Virtual Care and Medical Cannabis Access: A Geospatial Study of Telehealth’s Role in Reducing Socioeconomic Disparities

Abstract

Introduction:

Telehealth has the potential to improve health care access and reduce disparities. We examined whether the density of medical cannabis (MC) patients, stratified by those who were seen by a telemedicine provider or not, is associated with a concentrated disadvantage within Pennsylvania in 2022.

Methods:

This zip code-level analysis assessed associations between the concentrated disadvantage index (CDI) and counts of telemedicine-approved and all other MC patients. Total MC patient counts were obtained from Pennsylvania’s Department of Health, counts of telemedicine-approved MC patients were obtained from a telehealth provider (Leafwell), and CDI data were obtained from the 2022 American Community Survey. Multivariable negative binomial regression models with population offsets and robust standard errors were used, accounting for spatial autocorrelation through spatial lag adjustments.

Results:

The CDI was not associated with the number of telemedicine-approved MC patients (incidence rate ratios [IRR] = 0.962; p = 0.355) but was significantly negatively associated with all other MC patients (IRR = 0.904; p = <0.001). The number of in-office MC providers was significantly associated with the count of all other MC patients but not with telemedicine-approved patients. Spatial factors significantly influenced the distribution of both patient groups.

Discussion:

These findings suggest that telemedicine may play a crucial role in reducing access disparities for MC in socioeconomically disadvantaged areas. The lack of a significant association between CDI and telemedicine-approved MC patients indicates that telehealth services can overcome barriers such as provider scarcity and transportation issues. By facilitating remote consultations and approvals, telemedicine expands access for patients who might otherwise be unable to obtain MC.

Introduction

MEDICAL CANNABIS IN THE UNITED STATES: POLICIES AND PROMISES UNKEPT

In the United States, 38 states and the District of Columbia have legalized cannabis for medical use.¹ All states where cannabis is medically available have developed state-administered medical cannabis (MC) programs. State regulatory frameworks for MC vary, influencing which conditions qualify for treatment and the demographic profiles of users.^2,3 Prospective patients must obtain certification from a qualified physician verifying their diagnosis of one of these conditions to become eligible for MC use. As of 2022, there were nearly 4.1 million registered MC patients in the United States, a 33% increase from the previous year.²

Current research suggests that MC has beneficial properties.⁴ While it is not without risks, MC has been shown to assist in the discontinuation of high-risk medications.^5

–11 It has been shown to increase sleep quality and quality of life for people suffering from various conditions, including chronic pain, posttraumatic stress disorder, and epilepsy.^{12

–23} Examining the impact of medical and recreational cannabis laws through a societal lens shows evidence that MC laws do not generate negative outcomes in the labor market or lead to increased criminal activity.²⁴ However, the increased access to MC resulting from policy reform, and thus its potential benefits, have not been realized equally across racial and ethnic groups.^25,26

Despite the recent push toward making cannabis legal for medical and recreational purposes within the United States, the harms surrounding cannabis criminalization continue to disproportionately impact Black individuals and Hispanic individuals. A report from the American Civil Liberties Union notes that the war on cannabis continues despite law changes.²⁵ This report, and others specific to NY,²⁷ find persistent racial disparities in cannabis possession arrests in states that legalized cannabis.

These disparities exist despite policymakers making concerted efforts to pass cannabis reform legislation that includes diversity requirements in MC and adult-use laws. These measures aim to address social equity by ensuring that communities disproportionately affected by previous cannabis prohibition have opportunities to participate in the legal cannabis market.^28,29 But the efforts to implement diversity requirements in MC and adult-use cannabis laws have shown mixed progress as the legislative process has been slow, with many bills still awaiting committee approval.³⁰ The pace of legislative action and the complexity of regulatory frameworks continue to challenge the full realization of these diversity goals.

TELEHEALTH IN THE UNITED STATES

Telehealth has been recognized for its potential to deliver efficient and cost-effective care by reducing health care costs through decreased medication misuse, unnecessary emergency department visits, and prolonged hospitalizations.^31
–33 Patient preferences indicate a high likelihood of using telehealth for medication refills, preparing for visits, reviewing test results, and receiving education, showcasing its effectiveness in various health care services.³² Furthermore, telehealth interventions have shown promise in managing chronic conditions, suggesting potential long-term benefits for patient outcomes and health care cost reduction.^34,35

Telehealth also plays a crucial role in increasing access and alleviating disparities in health care. It offers significant benefits for socially or economically disadvantaged populations, such as those in rural areas, who face greater barriers to accessing traditional in-person care.^32,36 However, disparities in telehealth utilization persist, particularly among racial and ethnic minorities, individuals with low socioeconomic status, and those with limited technological access.^33,36,37 For instance, rural residents often report higher travel burdens and system-level barriers to accessing primary care.³¹

MC ACCESS AND TELEHEALTH

To date, there has been only one study that has examined the distribution of MC providers. Cunningham et al. examined the geographic distribution of in-office MC certifying providers in New York.³⁸ The study found that for every 10% increase in the percentage of Black residents, neighborhoods were 5% less likely to have at least one in-office MC provider. Conversely, they found that for every 10% increase in the percentage of residents with a bachelor’s degree or more, neighborhoods were 30% more likely to have at least one in-office MC provider. These findings suggest that in-office qualified providers are unevenly distributed across socioeconomic groups.

Telehealth offers a potential solution to address the uneven geographic distribution of MC providers. Its use has recently increased in popularity.^39,40 Currently, Rhode Island, South Dakota, and Utah do not allow for telehealth consultations for qualifying a patient for MC. Although limited research compares the two consultation modes, one study found similar effectiveness between telehealth and in-office provider consultations for chronic pain patients.⁴⁰

CURRENT CONTRIBUTION

We sought to understand whether telehealth consultation services addressed existing disparities in MC access found in a Cunningham et al. study of NY. To do so, we explored the relationship between telemedicine-approved MC patients, MC patients overall, and concentrated disadvantage across Pennsylvania zip codes in 2022, accounting for spatial distribution and in-office MC providers. Pennsylvania legalized MC on April 17, 2016. The first licensed sales occurred on February 15, 2018. As of July 2023, the Pennsylvania Department of Health (PA-DoH) announced a total of 942,231 registered patients and caregivers.

Methods

This cross-sectional study used publicly available 2022 data from PA-DOH, proprietary Leafwell data, and geospatial data from the National Historical Geographic Information System (NHGIS). The analysis was conducted at the zip code level, examining the associations between concentrated disadvantage and counts of patients, controlling for the density of in-office certifying medical providers.

STUDY VARIABLES

We examined two primary outcomes: counts of telemedicine-approved MC patients from Leafwell and counts of all other MC patients by zip code. We identified telemedicine-approved patients through the Leafwell Patient Database (LPD). Leafwell operates in 36 states and advertises on internet search engines and digital media to connect potential MC patients with physicians in their state. After a physician deems a patient qualified for MC, Leafwell assists patients in obtaining their medical card. Leafwell patients are asked to fill out a baseline questionnaire, providing their zip code of primary residence.

To achieve our research aims, we accessed and analyzed LPD data from January 1, 2022, to December 31, 2022. We identified all patients approved in Pennsylvania in 2022 and obtained their zip code information. For patient confidentiality, only de-identified data from the LPD were shared with internal researchers. Researchers did not have access to nonanonymized data. This project received exempt status from an external, third-party institutional review board (IRB), BRANY (IRB Number: IRB00000080). Patients consented to the use of their questionnaire data in aggregate form as part of the Leafwell terms of service.

Data on all other MC patients in Pennsylvania were obtained through a release resulting from the court case Department of Health v. Spotlight PA, Commonwealth Court of Pennsylvania, No. 660 C.D. 2021.⁴¹ Spotlight sued the PA-DOH for de-identified data related to MC patients. As a result, the PA-DOH provided anonymized MC patient data for 2017–2022 to Spotlight, which subsequently published the data. We downloaded the Spotlight data for 2022, which provided the zip code of patients.⁴² We aggregated both sets of patient data (LPD and PA-DOH) to the zip code level. We then subtracted the counts of telemedicine-approved patients from the counts of total Pennsylvania patients to create the outcome variable for all other MC patients.

Key independent variables include the concentrated disadvantage index (CDI), counts of in-office MC providers, and the percentage of the White, non-Hispanic population for each zip code. GIS zip code boundaries, as well as zip code-specific demographic data necessary to compute the CDI, were downloaded from the Integrated Public Use Microdata Series (IPUMS-NHGIS).⁴³ We downloaded data for the year 2022. The IPUMS-NHGIS provides summary tables of the 2022 American Community Survey (5-year average 2018–2022). We did not use data from 1- or 3-year estimates, as these are not available at the zip code level.

We defined the CDI variable following established literature.⁴⁴ Items were combined into an index by taking the average of their z-scores, per established literature. A higher value of the CDI variable indicates a more concentrated disadvantage within a zip code. We also included in the analysis a variable representing the percentage of the population that was White, non-Hispanic. White, non-Hispanics have been shown to utilize MC at higher rates compared with other races.^3,45

We obtained counts of in-office MC providers from the State of Pennsylvania’s MC program.⁴⁶ This information is publicly available through the PA-DOH. We downloaded the information, cleaned, and aggregated the counts of in-office MC providers to the zip code level for the analysis.

STATISTICAL APPROACH

The unit of analysis was the zip code level. To examine whether concentrated disadvantage was different for telemedicine-approved patients versus all other patients, we first conducted a multivariable negative binomial regression in which all independent variables were included in the model. We also conducted an analysis to estimate whether spatial autocorrelation impacted our findings. To achieve this, we first defined the spatial relationships between the observations, creating a spatial weight matrix using the four nearest neighbors. We then conducted Moran’s I tests for our two primary outcomes to test for autocorrelation. We extracted the residuals from the initial regression model, calculated a spatial lag, and fit it into the spatial regression model, which included the independent variables and lagged residuals to account for spatial dependence. For both the initial model and the spatial regression model, we included a zip code-specific population offset to express our estimates as incidence rate ratios (IRR). Both models also included standard errors clustered at the zip code level to adjust for potential correlation within clusters. All analyses were conducted using R version 4.3.1.⁴⁷ The multivariable negative binomial regression models were conducted using the glm.nb function in the R package MASS.⁴⁸

Results

Of the 2,167 total zip codes associated with Pennsylvania, we identified 1,458 standard zip codes, excluding all PO boxes and unique zip codes.⁴⁹ Of the 1,458 zip codes, 53 had a population of zero. Therefore, we excluded these zip codes for a total of 1,405 in the analyses.

We provided figures examining the distribution of telemedicine-approved MC patients, all other approved MC patients, in-clinic MC providers, and the CDI, respectively, in Figs. 1–4. Each figure provides zip code-level information for Pennsylvania, specifically highlighting Philadelphia and Pittsburgh. Figs. 1 and 2 show some overlap between telemedicine approved and all other patients, yet there is a wider overall distribution of the latter.

Fig. 1.

Distribution of telemedicine-approved patients within Pennsylvania in 2022. Light gray indicates no patients were present in the zip code.

Fig. 2.

Distribution of all other patients within Pennsylvania in 2022. There was at least one medical cannabis patient in each zip code.

Fig. 3.

Distribution of qualified medical providers within Pennsylvania in 2022. Light gray indicates no providers were present in the zip code.

Fig. 4.

Distribution of concentrated disadvantage index within Pennsylvania in 2022. Light gray indicates no providers were present in the zip code.

Fig. 3 illustrates the CDI at the zip code level. The map reveals that areas with higher CDI scores are located across Pennsylvania, with some concentrations in urban centers like Philadelphia and Pittsburgh. The presence of high CDI values in both urban and rural areas indicates pockets of socioeconomic disadvantage across diverse geographic regions. Fig. 4 shows the distribution of qualified in-clinic medical providers across zip codes. This figure highlights that the availability of qualified providers is unevenly distributed across the state, with significant concentrations in urban areas, particularly around Philadelphia and Pittsburgh. Rural areas and smaller towns show fewer providers, which might limit access to in-office services in those regions.

Table 1 presents the results of Moran’s I statistics, including Moran’s I, expectation value, variance, and p-value for both primary outcomes. Results suggest that the geographic distribution of both telemedicine-approved and all other patients is not random but rather exhibits a significant pattern of spatial clustering.

Table 1.

Moran’s I Statistic for Primary Outcomes, Telemedicine-Approved Patients, and All Other Patients

ZIP CODE-LEVEL VARIABLES	MORAN’S I	EXPECTATION	VARIANCE	p-VALUE
Telemedicine-approved patients	0.348813765	−0.00071225	0.0003136	<0.001
All other patients	0.350858382	−0.00071225	0.0003152	<0.001

In Table 2, we see that the CDI does not have a statistically significant association with the number of telemedicine-approved patients. Without spatial lags, the IRR for CDI is 0.988 (95% confidence interval [CI]: 0.881–1.107; p = 0.831), indicating no meaningful effect. When spatial lags are included, the IRR is slightly lower at 0.962 (95% CI: 0.885–1.045; p = 0.355) but still not significant. The number of providers does not significantly affect the number of telemedicine-approved patients in either model. The IRR is close to 1 in both cases (IRR = 1.005, p = 0.344 without spatial lags; IRR = 1.007, p = 0.129 with spatial lags). The inclusion of spatial lags in the model is significant, with an IRR of 1.040 (95% CI: 1.037–1.043; p = <0.001).

In Table 3, we see that the CDI shows a significant negative association with the number of all other MC patients in both models. Without spatial lags, the IRR is 0.891 (95% CI: 0.839–0.947; p = <0.001), indicating that higher CDI is associated with fewer non-telemedicine patients. This negative association persists when spatial lags are included, with an IRR of 0.904 (95% CI: 0.855–0.957; p = <0.001). Provider counts also show a significant positive association with the number of non-Leafwell patients in both models, in contrast to Table 2. The IRR is 1.009 without spatial lags (p = 0.017) and remains consistent with spatial lags included (IRR = 1.009, p = 0.010). The inclusion of spatial lags is significant in this model as well, with an IRR of 1.002 (95% CI: 1.002–1.003; p = <0.001).

Table 2.

Associations Between Total Telemedicine-Approved Medical Cannabis Patients and Concentrated Disadvantage Index Controlling for Qualified Medical Cannabis Providers, Percent Population White, Non-Hispanic, and Spatial Distribution

	WITHOUT SPATIAL LAGS			WITH SPATIAL LAGS
ZIP CODE-LEVEL VARIABLES	INCIDENCE RATE RATIO (IRR)	95% CONFIDENCE INTERVAL (CI)	p-VALUE	IRR	95% CI	p-VALUE
Concentrated disadvantage index	0.988	0.881–1.107	0.831	0.962	0.885–1.045	0.355
Provider counts	1.005	0.994–1.016	0.344	1.007	0.998–1.017	0.129
Percent White, non-Hispanic	1.194	0.918–1.555	0.187	0.929	0.749–1.152	0.503
Intercept	0.001	0.001–0.001	<0.001	0.001	0.001–0.002	<0.001
Spatial lags	—	—	—	1.040	1.037–1.043	<0.001

Models are negative binomial generalized linear models with cluster robust standard errors clustered at the zip code level with a population-level offset.

Table 3.

Associations Between All Other Medical Cannabis Patients and Concentrated Disadvantage Index Controlling for Qualified Medical Cannabis Providers, Percent Population White, Non-Hispanic, and Spatial Distribution

	WITHOUT SPATIAL LAGS			WITH SPATIAL LAGS
ZIP CODE-LEVEL VARIABLES	INCIDENCE RATE RATIO (IRR)	95% CONFIDENCE INTERVAL (CI)	p-VALUE	IRR	95% CI	p-VALUE
Concentrated disadvantage index	0.891	0.839–0.947	<0.001	0.904	0.855–0.957	<0.001
Provider counts	1.009	1.002–1.016	0.017	1.009	1.002–1.017	0.010
Percent White, non-Hispanic	1.040	0.886–1.221	0.632	0.835	0.726–0.961	0.012
Intercept	0.019	0.017–0.022	<0.001	0.023	0.02–0.026	<0.001
Spatial lags	—	—	—	1.002	1.002–1.003	<0.001

Models are negative binomial generalized linear models with cluster robust standard errors clustered at the zip code level with a population-level offset.

Discussion

Using a combination of publicly available data from PA-DOH, proprietary data from the LPD, and geospatial analyses, we examined how CDI, in-office MC provider counts, and racial composition influenced the distribution of these two patient groups, accounting for spatial autocorrelation. Our findings reveal significant differences in how socioeconomic factors and spatial distributions impact telemedicine-approved patients compared with all other patients. For telemedicine-approved patients, CDI did not show a statistically significant association with the number of patients, indicating that telemedicine access might mitigate some of the barriers typically associated with socioeconomic disadvantage. These findings suggest that telehealth services, specifically for MC patients, may be more equitably distributed across socioeconomic strata than in-office MC providers, as found in Cunningham et al. (2022).³⁸

On the other hand, CDI had a significant negative association with the distribution of all other patients, meaning that areas with higher socioeconomic disadvantage tended to have fewer other MC patients. This contrast underscores a crucial finding: telemedicine may serve as a critical tool in reducing disparities around obtaining a qualification for a medical card, particularly in socioeconomically disadvantaged areas where traditional, in-clinic services are less prevalent or harder to access.

In the context of the current literature, our findings align with the Cunningham et al. study, which found disparities along socioeconomic lines in the distribution of in-office MC providers.³⁸ Our research expands on this by differentiating between patients who obtained a medical card via a telemedicine provider and all other MC patients while controlling for the density of in-office qualified medical providers. Our findings, combined with Cunningham et al., suggest the benefits of MC may not be shared proportionately across different communities.

As states attempt to address the inequities perpetuated by the war on drugs through legislative means, this knowledge can help policymakers better understand ways of alleviating the existing disparities around MC service access. A recent article found racial and ethnic differences in cannabis use following legalization among U.S. states with MC laws. Martins et al. note that, while White, non-Hispanic, and Hispanic individuals saw cannabis use increase, Black, non-Hispanic individuals did not see similar gains. Part of the explanation for this difference may be the persistent disparities in cannabis arrest rates.²⁵ However, part of this difference may be due to differences in accessing MC services found in Cunningham et al. and further expanded on here.³⁸ Our findings suggest telemedicine services can alleviate some of the persistent disparities around access.

While our findings indicate that telemedicine can help alleviate some disparities in access to MC, the significant relationship between CDI and the distribution of all other patients suggests that broader health care inequalities still persist within Pennsylvania. Policymakers in Pennsylvania and policymakers in other states, should consider these disparities when crafting MC policies, ensuring that initiatives to expand telehealth do not replace but rather complement efforts to improve access to in-person care, particularly in areas with high levels of concentrated disadvantage.

Telehealth can address MC access challenges by connecting patients with cannabis-trained providers, regardless of location, thus improving access to expert advice in states with both medical and adult-use cannabis. It also helps bridge knowledge gaps among primary care providers by facilitating ongoing education and collaboration with specialists, ensuring cannabis use is safely integrated into treatment plans. Telehealth enables better communication and continuity of care, keeping all health care providers informed and up-to-date on patient cannabis use, which reduces unrecorded or unauthorized prescription medication substitution. This approach increases patient confidence, supports personalized care, and enhances overall safety and efficacy in cannabis therapy.

Limitations

Several limitations must be considered when interpreting these findings. Geospatial analyses, while powerful, have inherent limitations related to the accuracy of spatial data and the potential for ecological fallacies. The use of zip code-level data, in particular, may mask important variations within smaller geographic areas, leading to oversimplified conclusions about the relationship between CDI and patient distribution. Additionally, the creation of the all other patients variable, which subtracts Leafwell patients from the total patient population, does not account for individuals who may have used both telemedicine and in-person services, nor does it account for patients who may have used another telemedicine service. This could potentially distort the findings, particularly if a significant number of patients utilized another telemedicine approval company to obtain their medical card. Moreover, our study is cross-sectional, limiting our ability to infer causality from the observed associations. Longitudinal studies would be needed to fully understand how the relationship between CDI and patient distribution evolves over time, particularly as telehealth continues to expand.

Conclusions

In conclusion, this study underscores the potential of telemedicine to mitigate health care access disparities in the context of MC. While traditional in-office MC providers appear to be less accessible in these regions, telemedicine shows promise in reaching populations that might otherwise be underserved.

Footnotes

Authors’ Contributions

M.L.D.: Conceptualization, methodology, formal analysis, data curation, writing—original draft, writing—review and editing, visualization, project administration. M.K.: Writing—original draft, writing—review and editing. D.H.: Writing—original draft, writing—review and editing, data curation. E.F.: Project administration, writing—original draft, writing—review and editing. J.C.: Conceptualization, writing—original draft, writing—review and editing.

Data Availability

All publicly available data used as part of this study is available for use. Propriety data is not available.

Author Disclosure Statement

All authors are employers of Leafwell, a telemedicine company that connects medical cannabis patients to virtual health care providers.

Funding Information

No funding was received for this article.

References

National Conference of State Legislatures. State Medical Cannabis Laws [Internet]. 2024. Available from: https://www.ncsl.org/health/state-medical-cannabis-laws [Last accessed: March 11, 2024 ].

Boehnke

, Sinclair

, Gordon

, et al. Trends in U.S. Medical cannabis registrations, authorizing clinicians, and reasons for use from 2020 to 2022. Ann Intern Med, 2024; 177(4):458–466.

Fairman

. Trends in registered medical marijuana participation across 13 US states and District of Columbia. Drug Alcohol Depend, 2016; 159:72–79.

Committee on the Health Effects of Marijuana: An Evidence Review and Research Agenda, Board on Population Health and Public Health Practice, Health and Medicine Division, National Academies of Sciences, Engineering, and Medicine. The Health Effects of Cannabis and Cannabinoids: The Current State of Evidence and Recommendations for Research [Internet]. Washington, D.C.: National Academies Press; 2017. Available from: https://www.nap.edu/catalog/24625 [Last accessed: July 10, 2023 ].

Bradford

, Bradford

. The impact of medical cannabis legalization on prescription medication use and costs under Medicare part D. The Journal of Law and Economics [Internet], 2018; 61(3):461–487; doi: 10.1086/699620

Bradford

, Bradford

. Medical marijuana laws may be associated with a decline in the number of prescriptions for Medicaid enrollees. Health Aff (Millwood), 2017; 36(5):945–951.

Bradford

, Bradford

. Medical marijuana laws reduce prescription medication use in Medicare part D. Health Aff (Millwood), 2016; 35(7):1230–1236.

Bruce

, Grove

, Foster

, et al. Gender differences in medical cannabis use: Symptoms treated, physician support for use, and prescription medication discontinuation. J Womens Health (Larchmt), 2021; 30(6):857–863.

O’Connell

, Sandgren

, Frantzen

, et al. Medical cannabis: Effects on opioid and benzodiazepine requirements for pain control. Ann Pharmacother, 2019; 53(11):1081–1086.

10.

Purcell

, Davis

, Moolman

, et al. Reduction of Benzodiazepine use in patients prescribed medical cannabis. Cannabis Cannabinoid Res, 2019; 4(3):214–218.

11.

Charoenporn

, Charernboon

, Mackie

. Medical Cannabis as a substitute for prescription agents: A systematic review and meta-analysis. J Subst Use, 2023; 28(4):522–534.

12.

AminiLari

, Wang

, Neumark

, et al. Medical cannabis and cannabinoids for impaired sleep: A systematic review and meta-analysis of randomized clinical trials. Sleep, 2022; 45(2):zsab234.

13.

Bonn-Miller

, Brunstetter

, Simonian

, et al. The Long-Term, prospective, therapeutic impact of cannabis on post-traumatic stress disorder. Cannabis Cannabinoid Res, 2022; 7(2):214–223.

14.

Bruce

, Foster

, Shattell

. Perceived efficacy of medical cannabis in the treatment of Co-Occurring health-related quality of life symptoms. Behav Med, 2021; 47(2):170–174.

15.

Cahill

, Lunn

, Diaz

, et al. Evaluation of patient reported safety and efficacy of cannabis from a survey of medical cannabis patients in Canada. Front Public Health, 2021; 9:626853.

16.

Chan

, DeAngelis

, Ganesh

, et al. Medical cannabis use for patients with post-traumatic stress disorder (PTSD). J Pain Manage, 2017; 10(4):385–396.

17.

Drost

, DeAngelis

, Lam

, et al. Efficacy of different varieties of medical cannabis in relieving symptoms in post-traumatic stress disorder (PTSD) patients. J Pain Manage, 2017; 10(4):415–422.

18.

Harris

, Erridge

, Ergisi

, et al. UK Medical Cannabis registry: An analysis of clinical outcomes of medicinal cannabis therapy for chronic pain conditions. Expert Rev Clin Pharmacol, 2022; 15(4):473–485.

19.

Erridge

, Holvey

, Coomber

, et al. Clinical outcome data of children treated with cannabis-based medicinal products for treatment resistant epilepsy—analysis from the UK Medical cannabis registry. Neuropediatrics, 2023; 54(3):174–181.

20.

Nicholas

, Erridge

, Bapir

, et al. UK medical cannabis registry: Assessment of clinical outcomes in patients with headache disorders. Expert Rev Neurother, 2023; 23(1):85–96.

21.

Pillai

, Erridge

, Bapir

, et al. Assessment of clinical outcomes in patients with post-traumatic stress disorder: Analysis from the UK Medical cannabis registry. Expert Rev Neurother, 2022; 22(11–12):1009–1018.

22.

Erridge

, Kerr-Gaffney

, Holvey

, et al. Clinical outcome analysis of patients with autism spectrum disorder: Analysis from the UK Medical cannabis registry. Ther Adv Psychopharmacol, 2022; 12:20451253221116240.

23.

Mangoo

, Erridge

, Holvey

, et al. Assessment of clinical outcomes of medicinal cannabis therapy for depression: Analysis from the UK Medical cannabis registry. Expert Rev Neurother, 2022; 22(11–12):995–1008.

24.

French

, Zukerberg

, Lewandowski

, et al. Societal costs and outcomes of medical and recreational marijuana policies in the United States: A systematic review. Med Care Res Rev, 2022; 79(6):743–771.

25.

American Civil Liberties Union. A tale of two countries. Racially targeted arrests in the era of marijuana reform. 2020 [Internet]. 2020. Available from: https://assets.aclu.org/live/uploads/publications/marijuanareport_03232021.pdf [Last accessed: October 16, 2024 ].

26.

Martins

, Ghandour

, Lee

, et al. Sociodemographic and substance use correlates of gambling behavior in the Canadian general population. J Addict Dis, 2010; 29(3):338–351.

27.

Drug Policy Alliance and the Public Science Project. Inequitable marijuana criminalization, COVID-19 and socioeconomic disparities: The Case for Community Reinvestment in New York [Internet]. Available from: http://smart-ny.com/wp-content/uploads/2021/03/DPA_PSP_Marijuana_Disparities_COVID_3.16.2021.pdf [Last accessed: October 16, 2024 ].

28.

CT.gov. Policies and Procedures for Connecticut’s Adult-Use Cannabis Program. [Internet]. Available from: https://portal.ct.gov/cannabis/knowledge-base/articles/policies-and-procedures [Last accessed: October 14, 2024 ].

29.

Cannabis Regulatory Commission [Internet]. Cannabis Related Laws. Available from: https://www.nj.gov/cannabis/resources/cannabis-laws/. [Last accessed: October 14, 2024 ]

30.

Pa. marijuana laws: Here’s what you need to know [Internet]. WHYY. Available from: https://whyy.org/articles/pennsylvania-marijuana-laws-recreational-medical-legalization/ [Last accessed: October 14, 2024 ].

31.

Gajarawala

, Pelkowski

. Telehealth benefits and barriers. J Nurse Pract, 2021; 17(2):218–221.

32.

Ebbert

, Ramar

, Tulledge-Scheitel

, et al. Patient preferences for telehealth services in a large multispecialty practice. J Telemed Telecare, 2023; 29(4):298–303.

33.

Shaver

. The state of telehealth before and after the COVID-19 pandemic. Prim Care, 2022; 49(4):517–530.

34.

Omboni

, McManus

, Bosworth

, et al. Evidence and recommendations on the use of telemedicine for the management of arterial. Hypertension, 2020; 76(5):1368–1383.

35.

Tchero

, Kangambega

, Briatte

, et al. Clinical effectiveness of telemedicine in diabetes mellitus: A meta-analysis of 42 randomized controlled trials. Telemed J E Health, 2019; 25(7):569–583.

36.

Mahtta

, Daher

, Lee

, et al. Promise and perils of telehealth in the current Era. Curr Cardiol Rep, 2021; 23(9):115.

37.

Marcondes

, Normand

S-LT

, Le Cook

, et al. Racial and ethnic differences in telemedicine use. JAMA Health Forum, 2024; 5(3):e240131.

38.

Cunningham

, Zhang

, Hollins

, et al. Availability of medical cannabis services by racial, social, and geographic characteristics of neighborhoods in New York: A cross-sectional study. BMC Public Health, 2022; 22(1):671.

39.

DeWitt

. Cannabis and coronavirus: Impact on medical cannabis industries in three state [Internet]. 2020. Available from: https://europepmc.org/article/PPR/PPR244248 [Last accessed: October 16, 2024 ].

40.

Pankratz

. Impact of telehealth versus traditional visits on pain management patients utilizing medical cannabis during the COVID-19 pandemic [Internet] [PhD]. United States – Minnesota: Walden University; 2023. Available from: https://www-proquest-com-s.web.bisu.edu.cn/docview/2778660476/abstract/E9877EB46AB54A39PQ/1 [Last accessed: October 16, 2024 ].

41.

Mahon

. GET the data: Spotlight PA makes info on why patients qualify for medical marijuana publicly available [Internet]. Spotlight PA; 2023. Dec 20; Available from: https://www.spotlightpa.org/news/2023/12/pennsylvania-medical-marijuana-data-available/ [Last accessed: October 17, 2024 ].

42.

Mahon

. Spotlight-PA-marijuana-certifications [Internet]. Spotlight PA; 2023. Available from: https://github.com/ed-mahon/Spotlight-PA-marijuana-certifications [Last accessed: October 17, 2024 ].

43.

Manson

, Schroeder

, Van Riper

, et al. IPUMS National Historical Geographic Information System: Version 18.0 [dataset] [Internet]. IPUMS NHGIS: Minneapolis, MN; 2023; doi: 10.1812/8/D050.V18.0

44.

Jing

, Li

, Qiao

, et al. Investigating the relationships between concentrated disadvantage, place connectivity, and COVID-19 fatality in the United States over time. BMC Public Health, 2022; 22(1):2346.

45.

Mahabir

, Merchant

, Smith

, et al. Medical cannabis use in the United States: A retrospective database study. J Cannabis Res, 2020; 2(1):32.

46.

Pennsylvania Department of Health, Office of Medical Marijuana. Department of Health Medical Marijuana Approved Practitioners [Internet]. 2024. Available from: https://www.pa.gov/content/dam/copapwp-pagov/en/health/documents/topics/documents/programs/medical-marijuana/DOH%20Approved%20Practitioners.pdf [Last accessed: October 16, 2024 ].

47.

R Core Team. R: A language and environment for statistical computing. [Internet]. R Foundation for Statistical Computing: Vienna, Austria; 2024. Available from: https://www.R-project.org/

48.

Venables

, Ripley

. Modern Applied Statistics with S. 4th edition. Springer: New York, NY, USA; 2002.

49.

Listing of all Zip Codes in the state of Pennsylvania [Internet]. Available from: https://www.zip-codes.com/state/pa.asp [Last accessed: October 14, 2024 ].