To describe the opioid prescribing practices in opioid-naïve women undergoing elective gynecologic surgery for benign indications and identify risk factors associated with increased perioperative opioid use. We also explored factors associated with new persistent opioid use in women with perioperative opioid use.
Retrospective population-based cohort study.
We used linked administrative data from a government administered single-payer provincial healthcare system in Canada. This study was undertaken at ICES, a not-for-profit research institute in Ontario, Canada.
We followed opioid-naïve adult women who underwent benign elective gynecologic surgery between 2013 and 2018.
The primary outcome was peri-operative opioid use defined as ≥1 opioid prescription from 30 days before to 14 days after surgery. New persistent opioid use (NPOU) after gynecologic surgery was defined as having filled one or more opioid prescriptions between 91 days to 180 days post-operatively. Multivariable log-linear regression analyses were employed to adjust for clinical and demographic data.
Of the 132,506 patients included in our cohort, the majority (74.3%) underwent minor gynecologic procedures. Perioperative opioid use was documented in 27,763 (21.0%) patients and there was a significant decreasing trend (p<0.001) in the proportion of patients with perioperative opioid use from 21.8% in 2013 to 18.5% in 2018. Factors associated with increased perioperative opioid use included younger age, higher income quintile, urban dwellers, and diagnosis of infertility, endometriosis or adnexal mass. Perioperative opioid use was an independent risk factor for persistent use (adjRR 1.40, 95% CI 1.13-1.72) and for every 65 patients prescribed opioids associated with gynecologic surgery, one developed new persistent opioid use. The highest risk factor for developing persistent use were filling a high dose opioid prescription (adjRR5thQuintileOME 2.33, 95% CI 1.83-2.96).
One in five women who undergo a gynecologic procedure has a new exposure to opioids. For every 65 patients who fill an opioid prescription following their gynecologic surgery, one will experience prolonged opioid use.
In 2017 alone, the United States (US) reported over 70,000 deaths from drug overdose where approximately two-thirds were due to opioids 1. Prescription opioids accounted for 32% of the opioid-related death rate 2. Opioids play an important role in managing post-operative pain; however, their risks and benefits must be carefully weighed. There is abundant evidence that surgical patients are often over-prescribed opioids and that there is a disconnect between the quantity prescribed and the expected patient-reported pain level 3-5. It is also known that high opioid prescriptions lead to greater patient consumption and can consequently translate to opioid dependence 6. Studies have shown that patients undergoing major and minor surgical procedures are 15 times more likely to develop new persistent opioid use (NPOU) compared to their non-surgical counterparts 7. Reported rates of NPOU following surgery in opioid naïve patients ranges from 0.5-6.5%, depending on the definition used 7-9 .
Gynecologic procedures account for 7.7% of all surgeries performed in women representing an important generator of opioid prescriptions 10. Hysterectomy is the most common gynecologic surgery performed and the current literature largely explores opioid use exclusively for hysterectomy 5,9,11-13. Much less is known about opioid prescribing trends and rates of persistent use following non-hysterectomy gynecologic surgeries. Thompson et al. addressed temporal trends of opioid prescribing over a 10-year period and showed a 56% increase in opioid fills despite increase in minimally invasive (MIS) techniques. However, this study focused solely on hysterectomy and did not look beyond 2014 13. Other opioid literature in this field rely on the use of large administrative databases that only include commercially insured women, significantly limiting the generalizability of findings. As such, it is important to study the pattern of opioid prescribing for gynecologic surgeries during the opioid crisis (2013 and beyond) in a setting where all women and all surgeries are captured.
Our first objective was to identify risk factors associated with perioperative opioid use among opioid-naïve women undergoing elective gynecologic surgery. Second, in a restricted cohort of women who had perioperative opioid use, we explored factors that were associated with new persistent opioid use.
MATERIALS AND METHODS
Study Design and Data Sources
Exposure and Outcomes of Interest
Women exposed to any of the three categories of gynecologic surgery (major, intermediate and minor) were observed for the primary outcome of perioperative opioid use. Perioperative use was defined as filling ≥1 outpatient opioid prescription from 30 days before to 14 days after surgery 7,9. For each prescription, we calculated total oral morphine equivalent (OME) dispensed 17, daily dose OME and prescription duration. We also identified NPOU after gynecologic surgery, which was defined as having filled ≥1 opioid prescription between 91 to 180 days post-operatively. There is a consensus in the literature that post-operative pain related to surgery should have resolved by 90 days 7,13,16.
Table 1Baseline Characteristics of patients with and without perioperative use and relative risk of perioperative opioid use.
Perioperative opioid use was summarized for the entire cohort and by year. Prescription patterns (total OME, daily dose OME, duration) were presented as mean ± standard deviation (SD). The Cochrane Armitage Test for trend was used to evaluate differences in the proportion of perioperative opioid use over time. One-way ANOVA was used to test differences between mean OMEs across years. We presented descriptive statistics comparing baseline demographics of patients with and without perioperative opioid use. Standardized differences >0.10 were considered clinically important21. To explore factors associated with perioperative opioid use, we performed multivariable modified Poisson regression analysis where all baseline covariates were included in the model. The proportion of patients having NPOU was presented for those with and without perioperative opioid fill. The risk of NPOU was evaluated by estimating an unadjusted odds ratio (OR), absolute risk difference, number needed to harm, and an OR adjusted for age, mental health diagnosis, rurality and procedure type. Subsequently, we restricted the cohort to patients who had perioperative opioid use and a multivariable modified Poisson regression analysis was used to explore factors associated with persistent use. In this model, we added total OME and opioid class in addition to baseline characteristics. We reported 95% confidence intervals (CI) and a p-value < 0.05 was considered statistically significant. Analyses were conducted using SAS for UNIX, version 9.2 (SAS Institute).
Table 2Patterns of Perioperative Opioid Prescription by Procedure and Over Time
§ mean ± SD
* one-way ANOVA
OME = Oral Morphine Equivalent
NPOU was documented in 4,827 (3.65%) patients undergoing gynecologic surgery: 3,476 (3.32%) patients among those without perioperative opioid use and 1,351 (4.87%) among those with perioperative opioid use. Perioperative opioid use due to gynecologic surgery was significantly associated with developing NPOU (unadjusted OR 1.49, 95%CI 1.40-1.59, p<0.0001, number needed to harm 65). Perioperative opioid use was an independent risk factor for NPOU after adjusting for age, rurality, mental health diagnosis and procedure type, (adjusted OR 1.59, 95%CI 1.45-1.75, p<0.0001).
This retrospective cohort study of over 130,000 women who underwent gynecologic surgery in a single-payer healthcare system found that 21% filled an opioid prescription in the perioperative period. Perioperative opioid fill was more common in younger, more affluent women with diagnosis of infertility, endometriosis or adnexal mass. Perioperative opioid use was an independent risk factor for developing NPOU. Our data suggests that for every 65 patients who fill an opioid prescription due to their gynecologic surgery, one will develop NPOU. Among those with perioperative opioid use, factors most strongly associated with developing NPOU were a documented history of substance use disorder, prescription fills with high doses of opioids, and women undergoing minor procedures. We report a reassuring decreasing trend, albeit modest, in the rate of perioperative opioid use and quantity of prescribed opioids during the opioid crisis between 2013 to 2018.
The findings of our study build upon those of prior reports. Perioperative opioid use following hysterectomy in the US, regardless of surgical approach, has been reported at approximately 80% 13,22,23. Our rates of perioperative opioid use following hysterectomy were slightly slower at 72% for open hysterectomy and 67% for a MIS approach. Patients in our study received a lower quantity of OMEs and shorter duration of opioids compared to published reports. We report a mean OME of 193 mg for open hysterectomy and 170 mg for MIS hysterectomy and laparoscopic tubal/ovarian surgery with an average prescription duration of 4 days. Comparatively, the mean OMEs for hysterectomy in prior reports range from 225 to 250 mg with prescription duration between 5 and 11.5 days 9,12,13,22. These differences are even more striking for minor procedures. For example, for endometrial ablation and dilatation/curettage, we report a 30% and 1% rate of perioperative opioid use respectively, compared to 63% and 37% in the literature 22. The total OMEs for these minor procedures was 123mg in our study compared to 150mg in a prior report 22. Our findings underscore the importance of geographic differences in opioid prescribing patterns. Regional variations in prescribing patterns have been described where the North Eastern US demonstrate consistently lower rates of perioperative opioid use 13,22,23. Our data are consistent with reports of prescription variations between countries where mean dose of opioids after surgery was significantly higher in the US compared to Canada and Europe 24.
A unique aspect of our study is that it describes temporal trends in opioid prescribing during the era of the opioid crisis. We report a statistically significant difference in opioid prescriptions pattern with total OME, daily dose OME and prescription duration having shown a decrease by then end of our study period. Other studies have not reported similar reassuring trends, although their study periods were not as recent 13,22. Less quantity of opioids being prescribed towards the end of our study period may be a result of improved prescribing patterns where a better understanding of opioid risks translates to more appropriate prescriptions. Although these trends are reassuring, our data leaves us much room for improvement. For example, in 2018 patients filled an average of 23 pills of 5mg-oxycodone following open hysterectomy and 19 pills following MIS hysterectomy. This exceeds current prescribing recommendations that suggest 20 pills after open and 15 pills after laparoscopic hysterectomy should suffice 25,26. As clinical practice guidelines 26,27 are released, they must be accompanied by robust knowledge translation strategies to realign the way in which clinicians prescribe. This is highlighted by a study exploring prescribing practices among obstetricians-gynecologists that reported only 19% of practitioners adhered to guidelines on prescribing opioids28.
We found a NPOU rate of 3.32% and 4.87%, for patients without and with perioperative opioid use, respectively. The exact rate of NPOU is highly dependent on the definition. Our data is consistent with other studies which utilized a similar methodology: Brummett et al. NPOU rates between 5.9% to 6.5%; Wright et al. 6.8%; Peahl et al. 4% 7,16,22. Not surprisingly, studies with more stringent definitions reported lower rates of NPOU 8,9,23. However, all studies unequivocally show that perioperative opioid use and higher doses of prescribed opioids are associated with higher rates of NPOU. Thus irrespective of the true rate of NPOU, it is important for clinicians to appreciate that some proportion of prolonged opioid use is stemming from exposure to opioids prescribed following gynecologic surgery. This is likely one of the many contributors to the opioid crisis, considering that 8% of surgeries in the US are gynecologic. Based on the risk factors for NPOU we found, clinicians should make efforts to reduce the amount of opioid prescribed, especially in patients with previous mental health diagnoses, if we are to decrease the risk of chronic opioid use. Furthermore, the rate of opioid prescriptions generated after minor procedures such as hysteroscopy that are not thought to provoke significant pain was unexpectedly high. Avoiding unwarranted opioid prescription in such instances would further reduce the amount of opioid prescribed.
We were able to demonstrate that several patient characteristics are associated with risk of NPOU among women with perioperative opioid use. We showed that patients who have a mental health diagnosis, have a history of substance use disorder, or high amounts of prescribed OMEs had higher rates of NPOU and these findings are consistent with findings from similar studies 5,9,16,29. Given our large cohort, the magnitude of these risks is modest and likely merit further confirmation. Furthermore, the data becomes inconsistent between studies with respect to the association of age, income level, health status, indication for surgery and type of surgery on the risk of NPOU. This inconsistency in the literature is an obstacle for practitioners to be able to accurately identify risk factors for NPOU and develop individualized prescription plans to minimize risk of NPOU. Future studies to better evaluate these risks can help us better understand chronic post-surgical opioid use and devise strategies for prevention.
One of the biggest strengths of our study was that it included a wide range of gynecologic procedures, rather than focusing solely on hysterectomy. We also used data from a government administered, single-payer healthcare system that included all patients in the province and represented diverse socioeconomic and demographic groups, regardless of insurance status. This reduces the risk of selection bias and increases generalizability of our findings. Prior reports using commercial data usually study the most affluent in society and often exclude older patients as many transition to Medicare after age 65 and are lost to datasets like Data Mart or Marketscan, which are made up mostly of privately insured patients. Our results show that income is associated with both perioperative opioid use and NPOU, and using an inclusive dataset makes our findings more robust.
Our data must be interpreted in the context of the study design. Firstly, we use prescription fill as a surrogate for patient consumption. Although patients may be labelled as having high perioperative doses of OMEs, we are not able to ascertain if these were indeed consumed, which is important as it is known that more than half of opioid prescriptions following hysterectomy are unused 30. As with any other administrative dataset, we were also unable to ascertain certain specific variables such as ethnicity, smoking or body mass index, as well as prescriber and hospital attributes that could help clarify factors that were associated with NPOU. Administrative datasets in general cannot capture clinical factors with perfect accuracy, thus we endorse caution when implementing our findings into practice and attempting to identify patients at risk of increased perioperative opioid use. We were also limited in capturing specimen-specific variables such as final pathology results or uterine weight although this would be an interesting exploration for future studies. There is evidence that physician-specific attributes can affect a patient’s risk of long-term opioid use, however, we do not have any data on prescribers’ level of training or opioid-prescribing patterns 29,31. Previous studies have shown that the presence of co-existing pain conditions such as arthritis to be associated with increased opioid use7. We attempted to capture this association by identifying patients with pain/endometriosis as a co-variate, however, our findings are limited in determining the association between co-existing chronic pain syndromes and increased opioid use after gynecologic surgery and warrants further investigation. Lastly, our data are limited to a Canadian context and may not be applicable to other jurisdictions. However, the trends in Canada have tended to mirror those in the US. Indeed, our study demonstrates similar findings to the Center for Disease Control and Prevention (CDC), which reports a decrease in opioid prescriptions since a peak in 2012 2.
One in five women who undergo a gynecologic procedure is newly exposed to opioids. For every 65 patients who newly initiate opioids around the time of their gynecologic surgery, one will experience prolonged opioid use. Clinicians must understand the risks that any perioperative opioid prescription could lead to chronic use and contribute to our opioid crisis. We have a responsibility to focus our efforts on decreasing the number of opioid prescriptions that are written, especially for minor procedures which do not typically provoke sufficient pain to warrant an opioid prescription. When an opioid prescription is warranted, clinicians should be vigilant about the quantity of prescribed opioids. The risks and benefits of opioids for post-operative pain control should be balanced so that we can collectively prescribe responsibly.
This study was supported by ICES, which is funded by an annual grant from the Ontario Ministry of Health and Long-Term Care (MOHLTC). The study was completed at the ICES Western site, where core funding is provided by the Academic Medical Organization of Southwestern Ontario, the Schulich School of Medicine and Dentistry, Western University, and the Lawson Health Research Institute. Parts of this material are based on data and information compiled and provided by: MOHLTC, CCO, CIHI. The analyses, conclusions, opinions and statements expressed herein are solely those of the authors and do not reflect those of the funding or data sources; no endorsement is intended or should be inferred. We thank IMS Brogan Inc. for use of their Drug Information Database.
Scholl L, Seth P, Kariisa M, Wilson N, Baldwin G. Drug and Opioid-Involved Overdose Deaths – United States, 2013-2017. MMWR Morb Mortal Wkly Rep. 2018;67(5152):1419-1427.
Wilson N, Kariisa M, Seth P, Smith Ht, Davis NL. Drug and Opioid-Involved Overdose Deaths – United States, 2017-2018. MMWR Morb Mortal Wkly Rep. 2020;69(11):290-297.
Bicket MC, Long JJ, Pronovost PJ, Alexander GC, Wu CL. Prescription Opioid Analgesics Commonly Unused After Surgery: A Systematic Review. JAMA Surg. 2017;152(11):1066-1071.
Howard R, Fry B, Gunaseelan V, et al. Association of Opioid Prescribing With Opioid Consumption After Surgery in Michigan. JAMA Surg. 2019;154(1):e184234.
Wong M, Vogell A, Wright K, Isaacson K, Loring M, Morris S. Opioid use after laparoscopic hysterectomy: prescriptions, patient use, and a predictive calculator. Am J Obstet Gynecol. 2019;220(3):259.e251-259.e211.
Shah A, Hayes CJ, Martin BC. Characteristics of Initial Prescription Episodes and Likelihood of Long-Term Opioid Use – United States, 2006-2015. MMWR Morb Mortal Wkly Rep. 2017;66(10):265-269.
Brummett CM, Waljee JF, Goesling J, et al. New Persistent Opioid Use After Minor and Major Surgical Procedures in US Adults. JAMA Surg. 2017;152(6):e170504.
Clarke H, Soneji N, Ko DT, Yun L, Wijeysundera DN. Rates and risk factors for prolonged opioid use after major surgery: population based cohort study. Bmj. 2014;348:g1251.
Swenson CW, Kamdar NS, Seiler K, Morgan DM, Lin P, As-Sanie S. Definition development and prevalence of new persistent opioid use following hysterectomy. Am J Obstet Gynecol. 2018;219(5):486.e481-486.e487.
Oliphant SS, Jones KA, Wang L, Bunker CH, Lowder JL. Trends over time with commonly performed obstetric and gynecologic inpatient procedures. Obstet Gynecol. 2010;116(4):926-931.
Aarts JW, Nieboer TE, Johnson N, et al. Surgical approach to hysterectomy for benign gynaecological disease. Cochrane Database Syst Rev. 2015;2015(8):Cd003677.
Johnson CM, Makai GEH. A Systematic Review of Perioperative Opioid Management for Minimally Invasive Hysterectomy. J Minim Invasive Gynecol. 2019;26(2):233-243.
Thompson JC, Komesu YM, Qeadan F, et al. Trends in patient procurement of postoperative opioids and route of hysterectomy in the United States from 2004 through 2014. Am J Obstet Gynecol. 2018;219(5):484.e481-484.e411.
von Elm E, Altman DG, Egger M, et al. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies. Ann Intern Med. 2007;147(8):573-577.
Benchimol EI, Smeeth L, Guttmann A, et al. The REporting of studies Conducted using Observational Routinely-collected health Data (RECORD) statement. PLoS Med. 2015;12(10):e1001885.
Peahl AF, Morgan DM, Dalton VK, et al. New persistent opioid use after acute opioid prescribing in pregnancy: a nationwide analysis. Am J Obstet Gynecol. 2020.
Centers for Disease Control and Prevention. Calculating total daily dose of opioids for safer dosage. https://www.cdc.gov/drugoverdose/pdf/calculating_total_daily_dose-a.pdf. Published 2019. Accessed.
Charlson ME, Pompei P, Ales KL, MacKenzie CR. A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chronic Dis. 1987;40(5):373-383.
Kurdyak P, Lin E, Green D, Vigod S. Validation of a Population-Based Algorithm to Detect Chronic Psychotic Illness. Can J Psychiatry. 2015;60(8):362-368.
Steele LS, Glazier RH, Lin E, Evans M. Using administrative data to measure ambulatory mental health service provision in primary care. Med Care. 2004;42(10):960-965.
Austin PC. Using the Standardized Difference to Compare the Prevalence of a Binary Variable Between Two Groups in Observational Research. Communications in Statistics – Simulation and Computation. 2009;38(6):1228-1234.
Wright JD, Huang Y, Melamed A, et al. Use and Misuse of Opioids After Gynecologic Surgical Procedures. Obstet Gynecol. 2019;134(2):250-260.
Young JC, Wu JM, Willis-Gray M, Pate V, Jonsson Funk M. Persistent Opioid Use After Hysterectomy in the United States, 2005-2015. Obstet Gynecol. 2020;135(1):123-132.
Ladha KS, Neuman MD, Broms G, et al. Opioid Prescribing After Surgery in the United States, Canada, and Sweden. JAMA Netw Open. 2019;2(9):e1910734.
Vu JV, Howard RA, Gunaseelan V, Brummett CM, Waljee JF, Englesbe MJ. Statewide Implementation of Postoperative Opioid Prescribing Guidelines. N Engl J Med. 2019;381(7):680-682.
Opioid Prescribing Engagement Network (OPEN). Prescribing Recommendations. https://michigan-open.org/prescribing-recommendations/. Published 2020. Accessed2020.
Bree Collaborative WSAMD. Prescribing Opioids for Post-Operative Pain. http://www.agencymeddirectors.wa.gov/Files/FinalSupBreeAMDGPostopPain091318wcover.pdf. Published 2018. Accessed June 13, 2020.
Madsen AM, Stark LM, Has P, Emerson JB, Schulkin J, Matteson KA. Opioid Knowledge and Prescribing Practices Among Obstetrician-Gynecologists. Obstet Gynecol. 2018;131(1):150-157.
Barnett ML, Olenski AR, Jena AB. Opioid-Prescribing Patterns of Emergency Physicians and Risk of Long-Term Use. N Engl J Med. 2017;376(7):663-673.
As-Sanie S, Till SR, Mowers EL, et al. Opioid Prescribing Patterns, Patient Use, and Postoperative Pain After Hysterectomy for Benign Indications. Obstet Gynecol. 2017;130(6):1261-1268.
Badreldin N, Grobman WA, Chang KT, Yee LM. Patient and Health Care Provider Factors Associated With Prescription of Opioids After Delivery. Obstet Gynecol. 2018;132(4):929-936.
Appendix A. Data Sources
Canadian Institute for Health Information Discharge Abstract Database (CIHI-DAD)1 for all hospital visits.
National Ambulatory Care Reporting System (CIHI-NACRS) for emergency department (ED) visits.
Same Day Surgery Database (CIHI-SDS) for day surgeries.
Ontario Health Insurance Plan (OHIP) for physician billings.
Corporate Provider Database (CPDB) for physician demographic data.
Registered Persons Database (RPDB) for provincial vital statistics.
Ontario Cancer Registry (OCR) for all patients with a cancer diagnosis.
Narcotic Monitoring System (NMS), which records all legally dispensed opioids in the province, regardless of insurance status 2.
Appendix B. Code Lists
Received in revised form:
In Press Journal Pre-Proof
Conflict of interest statement: A.M. is a member of speaker bureaus and advisory boards for AbbVie, Allergan, Bayer, and Hologic, and also a member of the speaker bureau for Medtronic. The remaining authors report no conflicts of interest.
Source of funding: Financial support for this research was provided by the Department of Obstetrics and Gynaecology at Mount Sinai Hospital and University of Toronto. The sponsors did not have any role in study design, data collection/analysis/interpretation, writing of the manuscript and decision to submit for publication.
Statement of prior presentation: This study was presented at the AAGL20 Virtual Global Congress November, 2020.
IRB Exemption Statement: The use of data in this project was authorized under section 45 of Ontario’s Personal Health Information Protection Act, which does not require review by a Research Ethics Board.
Data available upon request.
Précis Perioperative opioid use after benign gynecologic surgery has significantly decreased from 2013 to 2018, however, remains an independent risk factor for persistent opioid use.
© 2021 Published by Elsevier Inc. on behalf of AAGL.