Abstract
Objectives
To evaluate national trends in opioid prescribing patterns by otolaryngologists for postoperative pain control after common otolaryngologic procedures.
Study Design
Cross-sectional; survey.
Subjects and Methods
A survey to determine opioid prescribing patterns for the treatment of postoperative pain following common otolaryngologic procedures was distributed to all members of the American Academy of Otolaryngology–Head and Neck Surgery.
Results
The most common pain medication prescribed for adults postoperatively was hydrocodone-acetaminophen (73%), followed by oxycodone-acetaminophen (39%). The most common pain medication prescribed postoperatively for children was acetaminophen (67%), followed by nonsteroidal anti-inflammatory drugs (65%). Overall, there was a wide variation in quantity of opioids prescribed for each surgery, ranging from 0 to more than 60 doses. Mean opioid prescriptions were greatest for tonsillectomy (37 tablets) and least for direct laryngoscopy (5.3 tablets).
Conclusion
This study identifies nationwide variations in opioid prescribing patterns among otolaryngologists. While otolaryngology is a relatively small specialty, we still have an obligation to work with all physicians to help combat the current opioid epidemic. By evaluating nationwide postoperative pain regimens, we are moving closer toward understanding how to reduce the opioid burden.
Opioids are pain-relieving medications with a long history dating back to use of the opium poppy during the Mesopotamian era. 1 Opioids were first popularized by the medical community with the introduction of morphine, which was then followed by heroin for the treatment of many common ailments, including analgesia and cough suppression. 2 While these medications were introduced with the hopes of significant therapeutic potential, the negative side effects and high rates of addiction began to overshadow the benefits.
In recent years, excess opioid prescribing is thought to have contributed to addiction, blood-borne diseases, and overdose-related mortality. 3 According to data from the Centers for Disease Control and Prevention (CDC), overdose deaths from opioids have more than quadrupled since 1999. 4 This high mortality rate has inspired both the government and medical community to examine methods to regulate opioid prescriptions. In 2016, the “Turn the Tide” campaign, headed by former US Surgeon General Vivek Murthy, sent recommendations and proposed guidelines for how to responsibly treat chronic pain. 5 A substantial amount of literature exists regarding use of opioids for chronic pain and associated guidelines for how to prescribe these medications. However, common stories of addiction begin with first use of these medications after acute pain from injury, trauma, or surgery.5-7
It is common practice for surgeons to prescribe opioid medications postoperatively for pain control. Despite the fact that opioids are routinely prescribed, there has been little investigation into postoperative opioid prescribing patterns and associated optimal pain control regimens for patients after undergoing surgery. Recent studies have demonstrated that surgical patients may not take half of the opioid pills prescribed to them after discharge. 8 This suggests that excessive narcotic prescriptions may not only be unnecessary but also may increase the number of narcotic medicines in circulation and available for unintended use. There is a paucity of surgical literature addressing opioid prescribing patterns, and those existing studies are specialty specific and, therefore, less applicable to a surgically diverse specialty like otolaryngology. This study aims to be the first to evaluate opioid prescribing patterns by currently practicing otolaryngologists for common otolaryngologic surgeries. Our intent is to fill a significant void in our specialty’s literature regarding this global health issue.
Methods
The study was approved by the Institutional Review Board at the University of Connecticut School of Medicine. A pilot study was first performed by surveying members of the Connecticut Ear Nose and Throat Society (CT ENT). 9 The data from this pilot study were not included in this current analysis. An online survey via SurveyMonkey.com was emailed to all otolaryngologist members of the American Academy of Otolaryngology–Head and Neck Surgery (AAO-HNS) (total number of 8025 members). The online survey consisted of 10 questions evaluating demographics (age, year of training completion, location of current practice, location of training, fellowship type) and opioid prescribing patterns (2 most common analgesic medications prescribed, refill frequency). The final question asked the participant to provide the average number of opioid doses prescribed for 12 of the most common surgeries: adult tonsillectomy, pediatric tonsillectomy, septoplasty alone, endoscopic sinus surgery, tympanoplasty alone, mastoidectomy, thyroidectomy/parathyroidectomy, head and neck cancer resection, neck dissection, soft tissue dissection (eg, lymph node or lipoma excision), rhinoplasty, and direct laryngoscopy. Participants were able to skip questions and choose multiple answers for certain questions. Participants were asked to provide information regarding opioid doses (including both tablet and liquid forms).
A dose is the quantity of a medication prescribed to be taken at one time (eg, for hydrocodone-acetaminophen, it is typically one 5 mg- to 325-mg tablet or 5 mL of a 5-mg to 325-mg/5-mL elixir) or total amount prescribed for a specific period. 10 Since medications are commonly prescribed in tablet form (a type of pill), we use the terms dose and tablet interchangeably in this article.
Demographic questions were associated with questions on prescribing patterns by simulation-based χ2 tests. In association analyses, ages “65 to 74” and “75 or older” were combined into “65 or older,” and year of training completion “1961-1970” and “1971-1980” were combined into “1961-1980.” Any “other,”“not applicable,” and responses to multiple items were excluded. Rarely used medications in adults and children with a frequency less than 5% were not considered in association analyses. To further investigate significant results, we replaced the dosage range by the mid-value (eg, if 1 to 10, replaced with 5) and summarized average opioid prescribing amounts for a surgery by mean and standard deviation. A P value less than .05 was considered statistically significant. All the statistical analyses were performed in R. 11
Results
There were a total of 596 respondents, a 7.4% response rate, for this AAO-HNS survey. Most respondents completed training after 1990 (74%). There was a proportional geographic variation among participants regarding residency training and current practice location. Approximately 41% of participants completed a fellowship ( Table 1 ).
Survey Demographics. a
The number of responses in each column may not sum to 596. Not all respondents answered each question. Values are presented as number (%) unless otherwise indicated.
Seventy-three percent of respondents reported prescribing hydrocodone-acetaminophen most commonly for pain control in adults. Other medications most frequently prescribed by respondents were oxycodone-acetaminophen (39% of respondents), nonsteroidal anti-inflammatory drugs (NSAIDs) (28% of respondents), and acetaminophen (18% of respondents). Unlike in adults, the most commonly prescribed medications for postoperative pain control in children were acetaminophen (67% of respondents) and NSAIDs (65% of respondents), followed by hydrocodone-acetaminophen (37% of respondents) and acetaminophen with codeine (9% respondents) ( Figure 1 ). Age and year of training completion were significantly associated with the status of prescribing NSAIDs (P = .014, .021) and acetaminophen with codeine (both P values <.001). Younger respondents and respondents who completed training later than 2000 were more likely to prescribe NSAIDs and less likely to prescribe acetaminophen with codeine to children than older respondents and respondents who completed training before 2000.
Adult and pediatric prescription patterns. Respondents most commonly prescribed hydrocodone-acetaminophen and oxycodone-acetaminophen for adults and acetaminophen and nonsteroidal anti-inflammatory drugs (NSAIDs) for children.
Nineteen percent of respondents reported never prescribing opioid refills, while the remaining 81% of respondents reported rarely (58%), sometimes (18%), or often (5%) prescribing refills ( Table 1 ). While frequency of prescribing refills was assessed, there was no information gathered regarding number of refills prescribed. The frequency of prescribing opioid refills was significantly associated with practice location (P = .02) and training location (P = .01), with reduced refills associated with the Northeast ( Table 1 ).
When quantifying average opioid amount prescribed based on surgery type, there was a wide variability even when controlling for one particular surgery. For example, when asked about endoscopic sinus surgery, 25 (5%) of respondents reported prescribing no narcotics, 66 (13%) reported prescribing 1 to 10 tablets, 162 (32%) reported prescribing 11 to 20 tablets, 175 (34%) reported prescribing 21 to 30 tablets, 60 (12%) reported 31 to 40 tablets, 14 (3%) reported prescribing 40 to 50 tablets, and 6 (1%) reported prescribing 51 to 60 tablets ( Figure 2 ).
Quantity of opioid tablets prescribed by respondents based on surgery type. Note the variation in opioid prescribing patterns by respondents for each surgery.
For the majority of otolaryngologic surgeries, the average opioid dose prescribed ranged from 14 to 25 tablets. On average, the greatest quantity of narcotics was prescribed for tonsillectomy (37 ± 17 tablets) and the least for direct laryngoscopy (5.3 ± 7.7 tablets) ( Table 2 ).
Mean Number of Opioid Tablets Prescribed by Respondents Based on Surgery Type.
Abbreviations: ESS, endoscopic sinus surgery; H&N, head and neck; ND, neck dissection.
For the majority of otolaryngologic procedures, the respondents who trained and practiced in the Northeast appeared to prescribe the least amount of opioids for each specific surgery. However, this was only statistically significant with endoscopic sinus surgery (P = .04, .026) and adult tonsillectomy (P = .014, .026) for those who trained and practice in the Northeast, respectively.
Discussion
Over the past couple of decades, opioid overdoses have significantly increased partly due to increased opioid prescribing by medical providers. 3 From 2000 to 2010, the average quantity of an oxycodone or hydrocodone prescription increased almost 70%. 12 Moreover, in 2013, it was reported that approximately 4.5 million persons age 12 and over were using prescription-type pain relievers without a medical indication. 13 Therefore, it is important to consider that any excess opioid medication that is prescribed may have potential for nonmedical use.
While primary care providers most often prescribe medicines for chronic pain conditions, it is estimated that surgeons account for over 36% of all dispensed opioid medications in the United States. 14 There are a limited number of studies evaluating how providers are prescribing pain medications for postoperative pain. Of those evaluating patient usage patterns for postoperative opioids, most demonstrate that patients may only use less than half of the pain medications prescribed to them after larger surgeries such as cesarean section, thoracic surgery, or upper extremity surgery.8,15
Even a short course of opioid use after acute surgery has the potential to contribute to long-term use. Use of opioids within 7 days of low-risk surgery has been associated with the increased likelihood of opioid use at 1 year. 7 Specifically, the likelihood of long-term opioid use increased with each additional day of medication supplied, starting after the third day for opioid-naïve patients. 16 Given this, the CDC guidelines recommend that clinicians should prescribe the lowest effective dose and no greater quantity than needed for the expected duration of severe pain. 17
Few studies have evaluated postoperative pain control after otolaryngologic surgery. Two studies suggested that no opioids were needed after sinus surgery and that acetaminophen was adequate in controlling postoperative pain.18,19 In addition, it was demonstrated that only approximately 20% of patients required any opioids after direct laryngoscopy. 20 Most recently, a study demonstrated that only 36% of opioid tablets prescribed were used by patients after rhinoplasty and that the number of pills prescribed was “arbitrarily selected” by the surgeons. 21 These studies are limited by low power and the fact that they address only single surgeries in a surgically diverse specialty like otolaryngology. This study is the first to evaluate how otolaryngologists are prescribing opioids for a variety of otolaryngologic surgeries on a national level.
Our results indicate that there is a large variation in the number of opioid doses prescribed for the same surgery based on an individual surgeon’s preference. For example, with adult tonsillectomy, 12% of participants typically prescribe 20 or fewer tablets, while 22% prescribe more than 50 tables. Since standard practice involves using “evidence-based medicine” for clinical decision making, it is important for us to objectively quantify these differences. Understanding this variability in practices can help guide future work in potentially standardizing postoperative pain management and lead to evidence-based decision making.
Our results are consistent with previous literature demonstrating that the most common opioids prescribed in the United States are hydrocodone, followed by oxycodone. 22 In addition, the use of acetaminophen and NSAIDs in children is consistent with previous literature reporting adequate pain control with alternating these 2 medications and is supported by the American Academy of Otolaryngology–Head and Neck Surgery Foundation (AAO-HNSF) guidelines.23,24 Our study demonstrates that younger providers may be more likely to prescribe nonopioid medications for children than older providers. This may be reflective of the recent emphasis by both undergraduate and graduate medical education regarding opioid dependence as well as awareness of current AAO-HNSF guidelines.
Based on our results, there also appeared to be a trend toward reduced prescribing by otolaryngologists who both trained and currently practice in the Northeast. CDC reports that 7 of the 10 Northeast states are listed in the top 13 states for the greatest number of deaths per 100,000 population based on age-adjusted rates. 4 It is possible that these providers may have reduced their prescribing quantities due to an increased awareness of state and regional issues with opioid abuse.
The US Food and Drug Administration placed a black box warning regarding the use of codeine in children younger than 12 years of age and a warning against its use for adolescents between 12 and 18 years of age. 25 Surprisingly, our data suggest that some otolaryngologists are still prescribing these medications to children. We anticipate this study will help raise awareness about these issues and prevent any future adverse events. These results are in contrast to previous literature, which reported close to 0% prescribing of codeine after the black box warning was issued. 26
With any survey-based study, there are intrinsic limitations. There is potential for recall bias for all respondents and selection (response) bias for those who chose to participate in the survey. This creates potential skewing of data toward those providers who are more cognizant of the physician role in the opioid epidemic. It is also possible that a small subset of providers may be prescribing the greatest quantity of opioids, which may misrepresent nationwide prescribing patterns. While there were a large number of respondents, our results still only represent 7.4% of the AAO-HNS membership.
In this study, we asked respondents to quantify the number of opioid pills prescribed based on a range. When analyzing the data, it was determined that significant results would be better quantified if the ranges were converted into a mid-value point (eg, 1 to 10, replaced with a 5). Although midpoint values allow for standardized results, it introduces the possibility that responses may not accurately reflect practice patterns.
While we asked respondents to quantify the number of opioid doses prescribed, the survey did not specify the definition of 1 opioid dose (eg, 1 dose is 1 hydrocodone-acetaminophen 5-mg to 325-mg tablet). This could potentially be a source of confusion for respondents and may have contributed to an underestimation of opioids prescribed by respondents. Furthermore, some respondents may prescribe refills; therefore, initial and total number of pills prescribed could be different values. We recognize that some of the answer choices are over-the-counter medications (ie, Tylenol and Advil) and do not require a prescription. This is another source of ambiguity that may promote underreporting of these medication choices if respondents “recommend” but do not prescribe these medications.
Finally, tonsillectomy patients and pediatric patients often require liquid pain medications. While respondents were asked to record the number of doses (including for liquid pain medications), many participants reported the total amount of liquid pain medication without reporting dosing frequency in the comments section and, therefore, these responses had to be excluded from data analysis.
Patient safety is the ultimate goal, and steps toward regulating opioid prescriptions have already been instituted on a legislative level. The goal of state prescription monitoring programs is to reduce opioid prescribing by limiting the maximum amount of opioids prescribed and by mandating physicians to review individual patient opioid use history via a centralized system. These state prescription programs have been shown to reduce the amount of opioids prescribed, which reduces the amount available to the public and subsequently lowers the associated opioid-involved overdose deaths.27,28
As physicians, we are encouraged to treat pain. In fact, pain has become a “fifth vital sign” in many clinical settings, and in many cases, the effectiveness of pain control subjectively reported by patients has been linked to patient satisfaction scores. 29 It is our responsibility, as physicians, to judiciously prescribe medications to allow for optimal pain control, but we also need to be cognizant of inappropriate prescribing patterns, which may place our patients at increased risk for opioid abuse and dependence. Future work should aim to evaluate ways to optimize opioid prescriptions while limiting excess. Our work provides preliminary data toward this end point and will hopefully influence other surgical specialties to critically evaluate prescribing patterns as well.
While otolaryngology is only a small part of the medical community, it is still important to work with other medical specialties to follow the trends of responsible opioid prescribing by having frank discussions with patients about the addictive risk profile, using prescription monitoring programs, and recommending over-the-counter pain relievers when appropriate. 30 We anticipate this study will promote more critical thought in managing postoperative pain, and we hope that it contributes to a growing literature base that addresses a significant global health issue.
Conclusion
This study reports opioid prescribing patterns among otolaryngologists nationwide for postoperative pain control. It further identifies variations in these opioid prescribing patterns. Future studies are needed to evaluate opioid usage patterns by patients with goals to propose more standardized prescribing patterns for postoperative opioid medication in the field of otolaryngology. As otolaryngologists, we should feel obligated and empowered to help address the opioid crisis in America, and further investigations determining appropriate postoperative pain regimens can help us attain this goal.
Author Contributions
Disclosures
Footnotes
No sponsorships or competing interests have been disclosed for this article.
This article was presented as a poster at the 2017 AAO-HNSF Annual Meeting and OTO Experience; September 10-13, 2017; Chicago, Illinois.