INTRODUCTION

Cognitive functioning is an important predictor of successful outcome after stroke.1 2 Cognitive impairment (eg mental slowness, memory and visuospatial problems) is, however, common in up to 50% of patients, in both the early and chronic phase after stroke.3^–5 Even small cognitive deficits can result in considerable functional problems and may hinder both the rehabilitation process and secondary stroke prevention.6

Given its impact, assessment of cognitive functioning in the first days or weeks after stroke is essential. Although the feasibility and reliability of assessment in this early phase is doubtful, its information is relevant to the early identification of patients at risk of adverse outcome.4 7 Additionally, early neuropsychological assessment can contribute to the rehabilitation strategy, including an appropriate allocation of discharge destination, and information and advice to patients and their caregivers. The current possibilities for neuropsychological assessment in the (sub)-acute phase are, however, controversial. Whereas standard and time-consuming neuropsychological assessment is supposed to be exhausting at this time due to general malaise, fatigue and other related conditions, frequently used screening instruments such as the Mini-Mental State Examination (MMSE)8 and the (R-) CAMCOG9 appeared not to be suitable for the identification of specific and subtle cognitive deficits.10

As suitable screening measures are not yet available, one could concentrate on subjective cognitive changes. The recently introduced Checklist for Cognitive and Emotional consequences following stroke (CLCE-24) assesses cognitive, emotional and behavioural complaints and takes only 10 minutes.11 12

This checklist is developed to identify patients with subjective psychological problems after stroke, without the intention to assess the presence and severity of impairment. The CLCE-24 appeared to be feasible and valid to identify psychological problems in chronic stroke patients and was also suitable for use by professionals other than the trained neuropsychologist.

The purpose of the present study is to evaluate the utility of the CLCE-24 in the early phase after stroke and, in particular, its diagnostic contribution to the identification of patients with cognitive impairment. The focus is on patients discharged home from the hospital, as these patients might be at higher risk of unnoticed cognitive impairment than those referred to other institutions such as a nursing home or rehabilitation centre. In particular, for those patients living at home after discharge from the hospital, the need for early identification of psychological problems is high as it may help to alert the responsible professionals before the patients and/or their families become unable to cope with their situation.

METHODS

Subjects

The sample consisted of all patients with a first-ever stroke discharged home from the stroke unit of the University Hospital Maastricht (UHM) in the early phase after stroke. All these patients were referred for neuropsychological examination by protocol. Patients were eligible if they had suffered an ischaemic stroke, were free of a neurological and psychiatric history, and had completed the CLCE-24. The diagnosis of ischaemic stroke was based on the presence of an acute focal deficit, and associated lesions on CT scan or MRI scan. Additional inclusion criteria were a stay of at least 1 day at the stroke unit, a Modified Rankin Scale score13 <4, command of the Dutch language and independent living prior to stroke. All patients gave written consent to retrieve data from their medical and psychological records. The data were collected retrospectively by means of file analysis.

Data from a historical control group were used as a reference sample for the neuropsychological examination. This group consisted of 40 healthy partners—17 men and 23 women with a mean age of 52 (SD = 9, range 28–70 years)—of stroke patients who were admitted to the Hoensbroeck Rehabilitation Centre between September 2001 and June 2002.14

Procedure

Basic demographic information was collected, including age, gender and level of education. Level of education was scored according to a Dutch classification system consisting of seven categories ranging from 1: primary school not finished, to 7: university degree obtained.15 A vascular neurologist (blinded for the neuropsychological results) judged the CT and/or the MRI scans that were made during admission for lesion characteristics, including left/right hemisphere and supra- and infratentorial.

Within 2 weeks after discharge, patients were interviewed at home by a nurse using the CLCE-24. All patients underwent neuropsychological examination at the department of Psychology of the UHM, irrespective of whether they filled in the CLCE-24. The mean time interval between the date of stroke and the date of neuropsychological examination was 5 (SD = 4, range 1–15) weeks.

Instruments

The CLCE-24 was filled in by interviewing the patient. The CLCE-24 assesses the subjective psychological changes after stroke and consists of 13 cognitive, 9 emotional and 2 non-specified items, which allow the addition of other problems (see online Appendix 1). For each item, the presence, absence and doubt—and, if present, also the severity (hindering daily life, or not)—could be indicated. For each item, scores range from 0 (no complaint) to 3 (complaint hindering daily life). In developing the CLCE-24, Cronbach’s alpha was computed and was 0.81 based on 69 stroke patients and 22 items.12

Neuropsychological examination covered the following main cognitive domains:16 (1) abstract reasoning (Raven’s Coloured Progressive Matrices); (2) memory (Rivermead Behavioural Memory Test); (3) executive functioning (Trail Making Test part B (preceded by A)), the Tower of London Test (Krikorian version)17 and Category (animals) Fluency from the Groninger Intelligence Test18).

Statistical analysis

To evaluate the CLCE-24, both (sub)total scores and separate items were considered. Scores were recoded as “0” and “1” for, respectively, no hindering complaints and complaints hindering daily life. Total scores, ranging from 0 to 22, are based on both the 13 cognitive and 9 emotional items, ranging from, respectively, 0 to 13 and 0 to 9.11 Frequencies of the separate items were conducted to explore the profile of complaints. To evaluate the cognitive performance between patients with at least one cognitive complaint and those without, between-group comparisons were made using the Mann–Whitney U test. Next, the performance on each neuropsychological task was compared with the reference data—except for fluency, which was not included in the assessment of the control group. Based on these data, patients were classified into “normal” (within 1.5 SD from the mean of the control group) and “impaired” (more than 1.5 SD from the mean of the control group). To test whether the cognitive complaints were associated with cognitive impairment, both chi-square analyses (Fisher’s Exact test if the minimum estimated value was less than 5) and Spearman correlation coefficients were used. All results were considered significant at p<0.05.

RESULTS

Patient characteristics

A total of 105 patients—43 men and 62 women with a mean age of 50 (SD = 13, range 29–83 years)—were directly discharged home in the early phase after stroke. All these patients were referred for neuropsychological testing by protocol. Sixty-one patients (58%) of this group—28 men and 33 women with a mean age of 54 (SD = 14, range 31–83 years)—were also assessed with the CLCE-24 and therefore included for data analyses. Most (97%) of these finished primary school and 50 (64%) patients had at least 10 years of education (median = 4, range 1–7). Twenty-six (43%) patients had a stroke in the left hemisphere and 54 (89%) patients suffered from a supratentorial stroke.

CLCE-24 response

From this sample, 55 (90%) patients reported at least one hindering psychological problem, as assessed by the CLCE-24. Considering the cognitive and emotional items, 45 (74%) patients (the cognitive subgroup) reported at least one cognitive problem and 53 (87%) patients (the emotional subgroup) reported one or more emotional problems. The number of individually reported psychological problems varied between 0 and 15, and the percentages of patients having problems per item in the total sample and both subgroups are shown in figure 1. In the total sample, attention problems (item 2: 38%), mental slowness (item 3: 46%) and recent memory problems (item 4: 38%) were the most reported cognitive complaints. Overall, the percentages of patients with complaints of mental slowness (item 3), feelings of anxiety (item 15), irritation (item 19) and fatigue (item 22) were the highest, varying between 46 and 79%.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1 Profile of psychological complaints of the total sample (n = 61) and the cognitive (n = 45) and emotional subgroup (n = 53). CLCE-24, Checklist for Cognitive and Emotional consequences following stroke.

Cognitive complaints versus performance

The mean performances and standard deviations on the neuropsychological tasks for the total sample, and those patients with and without cognitive complaints, are shown in table 1.

The different sample sizes are based on different numbers of patients who completed the tasks. Difficulties in completing the tasks were mainly due to poor understanding or production. The groups with and without cognitive complaints did not differ in performance on the respective neuropsychological tasks (see table 1).

Figure 2 shows the percentages of patients with impairment on the respective cognitive tasks for the total sample, and for those with and without cognitive complaints, as measured with the CLCE-24. When compared with the reference data, the percentages of patients with impairment varied between 16 and 66 in the total sample and the percentages of cognitive impairment appeared to be higher in patients with cognitive complaints than in those without, except for tasks measuring mental speed (see fig 2). However, no significance could be demonstrated based on both chi-square and correlation analyses. The correlations between cognitive impairment and both cognitive and emotional complaints showed weak and non-significant relationships (rho <0.30). On the other hand, a significant relationship was found between cognitive and emotional complaints (rho = 0.43, p<0.01).

• Download figure
• Open in new tab
• Download powerpoint

Figure 2 Cognitive impairment in patients with and without cognitive complaints. RBMT, Rivermead Behavioural Memory Test; Raven, Raven’s Coloured Progressive Matrices; TMT, Trail Making Test parts A and B; TLT, Tower of London Test. See table 1 for the sample sizes of the respective tasks.

DISCUSSION

The present results show that the majority of our sample reported at least one psychological problem hindering daily life in the early weeks after stroke. More specifically, 74% reported at least one cognitive and 87% at least one emotional problem, with mental slowness, feelings of anxiety, irritation and fatigue being the most frequently reported problems. Although there is knowledge about subjective changes after stroke, this is usually in terms of functional recovery or quality of life.19 20 Little is yet known about specific and subjective psychological problems or what cognitive and emotional changes patients themselves experience after stroke. Most studies explored cognitive consequences after stroke by neuropsychological examination,3 4 and although some studies explored other emotional consequences such as anger,21 the focus was often on depression.22^–25

The introduction of the CLCE-24 in a sample of 69 patients living at home, 6 months after stroke,12 showed the same cognitive and emotional changes but the percentage of patients reporting emotional problems (51%) was lower than that found in the present study. The reported problems in two other Dutch studies on subjective emotional and cognitive changes after stroke, at 326 and 9 months27, respectively, are also in line with our results despite differences in type and time of assessment. However, the respective percentages of changes found at 3 months were lower than those found in our study and also lower than those 9 months after stroke. Differences in percentages could be attributed to the degree of functional problems. When functional outcome is relatively good—such as, for example, in our sample—or has improved after rehabilitation, it can be expected that one will be more confronted or aware of changes in cognition and emotion.27 On the other hand, recovery of functions or successful coping can also reduce subjective cognitive or emotional problems,26 which may explain the difference between the present CLCE-24 response and that at 6 months post-stroke.12

In addition to subjective psychological consequences, we assessed cognitive performance and found that cognitive complaints hindering daily life are not necessarily an indication for low or impaired cognitive performance and visa versa. This accords with other studies on cognitive complaints and actual cognitive performance.28 29 It is suggested that cognitive complaints are determined by psychological well-being rather than cognitive performance but, in severe cases, the underlying problem of complaints may be an objective cognitive disorder such as dementia.30 In our sample, the classification of impairment is relatively stringent, based on 1.5 SD of the mean of the control group. Although little is known about the prevalence and nature of cognitive impairment in the early phase, the percentages of impairment are in line with those found in patients after stroke.3–5 31 Based on the present results and, in particular, the significant relationship between cognitive and emotional complaints, an explanation for the high rates of cognitive complaints seems to be in favour of the suggested and (in this case) poor psychological well-being. Moreover, given the symptoms of depression,32 some cognitive complaints (eg attention problems and mental slowness) might be a manifestation of depressive mood rather than cognitive impairment. However, given our cross-sectional design, no causal interferences can be made from the correlations found.

This study has several limitations that should be addressed. First, the patients included were retrospectively selected and living at home after discharge from the stroke unit. As a result, they do not represent a general stroke population. Given the high rates of problems, the risk of underestimation seems, however, low. Moreover, this selection was made with intent, based on the assumption that cognitive and emotional changes might be more easily neglected in patients discharged home than in those referred to a nursing home or rehabilitation centre. One might, however, argue that we included a selection bias, but all patients discharged home were referred for neuropsychological examination by protocol, irrespective of complaints. Only those patients who also completed the CLCE-24, irrespective of the results, were included to be able to answer our primary study objective. Second, the sample size is relatively small, thereby reducing statistical power. However, despite the small sample size, we could demonstrate that the CLCE-24 is not suitable for identifying patients with cognitive impairment. Third, there is no golden standard for cognitive performance—in particular, in cognitive-demanding real-life situations. Validation of the CLCE-2412 showed a significant relationship between complaints and performance, but performance was only based on global screening measures (ie MMSE and the CAMCOG). In the present study, tasks for reasoning and executive functioning were included, which have shown frequent impairment in the early phase after stroke.10 Finally, patients are only able to make relative decisions about their own cognitive functioning instead of absolute judgments.33 In some cases, patients are even not aware of their deficits, which is caused by either stroke lesion (ie anosognosia) or a psychological defensive mechanism to deny the consequences of stroke.34 So, both performance and complaints may not perfectly reflect the actual situation, which might explain the lack of association.

In conclusion, the CLCE-24 gives no more than subjective information about cognitive consequences. As such, it does not contribute to the identification of patients with cognitive impairment. To identify patients with cognitive impairment in the early phase after stroke, neuropsychological assessment is essential, which leaves us with the problems of feasibility, reliability and validity. The present findings do not mean, however, that subjective consequences have no impact on the outcome after stroke. Further research is needed to answer this question. At the moment, there should be awareness that the absence of complaints does not mean that there is no cognitive impairment. On the other hand, cognitive complaints do not always validate a request for cognitive rehabilitation.