Combined Pharmacotherapy and Psychotherapy Impact on Opioid Use Disorder Treatment

ABSTRACT

Objective: To examine whether patients with opioid use disorder (OUD) who receive both medications for OUD (MOUD) and psychotherapy experience lower rates of acute care utilization and total cost of care relative to periods when they receive either MOUD or psychotherapy alone.

Study Design: Retrospective claims data analysis of commercially insured adults in the US with a diagnosis of OUD between 2019 and 2022.

Methods: Multivariate linear regression models were estimated using rates of all-cause emergency department (ED) visits, inpatient admissions, and total cost of care as the outcome variables. Subsample analyses of urban and rural residents were also performed to examine whether the impacts differ by rurality.

Results: The sample included 18,316 patients representing 450,437 member-month observations. During the months when patients received both MOUD and psychotherapy, their per-member per-month rates of all-cause ED visits and inpatient admissions were lower by 33.5% (P < .001) and 47.6% (P < .001), respectively, relative to months when they received either MOUD or psychotherapy alone. This translated to a 38.3% lower total cost of care (P < .001). Similar magnitudes were noted among both rural and urban residents.

Conclusions: Receipt of both MOUD and psychotherapy is associated with reductions in the rates of ED visits and inpatient admissions along with lower total cost of care relative to receipt of either MOUD or psychotherapy alone. The results suggest that such a combined and complementary approach to OUD treatment may be justified due to its improved patient health outcomes and potential economic value.

Am J Manag Care. 2026;32(6):In Press

Takeaway Points

• Evidence-based psychotherapy as an alternative or supplement to medications for opioid use disorder (MOUD) is currently available.
• Empirical evidence on the combined effects of MOUD and psychotherapy is, however, mixed.
• This study reports that combined MOUD and psychotherapy is associated with lower all-cause acute care utilization and cost of care relative to either MOUD or psychotherapy alone.
• Such combined and complementary approaches to OUD treatment may be justified by their improved patient health outcomes and potential economic value to the health system.

Opioid use disorder (OUD) accounts for the majority of overdose cases in the US, surpassing all other substances.¹ OUD also has substantial economic costs; the cost of OUD treatment alone exceeds $34 billion annually.² There are currently 3 FDA-approved medications for OUD (MOUD): methadone, naltrexone, and buprenorphine. MOUD is associated with decreased all-cause and cause-specific (eg, overdose) mortality and improved health outcomes.³ Yet fewer than 1 in 5 patients with OUD receive MOUD.⁴ Barriers such as limited treatment programs, shortage of providers, and stigma interfere with MOUD utilization and increase reliance on acute care services, thus leading to higher cost of care.⁵ Furthermore, OUD treatment is often complicated by psychiatric comorbidity, which contributes to lower treatment adherence, poorer health, and greater health care costs.⁶

In addition to MOUD, evidence-based psychotherapy alternatives or supplements to MOUD currently exist. Cognitive behavioral therapy is an effective treatment option for OUD, with a focus on acquiring and practicing adaptive coping skills to replace maladaptive responses following exposure to proximal antecedents of substance use.⁷ Motivational interviewing is a set of psychotherapeutic techniques used to help patients resolve ambivalence about and enhance motivation to change substance use behaviors, particularly among those at earlier stages of readiness to change.⁸

Empirical evidence on the combined effects of MOUD and psychotherapy is mixed. Although earlier studies have indicated no significant impact,^9-11 several more recent studies have shown some benefits of combined MOUD and psychotherapy in certain contexts.^12-14 Moreover, the relationship between the treatments and outcomes may vary across different populations. Rural patients, for example, face unique treatment challenges compared with those in urban areas, such as limited access to OUD treatment due to low provider density. Rural patients with OUD also report higher rates of psychiatric comorbidity and greater mental health stigma compared with their urban counterparts.⁵

As such, this study assessed the impact of MOUD and psychotherapy on treatment costs associated with OUD among a large sample of commercially insured adults in the US. Specifically, it is hypothesized that OUD patients who received both MOUD and substance use disorder (SUD)–related psychotherapy during the same period would experience lower rates of emergency department (ED) visits and inpatient admissions as well as lower total cost of care vs periods when they received either MOUD or psychotherapy alone. Also, because it is expected that the impacts of MOUD and psychotherapy may differ depending on rurality of patients’ residence, separate subsample analyses were conducted for rural and nonrural patients.

This study did not attempt to answer whether MOUD or psychotherapy alone is superior. Given that MOUD is the commonly accepted standard of care in OUD treatment, this study considered only the patient population with OUD that had used any MOUD; no consideration was given to patients with OUD who might have used psychotherapy exclusively and never received any MOUD, even if such a subset ever existed. Within the population with OUD that has used MOUD, there is significant variation in timing, duration, and frequency of MOUD and psychotherapy receipt. Therefore, this study took advantage of this natural variation to tease out the hypothesized effects of combined pharmaco- and psychotherapy for OUD treatment.

METHODS

This study was an observational retrospective secondary data analysis using a limited data set obtained from the Health Care Cost Institute (HCCI),¹⁵ covering the period from January 1, 2019, through December 31, 2022. HCCI is a nonprofit independent research organization that collects and maintains a large national database of multipayer longitudinal health insurance claims information for research purposes. The HCCI database accounts for approximately 33% of the commercially insured population in the US and represents more than 50 million health plan members.¹⁵ Because HCCI only receives claims data from members of employer-sponsored health plans, the HCCI database does not include claims data from those enrolled in public health insurance programs such as Medicaid and Medicare. The University of Rochester’s Institutional Review Board approved this study as exempt (study No. 9338).

Data

The HCCI claims data obtained by the study team included masked plan member and claim identifiers; service begin and end dates; International Statistical Classification of Diseases, Tenth Revision (ICD-10) diagnosis codes; provider type identifiers; service location identifiers; Healthcare Common Procedure Coding System (HCPCS) codes; and National Drug Codes (NDCs). MOUD use (ie, buprenorphine, methadone, and naltrexone) was captured via encounters that involved any NDCs or HCPCS codes previously identified as MOUD (eAppendix Tables 1 and 2) (eAppendix available at ajmc.com).¹⁶ SUD-related psychotherapy was captured via encounters that involved any HCPCS codes previously identified as SUD-related psychotherapy¹⁷ (eAppendix Table 3) and the corresponding ICD-10 primary diagnosis code of F1X.XXX (mental and behavioral disorders due to psychoactive substance use).

In addition, the HCCI data included allowed amounts associated with each claim. Allowed amounts captured the dollar values that health care providers were entitled to receive from either the health plans or the patients based on the contract in place between the providers and the health plan. Therefore, allowed amounts were considered to represent the true cost of care incurred by providers.

The study team also merged in the corresponding member enrollment data file, which captured information on plan enrollment and disenrollment dates, plan type (eg, preferred provider organization, health maintenance organization), and indicators for whether the member had behavioral health and/or prescription drug coverage in each period. The data file also included some demographic information, such as age, sex, and 5-digit zip code of residence. Using the zip code, we also classified members as either rural or urban area residents via Rural-Urban Continuum Codes (codes ≥ 4 indicated rural area).¹⁸

The following inclusion criteria were applied: adults (aged ≥ 18 years) who had 1 or more ICD-10 codes for OUD (F11.XXX) and received any MOUD at any point between 2019 and 2022. Additionally, the following exclusion criteria were applied: members who had any gap in either prescription drug or mental health coverage during the study period. The requirement that all members of the sample had received MOUD was designed to remove potential confounding due to unobserved heterogeneity between patients with OUD who received MOUD and those who never received MOUD. Also, as described earlier, focusing only on patients who received MOUD was consistent with the overall aim of this study. Furthermore, the requirement that all patients had no gap in prescription drug and mental health benefit coverages was designed to ensure that the total cost of care estimates were not directly or indirectly influenced by changing benefits and to minimize unobserved self-selection biases among employees who opt out of such benefits.

Statistical Analysis

To test the hypothesis, the following equation was estimated:

Y_it = β₀ + β₁MOUD_it + β₂PT_it + β₃(MOUD_it ×PT_it) + β₄X_it + β₅Time_t + ε_it

where Y_it represents the dependent variables of interest (ie, ED visits, inpatient admissions, or total cost of care) for member i at time t; MOUD_it and PT_it are binary indicator variables for whether member i received any MOUD or psychotherapy at time t, respectively; MOUD_it × PT_it represents the interaction term between MOUD_it and PT_it that equals 1 only if member i received both MOUD and psychotherapy during time t and 0 otherwise; X_it represents a set of patient characteristics, such as age, sex, rural residence, and count of SUD diagnosis codes other than OUD (to capture comorbid SUD conditions); Time_t represents a set of indicator variables for calendar year and months to capture secular trends such as the period of the COVID-19 pandemic and yearly trends in health care cost increases; and ε_it represents the error term.

The main coefficient of interest was β₃, which represents the interaction effect of receiving both MOUD and psychotherapy above and beyond the main effects of receiving MOUD or psychotherapy alone. Statistically significant estimates of β₃ that are less than 0 would therefore be evidence in support of our hypothesis. On the other hand, the expected directions of the main effects of MOUD (β₁) and psychotherapy (β₂) were ambiguous because although MOUD and psychotherapy are both expected to reduce acute care utilization and total cost of care in general, the timing of their use might have coincided with periods of higher utilization and higher cost for the patients (ie, patients may take up MOUD or psychotherapy when they are already in a health crisis).

The previous equation was estimated via a set of ordinary least squares (OLS) linear regression models. The dependent variables were aggregated to a per-member per-month (PMPM) level and treated as continuous variables. Because the same members were observed multiple times in the data, clustered SEs were reported and used for tests of statistical significance. Moreover, to examine whether the results differed by rurality, the same regression models were run on subsamples of rural and urban residents.

For ease of interpretation, the β₃ coefficient estimates were translated into percent changes in terms of rates of ED visits and inpatient admits on a per-1000 per-month basis and also into percent changes in cost of care on a PMPM basis as observed in the data. This was accomplished as follows: For those member-month observations in which the patient received both MOUD and psychotherapy, the corresponding regression-adjusted dependent variable values were calculated as observed in the data via the predict command in Stata. Subsequently, the MOUD and psychotherapy interaction effect was then set to 0 (to simulate the hypothetical situation where the interaction effect was 0; ie, β₃= 0), and the corresponding regression-adjusted variable values were recalculated for the same member-month observations again via the predict command, keeping the main effect variables and the covariates unchanged. The latter values therefore represented the regression-adjusted values of the dependent variables under the hypothetical situation in which the same patients received either MOUD or psychotherapy alone but not both.

Furthermore, sensitivity analysis was conducted in which alternative regression models other than OLS were fit to the same data to determine whether the results and overall conclusions were sensitive to different choices of regression models. For the dependent variables that are count data in nature (ie, ED visits and inpatient admissions), negative binomial regression models were estimated. For the cost variable, a generalized linear model with a log link and a γ distribution was estimated. All statistical analyses were conducted via Stata 18 (StataCorp LLC).

RESULTS

The final sample included 450,437 member-month observations across 18,316 unique members over the 4-year period from 2019 to 2022 (Table 1). Due to low numbers of member-month observations associated with any MOUD or SUD-related psychotherapy use, less than 4% of the member-month observations in the sample were associated with receipt of both MOUD and SUD-related psychotherapy. Similar patterns were observed among the rural and urban subsample comparisons.

Table 2 summarizes the coefficient estimates corresponding to the main effects of MOUD (β₁) and psychotherapy (β₂) as well as the interaction effect between MOUD and psychotherapy (β₃). Consistent with the hypothesis, all estimates of the interaction effects were negative and statistically significant at the 1% level, except for inpatient admission rates among rural residents. Moreover, the magnitudes of the effect sizes did not seem to differ significantly between the urban and rural residents.

The Figure illustrates the differences in the regression-adjusted estimates of acute care utilization rates and total cost of care as translated from the β₃ coefficient estimates shown in Table 2 (corresponding full regression outputs are shown in eAppendix Tables 4-6). The Figure indicates that during the months when patients received both MOUD and psychotherapy, their PMPM rates of ED visits and inpatient admissions were lower by 33.5% (P < .001) and 47.6% (P < .001), respectively, relative to months when they received either MOUD or psychotherapy alone. Such reductions in the acute care utilization rates corresponded to 38.3% lower total cost of care during the same months (P < .001). Similar magnitudes were noted among both rural and urban residents. The results of the sensitivity analysis (shown in eAppendix Table 7) indicate that the overall conclusion is robust to different choices of regression models.

DISCUSSION

Individuals with OUD are often resistant to treatment, with a mean of 8.48 recovery attempts to achieve remission.¹⁹ As indicated by the findings of this study, however, a combined pharmacological and psychotherapeutic treatment option may enhance the overall effects of OUD treatments above and beyond what would be expected from each option alone. Similar results have been reported in previous studies: Data from a sample of Medicaid patients showed higher rates of treatment adherence among those who received MOUD and psychotherapy compared with MOUD alone.¹³ Using Veterans Health Administration data, Fairley and colleagues found that combining MOUD with psychotherapy led to fewer overdose deaths and lower health care costs than MOUD alone.²⁰

As mentioned earlier, other studies have found little incremental benefit from adding psychotherapy to MOUD.^9-11 A key difference between some of these studies and the current study is how the outcome variables were defined and operationalized. For example, some studies used treatment retention and drug use as outcomes, whereas the current study used acute care utilization and health care costs instead. Therefore, the conclusion about the effectiveness of combined MOUD and psychotherapy may depend on the chosen outcomes; thus, a broader consideration of the potential benefits of combined treatments is warranted.

As noted previously, the current study did not address the question of whether SUD-related psychotherapy alone is better or worse than MOUD alone because the sample was restricted to patients who had MOUD at some point but not restricted to those who had psychotherapy at some point. This implies that the unobserved process of psychotherapy take-up was not controlled in the same way that the process of MOUD take-up was controlled in the analysis. As a result, the coefficients on the main effects terms of the MOUD and psychotherapy variables in the regression models as shown in Table 2 cannot be used to directly compare MOUD and SUD psychotherapy treatment effects against each other.

The results of this study provide important insight, especially from the payer’s perspective. To put this into context, the estimated financial impact of OUD treatment alone exceeds $34 billion annually.² Acute care utilization accounts for a disproportionate share of the total cost of care among patients with OUD.²¹ The direct annual per-patient outpatient psychotherapy cost generally falls below $6000, whereas annual pharmacological treatment costs range from $6000 to $15,000 based on the type of medication and the frequency of use.²⁰ Therefore, improved OUD management at lower levels of care can lead to significant net total cost of care savings and alleviate strain on acute care services, and this may be achieved via integrated approaches to combining MOUD and psychotherapy.

Policy makers recently made changes to increase MOUD access. As of 2023, providers no longer require a specialized X-waiver to prescribe buprenorphine.²² Further, initial data suggest that efforts to expand access to methadone through pharmacy and dispensing programs have been well received by key stakeholders.²³ These efforts have greater potential to impact patients in rural areas, where access to care is disproportionately affected by restrictive policies and provider density.⁵ Further research is necessary to better understand how geographic factors impact the relationship between OUD treatment approaches and both acute care utilization and health care costs.²⁴

Limitations

One limitation of this study is a lack of detailed information on how combined MOUD and psychotherapy were implemented in actual clinical settings due to data limitations inherent in claims data. As such, more nuanced interpretations of the results are not feasible. Moreover, because this was an observational study reliant on secondary data analysis, it is difficult to make causal inferences. For instance, individuals receiving both MOUD and psychotherapy may represent a more engaged patient population, which may inherently lead to better outcomes and lower costs, rather than the combined treatment itself being the driver of these improvements. Also, this study relied on an algorithm (ie, psychotherapy claims plus primary ICD-10 code for SUD) to identify SUD-related psychotherapy, potentially resulting in inaccurate case identifications. Thus, future studies may include replicating the results of this study using different MOUD and psychotherapy definitions as well as stratifying the sample by specific MOUD type (eg, buprenorphine vs methadone or naltrexone).

Finally, because the claims data were limited to the commercially insured population only, the findings may not be generalizable to Medicare and Medicaid populations. The generalizability is further limited by the fact that the study period covers the COVID-19 pandemic period, which might have reflected idiosyncratic patterns of MOUD and SUD-related psychotherapy utilization no longer relevant during the postpandemic period. Although the regression models attempted to control for this via inclusion of a set of year indicator variables, further investigations are necessary.

CONCLUSIONS

Receipt of both MOUD and psychotherapy during the same period is associated with significant reductions in rates of all-cause ED visits and inpatient admissions as well as lower total cost of care relative to receipt of either MOUD or psychotherapy alone. Therefore, the results suggest that such a combined and complementary approach to OUD treatment may be justified not only for its improved patient health effects but also for its potential economic value to the health system.

Author Affiliations: Department of Psychiatry (DM, CAC, PW, LA) and Department of Public Health Sciences (DM, SC), University of Rochester Medical Center, Rochester, NY; School of Nursing, University of Rochester (HC), Rochester, NY.

Source of Funding: This article was funded by the HHS Health Resources and Services Administration under cooperative agreement number UD9RH33632. The views expressed in this publication are solely the opinions of the authors and do not necessarily reflect the official policies of HHS or the Health Resources and Services Administration nor does mention of the department or agency names imply endorsement by the US government.

Author Disclosures: The authors report no relationship or financial interest with any entity that would pose a conflict of interest with the subject matter of this article.

Authorship Information: Concept and design (CAC, PW); acquisition of data (DM); analysis and interpretation of data (DM, HC, PW, SC); drafting of the manuscript (DM, CAC, HC, PW, SC); critical revision of the manuscript for important intellectual content (CAC, LA); statistical analysis (DM, HC); obtaining funding (LA); administrative, technical, or logistic support (SC); and supervision (LA).

Address Correspondence to: Daniel Maeng, PhD, University of Rochester Medical Center, 300 Crittenden Blvd, Rochester, NY 14642. Email: daniel_maeng@urmc.rochester.edu.

REFERENCES

1. Spencer MR, Garnett M, Miniño AM. Drug overdose deaths in the United States, 2002-2022. National Center for Health Statistics data brief 491. March 2024. Accessed April 29, 2026. https://www.cdc.gov/nchs/products/databriefs/db491.htm

2. Florence C, Luo F, Rice K. The economic burden of opioid use disorder and fatal opioid overdose in the United States, 2017. Drug Alcohol Depend. 2021;218:108350. doi:10.1016/j.drugalcdep.2020.108350

3. Santo T Jr, Clark B, Hickman M, et al. Association of opioid agonist treatment with all-cause mortality and specific causes of death among people with opioid dependence: a systematic review and meta-analysis. JAMA Psychiatry. 2021;78(9):979-993. doi:10.1001/jamapsychiatry.2021.0976

4. Key Substance Use and Mental Health Indicators in the United States: Results From the 2023 National Survey on Drug Use and Health. Substance Abuse and Mental Health Services Administration; July 2024. Accessed April 29, 2026. https://www.samhsa.gov/data/sites/default/files/reports/rpt47095/National%20Report/National%20Report/2023-nsduh-annual-national.pdf

5. Stopka TJ, Estadt AT, Leichtling G, et al. Barriers to opioid use disorder treatment among people who use drugs in the rural United States: a qualitative, multi-site study. Soc Sci Med. 2024;346:116660. doi:10.1016/j.socscimed.2024.116660

6. Santo T Jr, Campbell G, Gisev N, et al. Prevalence of mental disorders among people with opioid use disorder: a systematic review and meta-analysis. Drug Alcohol Depend. 2022;238:109551. doi:10.1016/j.drugalcdep.2022.109551

7. Magill M, Ray L, Kiluk B, et al. A meta-analysis of cognitive-behavioral therapy for alcohol or other drug use disorders: treatment efficacy by contrast condition. J Consult Clin Psychol. 2019;87(12):1093-1105. doi:10.1037/ccp0000447

8. Miller WR, Rollnick S. Motivational Interviewing: Helping People Change. 3rd ed. The Guilford Press; 2012.

9. Weiss RD, Potter JS, Fiellin DA, et al. Adjunctive counseling during brief and extended buprenorphine-naloxone treatment for prescription opioid dependence: a 2-phase randomized controlled trial. Arch Gen Psychiatry. 2011;68(12):1238-1246. doi:10.1001/archgenpsychiatry.2011.121

10. Fiellin DA, Barry DT, Sullivan LE, et al. A randomized trial of cognitive behavioral therapy in primary care-based buprenorphine. Am J Med. 2013;126(1):74.e11-74.e17. doi:10.1016/j.amjmed.2012.07.005

11. Ling W, Hillhouse M, Ang A, Jenkins J, Fahey J. Comparison of behavioral treatment conditions in buprenorphine maintenance. Addiction. 2013;108(10):1788-1798. doi:10.1111/add.12266

12. Weimer M, Morford K, Donroe J. Treatment of opioid use disorder in the acute hospital setting: a critical review of the literature (2014-2019). Curr Addict Rep. 2019;6:339-354. doi:10.1007/s40429-019-00267-x

13. Eren K, Schuster J, Herschell A, et al. Association of counseling and psychotherapy on retention in medication for addiction treatment within a large Medicaid population. J Addict Med. 2022;16(3):346-353. doi:10.1097/ADM.0000000000000914

14. Gustafson DH, Landucci G, Vjorn OJ, et al. Effects of bundling medication for opioid use disorder with an mHealth intervention targeting addiction: a randomized clinical trial. Am J Psychiatry. 2024;181(2):115-124. doi:10.1176/appi.ajp.20230055

15. Gaynor M. Introducing the health care cost institute. Health Management, Policy and Innovation. 2012;1(1):35-36.

16. Maeng D, Russell HA, Conner KR, Malcho J, Cross W, Lee HB. 30-month impact of medications for opioid use disorder on acute care utilization in rural communities. J Addict Med. 2025;19(2):143-149. doi:10.1097/ADM.0000000000001385

17. Baller J, Barrett A, Gill P, et al. Identifying Beneficiaries With a Treated Substance Use Disorder (SUD): Technical Specifications. CMS. February 2021. Accessed April 29, 2026. https://www.medicaid.gov/medicaid/data-and-systems/downloads/macbis/sud_techspecs.docx

18. Rural-Urban Continuum Codes. Economic Research Service, US Department of Agriculture. Updated January 7, 2025. Accessed February 4, 2025. https://www.ers.usda.gov/data-products/rural-urban-continuum-codes

19. Kelly JF, Greene MC, Bergman BG, White WL, Hoeppner BB. How many recovery attempts does it take to successfully resolve an alcohol or drug problem? estimates and correlates from a national study of recovering U.S. adults. Alcohol Clin Exp Res. 2019;43(7):1533-1544. doi:10.1111/acer.14067

20. Fairley M, Humphreys K, Joyce VR, et al. Cost-effectiveness of treatments for opioid use disorder. JAMA Psychiatry. 2021;78(7):767-777. doi:10.1001/jamapsychiatry.2021.0247

21. Larochelle MR, Wakeman SE, Ameli O, et al. Relative cost differences of initial treatment strategies for newly diagnosed opioid use disorder: a cohort study. Med Care. 2020;58(10):919-926. doi:10.1097/mlr.0000000000001394

22. Stone EM, Xie F, Miles J, Samples H, Olfson M, Crystal S. Buprenorphine dispensation after X-waiver elimination by clinician specialty. Am J Prev Med. 2025;69(5):108055. doi:10.1016/j.amepre.2025.108055

23. Wu LT, Mannelli P, John WS, Anderson A, Schwartz RP. Pharmacy-based methadone treatment in the US: views of pharmacists and opioid treatment program staff. Subst Abuse Treat Prev Policy. 2023;18(1):55. doi:10.1186/s13011-023-00563-w

24. Watson DP, Staton MD, Gastala N. Identifying unique barriers to implementing rural emergency department-based peer services for opioid use disorder through qualitative comparison with urban sites. Addict Sci Clin Pract. 2022;17(1):41. doi:10.1186/s13722-022-00324-3