Evaluation of U.S. State Opioid Prescribing Restrictions Using Patient Opioid Consumption Patterns from a Single,Urban,Academic Institution

Introduction

In response to a worsening opioid epidemic and increasing awareness of the adverse effects of opioids, multiple states, insurance companies, and pharmacies have developed regulations on opioid prescribing since 2017. 1 ^, 2 While the effect of these regulations on prescribing and adverse events have been evaluated through the lens of large-scale claims data, prescription patterns are an indirect proxy and only weakly corrected with patient opioid consumption. 3 ^, 4 These policies have not yet been evaluated in the context of actual patient opioid consumption.

In particular, individual states have implemented several wide-ranging policies to reduce inappropriate prescribing of opioids. 1 ^, 2 These initiatives have included mandates that prescribers access state prescription drug monitoring programs (PDMP) or obtain education regarding evidence-based opioid prescribing, as well as more direct restrictions on opioid prescriptions. While efforts vary widely by state, twenty-six states had passed laws by the end of 2017 that impose mandatory limits on the prescribing or dispensing of opioids for acute pain. 1 This number had increased to 39 by the end of 2019. Fourteen states impose limits on the daily dosage of opioids that can be prescribed for acute pain. 1 In addition, multiple states assert limitations on the duration of opioid prescriptions, with the duration ranging from 3 to 7 days. 1 Importantly, these limitations generally apply to all patients, including post-operative patients, though limited and conditional exceptions exist in some states pending additional provider authorization.

Similar efforts to curb inappropriate opioid use have been implemented by pharmacies. In 2018, one of the United States’ largest for-profit pharmacies limited initial acute opioid prescriptions to seven days and daily doses of 50 morphine milligram equivalents (MME). 5 Multiple pharmacy benefit managers have similar restrictions, limiting the initial opioid prescription to a seven-day supply and imposing restrictions on refills. 6 Given the disparate nature of the state and pharmacy regulations, the combined impact of opioid prescribing policies may be extreme. For instance, the aforementioned pharmacy's daily 50 MME limit constrains total prescription size in most states with a seven-day limit to a maximum of 350 MME. However, if a patient resides in a state with a three-day limit, the total prescription would be limited to 150 MME.

Multiple entities have also instituted limits on opioid refills. However, evidence suggests that a significant fraction of post-surgical patients require opioid refills to adequately treat their pain; a seven-day prescription may be inadequate for some patients. 2 ^, 4 ^, 7 A recent study found that 11–39% of surgical patients require a refill after a seven-day prescription, particularly among those undergoing neurosurgical or orthopedic procedures. 7 In the same way, the potential unintended consequences of opioid refill restrictions have been examined with real-world data, we must evaluate the impact of prescription size limitations, as little is known about where actual patient consumption falls in the context of these limits, in part due to the scarcity of consumption data. 2

In the current study, we sought to examine how state prescription limits would affect opioid-naïve surgical patients. We utilized a unique resource—a prospective institutional database of surgical patients’ observed post-discharge opioid consumption—to understand how prescribing limits relate to patient's actual consumption patterns.

Methods

We prospectively evaluated opioid consumption of all opioid-naive post-operative patients at our large academic medical center from October 1, 2017 to June 30, 2018. 8 Pre-surgery opioid exposure was determined by examining the record of prescribed medications in the medical chart. Opioid naïve patients were defined as having at least 30 days without opioid exposure. Procedures with <10 patients were excluded. Prescribers were not given any prescribing guidelines or requirements. Variables collected by abstraction from the medical record included baseline demographics, preoperative medications, perioperative surgical data, discharge prescriptions, and opioids dispensed. Discharged post-surgical patients were called by research staff at defined intervals within a month after discharge to evaluate opioid consumption. Patients were verbally consented to the phone. Patients were asked to obtain their prescriptions and count the remaining tablets. The initial prescription size was confirmed via chart review. Consumption of opioid medications was calculated from the amount filled compared with the number remaining; the accuracy of this approach to measure opioid consumption has been validated. 8 ^, 9 Queries were continued at regular intervals, up to one month, until a patient indicated that they were no longer taking opioid medications or their initial prescription was depleted; no patients were consuming opioids from the initial prescription beyond the one-month follow-up. Consumption values were converted to oral morphine milligram equivalents (MMEs) by standard conversions established by the Centers for Disease Control and Prevention. 10

Patients were excluded if any of the following were present: no working telephone number; additional surgeries or reoperation identified; readmission unrelated to pain; or incomplete medical records. For this study, we only evaluated opioid-naive patients. Surgical procedures were excluded if they had <10 patients in our dataset. The percentage of patients, by surgical procedure type, whose opioid consumption exceeded 150 and 350 MME was calculated. We also identified procedures for which the median and 75th percentile of opioid consumption exceeded these thresholds. We chose to examine the 75th percentile as prior studies and guidelines have used this value as an outlier threshold for typical opioid consumption. 11 – 13

All data were collected and managed using Research Electronic Data Capture. 14

The study was reviewed and approved by the institutional review board at our institution.

Results

A total of 2,971 opioid naïve patients were called after surgery to measure opioid consumption after discharge. Our validated telephone survey had a response rate of 79%. After applying the inclusion and exclusion criteria described above, 1,849 patients undergoing one of 54 distinct surgical procedures were included in our analysis (Supplemental Figure 1).

OPEN IN VIEWER

Figure 1

The median reported opioid consumption for each procedure (blue circle) and 75th percentile of reported opioid consumption (yellow triangle) are shown.

Overall, 21% of patients consumed more than 150 MME after discharge, and 7.3% of patients consumed more than 350 MME (Table 1). Procedure-specific analysis showed that 41 (76%) procedures included some patients who consumed >150 MME. The median consumption exceeded 150 MME for 4 (7.4%) procedures (Figure 1). The 75th percentile of consumption exceeded 150 MME for 22 (41%) procedures. Orthopedic, neurosurgical, and complex abdominal procedures most commonly exceeded this limit.

OPEN IN VIEWER

Table 1

Numbers of patients and procedures that exceeded the 150 MME and 350 MME limits.

	Prescription size limit
	150 MME	350 MME
No. patients exceeding this limit (%)	388 (21)	135 (7.3)
No. procedures containing patients who exceeded this limit (%)	41 (76)	25 (46)
No. procedures with median consumption exceeding this limit (%)	4 (7.4)	2 (3.7)
No. procedures with 75th percentile consumption exceeding this limit (%)	22 (41)	4 (7.4)

Twenty-five (46%) procedures included some patients who consumed >350 MME. The median consumption exceeded 350 MME for only 2 (3.7%) procedures. The 75th percentile of consumption exceeded 350 MME for 4 (7.4%) procedures. Consumption after orthopedic and spinal procedures most commonly exceeded this limit.

Procedure-specific volume and consumption estimates are available in Supplementary Table 1.

Discussion

In the context of the country's worsening opioid epidemic, there has been a national push to reduce opioid prescribing after discharge—with the important goal of reducing opioid misuse and diversion to other individuals. As a result, many limits on initial opioid prescribing after hospital discharge have been enacted. While they represent important initial efforts to curb unnecessary prescribing, these limits may have unintended consequences for a specific subset of post-surgical patients. To our knowledge, we have performed the largest study evaluating these policies in the context of the patient's observed prescription opioid use. Encouragingly, we found that most patients are unaffected by these limitations: the median opioid consumption for almost all surgical procedures fell below both the 150 MME and 350 MME limits. Nonetheless, a significant fraction of patients may still be affected by these policies. We considered below the 75th percentile of opioid consumption to represent typical consumption for each procedure. For procedures where typical consumption exceeds both the 150 and 350 MME limits, there will likely be patients whose post-surgical pain may not be appropriately controlled with these limits. Prior literature has shown that undertreatment of post-surgical pain may lead to the development of chronic pain and patients seeking alternate, sometimes illicit, sources of analgesia. 15 ^, 16 Our findings are consistent with prior research showing that orthopedic, neurosurgical, and complex abdominal procedures are associated with the highest amounts of opioid consumption. 7

These findings have three important implications. First, providers should be aware of the subset of procedures for which patients may require more opioids than initial prescribing limits allow, and should monitor closely to see if a refill is appropriate. Second, consumption data should inform the evidence-based revision of these prescription limits to minimize their unintended consequences. Finally, blanket prescribing limits alone do not motivate optimal opioid prescribing in surgery; additional complementary efforts are needed to achieve broader results.

Conclusion

In summary, to appropriately treat post-surgical pain, patients undergoing select neurosurgical, orthopedic, and complex abdominal procedures may require more opioids than certain common prescription limits allow; providers should be attentive to the pain control needs of these patients who are more likely to require refills. Patient consumption data should be used to inform current prescribing limits and guide future opioid restrictions. While blanket prescribing limits were a powerful initial response to the opioid crisis, we now have the data to create better policy, prescribing recommendations, and clinical decision support tools.