The impact of physical therapy direct access policy on opioid shipments and opioid-related deaths: A difference-in-differences analysis Authors: Geronimo Bejarano, Peter Hull, James J. Young, David J. Meyers, Kevin N. Griffith, Francesca Baratta, Francesca Baratta, Francesca Baratta Published: 2026 ABSTRACT BackgroundWe exploit state-level policy changes for physical therapy direct access to estimate the effect of enacting direct access physical therapy policy on per-capita opioid pill volume (PCPV) and opioid-related deaths (ORD) compared to states without direct access to physical therapy.MethodsWe used a staggered difference-in-differences analysis to assess changes in PCPV and ORD for states that did (treatment states) and did not enact (control states) direct access physical therapy policies between 2006–2014. Various sensitivity analyses including synthetic control methods were used to check the robustness of our results.ResultsTreatment states had higher baseline PCPV (29.9 vs. 27.3) and ORD (27.2 vs. 16.9) compared to the control states. Pre-trends in PCPV and ORD were parallel between treatment and control states. After direct access physical therapy policies were enacted, PCPV was significantly reduced by −4.6 (95% confidence interval: −6.2 to −3.0) percentage points, a 15.4% relative reduction. There was no significant difference in ORDs (−2.3 [95% confidence interval: −7.2 to 2.6]).ConclusionDirect access physical therapy policies significantly reduced PCPV but not ORDs. Our results suggest that administrative barriers to physical therapy may lead to greater opioid reliance and reduced population health. Introduction The United States continues to have an ongoing public health crisis of opioid misuse and addiction with 17,000 prescription opioid-related deaths in 2021 and nearly 3 million people with opioid use disorder in 2020. [1–3] Musculoskeletal pain conditions are now the second leading cause of global disability, the most frequently cited reason to seek care and leading health care expenditure in the United States. [4–8] Opioids are most commonly prescribed in the treatment of musculoskeletal pain which unifies these two public health crises. [9,10] For patients with a newly diagnosed musculoskeletal pain condition, 21% are prescribed opioids compared to only 10% prescribed physical therapy. [11] Increased utilization of safe and cost-effective alternative treatments is needed to address the opioid crisis, reduce unnecessary spending, and lower the disability burden of musculoskeletal pain. [12–14] Non-pharmacological treatments (e.g., exercise and manual therapy) commonly delivered by physical therapists are consistently recommended by clinical practice guidelines as first-line treatment options for all musculoskeletal pain conditions prior to initiation of opioid prescriptions. [15–19] Likewise, the Centers for Disease Control recommended prioritizing non-pharmacological treatments for pain because evidence suggests they are as effective as opioids with a lower risk of adverse events. [20] If physical therapy for musculoskeletal pain does reduce opioid prescription volume, policymakers should consider removal of barriers to physical therapy access and encourage greater uptake of these services. Previous studies found an association between utilization of physical therapy for musculoskeletal pain and lower likelihood of opioid prescription. [21–23] However, these associations may be biased due to confounding from unobserved differences in the types and profiles of patient with musculoskeletal pain that seek care from physical therapists compared to those who seek care from medical doctors or other health care providers with prescription rights. Causal inference methods such as difference-in-differences (DID) allow us to derive causal estimates from policy change since they are not biased from unobserved time-invariant differences unlike regression and matching which have been used in previous studies. [24,25] DID methods are employed in other clinical areas to assess policies without bias from observed and unobserved time-varying confounders. [26] To our knowledge, no previous study has analyzed the state level variation in physical therapy direct access laws using DID. Therefore, we exploit state-level physical therapy direct access policy variation over time to compare trends in opioid shipments to retail pharmacies between states that have and have not enacted direct access of physical therapy. We also analyzed opioid-related deaths as a secondary outcome and performed several sensitivity analyses to test the robustness of our findings to changes in our study design. Methods Data sources We obtained data from a previous study that combined the 2006–2014 U.S. Drug Enforcement Administration’s Automation of Reports and Consolidated Orders System (ARCOS) pill shipment database and the Health Resources and Services Administration’s Area Health Resources Files (AHRF). [27–29] ARCOS contains detailed information on the distribution of controlled substances from manufacturers to retailers and was publicly released following a 2019 lawsuit by the Washington Post. AHRF contains county-level data on over 1000 variables including population demographics, healthcare infrastructure, health workforce, and health services utilization. We merged this dataset with information on state-level direct access physical therapy policy changes reported by the American Physical Therapy Association. [30] The results were reported in accordance with Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline. [31] Physical therapy direct access policy The exposure of interest was whether a state implemented a physical therapy direct access policy, applied to county-level outcome data within each state. Access differs between states as some require that a patient receives a referral from a physician prior to being able to receive care from a physical therapist. Several states enacted policies allowing patients direct access to physical therapy without a referral during our study period. In 2010, Hawaii passed physical therapy direct access followed by Kansas, California, and Indiana in 2013 therefore these states are our treated states. As of 2014, Alabama, Illinois, Louisiana, Mississippi, Missouri, New Mexico, Oklahoma, and Texas still did not have physical therapy direct access therefore these states served as the never treated control states. Any other states were excluded due to having partial or provisional direct access prior to the study periods. Outcomes Our primary outcome was the per capita pill volume (PCPV) of opioids, defined as the annual per-capita shipments of oxycodone and hydrocodone pill shipment to retail pharmacies, calculated at the county-level in the ARCOS database. Our secondary outcome was the annual age-adjusted opioid-related death rate per 100,000 population at the county-level as reported by the Centers for Disease Control. [28] Statistical analysis We conducted a DID analysis to compare differences in trends of PCPV before and after state-level physical therapy direct access policy changes using county and year fixed effects, where the estimate of interest is the interaction term between county and year fixed effects. DID analysis enables us to estimate the causal effect of policy changes on the treated states under the assumption their pre-trends would have continued parallel to the control group had the policy not been enacted. [24] Our primary analysis used a staggered DID estimator allowing us to account for differential timing (2010 and 2013) of physical therapy direct access policy changes. [25] We used four pre-periods and two post-periods in the analysis since there are only two post periods for the 2013 policy change. Standard errors were clustered at the county level in the primary analysis, the year before treatment served as the baseline, and regressions were weighted by state population. [32] In DID analyses, adjustment for covariates is used to relax the parallel trends assumption. [24] This differs from randomized controlled trials where covariate adjustment is used to increase precision of the estimate or observational regression studies where covariate adjustment is used to attempt to control for confounding. [33,34] Since our pre-trends appeared to be parallel based on event study plots, we consider the unadjusted model as the primary analysis. We conducted a DID analysis adjusting for continuous variables including age, income, uninsurance rates, and dual eligibility rates and another DID analysis with standard errors clustered at the state-level as sensitivity analyses. Our secondary analysis used the same methods with opioid related deaths as the outcome. To further check the robustness of our results, we conducted a synthetic control analysis to compare Hawaii (policy changed in 2010) to a synthetic Hawaii created using weights derived from the never treated states. [35] Synthetic control is more appropriate with one exposed state therefore, we used Hawaii as the treatment group since it is the only state to have changed policy in 2010. This method assigns weights to each control state; weights are determined via a data-driven optimization algorithm (root-mean square prediction error) to minimize pre-period differences in outcomes between treated and control states. The weighted sum of control states is then referred to as a synthetic control (hereafter “synthetic Hawaii”). All statistical analyses were conducted using Stata, version 18.0. The data is at the county-level and without any ability to identify individuals and was accessed for research purposes on February 19, 2024. Results The treatment group consisted of 2,268 counties which had a plurality aged 46–64 years (28.0%), were split on gender (50.0% female), and were majority White (91.0%) (Table 1). The control group consisted of 7,555 counties with a plurality aged 45–64 years (27.6%) and a majority that were female (50.4%) and White (82.9%). There was a greater proportion of dual-eligible (12.6% vs. 11.2%), uninsured individuals under 65 years old (20.0% vs. 16.3%), and lower per-capita income ($32,521 vs. $37,372) in the control group than the treatment group. Baseline PCPV (29.9 vs. 27.3) and ORD (27.2 vs. 16.9) was higher in the treatment group than the control group. 10.1371/journal.pone.0333292.t001Table 1Baseline Characteristics of Treatment and Control States.CharacteristicsTreatmentControlNumber of CountiesNo. (%)2,268 (23.1)7,555 (76.9)Age (%)25-44 years23.8%24.1%45-64 years28.0%27.6%65 + years15.9%15.8%Sex (%)Male49.96%49.76%Race (%)Asian2.7%0.8%Black2.6%12.5%White91.0%82.9%Other3.7%3.8%Ethnicity (%)Hispanic11.2%12.6%Insurance Status (%)Dual-Eligible3.2%4.4%Uninsured under 6516.3%20.0%Per-capita income ($)Mean (SD)$37,372 ($10,287)$32,520.70 ($7,712)Opioid PCPVMean (SD)29.9 (14.7)27.3 (14.7)Opioid-related deaths (per 100,000)Mean (SD)27.2 (79.2)16.9 (59.3) Baseline Characteristics of Treatment and Control States. Changes in opioid prescription volume Prior to the direct access physical therapy policy change, trends in PCPV were similar in the treatment group compared to the control group. (Fig 1 and Fig 2) In the DID analysis, we found that PCPV significantly decreased by −4.6 (−6.2 to −3.0) percentage points after the direct access physical therapy policy change, which is a 15.4% relative reduction compared to baseline. (Table 2) Our sensitivity analyses including covariate-adjusted DID models, state-level clustering models, and synthetic control analyses for Hawaii were consistent with primary analysis results showing a reduction in PCPV after change in physical therapy direct access. (S1 and S2 Tables, S1 and S2 Figures in S1 File) 10.1371/journal.pone.0333292.t002Table 2Staggered Difference-in-Differences estimates of direct access to physical therapy on PCPV and ORD of treated states.PCPVORDUnadjustedAdjustedUnadjustedAdjustedATT (95% CI)−4.6(−6.2 to −3.0)−5.02(−6.85 to −3.19)−2.3(−7.2 to 2.6)−8.81(−20.42 to 2.79)P-Values<0.001<0.0010.40.137Notes: ATT: Average treatment on the treated, PCPV: per-capita per volume of opioid pill shipments, ORD: opioid-related deaths. Difference-in-Difference models were estimated using CSDID package in STATA. Standard errors clustered at the county level. Adjusted estimates were adjusted for age, income, uninsurance, and dual eligibility. Staggered Difference-in-Differences estimates of direct access to physical therapy on PCPV and ORD of treated states. Notes: ATT: Average treatment on the treated, PCPV: per-capita per volume of opioid pill shipments, ORD: opioid-related deaths. Difference-in-Difference models were estimated using CSDID package in STATA. Standard errors clustered at the county level. Adjusted estimates were adjusted for age, income, uninsurance, and dual eligibility. Unadjusted staggered difference-in-differences estimates of the average treatment effect of physical therapy direct access policy on the PCPV of the treated states. Notes: Difference-in-Difference models were estimated using CSDID package in STATA. Dots represent the point estimate and shaded bars represent the 95% confidence intervals. Since physical therapy direct access policy was implemented in different years, the periods to treatment represent different years depending on the treated state. All estimates are derived using the period before baseline to improve interpretation of the figure. PCPV is per-capita per volume of opioid pill shipments. Standard errors clustered at the county level. Adjusted staggered difference-in-differences estimates of the average treatment effect of physical therapy direct access policy on the PCPV of the treated states. Notes: Difference-in-Difference models were estimated using CSDID package in STATA. Dots represent the point estimate and shaded bars represent the 95% confidence intervals. Since physical therapy direct access policy was implemented in different years, the periods to treatment represent different years depending on the treated state. All estimates are derived using the period before baseline to improve interpretation of the figure. PCPV is per-capita per volume of opioid pill shipments. Standard errors clustered at the county level. ORD outcome Prior to direct access physical therapy policies, trends in ORDs were similar between treatment and control states. (Fig 3 and Fig 4) We found no statistically significant difference in ORDs (−2.3, 95% CI: −7.2 to 2.6) between treated and control states after enacting direct access physical therapy policies. Unadjusted and adjusted staggered Difference-in-Differences estimates of the average treatment effect of physical therapy direct access policy on the ORD of the treated states. Notes: Difference-in-Difference models were estimated using CSDID package in STATA. Dots represent the point estimate and shaded bars represent the 95% confidence intervals. Since physical therapy direct access policy was implemented in different years, the periods to treatment represent different years depending on the treated state. All estimates are derived using the period before baseline to improve interpretation of the figure. ORD is opioid related deaths. Standard errors clustered at the county level. Adjusted staggered Difference-in-Differences estimates of the average treatment effect of physical therapy direct access policy on the ORD of the treated states. Notes: Difference-in-Difference models were estimated using CSDID package in STATA. Dots represent the point estimate and shaded bars represent the 95% confidence intervals. Since physical therapy direct access policy was implemented in different years, the periods to treatment represent different years depending on the treated state. All estimates are derived using the period before baseline to improve interpretation of the figure. ORD is opioid related deaths. Standard errors clustered at the county level. Discussion Reduced use of opioids is imperative in order to diminish the risks of addiction while providing guideline concordant care for musculoskeletal pain. [36] Our findings provide the first estimates using rigorous causal inference methods of the effect of direct access physical therapy policies on opioid use and mortality. We find that direct access to physical therapy significantly reduced PCPV but did not reduce ORDs. Several previous studies assessed the effect of direct access to physical therapy, however, the relevance of their findings for policymaking may be limited by selection bias due to unobserved confounders. [37,38] For example, direct access was associated with less visits, reductions in diagnostic imaging, and lower healthcare costs for low back pain compared to states without direct access to physical therapy. [39] However, the authors acknowledge that since they used logistic regression to compare differences between those with and without direct access to physical therapy, they were only able to establish association and not causation due to possible bias from unobserved confounders even after their adjustments (e.g., attitudes towards opioids where those who seek treatment from physical therapists would deny opioids even if they saw a medical doctor). By using DID methods, our findings add to the literature by providing rigorous quasi-experimental evidence on the effects of direct access physical therapy policies while accounting for both observed and time-invariant unobserved confounders. Increasing direct access to physical therapy can play a pivotal role in addressing both the opioid and musculoskeletal pain public health crises. Our findings strengthen the evidence indicating beneficial effects of physical therapists as first-line care providers to improve patient and public health. [13,14,40] As of 2024, only two states (Alabama and Mississippi) do not have direct access to physical therapy and rank amongst the highest for opioid dispensing rates and PCPV. [27,41] Our findings along with the current literature support consideration for direct access physical therapy policies to potentially reduce the burden of opioids. [14,42] Barriers remain for patients to access physical therapy beyond direct access policies including lack of insurance coverage for non-pharmacological care, limited number of physical therapists in their insurance networks, co-pay amounts greater for physical therapy than medical doctors, and lack of knowledge about the benefits and low risk of non-pharmacological compared to pharmacological care. [43,44] Limitations This is the first study to use a staggered DID method to assess direct access physical therapy policy and to our knowledge in musculoskeletal pain more broadly. Our findings were robust to several different sensitivity analyses including using a different novel causal inference methodology synthetic control that has been used to assess policy change in other health conditions. [45] Given Hawaii’s unique demographic characteristics, the synthetic control estimates provide the effect of physical therapy direct access policies for Hawaii which may not be generalizable to other states. However, given the alignment of the synthetic control results with the DID analyses, it provides further validity of the robustness of our findings. Our study design is observational and we cannot rule out that outcome trends would have diverged between treated and control states in the absence of policy changes. However, our pre-trends appear to be parallel between treatment and control states which strengthens our belief that the parallel trends assumption holds in our analysis. [24] While the DID analysis with county and year fixed effects helps account for common temporal shocks, differential adoption of other opioid related policies (e.g., pill mill laws and prescribing limits) may confound our estimates. Additionally, the treatment states have higher PCPV and ORD at baseline than control groups, which could bias the DID analysis if trends differed. However, trends between treated and control states during the periods before treatment were parallel suggesting that they would continue parallel if it had not been for the physical therapy direct access policy change. Another limitation is the outcome of PCPV may not be generalizable to opioid prescriptions, real-world opioid utilization or misuse. Our data also does not distinguish between appropriate and inappropriate opioid prescriptions or physical therapy use, therefore our PCPV findings are limited in that they cannot be interpreted as reducing inappropriate opioid prescribing behavior. Since our data do not include opioid prescription volume or use, future research is needed to see if these results translate to opioid prescriptions or use and other low value treatments for musculoskeletal pain (e.g., surgery or injections). Mortality rates can require large sample size to properly power analyses to identify statistically significant effects due to their rarity. ORD outcomes in our analysis have 95% confidence intervals that cannot rule out that the null effect is due to low power and not a lack of effect of physical therapy direct access policy on ORD. Our analysis does not account for cross-state spillovers, which could bias treatment effects toward the null if patients in control states use services in neighboring treated states or vice versa. Finally, physical therapy direct access policies can be heterogenous (e.g., amount of physical therapy visits restricted) which may limit the generalizability of our findings. Conclusion In this study, direct access physical therapy policies significantly reduced PCPV. This is the first study to use a difference-in-differences method to estimate the effects of direct access physical therapy policy, confirming previous studies that have provided correlational estimates. Our findings highlight the importance of access to physical therapy as a key contributor to improving opioid and musculoskeletal pain public health crises. Supporting information The Impact of Physical Therapy Direct Access Policy on Opioid Shipments: A Difference-in-Differences Analysis. (DOCX) 28 Jan 2026 -->PONE-D-25-49217-->-->The Impact of Physical Therapy Direct Access Policy on Opioid Shipments and Opioid-related Deaths: A Difference-in-Differences Analysis-->-->PLOS One Dear Dr. Griffith, Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process. Please submit your revised manuscript by Mar 14 2026 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file. Please include the following items when submitting your revised manuscript:--> A letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'. A letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'. A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'. A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'. An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'. An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'. If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter. If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: https://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols. Additionally, PLOS ONE offers an option for publishing peer-reviewed Lab Protocol articles, which describe protocols hosted on protocols.io. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols. We look forward to receiving your revised manuscript. Kind regards, Francesca Baratta, PharmD, PhD Academic Editor PLOS One Journal Requirements: When submitting your revision, we need you to address these additional requirements. 1.Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf 2. Please note that your Data Availability Statement is currently missing the repository name and/or the DOI/accession number of each dataset OR a direct link to access each database. If your manuscript is accepted for publication, you will be asked to provide these details on a very short timeline. We therefore suggest that you provide this information now, though we will not hold up the peer review process if you are unable. 3.If the reviewer comments include a recommendation to cite specific previously published works, please review and evaluate these publications to determine whether they are relevant and should be cited. There is no requirement to cite these works unless the editor has indicated otherwise. [Note: HTML markup is below. Please do not edit.] Reviewer's Responses to Questions -->Comments to the Author 1. Is the manuscript technically sound, and do the data support the conclusions? The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented. --> Reviewer #1: Yes Reviewer #2: No Reviewer #3: Yes ********** -->2. Has the statistical analysis been performed appropriately and rigorously? --> Reviewer #1: Yes Reviewer #2: Yes Reviewer #3: Yes ********** -->3. Have the authors made all data underlying the findings in their manuscript fully available? The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.--> Reviewer #1: Yes Reviewer #2: Yes Reviewer #3: No ********** -->4. Is the manuscript presented in an intelligible fashion and written in standard English? PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.--> Reviewer #1: Yes Reviewer #2: Yes Reviewer #3: Yes ********** -->5. Review Comments to the Author Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)--> Reviewer #1: (Suggestions for further Improvement) Abstract It is only wise that it expressly states that the policy exposure is measured at the state level, whereas the outcomes are measured at the county level. More so, a brief definition of the term per-capita volume of opioid pills in the form of annual per-capita shipments of oxycodone and hydrocodone pills to retail pharmacies should be given. Introduction The introduction of the knowledge gap can be made sharper; that is, one should add one or two sentences to the Introduction that will clearly indicate that no previous study has exploited inter-state variation in direct access physical therapy laws to measure the causal impact of the latter on the volumes of opioid pills shipped and opioid-related mortality. Methods (Data sources & exposure) 1. The analysis clearly indicates that the treated and control states of county-year observations are clustered, with standard errors being clustered at the county level; sensitivity analyses are also done by clustering at the state level. 2. The selection of only Hawaii, Kansas, California, and Indiana as treated states was based on each one of them implementing the policy of direct-access in the study period and on similar demographic and economic factors; the eight never-treated states were chosen as a control group to achieve matching on these factors, and any states that implemented a partial access policy, or whose timing did not coincide, or which had incomplete data were excluded. 3. A brief description is provided of the heterogeneity of direct-access policies, which include free access to the policy, procedures with visitor time restrictions, etc.; this diversity can affect the scale of the estimated impacts and constrain the extrapolation of the policy findings. Outcomes ORD outcome is measured in 100,000 inhabitants per 100,000 inhabitants, but these values are not directly indicated in the manuscript as being age-adjusted or unadjusted. The latter methodological ambiguity should be tackled because the age distributions of different counties may differ significantly, thus affecting the comparability of the reported rates. Statistical analysis 1. The Difference-in-Differences (DiD) model specification could be improved in terms of increased transparency, specifically in the clear definition of fixed-effects structures (e.g., state x year versus county x year) and the clear description of how time-varying covariates (including more than the usual controls of age, income, uninsurance, and dual eligibility) are operationalized (continuous or categorical). 2. A limitation of the study is that cross-state spillovers (like cross-border prescribing or mortality) may be present in the study. In addition, the synthetic-control approach is overly brevily treated, with more information given about how the pre-policy fit is assessed (as e.g. by root-mean-square prediction error), the use of either placebo-tests or permutation tests, and how the temporal window in which the synthetic control has been constructed and used is to be specified. Results 1. The differences in the baselines in the treatment and control states, including high PCPV and ORD value, differences in income and coverage of insurance, are accepted but not discussed in terms of the possible impact on cause interpretation or external validity. It would therefore be wise to give a brief evaluation on whether the treated states are highly-burdened contexts. 2. The results are typically presented clearly, but the specifics that are of interest are only located in the appendix (e.g., synthetic control results, sensitivity analysis). In the primary Results section, it would be a good idea to summarise how these sensitivity estimates match (i.e. concordant or divergent) the major findings of DID. 3. In the case of ORD, the specified non-significant effect with the wide confidence interval that includes both substantive decreases and increases is reported, yet not placed in a relative change context or in a statistical power setting. Discussion 1. The null effect of opioid-related mortality should be described in greater detail: the shortness of the follow-up, the research involving a high number of other mortality drivers, and the potential lack of statistical power in the form of county-level mortality results should be mentioned. 2. The impediments to PT access and the suggestion that the remaining states should emulate direct-access policies should be presented with more caution, clearly identifying those findings that are directly elucidated by this analysis as opposed to the conclusions made by the rest of the literature and the historical background of such policies. Limitations (Additional limitations to consider) The possibility of discrepancies between the distribution of the tablets with opioids and the trend of the real usage or intake of the prescription is real. In addition, the present study fails to involve data related to non-prescribed opioid use or poly-substance abuse. References It is seen that some of the web-based sources and Google Scholar entries, including the Washington Post database of ARCOS and the AHRF, are currently quoted by using only the path scholar.google.com and not by mentioning original reports or technical documents. To be on the safe side and assure verifiability in the long term, it would be wise to include direct quotes of the main sources. Reviewer #2: A good study has been conducted to control the use of opioids, but due to the differences in the need for opioids in different types of diseases, the need for physical therapy, the existence of specific surgeries, etc., it has not been considered. As a result, this comparison, in my opinion, cannot be a source of good judgment and decision-making. Reviewer #3: The research question is important and methods are generally appropriate. However, several core issues need clarification before the manuscript is suitable for publication. 1. PCPV outcome validity needs clarification (Approx. lines 102–110): PCPV reflects ARCOS pill shipments, not prescriptions or use. This limits interpretability. The authors should clearly justify why this metric is an appropriate proxy and strengthen the limitations. 2. Unadjusted DID as the “primary” model is not convincing (Approx. lines 117–121): Given major state-level changes during 2006–2014 (Medicaid expansion, PDMP rollout, ACA), relying on unadjusted DID is questionable. Authors should either justify this more strongly or elevate adjusted analyses. 3. Noise in ORD outcome (Approx. lines 168–177): ORD varies widely and is unstable at the county level. Mortality rates are highly noisy. Consider smoothing (e.g., 3-year averages) or acknowledge this more directly as a limitation. 4. Exposure variable too simplified (Approx. lines 93–100): Direct access laws vary across states (limited, provisional, unrestricted). Treating exposure as purely binary may oversimplify policy differences. Clarify or justify. 5. Synthetic control applied only to Hawaii is very limited (Approx. lines 124–131): Hawaii has unique demographic/geographic characteristics. Results from a single treated unit should not be generalized. Needs cautionary interpretation. 6. Missing discussion of co-occurring opioid policies: During this period, states implemented multiple opioid regulations (PDMP, pill mill laws, prescribing limits). These could confound effects unless addressed. At minimum, acknowledge as a limitation. ********** -->6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files. If you choose “no”, your identity will remain anonymous but your review may still be made public. Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.--> Reviewer #1: Yes:MOHD ASIF Reviewer #2: No Reviewer #3: No ********** [NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.] To ensure your figures meet our technical requirements, please review our figure guidelines: https://journals.plos.org/plosone/s/figures You may also use PLOS’s free figure tool, NAAS, to help you prepare publication quality figures: https://journals.plos.org/plosone/s/figures#loc-tools-for-figure-preparation. NAAS will assess whether your figures meet our technical requirements by comparing each figure against our figure specifications. Submitted filename: Report-PONE-D-25-49217.docx 20 Mar 2026 Please see attachment. Submitted filename: PLOS ONE Response Letter 2026 03 14.docx 8 May 2026 The Impact of Physical Therapy Direct Access Policy on Opioid Shipments and Opioid-related Deaths: A Difference-in-Differences Analysis PONE-D-25-49217R1 Dear Dr. Griffith, We’re pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it meets all outstanding technical requirements. Within one week, you’ll receive an e-mail detailing the required amendments. When these have been addressed, you’ll receive a formal acceptance letter and your manuscript will be scheduled for publication. An invoice will be generated when your article is formally accepted. Please note, if your institution has a publishing partnership with PLOS and your article meets the relevant criteria, all or part of your publication costs will be covered. Please make sure your user information is up-to-date by logging into Editorial Manager at Editorial Manager® and clicking the ‘Update My Information' link at the top of the page. For questions related to billing, please contact billing support. If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. If they’ll be preparing press materials, please inform our press team as soon as possible -- no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org. Kind regards, Francesca Baratta, PharmD, PhD Academic Editor PLOS One Reviewers' comments: Reviewer's Responses to Questions -->Comments to the Author 1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation.--> Reviewer #1: (No Response) Reviewer #3: All comments have been addressed ********** -->2. Is the manuscript technically sound, and do the data support the conclusions? The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented. --> Reviewer #1: Yes Reviewer #3: Yes ********** -->3. Has the statistical analysis been performed appropriately and rigorously? --> Reviewer #1: Yes Reviewer #3: Yes ********** -->4. Have the authors made all data underlying the findings in their manuscript fully available? The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.--> Reviewer #1: Yes Reviewer #3: Yes ********** -->5. Is the manuscript presented in an intelligible fashion and written in standard English? PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.--> Reviewer #1: Yes Reviewer #3: Yes ********** -->6. Review Comments to the Author Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)--> Reviewer #1: (No Response) Reviewer #3: The authors have adequately addressed all previous reviewer comments and concerns. The revisions have improved the clarity and overall quality of the manuscript, and I have no further major concerns. I recommend acceptance. ********** -->7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files. If you choose “no”, your identity will remain anonymous but your review may still be made public. Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.--> Reviewer #1: Yes:Mohd Asif Reviewer #3: No ********** PONE-D-25-49217R1 PLOS One Dear Dr. Griffith, I'm pleased to inform you that your manuscript has been deemed suitable for publication in PLOS One. Congratulations! Your manuscript is now being handed over to our production team. At this stage, our production department will prepare your paper for publication. This includes ensuring the following: * All references, tables, and figures are properly cited * All relevant supporting information is included in the manuscript submission, * There are no issues that prevent the paper from being properly typeset You will receive further instructions from the production team, including instructions on how to review your proof when it is ready. Please keep in mind that we are working through a large volume of accepted articles, so please give us a few days to review your paper and let you know the next and final steps. Lastly, if your institution or institutions have a press office, please let them know about your upcoming paper now to help maximize its impact. If they'll be preparing press materials, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org. You will receive an invoice from PLOS for your publication fee after your manuscript has reached the completed accept phase. If you receive an email requesting payment before acceptance or for any other service, this may be a phishing scheme. Learn how to identify phishing emails and protect your accounts at https://explore.plos.org/phishing. If we can help with anything else, please email us at customercare@plos.org. Thank you for submitting your work to PLOS ONE and supporting open access. Kind regards, PLOS ONE Editorial Office Staff on behalf of Dr. Francesca Baratta Academic Editor PLOS One