© Borgis - New Medicine 2/2010, s. 39-44
*Paula Madureira1, Bruno Peixoto2
Pain assessment in patients with dementia. Exploratory study in elderly people with Alzheimer's disease and vascular dementia
1Unidade Local de Saúde de Matosinhos (Matosinhos – Portugal)
2Sciences Department, Instituto Superior de Cięncias da Saúde-Norte (Gandra-Portugal);
Research Unit on Psychology and Health (UnIPSa)
Background. The assessment of pain in dementia patients is a controversial issue and we have witnessed an increasing number of studies in this field. Some advocate the impossibility of using direct pain measures in this population, while others support the view that elderly people with dementia are able to report their pain in a trustworthy manner.
Objective. This work aims to test the possibility of using direct measures of pain in patients with dementia and, at the same time, to characterize the experience of pain in these individuals and to correlate these measures with observational ones. Method.Self-report and behavioural pain measures were applied to three groups: an Alzheimer's Disease (AD) group (n=15); a Vascular Dementia (VD) group, (n=20); and a Control group (n=22).
Results. The obtained results suggest hypointensity in the pain experience in AD patients both in direct and observational measures. This group obtained lower scores on both direct and observational pain measures.
Conclusions. This study suggests that the use of pain self-report measures in early stages of dementia is perfectly plausible. The finding that the experimental groups did not differ from the control group in regard to comprehension of the scales, and the high positive correlation between this type of measure and the observations, support this idea.
Approximately one in ten people suffer from chronic pain requiring intensive care (1). In the specific case of the elderly, chronic pain is usually the result of structural and functional changes caused by diseases. The most common sources of chronic pain are the muscles and bones, although other organs and soft tissues can also be the cause of pain (2).
Recently there have been an increasing number of studies on pain in this population. Several studies have attempted to understand the effects of aging on the experience of pain. However, despite the growing number of investigations on the subject, much remains to be understood about the effects of aging on pain experience, the methods of evaluation and even the therapeutic measures to be applied in these patients (3). In addition, a significant proportion of elderly people exhibit changes in the cognitive, perceptual and sensory-motor domains, interfering with the ability to report and to measure the pain experience. The high incidence of dementia disorders is an additional factor that exacerbates the difficulties of pain assessment in this age group. The assessment of pain in these patients is currently a rather controversial issue where a conflict of opinions can be seen related to the type of instruments to be used. Several researchers assume that the cognitive difficulties in these patients are barriers to obtaining direct indicators of pain and therefore they support the use of measures of behavioural observation and recording of pain (4). Others have a different point of view, supporting the idea that the use of "self-report” measures is the ideal method for pain assessment, and opining that a large number of elderly people with dementia are able to report their own pain in a trustworthy manner (5, 6). According to Scherder and colleagues (7, 8, 9), it is possible to assess pain in patients with Alzheimer's disease and vascular dementia as long as they show understanding of the meaning of pain scales, including visual analogue scales and faces scales. However, most of these studies do not have groups with a precise dementia diagnosis, a fact that does not allow either the development of specific assessment methods for the different types of dementia or an understanding of the influence of different pathophysiological processes on the experience of pain (10).
A known fact is that Alzheimer's patients use fewer analgesics (11). However, there is no consensus on the grounds for this observation. On one side there are researchers who state that this is due only to the inability of these patients to communicate their pain (12), while on the other side some observations point to the independence of the use of analgesics and the stage of Alzheimer's disease (13), suggesting that the inability to report the pain is not a valid explanation by itself. In addition, patients with vascular dementia use about four times more painkillers than Alzheimer's patients (11), which is negatively correlated with the results on the Mini Mental State Examination. Taken together, these observations suggest a differential influence of the dementia diagnosis on the pain experience. It is known that the cerebral representations of pain include a vast network that includes cortical and subcortical, sensorial, limbic, associative and motor areas (14). Besides the obvious implications of primary and secondary sensorial areas, several other areas, such as the thalamic nucleus (15), the striatum (16), the hypothalamus, the basal nucleus of Meynert and the amygdala (17), and periaqueductal grey matter (14), are fundamental to the development of the pain experience. Therefore, it can be hypothesized that diverse types of dementia, with different pathophysiological processes, can in fact lead to specific changes in the experience of pain.
In this context, this work aims to test the possibility of the use of direct measures of pain in patients with vascular dementia and Alzheimer's disease and, at the same time, to characterize the experience of pain in these individuals in comparison to cognitively unharmed elderly people.
The participants in the study were recruited in the Medicine E admission service from the Pedro Hispano Hospital.
Our sample consisted of 57 participants grouped according to the following conditions: (1) the Alzheimer's Disease (AD) group, consisting of 15 elderly patients with neurological diagnosis of probable Alzheimer's disease according to the NINCDS-ADRDA criteria; (2) the Vascular Dementia (VD) group, made up of 20 elderly patients with a neurological diagnosis of vascular dementia; (3) the Control group made up of 22 cognitively unharmed elderly patients.
All of the subjects in the groups present a chronic pain condition and do not differ in regard to age, gender, education admission motive or chronic pain condition. In the AD group, many individuals (n = 8) do not take any kind of analgesic medication, which distinguishes this group from the others (tab. 1). The two experimental groups did not show differences regarding the results obtained on the Mini Mental State Examination (t = 1.046; sig. = 0.303).
Table 1. Characteristics of the three groups.
|Gender (M/F)||8; 14||5;15||3; 12|
|Age (M/SD)||71.27; |
|Mini Mental (M/SD)||26.45; |
|Elementary school incomplete||3||1||1|
|Elementary school complete||17||18||12|
|Diabetes mellitus ||0||2||3|
|Used Analgesics |
|Paracetamol + Tramadol||1||0||1|
|No medication ||0||0||8|
Patients with a history of psychological and psychiatric disturbances or other focal or diffuse neuropathology other than Alzheimer's or vascular dementia, disturbances of consciousness, alcoholism, illiteracy and problems of vision were excluded from the sample.
The pain assessment had the objective of evaluating different components of pain in a short period of time. The following instruments were used: the Coloured Analogue Scale (CAS), Facial Affective Scale (FAS), Faces Pain Scale (FPS), and Pain Assessment in Advanced Dementia (PAINAD).
The CAS was used because it is a sensitive, simple, reproducible and universal method to quantify the intensity of pain (18). We apply the CAS with the size of 10 cm in vertical orientation, with a colour spectrum from yellow to red, in which a greater intensity of pain corresponds to a greater intensity of the red colour on the scale. The results range between zero (no pain) and ten (maximum pain). This version has been frequently used in children, showing a positive correlation with behavioural measures and the evaluation made by parents and nurses (19). Some authors also suggest the use of this scale in patients with levels of mild and moderate Alzheimer's disease (7).
The FAS was used to rate the affective component of pain. This scale consists of nine faces ranging from a happy expression (no pain) to a very painful face (maximum pain). The score of each face is on the back of the scale, and it varies from 0.04 (very happy) to 0.97 (very painful).
The severity of pain was assessed by the use of the FPS. In this scale, the first face is neutral, meaning no pain, and is followed by six faces that express increasing extent of pain. The score ranges between 0 (no pain) and 6 (most severe pain).
The PAINAD is a pain behaviour checklist and it is used on individuals with difficulties in verbal communication (20). Its use enabled the identification of the behavioural manifestations of pain, and the establishment of correlations with the previous scales of self-report. This scale includes five items: breathing pattern, negative vocalization, facial expression, body language, and consolability. Each item is scored between zero and two and the total score ranges from 0 (no pain) to 10 (maximum pain) (20).
The pain assessment took place at the Medicine E admission service of the Pedro Hispano Hospital. The hospital Ethic Committee issued favourable reports related to the experimental design and informed consent was given. Participants were assessed with the Mini Mental State Examination in order to ensure the cognitively unharmed character of the control group and that the two experimental groups had similar global cognitive affectation.
After the cognitive screening, the understanding of self-report scales was tested. For the CAS the following questions were addressed to the subjects: "If you felt the worst pain ever, where might you place the pointer?”; "If you did not feel any pain, where would you put it?”. In the FAS and FPS the subjects were asked as follows: "Which face would you choose if you felt the worst pain possible? And if you did not feel any pain....?”. To the question "If you felt the worst pain, where might you place the pointer?”, it was expected that the patients would place the pointer at the upper end of the scale in the case of EVA, or the far right in the case of EFA and EFD. The opposite would be expected to the question: "And if you did not feel any pain....”. Only one person in each group showed no understanding of the scales and therefore they were excluded from the study. In the second phase, the pain assessment was carried out.
The patient's observation according to PAINAD was performed by a nurse not connected with the research.
The statistical analysis was carried out using the computer program SPSS(r) 15.0 for Windows. The differences in pain assessment between the three groups were determined by multivariate analysis of variance (MANOVA), followed by a univariate analysis of variance (ANOVA) and a comparison between the groups by the Scheffe's test. Finally, Pearson correlations between the PAINAD results and the scores obtained on self-report scales were made.
Differences were considered significant at p<0.05.
Table 2 shows the average results obtained by the groups on the pain measures. Wilks' lambda test (lambda = 0.348; F = 3.320; P = 0.000) showed a significant effect of the cognitive status on the test results. The univariate analysis revealed a significant effect of the independent variable on each of the dependent variables (tab. 2). In the CAS, the control and VD groups obtained pain intensity results significantly higher than the DA group (VD, Control vs AD; p=0.000). The same pattern was observed in the FAS, where the control and VD groups achieved significantly higher results in comparison to the AD group (VD, Control vs AD; p=0,000). The results on the EFD were different in the three groups; the VD group presented the highest scores and the AD group the lowest, while the control group was placed between the other two (VD vs Control; p=0.02; Control vs AD; p=0.00).
Table 2. Descriptive statistics and ANOVA of the results obtained by the three groups.
|CAS (M\SD)||7.71; 1.821||8.22; 1.927||4.87; 1.06||18.52||0.000|
|FAS (M\SD)||6.82; 2.039||8.35; 1.663||4.40; 0.986||20.92||0.000|
|EFD (M\SD)||4.59; 1.221||5.65; 0.988||3; 0. 845||24.22||0.000|
|Total PAINAD||6.227; 2.308||7.35; 1.565||3.4667; 1.125||18.348||0.000|
|PAINAD respiration||0.86; 0.64||0.95; 0.224||0.20; 0.414||11.435||0.000|
|PAINAD vocalization||1.55; 0.510||1.65; 0.489||1; 0.378||7.797||0.000|
|PAINAD behaviour||1.18; 0.664||1.60; 0.503||0.47; 0.516||14.902||0.000|
|PAINAD expression||1.23; 0.528||1.45; 0.510||0.87; 0.352||5.897||0.000|
|PAINAD consolability||1.45; 0.510||1.70; 0.470||1; 0.378||10.235||0.000|
Regarding the total score of the PAINAD scale, there is a significant difference between the AD group and the groups of control and VD (VD, Control Vs AD; p=0.00), with the latter two presenting l higher scores. The same pattern was seen on PAINAD breathing pattern (VD, Control vs AD; p=0.00), negative vocalization (VD, Control vs AD; p=0.00), pain behaviour (VD, Control vs AD; p=0.00) and consolability (VD, Control vs AD; p=0.00) subscales. This means that the AD group showed less breathing distress and negative vocalization, fewer pain behaviours, and a higher degree of consolability. On the PAINAD facial expression subscale, the AD group obtained lower results but with no significant difference from the control group (AD vs control; p=0.09), which in turn did not differ from the VD group (control vs VD; p=0.42). The correlation between the results on self-report scales and PAINAD total and partial (sub-scales) results revealed a positive and highly significant correlation between the two sorts of measures (tab. 3).
Table 3. Pearson's correlations between direct and observational pain measures.
|Total score||Sig. (2-tailed)||0.000||0.000||0.000||0.000||0.000||0.000||0.000||0.000|
|PAINAD ||Correl. Pearson||0.833||1.000||0.658||0.634||0.508||0.644||0.692||0.613||0.648|
|PAINAD ||Correl. Pearson||0.832||0.658||1.000||0.673||0.555||0.593||0.684||0.619||0.624|
|PAINAD ||Correl. Pearson||0.866||0.634||0.673||1.000||0.583||0.590||0.644||0.628||0.656|
|PAINAD ||Correl. Pearson||0.770||0.508||0.555||0.583||1.000||0.571||0.559||0.515||0.521|
|Facial Expression||Sig. (2-tailed)||0.000||0.000||0.000||0.000||0.000||0.000||0.000||0.000|
|PAINAD ||Correl. Pearson||0.806||0.644||0.593||0.590||0.571||1.000||0.711||0.662||0.731|
Discussion of results
This study has shown that the use of pain self-report measures in early stages of dementia is perfectly plausible. The finding that the experimental groups did not differ from the control group in regard to the comprehension of the scales, and the high positive correlation between this type of measure and the observation through the PAINAD, support this idea.
The results obtained on the CAS reflect a significantly lower intensity of pain in the AD group. This pain hypointensity in Alzheimer's disease has already been reported (21). However, the same researchers (9) point to a marked increase in pain intensity in subjects with vascular dementia compared with elderly controls. This observation was not confirmed by us. This may be due to the fact that in the research of Scherder and colleagues (9) the experimental group was composed of individuals with "possible” vascular dementia, and the type of dementia was not confirmed due to the lack of compelling additional means of diagnosis. The results obtained on the FAS support the idea of an affective hypointensity in the AD group. Despite the concordance of this observation with the results obtained by Scherder and collaborators (7, 21), our work does not confirm the increase of the pain affective component in the VD group, as reported by those researchers (9). Again, we believe that the inconsistency of the results is due to the fact that in the mentioned research the alleged group of vascular dementia patients did not complete all diagnostic criteria for the dementia category in question, which makes it difficult to distinguish between VD and AD and other dementia disorders (9), and therefore to make direct comparisons between studies. With respect to pain severity, the results obtained on the FPS point to higher levels in VD and, again, lower levels of pain severity in AD.
The behavioural aspects of pain checked through PAINAD follows the tendency of the self-report measures in which the AD group presented a lower number of pain behaviours, and control and VD groups did not differ, except for facial expression. According to that we can conclude that in Alzheimer's disease, chronic pain is experienced with lower intensity, affection, discomfort or severity and behaviour manifestations, when compared to VD patients and controls. This can be interpreted based on the fact that from the earliest stages of the disease there are changes in a multiplicity of structures of the central nervous system that are essential for the motivational-affective and cognitive-evaluative aspects of pain. They include the locus coeruleus, the peri-aqueductal grey matter, the thalamic intralaminar and medial nuclei, the anterior cingulate area, the insula, the amygdala, the hippocampus and histaminergic nuclei of the hypothalamus (8). Regarding vascular dementia we found that, contrary to what was hypothesized and to what Scherder and colleagues (9) observed, this group does not present an augmented pain experience in comparison to controls. Due to the distinct lesions in the white matter and the destruction of the cortical and subcortical circuits (deafferentation), as well as hyperactivity of the hypothalamic-pituitary-adrenal axis, with a concomitant increase of autonomous activity (8), an increased expression of pain in the different areas evaluated would be expected. This appeared only on the facial expression subscale. Possibly, the heterogeneity implied in the vascular pathology may account for these observations. In other words, despite the fact that all of the VD subjects had a well established diagnosis, the location and extent of cerebral lesions in the different patients were not specified; this could assist in a more effective interpretation of results. Despite the positive aspects of this study, there are some limitations that we would like to discuss. The first limitation of this study is the low number of participants; however, this is the result of a rigorous selection of the sample. The second limitation relates to the fact that the assessment took place at a single point in time and it would be more ecological if several assessments were made on different occasions. That would be enriching because the pain assessment could take place at different times, both favourable and adverse for the patients. Furthermore, in patients with AD, the memory factor is amended, so the monitoring of these individuals at more than one time would make the data more accurate. However, previous studies also had only one time of assessment, so our results can be compared. Despite these limitations, this study has enabled us to draw a picture of the experience of pain in elderly Portuguese, with different types of dementia and, at the same time, it allowed us to lay the foundations for the application of pain assessment tools in this population. In the future, it would also be advisable to extend the study of painful experience to other frameworks such as fronto-temporal dementia (8) and Parkinson's disease (22).
In accordance with the above, this study supports the use of pain self-report measures in early stages of dementia, and highlights the specificity of the pain experience according to the dementia diagnosis. The high correlation obtained by the direct and observational pain measures ensures that as long as the patients show understanding of the scales, the answers are reliable. Therefore, the use of pain scales will enable doctors to appraise the proper therapeutic pain methods, taking into consideration the affective and intensity aspects of pain.
1. WHO. People die in unnecessary pain - New WHO study calls for urgent development of palliative care. 2004. Retrieved February 3, 2007, from http://www.euro.who.int/PressRoom/pressnotes/20050621_1. 2. Pain Associates International Network. P.A.I.N. 2 PORTUGAL - Dor Crónica Músculo-esquelética. 2004. Retrieved February 3, 2007, from http://www.grunenthal.pt/cw/pt_PT/html/cw_pt_pt_aboutus.jhtml? CatId=cw_pt_pt_aboutus_b_05b_02. 3. Gagliese L, Katz J, Melzack R: Pain in the elderly. In Melzack & Wall (Eds.) Textbook of Pain. Edinburgh: Churchill Livingstone 1999 (pp. 991-1006). 4. LaChapelle D, Hadjistavropoulos T, Craig K: Pain measurement in persons with intellectual disabilities. The Clinical Journal of Pain 1999; 15(1): 13-23. 5. Pautex S et al.: Pain in Severe Dementia: Self-Assessment or Observational Scales? Journal of the American Geriatrics Society 2006; 54(7): 1040-1045. 6. Tsai P F, Chang J Y: Assessment of pain in elders with dementia. Medsurg Nursing 2004; 13(6): 364-390. 7. Scherder E J, Bouna A: Visual Analogue Scales for pain assessment in Alzheimer's disease. Gerontology: Behavioural Science Section 2000; 46: 47-53. 8. Scherder E J, Sergeant J A, Swaab D F: Pain processing in dementia and its relation to neuropathology. Lancet Neurol 2003; 2(11): 677-686. 9. Scherder EJ, Deijen D, Jan-Berend Y: Pain Assessment in Patients with Possible Vascular Dementia. Psychiatry: Interpersonal and Biological Processes 2003; 66(2): 133-145. 10. Huffman JC, Kunik ME: Assessment and understanding of pain in patients with dementia. The Gerontologist 2000; 40(5): 574-581. 11. Semla TP et al.: Drug use patterns of persons with Alzheimer disease and related disorders living in the community. Journal of the American Geriatrics Society 1993; 45(4): 408-413. 12. Marzinski LR: The tragedy of dementia: Clinically assessing pain in the confused, nonverbal elderly. Journal of Gerontology Nursing 1991; 17: 25-28. 13. Scherder EJ, Bouna A: Is decreased use of analgesics in Alzheimer disease due to a change in the affective component of pain? Alzheimer Disease Association Disorder 1997; 11: 171-174. 14. Bushnell MC, Apkarian AV: Representation of pain in the brain. In Melzack & Wall (Eds.), Textbook of Pain. Edinburgh: Churchill Livingstone 2006; (pp. 58-64). 15. Saab CY, Willis WD: The cerebellum: organization, functions and its role in nociception. Brain Research 2003; 42: 85-95. 16. Chudler: Response properties of neurons in the caudate-putamen and globus pallidus to noxious and non-noxious thermal stimulation in anesthetized rats. Brain Research 1998; 812: 283-288. 17. Becerra L, Breiter HC, Wise R: Reward circuitry activation by noxious thermal stimuli. Neuron 2001; 3: 927-946. 18. Sousa FF, Silva JA: A métrica da dor: problemas teóricos e metodológicos. Revista DOR 2005; 6(1): 469-513. 19. Savedra M et al.: Testing a tool to assess postoperative pediatric and adolescent pain. Advances in Pain Research Therapy 1990; 15: 85-93. 20. Leong IY, Chong MS, Gibson S: The use of self-reported pain measure, a nurse-reported pain measure and the PAINAD in nursing home residents with moderate and severe dementia: a validation study. British Geriatrics Society 2006; 35: 252-256. 21. Scherder E J et al.: Repeated pain assessment in Alzheimer's disease. Dementia and Geriatric Cognitive Disorders 2001; 12: 400-407. 22. Scherder EJ: Pain in Parkinson's disease. Parkinsonism & Related Disorders 2006; 12(1): 2.