Home > RNJ > 2005 > March/April > Psychometric Evaluation of Selected Pain Intensity Scales for Use with Cognitively Impaired and Cognitively Intact Older Adults

Psychometric Evaluation of Selected Pain Intensity Scales for Use with Cognitively Impaired and Cognitively Intact Older Adults
Laurie Jowers Taylor, PhD RN • Judy Harris, MSN ARNP CRRN CCR CCM • Cynthia D. Epps, PhD RN • Keela Herr, PhD RN FAAN

The purpose of this study was to determine the reliability and validity of selected pain intensity scales such as the Faces Pain Scale (FPS), the Verbal Descriptor Scale (VDS), the Numeric Rating Scale (NRS), and the Iowa Pain Thermometer (IPT) to assess pain in cognitively impaired older adults. A descriptive correlational design was used, and a convenience sample of 66 volunteers age 60 and older residing in assisted living facilities in the South was recruited for this study. The sample included 22 (33%) men and 44 (67%) women, with a mean age of 76. Ninety-eight percent (65) of the sample comprised Caucasian participants, with the exception of 1 African-American man. Seventy percent (47) completed high school and/or college. The mean Mini Mental State Exam (MMSE) score was 16, with a range of 1 to 29. Eighty-five percent scored 24 or lower, indicating some degree of cognitive impairment. The remaining 15% were cognitively intact. All but one participant could use each scale to rate their pain. Concurrent validity of the VDS, NRS, and IPT was supported with Spearman rank correlation coefficients ranging from .78 to .86 in the cognitively impaired group. The FPS, however, demonstrated weak correlations with other scales when used with the impaired group, ranging from .48 to .53. In the cognitively intact group, strong correlations ranging from .96 to .97 were found among all of the scales. Test-retest reliability at a 2-week interval was acceptable in the cognitively intact group (Spearman rank correlations ranged from .67 to .85) and unacceptable for most scales in the cognitively impaired group (correlations ranged from .26 to .67). When asked about scale preference, both the cognitively impaired and the intact groups preferred the IPT and the VDS. This study revealed that cognitive impairment did not inhibit participants’ ability to use a variety of pain intensity scales, but the stability issue must be considered.

Despite pain awareness in the older population, evidence suggests inadequate assessment and undertreatment of pain in older adults continues to be problematic (Bernebei, 1998; Ferrell, Ferrell, & Osterweil, 1990; Fries, Simon, Morris, Flodstrom, & Bookstein, 2001; Sengstaken & King, 1993). Feldt and Gunderson (2002) report that elders receive fewer analgesics upon transition to a nursing home or rehabilitation facility than they received in the hospital. Chronic pain in older adults who experience conditions such as osteoarthritis, osteoporosis, back pain, poststroke syndrome, and postherpetic neuralgia limits their activities of daily living and progress toward rehabilitation goals. Chronic pain can contribute to feelings of hopelessness and despair, which can enhance patients’ perceptions of pain (McCaffrey, Frock, & Garguilo, 2003). Satisfaction with treatment is an important measure of quality pain management; a strong predictor of satisfaction is improvement in daily activity (McCracken, Evon, & Karapas, 2002). Improving comfort is vital to reclaiming independence in the rehabilitation setting. Patients must be free of pain (or have a tolerable level of pain) to facilitate their participation in activities and exercise.

The common misconception that pain cannot be assessed in older adults with impaired cognition is another barrier to pain assessment and treatment across all settings. This lack of assessment and pain undertreatment (in both impaired and intact older adults) results in psychological, physical, and social sequelae that negatively affects quality of life (Ferrell & Ferrell, 1990; Lavsky-Shulan et al., 1985; Weiner, Peterson, & Keefe, 1998).

Accurate pain assessment and treatment across all ages and ethnic groups have been recommended by key policy, professional, and regulatory agencies, including the Agency for Health Care Policy and Research (U.S. Department of Health and Human Services, 1992, 1994), the American Geriatrics Society (2002), the American Pain Society (1999), and the Joint Commission on Accreditation of Healthcare Organizations (1999). Older adults with cognitive impairment present specific challenges to healthcare providers in terms of pain assessment because short-term memory deficits result in difficulty communicating pain. Preliminary research findings, however, suggest many cognitively impaired adults reliably can complete a basic Numeric Rating Scale (NRS) or Verbal Descriptor Scale (VDS; Ferrell, Ferrell, & Rivera, 1995; Herr, Spratt, Mobily, & Richardson, 2004; Manz, Mosier, Nusser-Gerlach, Bergstrrom, & Agrawal, 2000; Parmelee, Smith, & Katz, 1993).

Although empirical psychometric evaluation of pain assessment scales for use with older adults is growing, data are limited on pain scale use with cognitively impaired older adults in the rehabilitation setting. The lack of systematic psychometric evaluation of pain intensity scales and failure to report cognitive status and specific levels of cognitive status in research findings contribute to the confusion about appropriate scales with older adults experiencing some degree of cognitive impairment.

The purpose of this study was to determine the psychometric properties of four pain scales: the Iowa Pain Thermometer (IPT; Herr, 2000), the VDS, the 0–10 NRS, and the Faces Pain Scale (FPS; Bieri, Reeve, Champion, Addicoat, & Zeigler, 1990) as methods to assess pain intensity in older adults with and without cognitive impairment who reside in assisted-living facilities in Tallahassee, FL.

Literature review

Pain management in cognitively impaired older adults presents several challenges. Because some cognitively impaired individuals cannot report pain verbally, this group tends to underreport their pain. In addition, healthcare providers often use inappropriate pain intensity scales. Consequently, cognitively impaired older adults are at risk for undertreatment of pain (Horgas & Tsai, 1998). Additionally, some healthcare providers may be reluctant to use a pain intensity scale with cognitively impaired older adults, assuming they are incapable of using scales. Cognitively impaired older adults may not be able to verbalize their pain, but they usually are capable of describing pain or using a scale when prompted.

Verbal descriptor, numeric rating, and facial pain scales are the tools most commonly used to measure pain intensity in the United States (McCaffrey & Pasero, 1999). These scales have been shown to be valid and sensitive measures of pain intensity in general populations, and recent research has established the validity and reliability of these measures to assess pain intensity in older adults, including those with mild-to-moderate cognitive impairment (Feldt, Ryden, & Miles, 1998; Ferrell, Ferrell et al., 1995; Herr, 2000; Herr et al., 2004; Manz et al., 2000; Parmelee et al., 1993; Taylor & Herr, 2002; Taylor & Herr, 2003). The pain thermometer, an adapted VDS, also has been shown to be reliable in nonminority older adult samples (Herr, 2000; McCaffrey & Pasero, 1999; Weiner et al., 1998).

Although previous studies on use of pain scales with cognitively impaired older adults have shown that many can complete selected scales reliably, little evidence is available for differentiating scale use by cognitive level. Manz et al. (2000) found that 90% to 100% of mildly to moderately cognitively impaired nursing home residents could complete selected pain scales, and 30% of the severely impaired older adults could complete one or more tools. Feldt et al. (1998) reported 75% of cognitively impaired hospitalized older adults could complete the VDS. According to Dallam et al. (1995), cognitively impaired persons had an easier time completing a faces pain scale than the Visual Analog Scale (VAS). Ferrell et al. (1995) found that 83% of 217 nursing home residents could complete one or more of a variety of pain assessment tools. It would appear many cognitively impaired older adults can respond to pain scales if they are administered in a manner that addresses their unique needs.

Herr et al. (2004) evaluated the use of the FPS (Bieri et al., 1990) along with four other pain scales. This quasi-experimental study compared pain scale sensitivity among older and younger cohorts in response to experimentally induced painful stimuli. The FPS was a valid measure of pain intensity and was easy to use for both young and old subjects, but it was not the most preferred tool among older participants. The 21-point NRS and the VDS were the most sensitive and most preferred tools; however, the number of cognitively impaired participants was small.

Stuppy (1998) conducted a study to determine the reliability and validity of the FPS using a sample of 60 people age 55 or older, including 24 African-Americans. Subjects were asked to rate their pain using four scales: the FPS, the 0–10 NRS, the VAS, and the VDS. Subjects were asked to identify the scale they preferred to describe their pain. Findings revealed the FPS was reliable, valid, and sensitive to change in cognitively intact older adults. Most subjects (53%) preferred the FPS and the NRS (30%).

Carey et al. (1997) conducted a study to identify which of three pain intensity measurement scales was most appropriate and effective for hospitalized inpatients. The sample included 267 patients, of which 31% were older than 60 years of age. Three scales were used, including a VAS, a NRS, and the Wong and Baker (1988) FPS. Patients rated pain using all of the scales and then answered questions about ease of use and helpfulness in assessing pain. The scale selected most frequently was the FPS (48.6%), followed by the NRS (35.3%) and the VAS (16.1%). A reliability coefficient of 0.88 was reported, but it is unclear which data were used to determine this coefficient.

In a study with cognitively impaired long-term care residents, Wynne, Ling, and Remsburg (2000) investigated the usefulness of four pain intensity scales, including a Verbal Rating Scale (VRS), VAS, FPS and the McGill Word Scale. Sixty-one percent of the subjects were able to use the FPS to rate current pain, followed by the VAS (57%) and the VRS (51%). Residents with lower cognitive functioning experienced more difficulty completing the instruments than those with milder degrees of cognitive impairment.

Krulewitch et al. (2000) compared self-reports of pain by 156 cognitively impaired older adults (ages 65 to 98) living in the community with reports from their caregivers. The sample’s mean Mini Mental State Exam (MMSE) score was 15.7. The MMSE is a widely used clinical instrument for assessing cognitive status in the geriatric population, and is accepted as a reliable and valid tool (Folstein, Folstein, & Mchugh, 1975). Participants rated their pain using the nonverbal VAS, the FPS, and the Philadelphia Pain Intensity Scale (PIS). Findings revealed that 65 (42%) of the cognitively impaired subjects could complete all three of the scales. Of the 39 participants who completed either one or two scales, 17 (44%) completed the FPS, 17 (44%) completed the VAS, and 31 (79%) completed the PIS. The PIS was most likely to be completed regardless of impairment level. It is important to note, however, that as cognitive impairment increased, the ability to use the scales decreased.

Scherder and Bouma (2000) explored the use of three scales—the colored analogue scale (CAS), the FPS, and the Facial Affective Scale (FAS)—with early Alzheimer’s disease (AD) and midstage AD participants. Fifty percent to 100% of the early AD participants comprehended the scales, with 20% to 80% of the midstage AD group having comprehension. The researchers found that as cognitive impairment increased, reported pain severity decreased.

As the older population continues to increase, the number of older adults experiencing acute and chronic pain also will increase. Although evidence to support the use of pain intensity scales in cognitively intact older adults is accruing, there is limited empirical support for the use of pain scales with older adults with various levels of cognitive impairment. Identifying and using reliable and valid scales is a first step toward improved pain assessment and treatment.

Research questions

The focus of this research was to evaluate the reliability and validity of selected pain intensity scales for use with cognitively intact and cognitively impaired older adults residing in assisted living facilities. These were the research questions:

  1. Can cognitively impaired and cognitively intact older adults use selected pain intensity measures to rate their pain?
  2. What is the association between the FPS, the VDS, the NRS, and the IPT when used to rate current pain for cognitively impaired and intact older adults?
  3. What is the reliability of responses on all scales when used to rate a vividly remembered pain at a 2-week interval for cognitively impaired and intact older adults?
  4. What is scale preference, and is scale preference related to cognitive status, educational level, age, or gender?


Pain Intensity Measures and Cognitive Status Measure

Four pain intensity scales were used in this study: the IPT, the VDS, the NRS, and the FPS. The IPT was selected because of its increased sensitivity and response options identified in preliminary research. In an earlier study using verbal descriptor scales, Herr (2000) reported that subjects noted the need for more response options. In response to these findings and anecdotal clinical reports, the VDS was modified to provide more options between words and aligned with a pain thermometer to assist with pain conceptualization. This revised tool (the IPT) was tested in a sample of Caucasian older adults and found to be reliable and valid, as well as the preferred tool among the majority of elders in the study.

The VDS used in this study featured these terms to describe different levels of pain intensity: no pain, mild pain, moderate pain, severe pain, extreme pain, the most intense pain imaginable (Herr, et al., 2004) The words were scored by assigning numbers (0–6) to each adjective. This VDS was chosen because of established psychometric validation with older adults (Herr, 2000; Herr et al. 2004) and the belief that simple words were easier to understand for older adults.

A 0- to 10-point NRS was chosen, with word anchors of no pain at one end of the scale, moderate pain in the middle, and worst possible pain at the opposite end of the scale. The NRS used in this study is a combination of the horizontal numeric rating scale and word anchors. It was selected because it was believed fewer number options would be less challenging for older adults, and because the NRS is clinically the more commonly used numeric scale (U.S. Department of Health and Human Services, 1992, 1994).

Although the FPS has been revised, the revised tool was not available for this study’s data collection. The FPS used in this study was developed by Bieri et al. (1990) for use with children, and it features seven line-drawn faces presented in a horizontal format. Participants were instructed to point to the face that best reflected the intensity of their pain. The FPS has preliminary psychometric evaluation when used with older Caucasian and African-American samples with acceptable reliability and validity data (Herr, Mobily, Kohout, & Wagenaar, 1998; Stuppy, 1998). In the present study, the FPS was used as previously modified by Herr et al. (1998) and used in a study by Taylor and Herr (2002, 2003). The height of the faces was increased to 4 cm to allow for adequate visualization of facial characteristics. Facial markings also were darkened and slightly separated to enhance visualization.

Participants were administered the MMSE to determine the level of cognitive impairment. It provides a concise, yet formal and relatively thorough measure of cognition. The test has a score range of 0 to 30, with 30 being a perfect score. For purposes of this study, a score of 24 or lower was considered cognitively impaired, a score of 24 to 20 represented mild impairment, 19 to 16 was moderate impairment, and 15 or lower represented severe impairment. A score of 25 or higher was considered cognitively intact (Folstein et al., 1975).


Each potential participant was contacted by an agency staff member and then by the researcher. An interview was conducted individually with each participant. This study was an extension of earlier studies conducted by Taylor and Herr (2002, 2003). The initial phase of this research began with a study to determine the preliminary psychometric properties of the FPS with an African-American sample (Taylor & Herr, 2002). The next study examined the psychometric properties of selected pain intensity scales for use with African-American older adults (Taylor & Herr, 2003).

The present study focused on evaluating concurrent validity, test-retest reliability, ability to use selected scales, scale preference, and factors influencing scale preference. Procedural steps included the following:

  1. Subjects were shown the pain instruments and taught how to use them with the teaching guidelines developed by McCaffrey and Pasero (1999).
  2. Subjects were asked to rate their present pain (or lack of pain) using all four scales.
  3. Subjects were instructed to recall a vividly remembered pain and rate this remembered pain using all four scales. Recalled (remembered) pain was selected because it should not change over time, permitting a more valid retest at a 2-week interval for tool reliability.
  4. Subjects completed the MMSE after pain tools were administered to minimize anxiety or concern regarding their ability to complete the mental status questionnaire.
  5. The researcher returned 2 weeks later, reminded subjects of the “vividly remembered pain” identified earlier, and asked them to rate that pain again using all four scales.
  6. At the completion of the session, the subjects were asked to identify the tool “that helped you best describe the severity of the pain.”


The convenience sample comprised 66 subjects age 60 and older residing in assisted-living facilities in the South. The sample included 22 older men (33%) and 44 older women (67%). Ninety-eight percent (65) of the sample comprised Caucasian participants, with the exception of 1 African-American man. Seventy-one percent (n = 47) completed high school or had some higher education (e.g., college education or had acquired an undergraduate or graduate degree). Seventeen percent (n = 11) completed 7th through 11th grade, and 12% (8) completed 6 or fewer grades. The average age was 81, with an age range of 60 to 96. The mean score for the MMSE was 16 (SD = 7.62) with a range of 1 to 29. Twenty-one percent (n = 14) scored 24 to 20, indicating mild impairment, 17% (n = 11) scored 19 to 16, indicating moderate impairment, and 47% (n = 31) scored 15 and lower, indicating severe impairment. Eighty five percent (n = 56) of participants had some degree of cognitive impairment. Only 15% (n = 10) scored 25 or higher, indicating no cognitive impairment.

Approval was obtained from participants before data collection. Additionally, permission was obtained from the administrative personnel of each healthcare agency prior to contacting individuals receiving services. Participants were contacted, and the consent form was explained and discussed with each person. If the researcher noted the participant had difficulty understanding the consent process or form, that person was excluded. All subjects provided signed consent prior to participation.


Research Question # 1: Ability to Use the Scales to Rate Pain

All of the subjects, with the exception of one individual with severe cognitive impairment, were able to use all of the pain intensity scales to rate current pain in a manner that allowed interpretation of a single pain score. If participants selected more than one response, they were reminded to select only one option.

Reported severity of present pain using all of the pain intensity scales decreased as cognitive impairment increased, with cognitively intact subjects reporting more pain severity. Using the IPT as an example, cognitively intact subjects reported pain intensity ratings of 2.4 (SD = 3.2), the mildly impaired subjects reported 2.0 (SD = 2.3), the moderately impaired subjects reported pain intensity of 1.27 (SD = 2.4), and the severely impaired subjects reported pain intensity of 1.06 (SD = 2.18).

Research Question #2: Concurrent Validity

When subjects were initially contacted, they were asked to rate their present pain using the IPT, the VDS, the NRS, and the FPS. Spearman correlations between pain ratings on the selected scales were statistically significant (p < .01) and ranged between .96 and .97 in the cognitively intact group, and between .48 and .82 in the cognitively impaired group. The lowest correlations in the cognitively impaired group were found between the FPS and the other scales (.48 to .53). With the exclusion of the FPS, correlations were much stronger, ranging from .77 to .86.

Research Question #3: Test-Retest Reliability at a 2-Week Interval

Spearman rank correlation coefficients between the 2-week vividly remembered pain ratings ranged from .26 to .67 in the impaired group. The VDS had the strongest reliability coefficient (.67), followed by the IPT (.50), NRS (.47), and FPS (.26). In the intact group, the strongest correlation was the FPS (.85), followed by the VDS (.75), NRS (.68), and IPT (.67).

Research Question #4: Scales Preference and Scale Preference as Related to Age, Gender, Educational Level, and Cognitive Status

Eighty-three percent (n = 45) of the cognitively impaired group selected a preferred pain intensity scale, and 90% (n = 9) of the cognitively intact group indicated a preference. Overall, the participants in both groups preferred the VDS (n = 17, 31%) and the IPT (n = 16, 30%). The VDS (n = 4) and the NRS (n = 4) were the preferred scales in the intact group, and the IPT (n = 16, 36%) was the preferred scale in the impaired group.

Statistical analysis of the relationships between selected demographic variables and tool preference was not possible due to small numbers; however, trends are reported in Table 1, which summarizes responses relative to tool preference and selected demographic variables.


These findings demonstrate this sample of Caucasian, cognitively impaired and intact older adults was able to use each of the pain intensity scales to rate their pain. Although the majority of this sample was cognitively impaired and some severely impaired, participants could understand instructions, communicate, and provide consent for participation. The fact that they could use the tools is consistent with the findings of other researchers (Feldt et al., 1998; Ferrell et al., 1995; Manz et al., 2000; Weiner et al., 1998), who concluded that cognitively impaired elders can complete selected pain intensity scales. Although examining the severity of reported pain was not the focus of this study, it is interesting to note that the results support previous findings (Fries et al., 2001; Herr et al. 1998; Parmelee et al., 1993; Scherder & Bouma, 2000) regarding the tendency for those with increasing levels of cognitive impairment to report less pain severity. Because pain intensity scores were relatively low, the scales were not tested across a continuum of pain sensations, which may be a limitation in evaluating psychometric properties of the scales used in this study.

In terms of the concurrent validity, moderate to high intertool correlations for impaired and intact groups were found, with the exception of low correlations associated with the FPS. These findings are similar to previous findings using these tools with African-American older adults with cognitive impairment. In an earlier study, Taylor and Herr (2002) reported correlations ranging from .50 to .93 using the FPS, NRS, VDS, and the vertical visual analogue scale (v-VAS). When the FPS was excluded, the correlations between the remaining pain intensity scales ranged from .81 to .93. In a more recent study using the same scales, Taylor and Herr (2003) reported correlations ranging from .74 to .96 in both the impaired and the intact African-American groups with the lowest correlations noted between the FPS and the VDS (.74).

In studies with primarily cognitively intact older adults, the intertool correlations are considerably higher and consistent with the findings in this sample’s intact group. Stuppy (1998) reported inter-tool correlations of .81 to .95 between the FPS, 0–10 NRS, VDS, and VAS for current level of pain in African-American (n = 24) and Caucasian (n = 36) older adults. In a study by Herr and colleagues (2004), correlations between the FPS, IPT, VDS, and NRS ranged from .79 to .81 in the intact group.

Test-retest reliability coefficients for vividly remembered pain at a 2-week interval ranged from .26 to .67 in the cognitively impaired group, and .67 to .85 in the intact group. The weaker relationships found in the cognitively impaired group also have been noted by others, and are attributed to memory impairment deficits in this particular group. These findings are similar to an earlier study of African-American older adults with test-retest reliability coefficients ranging from .52 to .83 in cognitively intact and impaired older adults using the same tools (Taylor & Herr, 2002). In previous studies with cognitively intact and impaired older adults, test-retest reliability for the FPS was .72 (Taylor & Herr, 2002) and .70 (Stuppy, 1998). Correlation coefficients were not provided for other scales used in these studies because the overriding study purpose was to test the FPS; consequently, those data are not available for comparison. Given the higher stability measures for the VDS in both cognitively impaired (.67) and intact (.75) older adults when compared to the other scales, researchers should consider this scale when assessing pain intensity in older persons with cognitive impairment.

Overall, the findings do not support the stability of pain reports over a 2-week interval using these selected scales with cognitively impaired older adults, but the findings do support reproducibility in the intact group. Memory impairment in this primarily impaired sample is to be expected, however, and must be taken into consideration. Some of the participants had medical diagnoses contributing to memory impairment. Additionally, a study by Buffum, Miaskowski, Sands, and Brody (2001) has raised concern about the reliability of short-term pain reports in cognitively impaired older adults. As a result, further study in this area is warranted. Scherder and Bouma (2000) maintain that only pain at the moment can reliably be assessed in cognitively impaired older adults (specifically, AD patients) because of memory decline. Findings from the present study revealed the IPT and the VDS as the most preferred scales in a sample that included primarily impaired older adults. The VDS had acceptable stability over time, with a coefficient of .67 (impaired group) and .75 (intact group). The moderately high correlations between the VDS and remaining scales make the VDS a viable option for use with impaired older adults. In addition, women and all adults over the age of 85 preferred the VDS and the IPT. Cognitively impaired participants preferred the IPT and the VDS.

This research supports findings from Herr’s study (2004), in which participants selected the NRS and the VDS as preferred scales, lending support to the use of the VDS with older adults. In contrast, Taylor and Herr (2003) found that a sample of older African-American adults preferred the FPS. Stuppy (1998), who worked with a sample of African-American older adults, reported that 53% of subjects preferred the FPS, followed by the NRS (30%). The fact that African-Americans preferred the FPS suggests potential ethnic differences in tool preferences. Because the FPS did not correlate highly with other tools, however, it may be assessing a broader construct than pain, as suggested earlier by Taylor and Herr (2002).

Pain in older adults continues to be a problem (Fries et al., 2001), and can lead to impaired mobility and inadequate functional assessment by healthcare personnel. Failure to accurately assess pain’s influence on the function and activity levels of cognitively impaired older adults can lead to inappropriate treatment and inaccurate treatment goals. In addition, incorrectly assessing the functional abilities of a cognitively impaired older adult has serious implications for institutional placement and determining needs for services.

These findings show that cognitively impaired older adults would benefit from the inclusion of a pain assessment instrument with functional assessment. Functional abilities such as independence in feeding, transfer, and ambulation may very well relate to a patient’s pain, particularly pain with movement. The routine use of an appropriate pain assessment scale such as the IPT or VDS could result in a more accurate assessment of functional capacity, leading to improved treatment and enhanced function. A multidimensional patient-oriented approach that includes pain assessment is essential when performing the functional assessment of a cognitively impaired older adult. Additionally, research has suggested that a pain intensity rating above 3 may interfere with daily functioning (McCaffrey & Pasero, 1999). McCaffrey and Pasero suggest asking patients to identify a pain intensity rating that would allow sleep and participation in activities. It is important to consistently and accurately assess pain intensity in patients in the rehabilitation setting because participation in exercise and activities of daily living is vital to improved or maintained function. The consistent use of pain assessment scales provides a necessary communication tool for the healthcare team in the rehabilitation setting.

In summary, these findings reinforce that older adults with and without cognitive impairment are capable of using a variety of scales, and more than one scale may be appropriate for use in an older cognitively impaired group. The fact that a majority of the sample was cognitively impaired and could use the various scales suggests the FPS, VDS, IPT, and NRS are relatively easy to use. This finding further supports the mounting evidence about cognitively impaired individuals’ ability to respond accurately to assessment tools (Feldt et al., 1998; Ferrell et al., 1995; Herr et al., 2004; Parmelee et al., 1993; Taylor & Herr, 2002, 2003). If cognitively impaired older adults are not able to use these self-report scales, a checklist of nonverbal pain indicators, such as the one developed by Feldt (2000), may be helpful.

It is important to recognize the limitations of this study, which include small sample size, the lack of control for participant diagnoses, and the lack of scale differentiation among those with various levels of cognitive impairment. The fact that severely cognitively impaired participants were able to complete the tools suggests they could follow instructions, but comprehension may be in question because participants were coached or prompted if they needed help while completing the tools. Because 29% of the sample only completed 11 or fewer grades, those with impairment may reflect educational level rather than true cognitive impairment.


Based on this study’s findings, there is support for the use of the IPT and the VDS with cognitively impaired older adults. Moderate to high correlations found among these two scales (both groups) and the NRS supported the concurrent validity of the scales. Additionally, test-retest reliabilities were most acceptable using the VDS in the cognitively intact and impaired groups. Based on overall poor test-retest reliabilities in cognitively impaired elders (most likely related to memory impairment), it is recommended that these scales only be used to measure present pain intensity.

The VDS and the IPT were easy to use and understandable in both cognitively intact and impaired older adults, warranting their use in the clinical setting. Because few studies have addressed cognitive impairment and the use of pain assessment scales, continued research validating the usefulness of these pain intensity measures is warranted. When assessing pain in cognitively impaired older adults, remember to allow time to assimilate questions and formulate responses. Adequate lighting and hearing devices (if needed) also should be provided. Large print for ease of reading is helpful. The fact that coaching was helpful in this study may warrant investigation of coaching as a useful intervention in future research studies.

This study was limited by its small convenience sample and lack of control concerning medical diagnoses. Additional research that will allow in-depth analysis of the impact of cognitive level and disability, education, gender, ethnicity and cultural influences, and perceptions of various pain intensity scales is needed with larger samples. Tool preference among ethnic groups warrants continued investigation, as well.

About the Authors

Laurie Jowers Taylor, PhD RN, is professor and coordinator, graduate program, Department of Nursing, State University of West Georgia, Carrollton, GA. Judy Harris, MSN ARNP CRRN CCR CCM, is clinic director: nurse practitioner, St. Marks Powder, General Dynamics Co., St. Marks, FL. Cynthia D. Epps, PhD RN, is associate professor, Department of Nursing, State University of West Georgia. Keela Herr, PhD RN FAAN, is professor and chair, Adult & Gerontological Nursing, College of Nursing, The University of Iowa, Iowa City, IA.

Address correspondence to Laurie Jowers Taylor, 215 Shady Valley Drive, Carrollton, GA 30116.


American Geriatrics Society Panel on Chronic Pain in Older Persons. (2002). The management of chronic pain in older persons. Journal of the American Geriatric Society, 50, 1–20.

American Pain Society. (1999). Principles of analgesic use in the treatment of acute pain and chronic cancer pain. (4th ed.). Glenview, IL: Author.

Bernebei, R. (1998). Management of pain in elderly patients with cancer. Journal of the American Medical Association, 279, 1877–1882.

Bieri D., Reeve R., Champion G., Addicoat, L., & Zeigler, J. B. (1990). The Faces Pain Scale for the self-assessment of the severity of pain experienced by children: Development, initial validation, and preliminary investigation for ratio scale properties. Pain, 41, 139–150.

Buffum, M., Miaskowski, C., Sands, L., & Brody, M. (2001). A pilot study of the relationship between discomfort and agitation in patients with dementia. Geriatric Nursing, 22, 80–85.

Carey, S. J., Turpin, C. O., Smith, J., Whatley, J., Haddox, D., & Hunt, A. (1997). Improving pain management in an acute care setting: The Crawford Long Hospital of Emory University experience. Orthopedic Nursing, 16, 29–36.

Cockrell, J. R., & Folstein, M. F. (1988). Mini-Mental State Examination (MMSE). Psychopharmacology Bulletin, 24, 689–691.

Dallam, L., Smyth, C., Jackson, B.S., Krinsky, R., O’Dell, C., Rooney, J., et al. (1995). Pressure ulcer pain: Assessment and quantification. Journal of Wound Ostomy and Continence Nurses, 22, 211–218.

Feldt, K., Ryden, M., & Miles, S. (1998). Treatment of pain in cognitively impaired compared with cognitively intact older adult patients with hip-fracture. Journal of the American Geriatric Society, 46, 1079–1085.

Feldt, K. (2000). The checklist of non-verbal pain indicators. Pain Management Nursing, 1(1), 13–21.

Ferrell, B. A., & Ferrell, B. R. (1990). Geriatric pain management. Provider, 16, 31–32.

Ferrell, B. A., Ferrell, B. R., & Osterweil, D. (1990). Pain in the nursing home. Journal of the American Geriatric Society, 38, 409–414.

Ferrell, B. A., Ferrell, B. R., & Rivera, L. (1995). Pain in cognitively impaired nursing home patients. Journal of Pain and Symptom Management, 10, 591–598.

Folstein, M. F., Folstein, S. E., & McHugh, P. R. (1975). “Mini-mental state”: A practical method for grading cognitive state of patients for the clinician. Journal of Psychiatric Research, 12, 189–198.

Fries, B. E., Simon, S. E., Morris, J. N., Flodstrom, C., & Bookstein, F. L. (2001). Pain in U.S. nursing homes: Validating a pain scale for the minimum data set. Gerontologist, 41,173–179.

Herr, K. (2000, November). Assessment of pain intensity in older adults: Is there a better way? Paper presented at the meeting of the American Pain Society, Atlanta, GA.

Herr, K. A., Mobily, P. R., Kohout, F. J., & Wagenaar, D. (1998). Evaluation of the Faces Pain Scale for use with the elderly. Clinical Journal of Pain, 14, 29–38.

Herr, K., Spratt, K., Mobily, P. P., & Richardson, G. (2004). Pain intensity assessment in older adults: Use of experimental pain to compare psychometric properties and usability of selected scales in adult and older populations. Clinical Journal of Pain, 20(4), 207–219.

Horgas, A.L., & Tsai, P. (1998). Analgesic drug prescription and use in cognitively impaired nursing home residents. Nursing Research, 47, 235–242.

Joint Commission on Accreditation of Healthcare Organizations. (1999). Pain Management Standards. Retrieved March 20, 2003, from http://www.jcaho.org.

Krulewitch, H., London, M. R., Skakel,V. J., Lundstedt, G. J., Thomason, H., et al., (2000). Assessment of pain in cognitively impaired older adults: A comparison of pain assessment tools and their use by nonprofessional caregivers. Journal of the American Geriatric Society, 48, 1607–1611.

Lavsky-Shulan, M., Wallace, R. B., Korout, F. J., Lemke, J. H., Morris, M. C., & Smith, I. M. (1985). Prevalence and functional correlates of low back pain in the elderly: The Iowa 65+ Rural Health Study. Journal of the American Geriatric Society, 33, 23–28.

Manz, B. D., Mosier, R., Nusser-Gerlach, M. A., Bergstrrom, N., & Agrawal, S. (2000). Pain assessment in the cognitively impaired and unimpaired elderly. Pain Management Nursing, 1, 106–115.

McCaffrey, M., & Pasero, C. (1999). Pain: Clinical manual (2nd ed.). St. Louis, MO: Mosby.

McCaffrey, R., Frock, T., & Garguilo, H. (2003). Understanding chronic pain and the mind-body connection. Holistic Nursing Practice, 17(6), 281–289.

McCracken, L. M., Evon, D., & Karapas, E. T. (2002). Satisfaction with treatment for chronic pain in a specialty service: Preliminary prospective results. European Journal of Pain, 6, 387–393.

Parmelee, P. A., Smith, B. D., & Katz, I. R. (1993). Pain complaints and cognitive status among elderly institution residents. Journal of the American Geriatric Society, 41, 517–522.

Scherder, E., Bouma, A., Borkent, M., & Rahman, O. (1999). Alzheimer patients report less pain intensity and pain affect than non-demented elderly. Psychiatry, 62, 265–272.

Scherder, E. J., & Bouma, A. (2000). Visual analogue scales for pain assessment in Alzheimer’s disease. Gerontologist, 46, 47–53.

Sengstaken, E. A., & King, S. A. (1993). The problem of pain and its detection among geriatric nursing home residents. Jouranl of the American Geriatric Society, 41, 514–544.

Stuppy, D. (1998). The Faces Pain Scale: Reliability and validity with mature adults. Applied Nursing Research, 11, 54–89.

Taylor, L .J., & Herr, K. (2002). Evaluation of the FPS with minority elders. Journal of Gerontological Nursing, 27, 15–23.

Taylor, L. J., & Herr, K. (2003). Pain intensity assessment: A comparison of selected pain intensity scales for use in cognitively intact and cognitively impaired African American older adults. Pain Managment Nursing, 4, 87–95.

U.S. Department of Health and Human Services. (1992). Acute Pain Management: Operative or Medical Procedures and Trauma Clinical Practice Guidelines (AHCPR Publication No. 92-0032). Rockville, MD: Author.

U.S. Department of Health and Human Services. (1994). Management of Cancer Pain: Clinical Practice Guidelines (AHCPR Publication No. 94-0592). Rockville, MD: Author.

Weiner, D., Peterson, B., & Keefe, F. (1998). Evaluating persistent pain in long term care residents: What role for pain maps? Pain, 76, 249–257.

Wong, D. L., & Baker, C. (1988). Pain in children: Comparison of assessment scale. Pediatric Nursing, 14, 9017–9021.

Wynne, C. F., Ling, S. M., & Remsburg, R. (2000). Comparison of pain assessment instruments in cognitively intact and cognitively impaired nursing home residents. Geriatric Nursing, 21(1), 20–23.