Presented in part at the 108th Annual Meeting of the American Gastroenterological Association Institute, Washington, DC, May 19–24, 2007 (Gastroenterology 2007;132(4suppl 2):A-654).
Clinical features and outcome of patients with inflammatory bowel disease who use narcotics: A case–control study†
Article first published online: 23 DEC 2008
Copyright © 2008 Crohn's & Colitis Foundation of America, Inc.
Inflammatory Bowel Diseases
Volume 15, Issue 5, pages 772–777, May 2009
How to Cite
Hanson, K. A., Loftus, E. V., Harmsen, W. S., Diehl, N. N., Zinsmeister, A. R. and Sandborn, W. J. (2009), Clinical features and outcome of patients with inflammatory bowel disease who use narcotics: A case–control study. Inflamm Bowel Dis, 15: 772–777. doi: 10.1002/ibd.20847
- Issue published online: 7 APR 2009
- Article first published online: 23 DEC 2008
- Manuscript Accepted: 6 NOV 2008
- Manuscript Received: 6 AUG 2008
- Mayo Foundation for Medical Education and Research
- narcotic use;
- clinical features;
Background: The role of narcotic use in patients with inflammatory bowel disease (IBD) is poorly defined. We sought to determine the clinical features of patients with IBD who use narcotics and factors associated with the discontinuation of narcotics.
Methods: A centralized index was used to identify 100 cases (patients with IBD receiving narcotics) and 100 matched controls evaluated in an IBD clinic between 1999 and 2002. Associations of clinical factors with case or control status were summarized as overall percentages and assessed using conditional logistic regression. Associations within cases were assessed using Fisher's exact test.
Results: Review of 361 charts identified 100 patients with IBD receiving narcotics (cases). One hundred matched controls were then identified. Cases were significantly associated with: female gender (64% cases versus 45% controls, P = 0.01), ≥2 IBD-related surgeries (42% versus 17%, P < 0.001), moderate-to-severe pain (93% versus 20%, P < 0.001), and moderate-to-severe symptoms (83% versus 33%, P < 0.001). Significant associations were detected for depression (42% versus 19%, P < 0.001), anxiety (19% versus 7%, P = 0.02), history of abuse (sexual, emotional, or physical, 17% versus 3%, P = 0.006), and for substance abuse (excluding alcohol) (14% versus 1%, P = 0.01). Of 39 cases that returned for follow-up, 22 (56%) had discontinued narcotics. Discontinuing narcotics was associated with medical treatment adherence (100% versus 53%), none-to-mild pain (73% versus 18%), and none-to-mild clinical activity (80% versus 24%).
Conclusions: Narcotic use in IBD patients is associated with female gender, ≥2 previous surgeries, moderate to severe pain, and clinical disease activity, depression, anxiety, sexual, emotional and physical abuse, and substance abuse.
(Inflamm Bowel Dis 2008)
Patients with inflammatory bowel disease (IBD), ulcerative colitis (UC) or Crohn's disease, commonly present with symptoms of pain in the abdomen and/or pelvis, and in patients with associated rheumatologic extraintestinal manifestations of IBD, pain in the joints. The severity of pain experienced by patients is variable. It is expected that in patients with active IBD and pain, confirmatory symptoms of disease activity such as diarrhea, fecal urgency, abdominal cramping, fever, chills, night sweats, or weight loss, or confirmatory biochemical changes such as anemia, leukocytosis, increased sedimentation rate, increased C-reactive protein, or hypoalbuminemia should be present as well. Extraintestinal manifestations, including iritis or uveitis, arthritis, aphthous stomatitis, or skin lesions of erythema nodosum or pyoderma gangrenosum may also be present in parallel in patients with active IBD and pain. Improvement in abdominal pain is expected following medical or surgical treatment of active IBD. However, some patients with a confirmed diagnosis of IBD have persistent pain symptoms in the absence of any objective evidence of active disease as determined by endoscopic, histologic, or radiographic or biochemical studies.1, 2
Because IBD is a chronic and generally nonfatal condition, treatment of pain with narcotics is generally contraindicated for fear of addiction and fear of gastrointestinal side effects such as nausea and constipation, and in the case of UC, precipitation of toxic megacolon. Nevertheless, narcotics may be prescribed for short periods of time for specific indications expected to be of short duration such as acute pain during hospitalization following surgical therapy. Despite this general contraindication, some practitioners have prescribed narcotics for chronic pain in patients with IBD, even in the absence of objective evidence of disease activity. This may lead to narcotic bowel syndrome, an opioid-induced bowel dysfunction characterized by chronic or frequently recurring abdominal pain that worsens in spite of continued or escalating doses of narcotics.3
Withdrawal of narcotics may challenge the provider–patient relationship, especially if patients believe they need continued or increasing doses of narcotics for pain relief.4 Patients may also feel that their healthcare provider is questioning the legitimacy of their complaints or they may feel abandoned.4 Such difficulties in treating chronic pain with narcotics highlight the need for early recognition and possible preventive measures in at-risk patients. However, it is not clear whether early intervention could prevent the development of a chronic pain syndrome.5
Only a few studies have critically examined which factors might be associated with chronic narcotic use in IBD patients. Edwards et al1 found that these patients were significantly more likely to have a preexisting psychiatric disorder than those not on narcotics. Cross et al2 estimated that 13% of their CD patients were on chronic narcotics, and these patients were more likely to be female, have higher rates of disability and longer duration of disease, smoke cigarettes, and take at least 5 different medications. However, the number of patients on chronic narcotics in both studies was relatively low.
We therefore sought to identify clinical and demographic features associated with narcotic use among patients with IBD referred to a tertiary medical center and to identify factors associated with successful discontinuation of narcotics.
MATERIALS AND METHODS
Study Setting and Patients
This was a retrospective chart-based review of selected patients evaluated in the IBD Clinic at Mayo Clinic in Rochester, Minnesota, with a diagnosis of CD, UC, or ileal pouchitis between January 1, 1999, and December 31, 2002. The Institutional Review Board of Mayo Clinic approved the study. A centralized diagnostic index identified patients with a diagnosis of IBD and a brief review of the medical record identified those who were receiving narcotic medication at the time of their initial IBD Clinic evaluation. Both cases and controls had to have endoscopic, radiographic, or histologic confirmation of IBD in order to be included in the study group. The medical records of consecutive patients who had given research authorization were reviewed until 100 IBD patients who were being treated with narcotics for IBD-related pain at their initial visit (cases) were identified. Patients who were taking narcotic analgesics for other reasons (eg, non-IBD-related pain, cancer pain, postoperative pain control, diarrhea control) were excluded from the study group. A cohort of 100 IBD patients not being treated with narcotics at the time of their initial Mayo visit (controls) were then identified through medical indexing and matched 1:1 for subtype of IBD and for geographic location of residence.
Details about the subtype of IBD, age of onset, duration of disease, surgical interventions, severity and location of pain symptoms, and disease activity as assessed by clinical symptoms, endoscopy, histology, and radiography at the time of the initial Mayo evaluation were abstracted from the medical record of each patient. The clinical, endoscopic, histologic, and radiographic severity of IBD was categorized as none, mild, moderate, or severe based on the global assessment provided by the clinician, endoscopist, pathologist, or radiologist. Information regarding gender, age, comorbid medical and psychological diseases, history of substance abuse, and history of physical, emotional, or sexual abuse were collected. Much of this information was available in a questionnaire completed by each patient at the time of a new episode of care that is systematically administered to all patients seen at the Mayo Clinic as part of standard clinical practice. In addition, follow-up data regarding disease activity and narcotic use were collected.
Descriptive statistics were reported separately for cases (patients receiving narcotics at the time of presentation to Mayo) and controls, ignoring the paired nature of the data. The univariate association between individual variables and case/control status was assessed using conditional logistic regression to account for the matching of cases and controls. The results were presented as odds ratio (OR) with associated 95% confidence intervals (CIs).
Selected clinical and demographic characteristics were examined in a multiple variable model analysis using a backward elimination approach to identify a final set of variables independently associated with narcotic use at presentation. The alpha-level was set at 0.05 for statistical significance.
Discontinuation of narcotics use at latest follow-up among case patients was reported descriptively. A formal analysis of the association between discontinuation of narcotics and various clinical factors was not undertaken since a number of cases were lost to follow-up, the duration of follow-up in those not lost was variable, and the exact date of discontinuation was often not known.
A total of 361 charts indexed for new cases of CD or UC and narcotic use documented during the study period were reviewed to identify 100 cases (78 CD, 22 UC). Of the 261 excluded, 194 were due to use of narcotics for other reasons (ie, pain unrelated to IBD, cancer, postoperative IBD, or diarrhea control), and 67 did not have a diagnosis of IBD confirmed at Mayo. Patients with ileal pouch-anal anastomosis were included in the UC group. Most patients (90 cases and 74 controls) had an outside diagnosis of IBD at the time of presentation to our clinic. Subsequently, 100 controls, matched on IBD subtype and geographic location of residence were identified after review of 500 additional charts. Five case/control pairs were from Olmsted County or the surrounding 6 counties, 10 were from the remainder of Minnesota, 29 were from states surrounding Minnesota (Wisconsin, Illinois, Iowa, South Dakota, or North Dakota), and 56 were outside of this area. There were 4 cases and 4 controls from outside of the United States.
Most patients were either seen directly by the IBD clinic (46 cases, 49 controls) or general gastroenterology (35 cases, 24 controls), and all others were referred from elsewhere within the multispecialty clinic. Most patients were on medical therapy at presentation (78% of cases versus 62% of controls) (Table 1). Cases were significantly associated with: female gender, age at presentation to Mayo, duration of disease at presentation to Mayo, a history of 2 or more IBD-related surgeries, moderate to severe pain at presentation, and moderate to severe clinical disease, but not with endoscopic, histologic, or radiographic severity. Among cases, significant positive associations (ie, greater odds for a patient being on narcotics at presentation to Mayo) were detected for depression, anxiety, other gastrointestinal disorders, and a history of sexual, emotional, or physical abuse (Table 2). Case/control status was not significantly associated with fibromyalgia, interstitial cystitis, lung disease, tobacco use, or other psychiatric condition, or diabetes mellitus. Substance abuse (excluding alcohol and tobacco) was also associated with narcotic use. Tobacco and alcohol abuse were not associated with narcotic use. Patients with vascular disease had significantly lower odds of narcotic use at presentation to Mayo.
|Cases N = 100||Controls N = 100||P-value||OR (95%CI)|
|Medical therapy for IBD at presentation||78||62||0.02||2.1 (1.1–3.8)|
|History of ≥ 2 IBD surgeries||42||17||<0.001||2.9 (1.6–5.5)|
|Moderate to severe pain at presentation||93||20||<0.0001||38 (9.2–153)|
|Moderate to severe clinical activity at presentation||83||33||<0.0001||18 (5.5–57)|
|Moderate to severe endoscopic, histologic, or radiologic activitya||64||57||0.53||1.2 (0.7–2.3)|
|Age at Mayo presentation in years, median (range)||37.4 (1.4–70.5)||52.3 (12.0–81.4)||<0.001||0.96 (0.94–0.98)|
|Duration of disease in years, median (range)||6.2 (0–42.6)||2.6 (0–54.8)||0.42||0.99 (0.97–1.01)|
|Clinical History||Cases N = 100||Controls N = 100||P-value||OR (95% CI)|
|Other psychiatric condition||6||4||0.53||1.5 (0.4–5.3)|
|Vascular disease||21||37||0.019||0.5 (0.2–0.9)|
|Lung disease||9||10||1.0||0.9 (0.4–2.2)|
|Diabetes mellitus||9||5||0.29||1.8 (0.6–5.4)|
|Other gastrointestinal disorders||18||3||0.037||0.5 (0.2–0.96)|
|History of sexual, emotional or physical abuse||13||3||0.006||8.0 (1.8–34.8)|
|Tobacco abuse||53||51||0.79||1.1 (0.6–1.8)|
|Substance abusea||14||1||0.01||14 (1.8–106)|
Recommendations at the time of the Mayo presentation for psychiatric intervention including consultation, psychoactive medications, pain rehabilitation center, or inpatient addiction program were made for 15 cases and 2 controls either for IBD- or non-IBD-related concerns (OR, 7.0; 95% CI, 1.6–31) (Table 3). Seventy-one cases and 86 controls were prescribed medical intervention for IBD (OR, 0.4; 95% CI, 0.2–0.8), whereas a recommendation for surgical intervention was made in 35 cases and 18 controls (OR, 2.5; 95% CI, 1.3–5.1).
|Recommendation||Cases N = 100||Controls N = 100||P-value||OR (95% CI)|
|Psychiatric intervention||15||2||0.01||7.0 (1.6–31)|
|Medical intervention||71||86||0.01||0.4 (0.2–0.8)|
|Surgical interventiona||35||18||<0.01||2.5 (1.3–5.1)|
In a multiple variable model including clinical and demographic characteristics (listed in Table 1) and clinical history variables (listed in Table 2), those variables independently associated with an increased odds of the patient on narcotics at presentation to Mayo included: female gender (OR, 5.7; 95% CI, 1.3–26); moderate to severe pain at presentation (OR, 21; 95% CI, 4.5–95); and moderate to severe clinical activity at presentation (OR, 9.1; 95% CI, 1.5–5.4) (Table 4).
|Present: Present||Present: Absent||Absent: Present||Absent: Absent||P-value||OR (95% CI)|
|Moderate to severe pain at presentation||18||75||2||5||<0.001||21 (4.5–95)|
|Moderate to severe clinical activity at presentation||30||53||3||14||0.015||9.1 (1.5–5.4)|
Of the 200 total patients, 76 (38%) returned for 1 or more follow-up visits (39 cases, 37 controls). Of the 39 cases that returned, 22 (56%) were no longer receiving treatment with narcotics (Table 5). Endoscopic, histologic, or radiographic findings were not consistently performed at the follow-up visit. Among the 22 patients discontinuing narcotics at the latest follow-up, all were medically and surgically adherent, whereas in the 17 with continued narcotic use at their latest follow-up, 9 and 16 patients were medically and surgically adherent, respectively. Moderate to severe pain was reported in 27% and 82% of patients discontinuing and continuing narcotics, respectively. Among the patients who had discontinued narcotics at their latest follow-up, 80% reported none to mild clinical symptoms, while 24% of patients with continued narcotic use reported none to mild clinical symptoms at their latest follow-up.
|Narcotics Discontinued N = 22||Narcotics Continued N = 17|
|Medically adherent, n (%)||22 (100%)||9 (53%)|
|Surgically adherent n (%)||22 (100%)||16 (94%)|
|Mod/severe pain,an (%)||6 (27%)||14 (82%)|
|None/mild clinical symptoms,bn (%)||16 (80%)||4 (24%)|
In this referral center-based case–control study, we identified a number of factors associated with narcotic analgesic use, including female gender, a history of 2 or more surgeries for IBD, the presence of moderate to severe pain or other clinical disease activity, the presence of depression or anxiety, a history of sexual, emotional, or physical abuse, and a history of substance abuse. Endoscopic, histologic, and radiographic severity were not associated with narcotic use. IBD patients who were on narcotics at the time of their initial evaluation were almost twice as likely to undergo surgical treatment for their IBD. Among patients with sufficient follow-up, the discontinuation of narcotic analgesics was associated with medical adherence, little or no pain, and little or no clinical disease activity.
Providers have many concerns about narcotic use in patients with IBD. Use in the perioperative period, or with acute flares of perianal disease, are generally accepted, but chronic use leads to concerns of narcotic bowel syndrome, risk of toxic megacolon in patients with UC, narcotic dependence, development of hyperalgesia, and the potential to mask life-threatening complications. The recognition of narcotic bowel syndrome is made more difficult in the presence of underlying gastrointestinal disease.3 Further, concern about the use of narcotics in IBD patients has been heightened by the finding in a postmarketing registry of CD patients on infliximab of a significant association between serious infections and narcotic analgesics (OR 2.38; 95% CI, 1.56–3.63).6 The mechanism behind this association is not clear. Although narcotic use could have been a marker for increased disease severity, the possibility remains that the use of narcotics directly increased the risk of serious infections, perhaps by decreasing gut transit, increasing bacterial growth within the gut, or by some other mechanism.
Cross et al2 reported that narcotic use was associated with more severe disease based on the Harvey–Bradshaw Index (HBI), a primarily subjective symptom-driven index. In our study a significant association of narcotic use with subjective clinical disease but not with endoscopic, histologic, or radiographic disease activity was identified. Our study also demonstrated significant associations between depression, anxiety, history of physical, emotional and sexual abuse, and history of substance abuse. Walker et al7 noted that IBD patients with a concurrent psychiatric diagnosis reported more gastrointestinal and nongastrointestinal symptoms both associated and not associated with IBD. Patients with concurrent psychiatric and IBD diagnoses also had lower health-related quality of life as measured by scores on the SF-36 subscales of physical, emotional, and vocational role function, and vitality and health perception, despite similar objective findings compared to IBD patients without psychiatric diagnoses. They speculated that symptoms of unclear etiology might be the result of psychologic distress and maladaptive illness behaviors as well as functional disability in these patients. This suggests that patients presenting with psychiatric history are at an increased risk of narcotic dependence and may be best managed concomitantly with a psychiatrist or pain management specialist from the onset of care.
In this study, only 15% of patients on narcotics had a formal psychiatric or pain management recommendation or referral at presentation to our center. This may have been due to the lack of knowledge as to whether patients were truly at risk for narcotic dependence or merely had undertreated IBD. However, perhaps earlier multidisciplinary intervention would improve the rate of narcotic discontinuation in those patients identified early to be at risk, or at least improve the ability to optimally manage the special circumstances that arise in IBD patients on narcotics. Walker et al7 noted that “non-recognition of a psychiatric disorder may lead to unnecessary and aggressive interventions for IBD such as medication changes, invasive testing and surgery.” It would be important to further examine whether screening for psychiatric disease and subsequent cognitive behavioral intervention would reduce narcotic dependence, improve coping strategies, and improve quality of life in patients over the long term.
Patients with IBD are often relatively young and free of other chronic health conditions. We found no significant associations between narcotic use and conditions such as fibromyalgia, interstitial cystitis, lung disease, or diabetes mellitus. Therefore, the comorbid disease burden was not increased in narcotic users. Additionally, fibromyalgia and other functional gastrointestinal disorders associated with irritable bowel syndrome were not seen more often in narcotic users.
We found that a history of 2 or more IBD surgeries was significantly associated with narcotic use. Whether postsurgical adhesions can be the source of chronic abdominal pain remains controversial.8, 9 Herrick et al10 studied 29 patients undergoing laparotomy for a variety of conditions and found adhesions to be highly vascularized, cellular, and innervated. This suggests a possible relationship between postoperative peritoneal adhesions and abdominal pain. Such postoperative developments may lead to a poorer quality of life, as suggested in 1 study of 164 IBD patients using the Cleveland Clinic health-related quality of life instrument.11 In this study, patients with CD who had undergone surgery had the lowest quality of life scores.11 On the other hand, no significant association between adherence to surgical recommendations and the ability to discontinue narcotics in our study was observed.
Our study has several limitations. First of all, the retrospective nature of our study did not allow for a standardized assessment of clinical, endoscopic, histologic, or radiographic disease severity. We relied on the global assessment of the clinician, endoscopist, pathologist, or radiologist for these determinations. Second, the referral center-based aspect of this study did not permit a longitudinal assessment of narcotic use prior to the cases' first assessment at our institution. Some of these patients may have been given narcotic analgesics in the emergency room, while hospitalized, or by well-meaning providers without actually exhibiting drug-seeking behavior or risk factors for narcotic dependence. Perhaps these patients ultimately discontinued narcotics and were inherently different from those who became dependent. Third, we had adequate follow-up data in only 38% of cases and controls after the initial consult, likely because our institution is a tertiary care referral center.
In general, our practice has been to avoid prolonged use of narcotics in IBD patients for all of the reasons described. It is interesting to note that the number of cases and controls who returned for follow-up was similar, suggesting that narcotic use may not be a factor in healthcare-seeking behavior at our institution. One of the reasons for undertaking this study was to examine whether patients on narcotics were not being optimally managed medically or surgically, or whether they were endowed with characteristics known to put them at risk for narcotic abuse. The results were mixed. Patients in our study who were compliant with medical recommendations were more successful in discontinuing narcotics. Although adherence to surgical recommendations was not associated with discontinuation of narcotics, almost twice as many cases as controls received surgical recommendations, presumably because of greater disease severity. Therefore, a substantial minority of our patients were in need of more aggressive medical or surgical management.
In conclusion, narcotic use in IBD patients is associated with female gender, 2 or more previous IBD surgeries, moderate to severe pain, moderate to severe clinical disease activity, depression, anxiety, sexual, emotional and physical abuse, and substance abuse. Use of narcotics did not correlate with endoscopic, histologic, or radiographic severity of IBD. Adherence to medical therapy may aid in discontinuation of narcotics in some patients.
- 2Narcotic use in patients with Crohn's disease. Am J Gastroenterol. 2005; 100: 2225–2229., , .Direct Link:
- 5A biopsychosocial understanding of gastrointestinal illness and disease. In: Feldman, ed. Sleisenger & Fordtran's Gastrointestinal and Liver Disease ( 7th ed.). St. Louis: WB Saunders; 2002: 2373–2385..