Sub-Clinical Hepatic Encephalopathy in Cirrhotic Patients Subjected to Sedation with either Propofol or Midazolam‎

INTRODUCTION

Hepatic encephalopathy is an acute or chronic neurological disorder due to liver disease [1]. Sub-clinical hepatic encephalopathy or minimal hepatic encephalopathy (MHE), shows measurable neuropsychological disorders, but with normal neurological and mental status on clinical examination [2, 3].

Upper gastrointestinal endoscopy (UGE) is needed to be done regularly in chronic liver disease (CLD) patients for management and follow-up of esophageal varices, gastric varices and portal hypertensive gastropathy which are complications due to portal hypertension [4]. Sedation is needed during this procedure to help patient’s tolerance [5].

Patients with CLD are in risk of sedation complications during UGE. This is attributed to metabolization of some of these sedative drugs in the liver, as well as due to the possible hemodynamic changes that are associated with the anesthetic drugs, due to higher plasma level and prolonged effect of these drugs due to their delayed clearance [6].

Midazolam is a benzodiazepine that is frequently used for mild sedation in the general population undergoing UGE [7]. It is anxiolytic and amnesic with half-life

Propofol is a short-acting anesthetic drug that can be used drug during UGE. In comparison to benzodiazepines, it has convenient pharmacokinetic properties [10].

The aim of this study is to determine the impact of propofol in comparison to midazolam on the occurrence of sub-clinical hepatic encephalopathy in liver cirrhotic patients undergoing UGE.

PATIENTS AND METHODS

Study population

We conducted retrospective observational study, comprised data of 70 Egyptian patients; male and female (aged 17 to 65years) collected from endoscopy center of Ain Shams University Hospital, Cairo, Egypt, from December 2014; 60 patients were hepatic patients with CLD regardless of etiology (Child-Pugh class A, B or C) and 10 patients non hepatic who underwent diagnostic or therapeutic UGE.

Inclusion criteria

All patients were matched for gender, age, body mass index (BMI) and divided to three groups. Group A included 10 control non hepatic patients who underwent UGE without sedation (neither propofol nor midazolam) for medical condition other than liver diseases. Group B included 30 patients with CLD received midazolam for sedation. Group C included 30 patients with CLD received propofol for sedation.

All patients were subjected to complete blood count, liver and renal profile investigations, number connection test before and after the procedure, estimation of time to recovery and time of the procedure and hemodynamic parameters (blood pressure, pulse, respiratory rate, and oxygen saturation) before, during and after sedation.

Exclusion criteria

Patients with psychiatric disease, active neurological disease, overt hepatic encephalopathy, alcohol or drug abuse, active respiratory illness, gastrointestinal hemorrhage, hepatocellular carcinoma and allergy to sedative drugs were excluded from the study.

Technique

The number connection test (NCT) defines the time required to connect sequentially the randomly placed circles labeled from 1 to 25. The degree of encephalopathy was determined according to the time needed to complete the test (Table 1).

Table (1): The degree of minimal hepatic encephalopathy according to NCT.

NCT: Number connection test

Baseline vital signs including oxygen saturation, non-invasive blood pressure monitoring and 5-leads electrocardiogram before, throughout and post-procedure were recorded.

Intravenous bolus injection of midazolam 1 mg was given with increasing doses every 2–3 minutes till satisfactory sedation achieved suitable for carrying out the procedure or reaching total dose of 3 mg. Propofol was started with 1 mg/kg by an anesthesiologist followed by 0.5 mg/kg upon demand every 3 minutes. The patients were sedated by the aforementioned doses aiming to reach a moderate level of sedation where the patient respond appropriately to verbal command with or without light tactile stimuli. Patients who did not tolerate or the depth of the sedation was increased or even general anesthesia was ensured were excluded from the study.

A record of complications was made for the following events: desaturation <92% on pulse oximetry, decrease of systolic blood pressure

The degree of alertness and the time for full recovery were recorded. The full recovery was determined by the time when the patients will be oriented to the time, place and persons and achieving Aldrete score> 8. Aldrete scoring system is a 5 points scale including, activity, respiration, circulation, consciousness and oxygen saturation. the patients are assessed on a scale of 0 (worst), 1 (modest) and 2 points in the best condition.

The hemodynamic data, the ability to start oral fluids, the ability to stand-up without assistance and pain free were checked. Any pain encountered were managed using 1 gm of paracetamol intravenously. Patients who received narcotics as 0.5 mg/kg pethidine, were excluded from the study [11].

All patients repeated NCT after one hour, before discharge from the endoscopy post-anesthesia care unit (PACU) to ensure elimination of hypoxemia and antegrade amnesia which could influence the judgment.

After acceptance of all these parameters, the endoscopist discussed the procedure results with the patient and his family members.

Statistical analysis

Statistical Package for Social Science (IBM SPSS) version 23 was used for data analysis. Qualitative data was represented by number and percentages. Parametric quantitative data were represented by mean and standard deviations.  Non-parametric quantitative data represented by median and inter-quartile range (IQR). Chi-square test compared qualitative data between groups. One-way ANOVA compared parametric quantitative data between groups.  Mann-Whitney test compared non-parametric quantitative data. Sensitivity, specificity, positive and negative predictive values were determined. A receiver operating characteristic (ROC) analysis was formed and the best cut-off value for outcome was determined. The confidence interval was 95% and the accepted margin of error was 5%. P-value was considered significant if it is < 0.05.

RESULTS

Regarding the demographic data of studied population, the median age with inter-quartile range (IQR) for the studied cases in group A was 46.5 (40.5 –57.5) years old, 48.5 (38.25 – 42.5) years old in group B and 50 (45 – 57) years old in group C. Six of them (60%) were male patients in group A, 24 (80%) in group B and 20 (67%) in group C. The mean value  of BMI was 27.8±6 in group A, 27.7±5.4 in group B and 28.1±3.9 in group C.

The whole number of CLD in the studied population were 60 patients, 40 patients (66.67%) due to HCV, 5 patients (8.3%) due to HBV, 7 patients (11.67%) due to bilharziasis, 3 patients (5.0%) due to Budd-Chiari syndrome, 2 patients (3.3%) due to autoimmune hepatitis and 3 patients (5.0%) of unknown etiology.

The number of cirrhotic patients with abnormal pre-sedation NCT times in the midazolam group was 18 patients (60%), and this number was raised to 25 patients (83.3%) after one hour following sedation with midazolam (P<0.001). The number of cirrhotic patients with abnormal pre-sedation NCT times in the propofol group was 22 patients (73.3%), and this number did not change after one hour following sedation with propofol (P= 0.714).

Regarding NCT before, table (3) showed significant increase in group C versus group B, while both groups B and C showed significant increase versus group A.

Regarding NCT one hour after, table (3) showed significant increase in group B versus group C, while both groups B and C showed significant increase versus group A.

Regarding delta change of NCT, table (3) showed significant increase in group B versus both groups A and C.

Table (3) showed significant higher NCT value one hour after than NCT before in group B (P<0.001).

Times of procedure and recovery was significantly increased in group B in comparison to group C [median (IQR): 15 (9.25-20) and 39 (38-40.25) respectively, p<0.001).

A significant decrease of blood pressure associated with significant increase in pulse and decrease of oxygen saturation was recorded during procedure in group B (P<0.001) and C (P<0.004). These changes were improved significantly after procedure except blood pressure in group B (P<0.016). Oxygen saturation was significantly decreased in group C versus group B during procedure with significant lesser improvement in group C after applying rebreathing oxygen face mask at a flow of 7 L/min with no need for intubation (Table 4).

Table (2): Laboratory data of cirrhotic patients who received sedation.