JADA Continuing Education
The application of facial expressions to the assessment of orofacial pain in cognitively impaired older adults
Kuo-Tung Hsu, DDS, MS,
Stephen K. Shuman, DDS, MS,
Darryl T. Hamamoto, DDS, PhD,
James S. Hodges, PhD and
Karen S. Feldt, PhD, ARNP, GNP
 |
ABSTRACT
|
|---|
Background. The anticipated rapid growth in the number of cognitively impaired older adults, declining edentulism and increasing oral health expectations suggest a greater need for comprehensive dental care and effective ways to evaluate orofacial pain in people with compromised mental function and impaired communication skills. The authors conducted a study to evaluate facial expressions as a means of identifying orofacial pain in cognitively impaired and cognitively intact older adults, compared with other available pain assessment tools.
Methods. The authors conducted a prospective comparative study using three alternative pain measurement tools in a sample of 22 older adults. They divided subjects into cognitively impaired and cognitively intact groups on the basis of their mental status examination scores. The pain measurement methods evaluated were facial expressions quantified by the Facial Actions Coding System (FACS); self-reported pain via the Verbal Descriptor Scale; and physiological response to pain via changes in heart rate. The pain stimuli were local anesthetic injections in subjects who required them for routine dental procedures.
Results. The average FACS scores during anesthetic injections were significantly higher than those during the preinjection period (prebuccal versus buccal, P = .016; prepalatal versus palatal, P = .0002). The differences between preinjection and injection segments were even higher in cognitively impaired patients than in cognitively intact patients. There were no correlations between the three pain measurements (P > .05).
Conclusions. Changes in facial expression proved to be the most useful measure overall in identifying pain in both cognitively intact and cognitively impaired older patients. This measure appeared to be more sensitive in cognitively impaired patients because they demonstrated fewer facial movements in anticipation of pain stimuli.
Key Words: Orofacial pain; pain assessment; dementia; cognitive impairment; facial expressions; geriatrics; agingAbbreviations: AUs: Action units • FACS: Facial Actions Coding System • MMSE: Mini-Mental State Examination • VDS: Verbal Descriptor Scale
Pain is prevalent and undertreated among elderly people, especially in those with cognitive impairment.1 Communication of pain often relies on oral reports, but oral reports from cognitively impaired elderly patients may be difficult to obtain or unreliable because of short-term memory loss or impairment in language skills.2 Pain assessment tools have been developed to improve identification of pain in this population, but none is designed specifically for orofacial pain and few have been tested in dental research. Unrecognized and untreated oral pain can have an immediate and profound impact on a person’s well-being and lead to a diminished quality of life.3 The rapid growth of the elderly population,4 a growing prevalence of cognitive impairment5 and declining edentulism6 all suggest a need for practical, reliable pain assessment tools to evaluate orofacial pain for the increasing number of cognitively impaired people who will require dental care in the coming years.
Facial expressions during painful events have shown promise as a means of conveying important information about the presence of pain and have shown some consistent patterns.7 Since dental practitioners focus their attention on the head and neck area, facial pain expressions could be particularly useful in identifying orofacial pain in cognitively impaired people who have difficulty expressing themselves otherwise. In fact, dental practitioners probably already monitor facial expressions to some degree as evidence of pain in their patients, but this strategy has not been evaluated carefully or emphasized.
Therefore, we conducted an investigation to explore the utility of pain-related facial expressions scientifically for the purpose of evaluating orofacial pain in elderly subjects with and without cognitive impairment.
|
METHODS
|
|---|
Study design.
We used a prospective, comparative design to evaluate the experience of orofacial pain in a sample of older adults with and without cognitive impairment, using three pain measurement strategies:
- – changes in facial expressions, as quantified by the Facial Actions Coding System (FACS);
- – self-reported pain, as measured by the Verbal Descriptor Scale (VDS);
- – physiological response to pain, as measured by heart rate before and after the receipt of pain stimuli.
As the pain stimuli for this study, we chose buccal and palatal local anesthetic injections administered by one fellowship-trained geriatric dentist (D.T.H.) in conjunction with routine dental treatment for all subjects. We recruited a nonrandomized sample of older adults (aged 65 years or older) needing maxillary right quadrant dental treatment under local anesthesia via the enrollment process described below. We divided the subjects into two groups: cognitively impaired versus cognitively intact (as judged on the basis of scores on the Mini-Mental State Examination [MMSE]).8 We then compared the results of the three pain assessment strategies in these two groups.
Pain assessment instruments.
The pain assessment measures we chose helped us evaluate pain in three ways: via facial expression, via the subject’s self-report and via physiological changes.
FACS.
Since a structured approach was necessary to quantify the types of facial expressions that could be observed by dental professionals during examination and treatment, we used the FACS, a comprehensive anatomically based system for objectively describing facial activities. The FACS describes visually distinguishable facial activity based on 44 unique action units (AUs). The FACS has been used extensively in pain research9 and validated for pain assessment in older people with and without cognitive impairment,10 although not for orofacial pain as typically observed by dental professionals. Among the 44 AUs, six have the most consistent association with all varieties of pain: brow lowering, cheek raising, lid tightening, nose wrinkling, upper lip raising and eye closing.11 Therefore, we chose these facial AUs as the only facial actions assessed in this study. From the frequency and the intensity of these AUs, we generated a FACS score for a subject during a period we prescribed. A FACS score consisted of the sum of the intensity of brow lowering, orbital tightening, nose wrinkling and upper lip raising multiplied by the duration of these facial expressions and then added to the duration of the eyelid closing.10 Prior studies validated this index as a suitable method of quantifying pain expressions.11,12
VDS.
The VDS consists of seven levels of intensity, from "no pain" to "pain as bad as it could be." The VDS can be self-assessed or assessed by others, and in this study it was administered by the lead investigator (K.-T.H.). This instrument has been shown to be easier to complete for cognitively impaired, hospitalized older people,13 which is why we chose it as a second assessment instrument for this study.
Heart rate.
Physiological changes also can accompany the experience of pain. Since heart rate changes commonly are used to quantify this, we also included this measurement.14
Procedures.
The institutional review boards of the University of Minnesota approved the study protocol. We obtained additional approvals from two collaborating community organizations that sponsor university-affiliated dental clinics for older adults in the Minneapolis–St. Paul, Minn., metropolitan area: the Presbyterian Dental Program at Presbyterian Homes of Arden Hills, Minn., and the Amherst H. Wilder Foundation Senior Dental Clinic in St. Paul, Minn. We obtained written informed consent from subjects, if possible, or from their responsible parties otherwise. On each subject’s day of participation, he or she underwent the MMSE at his or her usual clinic. We then placed a finger-clip pulse oximeter to monitor heart rate and positioned a camcorder (Camcorder 1) in front of the pulse oximeter to videotape the subject’s heart rate readings. We asked the subject to rate his or her current magnitude of pain using the VDS. Another camcorder (Camcorder 2) was positioned at a fixed distance from the subject to videotape facial expressions during procedures.
A single dentist (D. T.H.) performed all procedures at both participating clinics. He administered all of the local anesthetic injections as pain stimuli in the maxillary right quadrant for all subjects, performing a buccal injection and then a palatal injection. Before the needle penetrated the tissue, the dentist administered a 10-second sham injection (a "prebuccal injection"). For this, the dentist held the syringe with the needle close to, but not penetrating, the oral tissue. After this sham injection, the dentist administered the buccal injection. The lead investigator (K.-T.H.) then asked each subject to rate the magnitude of pain of the injection using the VDS. The same procedures were repeated for the palatal injection.
Using computer software, we edited videotape of the subjects’ facial expressions to prepare for FACS coding. For every subject, we selected four 10-second segments: the 10-second prebuccal injection, the first 10 seconds of the buccal injection, the 10-second prepalatal injection and the first 10 seconds of the palatal injection. A final videotape was made showing, in randomized order, prebuccal, buccal, prepalatal and palatal injection video segments of each subject (Figures 1
and 2 are video images of preinjection and injection segments). We then sent videotapes for scoring by certified FACS coders at the University of Northern British Columbia, Prince George, British Columbia, Canada. We edited the film from Camcorder 1 for heart rate readings in the same way, and we adopted the concurrent video segment corresponding to the segment selected from the film in Camcorder 2 to represent heart rates for that period.
Data analysis.
For each subject, we calculated the effect of heart rates and the FACS scores for each of the four segments. The averages of VDS self-reported pain ratings (we assigned a score ranging from 0, meaning "no pain," to 6, meaning "pain as bad as it could be," to represent the seven-level pain rating) were available for buccal and palatal segments. We conducted two one-way analyses of variance (ANOVAs) with four levels (injection segments A = prebuccal, B = buccal, C = prepalatal and D = palatal) by two groups (cognitively intact and cognitively impaired), using FACS scores and heart rates as the dependent variables to analyze the differences among the cognitively intact and impaired groups and among the four injection segments. We conducted one-way ANOVA with two levels (buccal and palatal injection segments) by two groups (cognitively intact and cognitively impaired) using VDS as the dependent variable to analyze the differences among the cognitively intact and cognitively impaired groups and among the two injection segments. We calculated Pearson correlation coefficients to analyze the correlations among the three different pain measurements within each time segment for the subjects overall and separately among the cognitively impaired and cognitively intact groups.
|
RESULTS
|
|---|
Subjects’ characteristics.
We enrolled 22 subjects in this study; Table 1 summarizes their demographic and descriptive data. Using a mixed linear model, we tested the associations of sex and age with the three pain measurements, and we found that neither age nor sex was related significantly to any of the pain measurements (P > .05).
FACS scores.
Table 2 presents average FACS scores during each preinjection and injection segment, and Table 3 shows comparisons for each possible pair of injection segments. Table 3 reveals that the average FACS scores from the injection segments were significantly higher than those from the preinjection segments (segments A and B, P = .016; segments C and D, P = .0002), and palatal injection segments had the highest FACS scores among all segments.
Table 4 further shows the FACS score averages at each injection segment within the impaired and intact groups, respectively. On average, the FACS scores of the cognitively intact group were eight points higher than those of the cognitively impaired group, but the two groups did not differ significantly (P = .54). The pattern of FACS scores across the injection segments did not differ between the two groups (impairment by time interaction, P = .098). The scores during the preinjection segments for the impaired group also were much lower than those for the intact group. In addition, the difference between the preinjection and injection segments also varied between the groups (Figure 3, page 968). The scores of the cognitively intact group increased only about 14 points from the preinjection segments to the injection segments. However, the scores of the cognitively impaired group increased 26 points from the prebuccal to the buccal injection segment and 51 points from the prepalatal to the palatal injection segment.
View this table:
[in this window]
[in a new window]
|
TABLE 4 Facial Actions Coding System (FACS) score averages of cognitively intact and cognitively impaired subjects at each injection segment.
| |
VDS pain ratings and heart rates.
Mean VDS pain ratings were higher for palatal injections (2.9) than for buccal injections (2.4) to a statistically significant degree (P = .039). The cognitively intact group also tended to report less pain using the VDS for both buccal and palatal injections (Figure 4, page 968). The VDS pain ratings of the cognitively impaired group were 1.27 points higher on average, and the difference in mean VDS ratings between the two groups was statistically significant (P = .022).
In this study, we observed minimal average changes in heart rate from baseline. The mean heart rate change was 0.02 percent for the cognitively intact group and –0.10 percent for the cognitively impaired group, which was not statistically significant (P = .95). The heart rate changes for the four time segments also were minimal and not statistically significant (P = .76).
FACS versus VDS and heart rates.
We calculated the correlation coefficients between the three pain measurements within the cognitively intact and cognitively impaired groups (Table 5, page 969). None of the pairs of the three measurements used were correlated strongly as determined by either the magnitude of their correlation coefficients or P values.
|
DISCUSSION
|
|---|
Facial expression.
The finding of no significant differences in overall FACS scores between the cognitively impaired group and the cognitively intact group is consistent with previous studies examining facial pain expressions in elderly people9 or people with intellectual disability.10 This suggests that facial expressions relating to orofacial pain are similar in people with or without cognitive impairment. Average FACS scores were higher during injection periods than during preinjection periods, and the mean scores for the palatal injections were higher than those for buccal injections, although the latter difference was not statistically significant. These results supported the sensitivity of facial expressions in detecting pain for the stimuli we used.
FACS scores increased more between preinjection and injection segments in the cognitively impaired group than in the cognitively intact group. This may have been because cognitively intact subjects were more aware of the injection procedures and anticipated some pain, giving rise to more preinjection facial expressions and smaller net differences in FACS scores. Cognitively impaired subjects may have been less capable of predicting upcoming procedures, leading to fewer preinjection facial expressions and lower preinjection FACS scores. In addition, cognitively impaired subjects had relatively higher (but not statistically different) scores than did the cognitively intact subjects during the event expected to be more painful: the palatal injection. This result is consistent with the findings of Hadjistavropoulos and colleagues,9 who reported that more facial activity occurred in people with cognitive impairment during more painful events. It is possible that cognitively intact subjects try to mask their pain-related facial expressions for social purposes, while cognitively impaired people manifest reduced social inhibition in facial expression. These findings suggested that changes in facial expression, as quantified by the FACS, were more sensitive measures of pain among those with cognitive impairment, especially during events expected (by patients and dental professionals alike) to be more painful.
VDS pain ratings and heart rates.
The average VDS pain ratings were higher, to a statistically significant degree, for palatal injections than for buccal injections, suggesting that participants subjectively expressed more pain related to palatal injections. This is consistent with common clinical observations and expectations in dental practice. However, the cognitively intact group reported less pain on the VDS than did the cognitively impaired group, perhaps because the cognitively intact subjects were trying to be "polite" or were afraid of damaging their relationship with the dentist by reporting more pain.
Heart rate changes in response to pain in this study were minimal in magnitude and not statistically significant, which finding may have been influenced by the prevalence of cardiovascular medications in use among subjects (such as calcium channel blockers and digitalis glycosides). However, disqualification of those taking cardiovascular medications would have rendered this study virtually impossible to conduct, given the nature of the research question and the study population.
FACS versus VDS and heart rates.
FACS scores were not correlated significantly with VDS self-reports in either the cognitively impaired or the cognitively intact group. Similar results have been found in other studies using both facial expressions and self-report scales to measure pain.9,15 Hadjistavropoulos and colleagues9 suggested that self-report measures and nonverbal measures assess very different aspects of the pain experience. LeResche and colleagues15 also noted that no single measure could capture all of the variability associated with pain. In our study, FACS scores also were not correlated with heart rate changes, but this lack of correlation was expected given the insignificant alterations in this particular variable, as previously discussed.
|
CONCLUSIONS
|
|---|
Although our small sample size might not have generated enough statistical power to detect small differences in some pain measurements, and while local anesthetic injections cannot represent all orofacial pain, findings of this study suggest that using facial expressions to detect acute orofacial pain may be a useful objective tool in treating both cognitively intact and cognitively impaired elderly patients. This approach may even be more sensitive in cognitively impaired older patients because they make fewer anticipatory facial movements. An additional finding worthy of note here is that among the six facial AUs investigated in this study, upper-face expressions—brow lowering, lid tightening and cheek raising—appeared most highly correlated with orofacial pain.
While the complexity of the FACS would make it impractical for dental practitioners to use clinically on a day-to-day basis, we used it in this study to provide an objective means of analyzing facial expressions as potential indicators of acute orofacial pain symptoms of concern to dental professionals. In managing the care of a growing number of intellectually impaired patients, dental professionals will need to use such methods with patients who might not be able to report pain otherwise. The results of this study suggest that dental practitioners should remain sensitive to facial expressions as a potential means of identifying acute pain during the examination and treatment of their older patients, especially those with cognitive decline.
|
FOOTNOTES
|
|---|
When this study was conducted, Dr. Hsu was a dental fellow, Oral Health Services for Older Adults Program, Department of Primary Dental Care, School of Dentistry, University of Minnesota, Minneapolis. He now is a general dentist in private practice in Clearwater, Minn. Address reprint requests to Dr. Hsu at 6730 Troy Lane North, Maple Grove, Minn. 55311, e-mail "kuotunghsu{at}hotmail.com".
Dr. Shuman is an associate professor and the director, Oral Health Services for Older Adults Program, Department of Primary Dental Care, School of Dentistry, University of Minnesota, Minneapolis.
Dr. Hamamoto is an associate professor, Division of Oral Medicine, Department of Diagnostic and Biological Sciences, School of Dentistry, University of Minnesota, Minneapolis.
Dr. Hodges is an associate professor, Division of Biostatistics, School of Public Health, University of Minnesota, Minneapolis.
Dr. Feldt is an associate professor, College of Nursing, Seattle University.
The authors thank all those who played a pivotal role in completing the research project described in this article: Dr. Dzevdan Hodzic, who when the study was conducted was a dental fellow, Oral Health Services for Older Adults Program, Department of Primary Dental Care, School of Dentistry, University of Minnesota, Minneapolis, and now is in private practice, Westside Community Health Services, St. Paul, Minn., for his time and effort performing necessary dental procedures; Dr. Kenneth Prkachin, University of Northern British Columbia, Prince George, B.C., Canada, for his suggestions and help in FACS coding processes; and the staff of Presbyterian Homes Dental Program, Arden Hills, Minn., and Wilder Senior Dental Clinics, St. Paul, Minn., for their help and accommodations in the data collection phase of this study. The authors offer special thanks to Presbyterian Homes, Arden Hills, Minn., and the Amherst H. Wilder Foundation of St. Paul, Minn., for their tremendous support of this project and commitment to geriatric education and scholarship to improve the health care of older adults.
|
REFERENCES
|
|---|
- Gabre P, Sjöquist K. Experience and assessment of pain in individuals with cognitive impairments. Spec Care Dentist 2002;22(5):174–80.[Medline]
- Feldt KS. Improving assessment and treatment of pain in cognitively impaired nursing home residents. Annals Long-Term Care 2000;8(9):36–42.
- National Institute of Dental and Craniofacial Research. Oral health in America: A report of the surgeon general. Rockville, Md.: U.S. Department of Health and Human Services, U.S. Public Health Service; 2000.
- Centers for Disease Control and Prevention. Trends in aging: United States and worldwide. MMWR Morb Mortal Wkly Rep 2003;52(6):101–4, 106.[Medline]
- Hebert LE, Scherr PA, Bienias JL, Bennett DA, Evans DA. Alzheimer disease in the US population: prevalence estimates using the 2000 census. Arch Neurol 2003;60(8):1119–22.[Abstract/Free Full Text]
- Centers for Disease Control and Prevention (CDC). Public health and aging: retention of natural teeth among older adults—United States, 2002. MMWR Morbid Mortal Wkly Rep 2003;52(50):1226–9.[Medline]
- Zakrzewska JM, Harrison SD. Assessment and management of orofacial pain. Philadelphia: Elsevier; 2002:76.
- Folstein MF, Folstein SE, McHugh PR. Mini-mental state: a practical method for grading the cognitive state of patients for the clinician. J Psychiatr Res 1975;12(3):189–98.[Medline]
- Hadjistavropoulos T, LaChapelle DL, MacLeod FK, Snider B, Craig KD. Measuring movement-exacerbated pain in cognitively impaired frail elders. Clin J Pain 2000;16(1):54–63.[Medline]
- LaChapelle DL, Hadjistavropoulos T, Craig KD. Pain measurement in persons with intellectual disabilities. Clin J Pain 1999;15(1):13–23.[Medline]
- Prkachin KM. The consistency of facial expressions of pain: a comparison across modalities. Pain 1992;51(3):297–306.[Medline]
- Prkachin KM, Berzins S, Mercer SR. Encoding and decoding of pain expressions: a judgement study. Pain 1994;58(2):253–9.[Medline]
- Feldt KS, Ryden MB, Miles S. Treatment of pain in cognitively impaired compared with cognitively intact older patients with hip-fracture. J Am Geriatr Soc 1998;46(9):1079–85.[Medline]
- Katz J, Melzack R. Measurement of pain. Surg Clin North Am 1999;79(2):231–52.[Medline]
- LeResche L, Dworkin SF, Wilson L, Ehrlich KJ. Effect of temporo-mandibular disorder pain duration on facial expressions and verbal report of pain. Pain 1992;51(3):289–95.[Medline]
TOP
ABSTRACT
METHODS
