Abstract
Background
It is unknown if opioid naïve patients who undergo minimally invasive, benign foregut operations are at risk for progressing to persistent postoperative opioid use. The purpose of our study was to determine if opioid naïve patients who undergo minimally invasive, benign foregut operations progress to persistent postoperative opioid use and to identify any patient- and surgery-specific factors associated with persistent postoperative opioid use.
Methods
Opioid-naïve, adult patients who underwent laparoscopic fundoplication, hiatal hernia repair, or Heller myotomy from 2010 to 2018 were identified within the IBM® MarketScan® Commercial Claims and Encounters Database. Daily drug logs of the preoperative and postoperative period were evaluated to assess for changes in drug use patters. The primary outcome of interest was persistent postoperative opioid use, defined as at least 33% of the proportion of days covered by opioid prescriptions at 365-day follow-up. Patient demographic information and clinical risk factors for persistent postoperative opioid use at 365 days postoperatively were estimated using log-binomial regression.
Results
A total of 17,530 patients met inclusion criteria; 6895 underwent fundoplication, 9235 underwent hiatal hernia repair, and 1400 underwent Heller myotomy. 9652 patients had at least one opioid prescription filled in the perioperative period. Sixty-five patients (0.4%) were found to have persistent postoperative opioid use at 365 days postoperatively. Lower Charlson comorbidity index scores and a history of mental illness or substance use disorder had a statistically but not clinically significant protective effect on the risk of persistent postoperative opioid use at 365 days postoperatively.
Conclusions
Only half of opioid naïve patients undergoing minimally invasive, benign foregut operations filled an opioid prescription postoperatively. The risk of progression to persistent postoperative opioid use was less than 1%. These findings support the current guidelines that limit the number of opioid pills prescribed following general surgery operations.
Keywords: Foregut surgery, Fundoplication, Hiatal hernia, Heller myotomy, Opioid abuse, Outcomes
In the early 1990s, upwards of 80% of physician visits were addressing acute pain [1]. In an effort to improve the management of acute pain, thereby reducing the number of physician visits related to acute pain and the incidence of progression to chronic pain, pain was designated as the fifth vital sign in 1995 [1, 2]. While the designation of pain as a vital sign was well-intended, there were limited nonopioid medications available at that time to manage acute pain [3]. As a result, the quality of pain management did not improve [2]. Rather, there was an observed increase in opioid addiction and opioid related deaths, with pain as a fifth vital sign contributing to the opioid epidemic [4].
Opioid prescribing practices have changed significantly in the past decade due to an increased recognition of the opioid epidemic in the United States [5]. For example, there were 81 opioid prescriptions per 100 people in the United States in 2010 compared to 58 opioid prescriptions per 100 people in the United States in 2017 [6]. Nevertheless, the amount of opioids prescribed in morphine milligram equivalents (MME) per person is currently still three times greater now than it was in 1999 [6]. To put this into perspective, per prescription, the current average daily amount of opioids is more than 45 MME.
Surgery is a common setting in which patients are exposed to opioid medications. Opioids are effective at treating acute surgical pain [7]. However, up to 10% of opioid-naïve patients continue to take opioids 1 year after surgery [3]. One of the proposed benefits of the development of enhanced recovery after surgery (ERAS) pathways and the introduction of opioid-alternatives is to help reduce the opioid requirement after surgery [3]. Nevertheless, common surgical procedures and opioid prescribing practices vary significantly between surgeons, institutions, and geographically [8, 9].
To date, no study has investigated the risk of persistent opioid use for opioid-naïve patients following minimally invasive, benign foregut surgery. Further, there are currently no established guidelines for opioid prescribing practices following minimally invasive, benign foregut surgery. Therefore, the purpose of our study was to (1) detail opioid prescribing practices following minimally invasive, benign foregut surgery, (2) describe the prevalence of persistent postoperative opioid use among opioid-naïve patients who undergo minimally invasive, benign foregut surgery and (3) to identify any patient- and surgery-specific factors associated with persistent postoperative opioid use in opioid-naïve patients following minimally invasive, benign foregut surgery using the IBM® MarketScan® Commercial Claims and Encounters Database.
Materials and methods
Study design and population
We conducted a retrospective cohort study using IBM® MarketScan® Commercial Claims and Encounters data from 2010 to 2018. These data consist of medical encounter and prescription drug insurance claims for several million individuals enrolled in employer-sponsored health insurance, including employees, spouses, and their dependents. We identified a cohort of adults, ages ≥ 18 years of age to ≤ 64 years of age, who had a claim for elective, minimally invasive fundoplication [Current Procedural Technology (CPT) 43280], hiatal hernia repair without mesh (CPT 43281) or with mesh (CPT 43282), or Heller myotomy (CPT 43279) between 2010 and 2018. Patients were grouped based on the primary CPT code of their foregut operation performed in order to determine if there were any differences between the groups with respect to patient demographic variables, region and year of surgery, as well as the risk for progression to persistent postoperative opioid use. Because peroral endoscopic myotomy uses an unlisted CPT code and because the LINX procedure CPT code was established after 2010, patients who underwent either of these operations were excluded from our analysis. Additional patients excluded from our analysis included pediatric patients and patients who underwent either open or non-elective benign foregut surgery. Non-elective cases were captured based on CPT and IBM® MarketScan® codes for emergency services. These codes are listed in Supplemental Table 1. Table 1 summarizes the inclusion and exclusion criteria for this study.
Table 1.
Cohort selection criteria to identify patients undergoing foregut surgery in the IBM Watson health MarketScan commercial claims and encounters database from 2010 to 2018
| Step | Description | N |
|---|---|---|
| 1 | Claim for foregut surgery between 2001 and 2018a | 133,385 |
| 2 | At least 18 years of age on index date | 128,149 |
| 3 | Continuous enrollment during the 180 days prior to the index dateb | 81,171 |
| 4 | Continuous enrollment during the 30 days after the index date2 (perioperative period) | 78,072 |
| 5 | No preoperative opioid usec | 56,258 |
| 6 | Not an emergency procedure | 55,250 |
| 7 | No additional major invasive procedures on the index date | 24,371 |
| 8 | Procedure between 2010 and 2018 | 17,530 |
Claims for foregut surgery were identified based on Current Procedural Technology (CPT) codes 43,279 (Heller myotomy), 43,280 (fundoplication) and 43,281 and 43,282 (hiatal hernia repair). The date of the first claim with a CPT code for foregut surgery was the index date
Individuals were required to be enrolled in a fee-for-service plan with prescription benefits
Preoperative opioid use was defined as at least one claim for an opioid during the 180–15 days prior to the index date or at least two claims for an opioid during the 14 days prior to the index date. Claims for opioids were identified based on National Drug Codes
We required patients to have continuous insurance enrollment during the immediate 180 days prior to surgery to evaluate preoperative opioid use and baseline characteristics. Thirty days of continuous postoperative enrollment were also required to identify surgery-related complications. Individuals with preoperative opioid use, defined as at least one claim for an opioid prescription during the preceding 180–15 days prior to the surgery date or at least two claims during the 14 days prior to the surgery date, were excluded so as to minimize the likelihood of including non-naïve opioid patients in our analysis. In such patients, based on the limitations of the IBM® MarketScan® database, we would be unable to determine that persistent postoperative opioid use was related to their minimally invasive, benign foregut surgery. Individuals with a claim for another major invasive procedure on the date of surgery, identified using the Healthcare Utilization Project’s Clinical Classification Software, were also excluded. [10].
Opioid use assessment
Prescription fills for opioids were identified in the pharmacy claims using National Drug Codes and the Centers for Disease Control and Prevention Injury Center data resource files [11]. Proportion of days covered by opioids was estimated by creating daily logs of opioids presumed to be on hand based on opioid dispensing dates and the number of days supplied. For subsequent opioid prescriptions that occurred prior to the end of the number of days supplied for a prior prescription, the new prescription was assumed to begin after the end of the prior prescription. Proportion of days covered was then calculated as the total number of days supplied during the follow-up period, divided by the duration of the follow-up period. Individuals were censored at the end of continuous enrollment. Persistent postoperative opioid use was defined as at least 33% of the proportion of days covered by opioids during the following windows: 30-, 90-, 180-, and 365-days post-surgery. The definition of persistent postoperative opioid use is based on the article published by Lee et al. which used a cut-off of 120 opioid prescription days in 1 year to define chronic opioid use [12]. This proportion of opioid prescription days covered is equivalent to 33% (120/365) and therefore was used to determine persistent postoperative opioid use at additional postoperative time points. Only patients with complete follow-up during each time period window were included in the respective analyses.
Surgery complications
Surgery setting (inpatient vs. outpatient) and length of stay among admitted individuals were assessed using the inpatient and outpatient claims files. Surgery-related complications, including intubation, blood transfusion, infections (pneumonia, surgical-site, blood stream, and urinary tract), acute kidney injury, and pulmonary embolism (PE) or deep venous thrombosis (DVT), were identified based on International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM), ICD-10-CM codes, CPT codes, and Healthcare Common Procedure Coding System (HCPCS) codes in the 30 days following surgery. Emergency department visits, hospital readmission, reinterventions related to the index operation, including subsequent endoscopy or foregut surgery, and other invasive procedures during the 30 days following surgery were also abstracted. Please refer to Supplementary Table 1 for the full list of codes used.
Covariates
Patient demographics (age, sex, region of surgery) were identified on the date of the surgery. Comorbidities, including mental illness or substance use disorder, obstructive sleep apnea (OSA), hypertension (HTN), hyperlipidemia (HLD), history of venous thromboembolism (VTE), and the conditions of the Charlson Comorbidity Index (CCI) [13], were identified based on ICD-9-CM and ICD-10-CM codes during the 180 days prior to the surgery. Please refer to Supplementary Table 1 for the full list of codes used.
Statistical analysis
Patient demographics, indications for surgery, comorbidities, and 30-day postoperative complications were described overall and by type of foregut surgery using frequency statistics for categorical variables and means, standard deviations, and medians for continuous variables. Standardized differences were used to compare differences in patient characteristics between surgery groups [10]. An absolute value > 0.20 was considered a meaningful difference. We described perioperative opioid fills based on type of opioid dispensed (hydrocodone, oxycodone, codeine, tramadol, hydromorphone, morphine, fentanyl, oxymorphone, and other) and the days supplied for the first perioperative opioid fill. The number of unintentional opioid overdoses or deaths within the first 365 days following surgery were captured [14]. Persistent postoperative opioid use was described using frequency distributions. Crude risk ratios were estimated using bivariate log-binomial regression models with 365-day persistent opioid use as the dependent variable and patient and surgery characteristics as the independent variables. Only adults with ≥ 365 days follow-up were included in this analysis. Due to a low number of persistent opioid outcomes, adjusted models were not conducted.
All statistical analyses were completed using SAS version 9.4 (Cary, NC) and p < 0.05 was considered statistically significant. This study was exempt from review by the Institutional Review Board at the University of North Carolina at Chapel Hill (IRB # 19-2857).
Results
A total of 17,530 patients within the IBM® MarketScan® Commercial Claims and Encounters database underwent elective, minimally invasive benign foregut surgery from 2010 to 2018; 6895 (39.3%) underwent fundoplication, 9235 (52.7%) underwent hiatal hernia repair, and 1400 (8.0%) underwent Heller myotomy. Patients were divided into these groups in order to evaluate Patient demographic variables and comorbidities are displayed in Table 2. The median age at the time of surgery was 51 years of age and 66% of the patients were female. Differences between the surgical groups, as assessed by standard differences, existed. With respect to those patients who underwent hiatal hernia repair versus those patients who underwent fundoplication, patients who underwent hiatal hernia repair were more likely to have surgery from 2016 to 2018 and less likely to have surgery from 2010 to 2012 compared to patients who underwent fundoplication. With respect to those patients who underwent Heller myotomy versus those patients who underwent fundoplication, patients who underwent Heller myotomy were younger and were more likely to have surgery from 2010 to 2012 and less likely to have surgery from 2016 to 2018 compared to those patients who underwent fundoplication.
Table 2.
Demographic and surgical characteristics for patients undergoing foregut surgery in the IBM Watson health MarketScan commercial claims and encounters database from 2010 to 2018
| Type of foregut surgerya | ||||||
|---|---|---|---|---|---|---|
| Characteristic | Overall | Fundoplication | Hiatal hernia repair | Standardized Differencee | Heller myotomy | Standardized Differencee |
| N = 17,530 | N = 6895 | N = 9235 | Hiatal hernia repair vs. fundoplication |
N = 1400 | Heller myotomy vs. fundoplication |
|
| Age,b mean, SD (median) | 48.8, 11.2 (51.0) | 48.6, 11.2 (51.0) | 49.5, 11.0 (52.0) | 0.080 | 45.0, 12.0 (46.0) | 0.311 |
| Age category,b n (%) | 0.050 | 0.322 | ||||
| 18–29 | 1223 (7.0) | 514 (7.5) | 540 (5.8) | 169 (12.1) | ||
| 30–39 | 2600 (14.8) | 989 (14.3) | 1314 (14.2) | 297 (21.2) | ||
| 40–49 | 4164 (23.8) | 1681 (24.4) | 2114 (22.9) | 369 (26.4) | ||
| 50–59 | 6105 (34.8) | 2419 (35.1) | 3293 (35.7) | 393 (28.1) | ||
| 60–64 | 3438 (19.6) | 1292 (18.7) | 1974 (21.4) | 172 (12.3) | ||
| Female, n (%) | 11,476 (65.5) | 4269 (61.9) | 6548 (70.9) | 0.190 | 659 (47.1) | 0.301 |
| Year of surgery | 0.520 | 0.300 | ||||
| 2010–2012 | 6470 (36.9) | 3417 (49.6) | 2530 (27.4) | 523 (37.4) | ||
| 2013–2015 | 5830 (33.3) | 2037 (29.5) | 3339 (36.2) | 454 (32.4) | ||
| 2016–2018 | 5230 (29.8) | 1441 (20.9) | 3366 (36.4) | 423 (30.2) | ||
| Region of surgery, n (%) | 0.230 | 0.184 | ||||
| South | 9003 (51.8) | 3178 (46.6) | 5223 (56.9) | 602 (43.4) | ||
| North Central | 3503 (20.2) | 1679 (24.6) | 1520 (16.6) | 304 (21.9) | ||
| West | 2916 (16.8) | 1220 (17.9) | 1450 (15.8) | 246 (17.7) | ||
| Northeast | 1962 (11.3) | 748 (11.0) | 980 (10.7) | 234 (16.9) | ||
| Missing | 146 | 70 | 62 | 14 | ||
| Mental illness or substance use disorder,c n (%) | 4254 (24.3) | 1538 (22.3) | 2493 (27.0) | 0.110 | 223 (15.9) | 0.163 |
| Obstructive sleep apneac | 1768 (10.1) | 514 (7.5) | 1181 (12.8) | 0.180 | 73 (5.2) | 0.092 |
| Hypertensionc | 5822 (33.2) | 2043 (29.6) | 3438 (37.2) | 0.160 | 341 (24.4) | 0.119 |
| Hyperlipidemiac | 4837 (27.6) | 1830 (26.5) | 2749 (29.8) | 0.070 | 258 (18.4) | 0.195 |
| Venous thromboembolismc | 122 (0.7) | 35 (0.5) | 81 (0.9) | 0.040 | 6 (0.4) | 0.012 |
| CCI,d n (%) | 0.090 | 0.169 | ||||
| 0 | 12,027 (68.6) | 4823 (69.9) | 6123 (66.3) | 1081 (77.2) | ||
| 1 | 3839 (21.9) | 1470 (21.3) | 2160 (23.4) | 209 (14.9) | ||
| 2 | 1161 (6.6) | 421 (6.1) | 672 (7.3) | 68 (4.9) | ||
| 3 + | 503 (2.9) | 181 (2.6) | 280 (3.0) | 42 (3.0) | ||
Claims for foregut surgery were identified based on Current Procedural Technology (CPT) codes 43,279 (Heller myotomy; new in 2009), 43,281 and 43,282 (hiatal hernia repair; 43,281 was new in 2010), and 43,280 (fundoplication; new in 2010)
Age was evaluated on the date of the index surgery
Comorbidities were assessed using diagnosis codes in the 180 days prior to the index surgery
The Charlson Comorbidity Score (CCI) was calculated based on diagnosis codes in the 180 days prior to the index surgery. The CCI is comprise of 17 conditions with a weight assigned to each condition based on its potential influence on mortality. The weights are then summed to obtain a comorbidity index for each patient. The index accounts for both the number and severity of conditions, with higher scores indicating a higher burden of comorbid disease
Standardized mean difference > 0.2 represents a significant difference in the distribution of the variable between surgery types
Operative details and 30-day surgical complications are detailed in Table 3. The average length of hospital stay for all inpatient procedures was 2 days. No statistically significant differences existed between the three groups with respect to 30-day surgical complications.
Table 3.
Surgery characteristics and complications for patients undergoing foregut surgery in the IBM Watson health MarketScan commercial claims and encounters database from 2010 to 2018
| Characteristic | Type of foregut surgery | |||||
|---|---|---|---|---|---|---|
| Overall |
Fundoplication |
Hiatal hernia repair | Standardized differencec | Heller myotomy | Standardized Difference6c |
|
| N = 17,530 | N = 6895 | N = 9235 | Hiatal hernia repair vs. fundoplication |
N = 1400 | Heller myotomy vs. fundoplication |
|
| Inpatient index surgery, n (%) | 4,807 (27.4) | 1,569 (22.8) | 2,490 (27.0) | 0.035 | 748 (53.4) | 0.013 |
| Index hospitalization LOS,a mean, SD (median) | 2.1, 2.5 (2) | 2.0, 3.1 (1.0) | 2.1, 2.2 (2.0) | 0.035 | 2.1, 2.3 (2.0) | 0.013 |
| Intubation,b n (%) | 27 (0.2) | 9 (0.1) | 16 (0.2) | 0.011 | 2 (0.1) | 0.003 |
| Blood transfusion,b n (%) | 9 (0.1) | 2 (0.0) | 7 (0.1) | 0.020 | 0 | 0.024 |
| Infection,b n (%) | ||||||
| Urinary tract infection | 280 (1.6) | 99 (1.4) | 168 (1.8) | 0.030 | 13 (0.9) | 0.047 |
| Pneumonia | 226 (1.3) | 78 (1.1) | 135 (1.5) | 0.029 | 13 (0.9) | 0.020 |
| Surgical-site infection | 220 (1.3) | 78 (1.1) | 123 (1.3) | 0.018 | 19 (1.4) | 0.020 |
| Blood stream infection | 73 (0.4) | 19 (0.3) | 46 (0.5) | 0.036 | 8 (0.6) | 0.046 |
| Acute kidney injury,b n (%) | 81 (0.5) | 23 (0.3) | 55 (0.6) | 0.039 | 3 (0.2) | 0.023 |
| Pulmonary embolism or DVT,b n (%) | 180 (1.0) | 48 (0.7) | 128 (1.4) | 0.068 | 4 (0.3) | 0.059 |
| ED visit,b n (%) | 1372 (7.8) | 529 (7.7) | 760 (8.2) | 0.021 | 83 (5.9) | 0.069 |
| Hospital readmission,b n (%) | 758 (4.3) | 251 (3.6) | 467 (5.1) | 0.070 | 40 (2.9) | 0.044 |
| Reintervention,b n (%) | ||||||
| Foregut surgery | 957 (5.5) | 420 (6.1) | 474 (5.1) | 0.042 | 63 (4.5) | 0.071 |
| Endoscopy | 410 (2.3) | 149 (2.2) | 223 (2.4) | 0.017 | 38 (2.7) | 0.038 |
| Other invasive procedure,b n (%) | 339 (1.9) | 100 (1.5) | 214 (2.3) | 0.064 | 25 (1.8) | 0.027 |
Calculated only among those who were admitted to the hospital for the index surgery
Identified within 30 days following the index surgery
Standardized mean difference > 0.2 represents a significant difference in the distribution of the variable between surgery types
Perioperative and long-term opioid use is detailed in Table 4. Of the total patients included in this study, 9652 (55.1%) had at least one opioid prescription filled in the perioperative period. On average, patients were prescribed their first opioid prescription on or after their scheduled day of surgery. Hydrocodone was the most prescribed narcotic (40.4%) followed by oxycodone (39.7%). The median days’ supply of the initial perioperative narcotic prescription was 5 days (interquartile range 2–6 days) (Fig. 1). Persistent postoperative opioid use, defined as ≥ 33% of the proportion of days covered by opioids, was most common during the first 30 days postoperatively for all three groups (5.9%). Persistent postoperative opioid use declined from 30 days postoperatively to 90 days postoperatively for all three groups and remained < 1% of each surgical group at 180 days and 365 days postoperatively (Fig. 2). A total of 38 (0.4%) patients were diagnosed with opioid abuse or dependence or experienced opioid poisoning within the first year following their foregut surgery; 9 (0.1%) patients who underwent fundoplication, 25 (0.3%) patients who underwent hiatal hernia repair, and 4 (0.5%) who underwent Heller myotomy.
Table 4.
Perioperative, postoperative, and persistent opioid use for patients undergoing foregut surgery in the IBM Watson health MarketScan commercial claims and encounters database from 2010 to 2018
| Characteristic | Type of foregut surgery | |||
|---|---|---|---|---|
| Overall | Fundoplication | Hiatal hernia repair |
Heller myotomy |
|
| N = 6895 | N = 9235 | N = 1400 | ||
| Perioperative opioid fill (days – 14 to 14 days post surgery) | ||||
| Any opioid, n (%) | 9652 (55.1) | 3734 (54.2) | 5185 (56.1) | 733 (52.4) |
| Filled before index date | 1,036 (10.7) | 245 (6.6) | 769 (14.8) | 22 (3.0) |
| Filled on or after index date | 8616 (89.3) | 3489 (93.4) | 4416 (85.2) | 711 (97.0) |
| Generic name of first perioperative opioid fill,a n (%) | ||||
| Hydrocodone | 3901 (40.4) | 1707 (45.7) | 1989 (38.4) | 205 (28.0) |
| Oxycodone | 3834 (39.7) | 1483 (39.7) | 1924 (37.1) | 427 (58.3) |
| Codeine | 756 (7.8) | 171 (4.6) | 544 (10.5) | 41 (5.6) |
| Tramadol | 607 (6.3) | 176 (4.7) | 403 (7.8) | 28 (3.8) |
| Hydromorphone | 323 (3.3) | 111 (3.0) | 190 (3.7) | 22 (3.0) |
| Morphine | 19 (0.2) | 5 (0.1) | 11 (0.2) | 3 (0.4) |
| Fentanyl | 1 (0.0) | 0 | 1 (0.0) | 0 |
| Oxymorphone | 0 | 0 | 0 | 0 |
| Othera | 139 (1.4) | 60 (1.6) | 76 (1.5) | 3 (0.4) |
| Multipleb | 72 (0.7) | 21 (0.6) | 47 (0.9) | 4 (0.5) |
| Mean days supplied for first perioperative opioid fill,c median (IQR) | 5 (2.6) | 5 (2.6) | 5 (2.7) | 5 (2.7) |
| Any postoperative opioid fill during postoperative follow-up window (days), d n (%) | ||||
| 15–30 (n = 16,575) | 625 (3.8) | 239 (3.7) | 351 (4.0) | 35 (2.6) |
| 31–60 (n = 15,839) | 603 (3.8) | 236 (3.8) | 332 (4.0) | 35 (2.7) |
| 61–90 (n = 15,083) | 500 (3.3) | 221 (3.7) | 262 (3.3) | 17 (1.4) |
| 91–180 (n = 13,156) | 1354 (10.3) | 572 (11.0) | 717 (10.4) | 65 (6.0) |
| 181–365 (n = 10,235) | 1910 (18.7) | 806 (19.5) | 1015 (19.4) | 89 (10.2) |
| Persistent postoperative opioid use,d,e n (%) | ||||
| 30 days (n = 16,575) | 978 (5.9) | 333 (5.1) | 7559 (6.4) | 86 (6.4) |
| 90 days (n = 15,083) | 137 (0.9) | 63 (1.1) | 67 (1.0) | 7 (0.6) |
| 180 days (n = 13,156) | 75 (0.6) | 41 (0.8) | 32 (0.5) | 2 (0.2) |
| 365 days (n = 10,235) | 65 (0.6) | 33 (0.8) | 29 (0.6) | 3 (0.3) |
| Opioid abuse, dependence, or poisoning within 365 days following surgery,f n (%) | 38 (0.4) | 9 (0.2) | 25 (0.5) | 4 (0.5) |
The most frequently reported other opioid fills were propoxyphene napsylate and meperidine
Includes individuals who filled ≥ 2 prescriptions for different generic names on the date of the first perioperative prescription fill
For individuals who filled ≥ 2 prescriptions on the date of the first perioperative prescription fill, the maximum days supply is reported
Proportions are calculated among people with complete follow-up during the respective postoperative follow-up window. Follow-up time was censored at the date of a second invasive procedure
Persistent postoperative opioid use was defined as ≥ 33.3% proportion of days covered during the respective follow-up window. Proportion of days covered was estimated by creating daily logs of opioids presumed to be on hand based on opioid dispensing dates and days supply. For dispensings that occurred prior to the end of days supply for a prior prescription, the new prescription was assumed to begin after the end of the prior prescription. Proportion of days covered was then calculated as the total number of days supplied during the follow-up period, divided by the duration of the follow-up period
Individuals with less than 365 days of follow-up were excluded from the denominator
Fig. 1.
Any postoperative opioid fill during postoperative follow-up window (days)
Fig. 2.
Persistent postoperative opioid use at 30, 90, 180 and days postoperative
Table 5 details the risk factors for persistent opioid use in opioid naïve patients following elective, minimally invasive benign foregut surgery. For this analysis, only patients with 1-year follow-up data available were included (N = 10,235; 58.4%) A total of 65 (0.6%) patients were identified as having persistent postoperative opioid use. A lower CCI score and a mental illness or substance use disorder diagnosis was associated with a statistically but not clinically significant lower relative risk of persistent opioid use. The type of foregut operation performed, the indication for surgery, patient’s age at the time of surgery, patient sex, the region where surgery was performed, the year the operation was performed, and 30-day morbidity were not associated with persistent opioid use.
Table 5.
Risk factors for persistent opioid use following foregut surgery in the IBM Watson health MarketScan commercial claims and encounters databsae from 2010 to 2018
| Risk factor | Persistent opioid use (n) |
Total | Crude RRa (95% CI) | p value |
|---|---|---|---|---|
| Surgical procedure | ||||
| Fundoplication | 33 | 6895 | 1. (ref) | |
| Hiatal hernia repair | 29 | 9235 | 1.002 (0.999, 1.006) | 0.15 |
| Heller myotomy | 3 | 1400 | 1.004 (1.000, 1.009) | 0.06 |
| Age category | ||||
| 18–29 | 5 | 1223 | 1. (ref) | |
| 30–39 | 7 | 2600 | 1.003 (0.996, 1.011) | 0.40 |
| 40–49 | 18 | 4164 | 1.001 (0.993, 1.008) | 0.89 |
| 50–59 | 22 | 6105 | 1.002 (0.995, 1.010) | 0.56 |
| 60–64 | 13 | 3438 | 1.001 (0.998, 1.004) | 0.88 |
| Region | ||||
| South | 30 | 9003 | 1. (ref) | |
| North Central | 14 | 3503 | 1.082 (1.05, 1.11) | 0.70 |
| West | 15 | 2916 | 1.025 (1.00, 1.05) | 0.30 |
| Northeast | 5 | 1962 | 1.03 (1.00, 1.06) | 0.55 |
| Sex | ||||
| Male | 20 | 6054 | 1. (ref) | |
| Female | 45 | 11,476 | 1.001 (0.998, 1.004) | 0.52 |
| CCI | ||||
| 0 | 31 | 12,027 | 1. (ref) | |
| 1 | 23 | 3839 | 0.994 (0.989, 0.998) | 0.01 |
| 2 | 10 | 1161 | 0.989 (0.980, 0.999) | 0.03 |
| 3+ | 1 | 503 | 1.001 (0.994, 1.008) | 0.81 |
| Mental illness or substance use disorder | 30 | 4254 | 0.991 (0.986, 0.996) | 0.00 |
| Year of surgery | ||||
| 2010–2012 | 34 | 6470 | 1. (ref) | |
| 2013–2015 | 19 | 5830 | 1.000 (1.000, 1.001) | 0.06 |
| 2016–2018 | 12 | 5230 | 1.003 (1.000, 1.007) | 0.11 |
| Sugery setting | ||||
| Inpatient | 25 | 4807 | 1. (ref) | |
| Outpatient | 40 | 12,723 | 1.003 (1.000, 1.007) | 0.12 |
| ED visit within 30 days of surgery | 7 | 1372 | 0.997 (0.990, 1.004) | 0.37 |
| Hospital readmission within 30 days of surgery | 7 | 758 | 0.988 (0.975, 1.001) | 0.08 |
| Reintervention within 30 days | 3 | 1,239 | 0.987 (0.966, 1.009) | 0.24 |
Persistent postoperative opioid use was defined as ≥ 33.3% proportion of days covered during the 365 days following the index surgery. Proportion of days covered was estimated by creating daily logs of opioids presumed to be on hand based on opioid dispensing dates and days supply. For dispensings that occurred prior to the end of days supply for a prior prescription, the new prescription was assumed to begin after the end of the prior prescription. Proportion of days covered was then calculated as the total number of days supplied during the follow-up period, divided by the duration of the follow-up period. Individuals with less than 365 days of follow-up were excluded from the analysis
Crude risk ratios were estimated using log-binomial model with persistent opioid use as the dependent variable and the respective risk factor as an independent variable. Due to low number of events, no adjusted models were conducted
Discussion
Opioid overdose has been the leading cause of accidental death in the United States since 2010 [15, 16]. While there are several factors that have contributed to the rising number of opioid deaths in the United States, physician prescribing practices is one modifiable factor that warrants scrutiny. In a study by Lev et al. investigating drug overdose deaths in San Diego County, they found that surgeons accounted for 7% of all opioid prescriptions and that the average opioid prescription prescribed by a surgeon was for 123 tablets [17]. Further, a recent study by Lazar et al. found that for general surgery operations performed at their institution, the number of opioids prescribed at discharge increased as the week progressed and as the number of days in the hospital increased. [18].
Several studies have detailed opioid prescribing practices following general surgery operations. To our knowledge, however, minimally invasive benign foregut general surgery operations, including fundoplication, hiatal hernia repair, and Heller myotomy, have not been included in these analyses [8, 9, 19, 20]. Therefore, this is the first study to investigate the incidence of, and factors associated with, persistent opioid use in opioid-naïve patients following minimally invasive benign foregut operations. Our study found that only approximately half of all patients who undergo a minimally invasive benign foregut operation fill an opioid prescription in the perioperative period. Further, the rate of persistent postoperative opioid use in opioid-naïve patients at 1 year following minimally invasive benign foregut surgery is < 1% and is not associated with the type of procedure performed, the year of the operation, or the region where the surgery was performed.
It is well-accepted that there is significant in-state and regional variation in opioid prescribing practices [21]. One potential explanation for this variation is a difference in the proportion of specific surgical procedures performed both within different parts of a state as well as across geographic regions of the United States [6, 21-23]. For example, primary care physicians in the South and the West have previously been shown to be more likely to prescribe opioids for pain [17]. Furthermore, patients undergoing lumbar spine surgery in the South and the West have previously been shown to be more likely to progress to prolonged opioid use [22]. Interestingly, our study did not find an association between the region in which foregut surgery was performed and persistent postoperative opioid use. Based on the results of our study, it is possible that less regional variation exists for postoperative opioid prescribing in minimally invasive, benign foregut surgery. Additional studies are needed to further investigate regional opioid prescribing practices following benign foregut surgery.
Another finding of our study that we found interesting was the lack of association of a mental illness or substance use disorder with an increased risk of persistent postoperative opioid use in opioid-naïve patients. Several previous studies have shown an association between preoperative opioid use, a mental illness diagnosis, and/or a history of opioid abuse and an increased risk for chronic postoperative opioid use [24]. Almost one-quarter of the patients in our study had a prior mental illness or substance use disorder diagnosed within 180 days of their foregut operation. Nevertheless, only 55% of patients received an opioid prescription for postoperative pain. It is possible that with an increased awareness of the opioid epidemic and the known association of a history of mental illness and chronic postoperative opioid use, foregut surgeons have increased the utilization of multi-modal and opioid-sparing analgesia in addition to limiting narcotic prescriptions and refills. On the other hand, because risk factors for persistent opioid use were only investigated among those with complete 365-day follow-up, it is possible that patients with a severe mental illness or substance use disorder were lost-to-follow-up due to loss of insurance (e.g. unemployment) before this time period, therefore excluding them from this analysis and potentially biasing our results.
According to the Centers for Disease Control and Prevention, the treatment of acute pain should be limited to less than 7 days [9]. One risk factor that contributes to the variation in postoperative opioid prescribing practices includes prescribing a default number of opioid tablets for postoperative pain control without reviewing a patient’s opioid requirement while in the hospital [3]. In some cases, as seen with the study by Lev et al., a high number of opioids are prescribed in order to minimize postoperative patient phone calls and emergency department visits. Nevertheless, one of the tenants of opioid stewardship, which are efforts designed to promote the safe use of opioids, is a patient-centered approach to postoperative opioid prescribing practices [24]. It is our hope that the results from our study serve as a guide for the development of opioid stewardship and prescribing recommendations following foregut surgery. For example, in our study, the median number of days of opioid tablets prescribed postoperatively was 5 days, which is less than the recommended 7 days.
Our study is not without limitations. First, the definition of persistent postoperative opioid use is not well-established in surgical literature [3]. In fact, a recent systematic review found that there are currently 29 distinct definitions for persistent postoperative opioid use [25]. Therefore, while we found no clinically significant association between patient- or surgery-specific risk factors and the incidence of persistent postoperative opioid use following minimally invasive, benign foregut surgery, it is possible that these results could change based on the definition of persistent postoperative opioid use. Second, the CPT codes for hiatal hernia repair (CPT 43281 and 43282) and Heller myotomy (43279) include any fundoplication performed at the time of these operations and the CPT for fundoplication (43280) includes both partial and complete fundoplications. We therefore cannot determine the number of patients that underwent hiatal hernia repair or Heller myotomy with or without fundoplication and we cannot determine the number of patients that underwent partial versus complete fundoplication. Third, our study used an employee-sponsored, private insurance database. Therefore, the results of our study may not be applicable to patients with Medicaid or Medicare or those patients who are uninsured. Further, we were only able to capture outpatient opioid fills that were reimbursed through insurance. Fourth, we used diagnosis codes to capture comorbidities and opioid-related complications. This likely has led to an underestimation of the prevalence of these factors, as undiagnosed conditions or conditions not related to any health care billing during the study period would be missed. Additionally, opioid-related dependence and abuse are particularly under-estimated in claims data as it does not capture incidents occurring out of the hospital [26]. The transition between ICD-9-CM to ICD-10-CM codes in October 2015 could also impact our results, although we saw consistent trends over time. Finally, persistent use analyses were restricted to those who had follow-up data for the entire 1-year postoperative period, which may have introduced some selection bias. It is also possible that the outcomes for persistent postoperative opioid use were over- or underestimated due to lost-to-follow-up during this time period related to death, loss of insurance, or transition to a public insurance program.
Conclusion
Patients undergoing minimally invasive, benign foregut surgery are at low risk for persistent postoperative opioid use at 1 year postoperatively. Additional studies are needed determine what patient- and surgery-specific factors contribute to this relatively low risk of chronic postoperative opioid use in order to reduce the risk of chronic postoperative opioid use following other general surgery procedures.
Supplementary Material
Acknowledgements
The authors would like to acknowledge Virginia Pate, MS for her contributions in data preparation, definition development, and programming for this study.
Funding
Ivy N. Haskins, MD receives royalty from UpToDate, Inc. Emilie D. Duchesneau, MSPH is supported by the Cancer Care Quality Training Program at the University of North Carolina at Chapel Hill (Grant T32 CA 116339). Paula Strassle, PhD, MSPH is supported by the Division of Intramural Research, National Institute on Minority Health and Health Disparities, National Institutes of Health. The contents and views in this manuscript are those of the authors and should not be construed to represent the views of their employers.
Footnotes
Supplementary Information The online version contains supplementary material available at https://doi.org/10.1007/s00464-022-09123-y.
Disclosures Ivy N. Haskins, Emilie D. Duchesneau, Chris B. Agala, Stephanie T. Lumpkin, Paula D. Strassle, and Timothy M. Farrell have not disclosed any funding directly related to this study.
References
- 1.Walid MS, Donahue SN, Darmohray DM, Hyer LA, Robinson JS (2008) The fifth vital sign- what does it mean? Pain Pract 8(6):417–42218662363 [Google Scholar]
- 2.Mularski RA, White-Shu F, Overbay D, Miller L, Asch SM, Ganzini L (2006) Measuring pain as the 5th vital sign does not improve quality of pain management. J Gen Intern Med 21(6):607–612 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Kim MP, Godoy C, Nguyen DT, Meisenbach LM, Chihara R, Chan EY, Graviss EA (2020) Preemptive pain-management program is associated with reduction of opioid prescriptions after benign minimally invasive foregut surgery. J Thorac Cardiovasc Surg 159(2):734–744 [DOI] [PubMed] [Google Scholar]
- 4.Scher C, Meader L, Van Cleave JH, Reid MC (2017) Moving beyond pain as the fifth vital sign and patient satisfaction scores to improve pain care in the 21st century. Pain Manag Nurs 19(2):125–129 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Hill M, McMahon M, Stucke R, Barth R (2017) Wide variation and excessive dosage of opioid prescriptions for common general surgery procedures. Ann Surg 265(4):709–714 [DOI] [PubMed] [Google Scholar]
- 6.Prescription Opioid Data Prescribing Practices (2021) https://www.cdc.gov/drugoverdose/deaths/prescription/practices.html. Accessed 26 July 2021. [Google Scholar]
- 7.Clarke H, Soneji N, Ko DT, Yun L, Wijeysundera DN (2014) Rates and risk factors for prolonged opioid use after major surgery: population based cohort study. BMJ 348:g1251. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Overton HN, Hanna MN, Bruhn WE, Hutfless S, Bicket MC, Makary MA, Opioids after Surgery Workgroup (2018) Opioid-prescribing guidelines for common surgical procedures: an expert panel consensus. J Am Coll Surg 227(4):411–418 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Thiels CA, Ubl DS, Yost KJ, Dowdy SC, Mabry TM, Gazelka HM, Cima RR, Habermann EB (2018) Results of a prospective, multicenter initiative aimed at developing opioid-prescribing guidelines after surgery. Ann Surg 268(3):457–468 [DOI] [PubMed] [Google Scholar]
- 10.Yan D and Dalton JE (2012) A unified approach to measuring the effect size between two groups using SAS®. https://support.sas.com/resources/papers/proceedings12/335-2012.pdf. Accessed 20 Nov 2020. [Google Scholar]
- 11.Analyzing Opioid Prescription Data and Oral Morphine Milligram Equivalents (MME) (2021) https://www.cdc.gov/opioids/data-resources/index.html. Accessed 6 Aug 2021. [Google Scholar]
- 12.Deyo RA, Cherkin DC, Ciol MA (1992) Adapting a clinical comorbidity index for use with ICD-9-CM administrative databases. J Clin Epidemiol 45(6):613–619 [DOI] [PubMed] [Google Scholar]
- 13.Lee JSJ, Edelman AL, Brummett CM, Englesbe MJ et al. (2017) New persistent opioid use among patients with cancer after curative-intent surgery. J Clin Oncol 35(56):4042–4049 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Cochran G, Gordon AJ, Lo-Ciganic WH, Gellad WF et al. (2017) An examination of claims-based predictors of overdose from a large medicaid program. Med Care 55(3):291–298 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Injuries and Violence are Leading Causes of Death (2021) https://www.cdc.gov/injury/wisqars/animated-leading-causes.html#:~:text=from%201981%2D2019-,Unintentional%20injuries%20are%20the%20leading%20cause%20of%20death%20for%20Americans,unintentional%20drowning%2C%20and%20unintentional%20falls. Accessed 20 Aug 2021. [Google Scholar]
- 16.Kim B, Nolan S, Beaulieu T, Shalansky S, Ti L (2019) Inappropriate opioid prescribing practices: a narrative review. Am J Health Syst Pharm 76(16):1231–1237 [DOI] [PubMed] [Google Scholar]
- 17.Lev R, Lee O, Petro S, Lucas J et al. (2016) Who is prescribing controlled medications to patients who die of prescription drug abuse? Am J Emerg Med 34(1):30–35 [DOI] [PubMed] [Google Scholar]
- 18.Lazar DJ, Zaveri S, Khetan P, Nobel T, Divino CM (2021) Variations in postoperative opioid prescribing by day of week and duration of hospital stay. Surgery 169(4):929–933 [DOI] [PubMed] [Google Scholar]
- 19.Hill MV, McMahon ML, Stucke RS, Barth RJ (2017) Wide variation and excessive dosage of opioid prescriptions for common general surgical procedures. Ann Surg 265(4):709–714 [DOI] [PubMed] [Google Scholar]
- 20.McDonald DC, Carlson K, Izrael D (2012) Geographic variation in opioid prescribing in the U. S. J Pain 1310:988–996 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Sun EC, Darnall BD, Baker LC, Mackey S (2016) Incidence of and risk factors for chronic opioid use among opioid-naïve patients in the postoperative period. JAMA Intern Med 176(9):1286–1293 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Adogwa O, Davidson MA, Vuong VD, Desai SA et al. (2019) Regional variation in opioid use after lumbar spine surgery. World Neurosurg 121:e691–e699 [DOI] [PubMed] [Google Scholar]
- 23.Hah J, Bateman BT, Ratliff J, Curtin C, Sun E (2017) Chronic opioid use after surgery: implications for perioperative management in the face of the opioid epidemic. Anesth Analg 125(5):1733–1740 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Ardeljan LD, Waldfogel JM, Bicket MC, Hunsberger JB et al. (2020) Current state of opioid stewardship. Am J Health Syst Pharm 77(8):636–643 [DOI] [PubMed] [Google Scholar]
- 25.Jivraj NK, Raghavji F, Bethell J, Wijeysundera DM et al. (2020) Persistent postoperative opioid use: a systematic literature search of definitions and population-based cohort study. Anesthesiology 132(6):1528–1539 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Kirson NY, Shei A, Rice JB, Enloe CJ et al. (2015) The burden of undiagnosed opioid abuse among commercially insured individuals. Pain Med 16(7):1325–1332 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.