An Economic Analysis of the Cost of Mobile Units for Harm Reduction, Naloxone Distribution, and Medications for Opioid Use Disorder

Abstract

Background & Objective:

Mobile substance use treatment units are effective approaches to increase treatment access and reduce barriers to opioid use disorder (OUD) care. However, little is known about the economic costs of maintaining and operating these units. This study aimed to estimate the economic costs of starting and maintaining mobile units providing harm reduction, overdose education and naloxone distribution (OEND), and medication for opioid use disorder (MOUD).

Methods:

As part of the HEALing Communities Study, four communities in Massachusetts (Bourne/Sandwich, Brockton, Gloucester, Salem) implemented mobile units offering OEND and MOUD (buprenorphine and naltrexone only); each selected different services tailored to their community. All provided MOUD linkage via telehealth, but only one offered in-person MOUD prescribing on the unit. We retrospectively collected detailed resource utilization data from invoices to estimate the direct economic costs from August 2020 through June 2022. Cost components were categorized into start-up and operating costs. We calculated total economic cost over the study period and the average monthly operating cost.

Results:

Implementing a mobile unit offering OEND and MOUD required a one-time median start-up cost of $59,762 (range: $52,062 - $113,671), with 80% of those costs attributed to the vehicle purchase. The median monthly operating cost was $14,464. The largest cost category for all mobile units was personnel costs. The monthly ongoing costs varied by community settings and services: approximately $5,000 for two urban communities offering OEND and MOUD linkage via telehealth (Gloucester, Salem), $28,000 for a rural community (Bourne/Sandwich), and $23,000 for an urban community also providing in-person MOUD prescribing on the unit (Brockton).

Conclusion:

The economic costs of mobile substance use treatment and harm reduction units are substantial but vary by community settings and services offered. Our results provide valuable community-level economic data to stakeholders and policymakers considering establishing and/or expanding mobile units with OEND and MOUD services. Further exploration of cost-effectiveness and efficiency should be considered across different settings.

INTRODUCTION

The United States (US) continues to face an overdose crisis, with more than 105,000 drug-related and 80,000 opioid-related overdose deaths in 2022 (Centers for Disease Control and Prevention (CDC), 2023). The epidemic is heterogeneous across states. From 2016 to 2022, Massachusetts had a 9% increase in the rate of fatal opioid-related overdoses (from 30.7 to 33.5 per 100,000 people) (Massachusetts Department of Public Health, 2023). In contrast, Kentucky reported a rise from 22.9 to 39.3 per 100,000 people over the same time period (72% increase) (Kentucky Injury Prevention and Research Center, 2021, 2023). Illicitly manufactured fentanyl, a synthetic opioid, has been a major contributing factor to increases across states, albeit at different rates (Somerville et al., 2017; Spencer et al., 2019).

Historically marginalized individuals, such as people living below the poverty line, who have a history of incarceration, and who are experiencing homelessness, are at increased risk for overdose (Feng et al., 2016; Hedegaard et al., 2017; van Draanen et al., 2020) because of racialized drug laws and structural barriers that limit their access to and retention on medications for opioid use disorder (MOUD) (i.e., methadone, buprenorphine, and naltrexone) and other harm reduction services. Those barriers include but are not limited to stigma and discrimination (Bearnot et al., 2019) and a lack of accessible or affordable transportation, especially in rural communities (Hedegaard & Spencer, 2021; Kiang et al., 2021). As such, improving access to effective MOUD and other harm reduction services by bringing those services directly to these populations would help address the overdose crisis.

One effective, but underutilized, modality to deliver services and treatment is mobile units (Gibbons et al., 2022; Stewart et al., 2021). Such mobile units take several forms but are often vehicles (e.g., vans, campers, or buses) converted to accommodate patient exams and provision of other health services. Mobile units have been used to deliver a spectrum of services: street outreach among people experiencing homelessness (Buzza et al., 2019; Fine et al., 2023; O’Carroll et al., 2017; O’Gurek et al., 2021); syringe service programs (SSPs) that deliver harm reduction services (Islam & Conigrave, 2007; Iyengar et al., 2019; Strike & Miskovic, 2018); and infectious diseases testing (Rosecrans et al., 2022). Mobile units are effective at reducing HIV infections and overdoses, which they achieve by increasing the availability of high-quality overdose prevention services (Maxwell et al., 2006), addressing other injection-related harms (Bramson et al., 2015), and facilitating linkage to substance use treatment (Chan et al., 2021; Hall et al., 2014). Emerging evidence suggests that compared to fixed-site programs, mobile units that provide MOUD can increase treatment access, uptake, and retention for underserved populations (Greenfield et al., 1996). A recent study from Weintraub et al. (Weintraub et al., 2021) demonstrated the feasibility of combining telemedicine and mobile services to provide MOUD to individuals living in underserved rural areas, with 3-month treatment retention similar to those obtained from office-based telemedicine MOUD programs.

Despite their potential effectiveness, cost and lack of funding are frequently cited as barriers to incorporating mobile units into community care due to concerns about initial capital investment and operating resources required (Attipoe-Dorcoo et al., 2020; Stewart et al., 2023), which can lead to a lack of financial capacity (Yu et al., 2017). Since cost can, and often does, influence decision-making even when interventions are known to be effective, economic evaluations related to mobile units are necessary (Raghavan, 2017). Policymakers and other community stakeholders need to understand the resources required to start and implement such interventions. Limited research has examined the costs of starting and operating a mobile unit providing syringe services, wound care, and opioid overdose education and naloxone distribution (OEND) (Bartholomew et al., 2021) or providing buprenorphine treatment (Shah et al., 2024). We sought to use the experience of establishing mobile units in the context of the HEALing Communities Study (HCS) to estimate the economic costs of starting and maintaining mobile units providing harm reduction including OEND and MOUD.

METHODS

Setting

We estimated the direct economic costs of mobile units in four Massachusetts communities from August 2020 through June 2022.

Supported by the National Institutes of Health (NIH) and Substance Abuse and Mental Health Services Administration (SAMHSA), the HCS is a four-year multisite, parallel-group, community-engaged, cluster randomized waitlist-controlled trial (ClinicalTrials.gov identifier NCT04111939) implementing evidence-based practices (EBPs) to reduce opioid morbidity and mortality in 67 urban and rural communities in four states (HEALing Communities Study Consortium, 2020). In the HCS, different community organizations implemented a diverse set of EBP strategies over an 18-month period in the following areas: 1) OEND; 2) effective delivery of MOUD (expansion, linkage, retention); and 3) safer opioid prescribing.

Four communities in Massachusetts – Bourne/Sandwich, Brockton, Gloucester, and Salem – identified distance to services and poor public transportation as significant barriers to both harm reduction and engaging in OUD treatment. These communities chose to implement mobile units, each selecting different services to provide to their community (Table 1 and Table S1). All four mobile units offered linkage to MOUD via telehealth, OEND, and harm reduction (e.g., naloxone, fentanyl test strips, safer smoking supplies) and syringe services (e.g., distribution of syringes and other injection equipment). MOUD via telehealth included case managers facilitating patient registration with the health center, telemedicine consults, and MOUD prescriptions. Salem and Brockton also provided HIV and hepatitis C testing. The Brockton mobile unit was unique in that it prescribed MOUD on the unit in addition to linkage via telehealth. Mobile units in Massachusetts only included prescription of buprenorphine and naltrexone, since federal regulations did not permit methadone vans during the interventions. However, naltrexone was in fact rarely prescribed because both patients and providers felt that it was less effective than buprenorphine, and therefore less desired.

Table 1 –

Overview of the 4 Massachusetts HEALing Communities Study Mobile Units

Community	Urban / Rural	Implementation date	Hours/Days	MOUD services	Additional services
Bourne/Sandwich (Barnstable County Community Cluster)	Rural	September 2020	3 days per week	MOUD initiation and retention via telehealth	OEND and harm reduction services Syringe exchange services Coordinate transportation to/from the pharmacy Assistance with insurance barriers to telemedicine Short-term crisis management
Brockton	Urban	August 2020	32 hours per week (4 days per week)	In-person MOUD prescribing on the unit with telesupport	OEND and harm reduction services Syringe exchange services Infectious disease testing (with linkage to infectious disease treatment) Preventative medicine Recovery coaching and referrals for behavioral health support Health insurance enrollment Case management and referrals
Gloucester	Urban	October 2020	25 hours per week	MOUD initiation and continuation via telehealth	OEND and harm reduction services Syringe exchange services Delivery of “rescue kits” with harm reduction supplies and MOUD service information to residents upon release at Middleton Prison Refer and transport patients to same day methadone clinic intakes
Salem	Urban	October 2020	12 hours per week (2 days per week)	MOUD initiation and continuation via telehealth	OEND and harm reduction services Syringe exchange services HIV testing

Abbreviations: MOUD, Medication for Opioid Use Disorder; OEND, Opioid Education and Naloxone Distribution

Cost Estimation Procedure

Overview.

To estimate the mobile units’ start-up and operating costs, we used invoices submitted by Fiscal Agents. These fiscal agents provided funds to the community organizations responsible for operating the mobile units. We reviewed the invoices (see Tables S2–S5 for sample invoices), making sure to understand the meaning of each cost category. The review process employed a semi-structured method, categorizing information from invoices into distinct groups. These categories encompassed a range of expenditures, including costs related to staff personnel (including hiring processes), the van (purchase, maintenance, and insurance), supplies and services (medical supplies and pharmaceuticals), and miscellaneous items (patient transportation and travel expenses). To ensure data quality, we cross-referenced the invoice data with documentation provided by community members, evaluating both completeness and accuracy. In instances of data discrepancies, we convened to collectively determine a final consensus, clarifying the classification of unclear costs. All costs are presented in 2022 US dollars. The cost perspective for this analysis was that of the community organizations operating the mobile unit.

Based on the description of interventions, we identified three models of care for the mobile units:

• Category 1 (Bourne/Sandwich, Gloucester): baseline MOUD linkage and overdose prevention services (i.e., MOUD linkage via telehealth, OEND, and harm reduction including syringe services),

• Category 2 (Salem): category 1 + testing for infectious disease, and

• Category 3 (Brockton): category 2 + in-person MOUD prescribing on the van.

Start-up costs.

In general, start-up costs are initial investments necessary for deploying mobile units. They can be incurred before and during the initiation of the intervention, but they are generally one-time or short-term costs required to establish the mobile unit. For this study, start-up costs included training activities for the personnel and purchasing equipment and supplies (e.g., van purchase, van upfitting, and initial supplies such as naloxone distribution kits, information technology, and telehealth supplies). It is also very likely that some start-up costs were not described as such in the invoices, thus underestimating the total start-up costs. Thus, based on expert opinion with NL (Community Data Manager for the Gloucester mobile unit) and the Financial Manager for Massachusetts HCS, we assumed that 10% of the remaining total supplies cost (not explicitly defined as start-up) was start-up cost.

Operating costs.

Operating costs are recurrent costs that the unit accrues related to the daily management and provision of services. We divided operating costs into three categories: personnel and contracted services, equipment and supplies, and other costs. Personnel and contracted services included the personnel costs noted in the invoices (see Tables S2–S5), peer stipends and spending for consultants and contracted services, which are those from an internal or external entity to provide a service. Peer stipends in Gloucester involved compensating peers with unrestricted cash stipends for their expertise and to obtain and distribute naloxone and other harm-reduction supplies. Equipment and supplies costs included the spending for the van (maintenance, insurance, storage) and the various program supplies (medications). Other costs primarily included communication and advertisement. Operating costs can be further categorized as variable costs dependent on the number of persons served by the mobile units (e.g., syringes, pharmaceuticals) or time-dependent costs, which are ongoing operation costs that accrue on a regular schedule and are independent of the number of people that utilized the mobile unit services (e.g., personnel time completing different tasks, monthly payments for insurance on the van).

Total Economic Costs

We calculated the total costs for varying time periods, contingent on the availability of invoices (see Figure 1). For each of the four mobile units, the average monthly operating cost was approximated by dividing the total cost over the relevant time frame by the number of months data were available. We reported start-up costs separately.

Figure 1 –

Timeline of the four mobile units in Massachusetts HEALing Communities Study

Sensitivity analyses

The nature of the invoices, which main purpose was for reimbursement, led us to make several assumptions to further classify the costs as start-up or operating. Thus, our main analysis assumed that 0% of the personnel cost that was not specifically reported for “training” in the invoices and 10% of the remaining supplies cost (other than van purchase and upfitting, naloxone distribution kits, information technology or telehealth supplies) was start-up cost. Given that these assumptions could potentially underestimate personnel costs and overestimate supply costs among start-up cost, we performed sensitivity analyses around the percentage of the total personnel and supplies costs attributable to start-up rather than operating. First, we assumed 5% of the total supplies cost was start-up (instead of 10%). Second, we assumed 5% of the personnel cost was start-up (instead of 0%).

Some sites did not report specific start-up costs. An alternative hypothesis to the one adopted in the main analysis could be using a site’s data to inform resource utilization for another site providing the same service but for which no data was reported in the invoices. We performed another sensitivity analysis, assuming that all sites had non-zero start-up costs for all components. For example, Bourne/Sandwich was the only community that reported personnel start-up costs. These costs included $3,576 for the training of a Harm Reduction Specialist. Meanwhile, Gloucester and Salem’s mobile units also had a Harm Reduction Specialist each, but no training cost associated with their invoices. Thus, we added $3,576 in this sensitivity analysis as personnel start-up costs for Gloucester and Salem.

We also estimated the average monthly operating cost during the operating period for each mobile unit, by dividing the total cost by the number of months of operation (e.g., 11 months for Brockton).

RESULTS

Start-up costs

Table 2 presents the start-up cost estimates for each of the four mobile units as well as the mean and median. Median start-up costs were $59,762 (range: $52,062 for Brockton - $113,671 for Bourne/Sandwich). Overall, start-up costs were almost entirely equipment and supplies. Only Bourne/Sandwich had documented personnel start-up costs, which accounted for 17% of the total start-up cost in this community. Regarding equipment and supplies costs, the largest percentage of costs was associated with the purchase of a van (see Table S6 for more details on start-up costs). Purchasing the van accounted for 80% of the total start-up costs for Gloucester and Brockton, 43% for Bourne/Sandwich, and 99% for Salem. The median cost to purchase a van was $47,906.

Table 2 -.

Start-up Costs for the 4 Massachusetts HEALing Communities Study Mobile Units

Cost category	Gloucester	Brockton	Bourne/Sandwich	Salem	Mean (n = 4)	Median (n = 4)
Equipment and supplies	$56,128	$52,062	$94,255	$63,396	$66,460	$59,762
Personnel	$0	$0	$19,416	$0	$4,854	$0
Total start-up cost	$56,128	$52,062	$113,671	$63,396	$71,314	$59,762

Operating costs

Table 3 presents the economic cost estimates for operating the mobile units, monthly and for the full period following implementation to the end of the study. This time period differed according to the communities: 18 months for Gloucester and Bourne/Sandwich, 23 months for Brockton, and 7 months for Salem.

Table 3 -.

Operating Costs for the 4 Massachusetts HEALing Communities Study Mobile Units

Cost category	Gloucester (18 months)	Brockton (23 months)	Bourne/Sandwich (18 months)	Salem (7 months)
Total cost over the entire time period
Personnel and contracted services	$62,702	$456,209	$429,303	$21,541
Equipment and Supplies	$28,979	$74,583	$45,375	$19,416
Other costs (communications and ad)	$0	$0	$14,821	$0
Total operating costs	$91,681	$530,792	$489,499	$40,957
Monthly cost
Personnel and contracted services	$3,483	$19,835	$25,253	$3,077
Equipment and Supplies	$1,610	$3,243	$2,669	$2,774
Other costs (communications and ad)	$0	$0	$872	$0
Total operating costs	$5,093	$23,078	$28,794	$5,851

The median monthly cost was $14,464 (range: $5,093 - $28,794). The highest cost category was the personnel (and contracted services) cost, which counted for 68% (Gloucester), 86% (Brockton), 88% (Bourne/Sandwich), and 53% (Salem) of the total operating costs. For instance, the total personnel (and contracted services) cost over 18 months was $62,702 for Gloucester and $429,303 for Bourne/Sandwich (which involved three different community organizations).

The mobile units employed different staffing models. Table 4 presents the detailed staffing and associated costs (i.e., personnel only, excluding contracted services). Total staffing costs over 18 months were $37,702 and 424,653 for Gloucester and Bourne/Sandwich, respectively. This was $425,235 for Brockton (over 23 months) and $21,141 for Salem (over 7 months).

Table 4 –

Detailed Staffing Models and Associated Costs for the 4 Massachusetts HEALing Communities Study Mobile Units

Community (time period)	Job title	Job description (when available)	FTE	Total costs
Gloucester (18 months, Jan. 2021 to June 2022)	Harm Reduction Specialist	Implementation and coordination of harm reduction initiatives, particularly focused on treatment and outreach	1	$22,851
	MOUD Community Coordinator	Facilitates patient access to services, referrals, warm hand-offs to programs, and maintaining regular contact with patients referred to outpatient MOUD services.	0.2	$13,651
	Provider on the Pier (from February 2022)	Provide Narcan and other support services to members of the fishing community, refer individuals from the docks (collaborates with mobile van)	0.2	$1,200
Brockton (23 months, Aug. 2020 to June 2022)	Program Manager	Established mobile MOUD program	0.1	$28,920
	Project Manager	Manages the mobile unit	1	$149,237
	Outreach Worker	Substance use counseling. Provide rapid access to naloxone and MOUD	0.8	$43,204
	Nurse Practitioner	Provide MOUD on the van. Perform wound care and blood draws	0.8	$98,354
	Licensed practical nurse	Clinical coordinator	1	$94,676
	Community Health Worker (March 2022)	Substance use counseling. Provide rapid access to naloxone and MOUD	0.9	$10,844
Bourne/Sandwich (18 months, Jan. 2021 to June 2022)	Harm Reduction Specialists	Provide one-on-one OEND & harm reduction education, and supplies to participants (including syringes)	2	$75,103
	Director of prevention and screening services	Program oversight	0.2	$52,855
	Recovery Support Navigators	Engage and support retention in MOUD by connecting participants to a MOUD provider via telehealth on van. Coordinate transportation to/from the pharmacy to obtain medication. Assist insurance barriers and provide short-term crisis management	2	$161,730
	Recovery Coaches	Engage and retain clients in care.	2	$122,215
	Nurse Practitioner	For project management. Also provide training on MOUD and harm reduction	0.15	$12,750
Salem (7 months, June 2021 to Dec. 2021)	Harm Reductionist	Coordinate MOUD telemedicine visits, assess demand, and report implementation progress.	0.4	$12,855
	Program Director	Program oversight	0.2	$8,286

Abbreviations: FTE, Full Time Equivalent; MOUD, Medication for Opioid Use Disorder; OEND, Opioid Education and Naloxone Distribution

Total economic costs

The total economic cost of a mobile unit, including both start-up and operating costs, varied across the different communities. The total economic costs for Gloucester and Bourne/Sandwich mobile units over 18 months were $147,809 and $603,170, respectively. The total economic cost for the Brockton mobile unit over 23 months was $582,855, and it was $104,353 for Salem over 7 months.

Sensitivity analyses

The results of the sensitivity analyses on the start-up costs are provided in Table S7; those on the operating costs are presented in Tables S8–S11. Start-up cost estimates proved most sensitive to changes in our assumptions regarding the percentage of personnel cost (not specified as start-up in the invoices) considered being start-up costs (Table S7). Assuming that 5% of the personnel cost was start-up cost (Table S7, sensitivity analysis 2), median total start-up costs were $68,889 (range: $57,953 for Gloucester - $134,903 for Bourne/Sandwich). This is almost a $10,000 difference compared to the main analysis ($59,762). This was especially marked for Brockton and Bourne/Sandwich, with an increase in their start-up costs of $21,262 and $40,648, respectively (entirely due to the change in personnel start-up costs). Assuming that all sites had non-zero start-up costs for all components also increased the total start-up costs for all mobile units (Table S7, sensitivity analysis 4). This was particularly the case for Brockton, with an increase of nearly $30,000 (partly due to $12,076 personnel start-up costs, while no personnel start-up cost was identified in the main analysis for this mobile unit). In this sensitivity analysis, the Brockton mobile unit start-up costs ($81,686) were higher than those for Gloucester ($62,098) and Salem ($71,076), compared to the main analysis in which Brockton had the lowest start-up costs.

In our alternative calculation, we found that the overall monthly cost differed slightly for most of the mobile units when we averaged over the operation period for each (Table S11). Salem was the same, Gloucester was $1,000 higher, and Bourne/Sandwich was $6,000 higher compared to the main analysis. The most notable increase was observed for Brockton, for which the overall monthly operating cost doubled ($48,254) compared to the main analysis ($23,078).

DISCUSSION

Mobile units that address substance use and provide extensive services to people who use drugs can be an effective means of overcoming barriers to care (Berk, 2023; Messmer et al., 2023). This study estimated the costs associated with implementing and operating such mobile units in four Massachusetts communities. Overall, we found that starting a mobile unit providing harm reduction, OEND, and MOUD (at minimum via telehealth) required a one-time median start-up investment of approximately $60,000, though more than 75% of this investment is attributable to the purchase of the vehicle. The monthly average cost for sustaining these programs was approximately $15,000. Although this might seem like a considerable expense, such mobile units have the potential to serve a large number of individuals, and their services can help avert costly hospitalizations, emergency department care, and other societal costs such as those of an overdose estimated at around $4,000 (Behrends et al., 2019). In addition, a recent study modeling the potential impact of mobile methadone units showed that such interventions can be particularly effective in improving MOUD treatment rates, especially if they operate in rural communities (Gibbons et al., 2024). Published estimates suggest that MOUD prevents one fatal overdose death for every 40 patients treated (Schoenfeld et al., 2020). Depending on the number of individuals reached by the mobile units and subsequently treated, this may have substantial implications in terms of efficiency. Given the high economic burden of OUD and fatal overdose (Florence et al., 2021), the benefits of mobile units in increasing MOUD treatment uptake and decreasing the incidence of overdose could partially offset their cost.

Disaggregating the costs into resource categories, the cost of personnel was the highest component. Overall, personnel costs (including contracted services) accounted for nearly 85% of the total operating costs and 70% of the total economic costs (including start-up costs). Although purchasing a van may constitute the highest up-front expense and therefore dominate decision-making when considering implementation of a mobile unit, over even the medium term (first year of operations), ongoing personnel costs associated with the mobile unit operations emerged as a paramount issue. This underscores the need to expand the focus beyond only the initial high start-up costs, such as a vehicle purchase, to recognize how pivotal ongoing personnel costs are for sustaining such mobile units.

One major strength of this cost analysis is the use of cost data and the experience from four distinct mobile unit interventions in four different communities in Massachusetts. Each mobile unit intervention was context-specific, with the goal of providing services tailored to each community’s needs (Chatterjee et al., 2023). The variety of these models may provide stakeholders with a roadmap for their investments, should they wish to implement mobile units with comparable characteristics to those outlined in this study. The mobile units in urban communities (Brockton, Gloucester, and Salem) had similar start-up costs (approximately $50,000-$60,000). In contrast, Bourne/Sandwich, a rural community cluster involving three different community organizations, yielded a start-up cost approximately twofold higher than the other communities ($114,000). Regarding the operating costs, we also observed some notable differences. Costs were available for at least one year of operation for three communities: Gloucester (18 months), Brockton (23 months), and Bourne/Sandwich (18 months). The mobile units in Gloucester and Brockton illustrate two urban settings with very different operating costs. Indeed, monthly operating costs were $5,000 and $23,000 for Gloucester and Brockton, respectively. This difference is in part due to the fact that the Brockton mobile unit offered more services. The Gloucester mobile unit offered MOUD linkage via telehealth, OEND, and harm reduction (including syringe exchange services); while the Brockton mobile unit provided infectious disease testing and MOUD on the van, in addition to the services also offered in Gloucester. As such, the difference observed between Brockton and Gloucester mobile units’ operating costs could partially reflect the incremental cost that is necessary to include these two additional services to an existing mobile unit already providing OEND and MOUD via telehealth services. Although the Bourne/Sandwich mobile unit essentially provided the same services as the one in Gloucester (i.e., MOUD linkage via telehealth, OEND, harm reduction, and syringe exchange services), the monthly operating costs were closer to Brockton ($29,000). A potential explanation could be the larger scope of the intervention in Bourne/Sandwich involving three different community organizations with more personnel and the fact that this is a rural community. These findings provide insights into unique community needs potentially applicable to the implementation of similar interventions in communities outside of Massachusetts. In addition, it should be noted that this is one of the first studies evaluating the cost of mobile units offering MOUD (Shah et al., 2024), filling an important gap in the field.

Several challenges and limitations emerged within our cost analysis and should be acknowledged when interpreting the findings of this study. First, the primary source of data were invoices that were designed for reimbursement, not for research purposes. The invoices typically presented aggregated expenses rather than an itemized breakdown of economic costs by category. As a result, it was difficult to precisely estimate which and/or what portion of supplies and personnel costs were start-up costs. We are conscious of the potential for biased reporting of costs in the invoices, especially the possible reporting gaps between the different sites, introducing data interpretation and comparability challenges. This led us to make some assumptions, for example, that 10% of the total supply costs and 0% of the personnel costs, which were not clearly defined as start-up costs, were in fact start-up costs. These assumptions, however, were explored in sensitivity analyses. In particular, concerns regarding the heterogeneity in cost reporting were addressed by assuming that all sites had non-zero start-up costs for all components they were supposed to include (e.g., training for the personnel). When we used one community’s data to impute for another, it drove site start-up costs closer to each other but did not qualitatively alter our findings. We acknowledge that this may still lead to an underestimate and, thus, our estimates may be used as a benchmark for the starting point for a community, should they want to implement a similar unit. In particular, experience and training that staff members may have had prior to joining the mobile unit was probably underestimated. Given these complex factors, readers should interpret our results on start-up costs with caution. Second, it is entirely possible that we are missing some costs, due to the design of our study. For example, this is the case with the time spent planning activities and the associated salaries, which are important costs for implementing mobile units. Certainly, mobile units do not operate in isolation from the community but work in conjunction with other providers that may not be directly affiliated with the specific mobile unit’s intervention. This has significant implications for overall community costs, though it does not directly impact the funds allocated to the mobile units intervention in our research. In addition, personnel directly involved with the expenses of the mobile units (including peers) were costed and included in both start-up and operating costs. Third, there were several differences between the four communities, specifically regarding their rurality. While our cost estimates include a diversity of communities, this analysis was limited to one rural setting. As such, these results may not be generalizable to all rural communities aiming to implement mobile units offering OEND and MOUD services. More generally, our results are not generalizable to any kind of mobile units, but only to the ones with similar features and services offered (e.g., offering MOUD via telehealth in most cases). Comparing the costs between sites or interpreting aggregate values from our study should therefore be done with great caution. In addition, personnel cost data were derived from institution-specific salaries. Although this could be interpreted as a limitation to generalizability, it is also a sign of a community-engaged study design. Other institutions could use the full-time equivalent values from our detailed staffing models to calculate an estimate for their healthcare system based on local salary levels. Fourth, none of the Massachusetts Communities That HEAL intervention mobile units (included in our study) prescribed methadone. On the contrary, some Wave 2 mobile units hired an on-unit methadone clinician, and future studies should assess the costs related to offering all three MOUD. Finally, the analysis only provides implementation and first two years (maximum) operating costs for the different mobile units, and we did not examine the economic benefits of the mobile units nor any effectiveness outcome. In particular, we lacked valuable data on the number of people reached by the mobile units, which would be useful for describing variable costs and estimating the potential health benefit (e.g., increase in MOUD use, decrease in fatal overdoses). Future research should provide longer term results to get more information about the costs at a steady state. In addition, further exploration of cost-effectiveness and program efficiency should be considered across different community-based settings.

In conclusion, our study offers vital information for stakeholders considering mobile unit interventions that offer both OEND and MOUD services. Our results can serve as a starting point for future research and studies that seek a comprehensive understanding of the complete economic landscape of mobile unit operations. Furthermore, we provide insights regarding different types of communities (e.g., rural, urban) and mobile unit interventions providing different services. Our analysis demonstrates that different communities will have to contend with different costs, especially in providing tailored interventions for their communities.

Highlights.

1. This is the first study to provide an economic cost analysis of mobile units that provide a diverse array of health services for people who use drugs, particularly medication for opioid use disorder (MOUD).

2. The operating costs of mobile substance use treatment and harm reduction units are substantial but vary by community settings and services offered

3. These results provide stakeholders with a roadmap of resources that might be necessary to implement and maintain a mobile unit in other communities.