Improving pain management in elderly patients with dementia: validation of the Doloshort observational pain assessment scale

SIR—More than half of older adults report pain affecting their quality of life [1]. Self-assessment cannot be implemented in patients with limited communication abilities due to severe dementia [2, 3]. To address this issue, standardised observational pain scales have been designed but they may be relatively lengthy and their validity has not always been verified. A very brief validated tool could greatly enhance pain evaluation in busy clinical practices and could also help shorten more comprehensive geriatric and oncological assessments of such patients. In a prior study, we demonstrated that Doloplus-2 correlated with self-assessment and had adequate internal consistency and test–retest reliability. We constructed a short version of Doloplus-2, Doloshort, which includes the five items that were significantly associated with the visual analogue scale (VAS) score in a multiple regression model [4, 5]. We conducted the present prospective study to examine the validity of Doloshort and confirm its ease of use. Hundred and fifteen consecutively hospitalised French-speaking patients over the age of 65 years followed by the pain consultation (n: 81) or admitted to a specialised dementia unit (n: 34) were included. Exclusion criteria were delirium, acute psychiatric symptoms, end of life care and severe sensory impairment. Mini-Mental Status Examination, and the dementia rating scale (CDR), was used in all cases to rate cognitive status. The CDR assigns cognitive function to five levels defined as no dementia (CDR = 0), questionable dementia (CDR = 0.5), mild dementia (CDR = 1), moderate dementia (CDR = 2) and severe dementia (CDR = 3). The patients underwent a complete neuropsychological evaluation and appropriate laboratory testing including neuroimaging. Seventy-seven (67%) patients met DSMIV criteria for dementia [6]. Age, gender distribution, pain prevalence and pain aetiology were not significantly different in individuals with and without dementia (Table 1). Patient's characteristics †Chi-square test. ‡Fisher exact test. §One-way ANOVA. Patient's characteristics †Chi-square test. ‡Fisher exact test. §One-way ANOVA. Doloshort is an observational pain scale reflecting pain during usual care; it is completed by the nurse in charge of the patient after appropriate discussion with other involved team members (see Appendix 1 in the supplementary data on Age and Ageing online) [4, 5]. Self-assessment was completed with one study investigator in a quiet room. Patients were asked whether they experienced pain at the time of the assessment and to indicate the level of pain they were currently experiencing with the visual analogue scale (VAS). The patients were considered to have understood the VAS if they were able to explain its use and could correctly indicate which position corresponded to no pain at all and which position to the most severe pain. Study investigators and the nursing staff were blinded to each other's assessments. Convergent validity: bivariate correlational analysis using Kendall's tau statistic was used to assess the strength of the association between pain intensities measured by Doloshort and completed VAS scales. Convergent validity was said to be present, if there was at least a strong correlation (Kendall's tau >0.5 or <−0.5) [7]. Internal consistency: the Cronbach alpha was calculated to examine the homogeneity of Doloshort. Discriminant validity was established with two sub-samples of patients. Divergent validity was said to be present if there was less than minor correlation between two measures (Kendall's tau <0.3 or >−0.3) [7]. In 15 patients without dementia, the Doloshort score was compared to measures of anxiety, depression and appetite derived from the Edmonton Symptom Assessment System (ESAS) completed the same day by the patient. The ESAS consists of nine visual analogue scales measuring common symptoms in palliative care [8]. Furthermore, in 20 patients with moderate to severe dementia, the Doloshort score was compared to the Pittsburgh Agitation Scale (PAS) score determined by the team in charge of the patient [9]. The PAS assesses agitation in patients with dementia. Sensitivity to change of Doloshort was evaluated in a sub-sample of 34 patients with moderate to severe chronic pain. The first assessment was completed the day before the introduction of opioids (Day 1). The second was completed 3 days after their introduction (Day 4). After converting all scales to percent scores (no pain = 0% and maximum pain = 100%), the Wilcoxon matched-pairs signed-ranks test was used to evaluate whether both assessments were statistically different. All analyses were performed with the Stata 9.2 statistical package. The study protocol was approved by the local ethics committee, and all study participants or appropriate surrogates gave their informed consent. Scores in text represent mean (± standard deviation) unless stated otherwise. Fifty-nine (77%) patients with dementia demonstrated good comprehension of the VAS. The administration of the Doloshort was possible in all 115 patients and took 5.0 minutes (±0.9) (see Table 2). Pain intensity measured by visual analogue scale (VAS) and Doloshort aOnly patients that demonstrated good comprehension of VAS are included. Pain intensity measured by visual analogue scale (VAS) and Doloshort aOnly patients that demonstrated good comprehension of VAS are included. Convergent validity: Among patients demonstrating good comprehension of the VAS, convergent validity was established between pain intensity on the VAS and the Doloshort score. Kendall's tau-b was 0.523 with an asymptotic standard error (ASE) = 0.052. The strength of the correlation was similar in patients with (0.548 (ASE: 0.067)) and without dementia (0.445 (ASE: 0.112)). Internal consistency was adequate for all items (Cronbach alpha: 0.73) and similar in cognitively intact (0.68) and dementia patients (0.71). Discriminant validity: A score above 3 on the anxiety, depression and appetite sub-scores of the ESAS was present in respectively 7, 7 and 11 of the 15 patients without dementia. The mean scores were, respectively, 2.3 (±2.1), 3.3 (±2.4) and 5.3 (±2.8); Kendall's tau-b (ASE) between these scores and the Doloshort were: 0.031 (0.258), 0.248 (0.219) and 0.207 (0.252). Twelve of the 20 patients with dementia had a PAS greater than 0 (mean 2.7 ± 2.5). Kendall's tau-b between Doloshort and PAS was 0.139 (ASE = 0.155). Sensitivity to change was established in a sub-sample of 34 patients with moderate to severe chronic pain in whom opioid therapy was initiated. The intensity of pain measured by the Doloshort was 6.4 ± 2.6 on Day 1 and 3.3 ± 2.3 on Day 4 (P < 0.001). A score ≥3 on Doloshort had a sensitivity of 81.5% and a specificity of 70.5% for the detection of pain, with an area under the ROC curve of 0.76; this threshold correctly classified 76 of 100 patients. The Doloshort was easy to use, very quick to complete, correlated well with self-assessment and reached desired internal consistency levels for a new scale [10]. A strength of our study was the use of a self-assessment pain scale as a gold standard for patients that could still communicate. Although use of the VAS in dementia is controversial, we previously demonstrated its reliability in our population using standardised simplified instructions designed for use with cognitively impaired patients [11–14]. Although Doloshort has fewer items, its sensitivity and specificity were comparable to the longer Doloplus-2. Doloshort was also able to measure changes in pain intensity and discriminate pain from behavioural symptoms, anxiety and depression. Importantly, Doloshort scores decreased after treatment with opioid analgesics confirming its validity for pain evaluation. However, several limitations of our study should be kept in mind. Doloshort could only be compared to self-assessment in patients who understood the VAS, and it is possible that Doloshort performances may be different in patients who cannot communicate anymore. Unfortunately, in such cases there is no available gold standard [15–17]. Also, Doloshort was completed by nurses in charge of the patient who were not blinded to treatment status; this could have affected pain rating in patients receiving analgesics (placebo use was ruled out for ethical reasons and filming a patient during an entire shift was not possible). Finally, the study was performed in a hospitalised older population with a high prevalence of pain; Doloshort's performance may be different in other settings. Further studies are needed to confirm the generalisability of our findings and to compare Doloshort to other observational pain scales in multiple populations. What is known: older patients with dementia commonly experience pain but often cannot communicate it. A brief, observation-based, validated tool would greatly enhance pain control in this rapidly growing population. What this study adds: the Doloshort is a concise and reliable clinical pain assessment tool that is easy to use in patients with dementia and may be particularly useful in daily care. None. All authors were involved in the manuscript preparation and approved the final submitted version. Supplementary data are available at Age and Ageing online.

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.