Benzodiazepine

      Child-Pugh A Child-Pugh B Child-Pugh C
Alprazolam Safety additional risks known additional risks known additional risks known
Dosing start with half of the normal dose start with half of the normal dose no dosing advice possible
Bromazepam Safety additional risks known additional risks known additional risks known
Dosing no dosing advice possible no dosing advice possible no dosing advice possible
Brotizolam Safety additional risks known additional risks known additional risks known
Dosing start at the lower end of the normal dosing range start with half of the normal dose start with half of the normal dose
Chlordiazepoxide Safety additional risks known additional risks known additional risks known
Dosing start with 1/3 of the normal dose start with 1/3 of the normal dose no dosing advice possible
Clobazam Safety additional risks known additional risks known additional risks known
Dosing start at the lower end of the normal dosing range start at the lower end of the normal dosing range start at the lower end of the normal dosing range
Clorazepaat Safety additional risks known additional risks known additional risks known
Dosing no dosing advice possible no dosing advice possible no dosing advice possible
Diazepam Safety additional risks known additional risks known additional risks known
Dosing start with half of the normal dose start with half of the normal dose start with half of the normal dose
Flunitrazepam Safety additional risks known additional risks known additional risks known
Dosing start at the lower end of the normal dosing range start at the lower end of the normal dosing range start at the lower end of the normal dosing range
Flurazepam Safety additional risks known additional risks known additional risks known
Dosing no dosing advice possible no dosing advice possible no dosing advice possible
Loprazolam Safety additional risks known additional risks known additional risks known
Dosing no dosing advice possible no dosing advice possible no dosing advice possible
Lorazepam Safety additional risks known additional risks known additional risks known
Dosing start at the lower end of the normal dosing range start at the lower end of the normal dosing range start with half of the normal dose
Lormetazepam Safety additional risks known additional risks known additional risks known
Dosing start with half of the normal dose start with half of the normal dose no dosing advice possible
Midazolam Safety additional risks known additional risks known additional risks known
Dosing start at the lower end of the normal dosing range start with half of the normal dose start with half of the normal dose
Nitrazepam Safety additional risks known additional risks known additional risks known
Dosing start at the lower end of the normal dosing range start with half of the normal dose no dosing advice possible
Oxazepam Safety additional risks known additional risks known additional risks known
Dosing start at the lower end of the normal dosing range start at the lower end of the normal dosing range start with half of the normal dose
Prazepam Safety additional risks known additional risks known additional risks known
Dosing no dosing advice possible no dosing advice possible no dosing advice possible
Temazepam Safety additional risks known additional risks known additional risks known
Dosing start at the lower end of the normal dosing range start at the lower end of the normal dosing range start at the lower end of the normal dosing range
Zolpidem Safety unsafe unsafe unsafe
Dosing no dosing advice (unsafe) no dosing advice (unsafe) no dosing advice (unsafe)
Zopiclone Safety additional risks known additional risks known additional risks known
Dosing start at the lower end of the normal dosing range start with half of the normal dose no dosing advice possible
Explanation

All benzodiazepines are metabolized in the liver and clearance of most is reduced in cirrhosis. Benzodiazepines increase the risk of hepatic encephalopathy in patients with cirrhosis, particularly if patients have suffered a previous episode of hepatic encephalopathy. It is therefore recommended to avoid the use of benzodiazepines in these patients, if possible. If not possible, dose adjustments are often required and it is recommended to start low with dosing and uptitrate in small steps. Dose adjustments are mostly important when benzodiazepines are used for a longer period of time; in acute situations, e.g. epileptic insult, this is less relevant. Preferred benzodiazepines are those with a short half-life because of the risk of accumulation.

Information about the safety classification and the recommended actions can be found here.

 

Summary of literature

Considerations

Various studies were retrieved where benzodiazepines (BZDs) caused impairment of psychomotoric function and induced minimal or sub-clinical hepatic encephalopathy. This was also noted at peak plasma levels similar to healthy controls. The product information states that all BZDs are contra-indicated in severe hepatic impairment. Especially in patients with a previous episode of hepatic encephalopathy, BZDs should be avoided. For other cirrhotic patients it is also better to use another drug if possible. All benzodiazepines are classified as ‘additional risks known’ in cirrhosis, with the exception of zolpidem (unsafe, see following explanation).  

There are several scenarios imaginable in which there is no alternative drug possible. In these cases, start in the lowest possible dose for the shortest duration. The most favorable BZDs are those with a short half-life and a relatively simple metabolism (only glucuronidation), for instance oxazepam and temazepam. 

Alprazolam: in one single-dose study (evidence level 3) with 17 cirrhotic patients (CTP-class unknown) the exposure to alprazolam was more than doubled in patients compared to the healthy controls. Half-life increased from 11 hours in controls to 20 hours in patients. There were no studies about the safety of alprazolam. The advice from the product information is adopted to use a lower dose: in CTP A and B cirrhosis half of the normal starting dose is recommended. In patients with CTP C it is expected that exposure increases even more, however there is no data so no dosing advice is possible.

Bromazepam: no studies have been performed with bromazepam in patients with cirrhosis. Therefore, it is not possible to give dosing advice.  

Brotizolam: in one single-dose study (level 3) among 8 patients with cirrhosis (CTP-class unknown), the median half-life (13 h) was almost doubled compared to healthy controls (7 h). Exposure was increased by 40%. Two patients were excluded of data analysis because of greatly impaired pharmacokinetics. The authors describe that all patients fell asleep during the study, while none of the controls did, despite comparable peak plasma levels. In CTP A dose adjustment is not yet recommended, while in CTP B and C cirrhosis it is advised to start with half of the normal dose.

Chlordiazepoxide: in four pharmacokinetic studies (level 3-4) 36 patients with cirrhosis (CTP-class unknown) received chlordiazepoxide. Clearance was reduced by 50-66% and the half-life increased from 10-20 hours to 40-60 hours. In three studies (level 3-4) the safety of chlordiazepoxide was assessed in 17 cirrhotic patients. Fourteen suffered from adverse events (severely drowsy, hepatic encephalopathy). For patients with CTP A and B cirrhosis it is recommended to start with one third of the normal dose, for patients with CTP C no dosing advice could be given.  

Clobazam: in one single-dose study (level 3) among nine cirrhotic patients (5 CTP A/B, 4 CTP C), exposure and clearance were similar to healthy controls, but half-life was prolonged from 22 h to 51 h in patients. This is probably caused by an increased volume of distribution. The advice from the product information is taken to start at the lower end of the normal dosing range. It is possible to conduct therapeutic drug monitoring with clobazam, which could be of help when using clobazam in these patients.

Clorazepate: no study has been performed in patients with cirrhosis. Clorazepate has a long elimination half-life. No dosing advice could be given.

Diazepam: Seven studies (level 3-4) examined the pharmacokinetics of diazepam in 65 patients with cirrhosis (severity unknown). Clearance was halved in four studies and elimination half-life prolonged 2-5 times in patients with cirrhosis. In a study among patients with ascites, probably decompensated cirrhosis, clearance was reduced by 60%. Six studies (level 2-4) among 61 cirrhotic patients studied the safety of diazepam, just as four pharmacokinetic studies. Disturbances in psychomotor function and (minimal) hepatic encephalopathy were seen in three studies and two case-reports. In agreement with a review it is advised to use half of the normal dose in cirrhotic patients and proceed slowly.

Flunitrazepam: in one small single-dose study (level 3) the pharmacokinetics of flunitrazepam were comparable between healthy controls and 6 patients with cirrhosis (4 CTP A/B, 2 CTP C). However, large interindividual variations were observed. The product information advises to use a lower dose. We advise to start at the lower end of the normal dosing range.

Flurazepam: no studies have been performed with flurazepam in patients with cirrhosis. It has a high first pass effect and high hepatic clearance, so changes in pharmacokinetics can be expected and it is probably better to use another benzodiazepine. No dosing advice could be given.

Loprazolam: there were no studies found where loprazolam was used in patients with cirrhosis. Therefore, no dosing advice could be given.

Lorazepam: Two pharmacokinetic studies (level 3-4) have been performed in 19 cirrhotic patients. Clearance was similar to healthy controls and exposure was only slightly increased. Elimination half-life was increased from 16 to 20 h and from 20 to 40 h. A case-report (level 4) described a case of coma in patient using lorazepam and diazepam. The product information advises to use half of the normal dose. We advise to use half of the normal dose only in CTP C patients where glucuronidation is probably impaired.

Lormetazepam: In one pharmacokinetic study (level 3) five patients with CTP A/B cirrhosis received lormetazepam. Peak plasma levels, even as exposure was doubled compared to healthy controls. We advise to use half of the normal dose in patients with CTP A and B cirrhosis. No data is available about patients with CTP C, so dosing advice could be given.

Midazolam: There were 12 pharmacokinetic studies (level 3-4) found in which a total of 115 patients with cirrhosis (CTP A/B/C 18/15/3, rest unknown) received midazolam. After oral dosing, bioavailability was increased by 40% and clearance approximately halved. After intravenous dosing, clearance was reduced with increasing severity of cirrhosis and elimination half-life prolonged to 13 hours at longest in CTP C patients.  In 13 studies (level 1-4) the safety of midazolam in 815 patients with cirrhosis (all CTP-classes) was assessed and 4 pharmacokinetic studies also looked at the safety. In a meta-analysis patients undergoing sedation for endoscopy had a significant longer recovery time with midazolam than with propofol. Propofol also impacted minimal hepatic encephalopathy less than midazolam did. Other studies found an impairment in psychomotoric functions after midazolam use (significantly more than diazepam), and in some larger studies around 5% of patients developed overt hepatic encephalopathy after midazolam sedation. A study found that this was related to CTP-class and dose of midazolam. Cirrhotic patients show an increased cerebral sensitivity to midazolam and require less midazolam for sedation. Because of the altered pharmacokinetics and the risks of hepatic encephalopathy with higher doses, it is advised to start with half of the normal dose in CTP B and C cirrhosis.

Nitrazepam: In one single-dose study (level 3) 12 patients with cirrhosis (severity unknown) received nitrazepam intravenously. The total clearance was almost similar to healthy controls, while unbound clearance was reduced by 35%. For patients with CTP A it is advised to start at the lower end of the normal dosing range. For patients with CTP B more caution is advised and it is therefore recommended to start with half of the normal dose. No data is available for patients with CTP C, so no dosing recommendation could be given.  

Oxazepam: Three pharmacokinetic studies (level 3-4) have been performed in 18 patients with cirrhosis (11 severity unknown, 7 decompensated cirrhosis). It seems that clearance reduces with severity of the disease. In decompensated cirrhosis clearance is halved. Therefore, it is advised to use half of the normal dose in CTP C patients.

Prazepam: During the first pass through the liver, prazepam (prodrug) is metabolized to desmethyldiazepam (which is also the major metabolite of diazepam). There were no studies found with prazepam, but several studies with diazepam also assessed some pharmacokinetic parameters of desmethyldiazepam. Clearance of desmethyldiazepam was half of the value in healthy controls and elimination half-life approximately doubled. Because of the unknown influence of cirrhosis on this first-pass metabolism we could not give a dosing advice.

Temazepam: Pharmacokinetics of temazepam were studied in three studies (level 3-4) with 34 cirrhotic patients (CTP A/B/C, 2/7/6, 19 unknown). The only significant difference observed in cirrhotic patients is a postponed maximum plasma concentration (3 h instead of 1 h). No dose adjustment is necessary, however it is advised to start on the lower end of the normal dosing range.

Zolpidem: there is only one pharmacokinetic study (level 4) in six patients with cirrhosis (severity unknown) partly described in a pharmacokinetic review and in the FDA label. Peak plasma levels doubled in cirrhosis and the exposure (AUC) was fivefold increased compared to healthy controls. There were six cases of hepatic encephalopathy identified by the FDA associated with zolpidem, two were described in a case-report (level 4). In another cross-sectional study (level 4) six more cirrhotic patients had mental deterioration while using zolpidem. Because of the large pharmacokinetic alterations, the risk of hepatic encephalopathy and the availability of safer alternatives, zolpidem is classified as ‘unsafe’ in patients with cirrhosis. This is in contrast to the product information, which only advises one half of the normal dosage.

Zopiclone: the pharmacokinetics of zolpidem were studied in two single-dose studies (level 3 and 4) in 17 patients with cirrhosis (severity unknown). Peak plasma levels seemed comparable to healthy controls, but half-life was approximately doubled and AUC increased by 40%. The pharmacodynamic response was measured in one of these study and psychometric tests were more impaired in cirrhotics than in controls. The authors of one study and the product information advise to start with one half of the normal dose, i.e. 3.75 mg per day. We recommend to start at the lower end of the normal dosing range in CTP A and to start with half of the normal dose in CTP B. In patients with CTP C no data is available and no dosing advice could be given.

Pharmacokinetic data

Absorption: Most BZDs have a low hepatic extraction ratio and are almost completely absorbed after oral administration (F>70%). Exceptions are flurazepam, midazolam and zolpidem. Midazolam and zolpidem have an intermediate hepatic extraction ratio and flurazepam a high hepatic extraction ratio. The bioavailability of flurazepam is only 30% because of a large first-pass effect. Prazepam is a prodrug metabolized to desmethyldiazepam during the first pass through the liver.

Distribution: Almost all BZDs are highly bound to plasma proteins (fb=70-98%). Zopiclone is the only one with a lower protein binding (fb=45%). Most have a volume of distribution of around 1-3 L/kg. Chlordiazepoxide and loprazolam have a much smaller distribution volume (around 4 L), while prazepam (15 L/kg) and zopiclone (100 L/kg) have a much larger volume of distribution volume.  

Metabolism: All BZDs are metabolized in the liver. Most are metabolized via CYP450-enzymes (mainly CYP3A4, CYP2C19) and have one or more active metabolites. Some BZDs are only glucuronidated (i.e. lorazepam, lormetazepam, oxazepam and temazepam).

Elimination: Most BZDs and metabolites are excreted in urine. Zolpidem is also for a substantial amount biliary excreted (40%). The elimination half-life varies among BZDs. BZDs with a relatively short half-life (<6 h; without active metabolite) are: brotizolam, midazolam, zolpidem and zopiclone. A slightly longer half-life (6-12h) have: loprazolam, lormetazepam, oxazepam and temazepam. A long half-life (12-40 h) have: alprazolam, bromazepam, clobazam, flunitrazepam, lorazepam and nitrazepam. BZDs with a very long half-life (>40 h) are: chlordiazepoxide, clorazepate, diazepam, flurazepam and prazepam.

Pharmacokinetic studies:         

  • Alprazolam: One single-dose study measured a prolonged half-life (20 h vs. 11 h in controls) with clearance being halved in cirrhotics. Exposure was more than doubled (+140%) in patients compared to healthy controls.
  • Bromazepam: no study was conducted.
  • Brotizolam: In one single-dose study two cirrhotic patients showed hardly any decline in plasma concentration and were excluded from data analysis. For the other 6 patients, half-life was almost doubled to 13 h compared to healthy controls. The protein binding was decreased in cirrhotics with an increased volume of distribution. Exposure was increased by 42% in patients compared to healthy controls.
  • Chlordiazepoxide: In four single-dose studies pharmacokinetics of chlordiazepoxide IV and IM were examined. Clearance was reduced by 50-66% in all studies and the elimination half-life prolonged from 10-20 hours in healthy controls to 40-60 hours in cirrhotics. The peak plasma concentration and the exposure to the active metabolite was decreased.
  • Clobazam: In one single-dose study the volume of distribution in cirrhotics was more than doubled (180 L vs. 80 L in controls). Elimination half-life was also prolonged to 50 h (20 h in controls). Exposure, even as clearance, was comparable between cirrhotics and controls.
  • Clorazepate: no study was conducted.
  • Diazepam: there were seven clinical studies examining diazepam pharmacokinetics in cirrhotic patients. No study determined the AUC, but clearance was halved in four studies. Elimination half-life was prolonged in all studies to 50-160 hours in patients compared to 30-70 hours in controls. One multiple-dose study measured concentration at steady-state of diazepam, which was 70% higher in patients compared to controls. For the major active metabolite, steady state plasma concentration was doubled compared to controls. The free fraction of diazepam was doubled with a greater volume of distribution. There was also one pharmacokinetic modelling study that predicted exposure to diazepam after oral dosage. The total AUC was predicted to increase with 117% in CTP A, 155% in CTP B and 218% in CTP C. The unbound AUC was predicted to increase with 136% in CTP A, 231% in CTP B and 422% in CTP C.
  • Flunitrazepam: in one single-dose study pharmacokinetics of 6 cirrhotic patients were similar to that of healthy controls, although large variations were observed between individual patients.
  • Flurazepam: no study was conducted.
  • Loprazolam: no study was conducted.
  • Lorazepam: two studies examined lorazepam pharmacokinetics after a single IV-dose. In both studies half-life increased (from 16 to 20 h in one and from 20 to 40 hours in the other). Protein binding of lorazepam decreased from 93.2% to 88.6% in cirrhotics. Even so, the volume of distribution was increased in this study. Clearance was however similar to controls and the exposure was only increased by 30%.
  • Lormetazepam: in one, small, single-dose study lormetazepam was administered both PO and IV. After oral dosing, maximum plasma concentrations were doubled in the 5 cirrhotic patients (CTP A and B), even as the exposure, compared to healthy controls. Exposure to the active metabolite lorazepam was decreased. Following intravenous administration, exposure was only increased by 33%, and clearance reduced by 25%.
  • Midazolam: The pharmacokinetics of midazolam were examined in ten clinical studies:
    • Absorption: after oral administration the peak plasma level was increased by 40% in one study. Bioavailability was comparable in cirrhotic patients without a porto-systemic shunt (TIPS), but increased to 75% in cirrhotic patients with a TIPS (F=30% in controls), this was mainly caused by a diminished intestinal CYP3A activity. In a pharmacokinetic modelling study the predicted bioavailability was 49% in CTP A, 57% in CTP B and 66% in CTP C (F=37% in controls).
    • Distribution: The free fraction was doubled in two studies, while the third found an increase in free fraction with increasing severity of cirrhosis. The volume of distribution was larger in some studies, while no differences were found in others.
    • Metabolism: in one study CYP3A4 activity was measured by the ratio of [metabolite/midazolam]. They found a significantly decreased activity in patients with end-stage liver disease. Also, a relationship was found between ALAT and CYP3A4 activity and total bilirubin and CYP3A4 activity.
    • Elimination: In all eight studies that measured half-life, it was increased in patients compared to controls. One study looked per CTP-class and found a normal t1/2 in CTP A (3 h), prolonged in CTP B (11 h) and even more prolonged in CTP C (13 h). Other studies found a t1/2 in the range of 3-10 hours in cirrhotic patients. Clearance was determined in five studies. Most found a clearance decreased by 50% in patients with cirrhosis. In the study in patients with a TIPS, clearance was only one-fifth of healthy controls. Another study found a comparable clearance in CTP A and a decreased clearance with 66% in CTP B and C patients. Both the CTP score and the MELD score correlated well in this study with midazolam unbound clearance and half-life.
    • Exposure: two studies found an increase of resp. 43% and 57% in patients with cirrhosis compared to controls. In another study the AUC was 33% increased in CTP A and 92% in CTP B. A PK modelling study predicted after IV administration: a doubled unbound AUC in CTP A, +250% in CTP B and +500% in CTP C. After oral administration the predicted unbound AUC was +100% CTP A, +300% CTP B and +600% in CTP C. Another study predicted an increase of 150% in AUC after oral administration and 70% after intravenous administration.
  • Nitrazepam: in one single-dose study patients received nitrazepam intravenously. The unbound fraction was increased from 13.5% to 19% in cirrhotics. The unbound clearance was also decreased by about 35%. Total clearance and half-life were almost similar to healthy controls.
  • Oxazepam: in three studies patients received oxazepam orally as a single dose. One study in patients with decompensated cirrhosis found a prolonged half-life from 6 hours in controls to 17 hours in patients. This study also found an increased unbound fraction (15.4% vs. 4.6%) and an unbound clearance decreased by 80%. The total clearance was approximately halved. One of the other studies showed conflicting results with no major pharmacokinetic changes, while the third study found an almost similar elimination half-life, but did found an increased exposure (+70%) in patients. The study with decompensated cirrhotics, found a correlation between unbound clearance of oxazepam and antipyrine clearance (measure of hepatic function).
  • Prazepam: no pharmacokinetic studies have been performed.
  • Temazepam: three studies looked at the pharmacokinetics of a single-dose of temazepam in patients with cirrhosis. Two studies found a shorter elimination half-life, while the third noted a slight increase to 13 hours instead of 8.5 hours in healthy controls. The unbound fraction was increased to 5.5 % in one study (3.9% in controls), with a decrease in unbound clearance of approximately 33% in two studies. Total clearance was however similar in both groups in all 3 studies and no correlation was found with CTP score.
  • Zolpidem: in one review results are described which are also mentioned in the product information. The maximum plasma concentration was doubled in patients with cirrhosis compared to healthy controls and half-life prolonged to 10 hours (2 hours in controls). The total exposure was fivefold higher in patients with cirrhosis. In a modelling study unbound AUC was predicted to be +75% in CTP A, +200% in CTP B and +400% in CTP C.
  • Zopiclone: Two studies explored the pharmacokinetics of a single-dose of oral zopiclone in patients with cirrhosis. Bioavailability increased by 20% in patients with cirrhosis, but peak plasma level was comparable. Half-life was prolonged to 8.5 hours in the two studies (3.5-5 hours in controls). The exposure to zopiclone was increased by 43% in patients. One of the studies found a negative correlation between serum albumin and elimination half-life.

Safety data

General safety studies

In a recent case-control study, cirrhotic patients using BZDs were compared to patients not using BZDs. Continuous use of BZDs carried a higher risk of BZD-associated hepatic encephalopathy (duration-dependent risk). The risk was also higher with BZDs with a long t1/2 (>12 h) than a short t1/2 (dose-dependent increase in episodes). Multivariate analyses showed that long t½-BZD-use, long duration of use (>2 months) and high dosage (>1.5 DDD) were markers for development of BZD-associated hepatic encephalopathy.

Another study looked at plasma concentrations of BZDs and found these significantly increased in cirrhotic patients with severe hepatic encephalopathy compared to controls, cirrhotic patients with no encephalopathy and cirrhotic patients with moderate encephalopathy. Increased plasma BZD levels could be accounted for by prior exposure to BZD medication in all cases. Cerebrospinal fluid levels were not different.

Individual benzodiazepine safety studies

No specific safety study: alprazolam, bromazepam, clobazam, clorazepate, flunitrazepam, flurazepam, loprazolam, lormetazepam, nitrazepam, prazepam and temazepam.

Brotizolam: in the pharmacokinetic single-dose study, the authors noted that the sensitivity towards 0.5 mg brotizolam appeared larger for cirrhotics than patients (former all fell asleep, later not).

Chlordiazepoxide: In a clinical trial four cirrhotic patients received chlordiazepoxide (25-200 mg/d) for alcohol withdrawal. One became severely drowsy, but rapidly improved once the drug was withdrawn. In a case-report a cirrhotic patient developed hepatic encephalopathy after receiving chlordiazepoxide, and reacted well on flumazenil. Chlordiazepoxide and metabolites were still detectable after two months. In a prospective study among 100 cirrhotic patients with hepatic coma, chlordiazepoxide was the cause in 12 cases.

Diazepam: In three clinical trials the safety of diazepam in cirrhotics was examined. Two of these found disturbances in psychomotor function and minimal hepatic encephalopathy in cirrhotic patients on diazepam (0.1-0.15 mg/kg), and not in healthy controls or in patients on propofol. In one study this observed degree of impairment was greater after midazolam than following diazepam. In the last trial, after a single dose of 5 mg diazepam IV transient drowsiness occurred in all 16 cirrhotics, but none developed prolonged deterioration in conscious level. In two case-reports cirrhotic patients became comatose after diazepam treatment, which was in one case associated with diazepam plasma levels which did not decline. Both recovered after administration was stopped. In a cross-sectional retrospective study three patients on diazepam reported mental adverse events. Four pharmacokinetic studies also looked at safety, with three of these finding an increased sedation in cirrhotic subjects compared to controls, especially in patients who previously had hepatic encephalopathy. The last study (with only 3 control subjects) did not find this difference.

Lorazepam: in a case-report a cirrhotic woman was treated for delirium tremens with 12 mg lorazepam and 20 mg diazepam and became comatose. The following days she had 1-6 mg lorazepam per day. A spindle coma was detected, lorazepam was stopped and she recovered during the following days.   

Midazolam: 17 studies (partly) investigated the safety of midazolam in patients with cirrhosis. One meta-analysis (including five RCTs) compared midazolam with propofol for sedation during gastro-intestinal endoscopy. Patients on midazolam had a significantly longer recovery time than on propofol (but a high level of heterogeneity I2=82%). Propofol also impacted minimal encephalopathy less than midazolam did. In another RCT disturbances in psychomotor function in cirrhotic patients on midazolam, and not in healthy controls were found. In this study the observed degree of impairment was greater after midazolam than after diazepam. In 5 other clinical trials midazolam was used IV for sedation during an endoscopy. Two studies found differences in pre- and post-sedation psychomotoric function in cirrhotics, with an increase in subclinical hepatic encephalopathy. In the three others 19 patients developed overt hepatic encephalopathy (respectively 4.2%, 0.4% and 6.6%). In one study among 191 patients, overt hepatic encephalopathy was significantly related to CTP class and dose of midazolam. In a cross-sectional retrospective study 2 patients on midazolam reported mental adverse events and for one of these patients hepatic coma was the reason for hospital admission. In four pharmacokinetic studies, pharmacodynamics were also assessed, 1 study observed prolonged sedation in 1 CTP A patient and control and in 1 one the hypnotic effects were the same in patients and controls. The effective concentration of midazolam was 30% less in cirrhotic patients. In the last study the results of psychomotoric tests suggest there is increased cerebral sensitivity in cirrhotic patients to midazolam.

Oxazepam: in one of the pharmacokinetic studies, adverse events were also monitored. Apart from a light sedation, no other adverse events were seen after 15 mg oxazepam in decompensated cirrhotics. Peak plasma concentration was halved.

Zolpidem: in the pharmacokinetic study the onset of drowsiness and the duration were not different from healthy controls after a single dose. There were however 2 case-report in which hepatic encephalopathy developed in patients with cirrhosis. The dose was 10 mg in one and 5 mg in the other. In a cross-sectional retrospective study 6 patients on zolpidem reported mental adverse events.

Zopiclone: in a pharmacokinetic study pharmacodynamics were also tested. The maximum changes in psychometric tests appeared to be greater in cirrhotic patients than in healthy controls, although peak plasma levels were similar. This suggest patients may be more susceptible to the central nervous system depressant effects.