US Pharm. 2007;32(5):37-52.

In 1999, the Joint Commission on Accreditation of Healthcare Organizations (JCAHO) introduced new pain management standards. The goal of these standards was to make more health care professionals cognizant of the subjective nature of pain and to improve the efficiency of pain assessment interviews and the adequacy of management strategies. While improvement has been seen in many areas, the literature reveals that some health care professionals ineffectively assess and treat pain in the elderly.1

Adequate pain management should comprise both a thorough pain assessment and a highly effective and safe treatment strategy. Many factors contribute to the difficulty in obtaining an adequate assessment of pain in the elderly and providing optimal treatment to the patient.2 Among these factors are barriers to conducting a sufficient pain assessment interview, problems deciding which assessment tool to use based on a patient's degree of cognitive function, inadequacy in comprehending the important differences among tools, and choosing an inappropriate treatment strategy.1,2 This article will provide a review of the complexity of pain assessment in the cognitively impaired patient and improve the pharmacist's ability to assess pain.

Barriers to Pain Assessment
A thorough pain assessment is essential for effective pain management. If inaccuracies are made while assessing pain, it will be impossible for an optimal management regimen to ensue. Health care professionals must be aware of barriers that might negatively impact pain assessment, and they should make every effort to resolve these issues. Three main categories of barriers have been described in the literature: those concerning the patient or the patient's family/caregiver; those among health care professionals; and those within the health care system.2 Barriers conveyed by these three distinct groups have a vital role in the inadequacy of pain assessment and management. Identifying, anticipating, and resolving such barriers are important steps in improving a patient's pain experience.

The most prevalent barriers involving elderly patients, families, and/or caregivers include psychocultural biases and fears of side effects, addiction and tolerance, loss of independence, and medication costs.2-5 Psychocultural biases can be described as beliefs influenced by one's opinions or background. For example, many elderly patients falsely believe that pain is a normal process of aging. 2 This view can lead to an unintentional poor pain assessment by the health care professional, since the patient may be reluctant to admit to a "normal process." It is vital to inform patients that even though many disease processes such as osteoarthritis may be a consequence of aging, being in pain is not. Reinforcing the JCAHO's stance on the right of the patient to receive adequate pain management may also prove to be beneficial. This is especially true of patients who falsely believe that seeking pain relief is indicative of being a "weak" or "bad" patient. Health care professionals should be mindful that many elderly patients are extremely reluctant to report pain.4 Patients fear that a host of tests will lead to a diagnosis that may result in a loss of independence.2 Health care professionals should be empathetic to patients' fears and maximize patient education in order to correct misconceptions. When a treatment strategy is chosen, health care professionals should make every effort to include patients and their caregiver(s) as members of the pain management team. This will help to ease some of the anxiety associated with some treatment options. For example, patients may experience apprehension if treatment with opioids is necessary. 2 Identifying and discussing myths regarding opioid-related side effects and educating about the small incidence of addiction may allow patients to understand their regimen and become more comfortable with it. Patients should not fear addiction to the point where pain is left untreated. Lastly, health care professionals should anticipate that uninsured or underinsured elderly patients may fear health care or medication costs and become extremely apprehensive about placing the burden on others.2 This fear may govern a personal decision to not seek pain relief. Because psychocultural biases are unique to individual patients, discussing beliefs about pain and educating patients prior to initiating therapy will eradicate misconceptions and lead to a more comprehensive pain assessment interview.

Health care professionals may harbor prejudices about pain that can affect the success of pain assessments. Some health care professionals might fixate on opioid addiction and lack knowledge about it when assessing pain and choosing medications to treat pain. 4 These are among the most prevalent barriers affecting pain assessment and management. An overly conscious fear of addiction might hinder the selection of an appropriate treatment, such as opioid analgesics. However, research reveals that less than 1% of individuals using opioids develop addiction as a psychological dependence.4 Health care professionals should therefore acknowledge the significance of this research and spend more time meticulously assessing a patient's report of pain and less time anticipating an addiction that may not occur.

Health care professionals who are responsible for assessing patients' pain may inadvertently perpetuate barriers by confusing many pain behaviors. Some health care practitioners routinely confuse addiction with pseudoaddiction.6 The American Pain Society defines addiction as impaired control over drug use, compulsive drug use, and/or continued drug use despite harm.6 Pseudoaddiction, in many ways, is iatrogenic, meaning that it is induced inadvertently when health care professionals provide less than adequate relief of pain.6 For example, health care professionals may interpret the patient's early request for pain medications as a sign of possible addiction. However, patients with inadequately managed pain may tend to display "drug-seeking" behavior. Understandably, these patients may ask for their medications earlier than patients whose pain is adequately managed in an effort to gain relief. Therefore, health care professionals should research whether or not the initial pain assessment was comprehensive, whether the chosen treatment is optimal to manage the type of pain the patient is experiencing, whether medications are inadvertently being administered at intervals longer than the proven analgesic duration, and whether the patient has either a new injury or a new diagnosis that may generate or exacerbate pain.

In addition, some health care professionals confuse tolerance with pseudotolerance. Tolerance can be defined as needing increased amounts of drug to produce the same analgesic effect. 6 However, this increased need for medication should not be viewed as an addiction. For example, patients with type 2 diabetes may eventually need more exogenous insulin due to cell resistance. This does not mean that the patient is becoming "addicted" but rather that exposure to the drug results in changes that lead to a reduction in the drug's effects. Therefore, there is a need for more insulin to produce the same glycemic effect. In pseudotolerance, medication increases cannot be owed to a reduction in the drug's effects.6 Other factors, such as disease progression (i.e., radiating pain) or new disease (i.e., new injury), must be taken into consideration.6 Performing a thorough pain assessment interview should help to differentiate between the aforementioned phenomena.

Many clinicians share the erroneous belief that it is impossible to assess pain in elderly patients with cognitive impairment, and therefore, it is impossible to treat the pain. 4,7 In fact, a 1998 study by Horgas and Tsai concluded that the more disoriented, withdrawn, and functionally impaired nursing home residents were, the fewer the number of analgesics that were prescribed and administered.7 Pain assessment is complicated by its subjectivity. In fact, no one definition of pain clearly explains its multidimensional and complicated aspects. In 1968, Margo McCaffery best defined pain as "whatever the experiencing person says it is, existing whenever he/she says it does." 4 This flaccid definition represents an unbiased view of pain, which the pharmacist must possess.

Barriers involving the health systems have decreased substantially since the JCAHO standards were introduced, explaining the patient's right to adequate pain management. As a result of the guidelines, there has been more of an emphasis on the importance of pain assessment and management within institutions. However, systemic barriers still include a lack of institutional standards, a lack of systematic assessment, a lack of visibility of pain documentation, and the placement of pain at a lower priority level than other conditions.8

Pain Assessment Interview
Key questions have been developed to help conduct a thorough pain assessment interview in all patients. These questions focus on the intensity, location, and character of pain. Useful mnemonics such as "ABCDE" and "PQRST" prompt those conducting the pain interview to the pertinent questions that should be asked. Table 1 describes these mnemonics in more detail.9 The following questions should be included in all pain assessments.4,9
Where is the pain?
When did the pain start?
How often does the pain occur?
What does the pain feel like?
Has pain intensity changed?

Determining what makes the pain worse or better, whether the pain affects physical and social function, and what type of treatments have worked in the past is vital to adequate pain assessment interviews. For the cognitively impaired patient, questions asked during the pain interview should be limited to those with simple yes or no answers since changes in mental status may impair the patient's ability to process and relay information.10 Health care professionals must maintain keen pain assessment strategies and tailor pain assessments to meet individual patient needs based on the patient's level of cognition.

During the assessment of pain, health care professionals should also be cognizant of the effects of untreated or undertreated pain on patient behavior. Paying attention to changes in patient behavior is extremely important when dealing with cognitively impaired or non-verbal patients. A paradigm of inadequate pain management exists and includes negative effects on physical function, psychological morbidity, and social or societal sequelae.11 As a consequence of this, appropriate individuals should assess possible sleep disturbances, depression/anxiety, decreases in activities of daily living, disability, and changes in relationships with friends or family.11 Being aware of the sequelae of untreated or undertreated pain will lead to a more thorough pain assessment.

Pain Assessment Tools

Pain assessment tools are used to quantify a patient's report of pain. Accurate use of these tools is critical to optimal medication selection. There are numerous pain assessment instruments that evaluate pain intensity, pain location, pain behaviors, or a combination of these factors.1 These assessment tools can be further classified as unidimensional or multidimensional depending on the number of pain dimensions measured.

Health care professionals should not expect all pain-assessment tools to have the same rate of success in all patients. Data show that the usefulness of many tools declines in the cognitively impaired geriatric population.1,7 Because there is difficulty in using certain tools in cognitively impaired patients, discrepancies governing the successfulness of use exist.3 Many elderly patients are eliminated from studies that test clinical tools due to the difficulty noted in assessing pain in this population. As a consequence, very little definitive information is available on the usefulness of these instruments in patients with moderate to moderately severe cognitive impairment.7 It is important for health care professionals to be educated about the most widely used tools in clinical practice. Optimal tools are those that consistently produce similar results when readministered in the same patient (validity), have been evaluated as an appropriate representation of a patient's report of pain (test-retest reliability), have consistency in interpretation and scoring among/between raters (reliability), and are easy to administer.1 Therefore, a review of recent literature dealing with the use of different pain assessment tools in cognitively impaired elderly patients is warranted.

Pain Intensity: Self-report scales measure subjective pain intensity, are typically unidimensional, and include verbal descriptor scales, numeric rating scales, visual analog scales, and face scales. These self-report scales are the most common methods of assessing pain in clinical practice.7 They have demonstrated reliability and validity and are considered sensitive measures of pain intensity in verbal, cognitively intact older adults.7

The verbal descriptor scales exist in many different forms but maintain the consistent concept of providing a small list of adjectives to describe pain intensity. The advantages of verbal descriptor scales are that they are easy to understand for many patients, are quick to administer, and apply to different types of pain.7 The patient reads the descriptors, such as "no pain," "mild," "moderate," or "severe," and chooses the best one. In some instances, the clinician reads the list of adjectives and has the patient choose the one that best describes their current pain severity. One disadvantage is that some patients may not describe their pain using the exact adjectives used in the scales. For example, some patients may not admit to having pain but may respond better to terms such as "aching," "soreness," or "hurting." 7,12 For these patients, reports of pain intensity may not be very accurate.

To alleviate the problem of too few verbal descriptors, other descriptor scales with more adjectives may be used. The McGill Pain Questionnaire (MPQ) is one alternative self-report tool that uses more words to describe a patient's pain. The MPQ is a multidimensional scale that measures three dimensions of pain and contains several sections in addition to an extensive list of adjectives to describe pain. Difficulties in using this scale include its length and high literacy level.13 The reliability and validity has not been reported for elderly patients.13

Several studies have assessed the effectiveness of verbal descriptor scales in cognitively impaired patients. The general consensus of the studies is that while patients were consistent in reporting pain, there was not strong test-retest reliability associated with these methods.7 For example, the patients would consistently report that they had pain, but the adjectives chosen to describe the pain would vary when the test was repeated. The probable cause of this discrepancy is that cognitively impaired elderly may have difficulty in choosing one word to quantify their pain.12 This concept is very abstract and may be too difficult for a cognitively impaired patient to interpret.12 Therefore, health care professionals who use these scales for cognitively impaired patients should be conscious of this possible problem and use caution when interpreting the results. Use of this tool in cognitively impaired patients may lead to an incomplete pain assessment.

The simplest method, and perhaps the most universally used scale measuring pain intensity, is the numeric rating scale (NRS). This scale contains a numeric rating from 0 (indicating no pain) to 10 (indicating the worst pain imaginable), although different versions exist. The numeric scale, like the verbal descriptor scale, is easy to administer and can even be administered over the phone. This scale has been shown to be both valid and reliable in cognitively intact elderly individuals.7 There is conflicting literature relating to the use of this pain scale in cognitively impaired elderly patients.7 One reason for this conflict is a lack of standardization in clinical studies when interpreting Mini-Mental State Examination (MMSE) criteria to categorize levels of cognitive impairment. Some studies haven been conducted using patients with mild-to-moderate impairment, while others may have been conducted using patients with severe impairment.7 Some patients with mild-to-moderate cognitive impairment successfully used the numeric scale, while others with lower scores on the MMSE had extreme difficulty using the scale. The same reasoning for difficulty in using the descriptor scales can again be used to explain difficulty using the NRS. Choosing a number and directly correlating that number with one's pain is a complex task, which may be too difficult for the cognitively impaired patient to comprehend.7

The visual analog scale (VAS) is another type of self-report scale that is widely used to measure pain intensity. The VAS consists of a single 100-mm line with one end labeled "no pain" and the other end labeled "most pain imaginable." This scale is extremely difficult to use in the elderly with cognitive impairment. Individuals must be able to think abstractly and coordinate sensory-motor and visual skills. Successful use of this scale depends on placing a pencil mark on the line to accurately reflect one's pain intensity.7,12 Because the VAS is so complex, it is not an effective self-report tool for measuring pain in older patients with cognitive impairment.

Two face scales that are currently in use are the Wong-Baker FACES Pain Rating Scale (Figure 1) and the Faces Pain ScaleñRevised (FPS-R). Both were first developed for assessing pain in children and display either a series of faces with smiling to tearful, frowning expressions or a set of faces showing different degrees of distress.

When using the FACES Scale, the patient points to the face that best describes the severity of pain that is being experienced.14 For example, a smiling face is indicative of no pain and a face with tears indicates the worst pain possible. A numerical value is also associated with each face. This scale, in particular, has been criticized for assessing emotion rather than pain. Patients who interpret the FACES Scale as a way to assess emotion may have difficulty using the scale, because the expression of emotion is gender and culturally based. 7 Crying is not equally acceptable between genders and may not be an acceptable way of dealing with pain in some cultures.7,12 Moreover, a smiling face may not directly correlate to a report of "no pain." Although the patient may be in pain, he or she may be very satisfied with how the pain is being perceived and managed by the health care staff. As a result, the patient may mistakenly mark a smiling face to show approval of the pain management team.

In comparison, the FPS-R uses six to seven faces arranged with increasing looks of distress from "0 = no pain" to "6 = worst pain imaginable." One advantage of the FPS is that it is geared more toward adults than the FACES Scale. Some adults may respond better to faces that are more adultlike than to the animated happy/sad faces of the FACES Scale. Moreover, the rating system of the FPS-R does not confuse emotion with pain severity. The major disadvantage of using face scales for the cognitively impaired is that as cognitive impairment increases, the ability to use the scales decreases.7,15 Patients with severe cognitive impairment may point to the face they like the best rather than the one that correlates with pain intensity. Therefore, it is reasonable to conclude that the face scales are reliable and valid in patients with mild-to-moderate impairment, but not in those with severe impairment.7,16

Pain Location: Using both a pain map, which is typically a picture of a person or a doll, and one's own body allows the patient to show where the pain is located. Pain maps only allow for identification of changes in or consistency of pain location. They do not allow assessment of pain intensity. Cognitively impaired patients have had great success in using location instruments.7,17 Combining a pain intensity scale and a location tool might provide optimal pain assessment in the elderly.

Pain Behaviors: Behavioral pain scales are used most often in cognitively impaired patients in long-term care facilities. It is well known that untreated or inadequately treated pain can lead to behavioral disturbances.1,12 The American Geriatrics Society (AGS) has identified potential behaviors that may be an indicator of pain in cognitively impaired patients: facial expressions, negative verbalizations/vocalizations, specific agitated body movements, changes in interpersonal interactions, changes in activity patterns or routines, and further changes in mental status.1 A great deal of research has been conducted identifying behavioral patterns associated with pain, since addressing such factors could lead to an improvement in quality of life.1,12 Several tools have been developed to help with the identification of these behaviors. Although numerous scales exist, five types of behavioral tools will be discussed in this review: the Discomfort Scale for Dementia of the Alzheimer's Type (DS-DAT); the checklist of non-verbal pain indicators (CNPI); the Face, Legs, Activity, Cry, and Consolability (FLACC) scale; the Pain Assessment in Advanced Dementia (PAINAD) Scale; and the Assessment of Discomfort in Dementia (ADD) protocol.1,12 Ideal behavioral tools are those that have sound evaluation of usefulness, reliability, and validity, as well as equal efficacy when compared to other reputable tools.1 No specific tool is currently recommended for use by the AGS.12

The DS-DAT is a nine-item list requiring five minutes of observation to rate nonverbal behavior that is predictive of distress in patients at rest. Items measuring frequency, intensity, and duration are correlated with a score that corresponds to a patient's level of discomfort.2 This tool was developed with the severely cognitively impaired nonverbal patient in mind.1 Before observation begins, the rater must wait 15 minutes after an intervention that may have caused pain or discomfort for the patient, such as a position change or feeding. One limitation of this tool is that only trained professionals can administer the test; another is that it is extremely time consuming and complicated.1

Since the invention of the DS-DAT, a modified version, which contains only six items, has become available. Three of the items that were reportedly hard to interpret--sad facial expression, frowning, and relaxed body language--were removed.7 Although this scale was found to be just as useful as the original version, it too is complicated, time consuming, and requires expertise. Both tools are too labored for daily clinical practice, but have been shown to be extremely beneficial when assessing pain in clinical studies.7 Extensive training to guarantee consistent interrater reliability must be achieved before routine use in clinical practice is justified. Researchers also criticize that a scale that measures discomfort may not accurately and consistently represent pain. Their rationale is that not all conditions that cause discomfort result in pain.1

The CNPI is a six-item scale designed to measure pain behaviors in postoperative cognitively impaired adults both at rest and on movement. Like the DS-DAT, items are scored for an interpretation of pain severity. A study comparing verbal descriptor scales with the CNPI revealed a significant correlation in scoring among cognitively intact and cognitively impaired older people.7,18,19 As evidenced from clinical trials, facial grimacing and restlessness are the most easily identified behaviors among raters. The CNPI measures these behaviors, whereas other behavioral scales do not. This tool is both reliable and practical for assessing pain behaviors in both cognitively impaired and cognitively intact elderly patients in everyday clinical practice.18,19

The FLACC scale was first developed for use in nonverbal children, and its use has been validated in children ages 3 months to 7 years (Table 2).2,20 The tool includes five items that are scored and interpreted. Recently, this scale has been proposed for use in cognitively impaired elderly patients. However, many of the pain behaviors observed on the scale have not been reported in elderly patients with dementia and may not be representative of pain behaviors in this group.1 Preliminary data suggest that the use of this tool in elderly patients with dementia produces inconsistent reliability and validity. 1 Therefore, the routine usefulness of this scale in clinical practice cannot be substantiated.

The PAINAD scale (Table 3 ) was developed by the same team of professionals who developed the DS-DAT. 21 The purpose of the PAINAD scale was to avoid the limitations experienced with the DS-DAT.1 The PAINAD scale was developed using elements of the VAS, the FLACC scale, and the DS-DAT. It consists of five items: breathing, negative vocalization, facial expression, body language, and consolability. An objective description of each of the five items is outlined in the scale. Total scores range from 0 (no pain) to 10 (maximal pain). The tool may be simpler to understand and easier to use with limited training than the DS-DAT.1 In comparison to the FLACC scale, it includes more common pain behaviors displayed by the cognitively impaired elderly patient. Comparative studies between the DS-DAT and PAINAD scale have shown that the results of different raters are markedly similar, providing evidence of validity.1 The PAINAD scale has been criticized for not being able to detect pain in patients with subtle reactions to pain or subtle changes in pain behaviors. Additional research is needed to improve internal consistency among raters, though tool reliability is good for individual raters. This scale would be most useful for assessing pain in cognitively impaired geriatric patients who demonstrate significant changes in behavior secondary to pain.

The ADD is a protocol developed to assess and manage discomfort in long-term care facilities.1,7 This clinical tool consists of four steps, and assessment begins when a resident's behavior changes. The first step is to look for physical problems as a cause of behavioral changes. A physical examination should be conducted to rule out possible malignancies and new locations of pain. If no physical causes are noted during the physical examination, the second step is to perform an extensive review of the patient's medical history. If no causes of pain are identified, nonpharmacologic approaches to pain management should be initiated. These nonpharmacologic interventions include physical therapy, behavioral modification, distraction techniques, massage, meditation, transcutaneous electrical nerve stimulation, and acupuncture.4 If the nonpharmacologic intervention does not adequately control pain, the patient can be provided with an as-needed analgesic. Asking family members or caregivers if any treatment has been successful in the past may provide a good starting point for choosing an appropriate analgesic agent. This protocol demonstrates that the comfort needs of patients with dementia can be identified and managed comprehensively.7 While no specific time commitment has been described, it is reasonable to assume that the ADD protocol takes a considerable amount of time; routine use of this protocol in clinical practice may be too cumbersome.Ü Successful use requires making complex clinical decisions; therefore, an advanced education in pain assessment and management is required.1 Preliminary data support the validity of the ADD protocol, but its reliability remains to be established.

The patient's self-report of pain is the single most reliable factor in successful pain assessment. When the health care professional is equipped with an unbiased view and advanced pain assessment strategies, a thorough pain assessment will almost always ensue. Health care professionals must be cognizant of numerous factors that can affect a thorough pain assessment, including cognitive impairment, barriers to care, and nonverbalization. Tailoring the pain assessment to the individual patient and knowing when and how to use different assessment tools will substantially increase the probability of a comprehensive pain assessment. Combining the appropriate tools to identify pain intensity, location, and pain behavior is an appropriate strategy to improve assessment outcome. Health care professionals must be educated about the success and failure rates of widely used assessment tools in clinical practice, and make a habit of choosing the best tool. The accuracy of the assessment tool selection is critical to the adequacy of the pain management regimen.

1. Herr K, Bjoro K, Decker S. Tools for assessment of pain in nonverbal older adults with dementia: a state-of-the-science review. J Pain Symptom Manage . 2006;31:170-192.
2. Binaso K. Pain management for the geriatric patient. Presented at: 37th Annual American Society of Health System Pharmacists Midyear Clinical Meeting; December 11, 2002; Atlanta.
3. Dempsey S. Pain Assessment. Presented at: Annual Florida Society of Healthcare Pharmacists Meeting; August 18, 2000; Orlando.
4. Duquette C. Improving function and quality of life for older adults through pain assessment and management. Presented at: 21st Annual National Conference for Gerontological Nurse Practitioners: Advanced Practice Nursing: Making a Visible Difference; September 2002.
5. Freedman GM. Chronic pain. Clinical management of common causes of geriatric pain. Geriatrics. 2002;57:36-41.
6. American Pain Society. Definitions related to the use of opioids for the treatment of pain. A consensus document from the American Academy of Pain Medicine, the American Pain Society, and the American Society of Addiction Medicine. Available at: Accessed April 20, 2007.
7. Dewaters T, Popovich J, Faut-Callahan M. An evaluation of clinical tools to measure pain in older people with cognitive impairment. Br J Community Nurs . 2003;8:226-233.
8. Dempsey S. Pain assessment. Presented at Florida Society of Health System Pharmacists; August 18, 2000; Orlando.
9. Dipiro JT. Pharmacotherapy: A Pathophysiologic Approach. 6th ed. New York, NY: McGraw-Hill; 2005:1091.
10. Herr K, Decker S. Assessment of pain in older adults with severe cognitive impairment. Ann Long Term Care. 2004;12:46-52.
11. Argoff CE. Managing neuropathic pain: new approaches for today's clinical practice. Available at: Accessed April 20, 2007.
12. American Geriatric Society Panel on Persistent Pain in Older Persons. The management of persistent pain in older persons. J Am Geriatr Soc. 2002;50:S205-S224.
13. Duggleby W. The language of pain at the end of life. Pain Manage Nurs . 2002;3:154-160.
14. Hockenberry MJ, Wilson D, Winkelstein ML. Wong's Essentials of Pediatric Nursing. 7th ed. St. Louis: Mosby; 2005:1259.
15. Krulewitch H, London MR, Skakel VJ, et al. Assessment of pain in cognitively impaired older adults: a comparison of pain assessment tools and their use by nonprofessional caregivers. J Am Geriatr Soc. 2000;48:1607-1611.
16. Chibnall JT, Tait R. Pain assessment in cognitively impaired and unimpaired older adults: a comparison of four scales. Pain. 2001;92:173-186.
17. Weiner D, Peterson B, Keefe F. Evaluating persistent pain in long term care residents: what role for pain maps? Pain. 1998;76:249-257.
18. Lane P, Kuntupis M, MacDonald S. A pain assessment tool for people with advanced Alzheimer's and other progressive dementias. Home Healthc Nurse . 2003;21:32-37.
19. Feldt KS. The checklist of nonverbal pain indicators (CNPI). Pain Manage Nurs. 2000;1:13-21.
20. Merkel SI, Voepel-Lewis T, Shayevitz JR, et al. The FLACC: a behavioral scale for scoring postoperative pain in young children. Pediatr Nurs. 1997;23:293-297.
21. Warden V, Hurley AC, Volicer L. Development and psychometric evaluation of the Pain Assessment in Advanced Dementia (PAINAD) scale. J Am Med Dir Assoc . 2003;4:9-15.

To comment on this article, contact