Cost-Effectiveness of Integrated Behavioral Health for Depression, Anxiety, and Chronic Pain

ABSTRACT

Objective: To conduct a cost-effectiveness analysis comparing behavioral health integration (BHI) in primary care vs clinical decision support (usual care) for adult patients with depression and/or anxiety taking chronic opioid therapy for noncancer pain.

Study Design: Piggyback economic analysis of data collected for 632 adult patients during a pragmatic, stepped-wedge, type 2 effectiveness-implementation hybrid trial conducted in a health system in Louisiana between April 2019 and June 2022.

Methods: The study used decision tree analysis. The base case modeled study patients and assessed costs associated with interventions, acute care, ambulatory utilization, and prescriptions. Efficacy measures were modeled using quality-adjusted life-years (QALYs) and morphine equivalent daily dose (MEDD). Sensitivity analyses included 1-way sensitivity analysis and probabilistic sensitivity analysis (PSA). A US-based willingness to pay threshold range of $100,000 to $150,000 per QALY was used.

Results: In the base case, the BHI group incurred a cost of $10,489.19 per patient for 1 year compared with $5673.96 for usual care. BHI was associated with 0.0439 QALYs gained, which yielded an incremental cost-effectiveness ratio (ICER) of $108,784 per QALY. The BHI group had a MEDD reduction of 7.3 mg/d compared with an increase of 2.0 mg/d among usual care. This translates into an ICER of $513.51 per 1-mg/d reduction. One-way sensitivity analysis and PSA indicated that the cost of prescriptions for both study groups as well as the cost of primary care providers and licensed clinical social workers for the BHI group were the biggest drivers of cost-effectiveness.

Conclusions: BHI was cost-effective from the health system perspective, with reductions in prescription drug expenses being the primary driver of savings.

Am J Manag Care. 2026;32(7):In Press. https://doi.org/10.37765/ajmc.2026.89913

Takeaway Points

Compared with usual care, behavioral health integration delivered by a multidisciplinary team (community health worker, licensed clinical social worker, clinical pharmacist) via telemedicine had an incremental cost-effectiveness ratio of $108,784 per quality-adjusted life-year for adult patients with depression and/or anxiety taking chronic opioid therapy for noncancer pain.

• Unit cost investment in behavioral health integration needed to reduce opioid dosage was $514 per 1-mg/d reduction in morphine equivalent daily dose.
• Influential drivers of cost include prescription medications (eg, antidepressants, opioid and nonopioid analgesics) and encounters with primary care providers and social workers.
• Shifting care to lower-cost settings while minimizing acute care reliance supports the goals of expanding access and improving quality.

Major depressive disorder and generalized anxiety disorder frequently co-occur and have a high prevalence of comorbid conditions, including chronic noncancer pain.^1-4 Their associated economic burden related to increased health care costs with concomitant reduced quality of life is well documented.^1-5 Although prior study findings have demonstrated that collaborative care for depression, anxiety, and musculoskeletal pain disorders improves physical and mental health outcomes, quality of life, functional status, and satisfaction with care,^6-8 within the US, regulatory and financial barriers as well as historical separation of mental/behavioral health and primary care have created systemic barriers to effective integration of primary care and behavioral health.⁸

Health care organizations desperately need sustainable cost-effective solutions for integrating behavioral health into primary care given the up-front costs for implementing mental health services. Recent pay-for-performance health policies that incentivize high-value care and shared savings contracts between provider organizations and health plans may generate cost savings that could enable long-term investment in integrated care models. Health policy changes under the Medicare Access and CHIP Reauthorization Act of 2015 provided an avenue through which to incentivize behavioral health integration (BHI).⁹ In 2017, CMS began making payments for BHI services using the Psychiatric Collaborative Care Model.¹⁰ In 2018, CMS established additional payment for BHI models that do not involve a psychiatric consultant or designated behavioral health care manager.¹⁰ Between 2017 and 2021, the highest use of the new billing codes was for individuals enrolled in dual-eligible special needs health plans.¹¹ Medicare claims analysis further showed that providers significantly underutilize these care coordination codes and forgo considerable revenue.¹² Reasons for the lag in adoption rates are likely multifactorial, including the need for investments in care team and workflow redesign and related specifications for meeting billing requirements.

Multiple models of BHI in primary care have already demonstrated their ability to improve health outcomes,¹³ so it is imperative to implement specific models that fit local health system and population needs. In a recent comparative clinical effectiveness study of patients with depression and/or anxiety taking chronic opioid therapy for comorbid noncancer pain, a Louisiana-based health system implemented remote BHI collaborative care via telemedicine and/or telephone audio visits utilizing a multidisciplinary team (community health worker [CHW] for care coordination and care management, licensed clinical social worker [LCSW] for cognitive behavioral therapy, clinical pharmacist for medication management) during the COVID-19 pandemic.^14,15 The BHI intervention was compared with usual care, defined as electronic health record–based clinical decision support (EHR-CDS) for opioid risk mitigation. EHR-CDS included clinician prompts for assessing risk for opioid misuse, depression and anxiety screening, prescribing alerts for high-dose opioids (morphine equivalent daily dose [MEDD] ≥ 50 mg), indications for naloxone, and reviewing the state pharmacy monitoring program.¹⁶ Although the BHI intervention did not decrease rates of high-dose prescribing, it did reduce the mean daily dosage of opioids prescribed and increase prescriptions for naloxone and antidepressants.¹⁵ It also increased rates of referrals to psychiatry and other specialties. The BHI intervention did not generate differences in use of acute care services for this complex population. This follow-up study was conducted to assess cost-effectiveness of the BHI model compared with usual care.

METHODS

Study Setting, Design, and Population

This study used a piggyback economic evaluation approach to analyze data collected during a pragmatic, stepped-wedge, type 2 effectiveness-implementation hybrid trial conducted in 35 primary care clinics within Ochsner Health in Louisiana between April 2019 and June 2022. Detailed descriptions of the trial design, study interventions (BHI vs usual care), and main study outcomes are described in previous publications.^14-16 A total of 632 adult patients with depression and/or anxiety taking chronic opioid therapy for comorbid noncancer pain (221 BHI and 411 usual care matched controls) were included in the final data analysis. Most patients were female (72%), were White non-Hispanic (59%), and had multiple chronic medical conditions and 3 or more pain syndromes. The mean MEDD was 47.1 mg/d, and approximately 27% of patients were prescribed opioids with a mean MEDD of at least 50 mg/d. Baseline characteristics of the 2 groups were well balanced. The study was approved by the Ochsner Health Institutional Review Board (NCT03889418), and this cost-effectiveness report follows the Consolidated Health Economic Evaluation Reporting Standards 2022 statement for reporting health economic evaluations.¹⁷

Analytic Approach

All analyses followed the ISPOR—The Professional Society for Health Economics and Outcomes Research guidelines on cost-effectiveness analysis conducted alongside clinical trials and economic modeling.¹⁸ To conduct the cost-effectiveness analysis, a decision tree analysis model was developed alongside the clinical trial. A decision tree framework was chosen because it mimics the care patterns and pathways taken by patients in the real world.^19,20 The eAppendix Figure (available at ajmc.com) presents the decision tree framework to compare the cost-effectiveness impact of BHI vs usual care. Patients were modeled as traveling down a decision-making pathway and categorized into distinct groups with costs and efficacy measures applied. The results were then taken in aggregate across either decision path (BHI or usual care) and compared. The model was implemented using TreeAge Pro Healthcare (TreeAge Software, LLC).

Model Inputs

All variables used for the model analysis are provided in Table 1.^14,21-23 These model inputs can be divided into 3 broad categories. From the perspective of health systems, costs refer to the cost of care, medication, and treatments associated with each BHI intervention. Probabilities refer to the chance of a patient transferring to one group or another over the course of the decision tree, as shown in the eAppendix Figure. Probabilities across branches sum to 1.0, and patients do not cycle among branches, consistent with a 1-year decision tree horizon. Effectiveness measures were modeled using quality-adjusted life-years (QALYs) as well as MEDD. Under the base-case scenario, participants were followed for a single-year time horizon during which they received BHI and assessed for costs and effectiveness measures, after which they were compared with a calculated usual care model.

BHI intervention costs considered CHW-coordinated care management (CCM) as well as licensed pharmacist (PharmD) and LCSW encounters using the payment schedule of Orleans Parish in Louisiana. CCM costs used the US national mean hourly wage of a CHW in May 2023.²¹ PharmD encounters used the reimbursement of Current Procedural Terminology (CPT) code 99211 from the 2023 Medicare Physician Fee Schedule as a cost estimate.²² LCSW encounters used the reimbursement rate of CPT codes 90791, 90832, 90834, and 90837.²²

Outpatient (OP) encounter costs were calculated based on evaluation and management (E/M) codes documented in patients’ EHRs. Costs for OP clinic and emergency department (ED) encounters were estimated from reimbursement rates by E/M code.²² For remaining ED encounters with no documented E/M code, costs were imputed as the mean of the costed ED encounters with E/M codes. Inpatient (IP) costs were derived using Medicare claims from a representative sample of Louisiana Medicare beneficiaries who were similar to the study sample. Rather than using billing codes to assign costs, a mean cost per day of hospitalization was calculated from the Medicare claims data. IP costs for hospitalizations in our sample data were estimated by multiplying the flat daily rate by hospital length of stay (days). Only nonelective admissions were included.

We calculated medication costs using pricing information from UpToDate Lexidrug.²³ The price per unit was calculated for each combination of medication, route, strength, formulation, generic status, and package size. If a generic was available for a given medication combination, the generic pricing was used regardless of generic status of the prescribed medication. If generic pricing was unavailable, the nongeneric pricing was used.

Individual potential outliers were present in the upper tails of many of the cost metrics gathered from the clinical trial data. To mitigate the influence of these extreme observations, we used a winsorization process in which costs above the 99th percentile were replaced with the value of the 99th percentile.²⁴

The probabilities at each chance node in the decision tree were provided by patient demographic analysis of both intervention patients and controls. Of specific concern were the rate and dispositions of referrals, hospital utilization, and general primary care practitioner visits per year. Outcomes measures were assessed via pre- and postindex MEDD. The health utility scores were mapped from Patient-Reported Outcomes Measurement Information System Global Health–10 (PROMIS GH-10) assessment to estimate the daily mean changes in EuroQol–5 Dimension scores.²⁵ Due to the initial stepped-wedge design of the pragmatic clinical trial, the trial data set only had PROMIS GH-10 scores at baseline and post intervention for the BHI group. Using the preintervention measurements as a common baseline and the assumption of usual care as having no change, we estimated the daily impact of BHI on utility values and extrapolated across a year.

Willingness-to-Pay Threshold

In general health economics studies, willingness to pay (WTP) is defined as the maximum amount of money that an individual finds acceptable to pay for a specific health service or improvement. Traditionally in the United States, a WTP threshold of $50,000 per QALY was accepted as the standard since the early 2000s, but as time progresses and the costs of health care have risen, so too have WTP thresholds.^26,27 Per Vanness et al, health care interventions that exceed the range of $100,000 to $150,000 per QALY are unlikely to be successful in the US.²⁸ At the time this study was conducted, the health system had started hiring CHWs and LCSWs to increase access to behavioral health. The combined annual salaries and benefits for a CHW-LCSW team totaled $120,000. Clinical pharmacists were already integrated in primary care for chronic disease management and were not deemed a new expense. Therefore, for this analysis, a modern incremental cost-effectiveness ratio (ICER) threshold ranging from $100,000 to $150,000 per QALY was selected to represent the health system’s WTP.

Sensitivity and Scenario Analysis

Probabilistic and 1-way sensitivity analyses were conducted with input variables to identify which inputs drove the model results for the utility portion of the study (Figure 1, Figure 2, and eAppendix Figure). A probabilistic sensitivity analysis (PSA) was also conducted with 1000 iterations of randomly sampled values from estimated distributions of specific variables centered around cost (Table 1^14,21-23). In the 1-way sensitivity analyses, all inputs were varied by a set margin of 20% while holding other variables constant to generate a tornado diagram (Figure 3). Final sensitivity analyses used included 1-way sensitivity analysis and PSA to determine the most sensitive model parameters for cost-effectiveness and the robustness of results.

RESULTS

Table 2 presents the cost-effectiveness results. In the base-case scenario, BHI incurred a cost of $10,489.19 per patient over the course of the analyzed period from 2019 to 2022 compared with $5673.96 for usual care. This cost difference was divided by the annual QALY gain of 0.0439 for the BHI group vs the usual care group, yielding an overall ICER of $108,784 per QALY—above the standard WTP of $100,000 per QALY but below the upper limit of $150,000 per QALY for US-based health care interventions. The 2 scenarios had similar ICER measures. Using MEDD as the effectiveness measure, the BHI intervention group had a reduction in MEDD of 7.3 mg/d compared with a usual care increase in MEDD of 2.0 mg/d, which generated an ICER of –$513.51/MEDD (or $513.51 per 1-mg/d reduction in MEDD). In other words, for every $514 spent on BHI, an intervention patient’s MEDD dropped by 1.0 mg/d whereas a usual care patient had an increase in MEDD.

Figures 1 and 2 present the cost-effectiveness planes and cost-effectiveness acceptability curve. PSA was done over the course of 1000 simulations and at a standard WTP threshold of $100,000 per QALY. Under these conditions, BHI was considered cost-effective in approximately 48.4% of scenarios (Figure 2). When the analysis was repeated at the upper-limit WTP of $150,000, BHI was cost-effective in 51.1% of scenarios (Figure 2). At the extreme WTP of $0 per QALY, BHI was considered cost-effective in 25.8% of scenarios.

Figure 3 presents the tornado chart of 1-way sensitivity analysis results. Under 1-way sensitivity analysis, all relevant inputs were varied by 20% for lower and upper bounds. The most influential factors in the model were prescription medication costs (alternative nonopioid and opioid medications) for both study groups, and primary care providers and LCSWs for the BHI group were the biggest drivers of cost-effectiveness.

DISCUSSION

Behavioral health care management programs such as those run in the clinical trial at Ochsner Health¹⁴ may be a cost-effective option for managing comorbid depression and/or anxiety and chronic pain while mitigating risks associated with chronic opioid use. PSA showed a relative robustness of study results, with BHI showing a majority of simulations as cost-effective under the $150,000 WTP threshold and more than a quarter of simulations showing cost-effectiveness at a cost-neutral threshold. Deterministic sensitivity analysis also highlighted that the costs of prescriptions were the primary drivers of costs in the model. Use of MEDD as a cost-effectiveness measure also favored the BHI intervention compared with usual care.

The analysis for this study was conducted from the perspective of the health system launching a population health management program. Although $50,000 per QALY is still the commonly accepted WTP threshold, this study adopted a higher range based on up-front costs to the health system for the BHI team and on recent literature indicating a higher threshold for health care interventions in the US.^26-28 Various models of BHI and collaborative care management have already been shown to improve health outcomes in patients with depression, anxiety, and chronic pain^6-8; therefore, it is imperative that health systems choose models that best fit their population health strategies and local populations served.¹¹ Increasing patient access to psychotherapy for treating chronic mental and physical health conditions within primary care was the main driver of the model implemented in this study. The relative impact of prescription costs was not surprising given the medical complexity of the study population. A previous study of the intervention reported that BHI patients had higher rates of clinical guideline–concordant prescriptions for naloxone and antidepressants compared with the usual care group.¹⁴ Notwithstanding, prescription costs for the usual care patients still had a higher impact on cost-effectiveness compared with the BHI intervention components (ie, CHW-delivered care coordination, LCSW-delivered cognitive behavioral therapy [CBT], pharmacist-delivered medication management). Notably, this study further quantified the unit cost investment in BHI needed to reduce opioid medication dosage.

There is limited US-based evidence on the cost-effectiveness of the specific BHI model implemented in this study. However, a brief literature search did retrieve a 2019 meta-analysis of randomized controlled trials in which Ross et al compared the cost-effectiveness of guideline-concordant CBT and second-generation antidepressants (SGA) for initial treatment of major depressive disorder in the US using 1-year and 5-year time horizons.²⁹ CBT had a high likelihood of having an ICER of $100,000 or less per QALY at 5 years, whereas SGA had a high likelihood at 1 year; however, neither CBT nor SGA was consistently superior from a cost-effectiveness perspective.²⁹ Additionally, Ross et al reported that CBT produced greater QALYs at higher costs at 1 year and lower costs at 5 years.²⁹ In another US-based economic evaluation of a randomized clinical trial for chronic low back pain, Herman et al reported a 90% probability and an 81% probability of achieving less than $50,000 per QALY for mindfulness stress reduction and CBT, respectively, compared with usual care.³⁰ Both studies support the value of patients having access to psychotherapy.

Limitations

This study has several limitations with respect to the model inputs taken from the clinical trial. First, the clinical trial results showed that although the BHI and control groups were well balanced with respect to key characteristics (eg, demographics, comorbidities), patients within the BHI group had higher mean opioid dosage and lower inpatient and ED utilization at baseline compared with the controls, which could have impacted the final cost-effectiveness ratios. However, as previously reported, these baseline differences were not statistically significant. Quality-of-life measures reported by PROMIS GH-10 were potentially affected by multiple comorbidities, which may have impacted the final QALY calculations. The assumption that usual care produced no quality-of-life change may have overstated the incremental QALYs attributed to BHI. Limitations associated with using EHR data (eg, missing values, data entry errors) could have affected final cost-effectiveness ratios. The ICER of $108,784 per QALY is cost-effective in less than 50% of simulations at the lower bounds of the WTP range. Increased guideline-concordant prescribing of antidepressants and naloxone in the BHI group could have inflated pharmacy costs. These limitations complicate the interpretation of cost-effectiveness. Notwithstanding, the study’s analytic approach employed methods to reduce bias in results. Lastly, this clinical trial was conducted within one health system in the southern US, suggesting a limitation of generalizability. Future analysis should be conducted to improve and expand upon the work done here.

CONCLUSIONS

BHI, which included new staff hires, was cost-effective from the health system perspective. Cost of prescriptions was the most impactful driver in the analysis, and the increased cost from primary care encounters aligns with the goal of increasing access and improving quality of care within less expensive care settings. Future studies should examine cost-effectiveness from the perspective of health insurance plans and the societal perspective.

Author Affiliations: Tulane University Celia Scott Weatherhead School of Public Health and Tropical Medicine (DX, LS), New Orleans, LA; Center for Outcomes and Health Services Research, Ochsner Health (JB, EGP-H), New Orleans, LA; University of Queensland Ochsner Clinical School (EGP-H), New Orleans, LA; Ochsner-Xavier Institute for Health Equity and Research (EGP-H), New Orleans, LA.

Source of Funding: This study was supported by 1R01DA045029 from the National Institute on Drug Abuse of the National Institutes of Health.

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 (LS, EGP-H); acquisition of data (JB, EGP-H); analysis and interpretation of data (DX, LS); drafting of the manuscript (DX, JB, EGP-H); critical revision of the manuscript for important intellectual content (DX, JB, LS, EGP-H); statistical analysis (DX); provision of patients or study materials (EGP-H); obtaining funding (EGP-H); administrative, technical, or logistic support (EGP-H); and supervision (EGP-H).

Address Correspondence to: Eboni G. Price-Haywood, MD, MPH, MMM, Ochsner Health, 1401A Jefferson Hwy, Academic Center–2nd Floor, New Orleans, LA 70121. Email: eboni.pricehaywood@ochsner.org.

REFERENCES

1. IsHak WW, Wen RY, Naghdechi L, et al. Pain and depression: a systematic review. Harv Rev Psychiatry. 2018;26(6):352-363. doi:10.1097/HRP.0000000000000198

2. Mullins PM, Yong RJ, Bhattacharyya N. Associations between chronic pain, anxiety, and depression among adults in the United States. Pain Pract. 2023;23(6):589-594. doi:10.1111/papr.13220

3. Armbrecht E, Shah A, Schepman P, et al. Economic and humanistic burden associated with noncommunicable diseases among adults with depression and anxiety in the United States. J Med Econ. 2020;23(9):1032-1042. doi:10.1080/13696998.2020.1776297

4. Armbrecht E, Shah R, Poorman GW, et al. Economic and humanistic burden associated with depression and anxiety among adults with non-communicable chronic diseases (NCCDs) in the United States. J Multidiscip Healthc. 2021;14:887-896. doi:10.2147/jmdh.S280200

5. Zhdanava M, Kuvadia H, Joshi K, et al. Economic burden of treatment-resistant depression in privately insured US patients with co-occurring anxiety disorder and/or substance use disorder. Curr Med Res Opin. 2021;37(1):123-133. doi:10.1080/03007995.2020.1844645

6. Jacob V, Chattopadhyay SK, Sipe TA, Thota AB, Byard GJ, Chapman DP; Community Preventive Services Task Force. Economics of collaborative care for management of depressive disorders: a community guide systematic review. Am J Prev Med. 2012;42(5):539-549. doi:10.1016/j.amepre.2012.01.011

7. Thota AB, Sipe TA, Byard GJ, et al; Community Preventive Services Task Force. Collaborative care to improve the management of depressive disorders: a community guide systematic review and meta-analysis. Am J Prev Med. 2012;42(5):525-538. doi:10.1016/j.amepre.2012.01.019

8. Kwan BM, Valeras AB, Levey SB, Nease DE, Talen ME. An evidence roadmap for implementation of integrated behavioral health under the Affordable Care Act. AIMS Public Health. 2015;2(4):691-717. doi:10.3934/publichealth.2015.4.691

9. Medicare Access and CHIP Reauthorization Act of 2015, HR 2, 114th Cong (2015). Accessed May 15, 2025. https://www.congress.gov/bill/114th-congress/house-bill/2

10. Behavioral Health Integration Services. CMS; April 2025. Accessed May 15, 2025. https://www.cms.gov/files/document/mln909432-behavioral-health-integration-services.pdf

11. Coverage of Selected Behavioral Health Services Among Older Medicare and Medicaid Beneficiaries. Publication No. PEP24-06-008. Substance Abuse and Mental Health Services Administration; 2025. Accessed May 15, 2025. https://library.samhsa.gov/sites/default/files/cfri-older-adults-pep24-06-008.pdf

12. Agarwal SD, Basu S, Landon BE. The underuse of Medicare’s prevention and coordination codes in primary care: a cross-sectional and modeling study. Ann Intern Med. 2022;175(8):1100-1108. doi:10.7326/M21-4770

13. Westfall JM, Jabbarpour Y, Jetty A, et al. The State of Integrated Primary Care and Behavioral Health in the United States 2022. Robert Graham Center, HealthLandscape; 2022. Accessed May 15, 2025. https://www.graham-center.org/content/dam/rgc/documents/publications-reports/reports/state-of-integrated-pc-and-bh.pdf

14. Price-Haywood EG, Burton JH, Harden-Barrios J, et al. Decision support and behavioral health for reducing high-dose opioids in comorbid chronic pain, depression and anxiety: stepped-wedge cluster randomized trial. J Gen Intern Med. 2024;39(15):2952-2960. doi:10.1007/s11606-024-08965-7

15. Price-Haywood EG, Burton J, Harden-Barrios J, et al. Depression, anxiety, pain and chronic opioid management in primary care: type II effectiveness-implementation hybrid stepped wedge cluster randomized trial. Contemp Clin Trials. 2021;101:106250. doi:10.1016/j.cct.2020.106250

16. Price-Haywood EG, Robinson W, Harden-Barrios J, Burton J, Burstain T. Intelligent clinical decision support to improve safe opioid management of chronic noncancer pain in primary care. Ochsner J. 2018;18(1):30-35.

17. Husereau D, Drummond M, Augustovski F, et al; CHEERS 2022 ISPOR Good Research Practices Task Force. Consolidated Health Economic Evaluation Reporting Standards 2022 (CHEERS 2022) statement: updated reporting guidance for health economic evaluations. Value Health. 2022;25(1):3-9. doi:10.1016/j.jval.2021.11.1351

18. Ramsey SD, Willke RJ, Glick H, et al. Cost-effectiveness analysis alongside clinical trials II-an ISPOR Good Research Practices Task Force report. Value Health. 2015;18(2):161-172. doi:10.1016/j.jval.2015.02.001

19. Gupta N, Verma R, Dhiman RK, Rajsekhar K, Prinja S. Cost-effectiveness analysis and decision modelling: a tutorial for clinicians. J Clin Exp Hepatol. 2020;10(2):177-184. doi:10.1016/j.jceh.2019.11.001

20. Sado M, Koreki A, Ninomiya A, et al. Cost-effectiveness analyses of augmented cognitive behavioral therapy for pharmacotherapy-resistant depression at secondary mental health care settings. Psychiatry Clin Neurosci. 2021;75(11):341-350. doi:10.1111/pcn.13298

21. Occupational employment and wages, May 2023: 21-1094 community health workers. US Bureau of Labor Statistics. Updated April 3, 2024. Accessed October 1, 2025. https://www.bls.gov/oes/2023/may/oes211094.htm

22. PFS carrier specific files. CMS. Accessed November 10, 2023. https://www.cms.gov/medicare/payment/fee-schedules/physician/carrier-specific-files

23. Lexidrug. UpToDate. Accessed August 10, 2024. https://online.lexi.com/

24. Weichle T, Hynes DM, Durazo-Arvizu R, Tarlov E, Zhang Q. Impact of alternative approaches to assess outlying and influential observations on health care costs. Springerplus. 2013;2:614. doi:10.1186/2193-1801-2-614

25. Revicki DA, Kawata AK, Harnam N, Chen WH, Hays RD, Cella D. Predicting EuroQol (EQ-5D) scores from the Patient-Reported Outcomes Measurement Information System (PROMIS) global items and domain item banks in a United States sample. Qual Life Res. 2009;18(6):783-791. doi:10.1007/s11136-009-9489-8

26. Neumann PJ, Kim DD. Cost-effectiveness thresholds used by study authors, 1990-2021. JAMA. 2023;329(15):1312-1314. doi:10.1001/jama.2023.1792

27. Neumann PJ, Cohen JT, Weinstein MC. Updating cost-effectiveness—the curious resilience of the $50,000-per-QALY threshold. N Engl J Med. 2014;371(9):796-797. doi:10.1056/NEJMp1405158

28. Vanness DJ, Lomas J, Ahn H. A health opportunity cost threshold for cost-effectiveness analysis in the United States. Ann Intern Med. 2021;174(1):25-32. doi:10.7326/M20-1392

29. Ross EL, Vijan S, Miller EM, Valenstein M, Zivin K. The cost-effectiveness of cognitive behavioral therapy versus second-generation antidepressants for initial treatment of major depressive disorder in the United States: a decision analytic model. Ann Intern Med. 2019;171(11):785-795. doi:10.7326/m18-1480

30. Herman PM, Anderson ML, Sherman KJ, Balderson BH, Turner JA, Cherkin DC. Cost-effectiveness of mindfulness-based stress reduction versus cognitive behavioral therapy or usual care among adults with chronic low back pain. Spine (Phila Pa 1976). 2017;42(20):1511-1520. doi:10.1097/brs.0000000000002344