Screening And Behavioral Management: Medication Overuse Headache – The Complex Case


  • Conflict of Interest: None

Alvin Lake, Behavioral Medicine, Michigan Head-Pain and Neurological Institute, 3120 Professional Drive, Ann Arbor, MI 48104, USA.


The new appendix criteria for a broader concept of chronic migraine from the International Headache Society no longer require headache resolution or return to the previous headache pattern to confirm the diagnosis of medication overuse headache (MOH). MOH can be subdivided into simple (Type I) and complex (Type II). Complex cases may involve long-term use of daily opioids or combination analgesics, multisourcing, multiple psychiatric comorbidities, and/or a history of relapse. Daily use of opioids for other medical conditions, psychiatric comorbidity including borderline personality disorder, prior history of other substance dependence or abuse, and family history of substance disorders are risk factors for MOH. Relapse for analgesic overusers can be as high as 71% at 4-year follow-up. A case illustration spans 20 years from initial presentation through multiple periods of recovery and relapse to illustrate issues in the screening and management of complex MOH patients.

The definition of medication overuse headache (MOH) continues to evolve over time. The recently published new appendix criteria for a broader concept of chronic migraine1 of the International Headache Society define MOH as

  • 1Headache present on ≥15 days/month;
  • 2Regular overuse for ≥3 months of one or more acute/symptomatic treatment drugs, defined as
  • a. Ergotamine, triptans, opioids, or combination analgesic medication on ≥10 days /month on a regular basis for >3 months; or
  • b. Simple analgesics or any combination of ergotamine, triptans, analgesics, opioids on ≥15 days/month on a regular basis for >3 months without overuse of any single class alone.
  • 3Headache has developed or markedly worsened during medication overuse.

In contrast to the previous International Headache Society revised definition of MOH (2005),2 the new appendix criteria no longer require headache resolution or return to the previous headache pattern to confirm the diagnosis of MOH, and eliminate the term “probable MOH.” The authors note 3 primary reasons for the criteria revision: (1) some patients may not improve in headache frequency after withdrawal, but may become more responsive to prophylactic medication; (2) medication overuse could contribute to chronicity, but not be reversible after withdrawal; and (3) the new criteria would establish MOH as a default diagnosis in all patients with medication overuse, and thereby “encourage doctors all over the world to do the right thing, namely, to take patient off medication overuse as the first step in a treatment plan.”1 Although the new appendix criteria no longer require headache remission after withdrawal to confirm the MOH diagnosis, effective management still requires withdrawal of the overused medication or combination of medications, in conjunction with prophylactic treatment.

Analgesic overuse and subsequent MOH is a worldwide problem. Epidemiological data suggest that up to 4% of the population may overuse analgesics and other symptomatic medications for the management of pain disorders such as migraine.3 MOH occurs in about 1% of the adult population of Europe, North America, and Asia,3,4 and in 0.5% of adolescents.5 In specialized headache centers, the prevalence of MOH ranges from estimates of 15-20% in Europe6 to as high as 59-64% of patients seeking treatment at tertiary headache care centers in the United States.7

Several large population-based studies have identified the overuse of acute pain medication as a primary risk factor for the development of chronic daily headache (≥15 days/month).8 In the Head-HUNT study by Zwart et al, the use of analgesics daily or weekly elevated risk for chronic daily headache at an 11-year follow-up, with significantly higher relative risk (RR) for chronic migraine (RR = 13.3) and chronic nonmigrainous headache (RR = 6.2) than chronic neck pain (RR = 2.2) or chronic low back pain (RR = 2.3).4 Other risk factors for MOH include frequent to daily use of opioids for other medical conditions,9,10 psychiatric comorbidity11 including borderline personality disorder,12,13 and dependence on other psychoactive substances (including alcohol and nicotine)11 as well as a family history of substance disorders.11,14 In the retrospective study by Radat et al, psychiatric comorbidity was not only significantly more prevalent in MOH than episodic migraine, but was also significantly more likely to precede rather than follow the onset of MOH.11


Joel Saper and I have proposed a distinction between simple MOH (Type I) and complex MOH (Type II).15 Although currently not operationalized in detail, simple MOH refers to straightforward cases: short term overuse (eg, 3 months to 1 year), with modest doses (albeit daily) of analgesics or triptans, intervals between doses, psychiatric comorbidity limited to one or 2 Axis I clinical syndromes, and no history of relapse. Complex MOH involves a long-term history (>1 year) of daily opioids or combination analgesics, often with more that one doctor prescribing, multiple psychiatric comorbidities (including Axis II personality disorders), and/or a history of relapse after drug withdrawal.13 In general practice, simple MOH is probably the more common, certainly the easiest to manage, and may benefit from behavioral intervention. Although less frequently encountered, it is complex MOH that poses the greater clinical conundrum, frustrates the compassionate clinician, may lead well-meaning physicians to over-prescribe medications, and requires behavioral intervention.

The physician may screen for MOH (Type I) by directly asking the patient about medication use patterns, and then applying the criteria. Common pitfalls in failing to identify relatively simple cases of MOH include the following:

  • 1The assumption that the overused medication must be taken daily – note that the MOH criteria of 10 days/month for the indicated prescription acute medications is an average of less than 2.5 days/week;
  • 2The assumption that the medication must be taken in large quantities – the new appendix criteria only focus on days/month, not the number of tablets or capsules; and
  • 3The assumption that MOH can be avoided by “mixing and matching” medications – the appendix criteria are explicit that MOH is implicated when any combination of ergotamine, triptans, analgesics, or opioids is taken for 15 or more days/month, an average of only one such medication every other day; and
  • 4The assumption that medications taken for nonheadache conditions “don't count”– for example, in a 1-year follow-up of patients taking daily opiates to control bowel motility after surgery for ulcerative colitis, Wilkinson et al found that all patients with presurgery episodic migraine developed chronic daily headache.10

In complex cases (MOH, Type II), the health-care provider may need to employ multiple assessments to identify the problem: detailed questioning of the patient, interviews with family members, contact with other prescribing physicians, and review of records from insurance companies or state registry systems if available. For example, Michigan physicians can request a report from the Michigan Automated Prescription System, which documents all prescriptions for controlled substances that were filled by the patient over the past year. These reports can then be reviewed with the patient, and prescribing physicians can be contacted directly if there is evidence of multisourcing. Urine drug screens can be helpful, both in identifying drugs which were not reported by the patient, and in cases where biochemical markers of medications that the patient claims to be taking fail to show up in the urine – a red flag for duplicity.


To date, studies of MOH have not distinguished between simple and complex types. The following case illustrates our learning process related to issues of screening, management, and relapse that can occur. At initial evaluation, this 39-year-old woman reported a 4-year history of chronic daily headache, transformed from episodic migraine at age 18 years. She reported using 10 butalbital with codeine and acetaminophen tablets per day for pain control for the past year. She described a history of marital conflict, where her husband continually pushed her to function regardless of pain, which she described as a primary reason for her medication overuse. She was home-schooling her 4 daughters for religious reasons, and identified her faith as a major force in her life. When she was in acute pain, her children's education was on hold. Psychiatric comorbidity included one episode of major depression in remission, insomnia, a history of recreational drug use in adolescence (in sustained remission), and histrionic (but no apparent antisocial) personality traits. She relied on passive coping techniques, and said she had “no time for relaxation.”

She was first admitted to our inpatient unit in 1987, withdrawn from butalbital and codeine, placed on medical prophylaxis, and coached in relaxation techniques. It was only after discharge that we became aware of the extent of her previous deviant drug use. She was arrested for calling in her butalbital and codeine using the DEA of an “expendable” physician, whom she had consulted in the past but did not intend to see again. Convicted of a misdemeanor and fined $300, she pursued no-cost substance abuse counseling through her local church, in addition to intermittent cognitive-behavioral therapy at the time of return medical visits to our center. We had her sign a treatment contract, agreeing to the therapeutic plan and that she would only take headache or pain medication prescribed by our center. She eventually returned to her local physician.

Several years later (1994), she entered our inpatient unit for the second time, after escalating use of butorphanol to a daily basis. In 1995, she was readmitted after escalating cafergot. In late 1996, she was admitted again for withdrawal from alternating daily use of either oxycodone with acetaminophen or butalbital with codeine. At her first follow-up appointment in 1997, she volunteered how good she felt “working through the guilt” about the past falsified butalbital prescriptions from when we had first seen her 10 years ago. Then, at her next appointment 2 weeks later, she stated “I'm not sure you are going to want to see me anymore”– she had just been arrested by a state trooper for filing 41 false butorphanol prescriptions in 1996, and charged with 5 felonies. She added, “I would have told you (about the butorphanol), but no one asked.” She appeared contrite, and expressed a desire to remain in treatment at our center.

At that point, if not earlier in the game, we might have been justified in discontinuing her from our practice. However, after consultation with all members of the treatment team, we agreed to continue seeing her with a new and tougher treatment contract, requiring weekly psychotherapy at our center (despite her distance), compliance with a formal substance abuse (SA) program with observed urine drug screens that was required as part of her probation, and total proscription against any use of opioids or controlled substances for pain. She participated, but continued to express the belief that “I'm not like the recreational drug users” in her program.

She experienced significant episodes of headache without effective abortive treatment. Despite the contract and treatment program, she asked if we could – just once – prescribe an opioid so that she could attend her SA program graduation pain-free! The answer was “no.” After graduation, she missed several aftercare SA sessions due to headache, and was expelled from the SA program.

Nevertheless, over the next 36 months, she had clean urine drug screens at her monthly meetings with her probation officer, maintained consistent attendance at psychotherapy, made no further requests for opioids or butalbital, and required no further inpatient treatment despite family conflict, divorce, and significant financial problems. Cognitive-behavioral therapy focused on compassionate confrontation on the moral consequences of her past behavior. While much of relaxation-based therapies focus on reducing emotional distress, effective confrontation often requires creating enough emotional discomfort in a trusting relationship that the patient recognizes there is a significant problem with her behavior. In fact, this patient later identified our relationship as a primary reason she stuck with the therapy when her headaches were severe and she desperately wanted opioids. Pain management skills focused on the concept of developing pain tolerance, as opposed to her previous focus on pain relief, sedation, and escape.

When asked how she justified her illegal prescriptions with her deep-seated religious beliefs, she said “I just did it – I didn't think about it.” The availability of opioids or butalbital to a patient with a history of overuse, who is in pain or distressed, is a Siren call so seductive that the patient may catastrophically jump overboard despite the knowledge that disaster lies below. To avoid relapse into misuse of medications, the patient typically needs a proactive plan for restraint. Homer's Odyssey provides a metaphor for just such a plan. In order to avoid certain death but still allow himself to hear the beckoning beauty of the Sirens’ song – which inevitably led sailors to jump ship and perish – Odysseus had his men put wax in their ears, then tie him to the mast and contract not to release him no matter how hard he struggled, until they were well past the auditory range of his tempters.

The Figure depicts our patient's mean severe headache index over the course of 3 years of medication overuse, followed by 3 years of freedom from opioids and butalbital. Note that from 1997 to 1999 this patient experienced debilitating headaches, sometimes of several days’ duration, for which her available abortive medications were ineffective. However, the big picture of overall severe headache activity showed dramatic improvement. Therapy underscored the importance of looking at this big picture rather than judge her progress based on any individual headache, no matter how incapacitating. At the end of 1999, she was pleased with her headache control, and transferred care to her local doctor.

Figure Figure.—.

Migraine and Analgesic Abuse Severe Headache Index* (frequency/month × duration in hours)/28 days. *Medical record data (form completed each return visit). Mean (standard deviation).


Relapse – defined as the return to a pattern of acute medication overuse as defined by ICHD-II for 3 months or longer – can be a significant problem, as high as 71% for analgesics at a 4-year follow-up by Katsarava et al, vs 27% for ergots and 21% for triptans.16 In this study, the combined relapse rate for all acute medications with rebound potential was 45%. Grazzi et al found a comparable relapse rate of 42% for patients with MOH who had undergone inpatient drug withdrawal with the addition of prophylactic medication when they were re-assessed at a 3-year follow-up.17 However, for patients randomly assigned to additional coping skills training (biofeedback) after drug withdrawal, the rate of relapse dropped significantly to 12.5%. In the Grazzi study, biofeedback patients not only had fewer headaches at follow-up, but also managed the majority of their headaches with significantly less reliance on abortive medication.

Despite 3 years of freedom from MOH, this patient returned to our center in 2002. She had remarried her husband, developed fibromyalgia, and began consulting with a specialist who prescribed hydrocodone for her body pain, which she was using 3-4 days per week. Her headache returned to a more frequent pattern, consistent with data showing that use of opioids for nonheadache medical conditions can lead to headache exacerbation and interfere with effective prophylaxis.9,10 Treatment again involved discontinuation of her opioids in consultation with the other prescribing physician, followed by improvement. She returned again to her local physician.

She resurfaced again in 2006, after residential substance abuse treatment. Her mother had received home hospice care during a terminal illness. The patient had volunteered to be the sole nighttime caretaker, and only admitted to taking her mother's liquid morphine after she collapsed at the bedside following her mother's death. Toxicology at the hospital revealed morphine in her system. She divorced her husband a second time, and is now in the midst of addressing her estrangement from her dysfunctional family of origin – initially blaming them, with difficulty acknowledging her own behavior. She attends Narcotics Anonymous several times a week, and has restarted weekly psychotherapy at our center.


Complex medication overuse (Type II) requires both medical and behavioral intervention over a prolonged course of time. Patients who appear to be doing well may remain vulnerable to relapse when confronted with new medical problems or major stressful life events. There exists considerable pressure from multiple sources to treat pain as a “vital sign,” leading some caring physicians to prescribe frequent opioids for nonheadache pain that can significantly complicate headache outcome. The health-care provider retains the right to discharge a patient due to unacceptable behavior or failure to adhere to the treatment contract, which our center has performed on a number of occasions. On the other hand, relapse is not uncommon in MOH, and should not be the sole basis for terminating treatment. As in the treatment of chemical dependency in general, recovery from MOH can be a lifelong process, not a state once achieved, forever maintained.