| Contents | Previous | Next |
Most children with cancer experience pain (McGrath, 1990b), which, as does the progression of cancer in children, differs from that of adults. After diagnosis, the common childhood malignancies generally respond rapidly to treatment, and disease-related pain often remits. If the tumor recurs and is resistant to treatment, the disease progresses rapidly, resulting in early death (Miser, 1993).
Pain in children with cancer arises more often from the treatment than from the disease (Miser, Dothage, Wesley, et al, 1987). Aggressive multimodal treatment protocols for children have increased survival rates markedly for most types of cancer, but they often involve treatment toxicity that results in painful conditions, e.g., mucositis, peripheral neuropathy, and infection.
Children with cancer undergo procedures ranging from venipunctures to bone marrow aspirations and biopsies. Children with aggressive treatment protocols may have one or more venipunctures daily, lumbar punctures weekly, and bone marrow aspirations monthly. Unlike adults, infants and children do not provide consent for these procedures and often do not understand the reasons for them or realize their short duration. Although appropriate preparation and adequate analgesia are crucial for children undergoing procedures, often neither occurs or they occur in a haphazard fashion (Schechter, 1989).
The optimal treatment of a child's cancer-related pain requires an awareness of the many factors that shape that pain. Among these are the child's developmental level, emotional and cognitive state, personality traits, physical condition, and past experiences; the meaning of the pain for the child; the stage of the disease; the child's fears and concerns about illness and death; issues, attitudes, and reactions of the family; cultural background; and the environment (Hester, Foster, and Beyer, 1992). Clinicians should be aware that children with cancer experience many distressing symptoms such as pain, depression, anxiety, panic, pruritus, fatigue, nausea, constipation, insomnia, dyspnea, and the fear of abandonment and death.
Getting to know the child and having knowledge of developmental norms and behavioral competencies are important in the assessment and management of pain. Clinicians should tailor assessment and management strategies to the child's developmental level, personality style, and emotional and physical resources and to the context; tailoring is particularly necessary for children with developmental delays, learning disabilities, emotional disturbances, and language barriers.
Assessment is not only diagnostic but also therapeutic. Assessing the meaning of the pain to the child and the family, the effect of the pain on the activities of daily living and on mood, and the concurrent concerns and symptoms helps clinicians understand pain from the perspective of the child and the family. Asking about pain underscores the clinician's desire to ease pain and suffering and builds a therapeutic alliance with the child and family.
It is easier for clinicians to understand inherently subjective experiences, such as pain, anxiety, and despair, when the child can verbalize, but for some children, verbal communication is difficult or impossible. Therefore, the clinician should recognize the potential for pain and discomfort or suspect that the child is in pain even if the signs are not immediately apparent.
Whether the child is at home, in a clinic or clinician's office, or at the hospital, assessment of pain should occur during the course of his or her illness. Critical to assessment is open communication about pain among the child, the family, and the health care team.
An initial pain history focuses on understanding pain from the child's and family's perspective. Central to communicating with the child about pain is determining the language the child uses for pain (e.g., hurt, owie, boo-boo) and how and to whom the child communicates pain. Other issues include past pain experiences, the child's response to pain, expectations related to pain, and preferences for assessing and treating pain (see Attachment B). Clinicians should integrate this background information into subsequent assessments and treatment plans for the child.
A routine assessment of pain is critical to ongoing management. The frequency of assessment should be tailored to the severity of the pain, the context, and the preferences of the child and family. Frequent assessments are necessary when pain is being poorly managed or is not responding to the current treatment. Documentation of pain ratings on a chart or flowsheet, located in a visible place such as at the bedside, provides easy access for providers. The use of a flowsheet reduces the possibility of redundant questioning which can be overwhelming for the child and family and may interfere with the child's coping skills.
Assessment of children's pain involves one or more of the following approaches: self-reports, proxy reports, observations, and physiologic measures. "Because physiologic indicators such as heart rate, respiratory rate, blood pressure, and diaphoresis alter with a variety of stress-arousal events, they should not be used as measures of pain in the absence of other pain assessment methods or clinical indicators" (McGrath, de Veber, and Heam, 1985). A variety of assessment methods are available, but no one approach provides a complete picture of the pain experience. At least one method that is reliable, valid, and developmentally appropriate to the child should be used regularly for assessing pain. Because children with cancer may need assessment in settings other than hospitals (e.g., their homes), the methods should be affordable and easy for parents or other caregivers to use.
Self-Report. Self-report methods provide the most reliable and valid estimates of pain intensity and location. These methods are appropriate for children over the age of 4 who can verbalize (McGrath, 1990b). Rarely will children with cancer fabricate pain (Ross and Ross, 1988), but they may deny or underreport pain if they (1) fear that admission of pain will mean further painful procedures or treatments such as "a shot for pain"; (2) lack awareness that pain can be treated; (3) wish to protect parents from the reality of progressive disease; or (4) desire to please and placate others.
Self-report methods should be easy to administer with simple instructions for children. They should allow both verbal and nonverbal (e.g., pointing) responses. Often, children will not respond to questions verbally, especially if they are anxious or depressed or are experiencing severe pain.
Several self-report methods for pain intensity are available for use with children (see Attachment B). Although the psychometric adequacy of these methods for children with cancer has yet to be determined, reliability and validity estimates are available for other pain syndromes such as postoperative pain, procedural pain, and juvenile rheumatoid arthritis. Methods appropriate for children over the age of 4 years include the Oucher (Beyer, Villarruel, and Denyes, 1993) and the Poker Chip Tool (Hester, Foster, and Kristensen, et al., 1989). Some investigators have used cartoon faces as scales of measurement for young children with cancer who are undergoing procedures, but the construct being measured was not necessarily pain. One scale measures pain affect (McGrath, de Veber, and Heam, 1985), whereas others measure intensity of pain, anxiety, or distress (Adams, 1990; LeBaron and Zeltzer, 1984). Children over the age of 7 years who understand the concepts of order and number may prefer a numerical rating scale (NRS) (McGrath and Unruh, 1987), a horizontal word graphic rating scale (Savedra, Tesler, Holzemer, et al., 1989 [updated 1992]), or a VAS (McGrath, 1990b). A large study that included children and adolescents reported that the VAS was the least preferred of five horizontal pain scales (Tesler, Savedra, Holzemer, et al., 1991).
To determine the location of pain, children can be asked either to point to their body or use a body map (i.e., an outline). Children over the age of 4 can use crayons or colored markers to locate pain on a body map (Eland, 1989; Savedra, Tesler, Holzemer, et al., 1989 [updated 1992])(see Attachment B). The precision of the location will increase with the child's age. Children who are suffering may regress;similarly, children who are developmentally delayed or learning disabled may need assessment tools developed for younger children. If a child is unable or unwilling to provide pain ratings, parents or health care professionals can provide proxy reports. Proxy ratings, however, are inexact.
Behavioral Observation. Behavioral observation is the primary assessment approach for preverbal and nonverbal children and is an adjunct to assessment for verbal children. Observations focus on vocalizations (e.g., crying, whining, or groaning), verbalizations, facial expressions, muscle tension and rigidity, ability to be consoled, guarding of body parts, temperament, activity, and general appearance. Adequate reliability and validity documentation is lacking for behavioral observations; consequently, most such observations offer only a second-best approximation of the child's experience, even though clinicians often attribute greater importance to nonverbal expression than to self-report (Craig, 1992). Changes in how a child looks and acts may indicate the onset of pain or its increase (Hester and Foster, 1990) and warrant further investigation and documentation.
Observations are problematic in that the stimulus for behaviors or changes is not always clear. For example, children cry in response to pain, as well as fear, loneliness, and overstimulation. Clinicians may misinterpret behaviors such as sleeping, watching television, and using humor as the absence of pain when, in fact, the child is attempting to control pain. Moreover, behavioral responses may be absent or attenuated when vocalizations or movements cause or increase pain. Infants may become apathetic after only a few days of continuing severe pain, and suffering experienced by older children and adolescents with cancer may blunt behaviors and affect. Other factors that inhibit behavioral responses include intubation, use of paralyzing agents or sedatives, extreme illness, weakness, or depression. Therefore, the use of behavioral observation to guide analgesia requires close attention to the context. If caretakers are not sure whether a behavior indicates pain and if there is reason to suspect the presence of pain, a trial of analgesics can be diagnostic as well as therapeutic.
Most of the scales developed for measuring behaviors address postoperative pain or pain associated with invasive procedures (e.g., LeBaron and Zeltzer, 1984). Given the nature of cancer-related pain, behavioral scales for the assessment of acute pain problems are unlikely to be sensitive in assessing the child with cancer pain. The Gustave-Roussy Child Pain Scale (Gauvain-Piquard, Rodary, Rezvani, et al., 1987) is the only observation tool developed for children with cancer pain.
Pain is managed within a therapeutic alliance among the child, his or her parent(s), nurses, physicians, and other health care professionals. The beliefs and preferences of the child and family should be elicited, respected, and carefully considered. At the same time, the primary obligation of the health care professional is to ensure safe and competent care. The presence of divergent beliefs and goals among members of the team can interfere with effective pain and symptom management, but these can often be resolved through discussion and negotiation.
Medical Interventions. Medical interventions include analgesics, adjuvant agents (e.g., corticosteroids, tricyclic antidepressants, stimulants), palliative chemotherapy, radiation therapy, regional analgesia, and neurosurgical approaches. In most cases, analgesics—either alone or supplemented with chemotherapeutic agents, radiation therapy, and adjuvants—provide adequate pain relief. Regional analgesia is occasionally helpful.
Analgesics and Adjuvants. Health care professionals treating children often use medications that have not been specifically tested in children and that are thus not specifically indicated for pediatric use. This situation exists because children as a group are therapeutic "orphans"; the small number of children needing certain medications does not provide incentive for widescale testing. The administration of analgesics to children should follow the WHO'S ladder approach (see Chapter 3). Usual doses for children are listed in Tables 8 and 10.
Acetaminophen and NSAIDs. Acetamenophen is a useful and relatively safe analgesic that provides effective relief of mild pain and enhances opioid effectiveness (Tobias, 1992). The rectal route is available for children who cannot take me'dication orally; however, children do not like this route and may refuse to take the medication. Rectal administration is contraindicated for children who are neutropenic or thrombocytopenic and for those with mucositis. These contraindications and the irregular absorption of the rectal route limit its usefulness in treating severe pain (Miser and Miser, 1989).
Because children with cancer are often thrombocytopenic, NSAIDs frequently cannot be used. They do, however, provide excellent analgesia for the child who is not at risk for bleeding because of thrombocytopenia or a coagulopathy and who is not at unusual risk of gastritis or ulceration (e.g., from the concurrent use of corticosteroids). NSAIDs effects on inflammation can be salutary, especially in the presence of bone pain. Evidence suggests that NSAIDs such as choline magnesium trisalicylate and salsalate have little effect on platelet function in normal subjects not at risk for bleeding (Cronin, Edmiston, and Griffin, 1991). Even so, the use of these "platelet-sparing" NSAIDs in children at risk for bleeding is not recommended.
The administration of acetaminophen and NSAIDs varies according to the severity of the pain (Table 9). For very mild or intermittent pain, as-needed administration is appropriate. However, hospitalized children and their parents should be told to ask for the medication if pain occurs. An advocate may be appointed to assist the child and parents in requesting medications. For continuous or more severe pain, around-the-clock administration is necessary (Table 9).
Opioid analgesics. For moderate to severe pain, opioid analgesics are recommended. Some health care professionals are concerned about the potential for addiction in children, a risk that is rare in adult cancer patients (Kanner and Foley, 1981). Although studies of the risks in children are lacking, no known aspect of childhood development or physiology increases the risk of physiologic or psychological vulnerability to chemical dependence. Therefore, withholding opioids from children with cancer is unwarranted.
Route of administration. Whenever possible, opioids should be administered orally. Many are available in a liquid form or suspension: often, those that are not commercially available in this form can be pulverized (with the exception of controlled-release drugs) and administered in a small amount of liquid or soft food.
Parenteral administration is indicated when:
Even with severe pain, once the opioid dose requirement is ascertained, the route can be changed over a few days from the intravenous to the oral route by using equianalgesic conversions (Table 11).
When parenteral administration is required, the intravenous or subcutaneous route can be used (Miser, Moore, Greene, et al., 1986). Some children with cancer have implanted central catheters, which can be used; similarly, children who cannot take medication by mouth often have an intravenous catheter for the administration of fluids, chemotherapy, or antibiotics. Intramuscular injections should not be used, because "shots" are painful and frightening to children.
Transdermal fentanyl provides an alternative route for children with relatively constant pain who require larger doses of opioid analgesia. It is unsuitable for rapid-dose titration or for any patient with changing pain intensity. Currently available preparations do not permit the delivery of drug at dosage rates low enough for young children.
Schedule and dose. Severe pain is an emergency, requiring the rapid titration of analgesia to provide relief within a few hours. Rapid titration involves frequent assessments and dosage adjustments. For example, because the peak effect of intravenously administered morphine occurs about 15 minutes after administration, the patient whose dose requirement is unknown could be given 0.1 mg/kg of morphine and be assessed every 15 minutes, with additional increments of 0.05 mg/kg administered until relief is obtained.
Opioids can be administered by intermittent bolus injections, infusions, or infusions with "rescues." Unless the pain is truly intermittent and unpredictable, as-needed administration should not be used since delay between the request for medication and the administration results in poor pain management. In those isolated cases when as-needed administration is indicated and when "rescue" doses of medication are added to an infusion, the child needs an advocate (e.g., nurse, parent) to ensure prompt administration.
Intermittent bolus injections of morphine or its analgesic equivalent can be provided on an around-the-clock basis at a starting dose of about 0.1 mg/kg. Initial dose intervals are the same as those recommended for adults. Continuous infusion of morphine, at a starting dose of 0.02 to 0.04 mg/kg per hour for children over 6 months of age has been well studied in postoperative pain and described in cancer-related pain (Hendrickson, Myre, Johnson, et al., 1990; Miser, Moore, Greene, et al., 1986).
Continuous infusion avoids the extreme variations that may occur with intermittent intravenous doses and is indicated when intermittent doses:
Children receiving a continuous infusion should be offered "rescue" doses for breakthrough or poorly controlled pain regularly on the basis of their level of pain and the duration of the effect of the prescribed opioid.
Because of wide variability in opioid dose requirements (Nahata, Miser, Miser, et al., 1984), pain and side effects should be assessed frequently, with the dose and interval adjusted for optimal relief. Many children require large doses of opioids; the process for dose increase and titration to effect is the same as in adults (see Chapter 3). To titrate to effect when a continuous infusion with rescue doses is used, the total amount of opioid administered (including rescue doses) is calculated for a specific time period, usually 8 or 12 hours. This amount is then added to the total amount of opioid to be infused over the next 8- or 12-hour period. Because four or five half-lives are required to reach a new steady state, increasing the infusion when pain is poorly managed will not by itself provide adequate, immediate relief. In fact, it could result in the administration of an unnecessarily large dose as soon as the infusion reaches a new steady state.
Agent. Morphine is the preferred starting agent for severe pain. Codeine and oxycodone can be used for moderate pain, although morphine and hydromorphone may be better tolerated by some children. Opioids such as hydromorphone, methadone (Miser and Miser, 1986), and fentanyl may be preferable when side effects are not easily controlled. Methadone use requires careful titration and appreciation of the potential for delayed somnolence because of its long half-life. Meperidine should be used only in exceptional circumstances such as hypersensitivity to morphine and hydromorphone and when a single dose (e.g., for a procedure) or use for fewer than 2 days is anticipated.
Patient-controlled analgesia (PCA). PCA provides safe and effective analgesia in children old enough to understand the relationship among a stimulus (pain), a behavior (pushing the button), and a delayed response (pain relief). Most children over the age of 7 understand this concept, and sometimes even younger children can learn to use PCA, but some may not have the cognitive or emotional resources to use it. Only one postoperative pain study has focused on the effectiveness of PCA with and without a basal infusion in contrast to the effectiveness of intramuscular morphine (Berde, Lehn, Yee, et al., 1991). One study addressed the efficacy of PCA for adolescents (age 18 or older) with cancer in treating mucositis after bone marrow transplantation (Hill, Chapman, Komell, et al., 1990). Patient-controlled dose and basal infusion have not been explored systematically in children with cancer.
Monitoring. Regular assessment of a patient's vital signs and level of consciousness is necessary when parenteral opioids are used. Because of variable clinical situations and goals of treatment in children with cancer-related pain, professional judgment should be used to determine the presence, type, and frequency of monitoring. Because of wide interindividual and intraindividual variations in response, a child may have an adverse reaction, despite the most careful titration of doses and intervals.
Side effects. Young children may have difficulty communicating subjective symptoms like pruritus, nausea, constipation, and dysphoria; the preverbal child may show only generalized discomfort. If an infant or preverbal child becomes increasingly restless or irritable, despite an increased opioid dose, it is important to consider treatment of presumed side effects or a change to an alternative opioid. The pharmacologic approach to managing side effects in children is similar to that in adults. However, the assessment of side effects and pain relief should occur simultaneously.
One of the most feared side effects of opioid use is respiratory depression. In the dying patient, it may be due to the disease and not necessarily to the effects of opioids. In the patient who is not dying, naloxone may be indicated. If naloxone is used, it should be titrated incrementally (see Chapter 3) until the patient resumes adequate respiratory effort. The initial dose of naloxone in the child is about 0.5 to 2 u.g/kg, with the dose repeated about every minute. Physical stimulation, oxygen administration, and support of respiration can be used while titrating the naloxone to effect.
Adjuvants. Tricyclic antidepressants can be used as described for adults in Chapter 3. In general, the starting dose should be low, (e.g., about 0.2 mg/kg of amitriptyline), and then increased to about 1 to 2 nig/kg daily. Tricyclic antidepressants should be used with care in children who have received large doses of cardiotoxic anthracyclines. A baseline electrocardiogram may be useful but is not required.
Stimulants such as dextroamphetamine and methylphenidate can also be used for children, with the goal being to provide additional analgesia and increased quality of awake time. The starting dose for both stimulants, 0.05 mg/kg, is gradually increased to effect with an upper limit of roughly 0.25 mg/kg per dose. Stimulants are given at times of desired wakefulness, such as morning and midday.
Analgesics for Neonates and Young Infants. Acetaminophen can be safely administered to neonates and infants without concern for hepatotoxicity when given for short courses at the recommended doses (Berde, 1991). Acetaminophen can be given to augment analgesia.
The use of opioids in young infants requires special consideration and expertise. Young infants, especially premature babies or those who have neurologic abnormalities or pulmonary disease, are susceptible to apnea and respiratory depression when systemic opioids are used (Purcell-Jones, Dormon, and Sumner, 1987). The infant's metabolis altered so that the elimination half-life is longer and the blood-brain barrier is more permeable (Collins, Koren, Crean, et aI., 1985: Lynn and Slattery, 1987). Both factors result in young infants having high in-brain concentrations of opioids for a given dose than do mature infants or adults. Intensive monitoring is reasonable up to about 1yr ) of age for non ventilated infants who are receiving opioids because extreme sedation and decreased respiratory effort may be difficult to assess. Institutions where neonates and infants are treated for cancer should train personnel in the effective and safe administration of angesia and provide appropriate technologies for monitoring.
Some evidence suggests that the clearance of opioids increases rapidly over the first few weeks of life and approaches adult levels the time the infant is 1 to 2 months old (Hertzka, Gauntlett. Fisher aI., 1989; Koren, Butt, Chinyanga, et aI., 1985). Because available data are based on small numbers of infants, many practitioners reduce the initial dose and use intensive monitoring for infants up to 6 months age; this age is arbitrary and represents a cautious interpretation of the literature.
Although further research is necessary, apnea and respiratory depression appear to be dose related (Koren, Butt, Chinyanga, et. al. 1985). For non ventilated infants under 6 months of age, the initial opioid dose, calculated in milligrams per kilogram, should be about one-fourth to one-third of the dose recommended for older infants and children. For example. 0.03 mg/kg of morphine could be used instead of the traditional 0.1 mg/kg. Careful assessment is necessary that the optimal dose and interval of administration can be deter¬mined from clinical parameters (e.g., when pain occurs and whether the infant appeared comfortable). Many infants have inadequate pain relief after the initial small dose and require upward titration, sometimes to doses equivalent to those used for older children. Continuous infusions can be used as long as the infusion begins with a conservative starting dose, which is then titrated upward until pain is relieved.
Aggressive monitoring, when necessary, should include frequency assessments and close observation of heart and respiratory rates, respiratory effort, blood pressure, and level of alertness as determined by responsiveness to stimuli. Frequent or continuous assessment of arterial oxygen saturation with pulse oximetry is a valuable adjunct clinical observation. Because of delayed absorption, opioid levels in serum may increase many hours after a one-time intramuscular or subcutaneous dose in infants who are opioid naive: monitoring. therefore, should be continued for 12 hours after an opioid dose (Koehntop, Rodman, Brundage, et al., 1986).
Epidural Analgesia. The use of epidural analgesia is appropriate when systemically administered oral or parenteral analgesics do not achieve adequate pain relief without unacceptable sedation, respiratory depression, or other side effects. Epidural analgesia is now widely used for infants and children with postoperative pain. The hemodynamic and respiratory effects of major regional analgesia in infants with postoperative pain appear minimal (Meignier, Souron, and Le Neel, 1983). Systematic studies on epidural analgesia for children with cancer are unavailable, but experience in a few pediatric centers suggests that some children can be made comfortable with epidural or subarachnoid infusions of opioids and local anesthetics. For pediatric epidural infusion rates, the maximal recommended local anesthetic rates, per hour, are roughly 0.4 mg/kg for bupivacaine and 2 mg/kg for lidocaine. Epidural infusions that exceed those recommended rates may lead to convulsions (Berde, 1992). Epidural morphine has been used successfully even for young infants with cancer (Berde, Fischel, Filardi, et al., 1989). The proper use of infusions or intermittent doses of peridural opioids or local anesthetics requires expertise and close monitoring.
Although little research has focused on the use of cognitive-behavioral methods for children with cancer, they have a potentially important role in relieving pain, and the methods used by adults can be adapted for children. For example, distraction techniques might involve reading or telling a favorite story, talking about the story's characters, or playing video games. Preparation for painful events could involve the use of a puppet, a favorite cartoon character, or an animal. Stuffed animals with inserted audiotapes of stories or music may help to induce a state of relaxation. Something familiar to the child may facilitate participation in these nonpharmacologic strategies.
When a child is in pain, the presence of a parent is usually helpful. Other methods of psychological support include holding someone's hand, a stuffed toy, or favorite blanket; asking questions; using distraction; sleeping and resting; relaxing or using imagery; changing positions; and engaging in humor; these seemingly simple interventions can have powerful effects. Facilitating the child's usual strategies for decreasing pain is important. Although not well researched in child populations, measures such as physical therapy, TENS (Eland, 1989), splinting a painful extremity, positioning, the application of heat or ice, and massage may help relieve pain (see Chapter 4).
The most important consideration in the management of pain in children with cancer is the provision of a child-centered environment and attitude. Health care professionals should understand and consider developmental issues and problems that affect the integrity and structure of the family. Child-centered cancer care provides the necessary items and services to support the child and the family emotionally, socially, and spiritually throughout the treatment process. In such an environment, pain and suffering are continually assessed, and appropriate supportive measures are introduced. Otherwise, the treatment of pain with analgesics and other agents will often be inadequate.
Optimal assessment and treatment require a knowledge of children’s development, behavior, and physiology. Because children are less likely than adults to talk about pain, the responsibility to inquire about pain rests with the health care professional (Table 22). Some children will have pain that is particularly difficult to manage. In these situations. consultation with specialists in children's pain is recommended.
Elderly patients are often undertreated for cancer pain. Attitudes of health care professionals, the public, and patients toward pain can impede appropriate care; because many people consider acute and chronic pain to be a part of normal aging. In some instances, pain is not assessed because elderly patients, who may be confused, have difficulty communicating their pain to health professionals. In other instances, clinicians have mistaken beliefs about decreased pain sensitivity and heightened pain tolerance in the elderly. Frequently, the elderly are given nonopioids or weak doses of medications because their care providers mistakenly believe that they cannot tolerate opioid agents.
The elderly should be considered an at-risk group for the under-treatment of cancer pain because of inappropriate beliefs about their pain sensitivity, pain tolerance, and ability to use opioids. Elderly patients, like other adults, require aggressive pain assessment and management.
Pain management in the elderly presents several challenges, including the discrepancy between the high prevalence of pain in the elderly and the limited attention to this group in the research literature and in medical and nursing texts (Ferrell, 1991). Of all reports about pain published annually, fewer than 1 percent focus on pain experience or syndromes in the elderly (Melding, 1991). Current pharmacologic research is often limited to single-dose studies in young or middle-aged adults and does not assess the complications and side effects of medications in the elderly. Elderly patients who participate in pain clinics or studies are likely to be the mobile elderly. Furthermore, elderly patients are often excluded from rehabilitation programs and aggressive treatment of pain (Middaugh, Levin, Kee, et al., 1988; Sorkin, Rudy, Hanlon, et al., 1990).
| Table 22. Checklist for assessing adequacy of pain management in children | |
| Pharmacologic strategies | |
| Have the child and parent(s) been asked about their previous experiences with pain and their preferences for use of analgesics? | |
| Does the child or parent(s) have reservations about the use of opioids for pain treatment? | |
| Is the child being adequately assessed at appropriate intervals? Are analgesics ordered for prevention and relief of pain? | |
| Is the analgesic strong enough for the pain expected or the pain being experienced? | |
| Is the timing of drug administration appropriate for the pain expected or experienced? | |
| Is the route of administration appropriate for the child? | |
| Is the child adequately monitored for the occurrence of side effects? | |
| Are side effects appropriately managed? | |
| Has the analgesic regimen provided adequate comfort and satisfaction from the perspective of the child or parent(s)? | |
| Nonphannacologic strategies | |
|
Have the child and parent(s) been asked about their experience with and preferences for a given strategy? |
|
| Is the strategy appropriate for the child's developmental level, condition, and type of pain? | |
| Is the timing of the strategy sufficient to optimize its effects? Is the strategy adequately effective in preventing or alleviating the child's pain? | |
| Are the child and parent(s) satisfied with the strategy for prevention or relief of pain? | |
| Are the treatable sources of emotional distress for the child being addressed? | |
In spite of the lack of research, there is evidence that the elderly experience more pain than younger people. It has been estimated that the prevalence of pain in those older than 60 years of age (250 per 1.000) is double that in those younger than 60 (125 per 1,000) (Crook, Rideout, and Browne, 1984). Among the institutionalized elderly, the prevalence of pain may be over 70 percent (Ferrell, Ferrell, and Osterweil, 1990). Elderly patients with cancer often have other chronic diseases, more than one source of pain, and complex medication regimens that place them at increased risk for drug-drug as well as drug-disease interactions.
Cognitive impairment, delirium (common among the acutely ill elderly), and dementia (which occurs in as many as 50 percent of the institutionalized elderly) pose serious barriers to pain assessment (Kane, Ouslander, and Abrass, 1989). Psychometric properties of pain assessment instruments, such as VAS, verbal descriptor, and numerical scales, have not been established in this population. Moreover, a high prevalence of visual, hearing, and motor impairments in the elderly impede the use of these tools. Research on the nursing home population shows that many patients with mild to moderate cognitive impairment are able to report pain reliably at the moment or when prompted, although their pain recall may be less reliable. These findings suggest that this population may require more frequent pain assessment than patients who are not cognitively impaired (Ferrell, in press).
Nonopioid analgesics, including acetaminophen and other NSAIDs, are helpful adjuncts to opioids for cancer-related pain. The risk for gastric and renal toxicity from NSAIDs is increased among elderly patients, however, and unusual drug reactions including cognitive impairment, constipation, and headache are also more common (Roth, 1989). Factors that may contribute to altered side effects in the elderly include multiple medical diagnoses, multiple drug interactions, and altered pharmacokinetics. If gastric ulceration is a concern, NSAIDs with lower gastric toxicity (e.g., choline magnesium trisalicy-late) should be chosen. The coadministration of misoprostol should also be considered as a way to protect the gastric mucosa.
Opioids are effective for the management of cancer pain in most elderly patients. In the elderly, Cheyne-Stokes respiratory patterns are not unusual during sleep and need not prompt the discontinuation of opioid analgesia. Elderly people tend to be more sensitive to the analgesic effects of opioids, experiencing higher peak effect and longer duration of pain relief (Kaiko, 1980). The elderly, especially those who are opioid naive, also tend to be more sensitive to sedation and respiratory depression, probably as a result of alterations in metabolism and in the distribution and excretion of the drugs. For this reason, the prolonged use of longer acting drugs such as methadone requires caution (Ferrell, 1991).
Elderly people in general have increased fat-to-lean body mass ratios and reduced glomerular filtration rates. Opioids produce cognitive and neuropsychiatric dysfunction through poorly defined mechanisms that in part include the accumulation of biologically active metabolites such as morphine-6-glucuronide or normeperidine (Melzack, 1990). Opioid dosage titration should take into account not only analgesic effects but also side effects that extend beyond cognitive impairment. Such side effects may include urinary retention (a threat in elderly males with prostatic hyperplasia), constipation and intestinal obstruction, or respiratory depression.
Local anesthetic infusions, including lidocaine or opioids, may result in cognitive impairment if significant drug levels in the blood are reached. Orthostatic hypotension and clumsiness may result from tricyclic antidepressant administration and other medications used for pain management and concurrent medical illnesses. Precautions, such as assistance during ambulation, should be taken to prevent falls and fractures.
PCA was shown to be safe and effective for postoperative pain relief among some elderly patients (Egbert, Parks, Short, et al., 1990). PCA has not been extensively studied for long-term use in the elderly with cancer-related pain. The use of any "high-tech" pain treatment such as PCA or intraspinal analgesia should be titrated and monitored especially closely because of the elderly patient's increased sensitivity to drug effects (Ferrell, Cronin Nash, and Warfield, 1992).
Although patients with cancer and cancer-related pain commonly experience troublesome negative psychological consequences (Chapter 1), some patients require treatment beyond that provided by the patient's usual health care team. Three cancer centers (Derogatis, Morrow, Fetting, et al., 1983) reported that 53 percent of the patients were adjusting to the stresses of cancer with no diagnosable psychiatric disorder but that the remainder had clinically apparent psychiatric disorders. This study also noted that patients with pain were more likely to develop a psychiatric disorder; 39 percent of patients with a psychiatric diagnosis reported significant pain, but only 19 percent of patients without a psychiatric diagnosis had significant pain.
The incidence of pain, depression, and delirium all increase with higher levels of physical debilitation and advanced illness (Burkberg, Penman, and Holland, 1984). Approximately 25 percent of all patients with cancer experience severe depressive symptoms, with the prevalence increasing to 77 percent in those with advanced illness (Burkberg, Penman, and Holland, 1984). Among patients with cancer, the prevalence of organic mental disorders (delirium) requiring psychiatric consultation has been found to range from 25 to 40 percent and to rise to as high as 85 percent during the terminal stages of illness (Massie, Holland, and Glass, 1983).
Psychiatric symptoms in patients with pain should be viewed initially as a possible consequence of uncontrolled pain. Mood as well as personality characteristics may be distorted by the presence of uncontrolled pain, and relief of pain may result in the disappearance of a perceived psychiatric disorder (i.e., anxiety or depression) (Ahles, Blanchard, and Ruckdeschel, 1983). The patient's mental status should be reassessed after pain has been controlled to determine if a psychiatric disorder is present. The management of specific disorders such as depression, delirium, and anxiety in patients with cancer has been reviewed in detail elsewhere (Breitbart and Holland, 1988; Holland and Rowland, 1989).
The somatic symptoms of depression (e.g., anorexia, insomnia, fatigue, and weight loss) are less reliable and lack specificity in the patient with cancer (Endicott, 1984). Of greater diagnostic value are the psychological symptoms of depression: dysphoric mood, hopelessness, worthlessness, guilt, and suicidal ideation (Massie and Holland, 1990). A history of familial depression or of previous depressive episodes makes this diagnosis more probable.
An evaluation of cancer treatment-related organic factors that can present as depression should accompany treatment. Such factors include corticosteroids (Stiefel, Breitbart, and Holland, 1989), chemotherapeutic agents (Adams, Quesada, and Gutterman, 1984), whole-brain radiation (DeAngelis, Delattre, and Posner, 1989), CNS metabolic-endocrine complications (Breitbart, 1989), and paraneoplastic syndromes (Patchell and Posner, 1989).
Depressed patients with cancer are usually treated with supportive psychotherapy, cognitive-behavioral techniques, and antidepressant medications (Massie and Holland, 1990). The efficacy of tricyclic and other antidepressants in cancer patients is well established (Popkin, Callies, and Mackenzie, 1985). Psychostimulants are most helpful in the treatment of depression in patients with advanced disease and in those for whom dysphoric mood is associated with severe psychomotor slowing and even mild cognitive impairment. Clinicians are referred to the guideline on treatment of depression for further information on this subject (Depression Guideline Panel, 1993a, 1993b).
A patient's use of meperidine while on a monoamine oxidase inhibitor (MAOI) is absolutely contraindicated because it can lead to hyperpyrexia and cardiovascular collapse. One should be extremely cautious when using any opioid analgesics in patients on MAOIs, because myoclonus and delirium have been reported (Breitbart and Holland, 1988). Sympathomimetic drugs and other less obvious MAOIs, such as the chemotherapeutic agent procarbazine, can causea hypertensive crisis in patients taking an MAOI. If a patient has responded well to an MAOI for depression in the past, its continued use is warranted, but with caution.
| Table 23. Risk Factors that predispose cancer patients to depressive disorders |
| Social isolation. |
| Recent losses. |
| A tendency to pessimism. |
| Socioeconomic pressures. |
| A history of mood disorders. |
| Alcohol or substance abuse. |
| Previous suicide attempt(s). |
| Poorly controlled pain. |
| Source: Depression Guideline Panel, 1993a. |
Few patients with cancer commit suicide, but poorly controlled pain places them at increased risk. Tables 23 and 24 list factors that predispose cancer patients to depression and increase risk of suicide.
Although fleeting or occasional thoughts of suicide probably occur commonly in those with advanced illness, persistent and intense suicide thinking is rare in the absence of depression or of uncontrolled physical symptoms such as pain (Breitbart, 1990a). Suicide is often held as an option by the patient to retain a sense of control. Fear of unacceptable pain was a major component of requests to physicians for assisted death (Helig, 1988) and is so important to patients with cancer that 69 percent reported that they would consider committing suicide if their pain was not adequately treated (Levin, Cleeland, and Dar, 1985). In another study, the majority of patients who committed suicide had severe pain that was inadequately controlled (Bolund, 1985). Clinicians in a pain clinic report seeing many patients who considered suicide who changed their minds once given adequate pain relief (Foley, 1991). For at-risk patients, clinicians should be aggressive in the use of analgesics and other appropriate drugs as well as crisis intervention-oriented psychotherapeutic approaches that mobilize the patient's support system.
| Table 24. Suicide risk factors in cancer patients with pain |
| Depression. |
| Poorly controlled pain. |
| Previous suicide attempt(s). |
| Family history of suicide. |
| Delirium. |
| Substance abuse. |
| Prior psychiatric diagnosis (depression). |
| Advanced disease. |
| Increasing age. |
| Disfiguring disease or surgery. |
| Poor social support. |
| Source: Depression Guideline Panel, 1993a. |
Although it is appropriate to intervene when medical or psychiatric factors are clearly the driving force in a suicidal cancer patient, overly aggressive intervention may be less helpful in patients with advanced illness when comfort and symptom management are the primary concerns. As an alternative to suicide, the goal of intervention should be to establish rapport, to develop an alliance, and to provide effective management of symptoms.
Although the experience of anxiety is a normal response when people learn they have cancer or in the case of recurrence or treatment failure, patients who continue to experience high levels of anxiety for weeks or months should be referred to a psychiatrist, psychologist, psychiatric nurse, or psychiatric social worker for evaluation and possible treatment. Some patients with cancer have long-standing or chronic anxiety disorders, such as phobias and panic disorder, that can be exacerbated by the stressors of cancer illness or treatment. Claustrophobic patients may experience acute anxiety when confined in diagnostic scanning devices or radiotherapy treatment rooms (Brennan, Redd, Jacobsen, et al., 1988). Patients who experience such acute anxiety or exacerbations of chronic anxiety disorders may require the use of medications such as benzodiazepines or cognitive-behavioral therapies to help them tolerate procedures.
Cognitive failure (delirium), common in patients with advanced illness (Fleishman and Lesko, 1989), is an etiologically nonspecific, global, cerebral dysfunction, characterized by concurrent disturbances of level of consciousness, attention, thinking, perception, memory, psychomotor behavior, emotion, and the sleep-wake cycle. Disorientation, fluctuation, waxing and waning of the above symptoms, and acute or abrupt onset of such disturbances are other critical features of delirium. Delirium is reversible, even in advanced illness. Still, it may not be reversible in the last 24 to 48 hours of life, most likely because of irreversible processes such as multiple organ failure (Massie and Holland, 1990).
At times, it is difficult to differentiate delirium from dementia because they frequently share clinical features such as disorientation and impaired memory, thinking, and judgment. One difference is that the temporal onset of symptoms is more subacute or chronically progressive in dementia than in delirium. Occasionally, delirium is superimposed on an underlying dementia, such as in the case of an elderly patient, an AIDS patient, or a patient with a paraneoplastic syndrome
Delirium can be due to the direct effects of cancer on the CNS, to indirect CNS effects of the disease or treatments (medications, electrolyte imbalance, failure of a vital organ or system, infection, vascular complications) and to preexisting cognitive impairment or dementia.
Medical and nursing staff sometimes conclude that a new symptom is psychologically based without first ruling out all possible organic causes. Given the many drugs that cancer patients require ar the fragile state of their physiologic functioning, even routinely ordered hypnotics can be enough to precipitate delirium. Opioid analgesics, including levorphanol, morphine sulfate, methadone, and meperidine (Bruera, Macmillan, Hanson, et al., 1989), can cause confusional states, particularly in the elderly and terminally ill. Clinicians should correct those underlying causes of delirium (Adams, Fernandez, and Andersson, 1986; Fainsinger and Bruera, 1992; Fish, 1991; Lesko and Fleishman, 1991; Lipowski, 1987).
There is a low risk of iatrogenic addiction in patients with cancer (Kanner and Foley, 1981). For patients with cancer who are additionally diagnosed as substance abusers, the complex physiologic, behavioral, and psychological phenomena associated with drug addiction in no way exclude the ability to perceive painful stimuli. When opioids are required for pain management, the concurrence of substance abuse disorders and cancer produces a difficult and stressful situation for even the most experienced clinician. Nevertheless, certain principles can be followed to assure a careful and fair assessment of the pain complaint and to provide the best chance of achieving satisfactory pain relief in these circumstances (Portenoy and Payne, 1992). These principles are complementary to those discussed elsewhere in this guideline.
Tolerance and physical dependence are predictable consequences of chronic opioid use, but they do not equal addiction (Newman, 1983). Numerous clinical, epidemiologic, and pharmacologic studies now suggest that cancer patients may become tolerant and physically dependent on opioids if therapeutic doses are prescribed for several weeks. However, very few patients develop the loss of control and compulsive use patterns that characterizes addiction to opioids (and other prescribed or illicit substances) despite medical, social, legal, or emotional harm. This sort of behavior, more than tolerance and physical dependence, characterizes and defines addiction (Portenoy and Payne,1992).
An adequate assessment of the cause of pain is essential to the optimal treatment of the opioid addict with cancer. As with other cancer populations, specific antitumor treatments are indicated as the single best method of pain relief. Frequently, however, adequate anal-gesia has to be established before diagnostic studies and medical or surgical treatments can proceed. The appropriate management of the medical condition often decreases the requirements for opioids. Optimal pain treatment is essential to preventing the syndrome of "pseudoaddiction" (Weissman and Haddox, 1989) because inadequate pain management will invariably produce the manipulative behavior that the clinician wants to avoid.
Although not clearly substantiated by prospective clinical studies, it is common practice to make distinctions among (1) addicts who are actively abusing opioids and illicit narcotics at the time of their treatment for acute pain, (2) former addicts who no longer abuse drugs, and (3) addicts in methadone maintenance (Fultz and Sonay, 1975). Patients actively abusing heroin or prescription opioids (and those on methadone maintenance) should be assumed to have some degree of pharmacologic tolerance, which will be reflected in a need for higher-than-usual starting doses and dosing intervals shorter than those generally recommended in the nonaddicted population. Furthermore, because patients who are actively abusing drugs often manifest psychological disorders that influence pain perception (e.g., anxiety and depression), the treatment of concomitant psychiatric disturbances is often necessary and usually requires the consultation of experienced psychiatric clinicians. Patients with cancer who have abused drugs in the past (but who are not current abusers) or who are participating in methadone maintenance programs may have a higher degree of opioid tolerance than the general population. Among these patients, it is useful to anticipate that significant anxiety may accompany the stress of medical illness and acute pain and be manifested in a reappearance of drug abuse behaviors.
The use of opioid agonist-antagonist compounds in known or suspected active opioid addicts is absolutely contraindicated. Not only do these drugs have ceiling effects to their analgesic efficacy, and are therefore inappropriate for severe pain, but they may also precipitate withdrawal and increased pain in physically dependent patients. Likewise, nonopioid analgesic modalities should not be substituted for opioid analgesics to treat severe pain in patients who are suspected or known abusers of illicit substances. Tolerance to opioid analgesics decreases the duration of effective analgesia (Houde, 1979); therefore, "tolerant" patients require more frequent dosing than do nontolerant patients. For example, morphine, which has an average analgesic duration of 3 to 4 hours, may produce only 1 to 2 hours of pain relief in an opioid addict with a large degree of tolerance.
Paradoxically, PCAs are being used with increasing frequency when rapid titration of intravenously administered opioids is required in this population. Although it would seem that the administration of opioids could not be entrusted to an addicted individual, in fact, it has its advantages: With the appropriate prescription of doses, "lock-out" intervals, and instruction to the patient, this method of administration may reduce the confrontation and conflict inherent in clinician-administered analgesia. Opioid addicts may report a euphoric feeling or "high" coincident with an intravenous bolus injection of opioids, which presumably reinforces the need to self-administer drugs (Jaffe, 1985). Nevertheless, intravenous opioids can be used effectively (see below).
Appropriate PCA bolus doses and "lock-out" periods (i.e., the time that should elapse between the administration of one dose and the next) should be selected. The opioid addict may be easily underdosed and experience poor pain relief if the degree of the patient's tolerance is not accurately assessed. The commonly published "lockout" times and starting bolus doses are inappropriate for most opioid addicts and indeed for many patients whose prior opioid experience is such that they manifest opioid tolerance; in fact, the typical published parameters apply to the postoperative population of relatively opioid naive patients. The prolonged self-administration of morphine to cancer patients with acute recurrent pain caused by oral mucositis after bone marrow transplantation did not increase the risk for over-medication or addiction (Chapman and Hill, 1989), and compared with standard intravenous infusion, PCA decreased the requirements for morphine by 53 percent (Hill, Chapman, Kornell, et al., 1990). Although those patients were not addicts, the data nevertheless support the argument that PCA of intravenous morphine for pain does not invariably lead to ever-escalating dose requests.
Patients who are maintained on methadone for the treatment of addiction may also be treated with this agent for pain, if it is administered frequently enough. In this setting, methadone is useful in that the patient's dose may be easily tapered back to the level of the maintenance dose after the painful episode has been treated. In reality, however, most methadone maintenance programs do not have the flexibility to change the rules for individual patients to allow increases in the daily methadone dose or to increase the dosing frequency beyond once or twice a day. Unfortunately, then, the treatment of pain with methadone in this manner usually has to take place outside of the typical maintenance program.
For acute focal pain syndromes, regional anesthetic approaches such as somatic and sympathetic nerve block should be considered, unless contraindicated. These approaches are generally unsafe in patients who are septic, who have coagulopathy, or who are acutely confused and uncooperative. Nonpharmacologic methods can be useful adjuncts in the treatment of pain in this population.
One common characteristic of patients who are actively abusing opioids is a failure to set limits on their drug-seeking behavior, even in the presence of liberal uses of opioids for pain management. The clinician should discuss expectations and define limits of acceptable and unacceptable behaviors with the patient. The use of drug infusion pumps with security locks (available on almost all PCA pumps) should prevent dose escalation beyond what the clinician prescribes. If oral opioid analgesics are being administered, patients should be told that their ingestion will be witnessed and that routine precautions, such as searching the room for hidden pills or signs of hoarding, will be taken.
In the outpatient setting, clearly stated, written rules should cover prescription renewals, the procedure to be followed with lost or stolen prescriptions or medications, and procedures to ensure that only one clinician is prescribing analgesic medications. Prescription theft or forgery should lead either to the patient's admission to the hospital for continuation of opioid therapy, if still required, or to withdrawal of the therapy and referral to an appropriate drug treatment program, if opioid therapy for pain is no longer required. The patient should be seen frequently—daily, if necessary—and a limited quantity of opioids should be prescribed. In some States, the prescription of opioids to a patient known to be a "habitual user*' or "addict" must be reported to the State's regulatory agencies.
These general guidelines allow the clinician and the patient to establish behavioral expectations, which may be the only way to manage humanely. Patients with pain and substance abuse disorders require interdisciplinary assessment and care. These patients are generally not well managed by the traditional medical models of oncologic care, because the issues of pain management and substance abuse treatment together are almost always beyond the competence of a single clinician or clinical service and may often produce directly conflicting goals for treatment. On the one hand, for example, the traditional method of opioid addiction treatment is to detoxify the patient and provide pharmacologic and psychological therapies to maintain abstinence. On the other hand, in treating the addict with cancer-related pain, the avoidance of opioids is usually unacceptable, because there are few alternatives for effective pain treatment. However, the pain specialist usually has little ability or training to assess fully the behaviors manifested by addicts, particularly those actively abusing drugs. There is no obvious solution to this paradox, but clearly, clinical research is needed to develop a model for the care of the addict with pain that allows flexibility of traditional concepts of substance abuse and pain management and provides a mechanism that effectively integrates both disciplines.
In general, minority patients are likely to receive less adequate cancer treatment (Blendon, Aiken, Freeman, et al., 1989; Freeman and Wasfie, 1989). Indexing the adequacy of cancer pain management by WHO standards, a recent multicenter study found that minority patients (African-Americans and Hispanics) with pain due to metastatic cancer were three times more likely to have inadequate pain treatment than those seen in nonminority settings (Cleeland, Gonin, Hatfield, et al., in press).
The barriers that limit pain control in nonminority patients are even more likely to impede the pain management of minority patients. Cultural and linguistic differences may impair adequate assessment. Less frequent followup care will also lead to less adequate identification and monitoring of pain. Health care providers may be more concerned about the potential addiction of minority patients who need opioid medications on a chronic basis, and many minority patients may be more reluctant to use analgesics they see as potentially addicting. Because of concerns about theft and violence, pharmacies in neighborhoods where minorities live may not stock opioid analgesics (Kanner and Portenoy, 1986). Minority patients are more often economically disadvantaged, leading to inadequate reimbursement for pain control. Patients and families may face the dilemma of choosing between paying for pain medications or paying for other necessities.
Patients from special populations, because of their behavior when they have pain, may also be perceived as needing less aggressive pain management. Health care professionals frequently comment that such patients seem to have less pain even when they have equally severe tissue damage, because pain in these patients is inadequately assessed. but patients from diverse cultural and linguistic backgrounds will rate their pain at a similar level of severity when given the appropriate pain rating scales (Cleeland, 1989a). Judging pain by just observing the behavior of the patient will lead to consistent underestimation of its severity (Grossman, Sheidler, Swedeen, et al., 1991). Therefore, when developing a pain treatment plan, health care clinicians should be aware of the unique needs and circumstances of patients from various ethnic and cultural backgrounds.
Pain is a symptom commonly experienced by patients with HIV infection, even in the absence of an opportunistic cancer such as Kaposi's sarcoma (O'Neill and Sherrard, 1993). The principles of pain assessment and treatment in the patient with HIV positive/AIDS are not fundamentally different from those in the patient with cancer and should be followed for patients with HIV positive/AIDS.
The reported prevalence of pain in HIV-infected individuals varies depending on stage of disease, care setting, and study methods. Estimates of the prevalence of pain in HIV-infected individuals generally range from 40 to 60 percent with prevalence of pain increasing as disease progresses (Breitbart, Passik, Bronaugh, et al., 1991; Lebovits, Lefkowitz, McCarthy, et al., 1989; Schofferman and Brody, 1990; Singer, Zorilla, Fahy-Chandon, et al., 1993). Thirty-eight percent of ambulatory HIV-infected patients reported significant pain in a prospective study of current pain prevalence (Breitbart, Passik, Bronaugh, et al., 1991). Fifty percent of patients with AIDS reported pain, while 25 percent of those with earliest stages of HIV infection had pain. Patients had an average of pain from two or more sources at a time. A review of ambulatory HIV-infected men (Singer, Zorilla, Fahy-Chandon, et al., 1993) demonstrated that 28 percent of those who were asymptomatic seropositive, 55.6 percent of those with AIDS-related complex, and 80 percent of those with AIDS reported one or more painful symptoms over a 6-month period. A study of pain in hospitalized patients with AIDS revealed that over 50 percent of patients'required treatment for pain with pain the presenting complaint in 30 percent (Lebovits, Lefkowitz, McCarthy, et al., 1989). Schofferman and Brody (1990) reported that 53 percent of patients with far-advanced AIDS cared for in a hospice setting had pain. The most common pain syndromes reported in studies to date include painful peripheral sensory neuropathy, pain from extensive Kaposi's sarcoma, headache, pharyngeal and abdominal pain, arthralgias and myalgias, and painful dermatologic conditions (Breitbart, Passik, Bronaugh, et al., 1991; Lebovits, Lefkowitz, McCarthy, et al., 1989; Schofferman and Brody, 1990; Singer, Zorilla, Fahy-Chandon, et al., 1993).
HIV-related peripheral neuropathy is often a painful condition, affecting up to 30 percent of people with AIDS (Comblath and McArthur, 1988), and is characterized by a sensation of burning, numbness, or anesthesia in the affected extremity. Several antiviral drugs, such as didanosine or zaicitabine, chemotherapy agents used to treat Kaposi's sarcoma (vincristine), as well as phenytoin and isoni-azid, can also cause painful peripheral neuropathy.
Reiter's syndrome, reactive arthritis, and polymyositis are painful conditions reported in early HIV infection (Kaye, 1989). Other painful rheumatologic manifestations of HIV infection include various forms of arthritis (painful articular syndrome, septic arthritis, psoriatic arthritis), vasculitis, Sjogren's syndrome, polymyositis, zidovudine (AZT) myopa-thy, and dermatomyositis (Espinoza, Aguilar, Berman, et al., 1989).
Conditions associated with chronic or intermittent pain include intestinal infections with Mycobacterium avium-intracellulare and cryptosporidium, which cause cramping and intermittent abdominal pain; hepatosplenomegaly, resulting in abdominal distention and pain; oral and esophageal candidiasis, causing pain while the patient is eating and swallowing; and severe spasticity associated with encephalopathy, which causes painful muscle spasms.
HIV-related conditions that cause acute pain in children include meningitis and sinusitis, which result in severe headaches; otitis media; shingles; cellulitis and abscesses; severe Candida dermatitis; and dental caries.
The patient with HIV disease faces many stressors during the course of illness including dependency, disability, and fear of pain and painful death. Such concerns are universal; the level of psychological distress, however, is variable and depends on social support, individual coping capacities, personality, and medical factors, such as the extent or stage of illness. In a study of pain in ambulatory HIV-infected patients (Breitbart, 1993), depression was significantly correlated with the presence of pain. In addition to being significantly more distressed and depressed, those with pain (40 percent) were twice as likely to have suicidal ideation as those without pain (20 percent). HIV-infected patients with pain were more functionally impaired, were more depressed, were more likely to be unemployed or disabled, and reported less social support.
Children with HIV infection often come from multiproblem families (Boland, Mahan-Rudolph, and Evans, 1989). Many families have more than one infected member, and multiple losses from AIDS in one family are common. These issues affect how families deal with the disease and the pain it causes. Parental guilt, which often results in denial of the disease, can also cause denial of a child's pain and resistance to adequate pain management.
Fears of addiction and concerns regarding drug abuse affect both patient compliance and clinician management of opioid analgesics and often lead to the undermedication of HIV-infected patients in pain. Also problematic is the management of pain in the growing segment of HIV-infected people who are actively using illicit drugs.
The issue of drug abuse is also problematic in the pediatric HIV population. Many children with HIV infection live in families where intravenous drug abuse is or was a problem. Either they have parents who are actively using drugs or are recovered from drug abuse, or they live with extended family who have had experiences with their drug-abusing relatives. In these environments, questions arise about the safety of prescribing opioids for the child. Extended-family members are often anxious about the use of opioids for a child whose parent was a drug addict, fearing that the child will also become addicted. These fears and concerns should be anticipated and discussed, and explicit plans such as those discussed earlier should be put into place to minimize the risk of drug diversion.
The general management of pain in children with HIV is the same as that for children with cancer. The assessment of pain in HIV-infected children may be complicated by the frequency of encephalopathy and related developmental delays. It is often difficult to determine whether an encephalopathic infant or toddler who cannot talk is in pain. Observations of a child's response to a trial of pain medication may be the best means of assessing such a child's pain (see Chapter 6).