overdose education video was designed in an expedited fashion by HCS faculty and staff for immediate use during the COVID-19 pandemic; thus, people with lived experience were not consulted during the creation. The video featured two white women; however, additional materials displayed on the TVs in the community supervision office (e.g., communication campaign materials) featured a more racially and ethnically diverse population that were jointly designed with community coalitions and included images of local residents. Future initiatives should prioritize co-designing strategies and materials for a mail-based OEND program with diverse populations of people with lived experience. In addition, demographic data was shared with community supervision offices only in Wave 2, which may explain why more individuals who participated in the mail-based OEND program were non-white in Wave 2 (25%) as compared to Wave 1 (15%). It’s highly concerning that more than two-thirds had witnessed an overdose event. Naloxone access is crucial for people on community supervision because they are at elevated risk for an overdose and have high-risk social networks (Binswanger et al., 2013; Fiscella et al., 2020). Moreover, additional efforts are needed to address racial disparities as rates of drug overdose deaths among non-Hispanic Black Kentucky residents are increasing (Slavova et al., 2023).

Interestingly, more people were reached via the mail-based OEND program in Wave 1 counties as compared to Wave 2 counties, which is likely due to several factors. First, the urgency of COVID-19-related early releases of people from jails/prisons occurred during the Wave 1 intervention period. Due to state-wide shutdowns, people were less likely to be able to access naloxone in-person at local community-based agencies. Mail-based OEND programs directly addressed this need by virtually providing overdose education and shipping naloxone to individuals’ homes. Second, Wave 1 coincided with the onset of fentanyl, which was prominent in the illicit drug supply (Friedman & Akre, 2021; Irvine et al., 2022). Kentucky, alongside West Virginia and Tennessee, had the highest per capita overdose death rates at the beginning of COVID-19 (Friedman & Akre, 2021), highlighting the critical need for naloxone. The spike in overdoses associated with COVID-19 underscores the importance of mail-based and digital strategies for reaching groups at high-risk for overdose like individuals on community supervision in preparation and training for future pandemic responses. Third, HCS-KY hired, trained, and placed prevention specialists who directly delivered OEND in public and community settings, including community supervision offices once DOC approved the resumption of in-person research activities. At the end of Wave 1, prevention specialists were able to enter community supervision offices and provide in-person OEND services to 461 people on community supervision. This in-person strategy was more robust in Wave 2 as COVID-19 social distancing policies changed, resulting in 1,777 people receiving in-person OEND services in community supervision offices during the Wave 2 intervention period. The results in this paper only present the data on the number of naloxone units distributed via the OEND mail-based program.

Reach data collected during the Implementation phase also revealed that more people in rural counties partook in the mail-based OEND program than their urban counterparts. Rural individuals face transportation barriers to accessing programs and services (Oser, Rockett, et al., 2024); thus, a mail-based OEND program is ideal. In this study, technology was not a barrier to the online overdose education provided before ordering mail-based naloxone, which may be attributable to significantly increased access to digital technologies such as broadband and smartphones among rural Americans during the past decade (Vogels, 2021). Stigma may also be a deterrent to participation in face-to-face, direct-delivery OEND programs offered in CLS settings, especially in rural areas where there is less anonymity (Barnett et al., 2021; Hayes et al., 2021). In fact, prior HEALing Communities Study research across the 67 sites demonstrates higher perceived community stigma toward naloxone in rural than urban areas (Davis et al., 2024). The confidential distribution of naloxone via a mail-based program may have been more attractive to individuals on community supervision in less densely populated areas.

It is also worth noting there were disparities in the demographic characteristics of respondents participating in the mail-based OEND program, with a greater proportion of respondents being female and white compared to the U.S. community supervision population. Efforts are needed to promote greater reach across gender, racial, and ethnic lines for people on community supervision, especially when rates of opioid overdose events are increasing in non-Hispanic Black residents in Kentucky (Slavova et al., 2023) and across the four HEALing Communities Study states (Larochelle et al., 2021). Potential solutions to address inequities in the mail-based OEND program include targeting the implementation strategies to encourage more participation from diverse groups, such as revising print materials to draw in a more diverse set of respondents or creating an overdose education video that includes representation of persons from various genders, racial, and/or ethnic groups.

HCS-KY faculty and staff shared aggregate data with DOC leadership on the number of people who participated in the mail-based OEND program and discussed ideas for a sustainment action plan. Cost is often a barrier to sustainment (Grella et al., 2021); however, through the SAMHSA-funded state opioid response effort, the state supplied financial support to cover the cost of the naloxone, shipping, and personnel after the study intervention period. Specifically, the management of the OEND Typeform^™ website transitioned to the Kentucky Pharmacists Association, which has previously partnered with both the HCS-KY and the DOC. In the Sustainment phase, text messaging via the DOC supervision application was discontinued, as there was a cost associated with texting as well as a regulatory change in that the text messages are now classified as a solicitation and subject to fines. However, individuals are still driven to the Typeform^™ website through print materials and promotional videos played on televisions in community supervision offices and informational postings on the Kentucky DOC website. One strategy (i.e., individualized letters to people on community supervision) was not used during Wave 2 due to limited feasibility and appropriateness (i.e., resumption of in-person reporting) and was not sustained at the end of the study. This strategy produced an additional burden on community supervision administrative staff, which is a documented barrier to implementation and sustainment (Grella et al., 2021). While reach decreased during the Sustainment phase, the mail-based OEND program is still having an impact with minimal P&P and Division of Addiction Services staff burden and offering needed training on how to respond to an overdose event and obtain free naloxone for people on community supervision. This highly feasible, low-burden approach to passive OEND delivery in a specific CLS setting complements expanded direct and passive delivery of OEND by other community partner agencies.

Since implementation of the mail-based OEND program, naloxone nasal spray was approved for over-the-counter (nonprescription) use (U.S. Food and Drug Administration, 2023) and should be more easily accessible to those at risk for overdose. How OTC status will ultimately impact access to naloxone is not clear as concerns over affordability and barriers at the pharmacy have been expressed (Hetrick, 2024; Pérez-Figueroa et al., 2023). As such, continued provision of OEND through harm reduction programs is warranted to ensure access to this lifesaving EBP, especially in community supervision settings which reach people at-risk for overdose.

4.1. Limitations

This study is descriptive with limited measures during the implementation and sustainment periods and does not test the statistical significance of the effect of mail-based OEND programs in community supervision settings on overdose outcomes. Also, due to the public health threat of COVID-19, people on supervision were not consulted on implementation strategies which may have resulted in increased OEND reach, especially to minoritized racial/ethnic populations and men, which should be considered in future research. Additionally, we cannot contextualize the demographics of people reached by the mail-based OEND program relative to those of the community supervision site (e.g., % racial/ethnic minority, % female), because these data are not publicly available. The study also used self-reported OEND data, and individuals were able to skip questions on the demographic survey. It is possible that an individual may have completed the training and demographic survey more than once. Individuals were not asked to provide information on the strategy that led them to the website as strategies varied based on implementation setting (e.g., community supervision, emergency departments, syringe service programs, etc.); thus, we cannot determine which strategy was the most fruitful. Further, people with unstable housing and/or those lacking internet access may be missed with a mail-based OEND program; however other study efforts did specifically focus on reaching unhoused populations. Implementation occurred only in 16 Kentucky counties, so results may not be generalizable to OEND implementation efforts in other states; however, these 16 counties covered over 40% of Kentucky’s population.

5. Conclusion

Mail-based OEND programs are a feasible, novel delivery method for ensuring access to a life-saving medication for people on community supervision. Ideally, mail-based OEND programs complement, rather than replace, in-person OEND delivery by offering a confidential option to access naloxone with minimal agency effort. In-person OEND programs are more costly but provide the benefits of building rapport and countering stigma, answering questions in real-time, and potentially connecting individuals on community supervision to harm reduction and/or SUD treatment services. Strategies to promote the OEND program that were high-effort for agency and study staff, such as letters, or high-cost, such as texting, were not sustainable, but other low-effort promotional implementation strategies (televisions, flyers, website) and the website/mail-based distribution system were sustained after the intervention period. This study makes a needed contribution to the harm reduction, treatment, and implementation science literatures by mapping out implementation strategies along the EPIS framework that other community supervision agencies can use in planning and practice to better reach those at high risk for overdose upon release from incarceration and on probation. County- and state-level policymakers may consider mail-based OEND strategies when planning across CLS sectors to complement direct-delivery approaches to agency-based overdose prevention efforts.

Highlights:

• Mail-based OEND program overcame COVID-19 barriers in community supervision offices

• Mail-based OEND program had greater reach during the height of COVID-19

• Greater mail-based OEND program reach in rural community supervision settings

• EPIS model useful for explaining mail-based OEND programs in community supervision

Abbreviations:

CLS

Criminal legal system

DOC

Department of Corrections

EBP

Evidence-based practice

EPIS

Exploration, preparation, implementation, and sustainment

HCS-KY

HEALing Communities Study Kentucky site

OEND

Overdose education and naloxone distribution

P&P

Probation and Parole

SUD

Substance use disorders

U.S.

United States