Alexis Stoner, Lindsay Tjiattas-Saleski, Emma Padgett, Cole Harp, Martin Groke, David Redden
4 min
As opioid and stimulant overdose deaths remain a significant public health concern, researchers sought to identify the demographic and clinical factors associated with the concurrent prescribing of these two drug classes. While both substances are monitored individually, there is limited evidence regarding the patterns that lead to patients receiving both simultaneously.
This retrospective study analyzed data from the South Carolina Department of Public Health (SCDPH) Prescription Drug Monitoring Program (PDMP) between 2016 and 2021. The researchers examined over 9.7 million opioid prescriptions to identify those that overlapped with a stimulant prescription for at least 30 days. They used generalized estimating equations to assess the association between coprescribing and variables including patient age, sex, opioid supply length, and daily morphine milligram equivalents (MME).
Of the opioid prescriptions analyzed, 1.65% involved a concurrent stimulant prescription. The study found that patients aged 50 and younger were 2.74 times more likely to receive both medications compared to older patients. The strongest predictor was the duration of the opioid prescription; patients with an opioid supply of 30 days or longer had 4.38 times the odds of receiving a concurrent stimulant. Additionally, female patients and those with a daily MME between 50 and 100 mg showed a statistically significant, though smaller, increased likelihood of receiving both drugs.
Understanding the factors that drive opioid-stimulant polypharmacy is essential for clinicians to identify patients at higher risk of adverse outcomes, such as fatal overdose. The findings suggest that younger patients and those on longer-term opioid therapy are particularly vulnerable to this combination. These results provide a foundation for future research to investigate whether specific medical diagnoses—such as chronic pain or obesity—or socioeconomic factors drive these prescribing trends, ultimately helping to inform safer clinical practices.
OBJECTIVE: The South Carolina Department of Public Health (SCDPH) Prescription Drug Monitoring Program (PDMP) is used to monitor the prescribing patterns of controlled substances, including opioids and stimulants. Data collected by the National Center for Health Statistics show that the number of drug overdose deaths involving opioids and stimulants has been increasing since 1999. Although databases such as the PDMP monitor healthcare professionals’ prescribing habits regarding opioids and stimulants individually, there has been little research regarding opioid-stimulant coprescribing patterns. Considering increasing death rates and the lack of research, data from the SCDPH PDMP were used to identify factors that influence opioid-stimulant coprescribing. It was hypothesized that younger age, higher daily morphine milligram milliequivalents (MMEs), increased duration of prescription, and male gender would be associated with increased opioid-stimulant coprescriptions. METHODS: Data were gathered from the SCDPH PDMP from 2016 to 2021. The variables assessed included patient age, patient sex, filling date, supply length, daily MMEs, and Lexicomp Drug Classification. A linear regression statistical analysis was then performed to analyze the data. RESULTS: Between January 1, 2016 and December 31, 2021, a total of 9,785,146 opioid prescriptions were filled by 2,158,564 individuals, with 161,103 (1.65%) of them receiving a concurrent stimulant prescription. Patients who were 50 years of age and younger had a 2.74 greater odds of receiving a concurrent stimulant prescription (P<0.001), along with those who had an opioid prescription length of ≥ 30 days (odds ratio [OR] 4.38, P<0.001). Being female (OR 1.58) and a daily MME ≥ 50 mg but ≤100 mg (OR 1.24) were both significantly found to indicate an increased odds of having a concurrent prescription. CONCLUSIONS: This study indicates that those with longer opioid prescriptions, younger individuals, higher daily MMEs, and women are at an increased risk of opioid-stimulant coprescription. Future research should investigate other factors associated with opioid-stimulant coprescribing, such as education status, rurality, and whether specific opioid or stimulant agents are linked to coprescribing.
Sam: They ran a sensitivity analysis comparing the full dataset against a restricted version that excluded records with implausible values—extreme ages, outlier MME figures. The odds ratios barely moved, which suggests the main findings aren't artifacts of noisy data at the margins.
Alex: But there's a significant gap in what the data can actually tell you. Without diagnostic codes, how do you distinguish inappropriate polypharmacy from a medically necessary combination?
Sam: That's the primary limitation, and the authors are candid about it. Without ICD-10 codes, you can't separate a stimulant prescribed to counteract opioid-induced sedation from one prescribed for an unrelated comorbid condition. The study gives you the pattern—the demographic and duration-based markers—but the clinical intent behind each prescription remains unobserved.
Alex: So the "what" is reasonably well-characterized, but the "why" is still a black box. Where does that leave a clinician trying to act on this?
Sam: It points toward better integration at the point of care. The real utility would be in clinical decision support systems that don't just surface PDMP data, but cross-reference it with EHR diagnostic codes in real time. That would allow for targeted alerts when a new prescription creates a high-risk profile—giving the prescriber the context to judge whether the combination is evidence-based or a signal worth investigating.
Alex: It's a meaningful shift in framing—moving from retrospective mortality analysis toward prospective, patient-level prescribing safety. Thanks for walking through the model mechanics, Sam.
Sam: Understanding the structural patterns is the necessary first step. The clinical guardrails can only be as good as the data architecture supporting them. Thanks for listening to ResearchPod.