Intensive Care Med (1991) 17:$20-$23
IntensiveCare Medicine 9 Springer-Verlag 1991
Sedation in the ICU H.R. Vinik and I. Kissin Department of Anesthesiology, The University of Alabama at Birmingham, 619S 19th Street, Birmingham, Alabama 35233, USA
Abstract. The effect of midazolam on the induction doseresponse curve for alfentanil was studied in non-premedicated ASA physical status I or II patients. The response to the verbal command was used as an end point of anaesthesia. Dose-response curves for midazolam, alfentanil, and their combination were determined with a probit procedure, and compared with algebraic (fractional) analysis of drug interaction. Interaction between midazolam and alfentanil was found to be synergistic (supra-additive). The results suggest that the use of this combination is advantageous not only because it helps to achieve different anaesthetic goals with specific drugs (a benzodiazepine for unconsciousness and an opioid for blockade of the responses to noxious stimulation), but also because its components are complementary for unconsciousness. Key words: Intravenous anaesthetics - Analgesics Hypnotics - Benzodiazepines - Drug interactions Midazolam - Alfentanil
Appropriate sedation and analgesia in the intensive care setting is an essential part of good medical management. Patients are often agitated because of physical and/or mental distress caused by painful invasive procedures such as placement of or presence of intravenous access lines, catheters or drainage tubes. Frequently patients are unable to synchronize their respiration with the ventilator - commonly referred to as "fighting the ventilator". The consequences of uncontrolled anxiety, agitation and resultant stress are many. These include self-injury or indirect injury from disruption of vital lines or tubes, causing unstable haemodynamics, suboptimal gas exchange and therefore producing organ ischaemia. The need for sedation is to protect the patients from the consequence of their agitation in which they may injure themselves directly or indirectly. Sedation also has an important role to play in the prevention and treatment of psychosis which has been reported in the ICU with incidences varying
from 14-72%. The elimination of the triad of anxiety, pain and sleeplessness is the cornerstone of the successful reduction in the incidence of new psychosis [1]. Farina [2] surveyed intensive care institutions and confirmed that 40% received sedative drugs. The various drugs used (Table 1) have advantages and disadvantages. Three obvious problems exist: (1) The ideal drug, which should be soluble, potent, sedative, analgesic, and short-acting to allow for quick recovery withouth side effects, does not exist. Post-operative recovery after major surgery is often associated with increased metabolic activity [3]. These increased metabolic demands and oxygen consumption can be decreased with appropriate midazolam sedation [3]. (2) Inappropriate delivery of drugs is often used, e.g. dosages as required. Therefore, with intermittent and irregular i.v. bolus injections, sedation will be erratic. (3) Inadequate information regarding appropriate drugs, dosage and pharmacodynamic interaction. Answers to these problems, short of delivering the ideal agent, can be at least partially resolved by adopting the following principles: (1) Sedation should be obtained with short-acting drugs with a high therapeutic index. (2) These drugs should be delivered by an infusion regimen designed to achieve a steady state blood level. (3) This "steady state" should be tailored to achieve the pharmacodynamic effect required for the individual patient. (4) Sedation and analgesia can be best attained by using combinations of drugs which are synergistic. Midazolam has superseded the longer acting benzodiazepines in many Intensive Care Units (ICU). Studies [4] by Persson et al. have indicated that even short-acting benzodiazepines like midazolam, if given in large doses, may have a cumulative effect and delay recovery. For example, the blood level required for surgical anaesthesia is approximately 300 ng/ml. Correspondingly lower blood levels, 100-200ng/ml, produce sedation. The exponential curve in Fig. 1 from Persson's article indicates that blood concentrations will not decrease to awake levels (less than 100 ng/ml) before approximately 120 rain. Addition of alfentanil for analgesia will increase the sedation. This is demonstrated in Fig. 2 by the shift to the left,
H.R. Vinik and I. Kissin: Sedation in the ICU
$21
Table 1. See text
cally, the decreased blood levels would likely enable rapid awakening without the need for an antagonist drug.
Drug
Advantages
Disadvantages
Barbiturates
Quick onset, inexpensive
Opiates (morphine, fentanyl, alfentanil)
Analgesic
Etomidate
Quick onset and recovery Cardiovascular stability Neuroleptic Antipsychotic Quick onset, short duration Analgesic
Low therapeutic index Cardiovascular depression tolerance Respiratory depression Tolerance, constipation Venous irritant
Butyropherones Ketamine
Benzodiazepines Diazepam, Lorazepam
Amnesic, CV stability Short half life, amnesic CV stability
Midazolam
Adrenal suppression Dysphoria Hypotension Psychotomimetic Sympathomimetic
Methods Ninety healthy, drug naive women 18- 60 years of age were randomly assigned to receive intravenousiy, predetermined doses of either midazolam, alfentanil or a combination of both in a double-blind study design. The endpoint (unconsciousness) was determined by the abolition of the ability to open (move) eyes on command. Times between injection of drugs and determination of the endpoint were based on the times to peak effect for these agents: 3 min for midazolam [11] and 2 rain for alfentanil [9, 13]. With combined drug administration, both drugs were injected so that synchronization of the peak effects would occur. Dose ranges of 0.08-0.035 mg/kg were used for midazolam and 0.06-0.18rag/ks for alfentanil. In the midazolam-alfentanil group, midazolam was used in a fixed dose of 0.07 mg/kg and the alfentanil dose was varied. The fraction of ED50 midazolam alone was compared with the fraction of EDs0 midazolam in combination with alfentanil and subjected to statistical analysis for additivity.
Delayed recovery
Results
Tolerance and dose dependent recovery
group A (alfentanil) curve compared with the pure midazolam group B. Group A (alfentanil) were more sedated than group B at the same blood concentration of midazolam. Persson concluded that the blood concentrations required for anaesthesia would not enable the patient to recover quickly despite the relatively short halflife of midazoIam. He concluded that only an antagonist (e.g., flumazenil) would ensure quick recovery. This is demonstrated graphically in Fig. 2. The flumazenil curve is clearly shifted to the right of the midazolam curve indicating quicker awakening with the use of an antagonist. Our studies [6, 7] of synergism indicate that hypnotic endpoints can be achieved with much lower doses of midazolam and alfentanil if used in combination. The purpose of our study was to demonstrate synergism with a combination of midazolam and alfentanil. We contend that by using smaller doses acting synergisti-
Figure 3 shows the dose-response curves of alfentanil in patients who were injected intravenously with either alfentanil and saline, or alfentanil and midazolam using failure to respond to command as the endpoint. The vertical axis is the effect, i.e. the percent of patients who lost consciousness. The horizontal axis denotes the dose of alfentanil used in mg/kg. In the combined group, a fixed dose of midazolam 0.07 mg/kg and a variable dose of alfentanil (0.020-0.060 mg/kg) was used. Figure 4 shows the midazolam alone dose-response curve for induction of anaesthesia. The midazolam EDs0 value was 0.22 m s / k s (see Table 2). Figure 3 represents the alfentanil dose-response curve for induction of anaesthesia with and without the addition of midazolam 0.07 mg/kg. The alfentanil EDs0 value when given alone was 0.13 mglkg (95 % confidence limits 0.11 - 0.19). The addition of midazolam 0.07 mg/kg markedly shifted the alfentanil dose-response curve to the left. The alfentanil EDs0 was decreased to 0.028 mg/kg. A comparison of the combined and single EDs0 doses is presented in Table 2.
300 100 E 200 O3
E 7D
X3
~
Arousable
lOO . . . . . . . . . . . .
t
/
"d
....
] A
0
.....
Partial amr~esia. . . . . . . . . .
~--t [
1
25-
8
Fig. 1. See text
i
i
60
120
i
i
180 240 Time (rain)
i
i
300
360
/ A
/
i
7
/:
&
50.
~
/
/
75
o J
i
I
o Group A A Group B [] Ro 15-1788
/
Z/ 6Q
/ []
/
I
120
l
i
1B0
Midazolam (ng/mO Fig. 2. See text
]
i
240
I
1
300
$22
H.R. Vinik and I. Kissin: Sedation in the ICU
Table 2. Alfentanil-midazolam interaction Sum of fractional doses
Ratio b
0.00
i .00
-
0.22
1.00
-
0.54 p
IntensiveCare Medicine 9 Springer-Verlag 1991
Sedation in the ICU H.R. Vinik and I. Kissin Department of Anesthesiology, The University of Alabama at Birmingham, 619S 19th Street, Birmingham, Alabama 35233, USA
Abstract. The effect of midazolam on the induction doseresponse curve for alfentanil was studied in non-premedicated ASA physical status I or II patients. The response to the verbal command was used as an end point of anaesthesia. Dose-response curves for midazolam, alfentanil, and their combination were determined with a probit procedure, and compared with algebraic (fractional) analysis of drug interaction. Interaction between midazolam and alfentanil was found to be synergistic (supra-additive). The results suggest that the use of this combination is advantageous not only because it helps to achieve different anaesthetic goals with specific drugs (a benzodiazepine for unconsciousness and an opioid for blockade of the responses to noxious stimulation), but also because its components are complementary for unconsciousness. Key words: Intravenous anaesthetics - Analgesics Hypnotics - Benzodiazepines - Drug interactions Midazolam - Alfentanil
Appropriate sedation and analgesia in the intensive care setting is an essential part of good medical management. Patients are often agitated because of physical and/or mental distress caused by painful invasive procedures such as placement of or presence of intravenous access lines, catheters or drainage tubes. Frequently patients are unable to synchronize their respiration with the ventilator - commonly referred to as "fighting the ventilator". The consequences of uncontrolled anxiety, agitation and resultant stress are many. These include self-injury or indirect injury from disruption of vital lines or tubes, causing unstable haemodynamics, suboptimal gas exchange and therefore producing organ ischaemia. The need for sedation is to protect the patients from the consequence of their agitation in which they may injure themselves directly or indirectly. Sedation also has an important role to play in the prevention and treatment of psychosis which has been reported in the ICU with incidences varying
from 14-72%. The elimination of the triad of anxiety, pain and sleeplessness is the cornerstone of the successful reduction in the incidence of new psychosis [1]. Farina [2] surveyed intensive care institutions and confirmed that 40% received sedative drugs. The various drugs used (Table 1) have advantages and disadvantages. Three obvious problems exist: (1) The ideal drug, which should be soluble, potent, sedative, analgesic, and short-acting to allow for quick recovery withouth side effects, does not exist. Post-operative recovery after major surgery is often associated with increased metabolic activity [3]. These increased metabolic demands and oxygen consumption can be decreased with appropriate midazolam sedation [3]. (2) Inappropriate delivery of drugs is often used, e.g. dosages as required. Therefore, with intermittent and irregular i.v. bolus injections, sedation will be erratic. (3) Inadequate information regarding appropriate drugs, dosage and pharmacodynamic interaction. Answers to these problems, short of delivering the ideal agent, can be at least partially resolved by adopting the following principles: (1) Sedation should be obtained with short-acting drugs with a high therapeutic index. (2) These drugs should be delivered by an infusion regimen designed to achieve a steady state blood level. (3) This "steady state" should be tailored to achieve the pharmacodynamic effect required for the individual patient. (4) Sedation and analgesia can be best attained by using combinations of drugs which are synergistic. Midazolam has superseded the longer acting benzodiazepines in many Intensive Care Units (ICU). Studies [4] by Persson et al. have indicated that even short-acting benzodiazepines like midazolam, if given in large doses, may have a cumulative effect and delay recovery. For example, the blood level required for surgical anaesthesia is approximately 300 ng/ml. Correspondingly lower blood levels, 100-200ng/ml, produce sedation. The exponential curve in Fig. 1 from Persson's article indicates that blood concentrations will not decrease to awake levels (less than 100 ng/ml) before approximately 120 rain. Addition of alfentanil for analgesia will increase the sedation. This is demonstrated in Fig. 2 by the shift to the left,
H.R. Vinik and I. Kissin: Sedation in the ICU
$21
Table 1. See text
cally, the decreased blood levels would likely enable rapid awakening without the need for an antagonist drug.
Drug
Advantages
Disadvantages
Barbiturates
Quick onset, inexpensive
Opiates (morphine, fentanyl, alfentanil)
Analgesic
Etomidate
Quick onset and recovery Cardiovascular stability Neuroleptic Antipsychotic Quick onset, short duration Analgesic
Low therapeutic index Cardiovascular depression tolerance Respiratory depression Tolerance, constipation Venous irritant
Butyropherones Ketamine
Benzodiazepines Diazepam, Lorazepam
Amnesic, CV stability Short half life, amnesic CV stability
Midazolam
Adrenal suppression Dysphoria Hypotension Psychotomimetic Sympathomimetic
Methods Ninety healthy, drug naive women 18- 60 years of age were randomly assigned to receive intravenousiy, predetermined doses of either midazolam, alfentanil or a combination of both in a double-blind study design. The endpoint (unconsciousness) was determined by the abolition of the ability to open (move) eyes on command. Times between injection of drugs and determination of the endpoint were based on the times to peak effect for these agents: 3 min for midazolam [11] and 2 rain for alfentanil [9, 13]. With combined drug administration, both drugs were injected so that synchronization of the peak effects would occur. Dose ranges of 0.08-0.035 mg/kg were used for midazolam and 0.06-0.18rag/ks for alfentanil. In the midazolam-alfentanil group, midazolam was used in a fixed dose of 0.07 mg/kg and the alfentanil dose was varied. The fraction of ED50 midazolam alone was compared with the fraction of EDs0 midazolam in combination with alfentanil and subjected to statistical analysis for additivity.
Delayed recovery
Results
Tolerance and dose dependent recovery
group A (alfentanil) curve compared with the pure midazolam group B. Group A (alfentanil) were more sedated than group B at the same blood concentration of midazolam. Persson concluded that the blood concentrations required for anaesthesia would not enable the patient to recover quickly despite the relatively short halflife of midazoIam. He concluded that only an antagonist (e.g., flumazenil) would ensure quick recovery. This is demonstrated graphically in Fig. 2. The flumazenil curve is clearly shifted to the right of the midazolam curve indicating quicker awakening with the use of an antagonist. Our studies [6, 7] of synergism indicate that hypnotic endpoints can be achieved with much lower doses of midazolam and alfentanil if used in combination. The purpose of our study was to demonstrate synergism with a combination of midazolam and alfentanil. We contend that by using smaller doses acting synergisti-
Figure 3 shows the dose-response curves of alfentanil in patients who were injected intravenously with either alfentanil and saline, or alfentanil and midazolam using failure to respond to command as the endpoint. The vertical axis is the effect, i.e. the percent of patients who lost consciousness. The horizontal axis denotes the dose of alfentanil used in mg/kg. In the combined group, a fixed dose of midazolam 0.07 mg/kg and a variable dose of alfentanil (0.020-0.060 mg/kg) was used. Figure 4 shows the midazolam alone dose-response curve for induction of anaesthesia. The midazolam EDs0 value was 0.22 m s / k s (see Table 2). Figure 3 represents the alfentanil dose-response curve for induction of anaesthesia with and without the addition of midazolam 0.07 mg/kg. The alfentanil EDs0 value when given alone was 0.13 mglkg (95 % confidence limits 0.11 - 0.19). The addition of midazolam 0.07 mg/kg markedly shifted the alfentanil dose-response curve to the left. The alfentanil EDs0 was decreased to 0.028 mg/kg. A comparison of the combined and single EDs0 doses is presented in Table 2.
300 100 E 200 O3
E 7D
X3
~
Arousable
lOO . . . . . . . . . . . .
t
/
"d
....
] A
0
.....
Partial amr~esia. . . . . . . . . .
~--t [
1
25-
8
Fig. 1. See text
i
i
60
120
i
i
180 240 Time (rain)
i
i
300
360
/ A
/
i
7
/:
&
50.
~
/
/
75
o J
i
I
o Group A A Group B [] Ro 15-1788
/
Z/ 6Q
/ []
/
I
120
l
i
1B0
Midazolam (ng/mO Fig. 2. See text
]
i
240
I
1
300
$22
H.R. Vinik and I. Kissin: Sedation in the ICU
Table 2. Alfentanil-midazolam interaction Sum of fractional doses
Ratio b
0.00
i .00
-
0.22
1.00
-
0.54 p
