Hospitalizations During the Use of Rifaximin Versus Lactulose for the Treatment of Hepatic Encephalopathy
- First Online:
- Cite this article as:
- Leevy, C.B. & Phillips, J.A. Dig Dis Sci (2007) 52: 737. doi:10.1007/s10620-006-9442-4
- 864 Views
We sought to compare frequency and duration of hepatic encephalopathy-related hospitalizations during rifaximin versus lactulose treatment. Hospitalizations, clinical efficacy data, and adverse events obtained from charts of 145 patients with hepatic encephalopathy who received lactulose (30 cc twice daily) for ≥6 months and then rifaximin (400 mg 3 times a day) for ≥6 months compared last 6 months on lactulose (lactulose period) to first 6 months on rifaximin (rifaximin period). Fewer hospitalizations (0.5 versus 1.6; P < .001), fewer days hospitalized (2.5 versus 7.3; P < .001), fewer total weeks hospitalized (0.4 versus 1.8; P < .001), and lower hospitalization charges per patient ($14,222 versus $56,635) were reported during the rifaximin period. More patients had asterixis, diarrhea, flatulence, and abdominal pain during the lactulose period (P < .001). Treatment of hepatic encephalopathy with rifaximin was associated with lower hospitalization frequency and duration, lower hospital charges, better clinical status, and fewer adverse events.
Chronic liver disease affects approximately 5.5 million individuals in the United States , and data suggest that the majority of these patients develop hepatic encephalopathy ranging from minimal to severe [2–6]. Hepatic encephalopathy is manifested by a spectrum of neuropsychiatric and motor abnormalities. Depending on the stage of the illness, the neuropsychiatric impairment in hepatic encephalopathy ranges from mild alteration of cognition and consciousness to coma; neuromuscular symptoms range from mild tremor to decerebrate posture [7, 8]. Hepatic encephalopathy is associated with frequent use of health care, including emergency care and hospitalization. According to data from the 2003 Healthcare Cost Utilization Project, hepatic encephalopathy was listed as the primary diagnosis in 40,012 hospital discharges and as a diagnosis at any time during the hospital stay in 95,752 hospital discharges. Mean length of stay among patients with a primary diagnosis of hepatic encephalopathy was 5.7 days, and mean charge per patient per stay was $23,192 . The total hospital charges associated with a diagnosis of hepatic encephalopathy approached $1 billion in 2003 and seem to be increasing. These data underline the need for effective strategies for managing the clinical and economic consequences of hepatic encephalopathy.
Current approaches to management of hepatic encepha- lopathy include avoidance of precipitating factors, dietary protein restriction, bowel cleansing, nonabsorbable disaccharides, antibiotics, and experimental pharmacotherapies [10, 11]. The nonabsorbed ( < 0.4%) oral antibiotic rifaximin is an investigational orphan drug for hepatic encephalopathy in the United States, where it is also licensed for the treatment of travelers’ diarrhea. The drug was first introduced in Italy in 1987 and is now approved in 19 countries for uses including hepatic encephalopathy. On the basis of data from 20 clinical studies and experience in medical practice, rifaximin appears to have a favorable benefit:risk ratio in the treatment of hepatic encephalopathy . Rifaximin concentrates in the gastrointestinal tract, has broad spectrum in vitro activity against gram-positive and gram-negative aerobic and anaerobic enteric bacteria, and has a tolerability profile comparable to that of placebo with no known clinical drug interactions [11–14]. In comparator studies involving either nonabsorbable disaccharides or other antibiotics, rifaximin reduced blood ammonia levels and improved neurologic signs and symptoms earlier and/or more markedly than the comparator in patients with hepatic encephalopathy. The tolerability profile of rifaximin differentiates it from lactulose, which is associated with unpredictable, severe diarrhea, and often intolerably sweet taste, and from neomycin and paromomycin, which can cause toxicity manifested by effects such as hearing loss and renal injury.
Whereas the clinical efficacy and tolerability profiles of rifaximin in hepatic encephalopathy have been elucidated, the impact of rifaximin on hepatic encephalopathy-associated health care utilization and costs have not previously been assessed. This retrospective chart review was undertaken to compare the need for hospitalization and its associated charges for patients with hepatic encephalopathy during treatment with rifaximin with those during treatment with lactulose, the current mainstay of therapy. In addition, clinical status and adverse events during treatment with lactulose and rifaximin were assessed.
This study was a retrospective chart review conducted at a single study center in May 2005. The study compared frequency, duration, and hospitalization charges associated with a discharge diagnosis of hepatic encephalopathy, clinical status, and adverse events during treatment with lactulose with those during treatment with rifaximin.
Eligible patients were diagnosed with hepatic encephalopathy and were treated with lactulose 30 cc twice daily for ≥6 months followed by rifaximin 400 mg 3 times a day for ≥6 months. All patients diagnosed with hepatic encephalopathy and consulting the study center were switched to rifaximin upon its availability in the United States in 2004. Patients who had received a liver transplant during therapy with lactulose or rifaximin were excluded.
Charts of patients meeting inclusion and exclusion criteria were reviewed to compare outcomes during the last 6 months of therapy on lactulose (lactulose period) with the first 6 months of therapy on rifaximin (rifaximin period). The primary end point was the mean number of hospitalizations during each treatment period. Hospitalization data were obtained for all patients in the study sample regardless of whether they were hospitalized at the primary author's hospital or at other institutions. Therefore, all hospitalizations occurring during the study period in the sample were captured. Secondary end points compared between treatment periods included the average length of hospitalization, mean total time hospitalized, and hospitalization charges per patient.
Hospitalization charges were calculated in 2005 dollars based on information from HCUPnet, an electronic query system . HCUPnet provides access to cost and utilization data from U.S. community hospitals (including short-term, nonfederal, general, and other hospitals) in the Healthcare Cost and Utilization Project (HCUP), which maintains the largest set of publicly available all-payer health care databases in the United States. The nationwide inpatient sample collects data from 995 hospitals in 35 states and includes data from 7 million hospital stays. The sample represents approximately 90% of all discharges in the United States. The average charge per hospital day in 2003 was determined for the modified International Classification of Diseases, 9th Revision, Clinical Modification (ICD-9-CM) principal diagnosis code of 572.2 (hepatic encephalopathy). The charges did not include professional fees. A health care cost index was used to predict 2004 and 2005 charges from the 2003 costs in the HCUP database.
Clinical endpoints included hepatic encephalopathy grade based on modified West Haven criteria  at the end of the treatment period, presence of asterixis at the end of each treatment period, and adverse events that occurred any time during the treatment periods and that were considered by the investigator to be related to study medication. The investigator recorded each adverse event using a predefined descriptor (diarrhea, flatulence, abdominal pain, headache, or other) and rated its severity (very severe, severe, moderate, mild, or none). In addition, compliance during each treatment period was assessed based on patient reports of the percentage of prescribed doses taken (100%, 75–99%, 50–74%, 25–49%, or 0–24%).
All variables were summarized with descriptive statistics. Results were compared between treatment periods with paired t-tests for the mean number of hospitalizations, average length of hospitalization in days, the mean total days hospitalized, and the percentage of patients with each adverse event (diarrhea, flatulence, abdominal pain, headache, or other); with Bowker's test for symmetry for hepatic encephalopathy grade at the end of each treatment period, the maximum severity for all occurrences of each adverse event, and compliance categories; and with McNemar's test for the presence of asterixis at the end of the treatment period.
Demographics and baseline characteristics
Mean age, years (SD)
Male, n (%)
Cirrhosis, n (%)
On liver transplant waiting list, n (%)
Summary of clinical data and compliance
Hepatic encephalopathy grade,*n (%)
Asterixis present at end of treatment,*n (%)
Abdominal Pain,*n (%)
Headache, n (%)
Other, n (%)
Hepatic encephalopathy grade at the end of each treatment period reflected less severe illness with rifaximin than with lactulose (P < 0.001). The percentage of patients with stage 3 or stage 4 hepatic encephalopathy was 6% with rifaximin and 25% with lactulose (Table 2). Significantly fewer patients had asterixis at the end of the rifaximin period (63%) than the lactulose period (93%) (P < 0.001; Table 2).
The percentages of patients with diarrhea, flatulence, and abdominal pain were significantly higher during the lactulose period than the rifaximin period (P < .001 for each adverse event; see Table 2). The percentage of patients with headache did not differ between treatment periods (P=.718). No adverse events in the other category were reported during either treatment period.
This retrospective chart review demonstrates reductions in the need for hospital care and its associated charges during treatment of hepatic encephalopathy with rifaximin relative to treatment with lactulose. In this sample of patients treated for ≥6 months with lactulose followed by ≥6 months with rifaximin, both frequency and duration of hospitalization over 6 months were 3 times lower with rifaximin than lactulose. Furthermore, total time hospitalized over 6 months and hospitalization charges were approximately 4 times lower with rifaximin than lactulose. The charge per patient over 6 months was >$42,000 less during the rifaximin period than it was during the lactulose period.
The reductions in need for hospitalization and charges associated with hospitalization during the rifaximin period might be attributed to less severe illness with rifaximin than with lactulose as reflected by the findings of fewer patients with stage 3 or 4 hepatic encephalopathy and with asterixis at the end of the treatment period. The results are consistent with the possibility that the reductions in frequency, duration, and hospitalization charges are attributed to better efficacy of rifaximin. However, this conclusion is not unequivocally supported because of limitations in the study design, which was retrospective and observational. The study sample was selected on the basis of meeting the criterion that patients be treated for ≥6 months with lactulose followed by ≥6 months with rifaximin. Because patients were not randomized to a treatment group and the order of treatments was not counterbalanced, the possibility that factors other than rifaximin accounted for the lower need for hospitalization and the better clinical outcomes cannot be excluded. Time-related factors rather than treatment-related factors might explain these findings.
Better compliance with rifaximin than lactulose might have contributed to the better clinical outcomes during the rifaximin period. The majority (92%) of patients reported taking 75–100% of their prescribed dose during the rifaximin period whereas approximately one third (31%) of patients reported this compliance rate during the lactulose period. The compliance data should be interpreted cautiously because they are based on patient self-reports rather than objective measures.
The low compliance rate during the lactulose period might have resulted from poor tolerability of lactulose, a speculation consistent with previous research showing that adverse events such as diarrhea and nausea are primary reasons for discontinuation of lactulose therapy . In the current study, the majority of patients experienced diarrhea and abdominal pain of at least moderate severity during the lactulose period; these adverse events were severe or very severe in 32% of patients for diarrhea and 16% of patients for abdominal pain. The high incidence of diarrhea, abdominal pain, and flatulence during the lactulose period in this study is consistent with previous findings [7, 17, 18].
The tolerability profile of rifaximin in the current study is also consistent with previous reports [12–14]. In placebo-controlled studies of rifaximin for hepatic encephalopathy and travelers’ diarrhea, the adverse event profile of rifaximin was comparable to that of placebo [19, 20]. In comparator studies of rifaximin for hepatic encephalopathy, rifaximin was often reported to be better tolerated than nonabsorbable disaccharides or other antibiotics .
In conclusion, treatment of hepatic encephalopathy with rifaximin for 6 months compared with lactulose for 6 months was associated with lower frequency and duration of hospitalization, less total time hospitalized, lower hospital charges, better clinical status, and fewer adverse events. Results should be interpreted cautiously as the study was retrospective, patients were not randomized to treatment, and treatment order was not counterbalanced.