Assessment Tools

Nursing Admission Assessment

When patients are admitted to a health care facility, such as a hospital or an outpatient clinic, nurses perform general admission assessments. Along with such information as the patient’s self-care abilities and nutritional needs, a section should be included to identify pain problems. Patients may have chronic pain conditions for which they are already receiving treatment, or pain may be the primary reason for admission. Examples of questions that are appropriate for routine admission assessments are shown in Box 3-1.

The purpose of these questions is to identify new or ongoing pain problems. If a pain problem is ongoing and the patient already has an effective pain treatment plan, steps should be taken to ensure that the plan is continued. If a treatment plan must be developed, further assessments using the Initial Pain Assessment Tool (Form 3-1, p. 51) or the Brief Pain Inventory (BPI; Form 3-2, p. 53) may be indicated.

Tools for Initial Overall Pain Assessment

Two tools widely used for overall initial pain assessments are referred to as Initial Pain Assessment Tool (see Form 3-1) and the BPI (see Form 3-2). Both of these were included in the cancer pain clinical guidelines published by the Agency for Health Policy and Research (Jacox, Carr, Payne, et al., 1994a), and the BPI is included in the revised cancer pain guidelines published by APS (Miaskowski, Cleary, Burney, et al., 2005). The patient or the clinician may complete these tools. Another overall pain assessment tool often used in research and clinical practice is the short-form McGill Pain Questionnaire (Melzack, 1987).

These pain assessment tools need not be used for all patients with pain. The policies and procedures of a health care agency must identify the criteria necessary for these tools to be considered. Certainly if the patient has chronic pain problems that are not satisfactorily controlled, an overall pain assessment tool should be completed. Hospices and pain treatment centers may need far more extensive initial overall assessment tools than those offered here. For patients with acute pain that is not easily controlled by the usual pain treatments (e.g., intravenous patient-controlled analgesia [IV PCA] following surgery), an overall assessment also is indicated.

Both the Initial Pain Assessment Tool and the BPI, described later, attempt to assess some aspects of suffering. Suffering, like pain, is subjective, but it goes beyond simply feeling pain. Pain may exist without suffering. Suffering eludes definition but has been characterized as an individual’s experience of threat to self, a meaning given to events such as pain or loss (Kahn, Steeves, 1996). Suffering involves the person’s evaluations of the significance or meaning of pain (Spross, 1993). To some extent, suffering is similar to an impairment in quality of life. Items on the assessment tools directed at the effects of pain on various aspects of living are an attempt to assess some aspects of the patient’s suffering.

Criteria for Selecting Pain Rating Scales for Use in Daily Clinical Practice

Criteria for selecting self-report pain rating scales for use in daily clinical practice are summarized in Box 3-2. Most important, the pain rating scale selected must have been tested and found to be reliable and valid. Reliability means that the scale consistently measures pain intensity from one time to the next, and validity means that the scale accurately measures pain intensity. Probably the most fundamental aspect of validity for pain rating scales is that they demonstrate sensitivity to changes in the magnitude of pain (Herr, Spratt, Garand, et al., 2007). For example, they must be sensitive enough to measure effects of analgesic medication. (For more detailed information about the reliability and validity of a variety of pain measures, refer to Jensen [2003] and Herr, Spratt, Mobily, et al. [2004]).

Usually the main purpose of using pain rating scales in clinical practice is to identify the intensity of pain over time for the purpose of evaluating the effectiveness of interventions to relieve pain. Scales used to obtain self-reports of pain intensity fail to provide other information that is important to assessing pain and determining pain treatment. This is one of the reasons for remembering to use the Initial Pain Assessment Tool (see Form 3-1) or the Brief Pain Inventory (see Form 3-2).

Certain open-ended questions are also used in addition to the pain rating scale. Research with older adults (N = 312) has shown that an open-ended pain question such as “Tell me about your pain, aches, soreness, or discomfort” when initiating pain assessment discussions provided significantly more pain information than asking the patient to rate his or her pain or simply asking “How are you feeling?” (McDonald, Shea, Rose, et al., 2009). The latter question resulted in the least amount of information about pain. Thus, clinicians should be aware that simply asking a patient to rate the intensity of pain provides useful but not complete information about important aspects of the pain and the implications for treatment.

Unfortunately, pain rating scales seem to invite creativity. Sometimes colors are added to a preexisting scale such that different colors are used to indicate different intensities of pain, or colors are added to a faces scale. Some of the problems with this practice are that some health care facilities do not have copiers that duplicate in color and that research into colors as indicators of pain intensity is limited. Colors are subject to personal preferences and cultural differences and are best avoided. Another change to a preexisting valid and reliable scale is drawing hats on a faces rating scale. When a validated and reliable scale is changed in any of these ways, additional research is required to establish the reliability and validity of the new scale. Sometimes clinicians develop and use an entirely new scale without testing it for reliability and validity. Such energy is better spent on promoting the routine use of a single, simple pain rating scale that has already been established as being valid and reliable.

As stated in Box 3-2, the tool also should be sufficiently graded so it can capture changes in pain intensity and should be easily understood, easy to score, and liked by patients and staff. The tool should place a low burden on staff; that is, it should be quickly explainable and easily scored. It should also be inexpensive and easy to duplicate. The tool should also be appropriate for patients of various cultures and available in several languages.

Principles of Using Pain Rating Scales

Whichever pain rating scale is chosen for a patient should be used consistently with that patient. Using the same pain rating scale makes it possible to compare the pain ratings. Also, using the same scale is obviously easier for the patient than switching from one scale to another. Further, using the same pain rating scales and the same metric (that is, 0 to 10) throughout the institution facilitates communication about patients’ pain. Standardization is helpful within a given health care system, so that the same scale is used in the emergency department, the hospital, and the outpatient care services. A standard pain rating scale minimizes confusion for both patients and staff.

The primary reason for using pain rating scales is to evaluate the effectiveness of pain treatment plans. Reassessment is essential. The frequency with which pain ratings are obtained depends on the situation. The Guideline for the Management of Cancer Pain in Adults and Children (Miaskowski, Cleary, Burney, et al., 2005) recommends that at “a minimum pain should be reassessed and documented:

When pain is out of control, such as a 10 on a scale of 0 to 10, and a rapidly acting analgesic is used, pain ratings every 5 minutes may be appropriate. However, at the beginning of treatment, asking the patient for a pain rating may be entirely inappropriate. When a patient is obviously in pain or is not focused enough to learn to use a pain rating scale, pain treatment should proceed without pain ratings. Once a patient can use a pain rating scale, pain assessment may be necessary as often as every 15 minutes until pain is brought under control, then every 1 to 2 hours for 24 hours, followed by every 4 hours. Once pain has been well controlled, however, in hospitalized patients, obtaining pain ratings every 8 to 12 hours may be sufficient; in nursing homes a daily or even weekly pain assessment may be appropriate. In home care, once pain has been controlled, patients are not ordinarily asked to keep records of pain ratings.

An intricate part of using pain rating scales to evaluate pain management is to set goals for treatment. These may be referred to as comfort-function goals and are discussed later. Basically, it is essential to know what pain rating is necessary for individual patients.

Teaching Patients and Families How to Use a Pain Rating Scale

It is not unusual for clinicians to claim that patients cannot use pain rating scales. However, even in cognitively impaired elderly patients in nursing homes, 83% who report pain are able to use at least one type of pain rating scale (Ferrell, Ferrell, Rivera, 1995). Another study that included older cognitively impaired minority patients used a verbal descriptor scale (no pain, mild pain, moderate pain, severe pain, extreme pain, and most intense pain imaginable), a 0 to 10 NRS, a FPS-R, and an Iowa Pain Thermometer (Figure 3-9, p. 87) and found that all these scales were relatively easy for these patients to use (Ware, Epps, Herr, et al., 2006). Therefore, most adolescents and adults who are not cognitively impaired should be able to learn to use a pain rating scale. It is simply a matter of having a plan that ensures that when patients are admitted to clinical settings, someone is responsible for teaching all patients about the pain rating scale, having patients demonstrate their understanding of it, and documenting this in the records.

The steps for teaching the 0 to 10 NRS are in Box 3-4 on p. 86. Steps 1, 2, and 3 provide information; steps 4 and 5 ask the patient to demonstrate understanding of the information; and Step 6 focuses on the goals for comfort and function/recovery. Along with each step are specific examples of what may be said to the patient and family. A teaching brochure titled “Understanding Your Pain: Using a Pain Rating Scale” is patterned after the steps in Box 3-4 and is available at www.endo.com.

Patients tend to have a narrow concept of the word pain, often restricting its use to excruciating and intolerable sensations. The patient’s concepts of pain may be expanded by explaining that pain includes several different uncomfortable sensations, such as tightness and pressure, and by using the words aching and hurting (Step 3). In a study of four different ethnic groups—Hispanics, American Indians, blacks, and whites—participants were asked to describe and rate painful experiences. Findings revealed that in all groups, the word ache was used for mild pain, hurt for moderate pain, and pain for the most intense discomfort (Gaston-Johansson, Albert, Fagan, et al., 1990).

Patients Who Deny Pain or Refuse Pain Relief

When a patient denies pain, clinicians must accept this rating because a patient’s self-report is the single most reliable indicator of pain. However, when a patient’s behavior, known pathology, or other findings suggest the existence of pain, clinicians are responsible for exploring this seeming contradiction with the patient and family. Sometimes a patient acknowledges the pain but refuses analgesics. Clinicians must respect this decision but, again, the reasons should be discussed with the patient and family. Giving a patient information or considering other approaches to pain management may result in the admission of pain or the acceptance of measures to relieve it.

Reluctance to report pain or to use analgesics obviously results in poor pain relief. The reasons patients may act this way have been shown to be associated with a variety of specific concerns (Ward, Goldberg, Miller-McCauley, et al., 1993); many further reasons are supported by a survey of laymen’s attitudes (Mayday Fund, 1993). The relationships between patients’ attitudes about pain and its treatment and their willingness to report pain and take analgesics were investigated using the Barriers Questionnaire (BQ), an 8-item tool (Ward, Goldberg, Miller-McCauley, et al., 1993). The patients in the study had chronic cancer pain. The patients’ responses to the BQ revealed that poor pain control was related to concerns about addiction (the most problematic misconception), adverse effects, and tolerance (the need to “save” the pain medicine); the belief that increased pain meant increased disease; concern that complaining about pain would distract physicians from curing illness; and the belief that “good” patients do not complain about pain.

The BQ was later revised to reflect changes in pain management, such as patients’ believing that pain medicine could weaken the immune system, and the result was the 27-item Barriers Questionnaire-II (BQII), a reliable and valid measure of patient-related barriers to the management of cancer pain (Gunnarsdottir, Donovan, Serlin, et al., 2002). The BQ-II confirmed that inadequate analgesia was related to higher scores than those found in the group that used adequate pain medication. Further study of the BQ-II has resulted in the Barriers Questionnaire Short Form, a 9-item survey shown in Form 3-3 (p. 91)(Ward, personal communication, 2008).

The BQ Short Form is recommended for use with patients and their families to help identify what may be causing a patient with cancer to deny pain or refuse analgesics. For patients newly diagnosed with a disease that may become painful eventually, the BQ Short Form may be administered before pain occurs to identify barriers to pain control before problems arise. Depending on the patient population, clinicians may wish to adapt the BQ Short Form by omitting some items or adding others, such as concerns about finances, hiding pain from the family and others, distrust of caregivers, not being a “sissy,” or wanting to avoid highly technologic methods of pain control.

Immediately after the BQ Short Form is administered, clinicians should discuss the results with the patient. The patient and family should be assured that many others have similar concerns, but that the concerns need not be problems. Appropriate patient teaching should follow. However, one session of patient teaching probably will not be sufficient to reassure the patient and family.

Correct information about addiction is almost always needed. One study found that the impact of concern about addiction may be somewhat alleviated by the way the prescriber presents information about the medication (Dawson, Sellers, Spross, 2005). More than half of the patients who received prescriptions for opioids said they would be very likely to take them if the prescribers advised them that it would not be addictive if used according to instructions. However, this statement is not entirely true. Correct use of opioids does not protect patients from developing addictive disease. It is more accurate, as discussed in Chapter 2, to say that that far less than 1% of patients exposed to chronic opioid use will become addicted. Unfortunately, this may not be as reassuring to patients as being told that they simply will not become addicted if they follow instructions. For this reason, another strategy clinicians use is to tell patients to ask themselves a question: If you were not in pain would you want to take this medication? A “no” answer means they have not become addicted. (See Section IV for more information about talking with patients and families about addiction.)

Most literature that teaches patients about pain control covers many of the concerns listed in the BQ Short Form. A teaching brochure concerning taking oral opioids that addresses many of the concerns patients have, including addiction, adverse effects, and tolerance, is titled “Understanding Your Pain: Taking Oral Opioid Analgesics” and is available at www.endo.com. One use of the BQ Short Form is to evaluate patient-education interventions by having the patient or family complete the BQ Short Form before and after information has been provided.

A consistent finding has been that patients who used inadequate analgesics had higher barrier scores on the BQ than did those who used adequate analgesics (Lin, Ward, 1995; Ward, Goldberg, Miller-McCauley, et al., 1993; Ward, Hernandez, 1994). Although misconceptions about pain management exist in patients who do not have pain as well as in those who have cancer-related pain (Yeager, Miaskowski, Dibble, et al., 1997), some research has suggested that these misconceptions may exist because current pain is undertreated (Dawson, Sellers, Spross, 2005). Patients’ beliefs about pain management seem to be shaped in part by the fact that their pain management provides less than adequate pain relief. Thus, not only do patients’ education about misconceptions require attention, so do patients’ pain management regimens.

One of the most frustrating occurrences in pain management is knowing that a patient with pain does not have to suffer, yet the patient refuses help. The BQ Short Form may help clinicians begin the process of exploring why a particular patient has decided to endure pain. Despite receiving accurate information about misconceptions and being given choices regarding methods of pain management, some patients, for cultural or religious reasons, choose to experience pain. These decisions must be respected, and at some point clinicians must refrain from further efforts to change these patients’ minds.

It is interesting that denial of pain is not always deliberate or conscious. Sometimes patients have narrow concepts of pain and deny its presence when asked. Following-up with other terms, such as aching or hurting, may uncover the existence of pain.

Other situations in which pain is denied are related to the energizing effect of visits from physicians or health team members. Before and after such visits, the patient may report pain. But encounters during hospital rounds or clinic visits sometimes have powerful positive placebo effects, and a patient honestly reports feeling fine at that time. Clinicians must remember to question the patient further and to check the patient’s record to determine what occurred before the visit.

Some patients who have experienced pain for a long time seem to lose their frame of reference. They may forget what it is like not to have pain. As one patient reported after reluctantly taking an analgesic, “I didn’t know I had pain. I guess I had been in pain too long to know the difference.”

Sometimes pain is not reported not because of any wish to deny pain, but because the patient believes that the health care provider knows about the pain. The patient may reason that the clinicians knows about his or her disease or surgery and knows that pain is present, so it is unnecessary to tell them. For example, a study comparing adolescents’ pain ratings with their perceptions of nurses’ evaluations of their pain revealed that adolescents perceived that nurses know how much pain they are experiencing (Favaloro, Touzel, 1990).

Selection of Pain Rating Scales for Patients Who Have Difficulty with the Commonly Used Self-Report Scales

As cognitive impairment advances in the older patient, obtaining a reliable pain report becomes more difficult (Hadjistavropoulos, 2005; Kelley, Siegler, Reid, 2008). Many (Pesonen, Kauppila, Tarkkila, et al, 2009) will be unable to use the 0 to 10 pain rating scale previously described but will be able to self-report using other scales. No clear method has been determined to establish a patient’s ability to use a self-report scale reliably, but an attempt should be made. When reliability is an issue, it does not negate the importance of attempting to obtain a self-report in those with cognitive impairment. The MMSE is often used in research but is not readily reproducible for clinical use (Folstein, Folstein, McHugh, 1975). However, another simple tool may be used.

One approach is the Pain Screening Tool (Buffum, Miaskowski, Sands, 2001) (Box 3-5, #5). Focusing on their current pain, patients are asked to rate the intensity on a 0 to 3 scale and to say a word that describes their pain. Following 1 minute of distracting conversation, a patient is asked to recall the number and word. He or she receives 1 point for being able to provide the number and 1 point for providing the word. Half a point is given for recalling the word and half a point for recalling the number. A score of 3 suggests that the patient can use a self-report scale reliably. Most important, a score of less than 3 provides evidence that the patient cannot report pain reliably and that another pain assessment tool is necessary, such as a behavioral observation tool (see p. 126 for a discussion of behavioral tools). Another approach to evaluate whether the patient understands the use of the pain scale is to ask him or her to show on a pain scale where a severe pain would be located and to compare that to where the patient shows a mild pain would be located. This approach is discussed earlier in this chapter in the section about teaching patients how to use pain rating scales (see Box 3-4, p. 86).

Asking about pain in the present, that is “right now,” decreases issues related to memory losses that increase with advancing cognitive impairment. Studies have shown that asking older patients to identify pain severity that is dependent on memory, such as worst pain, least pain, or pain in the past week, may provide unreliable responses. (Bergh, Sjostrom, Oden, et al., 2000; Closs, Barr, Briggs, et al., 2004; Feldt, Ryden, Miles, 1998; Kelly, Siegler, Reid, 2008; Taylor, Herr, 2003). Soliciting reports of current pain and reassessing for change and for response to intervention are recommended. Box 3-5 illustrates strategies that can be used to increase success in soliciting self-report and pain scale use in older adults with cognitive impairment.

For older adults with mild to moderate cognitive impairment, a number of pain intensity scales have adequate reliability and validity (Hadjistavropoulos, Herr, Turk, et al., 2007; Weiner, Herr, 2002). There is wide individual variability in preference for and understanding of format, so no single tool is recommended for all patients. Organizations should identify several options that can be used when assessing pain severity in older patients with cognitive impairment. Tool formats that are acceptable and clinically useful include NRSs (Chibnall, Tait, 2001; Gagliese, Weizblit, Ellis, et al., 2005; Herr, Spratt, Mobily, et al., 2004); verbal descriptor scales (Closs, Barr, Briggs, et al., 2004; Herr, Spratt, Garand, et al., 2007; Herr, Spratt, Mobily, et al., 2004); and faces pain scales, such as FPS-R (Chibnall, Tait, 2001; Kaasalainen, Crook, 2003; Scherder, Sergeant, Swaab, 2003; Taylor, Herr, 2003). These scales are discussed earlier in the chapter (pp. 55-85) and are addressed here in relation to their use in older patients who have difficulty with commonly used scales such as the 0 to 10. Many older adults who are verbal simply have difficulty with scales that provide too many options, such as a 0 to 10 scale, but may be able to use a simpler scale such as a verbal descriptor scale that includes only four options (none, mild, moderate, severe) (Caraceni, Cherny, Fainsinger, et al., 2002; Closs, Barr, Briggs, et al., 2004; Dworkin, Turk, Farrar, et al., 2005). Although a 4-point scale is less sensitive to change, a scale with fewer options is preferable to no self-report at all. A substantial number of older adults (with and without cognitive impairment) have difficulty responding to an NRS, particularly if it is administered verbally and without a visual representation (Kaasalainen, Crook, 2003; Wynne, Ling, Remsburg, 2000). Vertical presentation (see Figure 3-5, A, p. 58) may be easier for persons with alterations in abstract thinking and often is preferred by older adults (Herr, Spratt, Mobily, et al., 2004).

Verbal descriptor scales (e.g., none, slight, mild, moderate, severe, extreme, most intense pain possible) have good reliability and validity, increased sensitivity to change, and a very low failure rate. Further, the verbal descriptor scale has been identified as the instrument preferred by many older adults, and it has strong clinical utility (Chibnall, Tait, 2001; Gagliese, Katz, 2003; Herr, Spratt, Mobily, et al., 2004; Kaasalainen, Crook, 2003; Pautex, Herrmann, Le Lous, et al., 2005; Pesonen, Kauppila, Tarkkila, et al., 2009; Taylor, Herr, 2003).

A recent quasiexperimental study compared the use of five pain intensity scales, including a novel pain thermometer called the Iowa Pain Thermometer (IPT; Figure 3-9, p. 87), with verbal descriptors in a sample of 61 younger and 36 older adults with osteoarthritic pain (Herr, Spratt, Garand, et al., 2007). The IPT provides a range of adjectives that have been supported in earlier research, including no pain, slight pain, mild pain, moderate pain, severe pain, very severe pain, the most intense pain imaginable (Herr, Spratt, Mobily, et al., 2004), along with a thermometer to assist understanding of the tool by older persons who have difficulties with abstract thinking. The IPT had the lowest failure rate of all pain intensity scales evaluated, was sensitive in detecting changes in pain following intervention, and was most preferred by both younger and older patients, including those with cognitive impairment. This and other studies provide further support for the IPT, as well as for the NRS (a 0 to 20 scale with 5-point intervals); the VDS (same adjectives as the IPT but no thermometer); and the FPS (7 line-drawn faces) in cognitively impaired elders (Taylor, Harris, Epps, et al., 2005; Ware, Epps, Herr, et al., 2006).

A simplified version of the verbal descriptor scale that was evaluated in long-term care includes only the adjectives none, mild, moderate, and severe (Closs, Barr, Briggs, et al., 2004). Treatment decisions are based on level of pain severity, and although this simple tool may be most useful for those with advanced dementia, it may be less sensitive than a verbal descriptor scale to detect changes in pain that move from severe to moderate or moderate to severe.

The FPS (7 faces) (Bieri, Reeves, Champion, et al., 1990) was revised in studies with children and presented in a 6-face version that is scored 0 to 10 (FPS-R), presented earlier in this chapter, p. 69) (Hicks, von Baeyer, Spafford, et al., 2001; Spagrud, Piira, von Baeyer, 2003). Subsequent studies in older populations have shown the FPS-R to be reliable and valid and to be the pain intensity scale preferred by cognitively impaired older African-Americans, Hispanics, and Chinese elders (Li, Liu, Herr, et al., 2007; Ware, Epps, Herr, et al., 2006). Although testing of the FPS-R in older Caucasian adults has not been published, the comparability of the FPS and FPS-R in children and the validity established in minority elders suggest that it would be appropriate in the general older-adult population. It should be noted that the FPS-R tends to correlate less strongly with other pain intensity scales, suggesting it may be capturing other characteristics besides intensity (e.g., pain affect). This does not negate its usefulness, but clinicians should be aware that it may be capturing a broader view of pain.

A consideration in tool selection that has not received much attention in the literature is the impact of a patient’s education and literacy on tool use and understanding. Low education level might suggest the need to focus tool selection on the simplest and most concrete options; however, education has not been a significant factor in failure to use pain intensity scales correctly in older adults (Gagliese, Weizblit, Ellis, et al., 2005; Herr, Spratt, Garand, et al., 2007). A study presenting a new scale for patients with low education, the Full Cup Test (FCT), demonstrated a statistically significant correlation between the FCT and the VAS (a 100-mm horizontal line with word anchors at the extremes; see p. 55 for discussion of and an image of VAS) and the ability to complete this alternative tool reliably in a sample of 114 patients with 9.23 years of education (Ergun, Say, Ozer, et al., 2007). Of 14 patients with low education, 3 could not understand the VAS, but all were able to complete the FCT. However, the literature shows that many older persons are unable to complete the VAS regardless of education level (Herr, Spratt, Mobily, et al., 2004). Calculation of a pain score using the FCT is cumbersome, so a copy of the tool is not included here, but it is mentioned and referenced, should the reader need such a tool.

Language challenges impact our ability to evaluate pain and are a consideration in tool selection. These challenges may relate to inability to read or write but also to those whose primary language is not English. A study of Portuguese patients with RA compared those who were literate (N = 66) with those who could not read or write in Portuguese (N = 25) using a VAS, VRS (separate boxes for no pain, mild, moderate, severe, unbearable), and NRS (boxes with 0 to 10) (Ferraz, Quaresma, Aquino, et al., 1990). The NRS that uses numbers rather than the VRS with word descriptors had higher reliability in both groups. Tools that avoid words, such as the FPS-R, may be useful options in older persons who are illiterate, dyslexic, or non-English speaking, although testing in these samples is needed.

In addition to pain intensity scales, the use of a pain map or figure drawing such as an enlargement of the figure in Form 3-1 (p. 51) can be a useful tool in identifying the location of pain. Even those with dementia have been able to use this approach to communicate the location of their pain (Weiner, Peterson, Logue, et al., 1998; Wynne, Ling, Remsburg, 2000). In those with cognitive impairment, information about pain location can guide approaches to care that minimize the activities and movements that precipitate discomfort. Knowledge of pain location can also help clinicians decide when anticipatory pain intervention would be useful (e.g., prior to range of motion exercises or transfer of a patient who located pain in the shoulders and hip joints).

In summary, the goals are to identify a pain assessment tool that older patients with cognitive impairment can use easily, to use the same tool consistently in each assessment, to establish a comfort-function goal, and to document the reports of pain in an accessible location in the medical records. The organization should identify valid and reliable tools from which to select for use with patients having varying levels of cognitive impairment (see Patient Example, below). A reasonable approach would be to have a vertical NRS such as 0 to 5, a VDS, and the FPS-R as options for this population. Unfortunately, no research has yet tested the 0-to-5 NRS in the cognitively impaired population. Development of documentation strategies that keep pain assessment information in a visible place is needed to monitor changes in pain and response to treatment. Table 3-1 (p. 96) provides a summary of research information on pain-intensity tools recommended for use in older adults with cognitive impairment.

Characteristics of Neuropathic Pain

The International Association for the Study of Pain defines neuropathic pain as “pain initiated or caused by a primary lesion or dysfunction of the nervous system” (Merskey, Bogduk, 1994, p. 212). This remains the current definition and includes neuropathies, postherpetic neuralgia, and radiculopathy. There are, however, a variety of clinical conditions that have neuropathic features but do not exactly fit this definition because they do not seem to involve an injury or dysfunction of the nervous system (Audette, Emenike, Meleger, 2005; Fishbain, Lewis, Cutler, et al., 2008). These conditions include complex regional pain syndrome, fibromyalgia, and others in which whether they represent neuropathic pain is controversial.

Neuropathic (pathologic) pain is distinctly different from physiologic (nociceptive) pain (see Section I). It is pain sustained by abnormal processing of sensory input by the peripheral or central nervous system, most often as a result of injury. There are numerous causes of nervous system damage that can result in neuropathic pain, including toxins, infection, viruses, metabolic disturbances, nutritional deficiencies, and trauma (Pasero, 2004).

The types of neuropathic pain are commonly divided into categories on the basis of the mechanism thought to be primarily responsible for causing the pain (i.e., peripheral or central nervous system activity) (see Figure I-1, p. 2). Note that in some cases the original injury occurs in the peripheral nerves (e.g., amputation), but the mechanisms that underlie the pain (e.g., phantom pain) seem to be generated primarily in the central nervous system.

Characteristics of Breakthrough Pain

BTP is divided into three subtypes: incident, idiopathic, and end-of-dose. Incident BTP is further divided into two types: predictable and unpredictable (Box 3-6). Idiopathic pain is not associated with any known cause. End-of-dose pain refers to pain returning before the interval between doses elapses. Defining these types of BTP may assist with selecting a form of treatment. Use of a short-acting opioid is commonly the treatment for incident pain and idiopathic pain, whereas an increase in the dosage of around-the-clock analgesic or a decrease in the interval between doses is usually the treatment for end-of-dose BTP.

In a review of 5 studies of cancer populations, BTP was found to occur in 50% to 90% of patients with persistent pain, confirming that BTP is a significant problem for patients with cancer-related pain (Portenoy, Bennett, Rauck, et al., 2006). In one study of BTP in 164 patients with persistent cancer pain, the median number of episodes was 6 per day, and the median interval from onset to peak was 3 minutes (Portenoy, Payne, Jacobsen, 1999). Almost two thirds (61.7%) could identify precipitants such as movement, but almost half (48.2%) stated that BTP was never predictable. Patients with BTP experienced more functional impairment and psychologic distress than did patients with controlled background pain and no BTP. BTP has not been as well studied in patients with persistent noncancer pain, but in a survey of 228 patients with noncancer pain, 74% had BTP (Portenoy, Bennett, Rauck, et al., 2006). The most common cause for their persistent pain was low back pain (52%). The median number of episodes of BTP per day was 2, median time to maximum intensity was 10 minutes, and median duration was 60 minutes. Patients identified a precipitant for 69% of the occurrences of BTP, and it was almost always activity related. Onset could not be predicted in 45% of bouts of BTP and could be predicted only sometimes in 31%.

A study of 43 patients with persistent noncancer pain assessed for impact of BTP on quality of life (Taylor, Webster, Chun, et al., 2007). In response to questions about how BTP affected patients’ lives, the two questions that 93% of the patients found affected their lives the most (“quite a bit or very much”) were related to general activity levels and the ability to work both outside the home and in the home doing housework. The next category most commonly rated (86%) as affecting their lives the most was enjoyment of life.

Tools for Assessment of Breakthrough Pain

Only a few tools have been developed for use in clinical practice to screen for or to assess BTP. Some have been used only for research purposes, and others are meant only for clinician assessment; that is, they guide a clinician’s interview of a patient. A list of questions that clinicians can use to interview patients about BTP has been published by a multidisciplinary panel (Bennett, Burton, Fishman, et al., 2005). A pain assessment algorithm has been published; it begins with the clinician asking a patient questions that screen for BTP and then follows up with specific questions about the BTP (Portenoy, Bennett, Rauck, et al., 2006). This was used in a research study of BTP in opioid-treated patients with noncancer pain and was modified from one originally developed for patients with cancer-related pain, neither of which has been validated (Portenoy, Payne, Jacobsen, 1999). Another published tool for assessing BTP, the Alberta Breakthrough Pain Assessment Tool, was developed for research with cancer patients (Hagen, Stiles, Nekolaichuk, et al., 2008). Some of the questions must be completed by the clinician as they interview the patient, and one section of the tool is suitable for completion by the patient. Evidence suggests that this tool is understandable by patients and clinicians, so further validation of the tool is warranted.

A screening tool for BTP is difficult to develop because BTP is not easily characterized when baseline pain has not been controlled. Baseline pain can be defined as almost always being present, continuous, steady, or constant or would be if medication did not control the pain. Thus, a screening tool for BTP in clinical practice is best designed to be used after baseline pain has been determined to exist and has been well managed.

Most patients with persistent pain who have controlled baseline pain experience only one type of BTP. In a study of 228 patients with persistent noncancer pain and controlled baseline pain, 74% had one or more types of BTP (Portenoy, Bennett, Rauck, et al., 2006). Of these patients, 88.7% had one type of BTP, and the remainder of the patients had up to three types of BTP. Precipitants were identified for 69% of the BTP episodes, and 92% of them were activity related. In a study of 164 patients with chronic cancer pain and controlled baseline pain, 51.2% experienced one or more types of BTP (Portenoy, Payne, Jacobsen, 1999). Of these patients, 83.1% experienced only one type of BTP, and the remainder experienced as many as three types of BTP. Precipitants of BTP could be identified by 61.7% of patients, with movement being the most common, 20.4%. However, almost half of the patients said that precipitants were never predictable.

Forms 3-6 (p. 103) and 3-7 (pp. 104-105) have been developed to provide simple, short forms to screen for and assess BTP. These forms have not been tested for reliability and validity, but they have been created in the absence of such forms in the hope that they will encourage assessment of BTP until simple, reliable, and valid forms are available. The short screening tool may be included in an already existing initial pain assessment tool, as is done with Form 3-1 (p. 51), or may be included in the clinician’s interview with the patient. The purpose of this form is to identify patients with persistent, constant, controlled baseline pain who also have pain that breaks through the controlled pain; that is, BTP. When such a patient is identified, he or she may complete the assessment form, or the clinician may use the form to ask the patient about BTP.

A detailed pain diary is commonly recommended as being the best account of the occurrence, predictability, severity, and duration of BTP (Payne, 2007). The Pain Control Diary, Form 3-11 (p. 116), may be used for this purpose and perhaps modified to include columns for noting physical activity at the time pain started and the time pain ended. A daily pain diary that may be more suitable for assessing BTP can be found at http://www.healthinaging.org/public_education/pain/my_pain_diary.pdf.

Flow Sheets for Analgesic Infusions in Acute Care

Obtaining pain ratings before and after the administration of analgesics is especially important because clinicians tend to assume that analgesics are safe and effective. Research has verified that the use of pain flow sheets, compared with the routine narrative charting in nursing notes, is effective in improving pain management (e.g., decreasing pain intensity and sedation) (Brown, 1992; Faries, Mills, Goldsmith, et al., 1991; McMillan, Williams, Chatfield, et al., 1988).

Flow sheets must be tailored to the patient population, clinical setting, and type of pain being managed. For example, some are designed for staff to complete for hospitalized patients using IV PCA for severe pain. Typically, flow sheets include, at a minimum, columns for time, pain ratings, facts about the analgesic administered (e.g., dose, route) or other pain treatment, and adverse effects pertinent to the situation. Adverse effects pertinent to the use of opioids for severe pain in opioid-naive patients (those not taking regular daily doses of opioids) include sedation level and respiratory status. Information about pain ratings should be obtained and recorded near the time of the event, whenever possible. For example, in a study of 60 patients being discharged from the postanesthesia care unit, many were unable to recall the pain ratings they gave when they were admitted to the unit (DeLoach, Higgins, Caplan, et al., 1998).

Frustration about the failure of staff nurses to reassess and document the effects of analgesics in acute care inpatient settings is frequently reported by nurses attending programs on pain management and is often a topic discussed on the nursing issues LISTSERV provided by the American Pain Society (APS) and the American Society for Pain Management Nursing. Research reflects this problem as well. In an Australian study of 52 nurses caring for 364 postoperative patients, 316 pain-related activities occurred in 74 observation periods (2 hours each), but only 4.4% were reassessments after analgesic administration (Bucknall, Manias, Botti, 2007). Before or during the observation period, 147 patients received analgesics but in only 13.6% was the pain reassessed. Most of these reassessments were opportunistic in that they were performed when the nurse was in the patient’s room carrying out other nursing activities such as administering other medications. Some have argued that reassessment following analgesic administration occurs but is not documented. However, this study was of actual nursing activity at the patient’s bedside, not documentation, and still reassessment of administered analgesics was substantially lacking.

In the United States, abstracts of the medical records of 709 older patients hospitalized for hip fracture were made for the purpose of determining nurses’ documentation of their pain assessments, including reassessments following administration of non-PCA analgesics (Herr, Titler, Schilling, et al., 2004). For the first 24-hour period following admission, 2965 analgesics were administered, and only 21.8% (646) were followed by reassessment of pain within 60 minutes. For the entire 72-hour period, only 15.3% (1497) of 9782 analgesic administrations were following by reassessments of pain within 60 minutes. In a study of frail elderly patients postoperatively who received analgesics, only 4% were reassessed following the analgesic (MacDonald, Hilton, 2001).

The medical records were reviewed for reassessment of 52 critically ill intubated patients who had had 183 pain episodes (Gelinas, Fortier, Viens, 2004). Pharmacologic interventions were used 89% of the time, and nonpharmacologic interventions were used less than 25% of the time. However, almost 40% of the time, pain was not reassessed after an intervention. Self-report was recorded for only 1 patient. Reassessment of patients consisted primarily of recorded behavioral and physiologic manifestations. Some of the physiologic indicators are questionable signs of pain, and although they were recorded within 1 hour of an intervention, it is not clear that they were done to reassess pain.

Another study of 117 charts of 80 inpatients and 37 outpatients with cancer whose pain had been documented in their charts revealed that reassessment after analgesic treatment was reported in 34% of the outpatient charts and 44% of the inpatient charts (Cohen, Easley, Ellis, et al., 2003). The more favorable results shown in this study may have occurred because charts that did not document pain at all were excluded from the sample, meaning that pain may have existed, and treatment may have been provided but not documented. For example, the researchers found one outpatient setting with no charts that documented pain at all, and the staff said they simply did not document pain.

Reassessment of pain after administration of an opioid is not just a documentation issue, it is also an issue of patient safety following administration of a drug that could lead to death. Reassessment should be the basis of planning for further intervention, in particular knowing when change in the treatment plan is needed. For example, measures such as decreasing the opioid dosage in a patient who is excessively sedated can be implemented to prevent life-threatening respiratory depression. Range orders for opioid analgesics simply cannot be effectively implemented unless the previous dose of analgesic is followed by reassessment of pain intensity and adverse effects. (See Section IV for additional information about opioid-range orders and treatment of opioid-induced adverse effects.)

Most health care facilities have incorporated guidelines for pain documentation into their policies and procedures and have provided space for documentation of basic pain assessments in standard nursing care forms; flow sheets are often used for analgesic infusions and IV bolus administration. Unit-specific quality improvement activities and ongoing education initiatives routinely include documentation of pain assessment following interventions for pain relief. In cases in which compliance is exceptionally low, management may decide to implement disciplinary measures such as linking documentation compliance to the nurses’ annual performance reviews.

In the initial treatment of moderate to severe pain with opioids, the patient’s pain rating is usually recorded every 1 to 2 hours during the first 24 hours, then every 4 hours if pain is controlled and other observations, such as sedation level, show the opioid is safe. When the patient appears to be sleeping at the time a pain rating should be obtained, it is often questioned whether to awaken the patient. Several factors should be considered. Not all patients who appear to be sleeping are actually sleeping. Further, sleep does not mean pain has been relieved. When a patient appears to be sleeping, his or her name may be called in a normal tone of voice. If the patient does not rouse and has acceptable respiratory status, pain assessment could be delayed until the patient awakens, and “appears to be sleeping” could be charted. However, even this approach has its limits, because sleep does not mean absence of pain. Therefore, if the medication has been given previously and the patient has rated relief as satisfactory, when that same medication is given again it is not necessary to awaken the patient after administration. However, if the patient has pain that is still out of control, discuss the situation with the patient and tell him or her that he or she may be awakened to determine whether the medication is providing effective pain relief.

Creative ways of reminding nurses to obtain and document pain ratings are often implemented as well. For example, hospital-wide competitions for prizes to be awarded to the unit demonstrating the highest compliance record can be good incentives for improving documentation. Some hospitals have installed, outside of patients’ rooms, colored flags (similar to those seen in clinics or office-based practices) that can be moved to a position perpendicular to the wall after an analgesic has been given so as to provide visual reminders to nurses of the need for reassessments. Enlisting the help of other staff members (e.g., nursing technicians or even patients can remind nurses of the need for reassessment at a specific time). (See the discussion of pain diaries on pp. 110-119 and oral PCA in Chapter 12.)

As health care facilities move toward the use of electronic health records, computer software has the potential to become an excellent tool for improving the documentation of pain. Carefully placed prompts and alerts can remind and even force clinicians to document pain assessment before moving on to the next computer screen. (See Appendix C for additional information about electronic health records.)

Following is a discussion of a flow sheet for use in acute care settings in which patients with severe pain are being treated with opioids. Examples are provided.

Pain Diaries for Outpatient Care

A pain control diary is a type of flow sheet for a patient to use in the outpatient setting, although there may be instances when it is appropriate to use it in inpatient settings too. Increasing numbers of patients with pain are being cared for on an outpatient basis. Patients with severe and sometimes escalating pain related to cancer are often cared for at home by family caregivers. Patients with moderate to severe pain related to surgical procedures are discharged from the hospital quickly, sometimes on the same day of the surgery. Both types of patients might benefit from the use of a pain diary.

Because the flow sheets or diaries of patients at home are kept by patients or their families, they must be particularly easy to use. Form 3-11 (p. 116) presents a paper pain diary designed for these circumstances. A patient teaching brochure titled “Understanding Your Pain: Using a Pain Rating Scale” contains an almost identical diary and is available at www.endo.com. It is similar to the diary presented in our previous book (McCaffery, Pasero, 1999). A nurse with chronic pain reported that she modified the diary for the purpose of clarifying for her physicians her pain intensity and the effectiveness of treatments (Aguirre, Nevidjon, Clemens, 2008). It helped her to identify trends in her pain and what worsened her pain. This adaptation of the diary is available at http://download.lww.com/wolterskluwer_vitalstream_com/PermaLink/NAJ_108_6_Aguirre_1059_SDC1.pdf.

As illustrated in Form 3-11, a paper diary for outpatients usually includes, at the minimum, columns or spaces for time, pain rating, medication taken or pain treatment, effect of the pain treatment on the pain rating, and comments, such as related activities or adverse effects. To avoid burdening patients and their families, certain criteria should be established regarding when to keep the record and for how long. The instructions for the pain diary in Form 3-11 stipulate that the record is to be kept only until an effective pain management plan has been developed. The example shown in Form 3-12 concerns a patient with chronic, progressive, cancer-related pain whose pain has escalated.

A paper diary may also be helpful for patients having ambulatory surgery. In a study of 20,817 patients who underwent same-day surgery, pain was the most common reason for unanticipated return to the hospital, occurring in 38% of the patients whose return was not anticipated (Coley, Williams, DaPos, et al., 2002). In ambulatory surgery, patients are usually discharged before there is time to establish the effectiveness of the oral analgesics they will be taking at home. Patients known to be at risk for moderate to severe pain, such as those having certain orthopedic procedures, may be asked to complete this diary about every 2 hours for the remainder of the day. The patients then have records that can be referred to when the clinicians contact them or the patients call with unrelieved pain or unmanageable adverse effects. Information from the pain control diary will be helpful in correcting the problem. For example, unrelieved pain may be due to failure to take the analgesic, or the analgesic prescription may be inadequate. Ideally, clinicians contact high risk patients the evening of discharge rather than the next day. Attempts to manage problems can be initiated before patients spend an uncomfortable night and develop complications.

In a hospitalized patient for whom the switch from parenteral or spinal analgesia to oral analgesia is not going smoothly, this paper pain diary may be modified and kept by the patients at the bedside. The information may be summarized in the nursing notes. Pain ratings every 2 hours may expedite identification of effective oral analgesic plans.

For certain periods of time, weeks or months, a daily pain diary may be indicated for a patient with chronic noncancer pain, such as headaches or low back pain. This type of record is especially important when medications or other treatments are being evaluated for continued use or are being adjusted to improve their effectiveness.

Poor recall of pain underscores the importance of asking some patients to keep records of pain management in the outpatient setting. The patient’s recall of pain often influences both the diagnosis and the treatment of pain. In one study of patients with chronic headache, pain ratings recorded in daily pain diaries kept over approximately 1 week were compared with patients’ memories of pain felt during that time (Eich, Reeves, Jaeger et al., 1985). Recall of pain intensity for maximum, usual, and minimum levels was influenced by the intensity of the pain felt at the time of the interview. The patients recalled their previous levels of pain as being more severe than they actually had been when the intensity of their present pain was high but recalled them as being less severe when their present pain intensity was low. In another study of patients with persistent pain, results with regard to recall of pain were similar (Smith, Safer, 1993). In addition, findings revealed that patients with lower present pain ratings also recalled their use of medication as having been less than it actually was.

In a study of patients receiving nerve blocks for pain, pain ratings were obtained before and after the blocks as well as 2 days and 2 weeks after the procedure (Porzelius, 1995). Findings revealed that memory distortions were common. At 2 days and 2 weeks after the nerve block, many patients recalled that the pain they had felt immediately after receiving the block was higher than they had reported at the time.

In outpatient care of patients with chronic pain, it is common to ask them, at each clinic visit, about their average or usual pain over the past week. Clinicians must realize that when precise information about pain ratings is needed so as to determine further treatment, daily records are more accurate than recall. Under these circumstances, a daily pain control diary is indicated.

Some patients will be asked to complete paper pain diaries only when problems are encountered, as in the example in Form 3-12. That patient was awakened by moderate to severe pain, and several supplemental doses were taken. Had the patient and family member not kept a diary, variations in pain ratings and the times breakthrough doses were taken, especially during the night, may not have been recalled accurately.

One of the limitations of paper diaries, when patients are asked to keep them over weeks or longer in the outpatient setting, is that some patients complete diaries immediately before they are to be turned in, or a week’s diary may be completed in advance (Gendreau, Hufford, Strone, 2003). Others have found that patients completing paper diaries are frequently noncompliant, tending to fabricate data if they have not completed the requested information at the times requested (Litt, Cooney, Morse, 1998). The use of electronic diaries that are programmed to prohibit invalid entries are helpful with respect to compliance (Hadjistravropoulos, Herr, Turk, et al., 2007). Momentary data are less likely to be falsified (Schwartz, Stone, 1998). The advantages of using electronic diaries are portability, low cost, ease of sharing data, and the availability of numerous software applications (Smith, Sheplock, 1999). Nevertheless their use is still limited by the much lower cost of paper dairies.

To study compliance with keeping paper diaries versus electronic diaries, 80 patients with chronic pain were asked to complete diaries; 40 patients kept paper diaries, and 40 patients kept electronic diaries (on palmtop computers) for 21 days, making entries three times a day (Stone, Shiffman, Schwartz, et al., 2002). Unknown to the patients, the paper diary was equipped with photosensors that identified the times at which entries were made so as to compare actual compliance with reported compliance. Among those who kept the paper diaries, reported compliance was 90%, but actual compliance was 11% to 20%. Actual compliance with the electronic diaries was 94%. The low compliance by those keeping the paper diaries is all the more remarkable considering the subjects were given $150 to participate in the study.

No other study has yet reported a comparison between compliance with electronic diary entries versus compliance with paper diary entries. The relatively higher compliance with paper diaries reported in the following studies may not be as high as actual compliance; that is, entries may not actually have been made at the times indicated in the paper diaries. In one study, 36 patients with chronic low back pain were asked to monitor their pain for 1 year (Jamison, Raymond, Levine, et al., 2001). Of the 36, 20 used both electronic diaries and paper diaries, and 16 used paper diaries alone. Two-way messaging available with the electronic diaries seemed to encourage the patients to use them, and they preferred them to the paper diaries. Further, patients were much more compliant with the monitoring of pain when using the electronic diaries (89.9%) than when using the paper diaries (55.9%). However, noncompliance resulted primarily from failure to return the paper diaries. Analysis of the data gathered by telephone calls and paper diaries suggested that the data in the electronic diaries were reliable and valid.

In another study, 155 cancer patients with metastasized bone pain were asked to complete daily paper pain diaries to determine the usefulness for themselves of doing so (Schumacher, Koresawa, West, et al., 2002). These were simple diaries that the patients were asked to complete once a day. (A copy of the diary is published in the Schumacher citation.) The pain diary was reported by 74% of the patients to be useful, and 10% felt it clearly was not. Very few patients and caregivers found the diary to be burdensome. No compliance data were reported. A group of 159 patients with cancer-related pain were asked to record their pain twice daily in a paper pain diary for 2 months at home (de Wit, van Dam, Hanneman, et al., 1999). Even in seriously ill patients, compliance was high (85.9%). High compliance may have been due in part to the guidance given by specially trained nurses and to the practice in completing the diaries before discharge. Although 37% of the patients reported that the diaries were of no benefit, 60% found that they helped them to cope with pain, giving them a sense of control over their pain. However, for patients with stable pain, the use of pain diaries may not be beneficial.

In a study of 91 patients with persistent noncancer pain, the patients used electronic diaries to monitor their pain during 2 weeks (Stone, Broderick, Schwartz, et al., 2003). Patients were divided into three groups and asked to monitor their pain 3, 6, or 12 times per day. Compliance was 94% or higher. Patients did not feel this was a burden and were willing to participate again in a similar study. One purpose of this study was to determine whether completing momentary diaries of pain would cause a shift in levels of pain over time (referred to as reactivity), and little support was found for such an occurrence.

In a crossover trial, 36 patients with persistent pain were asked to monitor their pain, mood, activity interference, medical use, and pain location in either paper or electronic diaries for 2 weeks (Marceau, Link, Jamison, et al., 2007). Patients reported that the electronic diary was easier to use than the paper diary, and 61% said that they would continue using it if given the choice. No compliance data were reported in this study. A serendipitous finding was that the patients using the electronic diary more frequently reported that their providers suggested medication changes based on information in the electronic diaries.

An emerging type of pain diary is an ordinary paper diary together with a digital pen and wireless mobile Internet technology. The digital pen looks, feels, and functions like an ordinary ballpoint pen. The strokes made by the pen are recorded and transferred by the Internet. In a small study of 12 palliative care patients using this technology, the patients reported that they found it easy to use despite severe illness and difficulties in comprehending the technology (Lind, Karlsson, Fridlund, 2008). The patients were asked to record immediate pain ratings three times a day and to complete a question about consumed doses. The patients could transmit their assessments at any time of any day. Caregivers gave the patients feedback on received assessments. It was this aspect of the methodology that the patients seemed to like the most. In postinterviews it was discovered that the patients believed that the transferred assessments were monitored by the caregivers around the clock, although the patients were not told that. Although the patients always recorded their pain at the requested times, initially they had difficulty using the VAS, and none of the patients understood how to mark the boxes to indicate the times at which they had taken their extra doses. Clearly, better directions should have been given. Nonetheless, the medical records indicated that the information from the assessments was useful and resulted in quick medical responses to changes in the patients’ statuses.

Research thus far suggests that electronic diaries and possibly digital pens are preferable to paper diaries when data are needed for 1 week or longer. Patients prefer electronic diaries to paper diaries, and compliance is better with electronic diaries. Based on the one study involving digital pens, it seems that this technology may be even better. Perhaps most important, it appears likely that data from electronic diaries or digital pens are most useful to physicians in making medication adjustments for patients.

When paper diaries, digital pens, or electronic diaries are used, the patient and family should receive explanations of the purpose of using them, along with information about how and when to complete the diaries and practice sessions with the diaries. Electronic diaries and digital pens are not always available in clinical practice, so paper diaries will serve as an alternative.

The low actual compliance with the use of paper diaries reported in one study (Stone, Shiffman, Schwartz, et al., 2002) suggests that when patients are asked to use paper diaries, it may be wise to caution them about the tendency to record information much later than it actually occurred, about the fact that recall is often inaccurate, and about the need to be honest about the times at which they actually make their entries. The most effective way to encourage compliance may be frequent feedback between clinician and patient. Calling the patient, perhaps every day or two, to remind him or her to complete the diary may help the patient to feel that the entries are of value. Perhaps the most important encouragement to complete the diaries lies in clinicians’ use of diary information to adjust care. Doing so gives the patient a reason to use the diary. In the studies mentioned, the patients responded very positively to having regular contact with clinicians and to feeling that their data were being used. Compliance may also be increased by decreasing the number of entries expected per day, although some patients did not feel overburdened by making as many as 12 entries per day (Stone, Broderick, Schwartz, et al., 2003).

Clinic Record for Reassessment of Patients with Persistent Pain Who Are Taking Opioids

A standardized tool for assessing patients with persistent pain who are receiving opioid therapy is almost essential to determine the effects of opioid management and to identify problems, such as undertreatment of pain, adverse effects, or problematic patient behaviors. Relevant domains for monitoring are often referred to as the Four As: Analgesia, Activities of daily living, Adverse effects, and Aberrant drug-related behaviors (Nicholson, Passik, 2007). The Pain Assessment and Documentation Tool (PADT), which assesses these four domains, has been published (Passik, Kirsh, Whitcomb, et al., 2004). The original, longer tool was field-tested by 27 clinicians who applied it to the assessments of 388 patients receiving long-term opioid therapy for persistent noncancer pain. The result is the PADT, a brief, two-sided chart note that can be included in a patient’s medical records (Form 3-13). It should require only a few minutes of a clinician’s time to complete.

The PADT assesses the four As in four sections of the tool. In the first section, Analgesia is assessed using the 0 to 10 scale. The second section, Activities of daily living, includes physical functioning, social relationships, and sleep patterns. The third section, Adverse effects, includes the effects that sometimes accompany chronic opioid use, such as nausea, constipation, and mental cloudiness. The fourth section assesses Aberrant drug-related behaviors, including those that may indicate undertreatment of pain, or abuse, or addictive disease, and they require differential diagnoses.

Aberrant drug-related behavior is too often interpreted as being indicative of abuse or addictive disease, and little thought may be given to the fact that the behavior may be caused by undertreatment of pain. The term aberrant drug-related behavior refers to behavior that is inconsistent with the expressed intentions of the prescriber, such as taking the drug for sleep when it was intended for pain relief (Fine, Portenoy, 2007). Another example of aberrant behavior is hoarding opioids during periods of reduced symptoms for later use when pain is worse or for occasions when the opioid is unexpectedly unavailable, such as a mail-order pharmacy’s being late in filling the prescription or the local pharmacy’s being temporarily out of the medication. Clearly, the presence of aberrant drug taking behavior should not automatically result in denying a patient opioid therapy.

The ultimate goals of using the PADT are to optimize analgesia, to improve activities of daily living, to minimize adverse analgesic effects, and to minimize aberrant drug behaviors. Further validation of the PADT is under way. The published tool is presented along with a survey of 19 clinicians who participated in the development of the PADT, revealing that 63.2% felt the tool would also be helpful in documenting behaviors or concerns for legal purposes (Passik, Kirsh, Whitcomb, et al., 2004).

1. Common goals. In each specific situation, the prescriber, nurse, patient, and others involved in the care must agree on a common goal, referred to as the comfort-function goals (discussed on pp. 85-88). This is often a simple matter of identifying activities, emotional states, or cognitive abilities that a patient values or needs to perform and determining the pain ratings that would be necessary for achieving the goal of performing. Simply put, if the team does not know where it is going, it will not know when it gets there or whether it has not yet arrived.

2. Common language. Obviously, communication is facilitated when all people involved are using the same words to refer to important elements of the situation. Pain rating scales are an example of using a common language. One of the most essential components of pain management is assessment of pain intensity. Pain rating scales that are valid, reliable, and easy to understand and use have been discussed in this chapter. For each individual patient, a single pain rating scale appropriate for that patient should be chosen and should be used by all team members when they discuss pain management among themselves and with the patient and family.

    Among clinicians, another component of a common language is using standardized assessment tools and formats for written communication and documentation. Examples of these, already discussed in this chapter, are the flow sheets.

3. Common knowledge. Because pain is a relatively new scientific discipline, lack of education about pain management is to be expected. Knowledge of pain management varies widely among clinicians and in various clinical situations. For this reason, clinical practice guidelines have been developed by professional organizations such as the APS. They may be used to provide clinicians with a quick summary of current recommendations for pain management. One or more of them can be distributed to clinicians and referred to for guidance each time a question arises about the assessment and treatment of pain. An example of a brief guideline that applies to many clinical areas is published by the APS and is titled “Principles of Analgesic Use in the Treatment of Acute Pain and Cancer Pain” (2003). This guideline might be given to prescribers, nurses, and pharmacists during orientation. Home care agencies might mail this guideline to referring physicians with a cover letter explaining that the agency is aiming to implement these recommendations.

    Other guidelines published by the APS include those that focus on pain management in sickle cell disease (Benjamin, Dampier, Jacox, et al., 1999), cancer in adults and children (Miaskowski, Cleary, Burney, et al., 2005), and various types of arthritis (2002). Although a considerable amount of essential information is contained in the clinical practice guidelines, the single most important directive is that clinicians must accept a patient’s report of pain. Repeated reference to this recommendation may well be necessary.

4. Regular communication. Common goals, language and knowledge must be shared by team members on a regular and predictable basis. In some instances, such as the care of patients with persistent pain, regular communication takes place at a weekly team meeting. In other situations, such as the care of surgical patients, communication make take the form of pain ratings recorded every 4 hours on a flow sheet and included at the change-of-shift or hand-off report. Home health care agencies may use a standardized format to fax or e-mail updates to referring prescribers and may follow up with telephone calls when there are problems. Care should be taken, however, to report the positive too, not just the problems.

Types of Behavioral Pain Assessment Tools

Tools for the observation of pain behaviors have been developed to assist in the recognition of pain in those unable to self-report. A number of reviews have critiqued existing behavioral tools and similarly concluded that no tool is currently recommended for general use across all populations and settings (Hadjistavropoulos, 2005; Herr, Bjoro, Decker, 2006; Stolee, Hillier, Esbaugh, et al., 2005; van Herk, van Dijk, Baar, et al., 2007; Zwakhalen, Hamers, Abu-Saad, et al., 2006). However, additional research into some existing tools and the development of new tools continue to broaden the selection of instruments for consideration. When selecting a standard observation tool for implementation, it is important for clinicians to consider the psychometric properties of a tool and its clinical utility and appropriateness for the setting and population. The next section provides information about existing tools that can guide clinicians in determining which tool to incorporate into a comprehensive approach to pain assessment in persons unable to self-report. Because of rapid changes in the field and ongoing research, monitoring for new developments is recommended.

The Mayday Fund provided support for the development of resources to assist clinicians in behavioral pain assessment tool selection, available in the resource section at the Pain Resource Center at the City of Hope (http://.prc.coh.org/PAIN-NOA.htm). A detailed critique of existing pain tools, a brief summary of tool strengths and challenges, information regarding tool access and permission, and a comparison across tools are available at the website.

There are a variety of ways to categorize the pain behavior assessment tools available for use with nonverbal older patients. One approach is to organize them as pain behavior checklists, pain behavior intensity scales, and a combination of both. Alternatively, tools can be organized by whether they are direct-observation tools or informant-based tools. Direct-observation tools can usually be completed by those without prior knowledge of the patients as they gather information about specific behaviors observed during a specified period or activity. They identify the presence or absence of the behavior and may include a rating of the behavior’s intensity or frequency (Feldt, 2000; Snow, Weber, O’Malley, et al., 2004; Warden, Hurley, Volicer, 2003). Informant-based tools (also called surrogate/proxy tools) rely on information from family members, caregivers, and health care providers familiar with the individuals’ baseline behaviors and they can detect changes in activities and behaviors from usual presentation. There also may be overlap in direct observation and informant report. Regardless of the categorization of tools, several considerations are necessary in determining the best approach for given populations and settings.

A key consideration is tool specificity versus sensitivity. Because of individual variability in behavioral presentation related to pain, short, direct observation tools that focus on key common pain behaviors (e.g., grimacing, moaning, bracing) may recognize pain in those who present with these common pain behaviors (specificity) but may fail to detect pain in those who present with more atypical changes in activity or behavior (sensitivity) (e.g., decreased appetite, increased sleeping, greater agitation, acting withdrawn, or increased irritability). Longer comprehensive tools that include more potential pain indicators may be more sensitive but less specific. This means that more patients who present with less common pain behaviors will be identified, but in some of those, pain may not be the cause of their behaviors. Follow-up evaluation is necessary to determine the causes of the behaviors observed.

One factor influencing tool use is whether behavior is observed at rest, during or following some activity, or during an activity intended to evoke pain. Studies have illustrated that observation at rest is misleading and can lead to incorrect judgment about the presence of pain (Feldt, 2000; Hadjistavropoulos, LaChapelle, MacLeod, et al., 2000; Husebo, Strand, Moe-Nilssen, et al., 2008). Thus, observation during movement or activity is recommended. The timing of the observation should be consistent and should be part of the assessment procedure in persons requiring behavioral observation (e.g., during morning care, transfers, or ambulation).

In patients with severe dementia, studies support the ability of surrogates to recognize pain’s presence, but not accurately rate its severity (Manfredi, Breuer, Wallenstein, et al., 2003; Pautex, Herrmann, Michon, et al., 2007; Shega, Hougham, Stocking, et a., 2004). This may be related to individual variability in patients’ behavioral presentations. For example, one patient with severe dementia may present with agitation and aggressive behavior, whereas another may become withdrawn and quiet. Their behavioral presentations are quite different, but they both could be experiencing moderate pain; this indicates that it is unreliable to determine the severity of pain on the basis of the severity of restlessness and agitation. Some authors have attempted to establish a direct relationship between behavior scores and pain intensity, such as adding up the score on a behavioral tool and equating it with pain intensity. Given current knowledge about this topic, number or intensity of behaviors cannot be added up to equal a pain score. We emphasize that no self-report means no pain-intensity rating (Pasero, McCaffery, 2005). The low correlations shown in most studies between surrogate judgment of pain severity and patient reports of severity contribute to concern about using behavioral scores as pain-intensity ratings (vanHerk, vanDijk, Biemold, 2009). Further, if clinicians could reliably determine the intensity of pain in patients who cannot self-report, there would be no need for behavioral assessment tools (Zwakhalen, Hamers, Abu-Saad, et al., 2006).

Certain studies that have focused on direct observation of typical pain behaviors (such as grimacing, vocalizations, and guarding) show stronger correlations between pain behaviors and pain severity (Horgas, Nichols, Schapson, et al., 2007; Husebo, Strand, Moe-Nilssen, et al., 2007). However, the behaviors observed were typical pain behaviors, and the circumstances were controlled, so the results might not apply to assessments involving broader circumstances and less typical behaviors. Future research may reveal approaches that guide judgments of pain’s severity on the basis of behaviors. However, until there is stronger evidence to support this relationship, the score on a pain behavior assessment tool is not the same as a pain-intensity rating scale, and the two should not be directly compared (Herr, Coyne, Key, et al., 2006). However, for an individual patient, an increasing score on a behavioral pain assessment tool usually indicates increasing pain, and a decreasing score probably represents a decrease in pain.

Selection of a tool is determined by the tool’s reliability and validity, the setting of the care and the population to be assessed, the availability of knowledgeable caregivers, and the clinical utility. Foremost, a pain assessment tool must have at least moderate reliability and validity. A tool that is not reliable and valid is not useful for supporting clinical judgments. It is very tempting to develop a new tool or adapt an existing tool to meet a clinical need. For example, a tool such as the Face, Legs, Activity, Cry, and Consolability (FLACC) may seem like a good tool to use with older persons. However, the FLACC was based on children’s development and has not yet been demonstrated to be reliable and valid for older persons (Baiardi, Parzuchowski, Kosik, et al., 2002). A recent study of the FLACC in citically ill patients who could not self-report included older adults, but it was not clear how many of the population were 65 to 70 years and older (Voepel-Lewis, Zanotti, Dammeyer, 2010). Therefore, further study of the FLACC in older patients is warranted. Considerable time and effort are required to establish a tool’s validity and reliability, which are essential before integrating a tool into clinical use.

One key aspect of a tool’s validity is its sensitivity to change—a critical aspect of a tool’s utility that has received limited attention in the research to date. Several of the existing tools have been evaluated for their abilities to detect responses to intervention (e.g., analgesia) (Cohen-Mansfield, Lipson, 2008; Morello, Jean, Alix, et al., 2007). Ensuring that tools can detect response to treatment is part of the development of a tool’s validity.

The setting of care is important in evaluating the appropriateness of a tool. For example, a number of tools that have been developed require observation over time and knowledge of patients’ usual behaviors, so they would not be useful in acute care settings to look for procedural pain or postoperative pain. In acute care settings, a short, direct-observation tool may be well suited for use with patients who have acute pain problems and are unknown to the nursing staff. In the long-term care setting, a short, direct-observation tool that can be used on a regular basis by CNAs during usual care combined with a more comprehensive screening tool that is completed at regular intervals may work well. Finally, a pain assessment tool that requires excessive training, is perceived as being too complex, or requires too much time to administer will not be helpful in promoting regular clinical use of the tool. Careful consideration of these characteristics can guide the selection of an appropriate tool. Detailed critiques and summaries of tools are available at http://pcr.coh.org/PAIN-NOA.htm.

Tools for Assessing Pain in Cognitively Impaired Patients and Others Who Cannot Self-Report

For those caring for older adults in long-term care facilities, the minimum data set (MDS) may be considered a source of assessment data. The MDS is a federally mandated system for the assessment of residents of long-term care nursing facilities. It systematizes the assessment of each resident’s functional, mental, psychosocial, and medical status at admission and at regular intervals thereafter (Kovach, Noonan, Griffie, et al., 2002). However, concerns have been raised regarding the underreporting of pain in cognitively impaired residents when reports are based on MDSs (Chu, Schnelle, Cadogan, et al., 2004; Cohen-Mansfield, 2004). A revised MDS 3.0 has been under development and evaluation and includes additional pain assessment content, including documentation of behaviors related to pain. Once implemented, the value of the revised MDS may be improved although evaluation of its ability to detect pain in nonverbal patients and to demonstrate response to intervention will be needed.

A number of standardized tools for pain assessment have been developed and are at various stages of testing and refinement. Following is a description of selected tools that merit consideration, depending on the setting and population. We have chosen not to include discussions of tools that do not have support for use with older adults with cognitive impairment or that focus primarily on the detection of discomfort rather than pain. Information about other tools can be obtained at http://prc.coh.org/PAIN-NOA.htm.

Recommended Tools for Patients Who Cannot Self-Report

No perfect behavioral assessment tool exists, but based on current research and clinical utility, the three tools we recommend for use in clinical practice with patients who cannot self-report are (1) Checklist of Nonverbal Pain Indicators (CNPI); (2) The Pain Assessment in Advanced Dementia (PAINAD); and (3) Pain Assessment Checklist for Seniors with Severe Dementia (PACSLAC). These tools are described, and an overview of them is provided in Table 3-3. Other tools are also described.

The tools presented here are in varying stages of development, and further testing may establish greater reliability and validity; however, considerable detail is presented regarding the current stages of the research that establishes their psychometric properties. Research terminology unfamiliar to some readers has been used. To facilitate understanding of the research findings, Table 3-4 provides definitions of selected research-related terms.

The Checklist of Nonverbal Pain Indicators (CNPI) (Feldt, 2000) is a brief, clinically useful observation list of six behavior items that are scored as being present or absent at rest and during movement in cognitively impaired older adults. (Form 3-14 on p. 131 shows the tool as it is used in clinical practice.) The tool includes only the common behaviors that reveal pain, thus limiting the measure’s ability to capture the full range of possible pain behaviors. Preliminary testing provided initial support for use of the tool with older adults in acute care settings, although internal consistency was low. Recent evaluation in long-term care (Jones, Fink, Hutt, et al., 2005) demonstrated support of the CNPI as a measure of pain’s severity; increases in CNPI scores were associated with increases in self-reports of pain (Jones, Fink, Hutt, et al., 2005). However, 50% of residents reporting pain had no visible indicators of pain, which raises concerns about the tool’s sensitivity and its ability to detect persistent pain in those unable to report. Cutoff scores for determining the presence and severity of pain are not available. Another study, conducted in a group of Norwegian nursing homes, provides evidence of acceptable test-retest and inter-rater reliabilities and concurrent validity, as well as the tool’s practicality when administered by various categories of nursing personnel (Nygaard, Jarland, 2006). Further testing in acute care and long-term care settings in the United States is recommended.

The Pain Assessment in Advanced Dementia (PAINAD) Scale (Warden, Hurley, Volicer, 2003) was developed as a short, easy-to-use observation tool for assessing pain in individuals with advanced dementia. (Form 3-15 on pp. 132-134 shows the tool as it is used in clinical practice.) The PAINAD is an adaptation of the Discomfort Scale for Dementia of the Alzheimer Type (DS-DAT) (Hurley, Volicer, Hanrahan, et al., 1992) and the FLACC (Merkel, Voepel-Lewis, Shayevitz, et al., 1997) and includes five items: breathing, negative vocalization, facial expression, body language, and consolability. These items are not comprehensive, and the ability to detect pain in those with less obvious changes in behavior may be compromised. No data attach level of pain severity (mild, moderate, or severe) to the number obtained in the tool, although higher scores do represent greater reports of pain.

The PAINAD has been translated into four languages and tested, and three additional English studies have provided additional information about the tool’s psychometrics and utility. Subsequent studies have provided support for its construct and its concurrent validity and ability to detect pain and to differentiate between groups with and without pain in both long-term and acute care settings (Costardi, Rozzini, Costanzi, et al., 2007; DeWaters, Faut-Callahan, McCann, et al., 2008; Leong, Chong, Gibson, 2006; Schuler, Becker, Kaspar, et al., 2007; van Iersel, Timmerman, Mullie, 2006; Zwakhalen, Hamers, Abu-Saad, et al., 2006).

One study that compared the PAINAD to other tools in terms of their abilities to detect change resulting from treatment raises questions regarding the tool’s sensitivity (Cohen-Mansfield, Lipson, 2008). Follow-up studies demonstrate good internal consistency, inter-rater reliability, and test-retest reliability. Several studies question the utility of including the items breathing and consolability, but removal of these items did not demonstrate improved internal consistency. Further study is needed to investigate the tool’s sensitivity to change in behavior in response to treatment.

The Pain Assessment Checklist for Seniors with Limited Ability to Communicate (PACSLAC) (Fuchs-Lacelle, Hadjistavropoulos, 2004), developed by a Canadian team, is an observational tool that uses direct observation and familiar caregiver information to assess both common and uncommon pain behaviors. (Form 3-16, pp. 135-136, shows the tool as it is used in clinical practice.) The PACSLAC is potentially a clinically useful behavior checklist that appears to be simple to use for assessing and monitoring changes in persons with dementia and diverse presentations of pain-related behavior. The tool is comprehensive; 60 indicators address all six pain behavior categories included in the AGS guidelines. Prospective evaluation has supported the tool’s reliability and validity (Fuchs-Lacelle, Hadjistavropoulos, 2005; Zwakhalen, Hamers, Berger, 2006; Zwakhalen, Hamers, Berger, 2007) and has done a factor analysis to determine the most efficient and useful indicator set for clinical use. Because of the length of the tool, it may not be suitable for acute care settings. However, clinicians report that it actually takes very little time to complete the form once they are familiar with it. Even if the tool is not formally used in the acute care setting, it can be posted in the clinical units as a resource for identifying behaviors.

A study was conducted to determine whether systematic pain assessment of older adults with severe dementia in a nursing home, using the PACSLAC, would improve pain management and decrease nursing stress in comparison with a control group (Fuchs-Lacelle, Hadjistavropoulos, Lix, 2008). The most common pain-related diagnosis was arthritis. The experimental group of nurses was asked to complete the PACSLAC regularly, that is, at least three times a week. The use of PRN medications by the experimental group increased more over time than it did in the control group. Further, the nurses in the experimental group reported a decrease in work-related emotional exhaustion over time; the control group did not. Rather than increasing stress in the experimental group, the addition of a 60-item assessment tool to their workload led them to find that regular use of the assessment tool reduced their stress. One explanation for this result is that pain caused disruptive and aggressive behavior, and assessment identified this pain and led to increased analgesic medication. This reduced patients’ pain and hence decreased disruptive behaviors, which in turn decreased the nurses’ stress.

The revised PACSLAC-D (Zwakhalen, Hamers, Berger, 2006) is a 24-item tool but does not include items based on changes in behavior or activity. Because of this revision, the PACLSAC and PACSLAC-D psychometric properties should be considered independently. The PACSLAC has good internal consistency, inter-rater reliability, and intra-rater reliability. The PACSLAC-D also shows good reliability estimates in preliminary testing. Preliminary normative data and cutoffs are provided but require validation in larger, more diverse samples. Further refinement of the PACSLAC has been reported, with a shorter tool including 18 valid and reliable indicators confirmed (van Nispen tot Pannerdan, Candel, Zwakhalen, et al., 2009). Additional factor analysis in English-speaking samples and evaluation of sensitivity in detecting treatment effects are recommended.

Other Tools in Development

Several other tools are in development for use in clinical practice. Some of them are discussed briefly; however, new tools and further evaluation of existing tools occur regularly in the literature. Readers are encouraged to study the literature so as to stay abreast of new research findings.

The Abbey Pain Scale (ABBEY) (Abbey, Piller, DeBellis, et al., 2004) is an Australian informant-based tool that measures pain intensity in people with late-stage dementia. The tool attempts to measure acute pain, chronic pain, and acute-on-chronic pain and produces ranges of scores to determine pain intensity. The rationale for mixing and differentiating pain types, pain behaviors, and pain causes is not clear nor is the justification for the pain intensity ranges. Although a recent evaluation of the use of the ABBEY with palliative care patients found that the staff perceived the tool to be helpful in judging pain (van Iersel, Timmerman, Mullie, 2006), facial expressions, vocalizations, and body language were identified as being the best and easiest pain indicators to observe. Information about the tool’s reliability and validity is limited, and tool revision and additional testing in well-designed studies are recommended.

The Doloplus 2 (Lefebvre-Chapiro, 2001; http://www.doloplus.com) is a French tool developed for the multidimensional assessment of pain in nonverbal older adults. It has been translated into several languages, including English, although none of the studies available were conducted using the English-translated tool. The Doloplus 2 is a comprehensive tool based on behavior change that addresses many key indicators noted in the literature and in the AGS persistent pain guidelines and that includes three subscales with 10 items focusing on somatic reactions, psychomotor reactions, and psychosocial reactions. The items are rated according to increasing severity of pain, and the result is a total score that ranges from 0 to 30, with 5 being identified as the threshold for pain. Justification of the cutoff score is not available in the English literature. Recent studies conducted in French, Norwegian, and Dutch samples have been published in English and provide strong support of the tool’s reliability, validity, and clinical utility (Holen, Saltvedt, Fayers, et al., 2005; Pautex, Herrmann, Michon, et al., 2007; Pautex, Michon, Guedira, et al., 2006; Zwakhalen, Hamers, Berger, 2006). Translation issues are evident, and further study or description regarding the use of Doloplus 2 in English-speaking populations is needed.

Elderly Pain Caring Assessment 2 (EPCA-2) (Morello, Jean, Alix, et al., 2007) was developed by a team of French physicians in long-term care facilities to evaluate both persistent and acute pain in non-verbal older adults. The tool relies on caregiver familiarity with a patient to report changes in behavior. The tool has eight items that comprise five of the six categories of nonverbal pain behaviors noted in the AGS persistent pain guidelines. The items are observed prior to and during caregiving; the time needed for observation and completion of the tool is 15 minutes. The hierarchy of pain behaviors according to pain intensity appears logical, but no conceptual basis for the ordering is offered. In other words, in the fifth subscale item (“Observations during caregiver intervention”), “restless” is rated lower than is “aggressive,” and this seems a logical order. However, validation of the ordering of items in each subscale of the tool has not been provided.

Preliminary internal consistency and inter-rater reliabilities are moderate to strong. There was high convergent and discriminant validity and strong responsiveness to treatment. The tool requires training and time for proper administration and has not been validated in English-speaking samples, limiting its clinical utility in the United States.

The CNA Pain Assessment Tool (CPAT) (Cervo, Raggi, Bright-Long, et al., 2007) is an informant-based assessment tool for use by certified nursing assistants (CNAs) for the assessment of pain in patients who have been diagnosed with severe dementia. The tool uses 12 items and collapses them into 5 categories, including 3 of the 6 categories of pain indicators consistent with AGS pain guidelines: facial expressions, body movements, and verbalizations. The tool requires approximately 1 minute to administer, with an uncomplicated forced choice numeric scoring of 0 to 5 and a score of 1 or greater requiring further action by the CNAs. Initial validity was established comparing ratings between patients with and without pain.

Subsequent evaluation of the CPAT (Cervo, Bruckenthal, Chen, et al., 2009; Cervo, Raggi, Bright-Long, et al., 2007) noted acceptable levels of both interrater reliability and test-retest reliability. Construct validity was established comparing ratings before and after a known painful or uncomfortable event and criterion validity established by comparing to another established pain scale, the DS-DAT. Based on a practicality survey, the CPAT was shown to be clinically useful and feasible. The CPAT requires further evaluation of ability to measure severity of pain and response to treatment.

The Mobilization-Observation-Behavior-Intensity-Dementia Pain Scale (MOBID) (Cervo, Raggi, Bright-Long, et al., 2007; Husebo, Strand, Moe-Nilssen, et al., 2007) is a nurse-administered instrument used to observe pain behaviors and infer pain intensity at rest and with standardized guided activity. It is intended for use with patients who have severe cognitive impairment and chronic musculoskeletal pain and are living in long-term care settings. A patient is guided through five structured activities (mobilization of both hands, both arms, and both legs; turning in bed; sitting at bedside) and the presence and intensity of pain are rated by nurses on an 11-point NRS for sounds, facial expressions, and defense, with increased behaviors indicating increased pain (e.g., combative behaviors when moved). The tool does not include subtle cues or changes in behavior and may limit recognition of pain in those presenting atypically. The MOBID has good internal consistency. There was a wide range of inter-rater reliability for the presence of pain behaviors, but reliability was better for pain intensity. The MOBID was able to detect increased pain with movement, and there was a linear trend among external raters toward more pain behaviors being associated with higher pain intensity. Recent follow-up evaluation supports reliability of MOBID scoring during video evaluation of behaviors in older persons with severe dementia (Huesebo, Strand, Moe-Nilssen, et al., 2009). More evidence is needed to support the validity of the tool for both the presence of pain behaviors and inferred pain intensity.

The Non-Communicative Patient’s Pain Assessment Instrument (NOPPAIN) (Snow, Weber, O’Malley, et al., 2004) is an instrument that focuses on nursing assistants’ (nursing technicians’) observations and assessments of pain in patients with dementia at rest and with movement. Nursing assistants observe for pain-related behaviors as they perform common caregiving tasks. The NOPPAIN contains common pain behaviors, not subtle cues or changes in behavior. Because the NOPPAIN is based on proxy reports of pain intensity, evidence is needed to support the validity of the judgments of intensity of pain behaviors as indicators of pain severity. Preliminary support for the tool’s reliability and validity was reported based on responses to video simulations, rather than on actual observation of patients. One follow-up study provides support for the convergent validity of proxy reports by means of detailed behavioral coding in patient videos (Horgas, Nichols, Schapson, et al., 2007). The tool’s reliability and validity have been established by means of video coding by nursing assistants and patient videos scored by nursing students in a laboratory setting. Although ease of administration by nursing assistants is a strength of the tool, evaluation of nursing assistants’ abilities to use the scale in actual patient care situations is needed.

Pain Assessment in Noncommunicative Elderly Persons (PAINE) (Cohen-Mansfield, 2006) is an informant-based assessment tool developed to assess pain in noncommunicative elders. The tool includes a comprehensive list of 22 behaviors representing 4 of the 6 pain behavior categories consistent with the AGS persistent pain guidelines. The tool has been administered only by research assistants and has not been studied for feasibility in the clinical setting. No information about the time or skill level needed to complete the PAINE is reported. Preliminary evaluation suggests good internal consistency and good inter-rater reliability, even among staff members who have varying familiarity with patients. Although procedures are unclear and difficult to follow, it appears that the PAINE has construct validity and good correlation with other tools testing responsiveness to intervention (Cohen-Mansfield, Lipson, 2008). However, PAINE is weakly correlated with other observational and self-report assessment tools. Further evaluation of reliability and validity when used by nursing assistants directly is needed.

Two newer instruments that warrant consideration are the Rotterdam Elderly Pain Observation Scale (REPOS) (van Herk, van Dijk, Tibboel, et al., 2009) and the Mahoney Pain Scale (Mahoney, Peters, 2008). These tools were not available to include in the critical review completed in 2008 on the City of Hope Pain Resource Center website and discussed in this chapter. However, the reader is encouraged to access the most recent information when reviewing possible tools for consideration.