Patient versus Caregivers/Family

Who is the authority on patients’ pain? Whose pain is it? Clinicians sometimes believe they know more about the patient’s pain than the patient does. No matter how appealing that belief may be, it is false. Nevertheless, privately or among themselves, clinicians may comment about a patient, “He doesn’t have as much pain as he thinks he does” or “The pain is not that bad,” implying that the clinicians are the true authorities on patients’ pain. In research cited earlier, clinicians’ tendencies to underestimate patients’ pain is noted.

No objective measures of pain exist. The sensation of pain is completely subjective. Pain cannot be proved or disproved. One definition of pain used in clinical practice says “Pain is whatever the experiencing person says it is, existing whenever he says it does” (McCaffery, 1968, p. 95). Statements by the American Pain Society (APS) have echoed the same approach to patients’ reports of pain by statements such as:

The gold standard for assessing the existence and intensity of pain is patients’ self-reports. No other source of information has ever been shown to be more accurate or reliable than what a patient says. Patients’ behaviors, the opinions of nurses and physicians delivering care, patients’ vital signs—none of these is as reliable as patients’ reports of pain and should never be used instead of what a patient says.

The Andrew-Robert survey presented in Box 1-1 (p. 18) illustrates what happens when clinicians do not adopt the patient’s self-report as the standard for assessment of pain intensity. This survey has been used by many nurse educators in hundreds of educational programs to explain the necessity of accepting the patient’s report of pain as the standard for assessment. The survey has become so familiar to staff nurse educators that it is often referred to simply as the Andrew-Robert survey. Several publications have reported the results of studies using this survey and modifications of it (McCaffery, Ferrell, 1991a, 1991b, 1992a, 1992b, 1994a, 1997a; McCaffery, Ferrell, O’Neil-Page, 1992). A summary of survey findings from 1990 to 2006 has been published by the originators of the survey (McCaffery, Pasero, Ferrell, 2007). The results of the Andrew-Robert survey as presented in Table 2-1 are based on the responses of 615 registered nurses who attended pain programs throughout the United States. The survey was administered to registered nurses before the pain conference began. In viewing the nurses’ responses to pain assessment (question 1), it is apparent that not all nurses understood that the patient’s self-report of pain is the single most reliable indicator of pain. Both of these patients reported their postoperative pain as 8, but 14% of the nurses did not record 8 for the smiling patient and 5% did not record 8 for the grimacing patient. Those nurses who did not record 8 falsified the record and made it impossible for the next nurse to evaluate previous treatment for pain.

The percentages of nurses who recorded what the patients said are high but there is no room for false entries in the chart. If these results were from a single institution, then one would have to say that up to 14% of the nurses were not recording what the patients said, and this is made more complex because we do not know who they are. Thus, all recordings of pain intensity would have to be viewed with some suspicion.

The Andrew-Robert survey is a quick and easy survey to use in small and large populations of nurses. The results of the survey help to generate discussion of common problems that arise in patient care. Another brief and more comprehensive survey of staff knowledge is titled Pain Knowledge and Attitude Survey and is available in Appendix A.

The survey findings illustrate what is sometimes seen in clinical practice—each person caring for a patient may have a different opinion about the intensity of a patient’s pain. Without a standard for assessing pain, chaos quickly ensues. For example, four different clinicians caring for the same patient may arrive at four different pain ratings, all of which are different from the patient’s own pain rating (and usually are underestimations of a patient’s pain). How do you resolve five different pain ratings so that intervention can be planned?

The need to establish the patient’s self-report as the standard becomes apparent. There appears to be no alternative. It is reassuring to realize that the validity and reliability of patients’ self-reports of pain are testified to in the numerous double-blind studies of analgesics, in which the patients’ pain ratings always determine the analgesic effect of the drugs being tested. Initial recommended doses of analgesics and equianalgesic charts have relied on the patient’s self-report for decades.

When a patient reports pain, the health care professional’s responsibility is to accept and respect the report and to proceed with appropriate assessment and treatment based upon the self-report. All reports of pain are taken seriously. When difficulties arise in accepting the patient’s report of pain, some of the strategies listed in Box 2-1 may be helpful.

Health Care Conditions that Influence Clinicians’ Judgments of Pain

Numerous factors influence clinicians’ tendencies to doubt patients’ reports of pain. This section discusses the effects of length of time caring for patients with pain, environmental cues that promote suspicion of patients’ pain reports, and the effects of the nature of one’s clinical practice.

Studies report conflicting results about the effects of length of time caring for patients with pain. As reported earlier, clinicians frequently underestimate patients’ pain ratings. This has been explained as being a process of habituation. Repeated exposure to patients’ pain seems to promote the development of insensitivity to pain (Kappesser, Williams, Prkachin, 2006). Yet other studies have not shown this (Everett, Patterson, Marvin, 1994; Hamers, van den Hout, Halfens, 1997). In a study of 50 registered nurses randomly selected from 2 general surgical units, two general medical units, and a combined intensive care unit/coronary care unit, participants completed an instrument that contained 60 vignettes describing a patient’s illness or injury (Dudley, Holm, 1984). Nurses were instructed to rate each patient on the degree of pain and psychologic distress. Nurses inferred significantly less pain than psychologic distress. Correlation analysis revealed a markedly weak and statistically insignificant correlation between years of practice and pain scores.

Other factors that may influence clinicians’ estimates of patients’ pain may be access to the patients’ self-reports and the atmosphere of suspicion surrounding particular patients. In one study, 60 physicians and nurses from the emergency department (ED) and 60 from the oncology setting watched videotapes of facial expressions of patients with shoulder pain undergoing range-of-motion exercises (Kappesser, Williams, Prkachin, 2006). Participants were divided into three groups and given different information: Group 1, videotape only; Group 2, the same videotape plus the patients’ reported pain intensity; and Group 3, the same videotape plus being told the patient pain reports and that some of the people were faking pain to obtain opioid drugs. All participants were asked to rate the patients’ pain, and participants in groups 2 and 3 were asked to state whether they had the impression that any of the patients had faked pain or hidden pain. Clinicians in all three groups underestimated patient pain intensity. The least discrepancy between clinicians’ and patients’ pain ratings occurred when the clinicians were given the patients’ self-reports but were not told that some patients were faking pain to obtain opioids.

The greatest discrepancy occurred when clinicians were given the patients’ pain reports but were told that patients might be faking pain to obtain opioids. This suggests that a clinical atmosphere of suspicion, such as sometimes occurs in trauma units or EDs, may cause an increase in clinicians’ doubting patients’ reports of pain. In all clinical areas clinicians need to be cognizant of the attitudes they convey when they discuss patient care with their colleagues. Managers and supervisors can be alert to stigmatizing labels and help staff to avoid them. For example, the term drug-seeking, which has no universally accepted definition, commonly conveys that a patient is addicted to opioids, abusing pain medicine, or manipulative (McCaffery, Grimm, Pasero, et al., 2005). In fact, the American Society for Pain Management Nurses (ASPMN; 2002) recommends that this term not be used because it creates bias and prejudice that subsequently have negative effects on pain management.

The nature of a particular clinical practice may influence clinician estimates of patient pain intensity. The effect of repeatedly inflicting pain on patients was studied by using brain scans to compare the responses of physicians who practice acupuncture with the responses of naïve participants while watching animated visual stimuli depicting body parts in both a nonpainful situation (being touched by a Q-tip) and a potentially painful situation (acupuncture [being pricked by needles]) (Cheng, Lin, Liu, et al., 2007). The acupuncturists rated these situations as being significantly less painful and unpleasant than did the naïve control participants.

In addition, the brain scans supported the notion that mechanisms are triggered in the areas of the brain that regulate emotions and cognitive control. The authors commented that without some regulatory mechanism, it is likely that clinicians would experience personal distress and anxiety that would interfere with their ability to perform. Acupuncturists know that they may be inflicting pain and seem to have learned throughout their training to inhibit the empathy-pain response.

Many other clinical practices, such as burn care and surgery, involve repeated infliction of pain. A study of burn care found that staff members who had spent more time working with burned patients believed débridement was less painful than did the staff members who had spent less time working with burned patients (Perry, Heidrich, 1982). To continue to manage pain effectively, such clinicians need to be aware of their tendencies to underestimate the painful impacts of what they do and to recognize that doing so may be an essential mechanism that allows them to continue to give care. Rather than be embarrassed or ashamed of this reaction, they can acknowledge it to themselves (and perhaps others) and try to compensate by obtaining and listening carefully to information from their patients.

Other examples are clinicians who perform venipunctures or bone marrow aspirations daily; they may become less sensitive to the amount of pain they are inflicting. Clinicians who work in a pain clinic several days a week may grow weary of and less sensitive to patients who continue to report pain despite their best efforts to relieve the pain. Again, to continue to practice effectively in such conditions it may be necessary to make a conscious personal inventory that honestly admits to having less empathy for patients as time goes by. Once this is acknowledged, the clinician can plan a systematic approach to obtaining information from the patient and ensure that it is acted upon.

Considerations When Doubts Arise

On occasion, accepting and acting on the patient’s report of pain are difficult. Because pain cannot be proved, health care professionals are vulnerable to being fooled by the patient who wishes to lie about pain. However, although accepting and responding to the report of pain will undoubtedly result in giving analgesics to some patients who do not have pain, doing so ensures that everyone who does have pain receives attentive responses. Health care professionals do not have the right to deprive a patient of appropriate assessment and treatment simply because they believe a patient is lying.

An important distinction exists between believing the patient’s report of pain and accepting the report. Following the recommendations of the clinical practice guidelines does not require that the clinician agree 100% with what the patient says. Clinicians are not required to believe a patient but are required to accept what a patient says, convey acceptance to the patient, and take the appropriate action. Clinicians are entitled to their personal doubts and opinions, but they cannot be allowed to interfere with appropriate patient care. Box 2-1 summarizes some strategies that can be used when the patient’s report of pain is not accepted.

Although accepting the patient’s report of pain occasionally results in being fooled, no stigma or blame should be attached to being duped. In any relationship, each party has certain responsibilities. Fault is assigned to the parties who fail to meet their responsibilities (Wesson, Smith, 1990). If the clinicians fulfills his or her responsibility to respond to all reports of pain with appropriate assessment and treatment, the clinician will be able to say, “Although I was probably fooled by some patients, I never failed to help those who did have pain. No one can find fault with my behavior or professional conduct.”

Furthermore, accepting the patient’s report of pain avoids an adversarial relationship. When a clinician conveys to a patient that a report of pain has not been accepted, it amounts to accusing the patient of lying. Understandably, this is upsetting and frightening to a patient who has asked the health care provider for help with pain. Much has been written in the literature about the distinction between suffering and pain. It is worth noting that in his analysis of suffering, Eric Cassell (1982) mentions that one source of suffering is physicians who do not validate a patient’s pain but rather ascribe it to psychologic causes or accuse the patient of faking. Clinicians ask patients to trust them. At some point clinicians must return the favor.

The term malingering may be encountered when a patient’s report of pain is doubted. Malingering can be defined as a conscious effort to produce symptoms such as pain for the purpose of deceiving or misrepresenting the facts, usually for monetary or other gains. Malingering seems to be suspected most commonly in patients with persistent pain, especially low back pain. A review of the literature prior to 1999 reveals 68 references to research of persistent pain malingering or disease simulation (Fishbain, Cutler, Rosomoff, et al., 1999). These studies conclude that malingering does occur, perhaps in 1.25% to 10.4% of patients, but these figures are not reliable because of the quality of the studies. This review concludes that there are no reliable methods of identifying malingering. The International Association for the Study of Pain notes that the process of identifying malingering is, in the final analysis, a legal, not a medical, process (Merskey, Bogduk, 1994).

On reflection, one might ask whether it is justifiable to be suspicious of all patients in an attempt to avoid being fooled by the few who lie. It is a burden to the clinician and an insult to the patient to wrestle with potential dishonesty at each encounter.

Acute Pain Model versus Adaptation

The acute pain model says that if the patient has pain, visible signs of discomfort, behavioral or physiologic, will be present. Examples of behavior usually expected of patients with pain include grimacing, rigid body posture, limping, frowning, or crying. Physiologically, elevated vital signs are commonly expected. Clinicians and laymen alike usually fail to appreciate that both physiologic and behavioral adaptation occurs, leading to periods of minimal or no signs of pain. Absence of behavioral or physiologic signs of pain does not necessarily mean absence of pain.

The acute pain model is of limited value for assessing pain. When pain is sudden or severe, behavioral and physiologic indicators may be present for a brief time. However, very quickly the patient may make an effort to cease behaviors such as crying or moaning. Behavioral adaptation or suppression of pain behaviors may occur because the patient values the stoic response or simply becomes exhausted. Physiologic indicators such as increased blood pressure (BP) or heart rate may also disappear. In a healthy individual, the body seeks homeostasis or equilibrium, returning to the former physiologic state despite severe pain. Or a patient may have a medical condition that causes low BP, such as hypothyroidism or dehydration, which has a much greater impact on vital signs than pain has. In such patients, sudden, severe pain may elevate the vital signs only briefly and minimally.

The APS addresses the misconceptions about the acute pain model by stating the following:

Responses Clinicians Expect of Patients with Pain

Nurses’ responses to the Andrew-Robert survey (see Table 2-1, p. 21) reveal that patients’ behavioral responses have significant effect on nurses’ pain assessments and treatment decisions (McCaffery, Pasero, Ferrell, 2007). The only difference between Andrew and Robert is their behavior—Andrew smiles and laughs with visitors, whereas Robert lies in bed and grimaces. This simple difference has a startling effect on nurses’ decisions about opioid doses.

Although the patients were exactly alike except for their behaviors, the nurses were more likely to increase the morphine dose for the grimacing patient. Both patients had received morphine, 2 mg IV, 2 hours before; half-hourly pain ratings had ranged from 6 to 8 out of 10 and were currently 8; and no clinically significant adverse effects such as sedation had occurred. The pain rating goal was 2. Nurses were given a choice of administering no morphine or 1 mg, 2 mg, or 3 mg IV. Morphine, 3 mg IV, was the correct choice for both patients because the previous dose of 2 mg was safe but ineffective. However, only 50.6% of the nurses would increase the dose for the smiling patient, whereas 64.3% would increase the dose for the grimacing patient. Both patients were undertreated, but at least 14% of the nurses knew that it was safe to increase the dose for the smiling patient, but they did not. Over the years since 1990, other similar vignette surveys conducted by these authors have shown similar findings, which show little improvement in nurses’ choices of opioid doses (McCaffery, Pasero, Ferrell, 2007).

The same discouraging results were found when the Andrew-Robert survey was adapted to reflect the needs of elderly patients in a long-term care facility (Katsma, Souza, 2000). The participants were 89 licensed nurses working in long-term care facilities. As in the original vignette, one patient was described as smiling and the other as grimacing. After receiving medication, hourly pain ratings were 6 to 8, and the patients showed no serious adverse effects. In response to the question about which analgesic dose they would now give, only 30% correctly selected a higher dose for the smiling patient, but 43% did so for the grimacing patient.

These same biases about behavior also exist in laypersons. The Andrew-Robert survey was revised to be appropriate for a nonnurse audience and was administered to 85 college students who were not enrolled as medical or nursing majors (McCaffery, Ferrell, 1996a). College students’ responses to assessment and relief of pain showed trends similar to those of practicing nurses. The smiling patient’s pain rating was accepted by 38% of the college students, whereas 55% accepted the grimacing patient’s report of pain.

Vital signs also influence nurses’ willingness to record a patient’s report of pain. A survey using the same format as the original Andrew-Robert survey was constructed; the only difference between the patients was their vital signs (McCaffery, Ferrell, 1992a, 1992b). One patient had low to normal vital signs, and the other patient had elevated vital signs. The responses of 166 nurses revealed that more nurses were willing to accept the report of severe pain from the patient with elevated vital signs than from the patient with low to normal vital signs.

As part of the Missoula Demonstration Project, a 15-year process to study end-of-life issues, surveys about pain knowledge and attitudes were sent to 942 nurses and produced 311 responses (Mayer, Torma, Byock, et al., 2001). The survey revealed that 93% of the nurses knew that vital signs were not reliable indicators of pain, but 75% said that vital signs moderately or greatly influenced their decisions to treat or not to treat pain.

In a small study designed to identify the criteria nurses use to assess postoperative pain, 10 nurses were interviewed about their pain assessments of 30 postoperative patients (Kim, Schwartz-Barcott, Tracy, et al., 2005). The strategy most frequently reported by the nurses relied on the patient’s appearance and drew on past experience of which physical signs to look for, such as facial expression, body movement, and heart rate. For example, nurses mentioned frowning or wincing as being indicative of pain and a patient’s being able to fall asleep as a sign of little or no pain. The researchers commented on the fact that the predominant strategy of looking for objective signs of pain was in marked contrast to current guidelines that emphasize the patient’s self-report of pain.

Expectations that certain behaviors indicate pain also influence the prescribing of analgesics. To identify factors that affect physicians’ decisions to prescribe opioids for persistent noncancer patients, the records of 191 patients referred to a pain center were examined to determine pain severity, physical findings, pain duration, age, gender, observed pain behaviors, reported functional limitations, and affective distress (Turk, Okifuji, 1997). Of all these variables, only observed pain behaviors were significantly related to receiving opioid prescriptions. In other words, patients who exhibited pain behaviors were most likely to receive opioids for pain relief. The extent of physical findings and the severity of the pain did not appear to influence the decision to prescribe opioids. Because opioids are prescribed for the purpose of pain relief, it would seem more logical to find that severity of pain determined prescription of opioids, but this was not the case.

Similar findings also were reported regarding patients with low back pain. Decisions about lumbar surgery were not made on the basis of physical pathologic conditions but rather on behaviors demonstrated by patients during their evaluations (Waddell, Main, Morris, et al., 1984). As summarized by Turk and Okifuji (1997), “physicians appear to believe that behavioral demonstrations of pain, such as limping and grimacing, indicate something important about the nature of the patient’s pain and the need for prescribing specific treatments such as surgery and opioids” (p. 334).

Patients’ Knowledge of Clinicians’ Expectations of Pain Behaviors

Interviews with patients who had used intravenous patient-controlled analgesia (IV PCA) for opioid administration after surgery revealed that a major reason for patients’ valuing PCA was related to the fact that PCA decreased the need to interact with the nursing staff regarding pain (Hall, Salmon, 1997; Taylor, Hall, Salmon, 1996a, 1996b). Not only did patients see PCA as being better than waiting for a nurse to administer opioids intramuscularly, they also believed it protected them from having to show distress to the nurses.

Apparently many patients with pain are aware of the behaviors expected of them, or they learn them quickly. Patients may learn from early childhood experiences with pain, television, the responses of their clinicians, and a variety of other sources how to behave to signal others that pain exists and that help is needed. It is a common observation that patients with pain change their behavior in the presence of clinicians and in other selected circumstances. Patients may appear to be calm and to read or have lively telephone conversations, but as soon as clinicians enter, patients may replace these activities with a solemn facial expression and may even grimace, moan, and restrict movement—just the behaviors clinicians want to see when patients report pain, but not necessarily the behavior patients prefer.

Consider what might happen if the patients in the Andrew-Robert survey were roommates. It would not take long for smiling Andrew to realize that grimacing Robert was receiving better pain relief, and the reason probably would be quite apparent to smiling Andrew. Andrew might take pride in looking energetic and happy in front of his visitors or may find that such distraction is very helpful as a coping mechanism. But if he wants better pain relief, he may decide to change that behavior, at least during the time clinicians are present. When clinicians see this change occur, they often regard patients as being manipulative, not realizing that it is the expectations that clinicians convey to patients that cause this behavior change.

Taking this hypothetical situation further, by succumbing to expectations of pain behavior to obtain relief, Andrew may begin to jeopardize his recovery. For example, both patients might be told to ambulate. Smiling Andrew may walk the hall until he begins to hurt and then go to the nurses’ station and ask for something for relief. The staff may feel that if Andrew can be this active, he could not hurt enough to require an analgesic. Grimacing Robert, on the other hand, may remain in bed grimacing rather than ambulating and may also ask for an analgesic. Robert may very well be more likely to receive the analgesic than Andrew. Andrew may then learn to stay in bed, prolonging his recovery time and increasing the risk of complications, but increasing his chances of receiving pain relief.

Comparable circumstances were observed in a study of cancer patients (Cleeland, Gonin, Hatfield, et al., 1994). The more active patients were more likely to receive inadequate pain management.

Because clinicians differ in how they expect patients to behave in response to pain, patients may have difficulty learning which behaviors will effectively convince which clinicians. An analysis of staff and patient behavior in an orthopedic unit revealed that the staff assessed pain by observing patients’ behaviors and that their expectations about how patients should express their pain varied within and between shifts (Wiener, 1975). Some, but not all, patients were adept at reading the explicit and implicit cues given by staff and changed their behaviors accordingly. Patients sometimes felt forced to use tactics that they believed were unacceptable but were expected by the staff and were necessary to obtain pain relief.

Expecting patients to behave in certain ways to verify their pain becomes more confusing when clinicians are especially particular about the intensity or type of behavior that a patient should display. Sometimes clinicians say that a patient does not appear to be in pain, but when a patient exhibits pain behavior, clinicians may say that a patient is making too much fuss about the pain.

Correction of the Misconceptions Concerning the Acute Pain Model

One of the most interesting aspects of the misconception that the patient’s behavioral response is more reliable than the patient’s self-report of pain is that it is totally illogical. Virtually every human being has had the personal experience of trying to hide pain and to function in spite of it, smiling when appropriate and even deliberately using humor as distraction. One of the oldest maxims in health care is that laughter is the best medicine. Why couldn’t the nurses and college students responding to the Andrew-Robert survey use this folk wisdom and their own personal experiences with pain to realize that patients with severe pain most certainly may smile and laugh? Nurses are probably the largest group of clinician advocates for nondrug pain relief measures, of which distraction and laughter are highly ranked. And no research has ever even suggested that smiling and joking are incompatible with feeling pain. Why clinicians seem to have such difficulty accepting and acting on reports of pain from smiling and active patients remains a mystery, at least to these authors.

A substantial amount of research refutes the value of the acute pain model, showing that neither behavioral indicators nor physiologic responses are dependably related to the intensity of a patient’s pain. Physiologic indicators appear to be even less valuable than a patient’s behavioral responses to pain.

Most nurses seem to have been taught to use elevated vital signs to assess or verify the presence of pain, especially severe pain. However, in literature reviews, investigators found very little research that supported using physiologic manifestations as specific indicators of pain (Herr, Coyne, Key, et al., 2006; van Cleve, Johnson, Pothier, 1996). In a study of 1063 patients admitted to the ED with conditions that could be verified to be painful, such as fractures and nephrolithiasis, no clinically significant associations were identified between pain scores self-reported on a scale of 0 to 10 and heart rate, BP, or respiratory rate (Marco, Plewa, Buderer, et al., 2006). In a study published in French, investigators found that absence of increased vital signs does not mean absence of pain (reported in Gelinas, Johnston, 2007).

Critical care nurses usually consider vital signs to be relevant indicators of pain, possibly because the signs are readily available, and it is often difficult to obtain pain reports from critically ill patients, whose behavioral responses may be compromised and who may be unconscious, intubated, or otherwise unable to communicate. In a study involving 30 conscious patients in an intensive care unit, no significant relationship was found between physiologic indicators and patients’ self-reports of pain, reinforcing the fact that physiologic indicators should not be considered primary indicators for pain assessment (Gelinas, Johnston, 2007). In other words, physiologic indicators were not related to patients’ reports of pain.

In another study of 755 patients, primarily in intensive care units, physiologic responses were monitored while the patients were undergoing tracheal suctioning (Arroyo-Novoa, Figueroa-Ramos, Puntillo, et al., 2007). Statistically significantly higher increases were noted in heart rate and systolic and diastolic BP, but these changes were not clinically significant. The authors suggest that methods of measuring these physiologic parameters may not be sensitive enough to capture the response to acute pain. For further discussion of the limited usefulness of physiologic measures in assessing pain, see the discussion of patients who are critically ill in Chapter 4, pp. 143-147.

The usefulness of physiologic measures of pain is further compromised by the presence of dementia in older adults. In a small study of adults 65 years old and older, 50 of whom were cognitively intact and 44 of whom had varying degrees of dementia, heart rate and other physiologic responses were monitored before and during venipuncture (Porter, Malhotra, Wolf, et al., 1996). Increasing severity of dementia was associated with the blunting of physiologic responses as measured by diminished heart rate increase in the preparatory phase and during venipuncture.

Thus, little research supports vital signs as being relevant indicators of pain, although they can be used as indicators of the need for further assessment of pain. A position statement by the ASPMN regarding assessment of pain in nonverbal patients (available at http://www.aspmn.org/Organization/documents/NonverbalJournalFINAL.pdf) suggests that physiologic indicators should not be used alone to assess pain but should be considered cues for further assessment of the possibility of pain (Herr, Coyne, Key, et al., 2006).

Considerable research demonstrates that the behavioral expressions of pain that clinicians expect to see in patients are often absent. In the late 1970s, investigators interviewed 102 adult patients with various types of pain, acute and persistent (Jacox, 1979). Many patients did not report pain and made strong efforts to conceal it. When the patients were asked whether they discussed pain with others, 70% said no or were ambivalent. When they were asked how they responded to pain, 66% said they tried to remain calm and not show pain. More than a decade later a similar finding was reported in a study of 45 patients with pain related to lung cancer (Wilkie, Keefe, 1991). Of the patients, 42% revealed that they coped with pain by trying not to let others know about it (Wilkie, Keefe, 1991).

Some researchers have attempted to use facial movements as a method of studying malingering. A review of the literature found that the results of these studies were inconsistent and concluded that it is unclear whether facial expressions of pain can be used as a reliable method for identifying malingering (Fishbain, Cutler, Rosomoff, et al., 1999). A few years after this literature review, facial expressions were studied in 40 patients with low back pain who were videotaped during rest and painful straight-leg raises (Hill, Craig, 2002). During painful movement they were asked to express their pain genuinely or to pretend that it did not hurt. Without moving, they were asked to fake pain. Although distinctions could be made between faked and genuine painful facial expressions, this research confirmed the difficulty of discriminating between them.

Patients with pain may deliberately engage in certain behaviors that are incompatible with those of the acute pain model but are helpful in coping with pain. In a study of 13 patients with pain related to advanced cancer, the patients reported that behaviors they used to control their pain included watching television (9 patients) or chatting with family and friends (Wilkie, Lovejoy, Dodd, et al., 1988). A questionnaire survey of 53 patients with persistent cancer pain asked them to identify and rate the effectiveness of the self-initiated, noninvasive pain control measures they used to cope with their pain (Fritz, 1988). Patients rated laughing as being the most effective.

Recently, researchers have identified that patients actually smile during painful situations, yet they did not find one tool for behavioral observation to assess pain that included smiling (Kunz, Prkachin, Lautenbacher, 2009). The facial expression of pain has been studied in both experimental and clinical research, and facial expression is often a part of behavioral observation tools to assess pain in both children and adults who cannot self-report. When these researchers viewed several videotapes of patients undergoing painful stimulation, one of the most unexpected findings in response to pain was the oblique raising of the lip that results in a smile. Their research revealed that the percentage of patients who smile at least once during painful stimulation ranges from 22% to 57%. A satisfactory explanation for this requires more research, but clinicians should not discount pain just because the patient smiles. Among other things, the smile could represent embarrassment over other behavioral responses to pain, an attempt to mask feelings of pain, or a willingness to endure the pain.

Sleep may be mistakenly equated with lack of pain, but even patients with severe pain may sleep. Some patients use sleep to help control their pain (Wilkie, Lovejoy, Dodd, et al., 1988). In one study, 100 patients were interviewed about the experiences of pain and sleep following abdominal surgery (Closs, 1992). Pain was the most common cause of sleep disturbance at night, demonstrating that pain occurs during sleep. Analgesics helped more patients get back to sleep than any other intervention. Also, about half of the patients felt that pain was worse at night.

Further, an appreciation of the fact that sleeping patients may have pain is demonstrated in analgesic research. When the effectiveness of analgesics is studied, trained observers ask patients to rate their pain at specific intervals, such as every hour, after the administration of the analgesic. When the observer finds the patient asleep at the time a pain rating is required, the observer awakens the patient to obtain a pain rating (Forbes, 1991).

This information does not necessarily indicate that in clinical practice sleeping patients should always be awakened to assess pain. If a patient has been given an analgesic and assessment of pain ratings afterward has shown that the analgesic has been effective, when the patient is given further doses, there usually is no need to awaken the patient to assess pain rating following a dose. Pain assessments can be made when the patient awakens. However, it may be wise to awaken the patient for the next analgesic dose if the analgesic lasts only 4 hours and the patient has continuous pain. This option can be discussed with the patient, explaining that if he or she is allowed to sleep beyond 4 hours, the pain may return to awaken him or her. If this happens, pain control is jeopardized. The patient must notify the nurse or obtain the analgesic on his or her own, take the analgesic, and then wait for it to be effective. In this scenario, a patient’s sleep is interrupted for longer than it would have been had the patient been awakened at 4 hours and given an analgesic before pain returned. Once this is explained to a patient, he or she may opt to be awakened or to wait to be awakened by pain. If the latter is chosen, a patient should be cautioned to notify the nurse immediately so the analgesic can be given. (If a patient’s sedation levels needs to be assessed, then it may be wise to awaken the patient. See the discussion of sedation levels on pp. 510-511.)

Very little research has been done to evaluate pain control using around-the-clock (ATC) dosing of analgesics versus administering them as needed, and the results are often inconclusive. However, one small study compared these two types of dosing schedules in 35 patients following abdominal surgery (De Conno, Ripamonti, Gamba, et al., 1989). Patients who received ATC analgesia had significantly better pain control, slept longer, and spent more time out of bed rather than lying down over the first postoperative week. Another study of medical inpatients compared ATC dosing of opioids to a control group and found that patients with ATC dosing reported lower pain intensity ratings than those reported by the control group yet did not take higher doses than the control group (Paice, Noskin, Vanagunas, et al., 2005).

Sedation is erroneously equated with analgesia. However, in a study of sedation and pain relief in the postanesthesia care unit, researchers found that opioid-induced sedation did not ensure adequate self-reported pain relief (Lentschener, Tostivint, White, et al., 2007). About half of the 26 patients who experienced opioid-induced sedation had persistently high pain scores in the postanesthesia care unit and during the initial 24-hour postoperative period. Further, morphine-induced sedation did not suppress the patients’ memories of early postoperative pain. Still another study confirmed that opioid-induced sedation could not arbitrarily be equated with analgesia (Paqueron, Lumbroso, Mergoni, et al., 2002). Of patients receiving intravenous morphine in whom morphine was discontinued because of sedation, 25% still had pain levels on the Visual Analog Scale (VAS) above 50. (In patients receiving morphine whose pain relief is unacceptable, a multimodal approach to analgesia should be considered; see pp. 9-10.)

Many sedating drugs, such as benzodiazepines and phenothiazines, are given to patients who are experiencing pain, but most of them provide no analgesia, and the resulting sedation may limit the amount of opioid that can be given safely to a patient in pain. Except for pain related to muscle spasm, benzodiazapines do not relieve pain. Further, available phenothiazines neither relieve pain nor potentiate opioid analgesia (APS, 2003).

When a patient’s report of pain is not accepted and acted on, an effective strategy is to ask, “Why is it so difficult to believe this patient has pain?” When the problem revolves around expecting the acute-pain model to exist, the answer is likely to be “He doesn’t look like he’s in pain.” When you find yourself thinking this or hearing someone say it, try asking, “How would this person have to act for us to believe he has pain?” A clinician might answer that a patient in that much pain would grimace or be less active. Unfortunately, expecting a patient to “act like he is in pain” may lead to those behaviors and contribute to manipulative behavior and physical harm such as disability or complications resulting from decreased function.