Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2022 Nov 1.
Published in final edited form as: Dig Dis Sci. 2021 Jan 1;66(11):3938–3950. doi: 10.1007/s10620-020-06742-0

High-dose opioid use among Veterans with unexplained gastrointestinal symptoms vs structural gastrointestinal diagnoses

Salva N Balbale 1,2,3, Lishan Cao 3, Itishree Trivedi 4, Jonah J Stulberg 1,5, Katie J Suda 6, Walid F Gellad 6, Charlesnika T Evans 1,3,7, Bruce L Lambert 8, Neil Jordan 1,3,9, Laurie A Keefer 10
PMCID: PMC8245587  NIHMSID: NIHMS1659472  PMID: 33385263

Structured Abstract

Background.

In a cohort of Veterans dually enrolled in the Department of Veterans Affairs (VA) and Medicare Part D, we sought to describe high-dose daily opioid use among Veterans with unexplained gastrointestinal (GI) symptoms & structural GI diagnoses and examine factors associated with high-dose use.

Methods.

We used linked national patient-level data from the VA and Centers for Medicare & Medicaid Services (CMS). We grouped patients into 3 subsets: those with unexplained GI symptoms (e.g. chronic abdominal pain); structural GI diagnoses (e.g. chronic pancreatitis); and those with a concurrent unexplained GI symptom & structural GI diagnosis. High-dose daily opioid use levels were examined as a binary variable (≥100 morphine milligram equivalents (MME)/day) and as an ordinal variable (50-99 MME/day, 100-119 MME/day, or ≥120 MME/day).

Results.

We identified 141,805 chronic GI patients dually enrolled in VA & Part D. High-dose opioid use was present in 11% of Veterans with unexplained GI symptoms, 10% of Veterans with structural GI diagnoses, and 15% of Veterans in the concurrent GI group. Compared to Veterans with only an unexplained GI symptom or structural diagnosis, concurrent GI patients were more likely to have higher daily opioid doses, more opioid days ≥100 MME, and higher risk of chronic use. Factors associated with high-dose use included opioid receipt from both VA & Part D, younger age, and benzodiazepine use.

Conclusions.

A significant subset of chronic GI patients in the VA are high-dose opioid users. Efforts are needed to reduce high-dose use among Veterans with concurrent GI symptoms and diagnoses.

Keywords: Gastroenterology, Health Services, Opioid Analgesics, Pain Management

Introduction

Ending the opioid epidemic in the United States is a national priority [1,2]. Driven by numerous factors [3–5], prescription opioid overdoses led to nearly 400,000 deaths nationally between 1999 to 2014 [6,7]. Although much attention has—understandably—been focused on the harms associated with long-term opioid use [8–10], a growing body of evidence suggests that the risks of opioid overdose, in particular unintentional overdose, and addiction may also increase with opioid dose [11–13]. Monitoring and decreasing high-dose daily opioid use are among the cornerstones of the Opioid Safety Initiative (OSI) implemented in 2013 in the Department of Veterans Affairs (VA) health care system [14]. These efforts are particularly important because Veterans remain disproportionately affected by the epidemic [15,16] and because Veterans who use high daily doses of opioids may have a higher likelihood of heroin use [17] and overdose-related death [18].

Some groups within the broader Veteran population are at especially high risk for the harmful effects of high-dose daily opioid use [19,20] and may benefit from risk stratification and tailored safety efforts [21]. Veterans with chronic gastrointestinal (GI) symptoms and disorders may represent one such group, but data are limited in this area. Although many GI symptoms and disorders are painful, opioid use may be ineffective in managing gastrointestinal pain [22], and prolonged use can potentially lead to bowel dysfunction, opioid escalation, and other serious adverse events [23–25]. In a recent study, Sayuk et al found that opioids were frequently prescribed to chronic GI patients receiving VA care. Moreover, Veterans with “unexplained GI symptoms”, such as chronic abdominal pain, irritable bowel syndrome, and other functional GI disorders, were considerably more likely to use opioids than Veterans with “structural GI diagnoses” such as inflammatory bowel disease (IBD) and gastroesophageal reflux [26]. This finding is alarming given that there is no clinical indication for opioids to address chronic abdominal pain and functional GI disorders [22,26–28]. And although the need to monitor and curb opioid use across GI care has been established [22,29], these findings raise new questions about the nuances of opioid use among subsets of GI patients.

The frequency of high-dose daily opioid use among individuals with unexplained GI symptoms and structural GI diagnoses is unclear, as are the factors that may be associated with high-dose use. In the broader GI literature, few prior studies have explored opioid dosage. Although limited to individuals with IBD and other structural diagnoses, these studies have shown that high-dose daily use may be associated with depression and substance abuse [30], as well as increased risk of hospitalization and death [31,32]. In the larger Veteran population, increased medication access through dual enrollment in the VA and Medicare Part D has been linked with high-dose and unsafe opioid use [33], but how dual enrollment may affect high-dose daily opioid use among the growing number of chronic GI patients in the VA is unknown.

To our knowledge, no published studies have focused on high-dose daily opioid use among individuals with unexplained GI symptoms and structural GI diagnoses. Characterizing this information will be critical to understand potential need for safer dosing practices that are tailored for chronic GI patients. Using a national cohort of Veterans dually enrolled in the VA and Medicare Part D, we aimed to (1) describe high-dose daily opioid use among Veterans with unexplained GI symptoms and structural GI diagnoses, and (2) identify factors associated with high-dose daily opioid use.

Materials and Methods

Study Design and Setting

For this retrospective cohort study, we linked national patient-level data (April 1, 2011 through October 31, 2014) from the VA and Centers for Medicare and Medicaid Services (CMS) to identify GI patients and observe high-dose daily opioid use. The VA data files were extracted from the VA Corporate Data Warehouse and included patient demographics, diagnosis codes for outpatient visits, and health care utilization and prescription records. CMS data included the Medicare Provider Analysis and Review file, the Beneficiary Summary File for patient enrollment and demographic information, outpatient and carrier files, and event files listing prescription drugs dispensed through Medicare Part D. We used Donabedian’s structure-process-outcome model [34,35] as a conceptual framework for the study. To explore high-dose opioid use as an outcome, we investigated associations between high-dose daily opioid use and (1) structural elements of the health care system such as patient and facility characteristics, and (2) health care process-related factors such as care utilization, treatments, and concurrent medication use.

This study was approved by the institutional review boards at VA Hines Hospital, VA Pittsburgh Healthcare System, and Northwestern University.

Study Population

The study population included Veterans with chronic GI symptoms and disorders who were enrolled in both the VA and Medicare Part D and filled at least one opioid prescription from either source in the six-month period following GI diagnosis. We included Medicare fee-for-service and managed care enrollees in our analysis. We focused on common chronic GI symptoms and disorders (chronic abdominal pain, IBD, pancreatic diseases, and functional GI disorders) that have comprised similar cohorts of chronic GI patients in previous research [26,36] and have been associated with higher rates of opioid use [29]. Using International Classification of Diseases–Ninth Revision–Clinical Modification (ICD-9-CM) diagnosis codes, we classified Veterans as chronic GI patients if they had two encounters for the same GI condition within a 12-month period; as in prior studies [26,36,37], at least one of these encounters had to be in an outpatient setting. GI conditions included chronic abdominal pain (789.0), chronic pancreatitis (577.1), functional GI disorders such as irritable bowel syndrome (536.0-536.9, 564.0-564.9), and IBD including Crohn’s disease (555.0-555.9) and ulcerative colitis (556.0-556.9). Veterans who met GI diagnosis criteria between April 1, 2011 and March 31, 2013 and received an opioid prescription within the six months following the second diagnosis code were then followed for a 12-month period or until death. Similar to Sayuk et al’s approach [26], we grouped GI patients into three groups: those with unexplained GI symptoms (including chronic abdominal pain, irritable bowel syndrome, functional dyspepsia, and other functional GI disorders), those with structural GI diagnoses (including chronic pancreatitis, Crohn’s disease and ulcerative colitis), and those with both unexplained GI symptoms and structural GI diagnoses. As in other studies, we excluded Veterans (1) if their only opioid prescription was in an injectable or suppository formulation that did not have a reliable morphine milligram equivalents (MMEs) conversion factor, or (2) for addiction treatment (i.e. methadone, or buprenorphine in combination with naloxone or naltrexone) [26,33,38]. We also excluded Veterans in hospice care.

Variable Definitions

Our primary outcomes of interest were outpatient opioid prescription dispensing, defined as receipt of an opioid prescription in an outpatient care setting, and high-dose daily opioid use, defined as daily mean of ≥100 MMEs during the 12-month observation period [16,18,20,39]. For our adjusted analyses, we further categorized high-dose daily use as a daily mean of 50-99 MMEs / day, 100-119 MMEs / day, and 120 or higher MMEs / day (vs. 0-49 MMEs / day) [18,19,33]. To examine these, we first combined VA and Part D data to determine the daily prescribed opioid dosage for each Veteran for all nonliquid opioid medications received during the observation period. We then calculated the oral MMEs for each opioid prescription using standard equianalgesic conversion factors [40]. Following a similar approach used by Gellad et al [33], we used this information to describe (1) opioid use measures, including total number of days receiving an opioid, types of opioid medications used, and mean daily MMEs for each Veteran, and (2) opioid safety measures, including the proportion of patients with any day ≥100 MMEs and the proportion of patients with a daily dosage greater than 120 MMEs for at least 90 consecutive days (defined as long-term, high-dose use). As in prior studies, we determined daily dosage assuming each Veteran took all prescribed opioids according to the dosage and schedule delineated in the prescription [33,38]. Opioids were defined as medications containing codeine, fentanyl, hydrocodone, hydromorphone, oxycodone, meperidine, morphine, tapentadol, or tramadol. As in other analyses, individuals with average daily doses greater than 1,000 MMEs were excluded from the analysis to address potential data entry errors or extreme cases [41].

We used previous literature and our conceptual framework to guide selection of independent variables. These variables included patient demographics such as age, gender, race / ethnicity (defined as white, black, Hispanic, or non-Hispanic / other), and region where the index GI diagnosis was received. We captured enrollment in Medicaid and low-income subsidies (LIS); these are programs developed by CMS to assist Medicare beneficiaries with limited resources in paying for prescription drugs and to reduce the costs of Medicare prescription drug coverage [42]. Enrollment in these subsidies were defined as (1) being enrolled in Medicaid and qualifying for fully subsidized Part D coverage, (2) qualifying for partially subsidized Part D coverage and receiving a LIS, or (3) being ineligible for subsidized Part D. We also assessed whether the original reason for Medicare enrollment was related to disability. We included type of opioid used and the source of opioid prescription (VA, Part D only, or both VA and Part D). We identified other medications used that have been included in previous analyses of ORADEs [20] (including sleeping aids, muscle relaxants, benzodiazepines, and antidepressants). We used current procedural terminology (CPT) codes as in previous research to assess whether the Veteran had any major surgery during the observation period [43]. Non-GI pain-related diagnoses outlined in other studies were also included using ICD-9 codes [26,44]. Finally, we included mental health disorders and substance use disorders defined using two outpatient or one inpatient ICD-9 codes for anxiety, non-opioid drug dependence, alcohol dependence, mood disorders, post-traumatic stress disorders, and personality disorders.

Statistical Analysis

Using frequencies, percentages, and means, we described patient characteristics as well as opioid use and safety measures by type of GI condition. We used the χ2 test for categorical variables and t-test for continuous variables to compare opioid use and safety measures among Veterans with unexplained GI symptoms, structural GI diagnoses, or both. To evaluate potential differences among Veterans with and without high-dose daily opioid use, we created a dichotomous variable indicating whether or not the individual had any high-dose use during the observation period. In our bivariate analyses, we used the χ2 test for categorical variables and analysis of variance test for continuous variables to compare patient characteristics by status of high-dose opioid use. A 2-sided p-value ≤.05 was considered significant. We used multinomial logistic regression modelling to generate adjusted relative risk ratios (RRR) and 95% confidence intervals to determine variables independently associated with individual categories of high-dose daily opioid use. The final model included variables significant in bivariate analyses that remained significant in the model with p-value ≤.05. We conducted analyses using SAS statistical software version 9.4 (SAS Institute Inc, Cary, NC) and Stata version 14 (StataCorp, College Station, TX).

Results

We identified 141,805 chronic GI patients who were (1) dually enrolled in VA and Part D, and (2) received at least one opioid prescription from the VA, Part D, or both between 2011 and 2014. The average age was 71 years (SD =12.6). Overall, approximately 85% of Veterans in this cohort were male and 82% were white. Patient characteristics by type of GI symptoms and diagnoses are shown in Table 1. Veterans predominantly had unexplained GI symptoms only (92%; n=130,749); a considerably smaller proportion of the overall cohort had a structural GI diagnosis only (1%; n=1,558), or an unexplained symptom and a concurrent structural GI diagnosis (herein referred to as the “concurrent GI group;” 7%; n=9,498). Of the Veterans with unexplained GI symptoms only, there were a total of 51,174 individuals with chronic abdominal pain only, 15,644 individuals with a functional GI disorder only, and 63,931 with an overlapping diagnosis of chronic abdominal pain and a functional GI disorder. Of the Veterans with structural GI diagnoses only, 215 Veterans had chronic pancreatitis only, 546 had Crohn’s disease only, 639 had ulcerative colitis only, and 158 had overlapping diagnoses of chronic pancreatitis, Crohn’s, and / or ulcerative colitis. Approximately 11% of Veterans with unexplained GI symptoms, 10% of Veterans with structural GI diagnoses, and 15% of Veterans with concurrent GI symptoms and diagnoses had a mean daily opioid dose ≥100 MMEs.

Table 1.

Patient characteristics by type of GI symptoms and diagnoses.

Type of GI Symptoms and Diagnoses All
Unexplained GI Symptoms Only Structural GI Diagnosis Only Concurrent Unexplained GI Symptoms & Structural GI Diagnosis
# of Patients % # of Patients % # of Patients % # of Patients %
All 130,749 92.2 1,558 1.1 9,498 6.7 141,805 100
Source of opioid prescription
  VA only 33,831 25.9 477 30.6 2,456 25.8 36,764 25.9
  Medicare Part D only 90,369 69.1 1,009 64.8 6,432 67.7 97,810 68.9
  Both VA & Medicare Part D 6,628 5.1 72 4.6 619 6.5 7,319 5.2
Age category
  18-39 2,109 1.6 24 1.5 240 2.5 2,373 1.7
  40-64 32,607 24.9 418 26.8 3,414 36.0 36,439 25.7
  65-84 76,958 58.8 999 64.1 5,060 53.2 83,017 58.5
  85+ 19,154 14.6 117 7.5 793 8.3 20,064 14.1
Gender
  Male 110,554 84.5 1,429 91.7 7,994 84.1 119,977 84.6
  Female 20,264 15.5 128 8.2 1,513 16.0 21,905 15.4
Race / ethnicity
  White 107,025 81.8 1,377 88.4 7,685 80.8 116,087 81.8
  Black 16,285 12.5 130 8.3 1,320 13.9 17,735 12.5
  Hispanic 4,565 3.5 19 1.2 310 3.3 4,894 3.5
  Non-Hispanic / Other 2,759 2.1 27 1.7 184 2.0 2,970 2.1
Region
  South 58,896 45.1 617 39.6 4,088 43.0 63,601 44.9
  Midwest 31,831 24.4 397 25.5 2,137 22.5 34,365 24.2
  West 22,453 17.2 287 18.4 1,690 17.8 24,430 17.2
  Northeast 17,258 13.2 256 16.4 1,568 16.5 19,082 13.5
Disability 54,128 41.4 596 38.3 4,917 51.7 59,641 42.0
Low-income subsidy
  No Medicaid or LIS 86,303 66.0 1,229 78.9 6,067 63.8 93,599 66.0
  Medicaid 39,817 30.4 281 18.0 3,070 32.3 43,168 30.4
  LIS only 4,700 3.6 48 3.1 370 3.9 5,118 3.6
Medication use based on active ingredient 0
  Antidepressants 58,763 44.9 589 37.8 4,629 48.7 63,981 45.1
  Benzodiazepines 38,670 29.6 405 26.0 3,336 35.1 42,411 29.9
  Muscle relaxants 27,008 20.6 243 15.6 2,061 21.7 29,312 20.7
  Sleeping aids 18,462 14.1 190 12.2 1,447 15.2 20,099 14.2
  Antipsychotics 16,443 12.6 107 6.9 1,295 13.6 17,845 12.6
Major surgery during the observation period 9,170 7.0 125 8.0 660 6.9 9,955 7.0
Mental health disorders
  PTSD 6,698 5.1 88 5.7 609 6.4 7,395 5.2
  Non-opioid drug dependence 5,896 4.5 63 4.0 665 7.0 6,624 4.7
  Mood disorder 5,812 4.4 73 4.7 544 5.7 6,429 4.5
  Anxiety 5,640 4.3 81 5.2 529 5.6 6,250 4.4
Number of nonopioid medications
  Mean (SD) 15.1 (9) 13.9 (8.0) 16.3 (9.4) 15.1 (9.0)
Open in a new tab
*

Disability refers to the original reason for Medicare enrollment.

Note. Missing data: gender (n = 11), race / ethnicity (n = 207), region (n = 8), Medicaid and LIS (n = 8), alcohol dependence medications (n = 369), and personality disorder (n = 944).

Table 2 includes opioid use and safety measures by type of GI symptoms and diagnoses. Veterans in the concurrent GI group were significantly more likely to have a mean daily dose ≥100 MMEs compared to Veterans with only an unexplained symptom (15.4% vs. 10.9%; p<0.001) or structural diagnosis (15.4% vs. 9.8%; p<0.001). The average number of nonconsecutive days on opioids was also highest in the concurrent GI group (mean = 166.3; SD = 134.1), followed by the unexplained group (mean = 151.7; SD = 131.7), and finally the structural group (mean = 150.6; SD = 135.7). Veterans in the concurrent GI group were more likely to have ≥90 nonconsecutive days on opioids (59%) compared to those with only an unexplained GI symptom (55%) or structural GI diagnosis (54%); this is depicted in Figure 1. Mean daily MMEs was also highest in the concurrent GI group (mean = 64.8; SD = 91.5), followed by those with only an unexplained GI symptom (mean = 53.5; SD = 75.0) and those with only a structural diagnosis (mean = 50.5; SD = 69.9). Veterans in the concurrent group were also more likely to have over 90 nonconsecutive days on opioids with daily MMEs ≥100 compared to Veterans with only an unexplained symptom (13.4 vs. 9.1; p<0.001) or structural diagnosis (13.4 vs. 8.7; p<0.001). Similarly, Veterans in the concurrent group were also more likely to have over 90 consecutive days on opioids compared to Veterans with only an unexplained symptom (59.1 vs. 54.7; p<0.001) or structural diagnosis (59.1 vs. 53.5; p<0.001). Long-term, high dose use (≥90 consecutive days with ≥120 MME / day) did not significantly differ between Veterans in the concurrent GI group and those with only an unexplained symptom or structural diagnosis.

Table 2.

Opioid use and safety measures by type of GI symptoms and diagnoses.

Unexplained GI Symptoms Only Structural GI Diagnosis Only Concurrent Structural GI Diagnosis & Unexplained GI Symptoms All P-value† P-value§ P-value‡
(n=130,749) (n=1,558) (n=9,498) (n=141,805)
Mean or n SD or % Mean or n SD or % Mean or n SD or % Mean or n SD or %
OPIOID USE MEASURES
Days on Opioids (nonconsecutive)
  Mean (SD) 151.7 131.7 150.6 135.7 166.3 134.1 152.7 132.2 0.12 <0.001 <0.001
Daily MME
  Mean (SD) 53.5 75.0 50.5 69.9 64.8 91.5 54.3 76.2 <0.001 <0.001 <0.001
Type of opioid used
  Hydrocodone 80,757 61.7 955 61.3 5,673 59.7 87,385 61.6 0.71 <0.001 0.24
  Oxycodone 42,371 32.4 459 29.5 3,720 39.1 46,550 32.8 0.01 <0.001 <0.001
  Tramadol 39,612 30.3 423 27.2 2,734 28.8 42,769 30.1 0.007 0.002 0.19
  Morphine 11,264 8.6 112 7.2 1,075 11.3 12,451 8.8 0.05 <0.001 <0.001
  Fentanyl 9,490 7.3 61 3.9 814 8.6 10,365 7.3 <0.001 <0.001 <0.001
  Hydromorphone 5,303 4.1 46 3.0 702 7.4 6,051 4.3 0.03 <0.001 <0.001
  Meperidine 3,597 2.8 42 2.7 382 4.0 4,021 2.8 0.89 <0.001 0.01
  Codeine 3,118 2.4 67 4.3 261 2.8 3,446 2.4 <0.001 0.03 <0.001
OPIOID SAFETY MEASURES
Mean daily MME >100 14,256 10.9 152 9.8 1,461 15.4 15,869 11.2 0.15 <0.001 <0.001
Days with high-dose daily use >100 MME*
  0 days 98,545 75.3 1,230 79.0 6,485 68.2 106,260 74.9 0.001 <0.001 <0.001
  1-6 days 6,001 4.6 62 4.0 521 5.5 6,584 4.6 0.25 <0.001 0.01
  7-29 days 7,819 6.0 73 4.7 615 6.5 8,507 6.0 0.03 0.05 0.007
  30-89 6,538 5.0 57 3.7 609 6.4 7,204 5.1 0.02 <0.001 <0.001
  ≥90 days 11,925 9.1 136 8.7 1,277 13.4 13,338 9.4 0.62 <0.001 <0.001
≥90 consecutive days 71,528 54.7 834 53.5 5,621 59.1 77,983 55.0 0.37 <0.001 <0.001
≥90 consecutive days > 120 MME 4,226 3.2 53 3.4 309 3.3 4,588 3.2 0.70 0.90 0.76
Open in a new tab
*

Categories of days with high-dose daily use >100MME indicate nonconsecutive days

†

Indicates p-value for chi-square test or t-test comparing opioid use / safety measures between Veterans with unexplained GI symptoms only and structural GI diagnoses only.

§

Indicates p-value for chi-square test or t-test comparing opioid use / safety measures between Veterans unexplained GI symptoms only and Veterans with concurrent GI symptoms / diagnoses.

‡

Indicates p-value for chi-square test or t-test comparing opioid use / safety measures between Veterans structural GI diagnoses only and Veterans with concurrent GI symptoms / diagnoses.

Note. Missing data: buprenorphine (n = 586) and tapentadol (n = 240).

Figure 1.

Figure 1.

Open in a new tab

Number of nonconsecutive opioid days with high-dose daily use by type of GI symptoms and diagnoses.

Results of the bivariate analyses (presented in Table 3) compare patient characteristics between Veterans in the cohort who had high-dose daily opioid use (n=15,869) and those who did not (n=125,936). Compared to Veterans who did not have high-dose use, high-dose users differed significantly in terms of source of opioid prescription, age, race / ethnicity, region, enrollment in a LIS, type of GI symptom or diagnosis. A greater proportion of Veterans with disability as the reason for Medicare enrollment used high-dose opioids compared to those without high-dose use (65% vs. 39%; p<0.001). With respect to concurrent medications, we noted a significant association between high-dose use and use of antidepressants, benzodiazepines, muscle relaxants, antipsychotics, and alcohol dependence drugs. Undergoing major surgery and having a mental health disorder were also significantly associated with high-dose use. Finally, use of opioids for over 90 consecutive days was also significantly associated with high-dose use.

Table 3.

Bivariate analysis of patient characteristics by status of high-dose daily opioid use and unadjusted odds ratios.

Bivariate Analysis Unadjusted Odds Ratio
Status of high-dose daily opioid use P-value
(χ2)		 (95% CI)
Did not have high-dose daily opioid use (n=125,936) Had high-dose daily opioid use (n=15,869)
# of Patients % # of Patients %
Source of opioid prescription
  Medicare Part D only 87,033 69.1 10,746 67.7 Ref Ref
  VA only 33,248 26.4 3,459 21.8 <0.001 0.84 (0.81, 0.88) †
  Both VA & Medicare Part D 5,655 4.5 1,664 10.5 <0.001 2.38 (2.25, 2.53)†
Age category <0.001
  18-39 1,968 1.6 395 2.5 Ref Ref
  40-64 28,982 23.0 7,414 46.7 <0.001 1.27 (1.14, 1.42)
  65-84 75,851 60.2 7,138 44.9 <0.001 0.47 (0.42, 0.52) †
  85+ 19,135 15.2 922 5.8 <0.001 0.24 (0.21, 0.27) †
Gender 0.16
  Male 106,582 84.6 13,395 84.2 0.96 (0.92, 1.01)
  Female 19,383 15.4 2,522 15.8 Ref
Race / ethnicity
  White 102,944 81.7 13,070 82.4 Ref Ref
  Black 15,736 12.5 1,992 12.6 0.91 0.99 (0.95, 1.10)
  Hispanic 4,435 3.5 454 2.9 <0.001 0.81 (0.73, 0.89) †
  Non-Hispanic / other 2,636 2.1 331 2.1 0.85 0.99 (0.88, 1.11)
  Missing 185 0.2 22 0.1
Region <0.001
  Midwest 30,612 24.3 3,753 23.7 Ref Ref
  Northeast 16,661 13.2 2,421 15.3 <0.001 1.19 (1.12, 1.25) †
  PR/VI/GU 302 0.2 17 0.1 0.001 0.46 (0.28, 0.75)
  South 57,401 45.6 6,200 39.1 <0.001 0.88 (0.84, 0.92)
  West 20,953 16.6 3,477 21.9 <0.001 1.35 (1.29, 1.42)†
Disability 49,226 39.1 10,415 65.4 <0.001 2.95 (2.84, 3.06) †
  Missing 8,323 6.6 605 3.8
Low-income subsidy
  No Medicaid or LIS 84,951 67.5 8,589 54.2 Ref Ref
  Medicaid 36,694 29.1 6,449 40.6 <0.001 1.74 (1.68, 1.80) †
  LIS only 4,283 3.4 831 5.3 <0.001 1.91 (1.78, 2.07) †
Type of GI symptoms and diagnoses
  Unexplained GI symptoms only 116,493 92.5 14,256 89.8 Ref Ref
  Structural GI diagnosis only 1,406 1.1 152 1.0 0.15 0.88 (0.75, 1.05) †
  Both 8,037 6.4 1,461 9.2 <0.001 1.49 (1.40, 1.57) †
Medication use based on active ingredient
  Antidepressants 54,970 43.6 9,011 56.6 <0.001 1.69 (1.63, 1.74) †
  Benzodiazepines 35,545 28.2 6,866 43.1 <0.001 1.93 (1.87, 2.00) †
  Muscle relaxants 24,305 19.3 5,007 31.5 <0.001 1.92 (1.85, 1.99) †
  Sleeping aids 17,133 13.6 2,966 18.6 <0.001 1.46 (1.40, 1.52) †
  Antipsychotics 15,484 12.3 2,361 14.8 <0.001 1.24 (1.18, 1.30) †
  Alcohol dependence medications 271 0.2 98 0.6 <0.001 2.81 (2.23, 3.56) †
Major surgery 8,755 7.0 1,200 7.5 0.006 1.08 (1.02, 1.15) †
Mental health disorders
  PTSD 6,329 5.0 1,066 6.7 <0.001 1.34 (1.25, 1.43) †
  Non-opioid drug dependence 5,384 4.3 1,240 7.8 <0.001 1.86 (1.74, 1.98) †
  Mood disorder 5,357 4.3 1,072 6.7 <0.001 1.60 (1.50, 1.72) †
  Anxiety 5,214 4.1 1,036 6.5 <0.001 1.58 (1.48, 1.70) †
  Personality disorder 759 0.6 185 1.1 <0.001 1.92 (1.63, 2.25) †
Type of opioid used
  Hydrocodone 81,292 64.5 6,093 38.3 <0.001 0.34 (0.33, 0.35) †
  Oxycodone 36,857 29.3 9,693 60.9 <0.001 3.79 (3.67, 3.92) †
  Tramadol 41,004 32.6 1,765 11.1 <0.001 0.25 (0.25, 0.27) †
  Morphine 7,165 5.7 5,286 33.2 <0.001 8.28 (7.94, 8.62) †
  Fentanyl 5,008 4.0 5,357 33.7 <0.001 12.3 (11.78, 12.85) †
  Hydromorphone 3,801 3.0 2,250 14.1 <0.001 5.31 (5.02, 5.61)†
  Meperidine 1,759 1.4 2,262 14.2 <0.001 11.7 (11.0, 12.52) †
  Codeine 3,320 2.6 126 0.8 <0.001 0.30 (0.25, 0.35) †
  Tapentadol 189 0.2 51 0.3 <0.001 2.15 (1.57, 2.92) †
Opioid use for ≥90 consecutive days 64,918 51.5 13,065 82.1 <0.001 4.30 (4.12, 4.49) †
Open in a new tab
*

Disability refers to the original reason for Medicare enrollment.

†

Significant at the 0.05 level.

Note. Missing data: gender (n = 11), region (n = 8), and Medicaid and LIS (n = 8).

Table 4 presents the results of the multinomial logistic regression analysis and the adjusted RRR of high-dose daily opioid use by high-dose category (50-99 MMEs / day, 100-119 MMEs / day, or ≥120 MMEs / day) compared to lower daily doses between 0-49 MMEs / day. In these adjusted analyses, Veterans who received opioid prescriptions from both the VA and Medicare Part D were significantly more likely to have any level of high-dose daily opioid use compared to those who received opioids from Part D alone. For these Veterans who received opioids from both sources, the RRR was highest for high-dose daily use ≥120 MMEs / day (ARR=2.06; 95% CI: 1.87, 2.27). Veterans who were prescribed benzodiazepines were also significantly more likely to have any level of high-dose daily opioid use than those who did not use benzodiazepines (compared to 0-49 MMEs / day). Among benzodiazepine users, the RRR was highest for high-dose daily use ≥120 MMEs / day (ARR=1.29; 95% CI: 1.22, 1.37). Muscle relaxant use and sleeping aid use were associated with a significantly higher RRR of high-dose daily opioid use between 50 and 99 MMEs / day and ≥120 MMEs / day, but were not significantly associated with high-dose use between 100 and 119 MMEs / day. Among mental health disorders, Veterans with a diagnosis of non-opioid drug dependence were significantly more likely to have any level of high-dose daily opioid use (than those without non-opioid drug dependence); in cases of non-opioid drug dependence, the RRR was highest for high-dose daily use ≥120 MMEs / day (ARR=1.22; 95% CI: 1.09, 1.35). Among types of opioids used, Veterans were substantially more likely to have high-dose daily opioid use if they used oxycodone, morphine, fentanyl, hydromorphone, meperidine, or tapentadol. Among users of these specific opioid medications (with the exception for oxycodone and tapentadol), the RRR was highest for high-dose daily use ≥120 MMEs / day; the greatest RRR among these were associated with fentanyl use (ARR=67.82; 95% CI: 62.16, 73.99), meperidine use (ARR=26.45; 95% CI: 23.49, 29.77), and morphine use (ARR=16.93; 95% CI: 15.73, 18.23). Veterans with chronic opioid use (90 or more consecutive days on opioids) were significantly more likely to have high-dose daily use; among Veterans with chronic use, the RRR was highest for high-dose daily use ≥120 MMEs / day (ARR=2.52; 95% CI: 1.09, 1.35). Factors that were significantly associated with a lower RRR of any level of high-dose daily use included receipt of opioids from the VA only, older age (85 years or higher), black race, antipsychotic medication use, undergoing major surgery during the observation period, as well as hydrocodone, tramadol, and codeine use.

Table 4.

Multinomial logistic regression analysis assessing the association between patient characteristics and individual categories of high-dose daily opioid use (N = 141,805).

0 – 49 vs. 50 – 99 MME / day 0 – 49 vs. 100 – 119 MME / day 0 – 49 vs. ≥120 MME / day
Variable Relative Risk Ratio
(95% CI)		 Relative Risk Ratio
(95% CI)		 Relative Risk Ratio
(95% CI)
Source of opioid prescription
  Medicare Part D only Ref Ref Ref
  VA only 0.45 (0.42, 0.47) † 0.44 (0.38, 0.50) † 0.52 (0.49, 0.57) †
  Both VA & Medicare Part D 1.56 (145, 1.68) † 2.04 (1.76, 2.37) † 2.06 (1.87, 2.27) †
Age category
  18-39 Ref Ref Ref
  40-64 1.08 (0.90, 1.30) 1.22 (0.82, 1.82) 1.23 (0.98, 1.55)
  65-84 0.90 (0.75, 1.08) 0.85 (0.56, 1.27) 0.72 (0.57, 0.91) †
  85+ 0.52 (0.44, 0.64) † 0.41 (0.26, 0.64) † 0.29 (0.22, 0.38) †
Race / ethnicity
  White Ref Ref Ref
  Black 0.84 (0.79, 0.89) † 0.85 (0.73, 0.99) † 0.91 (0.84, 0.99) †
  Hispanic 0.86 (0.76, 0.97) † 0.91 (0.68, 1.21) 0.74 (0.62, 0.89) †
  Non-Hispanic / other 0.98 (0.85, 1.11) 0.90 (0.65, 1.25) 0.94 (0.77, 1.14)
Medication use based on active ingredient
  Benzodiazepines 1.08 (1.03, 1.12) † 1.24 (1.12, 1.37) † 1.29 (1.22, 1.37) †
  Muscle relaxant 1.13 (1.07, 1.18) † 1.07 (0.95, 1.20) 1.21 (1.13, 1.30) †
  Sleeping aids 1.06 (1.00, 1.11) † 1.07 (0.94, 1.21) 1.08 (1.01, 1.17) †
  Antipsychotics 0.89 (0.83, 0.94) † 0.80 (0.69, 0.91) † 0.70 (0.65, 0.76) †
Major surgery 0.84 (0.78, 0.90) † 0.80 (0.68, 0.95) † 0.68 (0.62, 0.76) †
Mental health disorders
  Non-opioid drug dependence 1.15 (1.06, 1.25) † 1.21 (1.01, 1.44) † 1.22 (1.09, 1.35) †
Type of opioid used
  Hydrocodone 0.74 (0.70, 0.77) † 0.57 (0.51, 0.63) † 0.31 (0.29, 0.33) †
  Oxycodone 3.06 (2.94, 3.20) † 4.06 (3.68, 4.49) † 3.68 (3.46, 4.89) †
  Tramadol 0.32 (0.31, 0.34) † 0.26 (0.23, 0.30) † 0.18 (0.16, 0.19) †
  Morphine 7.23 (6.80, 7.70) † 17.27 (15.47, 19.27) † 16.93 (15.73, 18.23) †
  Fentanyl 12.03 (11.15, 12.98) † 33.28 (29.33, 37.76) † 67.82 (62.16, 73.99) †
  Hydromorphone 2.67 (2.44, 2.91) † 3.57 (3.04, 4.19) † 4.41 (3.97, 4.89) †
  Meperidine 7.14 (6.36, 8.01) † 17.97 (15.06, 21.45) † 26.45 (23.49, 29.77) †
  Codeine 0.36 (0.30, 0.42) † 0.35 (0.23, 0.52) † 0.22 (0.17, 0.28) †
  Tapentadol 17.91 (13.15, 24.39) † 24.26 (12.84, 45.82) † 17.35 (11.25, 26.77) †
Opioid use for ≥90 consecutive days 1.10 (1.06, 1.15) † 1.47 (1.31, 1.65) † 2.52 (2.35, 2.71) †
Open in a new tab
†

Significant at the 0.05 level.

Note: Models were adjusted for low-income subsidy, disability, region, and non-GI pain diagnoses (i.e. back pain, arthritis, and headache).

Discussion

We examined high-dose opioid use among dually enrolled Veterans with unexplained GI symptoms and structural GI symptoms in this study. This is a vulnerable population at high risk for overdose, and we must better understand the predictors to address our current crisis. In this cohort, approximately 11% of Veterans with unexplained GI symptoms, 10% of Veterans with structural GI diagnoses, and 15% of Veterans with concurrent GI symptoms and diagnoses had a mean daily opioid dose ≥100 MMEs. Compared to Veterans who had solely an unexplained GI symptom or a structural GI diagnosis, Veterans with both an unexplained GI symptom and a structural GI diagnosis had a significantly greater likelihood of having higher daily opioid doses, higher number of nonconsecutive days on opioids ≥100 MMEs, and higher risk of long-term opioid use (≥90 consecutive days). We identified several factors that were independently associated with all levels of high-dose daily opioid prescription use (>50 MMEs / day). These factors included receipt of opioid prescriptions from both the VA and Medicare Part D, younger age (18-39 years old), benzodiazepine or muscle relaxant medication use, and having a diagnosis for non-opioid drug dependence. Except for hydrocodone, codeine, and tramadol, use of all other opioids—and most notably, fentanyl, meperidine, and morphine—were also independently associated with high-dose daily use. Taken together, these findings suggest that a substantial proportion of chronic GI patients in the VA may be high-dose opioid users and may be at increased risk for unsafe opioid use outcomes.

High-dose opioid use is an important measure of opioid safety [8,14,20]. Higher doses of opioids are associated with greater likelihood of long-term use and unintentional opioid overdose-related death [38,45], and even shorter courses at high doses may contribute to the development of addiction [46]. Although efforts to reduce high doses of opioids should be an integral part of comprehensive strategies to decrease opioid use [8,18], there remain questions about how to optimize safety while treating pain. For individuals with unexplained GI symptoms and structural GI diagnoses, these questions are especially relevant given the need to address abdominal pain while mitigating the risk of opioid escalation, bowel dysfunction, and other serious harms [22,25,47], but there is a lack of clinical evidence on dosing safety and high-dose use in this patient population. This study begins to fill these gaps. We learned that 15% of Veterans with a concurrent unexplained GI symptom and a structural GI diagnosis used high-dose opioids; these Veterans were more likely to have high-risk opioid exposure than those with only an unexplained symptom or a structural diagnosis. In one prior study of Crohn’s disease patients with a concurrent functional GI disorder, Crocker et al found that 20% of patients were using opioids chronically and were significantly more likely to misuse opioid prescriptions [48]. Our study builds on these findings and demonstrates that—in addition to chronic use—patients with concurrent GI symptoms and diagnoses may also be at increased risk of high-dose daily use. The reasons underlying this finding remain unclear, but it is possible that complex pain management needs and fragmented care among concurrent GI patients may be contributing to elevated risk of high-dose opioid use. Given the prevalence of functional GI disorders such as irritable bowel syndrome—both in the Veteran and general population [26,49,50], there is a need to better monitor the relationship between potentially unsafe opioid use and the presence of concurrent GI symptoms and diagnoses.

Our analysis also revealed a complex set of factors associated with high-dose use among chronic GI patients that has not been documented previously. In this cohort, over 70% of Veterans received opioids from Medicare Part D alone or via both the VA and Part D. Our findings align with other studies in the broader Veteran population indicating that dual users are more likely to use high-dose opioids [33]. It is concerning, however, because receipt of opioids from multiple sources and high daily doses of opioids are substantially associated with higher risk of overdose death [51,52]. We also found that Veterans who used benzodiazepines were more likely to use high-dose opioids. Although no prior study to our knowledge has examined concurrent opioid and benzodiazepine use among individuals with the GI conditions included in this study, one analysis of cirrhosis patients found that rates of overlapping opioid and benzodiazepine use were higher among cirrhosis patients compared to those with other chronic diseases [53]. Given the safety risks associated with opioid and concurrent benzodiazepine use [54], further studies are needed to explore the potentially overlapping use of these medications among GI patients.

Some of the most prominent factors linked with high-dose use in our study were related to the duration of opioid use and the type of individual opioid prescriptions. Veterans with long-term opioid use in this cohort had more than twice the risk of high-dose use; this is alarming given that long-term, high-dose opioid use is considered particularly dangerous [38,39,51]. We noted the highest risk of high-dose opioid use (particularly for doses ≥120 MMEs / day) among Veterans with prescriptions for fentanyl, meperidine, and morphine use. Because these particular medications are considered strong opioid analgesics intended to treat moderate to severe pain [11,55], perhaps their association with high-doses is conceivable. Nevertheless, it raises concern particularly in light of evidence that opioid prescriptions such as fentanyl are linked with illicit opioid use and death [56]. Use of weaker opioids, including tramadol and codeine, was associated with a lower likelihood of high-dose use. For some GI patients, use of these medications may be safer than other opioids and may help to address pain [57–59].

Our study underscores a potentially complicated challenge for gastroenterologists and other GI providers who need to carefully manage chronic illness and relieve pain while being attuned—from a safety perspective—to a patient’s individual needs and risks. This is particularly relevant for GI providers who care for Veterans that are dually enrolled in Medicare and may have access to medical care and prescriptions beyond the VA. Although further investigation is needed, GI providers who treat Veterans should be aware of potentially overlapping opioid receipt from the VA and other sources. Qualitative studies of GI provider perspectives and pain management practices are needed to characterize these and other challenges and may help to determine the ways in which we can enhance the safety of GI care and pain management. Among these challenges may also be the possibility for GI patients to receive opioids for concurrent comorbidities from health care providers outside of gastroenterology. Data are needed to understand prescribing practices more broadly for GI patients and, subsequently, to identify potential coordination of care issues that may result in unique patient safety risks for this population. In addition, there has been limited evidence to direct the appropriate use and dosage of opioids for chronic GI patients. While opioids are essential for short term management of pain, opioid use for unexplained symptoms—including chronic abdominal pain and functional GI disorders—should be limited [22,26]. Evidence-based guidance and provider education in this area is needed [60].

This study had some limitations. First, our findings may not reflect the current state of opioid use among Veterans with chronic GI symptoms and disorders. Second, we used administrative data for our analyses; administrative data carries the potential risk for misclassification. To address this, we relied on previously published data definitions. Additionally, we were unable to verify whether filled opioids were used as prescribed and were not able to determine whether each opioid prescription was specifically attributed to a GI condition. Although we did not know the indication for opioid prescriptions, we attempted to address this by observing opioid use for only those GI patients who had an opioid fill within six months of GI diagnosis. Our findings reflect a predominantly male population and may not generalize to non-Veteran populations. Moreover, because our focus was on a dually enrolled group, our findings may not generalize to Veterans who are not dually enrolled in the VA and Medicare Part D.

In this cohort of dually enrolled Veterans with chronic GI symptoms and disorders, approximately 11% of Veterans with unexplained GI symptoms, 10% of Veterans with structural GI diagnoses, and 15% of Veterans with concurrent GI symptoms and diagnoses had a mean daily opioid dose ≥100 MMEs. Compared to Veterans who had solely an unexplained GI symptom or a structural GI diagnosis, Veterans with both an unexplained GI symptom and structural GI diagnosis had a significantly greater likelihood of having a higher daily opioid dose, higher number of nonconsecutive days on opioids ≥100 MMEs, and higher risk of long-term opioid use. Factors independently associated with high-dose daily opioid included fentanyl, meperidine, and morphine use, receipt of opioid prescriptions from both the VA and Medicare Part D, benzodiazepine or muscle relaxant medication use, and having a diagnosis for nonopioid drug dependence. Efforts are needed to carefully monitor and reduce high-dose opioid use, particularly among Veterans with concurrent GI symptoms and disorders.

Acknowledgements

Funding / Support: This study was conducted while S.N.B. was a National Research Service Award pre-doctoral fellow under an institutional award from the National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK), T32DK101363 (PI: Pandolfino). Additionally, this work was supported by the Department of Veterans Affairs (VA) Health Services Research & Development grant I01 HX001765-01. Support for VA–Centers for Medicare & Medicaid Services data is provided by VA, Veterans Health Administration, Office of Research & Development, Health Services Research & Development, and VA Information Resource Center (SDR 02-237 and 98-004).

Footnotes

Publisher's Disclaimer: This Author Accepted Manuscript is a PDF file of an unedited peer-reviewed manuscript that has been accepted for publication but has not been copyedited or corrected. The official version of record that is published in the journal is kept up to date and so may therefore differ from this version.

Disclosures: All authors declare that they have no conflicts of interest.

Publisher's Disclaimer: Disclaimer: The views expressed in this article are those of the authors and do not necessarily represent the views of the U.S. Department of Veterans Affairs.

References

  • 1.Report of the President’s Commission on Combatting Drug Addiction and the Opioid Crisis. November 1, 2017. [Google Scholar]
  • 2.Blendon RJ, Benson JM. The public and the opioid-abuse epidemic. New Eng J Med. 2018;378(5):407–411. [DOI] [PubMed] [Google Scholar]
  • 3.Beauchamp GA, Winstanley EL, Ryan SA, Lyons MS. Moving beyond misuse and diversion: the urgent need to consider the role of iatrogenic addiction in the current opioid epidemic. Am J Pub Health. 2014;104(11):2023–2029. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Volkow ND. America’s addiction to opioids: Heroin and prescription drug abuse. Senate Caucus on International Narcotics Control. 2014;14. [Google Scholar]
  • 5.Webster LR, Cochella S, Dasgupta N, et al. An analysis of the root causes for opioid-related overdose deaths in the United States. Pain Med. 2011;12(suppl_2):S26–S35. [DOI] [PubMed] [Google Scholar]
  • 6.Scholl L, Seth P, Kariisa M, Wilson N, Baldwin G. Drug and opioid-involved overdose deaths—United States, 2013–2017. Morbidity and Mortality Weekly Report. 2019;67(5152): 1419. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Centers for Disease Control and Prevention NCfHS. Number and age-adjusted rates of drugpoisoning deaths involving opioid analgesics and heroin: United States, 2000-2014. [Google Scholar]
  • 8.Chou R, Fanciullo GJ, Fine PG, et al. Clinical guidelines for the use of chronic opioid therapy in chronic noncancer pain. J Pain. 2009; 10(2): 113–130. el22. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Edlund MJ, Martin BC, Fan M-Y, Devries A, Braden JB, Sullivan MD. Risks for opioid abuse and dependence among recipients of chronic opioid therapy: results from the TROUP study. Drug and Alcohol Depend. 2010;112(1-2):90–98. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Juurlink DN, Dhalla IA. Dependence and addiction during chronic opioid therapy. J Medical Toxicol. 2012;8(4):393–399. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Chou R, Turner JA, Devine EB, et al. The effectiveness and risks of long-term opioid therapy for chronic pain: a systematic review for a National Institutes of Health Pathways to Prevention Workshop. Ann Intern Med. 2015;162(4):276–286. [DOI] [PubMed] [Google Scholar]
  • 12.Zhang Y, Johnson P, Jeng PJ, et al. First opioid prescription and subsequent high-risk opioid use: a national study of privately insured and Medicare Advantage adults. J Gen Intern Med. 2018;33(12):2156–2162. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Edlund MJ, Martin BC, Russo JE, DeVries A, Braden JB, Sullivan MD. The role of opioid prescription in incident opioid abuse and dependence among individuals with chronic non-cancer pain: the role of opioid prescription. Clin J Pain. 2014;30(7):557. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Gallagher RM. Advancing the pain agenda in the veteran population. Anesthesiol Clinics. 2016;34(2):357–378. [DOI] [PubMed] [Google Scholar]
  • 15.Baser O, Xie L, Mardekian J, Schaaf D, Wang L, Joshi AV. Prevalence of diagnosed opioid abuse and its economic burden in the Veterans Health Administration. Pain Pract. 2014;14(5):437–445. [DOI] [PubMed] [Google Scholar]
  • 16.Bohnert AS, Ilgen MA, Galea S, McCarthy JF, Blow FC. Accidental poisoning mortality among patients in the Department of Veterans Affairs Health System. Med Care. 2011:393–396. [DOI] [PubMed] [Google Scholar]
  • 17.Banerjee G, Edelman EJ, Barry DT, et al. High-dose prescribed opioids are associated with increased risk of heroin use among US military veterans. Pain. 2019. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Bohnert AS, Logan JE, Ganoczy D, Dowell D. A detailed exploration into the association of prescribed opioid dosage and overdose deaths among patients with chronic pain. Med Care. 2016;54(5):435–441. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Seal KH, Shi Y, Cohen G, et al. Association of mental health disorders with prescription opioids and high-risk opioid use in US veterans of Iraq and Afghanistan. J Am Med Assoc. 2012;307(9):940–947. [DOI] [PubMed] [Google Scholar]
  • 20.Zedler B, Xie L, Wang L, et al. Risk factors for serious prescription opioid-related toxicity or overdose among Veterans Health Administration patients. Pain Med. 2014;15(11):1911–1929. [DOI] [PubMed] [Google Scholar]
  • 21.Oliva EM, Bowe T, Tavakoli S, et al. Development and applications of the Veterans Health Administration’s Stratification Tool for Opioid Risk Mitigation (STORM) to improve opioid safety and prevent overdose and suicide. Psychol Services. 2017; 14(1):34. [DOI] [PubMed] [Google Scholar]
  • 22.Szigethy E, Knisely M, Drossman D. Opioid misuse in gastroenterology and non-opioid management of abdominal pain. Nature Rev Gastroenterol Hepatol. 2018;15(3):168–180. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Drossman D, Szigethy E. The narcotic bowel syndrome: a recent update. Am J Gastroenterol Suppl. 2014;2(1):22. [DOI] [PubMed] [Google Scholar]
  • 24.Grunkemeier DM, Cassara JE, Dalton CB, Drossman DA. The narcotic bowel syndrome: clinical features, pathophysiology, and management. Clin Gastroenterol and Hepatol. 2007;5(10):1126–1139. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Jones JL, Loftus EV Jr. Avoiding the vicious cycle of prolonged opioid use in Crohn’s disease. Am J Gastroenterol. 2005;100(10):2230. [DOI] [PubMed] [Google Scholar]
  • 26.Sayuk GS, Kanuri N, Gyawali CP, Gott BM, Nix BD, Rosenheck RA. Opioid medication use in patients with gastrointestinal diagnoses vs unexplained gastrointestinal symptoms in the US Veterans Health Administration. Aliment Pharmacol Therapeutics. 2018;47(6):784–791. [DOI] [PubMed] [Google Scholar]
  • 27.Dorn SD, Meek PD, Shah ND. Increasing frequency of opioid prescriptions for chronic abdominal pain in US outpatient clinics. Clin Gastroenterol Hepatol. 2011;9(12): 1078–1085.e1071. [DOI] [PubMed] [Google Scholar]
  • 28.Drossman DA. Severe and refractory chronic abdominal pain: treatment strategies. Clin Gastroenterol Hepatol. 2008;6(9):978–982. [DOI] [PubMed] [Google Scholar]
  • 29.Balbale SN, Trivedi I, O’dwyer LC, et al. Strategies to identify and reduce opioid misuse among patients with gastrointestinal disorders: a systematic scoping review. Dig Dis Sci. 2017;62(10):2668–2685. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Targownik LE, Nugent Z, Singh H, Bugden S, Bernstein CN. The prevalence and predictors of opioid use in inflammatory bowel disease: a population-based analysis. Am J Gastroenterol. 2014;109(10): 1613. [DOI] [PubMed] [Google Scholar]
  • 31.Gao Y, Khan S, Akerman M, Sultan K. Analysis of the clinical indications for opiate use in inflammatory bowel disease. Intestinal Res. 2017;15(1):83. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32.Olesen SS, Poulsen JL, Broberg MC, Madzak A, Drewes AM. Opioid treatment and hypoalbuminemia are associated with increased hospitalisation rates in chronic pancreatitis outpatients. Pancreatol. 2016;16(5):807–813. [DOI] [PubMed] [Google Scholar]
  • 33.Gellad WF, Thorpe JM, Zhao X, et al. Impact of Dual Use of Department of Veterans Affairs and Medicare Part D Drug Benefits on Potentially Unsafe Opioid Use. Am J Public Health. 2018;108(2):248–255. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34.Battles J, Lilford R. Organizing patient safety research to identify risks and hazards. BMJ Qnal Saf. 2003;12(suppl 2):ii2–ii7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Donabedian A The quality of medical care. Science. 1978;200(4344): 856–864. [DOI] [PubMed] [Google Scholar]
  • 36.Nguyen NH, Khera R, Ohno-Machado L, Sandborn WJ, Singh S. Annual Burden and Costs of Hospitalization for High-Need, High-Cost Patients With Chronic Gastrointestinal and Liver Diseases. Clin Gastroenterol Hepatol. 2018; 16(8): 1284–1292 e1230. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Hou JK, Tan M, Stidham RW, et al. Accuracy of diagnostic codes for identifying patients with ulcerative colitis and Crohn’s disease in the Veterans Affairs Health Care System. Dig Dis Sci. 2014;59(10):2406–2410. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38.Bohnert AS, Valenstein M, Bair MJ, et al. Association between opioid prescribing patterns and opioid overdose-related deaths. J Am Med Assoc. 2011. ;305(13): 1315–1321. [DOI] [PubMed] [Google Scholar]
  • 39.Dunn KM, Saunders KW, Rutter CM, et al. Overdose and prescribed opioids: Associations among chronic non-cancer pain patients. Ann Intern Med. 2010;152(2):85.20083827 [Google Scholar]
  • 40.Prevention CfDCa. Opioid morphine equivalent conversion factors. 2018. [Google Scholar]
  • 41.Edlund MJ, Austen MA, Sullivan MD, et al. Patterns of opioid use for chronic noncancer pain in the Veterans Health Administration from 2009 to 2011. Pain. 2014;155(11):2337–2343. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 42.Services USCfMM. Limited Income and Resources. 2019; https://www.cms.gov/Medicare/Prescription-Drug-Coverage/LimitedIncomeandResources/index.html.
  • 43.Johnson RG, Arozullah AM, Neumayer L, Henderson WG, Hosokawa P, Khuri SF. Multivariable predictors of postoperative respiratory failure after general and vascular surgery: results from the patient safety in surgery study. J Am College Surgeons. 2007;204(6):1188–1198. [DOI] [PubMed] [Google Scholar]
  • 44.Edlund MJ, Steffick D, Hudson T, Harris KM, Sullivan M. Risk factors for clinically recognized opioid abuse and dependence among veterans using opioids for chronic non-cancer pain. Pain. 2007;129(3):355–362. [DOI] [PubMed] [Google Scholar]
  • 45.Deyo RA, Hallvik SE, Hildebran C, et al. Association between initial opioid prescribing patterns and subsequent long-term use among opioid-naïve patients: a statewide retrospective cohort study. J Gen Intern Med. 2017;32(1):21–27. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 46.Butler MM, Ancona RM, Beauchamp GA, et al. Emergency department prescription opioids as an initial exposure preceding addiction. Ann Emerg Med. 2016;68(2):202–208. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 47.Szigethy E, Schwartz M, Drossman D. Narcotic bowel syndrome and opioid-induced constipation. Curr Gastroenterol Rep. 2014;16(10):410. [DOI] [PubMed] [Google Scholar]
  • 48.Crocker JA, Yu H, Conaway M, Tuskey AG, Behm BW. Narcotic use and misuse in Crohn’s disease. Inflamm Bowel Dis. 2014;20(12):2234–2238. [DOI] [PubMed] [Google Scholar]
  • 49.Peery AF, Crockett SD, Barritt AS, et al. Burden of Gastrointestinal, Liver, and Pancreatic Diseases in the United States. Gastroenterol. 2015;149(7):1731–1741 e1733. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 50.Tuteja AK, Talley NJ, Stoddard GJ, Samore MH, Verne GN. Risk factors for upper and lower functional gastrointestinal disorders in Persian Gulf War Veterans during and post- deployment. Neurogastroenterol Motil. 2019;31(3):e13533. [DOI] [PubMed] [Google Scholar]
  • 51.Baumblatt JAG, Wiedeman C, Dunn JR, Schaffner W, Paulozzi LJ, Jones TF. High-risk use by patients prescribed opioids for pain and its role in overdose deaths. JAMA Intern Med. 2014;174(5):796–801. [DOI] [PubMed] [Google Scholar]
  • 52.Rice JB, White AG, Bimbaum HG, Schiller M, Brown DA, Roland CL. A model to identify patients at risk for prescription opioid abuse, dependence, and misuse. Pain Med. 2012;13(9):1162–1173. [DOI] [PubMed] [Google Scholar]
  • 53.Konerman MA, Rogers M, Kenney B, et al. Opioid and benzodiazepine prescription among patients with cirrhosis compared to other forms of chronic disease. BMJ Open Gastroenterol. 2019;6(1):e000271. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 54.Gressler LE, Martin BC, Hudson TJ, Painter JT. Relationship between concomitant benzodiazepine-opioid use and adverse outcomes among US veterans. Pain. 2018;159(3):451–459. [DOI] [PubMed] [Google Scholar]
  • 55.Dowell D, Haegerich TM, Chou R. CDC guideline for prescribing opioids for chronic pain—United States, 2016. J Am Med Assoc. 2016;315(15): 1624–1645. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 56.Schuchat A, Houry D, Guy GP. New data on opioid use and prescribing in the United States. J Am Med Assoc. 2017;318(5):425–426. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 57.Burr NE, Smith C, West R, Hull MA, Subramanian V. Increasing prescription of opiates and mortality in patients with inflammatory bowel diseases in England. Clin Gastroenterol Hepatol. 2018; 16(4):534–541. e536. [DOI] [PubMed] [Google Scholar]
  • 58.Iacobellis A, Seripa D, Palmieri O, Andriulli N, Latina P. Efficacy and Safety of Long-Term Administration of Tapentadol in Relieving Chronic Pancreatitis Pain. Pain Med. 2017; 18(4):815–817 [DOI] [PubMed] [Google Scholar]
  • 59.Wilder-Smith CH, Hill L, Osler W, O’keefe S. Effect of tramadol and morphine on pain and gastrointestinal motor function in patients with chronic pancreatitis. Dig Dis Sci. 1999;44(6): 1107–1116. [DOI] [PubMed] [Google Scholar]
  • 60.Reynolds C, Esrailian E, Hommes D. Quality improvement in gastroenterology: a systematic review of practical interventions for clinicians. Dig Dis Sci. 2018;63(10):2507–2518. [DOI] [PubMed] [Google Scholar]

RESOURCES