- Email: [email protected]
Blood propofol concentration and psychomotor effects on driving skills
British Journal of Anaesthesia 85 (3): 396±400 (2000)
Blood propofol concentration and psychomotor effects on driving skills S. A. Grant1
3 *
, J. Murdoch1, K. Millar2 and G. N. C. Kenny1
1
Glasgow University Department of Anaesthesia, Royal In®rmary, 8±16 Alexandra Parade, Glasgow G31 2ER, UK. 2University Department of Psychological Medicine, Division of Behavioral Sciences, Academic Centre, Gartnaval Royal Hospital, 1055 Great Western Road, Glasgow G12 0XH, UK 3
Present address: Division of Ambulatory Anesthesia, Department of Anesthesiology, Duke University Medical Center, Box 3094, Durham, NC 27710, USA *Corresponding author We studied psychomotor performance in 10 healthy volunteers during recovery after a targetcontrolled infusion of propofol. Choice reaction time, dual task tracking with secondary reaction time and a within-list recognition task were assessed at target blood propofol concentrations of 0.8, 0.4 and 0.2 mg ml±1. Performance was impaired most at the highest blood propofol concentration (choice reaction time increased by a mean of 247 ms and secondary reaction time by a mean of 178 ms). Choice reaction time and dual task tracking with secondary reaction time were the most sensitive and reliable methods of assessment (signi®cant difference from baseline (P<0.05) at a propofol concentration of 0.2 mg ml±1 with choice and secondary reaction time testing). Within-list recognition assessment of memory was not suf®ciently sensitive at very low propofol concentrations. The impairment in choice and secondary reaction time with a blood propofol concentration of 0.2 mg ml±1 was less than that observed with a blood alcohol concentration of 50 mg 100 ml±1 and no greater than that observed with a blood alcohol concentration of 20 mg 100 ml±1 in a previous study involving healthy volunteers. Br J Anaesth 2000; 85: 396±400 Keywords: recovery; anaesthetics i.v., propofol Accepted for publication: March 30, 2000
With the development of new shorter-acting anaesthetic and analgesic drugs, the recovery of psychomotor function and ®tness to operate a motor vehicle after day-case anaesthesia have become important issues. There is ample experimental evidence that function is impaired after anaesthesia. For example, Whitehead and co-workers1 have shown a doserelated impairment by propofol of choice reaction time. Dose-related impairment has also been reported with older anaesthetic agents such as thiopental.2 Although it is possible to show that skills related to those required while driving are impaired, the criteria to be used in establishing whether a patient is ®t to drive have not been established. Our previous study, which examined the relationship between blood alcohol concentration (BAC) and psychomotor effects,3 attempted to provide a standard by suggesting that the maximum permitted BACs for driving in three European countries (20, 50 and 80 mg 100 ml±1) be used as reference points with which recovery from different anaesthetic techniques could be compared.
Propofol was chosen for this study because of its frequent use during day surgery. In this prospective cohort study, psychomotor function and memory were assessed in volunteers during recovery after a target-controlled infusion of propofol using the Diprifusor target-controlled infusion system. Assessments were made using tests of choice reaction time, tracking with secondary reaction time and within-list recognition.
Methods After obtaining approval from the ethics committee and written informed consent, 10 healthy volunteers were studied (nine male and one female; mean age 32 yr, range 23±40 yr; mean body weight 75.5 kg, range 57±86 kg). The subjects were informed that they would be deeply sedated using propofol administered via the Diprifusor and would slowly recover consciousness in a stepwise fashion. They were also told that, during the recovery period, they would be asked to repeat psychomotor tests.
Ó The Board of Management and Trustees of the British Journal of Anaesthesia 1999
Blood propofol and psychomotor impairment
Subjects with a history of cardiac, pulmonary, neurological, hepatic or renal disease, psychiatric disorder or substance misuse were excluded. Subjects were instructed not to eat for 6 h or drink for 4 h before the study nor to drive a car or operate any machinery in the evening after the study. All subjects were driven home and were accompanied by a competent carer that evening. Assessments took place during working hours in a laboratory in the Department of Anaesthesia, and lasted approximately 4 h. An intravenous cannula was inserted in the dorsum of the non-dominant hand. Full continuous monitoring, consisting of ECG, non-invasive arterial pressure and peripheral arterial oxygen saturation (SpO2), was instituted. The volunteers then repeated the psychomotor tests to reduce practice effects and to establish a performance baseline. All subjects received a target-controlled infusion of propofol using the `Diprifusor' target-controlled infusion system which was increased until the subjects were deeply sedated. A minimum score of 6 using the modi®ed Steward coma scale4 (Table 1) was used as the criterion for the endpoint of sedation. The propofol concentration was then reduced to a target of 0.8 mg ml±1, maintained at that concentration for 30 min, then reduced to targets of 0.4 and 0.2 mg ml±1 for 30 min each. Supplemental oxygen was administered via nasal prongs if the oxygen saturation dropped below 96%. All three assessments of psychomotor performance were carried out three times each over the 30 min periods when the target blood propofol concentration was 0.8, 0.4 and 0.2 mg ml±1. This gave a total of 10 assessments for each test, including baseline measurements. At the end of the study period the propofol infusion was discontinued and the patients recovered in the Department until ®t enough to be escorted home. Subjects performed two psychomotor tests and one memory test during the period of the study. Computerized tasks5 to measure choice reaction time, and dual-task tracking and secondary reaction time were chosen for their known sensitivity to the sedative effects of anaesthetics and alcohol6 7 and had been shown to be sensitive in our previous study.3 Within-list recognition memory has been shown sensitive to sedative effects8 and, as memory is a central component of cognitive competence, was therefore included to extend the range of the test battery.
Primary tracking and secondary visual reaction time Subjects operated a computer mouse to control an on-screen icon (a cross) with the task of maintaining it in contact with a `target' circle moving at varying velocity and direction across the screen of a visual display unit (VDU). The primary task score was the time spent `on target' (i.e. the proportion of the total time spent tracking during which the cross was kept in contact with the circle) expressed as the root mean square error (r.m.s.). While performing the tracking test, subjects had also to respond by pressing the keyboard spacebar when visual signals (star-shaped icons) appeared unpredictably in the periphery of the VDU. Secondary task performance was recorded in milliseconds.
Choice reaction time Five numbered circles were displayed on the VDU, corresponding spatially to response keys 1±5 on the adjoining keyboard. During each trial, the subject's dominant hand rested on the keyboard spacebar. After a variable interval, one of the circles randomly changed colour, requiring the subject to lift their hand from the spacebar and press the appropriate response key. Reaction time was expressed in milliseconds and consisted of two components; `decision time' (time taken to lift hand from the spacebar) and `movement time' (time taken to move to the response key).
Within-list recognition Subjects were presented with lists of 23 words, spoken at a rate of one word every 2 s. Seven of the words in each list were each repeated once at some point in the list; the subjects' task was to detect these repetitions. By varying the number of words occurring between the repeated words, the detection task also becomes one of memory retention. Previous research has shown that the probability of detecting a repetition declines as the number of intervening words increases; this effect is particularly evident in sedated subjects, suggesting that encoding and retrieval processes are impaired. The number of words intervening between the repetitions of the `target' words was 0, 1, 2, 4, 8 or 16. The
Table 1 Modi®ed Steward coma scale Consciousness
Airway
Activity
4 3 2 1 0 3 2 1 0 2 1 0
fully awake; eyes open; conversive lightly asleep; eyes open intermittently eyes opened on command or in response to name responds to ear pinching does not respond opens mouth and/or coughs on command no voluntary coughing but clear unsupported airway airway obstructed on neck ¯exion, but clear on extension unsupported airway obstructed without support raises one arm on command no purposeful movement not moving
397
Grant et al.
subjects responded to repeated words by raising their thumb. Two lists were presented in each test period. Data were analysed with balanced analysis of variance using Minitab software version 10. Signi®cant main effects were investigated further by pairwise comparisons of means using t-tests and Bonferroni corrections as appropriate. Within-list recognition data were analysed by non-parametric tests. All-alpha levels were two-tailed; those less than 0.05 were considered to indicate statistical signi®cance.
Results The maximum target blood propofol concentration required to sedate the subjects to the required level of a modi®ed Steward coma score of <6 varied from 1.4 mg ml±1 to 3.0 mg ml±1. At maximum sedation level, all subjects maintained their airway. No subject required supplemental oxygen for a fall of oxygen saturation below 96%.
Psychomotor testing Choice reaction time
Figure 1 illustrates the statistically signi®cant (P<0.01) increase in choice reaction time as blood propofol concentration increased. Comparison of means by t-test showed that reaction times at all propofol concentrations were signi®cantly longer than at baseline, and also signi®cantly longer at 0.8 mg ml±1 than at 0.4 mg ml±1. Dual task performance
Fig 1 Mean (5±95% con®dence interval) increase in choice reaction time, in milliseconds, recorded at blood propofol concentrations of 0.2, 0.4 and 0.8 mg ml±1.
(i) Secondary visual reaction time. Figure 2 shows the statistically signi®cant effect of propofol concentration (P<0.01) re¯ected in the increase in secondary reaction time as blood propofol concentration increased. t-Tests con®rm that reaction time was signi®cantly longer at all blood propofol concentrations than at baseline, and that all pairwise comparisons between concentrations were also signi®cant (Table 2). (ii) Tracking. Figure 3 illustrates the deterioration in tracking performance seen with increasing blood propofol concentration (P<0.01). There was a signi®cant difference between the 0.8 mg ml±1 propofol group and baseline and also between the 0.2 and 0.8 mg ml±1 groups (Table 2).
Within-list recognition performance
Fig 2 Mean (5±95% con®dence interval) increase in secondary reaction time, in milliseconds, recorded at blood propofol concentrations of 0.2, 0.4 and 0.8 mg ml±1.
The greatest deterioration in recognition performance occurred with the highest blood propofol concentration, 0.8 mg ml±1 (Figure 4). The effect was not, however, signi®cant. There was not always a detectable deterioration in performance at the lower blood propofol concentration. With a blood propofol concentration of 0.4 mg ml±1, the recognition performance was better than baseline when the repeated words were 4, 8 or 16 words apart. At a blood propofol concentration of 0.2 mg ml±1, the performance was better than baseline at the 8 and 16 word intervals.
Table 2 Deterioration in performance observed with choice reaction time (CRT), secondary reaction time (SRT) and tracking at the three propofol concentrations tested. *Signi®cantly different from baseline (P<0.05). **Signi®cantly different from baseline and from 0.2 and 0.4 mg ml±1 groups (P<0.05) Propofol concentration (mg ml±1)
Mean (95% CI) deterioration in CRT (ms±1)
SRT (ms±1)
Tracking (%)
0.2 0.4 0.8
46 (15±77)* 97 (33±160)* 247 (162±332)**
29 (4±53) 76 (46±105)* 178 (127±229)**
2.5 (±2 to ±7) ±3.5 (±8.5 to 1.5) ±19 (±27 to ±12)**
398
Blood propofol and psychomotor impairment
Fig 3 Change in tracking performance (expressed as a percentage) with increasing blood propofol concentration.
Fig 4 Within-list recognition (WLR) score (median) expressed as a percentage of correct responses at baseline (r) and with three different blood propofol concentrations, 0.2 (s), 0.4 (m) and 0.8 mg ml±1 (h).
Discussion We have examined the relationship between blood propofol concentration and psychomotor performance in healthy volunteers using a decreasing concentration of blood propofol to simulate the decreasing concentrations that would be seen during recovery from anaesthesia. The ®ndings of this study show that choice reaction time and dual task tracking with secondary reaction time provide sensitive assessments of progressive performance impairment with increasing blood concentrations of propofol. In contrast, within-list recognition did not discriminate signi®cantly between propofol concentrations. The fact that memory performance at low concentrations of propofol was, on occasion, better than at baseline might suggest an
unresolved practice effect. Equally, however, it might also re¯ect better concentration in the lightly sedated state because of reduced distractibility. Common criticisms of psychomotor tests are that they often lack sensitivity and validity and that there are no wellde®ned `gold-standard' tests with normative databases to de®ne criteria for impaired performance. Such criticisms are well founded but, none the less, the present results indicate that, even at low blood propofol concentrations, choice reaction time and a dual task are sensitive to quite subtle impairment. Furthermore, although such tests are `arti®cial', they do re¯ect critical aspects of skills required when driving and are sensitive to sedative effects in the context of real-life environments requiring skilled performance.9 The question of a meaningful standard against which to compare performance impairment has been addressed by Tiplady10 and by our recent empirical work.3 Tiplady has noted that dose-related effects of alcohol upon psychomotor performance are reasonably well established and has described the consequent potential value of alcohol as a comparator to assess the effects of anaesthetic agents. He has also noted the virtue that BACs are set as arbitrary threshold standards for ®tness versus non-®tness to drive. Using the present psychomotor tests, we have previously assessed impairment at BACs re¯ecting three current European drink-driving limits (20, 50 and 80 mg 100 ml±1). The impairment found with propofol 0.2 mg ml±1, the lowest dose used in the present study, was comparable to that measured at 20 mg 100 ml±1 BAC in our previous experiment. The latter BAC is equivalent to the Swedish drink driving limit and marks evidence for impairment at a considerably lower BAC than the limit of 80 mg 100 ml±1 set in the UK. On the basis of such experimental results, one might therefore generalize and conclude that affected individuals may be vulnerable to mishap in real-life situations. Tiplady, however, provides the important caveat that test performance `cannot be validated in a strict sense against the likelihood of road accidents (or of accidents at home, for that matter)' (reference 10, p. 33). Given suf®ciently sensitive testing conditions and appropriate tests, impairment of mental abilities can be demonstrated for four or more post-operative days.11 The question remains open as to whether driving after receiving an anaesthetic should be banned until psychomotor performance on all tests has returned to normal. It could be argued that a small degree of impairment, similar to that seen with a BAC of 20 mg 100 ml±1 (or propofol 0.2 mg ml±1 in volunteers), would be acceptable in drivers. The impairment in psychomotor performance found at this concentration must be compared with that measured in patients receiving, for example, tricyclic antidepressants, antihistamine agents for hay fever or analgesics for back pain. This is obviously an issue that will need full and widespread debate within the profession, but is of growing importance with an expanding and increasingly mobile day-case population.
399
Grant et al.
References 1 Whitehead C, Sanders LD, Oldroyd G et al. The subjective effects of low-dose propofol. A double-blind study to evaluate dimensions of sedation and cosnsciousness with low-dose propofol. Anaesthesia 1994; 49: 490±6 2 Kortilla K, Linnola M, Ertama P, Hakkinen A. Recovery and simulated driving after intravenous anaesthesia with thiopental, methohexital, propanidid or alphadione. Anesthesiology 1975; 43: 291±9 3 Grant S, Millar K, Kenny GNC. Blood alcohol concentration and psychomotor effects. Br J Anaesth 2000; 85: 401±6. 4 Steward DJ. A simpli®ed scoring system for the post-operative recovery room. Can Anaesth Soc J 1975; 22: 111±13 5 Hope AT, Woolman PS, Gray WM, Asbury AJ, Millar K. A system for psychomotor evaluation; design, implementation and practice effects in volunteers. Anaesthesia 1998; 53: 545±50 6 Millar K. Effects of anaesthetic and analgesic drugs. In: Smith AP,
400
7 8
9 10 11
Jones DM, eds. Handbook of Human Performance. Academic Press, 1992; 337±85 Moss E, Hindmarch I, Pan AJ, Edmondson RS. A comparison of recovery after halothane and alfentanil after minor surgery. Br J Anaesth 1987; 59: 970±77 Munglani R, Andrade J, Sapsford D, Baddeley A, Jones JG. A measure of consciousness and memory during iso¯urane administration: the coherent frequency. Br J Anaesth 1993; 71: 633±41 Millar K, Finnigan F, Hammersley RH. Is residual impairment after alcohol an effect of repeated performance? Aviat Space Env Med 1999; 70: 124±30 Tiplady B. Alcohol as a comparator. In: Klepper ID, Sanders LD, Rosen M, eds. Ambulatory Anaesthesia and Sedation: Impairment and Recovery. London: Blackwell, 1991; 26±37 Flatt JR, Birrell PC, Hobbes A. Effects of anesthesia on some aspects of mental functioning of surgical patients. Anesth Crit Care 1984; 12: 315±24
Blood propofol concentration and psychomotor effects on driving skills S. A. Grant1
3 *
, J. Murdoch1, K. Millar2 and G. N. C. Kenny1
1
Glasgow University Department of Anaesthesia, Royal In®rmary, 8±16 Alexandra Parade, Glasgow G31 2ER, UK. 2University Department of Psychological Medicine, Division of Behavioral Sciences, Academic Centre, Gartnaval Royal Hospital, 1055 Great Western Road, Glasgow G12 0XH, UK 3
Present address: Division of Ambulatory Anesthesia, Department of Anesthesiology, Duke University Medical Center, Box 3094, Durham, NC 27710, USA *Corresponding author We studied psychomotor performance in 10 healthy volunteers during recovery after a targetcontrolled infusion of propofol. Choice reaction time, dual task tracking with secondary reaction time and a within-list recognition task were assessed at target blood propofol concentrations of 0.8, 0.4 and 0.2 mg ml±1. Performance was impaired most at the highest blood propofol concentration (choice reaction time increased by a mean of 247 ms and secondary reaction time by a mean of 178 ms). Choice reaction time and dual task tracking with secondary reaction time were the most sensitive and reliable methods of assessment (signi®cant difference from baseline (P<0.05) at a propofol concentration of 0.2 mg ml±1 with choice and secondary reaction time testing). Within-list recognition assessment of memory was not suf®ciently sensitive at very low propofol concentrations. The impairment in choice and secondary reaction time with a blood propofol concentration of 0.2 mg ml±1 was less than that observed with a blood alcohol concentration of 50 mg 100 ml±1 and no greater than that observed with a blood alcohol concentration of 20 mg 100 ml±1 in a previous study involving healthy volunteers. Br J Anaesth 2000; 85: 396±400 Keywords: recovery; anaesthetics i.v., propofol Accepted for publication: March 30, 2000
With the development of new shorter-acting anaesthetic and analgesic drugs, the recovery of psychomotor function and ®tness to operate a motor vehicle after day-case anaesthesia have become important issues. There is ample experimental evidence that function is impaired after anaesthesia. For example, Whitehead and co-workers1 have shown a doserelated impairment by propofol of choice reaction time. Dose-related impairment has also been reported with older anaesthetic agents such as thiopental.2 Although it is possible to show that skills related to those required while driving are impaired, the criteria to be used in establishing whether a patient is ®t to drive have not been established. Our previous study, which examined the relationship between blood alcohol concentration (BAC) and psychomotor effects,3 attempted to provide a standard by suggesting that the maximum permitted BACs for driving in three European countries (20, 50 and 80 mg 100 ml±1) be used as reference points with which recovery from different anaesthetic techniques could be compared.
Propofol was chosen for this study because of its frequent use during day surgery. In this prospective cohort study, psychomotor function and memory were assessed in volunteers during recovery after a target-controlled infusion of propofol using the Diprifusor target-controlled infusion system. Assessments were made using tests of choice reaction time, tracking with secondary reaction time and within-list recognition.
Methods After obtaining approval from the ethics committee and written informed consent, 10 healthy volunteers were studied (nine male and one female; mean age 32 yr, range 23±40 yr; mean body weight 75.5 kg, range 57±86 kg). The subjects were informed that they would be deeply sedated using propofol administered via the Diprifusor and would slowly recover consciousness in a stepwise fashion. They were also told that, during the recovery period, they would be asked to repeat psychomotor tests.
Ó The Board of Management and Trustees of the British Journal of Anaesthesia 1999
Blood propofol and psychomotor impairment
Subjects with a history of cardiac, pulmonary, neurological, hepatic or renal disease, psychiatric disorder or substance misuse were excluded. Subjects were instructed not to eat for 6 h or drink for 4 h before the study nor to drive a car or operate any machinery in the evening after the study. All subjects were driven home and were accompanied by a competent carer that evening. Assessments took place during working hours in a laboratory in the Department of Anaesthesia, and lasted approximately 4 h. An intravenous cannula was inserted in the dorsum of the non-dominant hand. Full continuous monitoring, consisting of ECG, non-invasive arterial pressure and peripheral arterial oxygen saturation (SpO2), was instituted. The volunteers then repeated the psychomotor tests to reduce practice effects and to establish a performance baseline. All subjects received a target-controlled infusion of propofol using the `Diprifusor' target-controlled infusion system which was increased until the subjects were deeply sedated. A minimum score of 6 using the modi®ed Steward coma scale4 (Table 1) was used as the criterion for the endpoint of sedation. The propofol concentration was then reduced to a target of 0.8 mg ml±1, maintained at that concentration for 30 min, then reduced to targets of 0.4 and 0.2 mg ml±1 for 30 min each. Supplemental oxygen was administered via nasal prongs if the oxygen saturation dropped below 96%. All three assessments of psychomotor performance were carried out three times each over the 30 min periods when the target blood propofol concentration was 0.8, 0.4 and 0.2 mg ml±1. This gave a total of 10 assessments for each test, including baseline measurements. At the end of the study period the propofol infusion was discontinued and the patients recovered in the Department until ®t enough to be escorted home. Subjects performed two psychomotor tests and one memory test during the period of the study. Computerized tasks5 to measure choice reaction time, and dual-task tracking and secondary reaction time were chosen for their known sensitivity to the sedative effects of anaesthetics and alcohol6 7 and had been shown to be sensitive in our previous study.3 Within-list recognition memory has been shown sensitive to sedative effects8 and, as memory is a central component of cognitive competence, was therefore included to extend the range of the test battery.
Primary tracking and secondary visual reaction time Subjects operated a computer mouse to control an on-screen icon (a cross) with the task of maintaining it in contact with a `target' circle moving at varying velocity and direction across the screen of a visual display unit (VDU). The primary task score was the time spent `on target' (i.e. the proportion of the total time spent tracking during which the cross was kept in contact with the circle) expressed as the root mean square error (r.m.s.). While performing the tracking test, subjects had also to respond by pressing the keyboard spacebar when visual signals (star-shaped icons) appeared unpredictably in the periphery of the VDU. Secondary task performance was recorded in milliseconds.
Choice reaction time Five numbered circles were displayed on the VDU, corresponding spatially to response keys 1±5 on the adjoining keyboard. During each trial, the subject's dominant hand rested on the keyboard spacebar. After a variable interval, one of the circles randomly changed colour, requiring the subject to lift their hand from the spacebar and press the appropriate response key. Reaction time was expressed in milliseconds and consisted of two components; `decision time' (time taken to lift hand from the spacebar) and `movement time' (time taken to move to the response key).
Within-list recognition Subjects were presented with lists of 23 words, spoken at a rate of one word every 2 s. Seven of the words in each list were each repeated once at some point in the list; the subjects' task was to detect these repetitions. By varying the number of words occurring between the repeated words, the detection task also becomes one of memory retention. Previous research has shown that the probability of detecting a repetition declines as the number of intervening words increases; this effect is particularly evident in sedated subjects, suggesting that encoding and retrieval processes are impaired. The number of words intervening between the repetitions of the `target' words was 0, 1, 2, 4, 8 or 16. The
Table 1 Modi®ed Steward coma scale Consciousness
Airway
Activity
4 3 2 1 0 3 2 1 0 2 1 0
fully awake; eyes open; conversive lightly asleep; eyes open intermittently eyes opened on command or in response to name responds to ear pinching does not respond opens mouth and/or coughs on command no voluntary coughing but clear unsupported airway airway obstructed on neck ¯exion, but clear on extension unsupported airway obstructed without support raises one arm on command no purposeful movement not moving
397
Grant et al.
subjects responded to repeated words by raising their thumb. Two lists were presented in each test period. Data were analysed with balanced analysis of variance using Minitab software version 10. Signi®cant main effects were investigated further by pairwise comparisons of means using t-tests and Bonferroni corrections as appropriate. Within-list recognition data were analysed by non-parametric tests. All-alpha levels were two-tailed; those less than 0.05 were considered to indicate statistical signi®cance.
Results The maximum target blood propofol concentration required to sedate the subjects to the required level of a modi®ed Steward coma score of <6 varied from 1.4 mg ml±1 to 3.0 mg ml±1. At maximum sedation level, all subjects maintained their airway. No subject required supplemental oxygen for a fall of oxygen saturation below 96%.
Psychomotor testing Choice reaction time
Figure 1 illustrates the statistically signi®cant (P<0.01) increase in choice reaction time as blood propofol concentration increased. Comparison of means by t-test showed that reaction times at all propofol concentrations were signi®cantly longer than at baseline, and also signi®cantly longer at 0.8 mg ml±1 than at 0.4 mg ml±1. Dual task performance
Fig 1 Mean (5±95% con®dence interval) increase in choice reaction time, in milliseconds, recorded at blood propofol concentrations of 0.2, 0.4 and 0.8 mg ml±1.
(i) Secondary visual reaction time. Figure 2 shows the statistically signi®cant effect of propofol concentration (P<0.01) re¯ected in the increase in secondary reaction time as blood propofol concentration increased. t-Tests con®rm that reaction time was signi®cantly longer at all blood propofol concentrations than at baseline, and that all pairwise comparisons between concentrations were also signi®cant (Table 2). (ii) Tracking. Figure 3 illustrates the deterioration in tracking performance seen with increasing blood propofol concentration (P<0.01). There was a signi®cant difference between the 0.8 mg ml±1 propofol group and baseline and also between the 0.2 and 0.8 mg ml±1 groups (Table 2).
Within-list recognition performance
Fig 2 Mean (5±95% con®dence interval) increase in secondary reaction time, in milliseconds, recorded at blood propofol concentrations of 0.2, 0.4 and 0.8 mg ml±1.
The greatest deterioration in recognition performance occurred with the highest blood propofol concentration, 0.8 mg ml±1 (Figure 4). The effect was not, however, signi®cant. There was not always a detectable deterioration in performance at the lower blood propofol concentration. With a blood propofol concentration of 0.4 mg ml±1, the recognition performance was better than baseline when the repeated words were 4, 8 or 16 words apart. At a blood propofol concentration of 0.2 mg ml±1, the performance was better than baseline at the 8 and 16 word intervals.
Table 2 Deterioration in performance observed with choice reaction time (CRT), secondary reaction time (SRT) and tracking at the three propofol concentrations tested. *Signi®cantly different from baseline (P<0.05). **Signi®cantly different from baseline and from 0.2 and 0.4 mg ml±1 groups (P<0.05) Propofol concentration (mg ml±1)
Mean (95% CI) deterioration in CRT (ms±1)
SRT (ms±1)
Tracking (%)
0.2 0.4 0.8
46 (15±77)* 97 (33±160)* 247 (162±332)**
29 (4±53) 76 (46±105)* 178 (127±229)**
2.5 (±2 to ±7) ±3.5 (±8.5 to 1.5) ±19 (±27 to ±12)**
398
Blood propofol and psychomotor impairment
Fig 3 Change in tracking performance (expressed as a percentage) with increasing blood propofol concentration.
Fig 4 Within-list recognition (WLR) score (median) expressed as a percentage of correct responses at baseline (r) and with three different blood propofol concentrations, 0.2 (s), 0.4 (m) and 0.8 mg ml±1 (h).
Discussion We have examined the relationship between blood propofol concentration and psychomotor performance in healthy volunteers using a decreasing concentration of blood propofol to simulate the decreasing concentrations that would be seen during recovery from anaesthesia. The ®ndings of this study show that choice reaction time and dual task tracking with secondary reaction time provide sensitive assessments of progressive performance impairment with increasing blood concentrations of propofol. In contrast, within-list recognition did not discriminate signi®cantly between propofol concentrations. The fact that memory performance at low concentrations of propofol was, on occasion, better than at baseline might suggest an
unresolved practice effect. Equally, however, it might also re¯ect better concentration in the lightly sedated state because of reduced distractibility. Common criticisms of psychomotor tests are that they often lack sensitivity and validity and that there are no wellde®ned `gold-standard' tests with normative databases to de®ne criteria for impaired performance. Such criticisms are well founded but, none the less, the present results indicate that, even at low blood propofol concentrations, choice reaction time and a dual task are sensitive to quite subtle impairment. Furthermore, although such tests are `arti®cial', they do re¯ect critical aspects of skills required when driving and are sensitive to sedative effects in the context of real-life environments requiring skilled performance.9 The question of a meaningful standard against which to compare performance impairment has been addressed by Tiplady10 and by our recent empirical work.3 Tiplady has noted that dose-related effects of alcohol upon psychomotor performance are reasonably well established and has described the consequent potential value of alcohol as a comparator to assess the effects of anaesthetic agents. He has also noted the virtue that BACs are set as arbitrary threshold standards for ®tness versus non-®tness to drive. Using the present psychomotor tests, we have previously assessed impairment at BACs re¯ecting three current European drink-driving limits (20, 50 and 80 mg 100 ml±1). The impairment found with propofol 0.2 mg ml±1, the lowest dose used in the present study, was comparable to that measured at 20 mg 100 ml±1 BAC in our previous experiment. The latter BAC is equivalent to the Swedish drink driving limit and marks evidence for impairment at a considerably lower BAC than the limit of 80 mg 100 ml±1 set in the UK. On the basis of such experimental results, one might therefore generalize and conclude that affected individuals may be vulnerable to mishap in real-life situations. Tiplady, however, provides the important caveat that test performance `cannot be validated in a strict sense against the likelihood of road accidents (or of accidents at home, for that matter)' (reference 10, p. 33). Given suf®ciently sensitive testing conditions and appropriate tests, impairment of mental abilities can be demonstrated for four or more post-operative days.11 The question remains open as to whether driving after receiving an anaesthetic should be banned until psychomotor performance on all tests has returned to normal. It could be argued that a small degree of impairment, similar to that seen with a BAC of 20 mg 100 ml±1 (or propofol 0.2 mg ml±1 in volunteers), would be acceptable in drivers. The impairment in psychomotor performance found at this concentration must be compared with that measured in patients receiving, for example, tricyclic antidepressants, antihistamine agents for hay fever or analgesics for back pain. This is obviously an issue that will need full and widespread debate within the profession, but is of growing importance with an expanding and increasingly mobile day-case population.
399
Grant et al.
References 1 Whitehead C, Sanders LD, Oldroyd G et al. The subjective effects of low-dose propofol. A double-blind study to evaluate dimensions of sedation and cosnsciousness with low-dose propofol. Anaesthesia 1994; 49: 490±6 2 Kortilla K, Linnola M, Ertama P, Hakkinen A. Recovery and simulated driving after intravenous anaesthesia with thiopental, methohexital, propanidid or alphadione. Anesthesiology 1975; 43: 291±9 3 Grant S, Millar K, Kenny GNC. Blood alcohol concentration and psychomotor effects. Br J Anaesth 2000; 85: 401±6. 4 Steward DJ. A simpli®ed scoring system for the post-operative recovery room. Can Anaesth Soc J 1975; 22: 111±13 5 Hope AT, Woolman PS, Gray WM, Asbury AJ, Millar K. A system for psychomotor evaluation; design, implementation and practice effects in volunteers. Anaesthesia 1998; 53: 545±50 6 Millar K. Effects of anaesthetic and analgesic drugs. In: Smith AP,
400
7 8
9 10 11
Jones DM, eds. Handbook of Human Performance. Academic Press, 1992; 337±85 Moss E, Hindmarch I, Pan AJ, Edmondson RS. A comparison of recovery after halothane and alfentanil after minor surgery. Br J Anaesth 1987; 59: 970±77 Munglani R, Andrade J, Sapsford D, Baddeley A, Jones JG. A measure of consciousness and memory during iso¯urane administration: the coherent frequency. Br J Anaesth 1993; 71: 633±41 Millar K, Finnigan F, Hammersley RH. Is residual impairment after alcohol an effect of repeated performance? Aviat Space Env Med 1999; 70: 124±30 Tiplady B. Alcohol as a comparator. In: Klepper ID, Sanders LD, Rosen M, eds. Ambulatory Anaesthesia and Sedation: Impairment and Recovery. London: Blackwell, 1991; 26±37 Flatt JR, Birrell PC, Hobbes A. Effects of anesthesia on some aspects of mental functioning of surgical patients. Anesth Crit Care 1984; 12: 315±24