Because chronic pain often has multiple causes or contributing factors, it often does not respond to either purely somatic or purely psychological modes of treatment. Persistent pain can set a vicious cycle of reinforcing features into motion that then becomes a self-perpetuating problem independent of the initiating illness or injury. Deactivation, depression, disuse, medication misuse, and vocational dysfunction are all common contributing factors to the suffering and disability associated with chronic pain. Although simpler cases of chronic pain may respond to an approach based on the biomedical model, this is not true for the extremely disabled or prolonged cases likely to be referred to psychiatrists.

Patients with disabling chronic pain are prone to “doctor shopping,” in which medications and procedures are obtained from several physicians without the knowledge of the other physicians. A patient with chronic pain who has not formed a solid and honest therapeutic alliance with his or her treating physician cannot be successfully treated. When evidence of doctor shopping becomes apparent, the patient should be confronted immediately and a conversation begun about the doctor-patient relationship. If the patient does not agree to stop unannounced visits to other physicians, he or she should be dismissed from care because it is impossible to provide appropriate care in this situation.

The needs of patients with disabling chronic pain often outstrip the resources of the most enlightened and eager primary care physician. These patients are most appropriately treated by a multidisciplinary team experienced in the treatment of chronic pain. Members of this team may include a psychiatrist, psychologist, physiatrist, neurologist, neurosurgeon, physical therapist, occupational therapist, nurse, and vocational counselor. Each patient will not require the expertise of all of these disciplines, but all of these disciplines have expertise relevant to the management of chronic pain.

Chronic pain patients frequently present to physicians with the request for diagnosis and cure, but this is generally not possible. The patient generally has no broken part that can be fixed. Repeated attempts to find and repair the broken part in the chronic pain patient can lead to permanent iatrogenic injury rather than cure. Clinicians at the Multidisciplinary Pain Center at the University of Washington treat many patients who have had multiple back operations. For example, one patient had 54 procedures on his back for the relief of pain. One of the important, and undervalued, goals of chronic pain treatment is to protect patients from further iatrogenic injury at the hands of well-meaning but overly aggressive physicians.

One of the first steps in the successful treatment of chronic pain is to stop the search for diagnosis and cure in favor of an attempt at rehabilitation. In this attempt, increase in functional capacity must take precedence over pain relief. Attempts to provide immediate pain relief for the chronic pain patient will almost always have negative long-term physiological or psychological consequences. It is essential to remind the desperate patient with chronic pain that he or she has a chronic problem that must be managed with the next month and the next year in view. It is also important to educate the patient that his or her problem does not have a single cause or a single cure. Although an injury may have initiated a chronic pain problem, many of the effects of pain, such as inactivity, have now become causes of the pain’s perpetuation. Treatment succeeds more often if it targets these multiple factors. Because there is no “magic bullet” for chronic pain, a “shotgun approach” often must be used.

Pharmacological Management of Chronic Pain

Opioid Analgesics
Few psychiatrists will be prescribing opioid analgesic medications for their patients. But patients with chronic pain often will be receiving these medications from other providers, so it is important that psychiatrists are aware of the properties of these medications. Opioid analgesics have a primary role in the management of moderate to severe postoperative and other acute pain states and cancer-related pain. Opioid use in chronic non-cancer-related pain is the subject of current controversy. Traditional concerns have included addiction, drug dependence, neuropsychological impairment, and functional decline. Several investigators have reported that opioid medications can be prescribed safely for patients with non-cancer-related pain; it has also been suggested that physicians may be unnecessarily withholding opioid analgesics from patients whose pain cannot be successfully managed without them.

The American Pain Society and the American Academy of Pain Medicine have recently issued a consensus statement, “The Use of Opioids for the Treatment of Chronic Pain”. In this statement, the authors noted that several of the dangers thought to be associated with chronic opioid use, including addiction, respiratory depression, analgesic tolerance, and diversion of controlled substances, may not be as significant as has been assumed. They emphasized that policy in this area is evolving but did propose some principles of good medical practice in this area. The prescribing physician should do a comprehensive evaluation of the patient, including pain history and effect, directed physical examination and review of diagnostic studies, a review of previous interventions, a drug use history, and an assessment of comorbid medical and psychiatric conditions. A treatment plan should be formulated, generally incorporating multiple treatment modalities. An opioid contract is often useful. Special caution is urged for patients with histories of substance abuse. Periodic review and documentation of treatment efficacy are essential. Reports of not only pain intensity but also daily functioning and other domains of quality of life should be included.

Critical questions remain about how to assess the effectiveness of chronic opioid use in the non-cancer-related pain population. Recent randomized trials generally have shown positive effects of opioids on pain and mood but no effect on function or sleep. Opioid abuse has not been a significant problem in these trials. These results raise important questions. Should the patient or the physician be the ultimate decision maker about whether the medication is working? How should effects on symptoms and function be balanced in an overall assessment of the effects of opioids on quality of life? Future research and debate on these issues are necessary to define the best clinical practices in this area.

For now, clinical consensus supports the following practices. First, opioids should be used as a treatment of last resort. A systematic investigation of other physical, pharmacological, psychological, and behavior therapies with fewer long-term risks should be done first. Second, opioids (and other analgesics) generally should be given on a fixed schedule rather than on an as-needed basis. Among the advantages of a fixed schedule are better analgesic effects and the elimination of problems related to as-needed dosing. The patient is not required to decide whether the pain is sufficient to take the medication, whether to take it as the pain is beginning or to wait until it becomes unbearable, or whether to delay taking it so that it will remain available. These decisions tend to focus the patient’s mind on the pain rather than away from it, the opposite of what is desired in the management of most cases of chronic pain. Third, oral opioids with a long duration of action are generally preferred over shorter-acting formulations and parenteral formulations. These provide more consistent analgesia and are less prone to abuse. Fourth, periodic review is essential. Every 3 months or so, effects on symptoms, function, cognition, and quality of life should be monitored. Any dose escalation should prompt a thorough in-person review.

In recent years, tramadol has been heavily marketed for use in chronic pain. It is a novel medication that is part opioid (a weak agonist at the μ opioid receptor) and part antidepressant-like drug (a serotonin and norepinephrine reuptake inhibitor). Whether tramadol offers advantages over the much cheaper codeine-acetaminophen or hydrocodone-acetaminophen combinations is unclear. Contrary to initial reports, tramadol does appear to cause some problems with dependence, withdrawal, and dose escalation. These problems may be limited to those with previous experience with opioids. Until further research is done, caution is advised in using tramadol in patients at risk for opioid abuse.

Opioid detoxification Multiple protocols for detoxifying patients from opioid analgesics are available. The simplest method is gradually tapering the dosage of the opioid being used by placing the patient on a fixed schedule of medication, followed by a reduction in dosage. Several factors should be considered in determining the appropriate schedule: the length of time the patient has been taking the medication, the dosage level, the patient’s desire for detoxification and willingness to cooperate with the process, and the degree of pain still being experienced. Generally, the dosage of the opioids can be safely reduced by 10%-20% each day with minimal risk of withdrawal.

When the patient is taking large dosages of an opioid, a substitution-detoxification schedule with methadone may be needed. In this case, the patient is switched from the opioid currently prescribed to an equianalgesic dosage of methadone, usually provided four times a day. To avoid the necessity of waking the patient during the night, an 8:00 a.m., 12:00 p.m., 5:00 p.m., and 9:00 p.m. schedule is often used. The methadone is then reduced by 10%-20% at daily, weekly, or monthly intervals. Patient compliance may be improved by providing the methadone in a “pain cocktail,” in which the medication is mixed with syrup so that the patient is unaware of the amount of medication being taken. Patients are informed of the general strategy but not of the specific dose being given. This helps dissociate physical from psychological withdrawal problems. This method is commonly provided by inpatient pain management programs and, with the help of cooperative pharmacies, in outpatient settings.

An alternative method of opioid detoxification is through the use of clonidine, an α₂ - adrenergic agonist. However, this method has been reported to be less effective than methadone. Although clonidine or dosage tapering can be used to detoxify a patient from a mixed agonist-antagonist opioid, the methadone withdrawal protocol should not be followed for patients taking this class of medications.

Although successful detoxification is sometimes thought to represent the end of concern about medication use, it is important that physicians address the reasons the opioid analgesic was initially required and why the patient continued to need it. Otherwise, the patient may simply seek care from another physician who may restart the medication, thereby resulting in a recurrence of the problem.

Nonsteroidal Anti-Inflammatory Drugs
The nonsteroidal anti-inflammatory drugs (NSAIDs) are efficacious for many acute and chronic pain conditions. They are the mainstays in the treatment of pain caused by inflammatory arthritic conditions. Little evidence indicates that any NSAID is more effective for pain than any other, although price, half-life, and dosing frequency vary considerably. Although this class of medications has been available for a long time, some new additions are worthy of note.

Ketorolac is an NSAID that does not show the efficacy “ceiling” typical of the NSAIDs. This is true, however, only when the drug is administered parenterally. In these cases, it has expanded the indications for NSAID therapy from mild-to-moderate to moderate-to-severe pain. This has broadened the use of this class of medications in postoperative pain and in migraine treatment.

All NSAIDs exert their therapeutic effects through the inhibition of cyclooxygenase. Recently, two medications that selectively inhibit cyclooxygenase-2 have been released. Celecoxib and rofecoxib produce fewer gastrointestinal side effects (including ulcers) than do the less selective NSAIDs. However, these drugs are no more potent analgesics than the traditional NSAIDs, are vastly more expensive, and have side effects of their own.

Acetaminophen has comparable analgesic potency to the NSAIDs but does not operate through inhibition of cyclooxygenase and does not inhibit inflammation. It does not produce ulcers, but it has hepatic toxicity in overdose or in combination with alcohol. In many noninflammatory conditions, including osteoarthritis, it appears to provide pain relief equal to the NSAIDs and may be the preferred drug.

Antidepressants
Antidepressants are frequently prescribed for patients with chronic pain. Three good reasons can be offered for these prescriptions: 1) patients with chronic pain have high rates of depressive disorders, 2) patients with chronic pain who do not meet diagnostic criteria for depression have high rates of sleep disturbance, and 3) evidence supports the analgesic effects of antidepressant medications. Although the analgesia provided by these medications was originally considered to be attributable to their antidepressant properties, it is now generally accepted that they exert separate analgesic effects in at least some chronic pain conditions. Thus, they also may be beneficial for patients whose pain is not accompanied by depression.

Antidepressant medication can effectively treat depression in the presence of chronic pain. With chronic pain, as with other chronic medical disorders, treatment may have to overcome some extra hurdles, including aversive physical symptoms, severe deactivation, vocational dysfunction, marital conflict, social isolation, and concurrent medications. Comprehensive assessment of these issues and formulation of a treatment plan that takes them into account will increase the likelihood of successful depression treatment in the chronic pain patient. When antidepressant treatment is suggested, patients often will respond that their depression will resolve when their pain decreases. This may be true; nonetheless, depression is often more accessible to treatment than an entrenched and multidetermined pain problem. If depression can be relieved, many other aspects of rehabilitation, such as physical therapy, are often much more easily accomplished. However, pain will usually, but not necessarily, subside with improvement in depressive symptoms.

All currently marketed antidepressants are equally effective for the treatment of depression. Whatever differences may exist among antidepressants in efficacy for neuropathic pain do not appear to affect their ability to treat depression. Given the lack of greater efficacy of any one antidepressant compared with another, the medication should be selected based on its side effects.

The clinical art of depression treatment for those with chronic pain consists of establishing a solid therapeutic alliance around the problem of depression and finding a medication regimen with a side-effect profile that the patient can tolerate. Because patients with chronic pain can be vigilant and catastrophizing about somatic symptoms, care must be taken to educate them about antidepressant side effects. Sometimes an antidepressant regimen must be initiated at the lower doses used for geriatric patients in order to ease habituation to side effects. Several case reports have described relief of chronic pain in patients with treatment-resistant depression with electroconvulsive therapy; however, no carefully controlled studies have shown the effectiveness of electroconvulsive therapy for the treatment of chronic pain.

The mechanism by which antidepressants provide analgesia is unclear. In the past, most theories focused on antidepressant effects on serotonin, but the efficacy of antidepressants with primarily noradrenergic actions (e.g., desipramine) indicates that the analgesia is probably not related to any one single neurotransmitter system. The interdependence of the opioid and nonopioid endogenous pain modulation systems has been suggested by studies that show enhanced opioid analgesia in the presence of antidepressant treatment (Gordon et al. 1993) and decreased opioid analgesia after serotonin and norepinephrine depletion. Therefore, biogenic amines appear to play a critical role in endogenous pain modulation. To the extent that there is depletion or impaired function of amines such as serotonin and norepinephrine in depression, this may contribute to the pain experienced and reported by those with major depression. It is not clear why this is not also manifested as increased sensitivity to external noxious stimuli.

Antidepressants have been reported to be efficacious for many types of pain. In their meta-analysis of 39 placebo-controlled studies, Onghena and Van Houdenhove found that antidepressant medications appear to be especially beneficial for neuropathic pain, headache (including migraines), and facial pain. Similar results were found in the review by Magni. The tricyclic antidepressants are now considered to be first-line drugs for the treatment of pain related to neuropathies, including postherpetic neuralgia and diabetic neuropathy. A recent meta-analysis suggested effectiveness in patients categorized as having psychogenic pain or somatoform pain disorder. Evidence for pain relief independent of depression relief in musculoskeletal pain syndromes such as low back pain has been weak, although a recent randomized trial showed a modest effect for nortriptyline.

The question of which antidepressant is most effective for pain is unclear. Many tricyclic antidepressants, including amitriptyline, nortriptyline, imipramine, desipramine, clomipramine, and doxepin, have been reported to have analgesic effects. Amitriptyline appears to be the most frequently used in clinical practice, but it is questionable whether amitriptyline provides more analgesia than the other tricyclic antidepressants. Although its anticholinergic and sedating actions may make it difficult for patients to tolerate at higher doses, its sedative effects may be sought for many patients with pain. Because sleep problems often accompany and can exacerbate chronic pain, sedating antidepressants are frequently more beneficial. In our clinical experience, nortriptyline (the desmethylated metabolite of amitriptyline) and desipramine have analgesic efficacy equivalent to that of amitriptyline and are often better tolerated by patients. Although studies suggest that nontricyclic antidepressants, including trazodone and the monoamine oxidase inhibitors, may also provide analgesia, the evidence is not as strong as that supporting the use of the tricyclic antidepressants. A recently completed study of patients with chronic low back pain showed that maprotiline, which operates primarily through blocking norepinephrine uptake, was superior to paroxetine and placebo for pain relief.

The selective serotonin reuptake inhibitors may prove to be beneficial for certain pain conditions. However, Max et al. reported that fluoxetine appeared to be no more effective than placebo for pain related to diabetic neuropathy and that amitriptyline and desipramine were more efficacious. Furthermore, fluoxetine exerted an analgesic effect only in those patients who had comorbid depression, whereas the tricyclic antidepressants reduced pain even in the absence of depression. Max hypothesized that norepinephrine reuptake inhibition may be critical for antidepressant analgesia in neuropathic pain. In a recent review of the published English literature on the use of selective serotonin reuptake inhibitors for the treatment of chronic pain, 12 studies were identified, including 6 on the treatment of headache, 3 on diabetic neuropathy, and 3 on fibromyalgia. Two controlled trials suggested that selective serotonin reuptake inhibitors may be useful in patients with chronic tension headache, but controlled trial results for migraine headache, diabetic neuropathy, and fibromyalgia are mixed. For patients with these conditions, it is reasonable to reserve selective serotonin reuptake inhibitors for those who have concurrent major depression, who fail to respond to other medications, or who are intolerant of their side effects. Venlafaxine, a new antidepressant with both serotonergic and noradrenergic reuptake inhibition, lacks the troublesome anticholinergic and antiadrenergic actions of the tricyclic antidepressants. Although, at present, only case reports support venlafaxine’s use in chronic pain, a number of clinical trials testing venlafaxine in chronic pain syndromes are under way, and a recently completed migraine prophylaxis trial reported promising results.

The optimal dose of antidepressant medications for analgesia is the subject of controversy. The tricyclic antidepressants can exert their analgesic effects at lower dosage levels than are required for their antidepressant actions. Studies that have examined the benefits of the selective serotonin reuptake inhibitors for pain have generally used dosages equivalent to those used for depression. There is a great deal of individual patient variability regarding the dosage of the antidepressant medications that will provide maximum relief and can be tolerated. To some degree, analgesia appears to be dosage related, although it is unclear whether there is a correlation between drug plasma concentrations and pain relief. For many patients, the side effects that accompany these medications, especially the tricyclics, may be the determining factor in dosing. The best recommendation when using tricyclic antidepressants as analgesics is to initiate treatment at relatively low dosages and then gradually increase the dosage until pain relief is achieved or dose-limiting side effects, most often daytime sedation, develop. A usual starting dosage of amitriptyline would be 25 mg 1-2 hours before bedtime or 10 mg in a debilitated patient, with gradual increases of 10-25 mg every second to third day until side effects prevent the use of larger dosages. If analgesia is not achieved or if patients are unable to tolerate the side effects from one antidepressant, an alternative should be tried. Tricyclics used for analgesia, as for depression, generally can be given once a day.

The current literature provides limited suggestions regarding the duration of treatment with antidepressants for chronic pain. The long-term side-effect profiles of both the tricyclics and the selective serotonin reuptake inhibitors are generally relatively mild, especially when compared with those of other medications commonly used for chronic pain, including the NSAIDs. One of the great benefits of the antidepressants is that tolerance does not develop with prolonged use. Therefore, if they are providing effective analgesia, they should be among the last medications to be discontinued. Continued use of these medications should be reassessed every few months, and, as with all treatments for chronic pain, both functioning and levels of pain should be examined. Only after the patient’s pain and functional capacity have stabilized should tapering and discontinuation of these medications be considered.

Benzodiazepines
Benzodiazepines can be effective adjuncts in the management of acute pain conditions because of their anxiolytic and muscle relaxant properties. However, benzodiazepines show scant evidence of analgesic effects, and they may antagonize opioid analgesia. Chronic pain is frequently associated with insomnia and anxiety. Therefore, patients commonly receive benzodiazepines or other sedatives (such as the “muscle relaxers”). Some patients start these medications during the acute phase of their pain problem and then continue to take them for many months or even years. It is very important for chronic pain patients taking benzodiazepines to be assessed or depression. These medications will mask some symptoms of depression (e.g., initial insomnia, agitation), but they are not adequate treatments for depression. Indeed, dangerous levels of depression can develop under the cover of benzodiazepines. It has been suggested that benzodiazepines can induce depression with chronic use, but the evidence for this is not strong. More important is the potential for masking of depression by benzodiazepines.

Nearly all patients with chronic pain should be tapered off benzodiazepines. Chronic benzodiazepines are the treatment of choice for very few conditions. The treatment of choice for chronic anxiety disorders in this population is antidepressant medication. Buspirone (a serotonin type 1A receptor partial agonist) is marketed as an anxiolytic but is more similar to the antidepressants in its pharmacology and side-effect profile. It is a reasonable alternative to the benzodiazepines for the treatment of chronic anxiety, particularly for those who experience agitation while taking antidepressants.

Anticonvulsants
Anticonvulsants, including phenytoin, carbamazepine, and sodium valproate, have been found to be efficacious for neuropathic pain. The anticonvulsants are less stigmatized than the antidepressants and may be chosen first. But for pain associated with peripheral neuropathy, the “number needed to treat” to achieve adequate pain relief is probably lower for the antidepressants. Therefore, for conditions such as diabetic neuropathy, the antidepressants are preferred.

Carbamazepine is considered the treatment of choice for pain related to trigeminal neuralgia. Other syndromes for which anticonvulsants have been used include thalamic pain, diabetic neuropathy, and postherpetic neuralgia. Valproate has a role in migraine prophylaxis. However, because antidepressant medications also may be efficacious for treating these forms of pain and are less likely to cause significant side effects, they are usually tried before the anticonvulsants. Few dose-response studies have investigated the analgesic properties of the anticonvulsants. But clinical experience suggests that the required dosage of the anticonvulsants, when used for analgesia, is similar to that when they are used for seizures.

A new anticonvulsant of low toxicity proven to be effective in neuropathic pain is gabapentin. It has a benign side-effect profile and can be tolerated by most patients in large doses (3-4 g/day). It is being investigated for a range of psychiatric uses, including bipolar disorder, panic disorder, and other forms of affect dysregulation. One recent randomized trial found it to be equally effective to amitriptyline in diabetic neuropathy although much more expensive.

Additional Medications Used for Pain Management
A wide variety of other medications have been shown to be useful for neuropathic pain. Among the antihypertensive medications, the α₂ antagonist clonidine has been effective for peripheral neuropathic pain and pain associated with spinal cord injury. Our clinical experience with oral and transdermal preparations have been disappointing, however, and it may be that clonidine is most effective when administered epidurally or intrathecally. The antiarrhythmic drug mexiletine is also effective in neuropathic pain. Clinical response to this oral analogue of lidocaine can be fairly well predicted by a favorable response to an intravenous lidocaine infusion. Capsaicin, a topical cream, works by depleting substance P, a neuropeptide involved in the transmission of pain. Because it is applied externally to painful regions, capsaicin has the additional benefit of being without any apparent systemic side effects. It can be quite painful to apply initially and has not had good long-term results in our clinical experience.

Ergot derivatives have been used for many years to treat migraine headaches. These have been largely replaced in recent years by the “triptans,” which are selective agonists at serotonin type 1-like receptors, including serotonin type 1B/1D receptors . These are effective treatments for acute migraine attacks, and the injectable sumatriptan form has also shown efficacy in the treatment of cluster headaches. Headache recurred in 21%-57% of the patients who received oral or subcutaneous sumatriptan, but most patients responded to a second dose of the drug. Response rates achieved after oral sumatriptan were similar to those reported after treatment with oral naratriptan, rizatriptan, or lysine acetylsalicylate plus metoclopramide. Sumatriptan is generally well tolerated. Nausea, vomiting, malaise, and fatigue are the most common adverse events with oral sumatriptan. The “triptans” are all quite expensive but can be indicated for those with frequent migraine attacks and significant disability or health care use. Recurrent or rebound headaches can become a problem for those patients with daily or near daily headaches. In these patients, headache prophylaxis should take precedence over treatment after the headache has started.