1Department of Psychological Medicine, University of Glasgow Academic Centre, Gartnavel Royal Hospital, 1055 Great Western Road, Glasgow G12 0XH, 2University Department of Anaesthesia, Western Infirmary, Glasgow G11 6NT and 3Department of Community Health Sciences, Epidemiology and Statistics, University of Edinburgh Medical School, Teviot Place, Edinburgh EH8 9AG, UK*Corresponding author
Accepted for publication: September 1, 2000
![]() |
Abstract |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Br J Anaesth 2001; 86: 637
Keywords: surgery, cardiovascular; complications, cognitive impairment; complications, depression
![]() |
Introduction |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
The assessment of intellectual deficits after CABG frequently employs performance tests which lack performance norms and hence cannot establish in absolute terms whether a patient is impaired. Impairment can only be established in relative terms from differences in performance before and after surgery (either statistically different or some arbitrary difference between pre- and post-scores such as 1 SD, or 20% decline, etc.). As the approach neglects whether a patient may have been cognitively impaired prior to the procedure, two difficulties arise. First, as patients with pre-existing impairments may be more vulnerable to the adverse effects of CABG, their inclusion in a sample may then inflate the apparent risk attached to the procedure. Conversely, patients who are already impaired may be performing so poorly that a floor effect occurs whereby there is little scope for their performance to become worse; hence giving the misleading impression that the procedure has little adverse effect.
Such difficulties may be addressed by using tests that have norms to establish absolute impairment and thus detect pre-existing deficits. The present study makes such an assessment by using the Stroop Neuropsychological Screening Test5 which has age-related impairment norms to define absolute impairment. The Stroop assesses attention and the ability to resist distraction, and is sensitive to impairment as a result of mild brain damage and to the subtle changes seen in mild to moderate dementia.6 Performance on the Stroop prior to CABG has already been shown to be predictive of post-operative outcome.7 However, critically, the investigators did not take account of whether the pre-operative performance was indicative of absolute impairment. Other studies which have employed the Stroop to investigate both CABG-related impairment810 and post-anaesthetic recovery11 have confirmed the sensitivity of the task to change from pre- to post-operative state but, again, without accounting for pre-operative impairment.
The potential importance of accounting for the patients pre-existing cognitive state in interpreting post-CABG deficits is not speculative but is confirmed by parallel research concerning the effects of CABG upon emotional state. Whilst the reported significant increases in depression after CABG have commonly been assumed to be a direct consequence of the procedure, research has shown that the presence of pre-existing depression is a significant determinant of the post-CABG state: patients who are not depressed beforehand are at substantially lower risk.1215
The methodology of the present study involves assessment of both absolute cognitive impairment and presence of depression prior to CABG, hence clarifying whether pre-existing impairments predict post-operative outcome, and replicating the existing results concerning depression and CABG outcome.
![]() |
Methods |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Psychometric assessment
Patients were assessed 3 days before, and 6 days and 6 months after CABG. The Stroop Neuropsychological Screening Test5 consists of a sheet presenting a list of 112 colour names printed in incongruous colours (e.g. the word blue printed in green ink). The subject was required to read through the list as quickly as possible, reading out the name of the colour in which the word is printed. The interference effect produced by the incongruity of the word name and the colour of its presentation renders the task very demanding of attention and concentration. The score is the number of words read correctly in 2 min. The task norms allow an individuals score to be categorized as indicative of likely brain damage as a function of their age. Depression was assessed with the self-report Beck Depression Inventory16 using the conventional criterion of a score 9 as indicative of depression. Years of education are known to have a protective function from the adverse effects of CABG, 17 hence pre-morbid intellectual status was assessed prior to CABG with the National Adult Reading Test (NART).18 Personality characteristics of extroversion and neuroticism can predict post-surgical outcome19 and were assessed prior to CABG by the Eysenck Personality Questionnaire.20
Statistical analysis was conducted with Minitab version 12 via Windows 95 on a Viglen Contender PC.
![]() |
Results |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Table 1 shows the number and percentage of patients suffering cognitive impairment at 6 days or 6 months follow-up, or at both follow-ups, as a function of their pre-CABG state. Prior to CABG, 13 (16%) patients had pre-existing impairment. Postoperatively, 11 (85%) of the latter patients were impaired at 6 days, a total comprising six (46%) who were impaired only at 6 days and five (39%) who were impaired at 6 days and remained so at 6 months. In contrast, of the 68 (84%) who were not impaired prior to the procedure, a total of nine (14%) became impaired at 6 days, these being eight (12%) who were impaired only at 6 days and one (2%) who was impaired at both follow-ups. By contrasting the proportion of patients who were impaired with those unimpaired before CABG and at follow-up, the presence of pre-CABG impairment was found to be a significant predictor of the post-procedure state (Fishers exact test, both P<0.001). There was no evidence of late-onset impairment: no patient was impaired at 6 months without having previously been impaired.
|
The prediction of post-CABG impairment on the Stroop task was assessed further by multiple linear regression with the independent continuous variables of age, NART error score, and scores of extroversion and neuroticism. The regression showed the variables of age and NART to be significant predictors, together accounting for 34% of variance in Stroop performance at 6 days (P<0.001) and 24% at 6 months (P<0.01). Thus, patients who were older, and those of lower pre-morbid intellectual capacity, were more likely to suffer post-CABG deficit. The personality factors were not predictive.
Table 1 also shows mean Stroop performance at each time point as in a conventional analysis to describe impairment as a change from baseline. The performance profile shows an immediate post-operative deterioration in performance, which is then followed by a marked recovery at 6 months to a level substantially above that of the pre-operative baseline. The significant main effect of time (ANOVA: F2,240=8.70, P<0.001) is located in significant differences between performance at 6 months and that at 6 days and baseline (Tukey post-hoc comparisons). The outcome of this mean-based analysis suggests an overall recovery of function at 6 months and is in contrast to the preceding analysis, based on assessment of absolute impairment, which shows some 40% of patients to be impaired at the second follow-up.
![]() |
Discussion |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
The results also fully replicate earlier findings1215 showing that post-CABG depression, too, is significantly determined by the patients state prior to the procedure. More generally, the results support previous findings that CABG outcome is predicted by age and pre-morbid intellectual status.16 24 Older patients, and those with lower pre-morbid intellectual function, were more likely to suffer post-CABG impairment.
Caution is required in extrapolating from the present study where only one cognitive assessment has been employed. Nonetheless, the obvious corollary of the results is that, unless allowance is made for pre-existing impairments, there is the risk of erroneously ascribing some adverse outcomes to a procedure and hence over-estimating its potential risks. Furthermore, where a procedure does exacerbate a pre-existing deficit, the full impact may be missed when using purely relative measures. For example, even a slight decline in the performance of an individual with a pre-existing impairment may be of considerable clinical consequence when compared with that of an individual without such impairment.
The latter issue of individual variation is also critical to full understanding of CABG-related impairment. It has been widely observed, both in the case of CABG and recovery from general anaesthesia, that individual variation is often masked when results are expressed in terms of group means and that this may lead to failure to detect both decline and improvement in a subgroup of patients.2 22 2527 The present analysis demonstrates that simply comparing average pre- and post-CABG performance does not accurately describe the adverse impact of CABG upon Stroop performance and the recovery of function. The mean-based analysis gives the impression of overall recovery of function at 6 months whereas the assessment of absolute impairment shows a substantial proportion of the patients still to be impaired.
Whilst the detection of impairment is a principal concern in studying outcome after CABG, it is important also to be aware of potential improvements in function because of the procedure. Eight of the present patients who were impaired prior to CABG improved significantly to unimpaired status 6 months after surgery. Their mean change in Stroop score of 18.5 (SD 9.47) corresponds to a 28% improvement from their baseline (t7=4.55, P=0.003) and contrasts markedly with the 5% improvement due to practice typically seen in normal subjects.5 Whilst speculative given the present small sample, it has been proposed that such improvement reflects both practice and the beneficial effects of improved perfusion upon cognition.23
The present attrition rate of 32% is virtually identical to that of a very recent study of pre-existing depression and CABG14 but it is higher than that of others.12 13 15 Attrition commonly affects studies involving long-term follow-up and may result in fundamental differences between those who continue with a study and those who drop out. Benedict22 has observed that when a group of particularly impaired individuals drop out of a CABG data set there is then the potential for the procedure to appear to exert a less adverse effect. Further analysis of the present data confirms that, pre-operatively, the proportions of those impaired and depressed were considerably higher amongst the drop-outs than those who continued with the study (36 vs 16%, and 44 vs 28%, for impairment and depression, respectively). Whilst the differences are not statistically reliable because of the small numbers involved, they show that attrition has the potential to distort the results of such a study (P=0.21 and P=0.10, for impairment and depression, respectively). A similar effect of attrition is seen in the results of Vingerhoets and colleagues.8 Some 48% of one of their CABG study groups dropped out between baseline and follow-up, the consequence being that the baseline Stroop performance of those who continued with the study (and against which long-term follow-up was evaluated) was some 23% better than that of the original sample.
Extensive studies of CABG-related cognitive deficits have led to detailed Consensus Statements by Murkin and his colleagues regarding the tasks and methodology for such research.4 25 28 29 The present results confirm the specific consensus regarding pre-operative assessment: A number of patients presenting [for CABG] may have pre-existing CNS abnormalities. [The] neuropsychologic state needs to be assessed[...] prior to operation to provide accurate baseline information. (p. 1289).25 The present results suggest that baseline information would be obtained more accurately with tasks that provide evidence of absolute functional status than the commonly used relative measures. They have the further virtue of being predictive of outcome.
The present results do not call into question the general principles outlined in the Consensus Statement for research into CABG impairment.25 There would seem merit, however, in a series of studies to contrast absolute and relative measures of cognition in their ability both to detect pre-existing deficits and to predict subsequent CABG-related impairment. Indeed, retrospective analysis of some published data sets might be capable of addressing this question where those studies have involved tasks with absolute performance norms which were not employed in the original analysis.
![]() |
Acknowledgement |
---|
![]() |
References |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
2 Newman SP. The incidence and nature of neuropsychological morbidity following cardiac surgery. Perfusion 1989; 4: 93100
3 Newman SP. Neuropsychological and psychological changes. In: Smith P, Taylor K, eds., Brain and Cardiac Surgery. London: Edward Arnold, 1990; 4354
4 Blumenthal JA, Mahanna EP, Madden DJ, White WD, Croughwell ND, Newman MF. Methodological issues in the assessment of neuropsychologic function after cardiac surgery. Ann Thorac Surg 1995; 59: 134550
5 Trenerry MR, Crosson B, DeBoe J, Leber WR. Stroop Neuropsychological Screening Test Manual. Odessa, Florida: Psychological Assessment Resources, 1989
6 Lezak MD. Neuropsychological Assessment (3rd edition). Oxford: Oxford University Press, 1995; 3735
7 Sotaniemi KA, Juolasmaa A, Hokkanen ET. Neuropsychologic outcome after open-heart surgery. Arch Neurol 1981; 38: 28[Abstract]
8 Vingerhoets G, Van Nooten G, Jannes C. Effect of asymptomatic carotid artery disease on cognitive outcome after cardiopulmonary bypass. J Int Neuropsych Soc 1996; 2: 2369[Medline]
9 McKhann GM, Goldsborough MA, Borowicz LM et al. Cognitive outcome after coronary artery bypass: a one-year prospective study. Ann Thorac Surg 1997; 63: 5105
10 McDaid CM, Lewis SA, McMurray T, Phillips SA. Heart surgery: what are the neuropsychological consequences? Irish J Psychol 1994; 15: 11025[ISI]
11 Dijkstra JB, Houx PJ, Jolles J. Cognition after major surgery in the elderly: test performance and complaints. Br J Anaesth 1999; 82: 86774
12 Channer KS, OConnor S, Britton S, Walbridge D, Russell Rees J. Psychological factors influence the success of coronary artery surgery. J R Soc Med 1988; 81: 62932[Abstract]
13 Timberlake N, Klinger L, Smith P, Venn G, Treasure T, Harrison M, Newman SP. Incidence and patterns of depression following coronary artery bypass graft surgery. J Psychosom Res 1997; 43: 197207[ISI][Medline]
14 McKhann GM, Borowicz LM, Goldsborough MA, Enger C, Selnes OA. Depression and cognitive decline after coronary artery bypass grafting. Lancet 1997; 349: 128284[ISI][Medline]
15 Arrowsmith JE, Stygall J, Timberlake N, Pugsley WB, Harrison MJG, Newman SP. New depression following coronary artery bypass surgery. Anesth Analg 1999; 88: SCA68
16 Beck AT, Steer RA. Beck Depression Inventory Manual. New York: Psychological Corporation (Harcourt Brace Jovanovich), 1987
17 Newman MF, Croughwell ND, Blumenthal, JA et al. Predictors of cognitive decline after cardiac operation. Ann Thorac Surg 1995; 59: 132630
18 Nelson HE. National Adult Reading Test (NART): Test Manual. Windsor: NFER-Nelson, 1982
19 Thorp JM, Kennedy BW, Millar K, Fitch W. Personality traits as predictors of anxiety prior to Caesarean section under regional anaesthesia. Anaesthesia 1993; 48: 94650
20 Eysenck HJ, Eysenck SB. The Eysenck Personality Inventory. London: Hodder and Stoughton, 1992
21 Toner I, Hamid SK, Peden CJ, Taylor KM, Smith PLC, Newman SP. Magnetic resonance imaging and P300 (event-related auditory evoked potentials) in the assessment of postoperative cerebral injury following coronary artery bypass surgery. Perfusion 1993; 8: 3219[Medline]
22 Benedict RHB. Cognitive function after open-heart surgery: are postoperative neuropsychological deficits caused by cardiopulmonary bypass? Neuropsych Rev 1994; 4: 22355
23 Deshields TL, McDonough EM, Mannen RK, Miller LW. Psychological and cognitive status before and after heart transplantation. Gen Hosp Psychiat 1996; 18: 62S9S
24 Roach GW, Kanchuger M, Mora Mangano C et al. Adverse cerebral outcomes after coronary artery bypass surgery. New Engl J Med 1996; 335: 185763
25 Murkin JM, Newman SP, Stump DA, Blumenthal JA. Statement of consensus on assessment of neurobehavioural outcomes after cardiac surgery. Ann Thorac Surg 1995; 59: 128995
26 Millar K. Effects of anaesthetic and analgesic drugs. In: Smith AP, Jones DM, eds. Handbook of Human Performance Volume 2. London: Academic Press, 1992; 33785
27 Hickey S, Asbury AJ, Millar K. Psychomotor recovery after outpatient anaesthesia: individual impairment may be masked by group analysis. Br J Anaesth 1991; 66: 34552[Abstract]
28 Murkin JM, Stump DA, Blumenthal JA, McKhann G. Defining dysfunction: group means versus incidence analysis a statement of consensus. Ann Thorac Surg 1997; 64: 9045[ISI][Medline]
29 Newman SP. Analysis and interpretation of neuropsychological tests in cardiac surgery. Ann Thorac Surg 1995; 59: 13515