Development of the Japanese DOLOPLUS-2: A pain assessment scale for the elderly with Alzheimer s diseasepsyg_

Transcription

1 doi: /j x PSYCHOGERIATRICS 2010; 10: ORIGINAL ARTICLE Development of the Japanese DOLOPLUS-2: A pain assessment scale for the elderly with Alzheimer s diseasepsyg_ Chiaki ANDO and Michiko HISHINUMA Yokufukai Geriatric Hospital and Department of Nursing Science, St Luke s College of Nursing, Tokyo, Japan Correspondence: Ms Chiaki Ando RN, Yokufukai Geriatric Hospital, , Takaido-nishi, Suginami-ku, Tokyo , Japan. bi-ban@ab.auone-net.jp Received 14 November 2009; accepted 21 April Key words: Alzheimer s disease, observational instrument, pain assessment. Abstract Background: Pain assessment in non-communicative patients relies primarily on observation scales. The DOLOPLUS-2 is a behavioral pain assessment scale for the elderly with impaired verbal communication. It rates five somatic items, two psychomotor items, and three psychosocial items as indicators of pain. The aims of the present study were to develop a Japanese version of the DOLOPLUS-2 and to apply it to elderly patients with Alzheimer s disease (AD). Methods: The translated instrument was evaluated with reference to Japanese patients with AD experiencing pain during post-surgery rehabilitation for hip fractures and who were unable to use any self-reported pain measures. After translation and back-translation, 21 registered nurses tested Version 1 of the pain assessment scale on three patients. The ratings of the nurses and researcher were compared and the number of matching scores determined. Semistructured interviews were conducted with the nurses and patient case studies were recorded. The results of the interviews provided the basis for the development of Version 2 of the scale, which was then tested with six patients and 31 nurses using the same procedures as for the testing of Version 1. Results: The intraclass correlation coefficient for inter-rater reliability for the Version 2 administrators was 0.90 (P < 0.001), with a 95% confidence interval of ; the degree of agreement by items ( ) was excellent. Nurses comments for Version 1 revealed that it was difficult to use and that some Japanese expressions and explanations needed improvement. In contrast, the nurses comments on Version 2 indicated that there were no problems. Analysis of patient case studies in Version 2 indicated that pain scores were high only when the patients clearly would have had pain, such as when they started with full weight bearing. On the basis of these results, we developed a final version of the Japanese DOLOPLUS-2. Conclusion: The findings of the present study suggest that Version 2 of the Japanese DOLOPLUS-2 can be used reliably in clinical settings to measure the pain AD patients may be experiencing. However, the final version of the Japanese DOLOPLUS-2 needs to be validated in a larger patient cohort in future studies. INTRODUCTION Pain is older adults is very common and is often complicated by the concomitant presence of different types, multiple locations, and varying causes of pain. 1 In the nursing literature, pain is defined as whatever the experiencing person says it is, existing whatever he says it does. 1 Thus, patients self-reports are considered the gold standard of pain assessment. 131

2 C. Ando and M. Hishinuma However, the ability to self-report in elderly patients with Alzheimer s disease (AD) may be diminished because of a global decline in cognition. A recent positron emission tomography (PET) study revealed that the SI and SII cortices appear to be involved in somatosensory discriminative pain, whereas the anterior cingulate cortex (ACC Area 24) is involved in the affective cognitive aspect of pain. 2 The ACC participates in relating to affectivemotivated pain or affective-cognitive pain stimuli, choosing the type of response to pain stimuli, and learning to recognize and subsequnetly avoid pain stimuli. 2 In patients with AD, the pain thresholds (a sensory discriminative aspect) do not differ from those in elderly people without dementia, whereas pain tolerance (a motivational affective aspect) is increased in patients with AD. 3 That is, areas that belong to the medial pain system (such as the thalamic intralaminar nuclei) and that play an important role in the motivational affective processing of pain are severely affected in AD. 4 The increased pain tolerance in patients with AD may explain the evidence suggesting that cognitively impaired older adults verbally report less pain than the non-impaired elderly. 1 Pain assessment in non-communicative patients relies primarily on observation scales. 5 Zwakhalen et al. reviewed the extant research to identify which behavioral pain assessment tools were available to assess pain in elderly people with dementia and concluded DOLOPLUS-2 is one of the most appropriate scales currently available. 6 DOLOPLUS-2 (from DOLO, an abbreviation of la douleur, meaning pain, and PLUS, meaning that pain is, indeed, felt) is based on the Douleur Enfant Gustave Roussy (DEGR) scale for young children and has been adapted for use in the elderly. 7 Several studies conducted in geriatric centers and palliative care units have validated this scale. 8 Thus, the DOLOPLUS-2 was tested in elderly patients with AD who were unable to use any selfreported pain scales. The aims of the present study were to develop a Japanese version of the DOLOPLUS-2 and to apply it to elderly patients with Alzheimer s disease. METHODS The methodology used to develop the Japanese version of the DOLOPLUS-2 is outlined in Fig. 1 and described in detail below. Step I Step II Translation: Version 1 developed Version 1 implemented (Scored on surgical ward patients) 1. Inter-rater reliability (between 21 nurses and the researcher) 2. Three case studies 3. Group interview of five surgical nurses who used Version 1 Version 2 developed Version 2 implemented (Scored on surgical and psychiatric ward patients) 1. Inter-rater reliability (between 21 surgical ward nurses or 10 psychiatric ward nurses and the researcher) 2. Six case studies 3. Group interview of four surgical nurses who used Versions 1 and Individual interview of 10 psychiatric ward nurses who used Version 2 only. The Japanese DOLOPLUS-2 Figure 1 Methodology for the development of the Japanese DOLOPLUS-2. Instrument Deficiencies in understanding, participation, and communication, sensory disorders, coma, aphasia, dementia, and character disorders, among other things, frequently make self-assessment instruments unusable in elderly patients with AD. 8 Therefore, in the elderly, hetero-assessment of pain is frequently required. 8 The DOLOPLUS-2 is a behavioral pain assessment scale for elderly patients with impaired verbal communication. Indicators of pain are assessed in three domains: (i) five somatic items (verbal complaints, facial expressions, protective body postures, protection of sore areas, and disturbed sleep); (ii) two psychomotor items (functional impairment in daily living, namely washing and dressing and general mobility); and (iii) three psychosocial items (behavioral problems, changes in communication, and changes in social life). Each item has four levels of pain intensity ranging from 0 to 3, yielding an overall score between 0 and 30. For example, expression can be recorded as the patient showing his/her usual expression (0 points), expression showing pain when approached (1 point), expression showing pain even without being approached (2 points), or permanent and unusually blank look (3 points). 7 A score 5 (out of 30) indicates the presence of pain. 7 The use of the DOLOPLUS-2 scale is not limited by what the pain intensity or type is, or the setting. 8 Hølen found that for the best results, the rater should know the patient s habits and regular patterns, should have 132

3 Development of the Japanese DOLOPLUS-2 received training, and should have read the instruction manual before scoring any behaviors as pain related. 9 It is preferable that several health caregivers (e.g. physician, nurse, and nursing assistant) who are trained in advance about how to use the scale administer the scale to a patient. 8 At home, the family and other people can contribute using a liaison notebook or a bedside meeting. 8 For the trained health caregiver, scoring pain with the DOLOPLUS-2 approximately takes 2 5 min at most. 8 Translation procedure First, permission was granted by the DOLOPLUS-2 group (from Bernard Wary, DOLOPLUS-group coordinator) to translate the scale from French into Japanese. The DOLOPLUS-2 scale was then translated into Japanese from French by a person with Japanese as her first language and a good knowledge of French. Next, a Japanese psychiatrist who had studied in France performed the back-translation. Finally, Gisèle Pickering, MD, from the DOLOPLUS-2 group, checked the back-translated documents against the original and confirmed that the translation was satisfactory. Although the French DOLOPLUS-2 has a separate lexicon and instrument, the DOLOPLUS-2 group granted permission for us to blend the lexicon with the instrument for the Japanese version and verified the accuracy of this modified Japanese version. This version was named Version 1, with the next iteration named Version 2. Implementation of Versions 1 and 2 We developed Versions 1 and 2 to evaluate the validity of the translated DOLOPLUS-2. Version 1 was scored on patients from a surgical ward, whereas Version 2 was tested on patients from a surgical and psychiatric ward. The subjects in Version 1 were three patients and 21 registered nurses. In Version 2, the subjects were six patients and 31 registered nurses (Table 1). The inclusion criteria for patients were: (i) admission to a general hospital; (ii) a diagnosis of AD; 10 (iii) a Mini- Mental State Examination (MMSE) 11,12 score of 15 (i.e. subjects who were unable to self-report pain); and (iv) scheduled for hip fracture surgery. Before scoring, all nurses were provided with in-depth instructions regarding DOLOPLUS-2 scoring, which included the instruction do not score if the item is inappropriate. At the time patients were admitted to hospital, their behavior was observed and the nurses attempted to Table 1 Patient and nurse data for Versions 1 and 2 Version 1 Version 2 No. patients 3 6 No. men : women 2:1 3:3 Age (years) MMSE score HDS-R score BI No. nurses Years of nursing experience Years of experience caring for the elderly with dementia Unless indicated otherwise, data are presented as the mean 1 SD. MMSE, Mini-Mental State Examination; HDS-R, Hasegawa Dementia Scale- Revised; BI, Barthel Index. learn the patient s habits and usual condition by conferring with family or healthcare workers who were familiar with the patient. This critical step established usual or normal behavior patterns that were to be the instrument s baseline data. Before surgery, a clinical psychologist rated each patient s cognitive function using the MMSE and Hasegawa Dementia Scale- Revised (HDS-R) The Barthel Index 14 was used by physical therapists to rate each patient s basic activities of daily living. All patients fulfilled the diagnostic criteria for probable AD 10 and were diagnosed as having moderate or severe AD by a geriatric psychiatrist. When the patients started rehabilitation, nurses scored the patients pain levels each day and one of the researchers (CA) scored it 5 days/week prior to the removal of stitches because that would be the most likely time when the patients would experience pain. The researcher (CA) joined the nurses in caring for the patients and observed their daytime behavior (Fig. 2). Nurse s experience implementing Versions 1 and 2 Following the completion of the measurement period, semistructured interviews that took approximately 30 min were conducted with the nurses and tape recorded. The selection criteria for nurses to be interviewed were: (i) they had rated the patients using Version 1; (ii) they had had at least 3 years experience as a registered nurse and at least 1 year experience in caring for elderly patients with impaired verbal communication; and (iii) they had volunteered to partake in the study. For Version 1, five nurses in the surgical ward were interviewed. For Version 2, 14 nurses, who 133

4 C. Ando and M. Hishinuma The rater knows the patient's habits and regular behavioral patterns MMSE, HDS-R, brain CT Nurses scoring (every day) Researcher s scoring (Mon Fri) Admission to hospital Surgery POD 3 POD 7 POD 10 Start of rehabilitation Start full (begin scoring) weight bearing had used both Versions 1 and 2 in the surgical ward or used Version 2 only in the psychiatric ward, were interviewed. For each item, the nurses were asked if there were any practical problems, if the wording was confusing or ambiguous, and whether it was difficult to understand the Japanese expression or the content of the item. Data analysis Inter-rater reliability The inter-rater reliability was tested by the ratio of matching scores and intraclass correlation coefficients, 15 determined using SPSS PASW Statistics 17 (SPSS Institute, Chicago, IL, USA). Furthermore, consistency between Versions 1 and 2 was compared with the Chi-squared test. Face validity The tape-recorded interviews were subsequently transcribed. The raw data were grouped into meaningful sentence clusters and counted as one code. The codes were clustered together depending on similar content, developing subcategories. Finally, the subcategories were reorganized into categories. Case study The DOLOPLUS-2 instructions explain that the comparison of scores between patients makes no sense because pain is a subjective and personal sensation and emotion. 8 All patients scores were examined to evaluate whether the scale could measure each patient s pain. Ethics Following the guidelines of the Helsinki Declaration, oral and written informed consent was obtained from BI Removal of stitches (end scoring) Figure 2 Data collection methods. MMSE, Mini-Mental State Examination; HDS-R, HDS-R, Hasegawa Dementia Scale-Revised; BI, Barthel Index. surrogate decision makers of those patients with dementia. All nurses who participated in the study provided oral and written informed consent. The nurses understood they could withdraw from the study for any reason without penalty. The Ethics Board of St Luke s College of Nursing approved the research. RESULTS Implementation of Version 1 Inter-rater reliability The total number of observations made by the nurses and researcher was 199. There were 119 matching scores (59.8%). The intraclass correlation coefficient for inter-rater reliability for the Version 1 administrators was 0.52 (P < 0.001), with a 95% confidence interval of Interview Twenty-five codes made up four subcategories. From that distillation two categories emerged: (i) difficult to use; and (ii) some Japanese expressions need improvement. Case study All nurses provided the first measurement. The nurses and researcher s scores were not always the same on the day that the patients were likely to have pain, such as when they started full weight bearing or rehabilitation. Revisions to create Version 2 Based on interviews with nurses regarding Version 1, some items were changed as follows. 1. Somatic complaints: Pain degree and type of reaction were added to the top section so as to be more visible. 2. Protective body postures adopted at rest: At rest was defined as lying and sitting because the interpretation of the term at rest differed between nurses. 3. Expression: The usual section was added because it was missed in the table. 4. Sleep pattern: The use of antipsychotics was specified so as to be more practical. 5. Instructions for use: The explanation that the rater could check the section many times if it was appropriate was added. 134

5 Development of the Japanese DOLOPLUS-2 Table 2 Version 2: Agreements and intraclass correlation coefficients according to items Hypothesized Domains for Pain Assessment Items % Agreement (n = 333) Intraclass correlation coefficient (95% CI) Somatic 1. Somatic complaints 72.2** 0.73*** ( ) 2. Protective body postures adopted at rest *** ( ) 3. Protection of sore areas *** ( ) 4. Expression 72.2** 0.86*** ( ) 5. Sleep pattern *** ( ) Psychomotor 6. Washing and/or dressing *** ( ) 7. Mobility 73.5* 0.90*** ( ) Psychosocial 8. Communication 74.3** 0.67 ** ( ) 9. Social life *** ( ) 10. Problems of behavior *** ( ) Total 77.5** 0.90*** ( ) *P 0.05, **P 0.01, *** P < Items were changed to more appropriate Japanese expressions in Version 2. CI, confidence interval. Implementation of Version 2 Version 2 was implemented not only on the surgical ward, but also on the psychiatric ward because two patients were transferred from the surgical ward to the psychiatric ward after surgery. Inter-rater reliability The total number of observations made by the nurses and researcher was 333, with 258 matching scores (77.5%; P < 0.01). The intraclass correlation coefficient for inter-rater reliability for the Version 2 administrators was 0.90 (P < 0.001), with a 95% confidence interval of and an excellent degree of agreement by items ( ; Table 2). Interviews Data analyses yielded 48 codes sorted into 11 subcategories, which were then grouped into four categories: (i) difficult to use; (ii) some Japanese expressions need improvement; (iii) no changes needed; and (iv) measure is feasible. Case study All surgical ward nurses using Version 2 assessed the patients more than once. In three cases, both the nurses and the researcher s scores were high and almost the same on the day that the patients were likely to have pain, such as starting full weight bearing or rehabilitation. In Case A, the scores were high (5 8) after surgery. As a result, analgesics were prescribed on the ninth evening after surgery. The following day the score was down to 0 (Fig. 3). Pain scores Start rehebilitation Start full weight bearing Analgesics started on evening of POD Post-operative day Figure 3 Comparison of pain scores in Case A on post-operative Days (POD) 5, 7, 9, and 11 as determined by the researcher ( ) and the nurses ( ). Revision of Version 2 to create the Japanese DOLOPLUS-2 Based on inter-rater reliability results and outcomes of the semistructured interviews, some items were changed to more appropriate Japanese explanations and the Japanese DOLOPLUS-2 was developed. 1. Protective body postures adopted at rest: The inter-rater reliability essentially did not change from Version 1 to Version 2, even though the item was redefined in Version 2. In addition, during the interviews, the nurses said they could understand the item s meaning without a definition of the item at rest. Ultimately, the definition of the words was deleted. 2. Sleep pattern: Agreements improved from 60.0% to 75.0% after the use of antipsychotics was specified so as to be more practical in Version 2. However, it became clear in the interviews that this item was difficult to measure in those cases in 135

6 C. Ando and M. Hishinuma which patients would not take antipsychotics. In addition, the instructions for the original French version explained that the user should document the behavior regardless of whether patients were taking antipsychotics. Ultimately, information about the use of antipsychotics was deleted. Comparisons of Version 1 and Version 2 The inter-rater reliability of Versions 1 and 2 was compared to clarify whether Version 2 was the more reliable instrument. The ratio of matching scores for Version 1 was 59.8% compared with 77.5% for Version 2. Furthermore, the Chi-squared test was more significant for Version 2 than for Version 1(c 2 [1, n = 532] = 18.9; P < 0.01). Results of the Chi-squared test by item for Versions 1 and 2 were as follows: Somatic Complaints, c 2 = 17.8, P < 0.01; Expression, c 2 = 8.2, P < 0.01; Mobility, c 2 = 3.8, P < 0.05; and Communication, c 2 = 6.4, P < DISCUSSION The aims of the present study were to develop a Japanese DOLOPLUS-2 and to apply it to elderly patients with AD who were unable to use any selfreporting pain scales. The intraclass correlation coefficient for the Version 2 administrators was excellent, with the agreement between administrators of scores obtained using Version 2 significantly higher than for Version 1. One of the reasons for the increased number of matching scores in Version 2 could be improvements in the items of Somatic Complains, Expression, and Sleep Patterns. Although the Washing and/or Dressing, Mobility, and Social Life items were not revised in Version 2, the number of matching scores changed compared with Version 1. This may be due to our making comprehensive notes in sections regarding usual or normal behavior patterns that subsequently used as the instrument s baseline data. Conversely, it cannot be denied that an increase in the frequency of nurse assessments may have contributed to the increase in inter-rater reliability. The DOLOPLUS-2 instructions explain that pain is indeed involved if the patient s behavior changes following the administration of analgesics. 8 In Case A, the total score on Version 2 was high from postoperative Days (POD) 5 8. After analgesics had been prescribed from the evening of POD 9, the total score on Version 2 decreased to 0. Moreover, the total pain score in Version 2 was high only when patients placed a weight load on the hip joint, such as when they started rehabilitation or started full weight bearing, even though antipsychotics had been prescribed on the basis of the patients behavioral and psychological symptoms of dementia (BPSD). These results suggest that the pain levels in patients with AD could be differentiated from BPSD or antipsychotics in the present study. Furthermore, this highlights the importance of knowing a patient s habits and regular behavioral patterns before scoring, although Hølen et al. recommended that psychosocial items should be removed from this scale. 9 There are several limitations to the present study. First, the number of patients evaluated using both Versions 1 and 2 was very small. In addition, it may have been better to evaluate four patients rather than three with Version 1 because the scale s scores cover four levels. In future, the final version of the Japanese DOLOPLUS-2 needs to be validated in a larger cohort of patients. Second, this instrument is a structured observational rating scale and human perceptual errors and inadequacies will impact on the quality of information obtained. 16 Proper training and preparation of observers is very important in minimizing biases in using this instrument. 16 Five important points were identified for those intending to use the DOLOPLUS-2: (i) a patient s usual or normal parameters must be established before scoring; (ii) a complete written description of usual or normal behavior must be provided before scoring; (iii) the patient s behavior itself must be scored, rather than trying to interpret the meaning of a patient s behavior; (iv) the pretest must be performed more than once; and (v) proper training must be undertaken prior to scoring. ACKNOWLEDGMENTS The authors sincerely thank all the patients and nurses who participated in the present study. In addition, the authors thank Tomoko Shimanouchi, MD (Department of Psychiatry, Sakuragaoka-Memorial Hospital, Tokyo, Japan), and Gisèle Pickering, MD (Department of Clinical Pharmacology School of Medicine, Clermont-Ferrand, France). REFERENCES 1 Horgas AL, Elliott AF. Pain assessment and management in persons with dementia. Nurs Clin North Am 2004; 39:

7 Development of the Japanese DOLOPLUS-2 2 Kuroda R, Kawabata A. Pain information pathways from the periphery to the cerebral cortex. Yakugaku Zassh 2003; 123: (in Japanese with an English abstract). 3 Scherder E, Oosterman J, Swaab D et al. Recent development in pain in dementia. BMJ 2005; 330: Rub U, Del Tredici K, Del Turco D, Braak H. The intralaminar nuclei assigned to the medial pain system and other components of this system are early and progressively affected by AD-related cytoskeletal pathology. J Chem Neuroanat 2002; 23: Herr K, Decker S. Assessment of pain in older adults with severe cognitive impairment. Ann Long Term Care 2004; 12: Zwakhalen S, Hamers JP, Abu-Saad HH, Berger MP. Pain in elderly people with severe dementia: A systematic review of behavioral pain assessment tools. BMC Geriatr 2006; 6: Lefebvre-Chapiro S, the DOLOPLUS group. The DOLOPLUS-2 scale: Evaluating pain in the elderly. Eur J Palliat Care 2001; 8: The DOLOPLUS Group. The DOLOPLUS-2 Scale. Available from: [Accessed 2 March 2008]. 9 Hølen JC, Saltvedt I, Fayers PM et al. The Norwegian Doloplus-2, a tool for behavioural pain assessment: Translation and pilot-validation in nursing home patients with cognitive impairment. Palliat Med 2005; 19: McKhann G, Drachman D, Folstein M, Katzman R, Price D, Stadlan EM. Clinical diagnosis of Alzheimer s disease: Report of the NINCDS-ADRDA Work Group under the auspices of Department of Health and Human Services Task Force on Alzheimer s Disease. Neurology 1984; 34: Kim KW, Lee DY, Jhoo JH et al. Diagnosis accuracy of minimental status examination and revised Hasegawa dementia scale for AD. Dement Geriatr Cogn Disord 2005; 19: Folstein MF, Folstein SE, McHung PR. Mini-Mental State. A practical method for grading the cognitive state of patients for clinician. J Psychiatr Res 1975; 12: Kato S, Hasegawa K. Hasegawa dementia scale-revised. Jpn J Geriatr Psychiatry 1991; 2: (in Japanese with an English abstract). 14 Mohoney FI, Barthel DW. Functional evaluation: The Barthel Index. Maryland State Med J 1965; 14: Fleiss JL. Statistical Methods for Rates and Proportions. New York: John Wiley & Sons, Polit DF, Beck CT. Nursing Research Principles and Methods, 7th edn. Philadelphia: Lippincott, Williams & Wilkins,