bs_bs_banner
Geriatr Gerontol Int 2016; 16: 1028–1035
ORIGINAL ARTICLE: EPIDEMIOLOGY, CLINICAL PRACTICE AND HEALTH
Evaluating attention in delirium: A comparison of bedside tests of attention Dimitrios Adamis,1,2 David Meagher,3 Orla Murray,4 Donagh O’Neill,1 Edmond O’Mahony,1 Owen Mulligan1 and Geraldine McCarthy1 1 Department of Psychiatry, Sligo Mental Health Services, 4Medical Student Sligo Medical Academy, NUI Galway, Sligo, 3Cognitive Impairment Research Group (CIRG), Graduate-Entry Medical School, University of Limerick, Limerick, Ireland; and 2Department of Psychiatry, Research and Academic Institute of Athens, Athens, Greece
Aim: Impaired attention is a core diagnostic feature for delirium. The present study examined the discriminating properties for patients with delirium versus those with dementia and/or no neurocognitive disorder of four objective tests of attention: digit span, vigilance “A” test, serial 7s subtraction and months of the year backwards together with global clinical subjective rating of attention. Methods: This as a prospective study of older patients admitted consecutively in a general hospital. Participants were assessed using the Confusion Assessment Method, Delirium Rating Scale–98 Revised and Montreal Cognitive Assessment scales, and months of the year backwards. Pre-existing dementia was diagnosed according to the Diagnostic and Statistical Manual of Mental Disorders fourth edition criteria. Results: The sample consisted of 200 participants (mean age 81.1 ± 6.5 years; 50% women; pre-existing cognitive impairment in 126 [63%]). A total of 34 (17%) were identified with delirium (Confusion Assessment Method +). The five approaches to assessing attention had statistically significant correlations (P < 0.05). Discriminant analysis showed that clinical subjective rating of attention in conjunction with the months of the year backwards had the best discriminatory ability to identify Confusion Assessment Method-defined delirium, and to discriminate patients with delirium from those with dementia and/or normal cognition. Both of these approaches had high sensitivity, but modest specificity. Conclusion: Objective tests are useful for prediction of non-delirium, but lack specificity for a delirium diagnosis. Global attentional deficits were more indicative of delirium than deficits of specific domains of attention. Geriatr Gerontol Int 2016; 16: 1028–1035. Keywords: attention, attention tests, delirium, older people, psychometrics.
Introduction Delirium is an acute neuropsychiatric syndrome that is both common and commonly missed in older hospitalized patients. It is associated with a range of adverse outcomes that include functional decline, institutionalization, long-term cognitive deficits and elevated mortality.1–3 Impaired attention is a core diagnostic feature in International Classification of Diseases-10, Diagnostic and Statistical Manual of Mental Disorders Accepted for publication 25 June 2015. Correspondence: Dr Dimitrios Adamis MSc, GradStat, MD, Department of Psychiatry, Sligo Mental Health Services, Clarion Rd, Sligo, Ireland. Email:
[email protected]
1028 | doi: 10.1111/ggi.12592
fourth edition and Diagnostic and Statistical Manual of Mental Disorders fifth edition classification systems. However, attention is difficult to define precisely. Recent definitions emphasize the capacity of the brain to support goal-directed behavior when distraction occurs.4 Attention is fundamental to the executive functions of the prefrontal cortex to execute purposeful behavior, such as problem solving, planning and decision making.5 Knudsen outlined four component processes in attention in terms of underlying neural mechanisms: (i) working memory; (ii) competitive selection; (iii) top-down sensitivity control; and (iv) filtering for stimuli that are likely to be behaviorally important (salience filters).6 However, it is not clear which aspects of attention are affected in delirium or whether it involves a global disturbance of attention.7 Tieges et al.
© 2015 The Authors. Geriatrics & Gerontology International published by Wiley Publishing Asia Pty Ltd on behalf of Japanese Geriatrics Society.
This is an open access article under the terms of the Creative Commons Attribution-NonCommercial-NoDerivs License, which permits use and distribution in any medium, provided the original work is properly cited, the use is non-commercial and no modifications or adaptations are made.
Evaluation of attention tests in delirium
highlighted how tests of attention used in delirium research also engage other cognitive domains (e.g. comprehension, working memory).8 O’Keeffe and Gosney compared a variety of approaches to assessment of attention in older medical patients and reported that although objective tests of attention could distinguish delirium from nondelirium, a clinical global assessment of attention was the most distinguishing approach to delirium detection.9 In the present study, we examined the ability of four objective tests of attention (digit span, vigilance “A” test, serial 7s subtraction and months of the year backwards [MOTYB]) together with the global clinical subjective rating of attention in terms of discriminating properties for patients with delirium versus those with dementia and/or no neurocognitive disorder in older medical inpatients.
Material and methods The present study was carried out in a university teaching general hospital. Patients were admitted for acute assessment and treatment to a medical care unit for older people.
Patients All acute medical admissions of patients aged 70 years and older were approached for permission to enter the study within 72 h (3 days) of admission. Exclusion criteria were: (i) longer than 3 days of hospitalization at study commencement; (ii) included in a previous admission; and (iii) severe communication difficulties due to any of; non-English speaking, severe aphasia, intubation or severe sensory problems.
Assessments Demographics Demographic data (sex, age) were collected from medical records (casenote files and the hospital computerized database).
Delirium assessment Delirium presence was assessed with the Confusion Assessment Method (CAM). The four “cardinal” criteria of CAM are: (i) acute onset and fluctuating course; (ii) inattention; (iii) disorganized thinking; and (iv) altered level of consciousness. These contribute to an algorithm for diagnosis of delirium. The presence of (i) and (ii) with either (iii) or (iv) are required for a diagnosis of delirium.10
Diagnosis of pre-existing cognitive impairment (dementia) Cases with pre-existing cognitive impairment were identified using all available sources (medical files, GP files, relatives’ history and neuroimaging) according to Diagnostic and Statistical Manual of Mental Disorders fourth edition clinical criteria.
Measurement of attention Digit span test. This involved a combination of both forwards and backwards digit span, as described in the Montreal Cognitive Assessment (MoCA) scale.11 Patients were asked to repeat a sequence of five numbers (forwards) followed by a sequence of three numbers to repeat backwards. Total score ranged from 0 to 2 according to how many elements of the test were correctly completed; if they performed correctly in both tasks, a maximum score of 2 was given; if they completed only one test, a score of 1 was given; and if they were unable to complete either task, a score of 0 was assigned. Vigilance “A” test. The vigilance “A” test was also derived from the MoCA scale. A list of 29 letters with the letter “A” included on 11 occasions was presented to the patient and they were asked to indicate each time the letter “A” was mentioned. Scores ranged from 0 to 1 depending on whether patients could complete the test without making more than two errors. Error was considered either commissions or omissions of the letter “A”. Serial sevens subtraction test. This task was scored according to the system outlined in the MoCA. Participants were asked to sequentially subtract 7 starting at 100, and were scored 0–3 according to the number of successful trials: no correct subtractions (0); one correct subtraction (1); two or three correct subtractions (2); and four or five correct subtractions (3). MOTYB test. In this test, the participant was asked to recite the months of the year forwards from January to December and then in reverse order starting from December. Scoring was according to that proposed by O’Regan, patients were considered to have passed this test on reaching July without error.12 Hence, inattention was deemed present in those who were unable to correctly recite the MOTYB as far as July (categorical scoring 0 for inattention, 1 for intact attention). Global clinical subjective rating of attention. The clinical rate of overall attention was carried out according to item 10 of the Delirium Rating Scale–98 Revised (DRS-R98). Patients were rated according to a brief (30–60 s) conversation about their hospitalization and treatment. A subjective rating of attention was made ranging from (0)
© 2015 The Authors. Geriatrics & Gerontology International published by Wiley Publishing Asia Pty Ltd on behalf of Japanese Geriatrics Society
|
1029
D Adamis et al.
for those who were fully alert and attentive to (3) for those with severe difficulty focusing and/or sustaining attention such that an interview was severely affected. Of note, in contrast to the other approaches to assessing attention, better performance is associated with lower scores on this rating system. This rating was interdependent from the diagnosis of delirium. Assessments were carried out by two raters with experience and expertise in administration of the scales who received specific training to establish high interrater reliability before the study. Each participant was assessed by the same rater for all the scales (including the CAM) not in any particular order.
Ethics The Local Research Ethics Committee approved the study. The procedures and rationale for the study were explained to all patients, but because many patients had cognitive impairment at entry into the study, it was presumed that many might not be capable of giving informed written consent. Patients with capacity who refused to participate and patients without capacity whose relatives refused assent were excluded from the study. Patients in agreement, but lacking capacity and with no available relatives, were entered in the study until they regained capacity or relatives were contactable. This was in accordance with previous research in delirium and the ethics committee approval.13,14
Statistical analysis The SPSS version 19 (IBM, Armonk, NY, USA) and STATA version 11 (StataCorp, College Station, TX, USA) software were used for data analysis. Descriptive statistics were reported as means with standard deviations for continuous variables, and as proportions and percentages for categorical variables. In order to evaluate attention tests (and their combinations) in terms of distinguishing delirium versus no delirium (as defined by the CAM), a stepwise discriminant analysis was carried out with the binary classification (CAM+/CAM−) as the dependent variable and the five attention measurements as independent variables. Similarly, stepwise discriminant analysis was used to investigate how the attention tests distinguish delirium, dementia without delirium and participants with no neurocognitive disorder. In addition, each test was examined with receiver operating characteristic analysis, and the area under the curve was calculated. The relationship between the five measurements of attention was examined using nonparametric correlation testing (Spearman’s rho), as some variables were categorical and others continuous variables. Finally, for the binary tests of attention, sensitivity and specificity, as well as positive and negative 1030
|
likelihood ratio, positive predictive value, and negative predictive value were calculated.
Results The studied sample consisted of 200 participants (93% of the total population approached, n = 215).
Sample description The mean age of the sample was 81.1 years (SD 6.5 years, median 82 years, interquartile range 10). Of these, 100 (50%) were women. A previous history of cognitive impairment was identified in 126 (63%) participants. Using the CAM, 34 (17%) of the participants were identified with delirium (CAM+).
Attention tasks Descriptive statistics for each of the attention tasks in those with and without delirium, as well as for those with delirium, dementia or no neurocognitive disorder, are shown in Table 1.
Approaches to assessment of attention Correlation between tests of attention The correlations between performance on the tests of attention were examined using Spearman’s rho coefficients (see Table 2). Some correlations are negative (−sign) due to different coding of the measurement (e.g. high score in the clinical rate of attention means more attention problems but higher scores in the serial 7 scores means a better performance in the attention tests). All the measurements correlated highly.
Discriminant analysis Delirium status (CAM+/CAM−). A stepwise discriminant function analysis was carried out using the five tests of attention as predictors of classification as delirium or not delirium. ANOVA testing of equality of means showed that two approaches – the global clinical subjective rating of attention and the MOTYB (Wilks’ Lambda = 0.68, F = 93.21, d.f. 1:1, d.f. 2:198, P < 0.001, and Wilks’ Lambda = 0.66, F = 50.52, d.f. 1:2, d.f. 2:197, P < 0.001, respectively) significantly predicted the delirium/not delirium classification. Combining these variables allowed for correct identification of 85.5% of the CAM+ classified cases of delirium. Using more statistically stringent cross-validation, in which each case was reclassified by the functions derived from all cases other than itself, these variables correctly identified similar percentages of the delirium cases. Box’s test
© 2015 The Authors. Geriatrics & Gerontology International published by Wiley Publishing Asia Pty Ltd on behalf of Japanese Geriatrics Society
Evaluation of attention tests in delirium
Table 1 Descriptive statistics of the five attention measurements in each of the neurocognitive groups (delirium/dementia/none) and for delirium versus no delirium)
Vigilance “A” test
Inattention Attention intact
MOTYB
Inattention Attention intact
Serial sevens*
Digit span *
Global clinical subjective rating of attention *
n % n % n % n %
Mean Median Standard deviation Mean Median Standard deviation Mean Median Standard deviation
Delirium/dementia None Dementia
Delirium
CAM No delirium
Delirium
23 32.4% 48* 67.6% 14 19.7% 57* 80.3% 2.23 3.00 1.20 1.63 2.00 0.66 0.27 0.00 0.58
28* 82.4% 6 17.6% 28* 82.4% 6 17.6% 0.71 0.00 1.03 0.85 1.00 0.82 1.74 2.00 0.67
66 39.8% 100 60.2% 57 34.3% 109 65.7% 1.81 2.00 1.26 1.48 2.00 0.73 0.46 0.00 0.71
28* 82.4% 6 17.6% 28* 82.4% 6 17.6% 0.71 0.00 1.03 0.85 1.00 0.82 1.74 2.00 0.67
43 45.3% 52 54.7% 43 45.3% 52 54.7% 1.51 1.00 1.22 1.36 2.00 0.76 0.60 0.00 0.76
*Statistical significant, P < 0.001 (χ2-test, Kruskal–Wallis test, Mann–Whitney test). CAM, Confusion Assessment Method.
Table 2 Correlation between different approaches to assessing attention Digit span Digit span Vigilance “A” test Serial sevens MOTYB Global clinical subjective rating of attention †
Correlation Coefficient Correlation Coefficient Correlation Coefficient Correlation Coefficient Correlation Coefficient
Vigilance “A” test
Serial sevens subtraction
MOTYB
Global clinical subjective rating of attention
1.000
0.414†
0.556†
0.584†
−0.412†
0.414†
1.000
0.487†
0.528†
−0.372†
0.556†
0.487†
1.000
0.634†
−0.435†
0.584†
0.528†
0.634†
1.000
−0.414†
−0.412†
−0.372†
−0.435†
−0.414†
1.000
Correlation is significant at the 0.001 level (two-tailed); n = 200.
shows that variance–covariance matrices are equal (Box’s M = 3.673, F = 1.197, d.f. 1:3, d.f. 2:46 200.192, P = 0.309) and thus the assumptions for discriminant analysis were held. Delirium, dementia and no neurocognitive disorder. We carried out a stepwise discriminant function analysis with the categorical variable (delirium/dementia/none of them) as the dependent variable and again found that two tests – global clinical subjective rating of attention
and the MOTYB (Wilks’ Lambda = 0.65, F = 53.34, d.f. 1:2, d.f. 2:197, P < 0.001, and Wilks’ Lambda = 0.81, F = 28.0, d.f. 1:4, d.f. 2:392, P < 0.001, respectively) significantly predicted the delirium/dementia/no neurocognitive disorder classification. The assumptions for discriminant analysis were again held for this analysis (Box’s M = 12.93, d.f. 1:1, d.f. 2:106 269.93, P = 0.048). However, in this case, just 58.0% of original grouped cases were correctly classified by both measurements.
© 2015 The Authors. Geriatrics & Gerontology International published by Wiley Publishing Asia Pty Ltd on behalf of Japanese Geriatrics Society
|
1031
D Adamis et al.
Figure 1 Receiver operating characteristic (ROC) curves for accuracy of each individual test of attention in identifying Confusion Assessment Method-defined delirium. , Serial 7s subtraction; , digit , vigilance “A” text; , span; months of the year backwards , clinical rate of (MOTYB); , reference line. attention;
Table 3 Area under the receiver operating characteristic curve for each individual test of attention in terms of identifying Confusion Assessment Method-defined delirium Test result variable(s)
Area
Standard error
Asymptotic significance
Asymptotic 95% confidence interval Lower bound Upper bound
Serial sevens test Digit span Vigilance “A” test MOTYB Global clinical subjective rating of attention
0.738 0.703 0.713 0.740 0.878
0.043 0.051 0.045 0.044 0.028
