As MC becomes increasingly available, it is essential to study the geographical distribution and identify factors that influence access to MC for the treatment of pain. We conducted the first analysis of geographical patterns and barriers to access of utilization of MC for pain. In line with the findings of the State Epidemiological Outcomes Workgroup in 2022 reporting the patient density of total MC certifications per county for all indications as of August 2021, we found that Bradford County and Tioga County also had a significantly lower proportion of adults with a MC certification for pain as compared to the county-wide average in 2024 [19]. The four-fold disparities between the highest and lowest counties may be concerning because chronic pain in adults is one of the few qualifying medical conditions in PA where the 2017 NAS report rated the quality of the evidence base that cannabis is effective as “substantial” [4].

Although there has been little research on the associations between sociodemographic variables and the frequency of MC certifications for pain, previous reports analyzing geographic access to MC services across states have found mixed results regarding the influence of median household income on MC accessibility [20–23]. While prior studies of access to MC dispensaries in California and Colorado found no significant association between income and geographic access to MC dispensaries, a report in Oklahoma found that higher proportions of residents who were uninsured and living in poverty were associated with less availability of MC services [20–22]. In contrast, an investigation from New York found that MC services were least available in neighborhoods with highly educated residents [23]. Our study found a significant correlation between the percent of non-White individuals including those of Hispanic ethnicity and the proportion of adults with a MC certification for pain at the ZCTA level.

From an epidemiological perspective where those residing in rural areas are older, more likely to work in manufacturing or agricultural jobs, and more likely to report chronic pain [24], one might anticipate that MC for pain would be higher in rural vs urban areas. However, marijuana is currently a Schedule I substance so MC is not covered by insurance. Chronic pain patients in New England spent a considerable amount (median = $2,320, mean = $3,064) of their disposable income on MC [25]. Dispensaries tend to be localized in more urban areas [15]. In our analysis, we found a significant positive correlation between median household income and the proportion of adults with MC certifications for pain in 2024 at the county level. This suggests that patients’ propensity to utilize MC for pain relief may be influenced by their median household income in addition to their geographical location. The cost of a high-quality ounce of MC in PA ($384) was well above the national average ($326) and ranked seventh in the US [26]. The annual cost ($50) for registering for a medical marijuana identification card can be waived for those enrolled in Medicaid, Children’s Health Insurance Program, Supplemental Nutrition Assistance Program, and the Supplemental Nutrition Program for Women, Infants, and Children [27]. As such, our findings suggest that it could be beneficial for primary care providers to consider the ability of patients to afford medical cannabis and geographic barriers to utilization when incorporating MC into the treatment plan of patients with severe chronic or intractable pain.

There is a long and unfortunate racialized history in this field with some perceiving the term “marijuana” as racist [28]. Doctors from a large health system in western PA were almost three-fold more likely to test a Black than a White pregnant woman for marijuana and other drugs despite the urine toxicology being less likely to be positive for Black than White patients [29]. A Black person was three times more likely to be arrested for marijuana possession than a White person in PA [30]. The disparity was even larger in 2018 in Clarion (25.6 fold) and Parry (28.4 fold) counties [30]. Nationally, past month marijuana use among those greater than or equal to 12 years of age were similar among Black (18.1%) and White (16.3%) people [31]. Although pain is a biopsychosocial condition, reporting pain severity is subjective. We can not discount the possibility that some patients obtain a MC certification for pain or other medical conditions [15] in an attempt to minimize their legal risk. Although we found that ZCTAs with greater non-White populations had significantly fewer pain certifications, the percentage of variance accounted for (0.6%) was so modest as to be practically meaningless [18].

There are some caveats to this report. One limitation of this analysis is that the de-identified data was submitted to us from a third-party source (Supplementary Fig. 1). There are multiple versions of MC certification datasets as the PDOH has updated the data several times. We anticipate that it is likely that the PDOH will make amendments to the datasets that were provided to us [14]. Another limitation is the potential for confounding, as counties with higher median household incomes and higher proportions of non-White individuals may differ in other relevant ways—such as adult population size, provider density, and healthcare access —that could influence certification rates. An additional limitation includes the potential of zip codes to map to multiple geographic units, leading to inaccuracies in mapping. In addition, ACS estimates at the local level are often less reliable with larger margins of error which may significantly influence the results of our ZCTA-level analysis [32]. This investigation was completed in one state (PA) and assessed medical but not recreational cannabis. These findings may not generalize to other states with different MC policies [28]. A prior report from a resource limited population in the Bronx determined that over half (55%) of newly certified patients, predominantly for pain, did not purchase MC [33]. We do not know how many of the certified patients in PA were regularly using MC. Additional research is necessary to examine whether the population size or income geographical differences also apply to illicit cannabis or to delta-8 tetrahydrocannabinol use [34].

In conclusion, we have mapped the geographic distribution of the percent of adults with a MC certification for pain in PA from 2018 to 2024 at the county and ZCTA levels. We found approximately a four-fold difference in the percent of adults with a MC certification for pain between the county with the highest and lowest certification rate in 2024. In addition, we report significant correlations between 1) income and 2) ethnicity and the percent of adults with a MC certification for pain in 2024.