The prognosis of cirrhosis has improved following the development of a number of effective interventions [
Identifying factors associated with HRQOL may help improve patient care and guide future research [
From February 2008 to May 2012 we conducted a prospective cohort study on patients with cirrhosis recruited from two Danish liver outpatient clinics. Patients were eligible for inclusion if they had histological or clinical cirrhosis and were able to read Danish. Patients with overt HE (West Haven Criteria Grades 2 to 4) and concurrent malignancy were excluded. The study protocol conformed to the ethical guidelines of the 1975 Declaration of Helsinki and was approved by the Danish ethics committee.
Patients were identified at their regular visits to the participating clinics or through the Danish case-mix system of diagnostic codes (dkDRG; Diagnosis Related Groups, based on the ICD-10 classification system). Patients identified electronically were invited to participate by telephone. Patients who agreed to participate completed a written informed consent at the inclusion visit.
At inclusion, demographic data, the patient history, and medication were recorded. A full physical examination and a broad screening panel of tests (including liver function tests, haematology, creatinine, and electrolytes) were performed. The Child-Pugh score was calculated. Evidence of covert HE [
HRQOL was evaluated using the six-point, 29-item Chronic Liver Disease Questionnaire (CLDQ) [
Patient characteristics were summarized as proportions with means and standard deviations/range. The CLDQ score was classed as low (<4) or high (4–6). Predictors of CLDQ scores and mortality were analyzed using binary logistic regression. Multivariable analyses were performed using backward elimination. The predictors were age, gender, body mass index (BMI), etiology of liver cirrhosis, employment, marital status, comorbidities, previous hepatic decompensation, covert hepatic encephalopathy, Child-Pugh score, ascites, ongoing alcohol abuse, albumin, and hyponatremia. Statistical analyses were performed using STATA version 12 (Stata Corp., College Station, Texas, US).
A total of 92 patients were included and followed for a mean duration of 20 months (range 3 to 52 months). Four patients withdrew their consent regarding the CRT test and the HRQOL questionnaire. Five patients did not complete the CRT test or the HRQOL scores because they had to be hospitalized due to worsening of their underlying liver disease. All 92 patients continued in the follow-up cohort and were included in the outcome analysis.
The mean age of included patients was 61 (SD 8.7; range 41 to 83 years) and 54 (59%) were men. The mean body mass index was 24 (SD 4.0; range 16 to 35). Most patients had cirrhosis due to alcoholic liver disease (Table
Patient characteristics at inclusion.
Variable | Mean ± SD (range) |
---|---|
Followup (months) | 19.9 ± 16.0 (3–52) |
Male gender | 54 (59) |
Employed | 14 (15) |
Married or similar | 51 (55) |
Age (years) | 61.5 ± 8.7 (41–83) |
Alcoholic liver cirrhosis | 79 (86) |
Child-Pugh score | 6.8 ± 1.6 (5–12) |
Child-Pugh class A/B/C | 46/38/8 |
International normalized ratio | 1.28 ± 0.3 (0.9–3.1) |
ALT (international units) | 32 ± 25 (3–163) |
Creatinine ( |
86 ± 35 (43–221) |
Sodium (mmol/L) | 137 ± 6 (101–146) |
Ammonia ( |
37 ± 34 (0–91) |
Albumin (g/L) | 39 ± 6 (27–50) |
Ongoing alcohol abuse | 39 (42) |
Ascites present | 26 (28) |
Prior decompensation | |
Ascites | 59 (64) |
Hepatic encephalopathy | 21 (23) |
Varices | 41 (45) |
Prior variceal bleeding | 13 (14) |
Comorbidities | |
Lung disease | 11 (12) |
Heart disease* | 30 (33) |
Kidney disease | 12 (13) |
Diabetes | 18 (20) |
Prior malignancy | 8 (9) |
aminotransferase.
Fifty-two percent of the patients were classed as Child-Pugh class A. The mean Child-Pugh score was 6.9 (SD 1.7; range 5 to 12). Seventy-seven patients had clinical signs of decompensation prior to inclusion. At baseline, 26 patients (28%) had ascites and 41 patients (45%) were diagnosed as having covert HE. Sixty-one patients were treated with lactulose, 61 with beta-blockers, 57 with loop diuretics, and 35 with spironolactone. Eighty-three received vitamins.
During followup, 19 patients died (21%), including seven classed as Child-Pugh class A at baseline (Table
Causes of death during followup.
Causes | Child-Pugh class |
---|---|
Gastrointestinal bleeding (1), variceal bleeding (2), hepatorenal syndrome (1), metastatic oropharynx cancer (1), and HCC (2). | A |
Metastatic lung cancer (2), progressive liver failure (2), HCC (1), sepsis (1), and unknown (1). | B |
Gastrointestinal bleeding (1), progressive liver failure (2), HRS (1), and HCC (1). | C |
HCC: hepatocellular carcinoma; HRS: hepatorenal syndrome.
Clinical outcomes during followup.
Event |
|
---|---|
Death | 19 (21) |
Transplantation | 1 (1) |
TIPS | 2 (2) |
Hepatic encephalopathy | 25 (27) |
Hepatorenal syndrome | 6 (7) |
Nonvariceal gastrointestinal bleeding | 17 (18) |
Variceal bleeding | 5 (5) |
Hepatocellular carcinoma | 4 (4) |
Spontaneous bacterial peritonitis | 6 |
Bacterial infections | 34 |
Other events, requiring hospitalisation | 25 |
TIPS: transjugular intrahepatic portosystemic shunt.
In univariable logistic regression, the Child-Pugh score and albumin predicted mortality (regression coefficient (
Univariable regression analysis of potential predictors for mortality and health-related quality of life.
Variable | Mortality | CLDQ score < 4 | ||
---|---|---|---|---|
Regression coefficient |
|
Regression coefficient |
|
|
Age | 0.017 | 0.576 | −0.020 | 0.507 |
Gender | −0.580 | 0.264 | −0.945 | 0.091 |
Body mass index | −0.147 | 0.083 |
|
|
Employment | −0.405 | 0.619 | −0.539 | 0.519 |
Marital status | 0.095 | 0.855 | −0.841 | 0.133 |
Nonalcoholic etiology of cirrhosis | −0.891 | 0.196 |
|
|
Comorbidities | ||||
Heart disease | 0.598 | 0.267 | 0.270 | 0.634 |
Pulmonary disease | — | — | −0.219 | 0.798 |
Renal disease | 0.758 | 0.263 | −0.515 | 0.537 |
Diabetes | 0.557 | 0.360 | −0.644 | 0.437 |
Previous malignancy | 0.921 | 0.238 | −0.348 | 0.763 |
Previous hepatic decompensation | ||||
Ascites | 0.459 | 0.425 | −0.270 | 0.634 |
Hepatic encephalopathy | 0.642 | 0.267 | −0.707 | 0.315 |
Esophageal varices | 0.118 | 0.820 | −0.879 | 0.140 |
Child-Pugh score |
|
|
|
|
Ascites at inclusion | −0.142 | 0.807 | 0.752 | 0.199 |
Minimal hepatic encephalopathy | −0.013 | 0.982 | 0.160 | 0.778 |
Hyponatremia | −0.002 | 0.679 | −0.003 | 0.521 |
Albumin |
|
|
−0.083 | 0.109 |
Ongoing alcohol abuse | −0.064 | 0.903 | −0.491 | 0.379 |
CLDQ: Chronic Liver Disease Questionnaire.
Results in bold refers to statistically significant predictors of mortality and CLDQ.
The mean CLDQ score was 4.4 (SD 0.7; range: 2.8 to 5.9). Eighteen patients (20%) had low CLDQ scores. The most frequent complaint was fatigue (mean score 3.7; SD 1.0; range 1 to 6 points) and most commonly patients had trouble lifting or carrying heavy objects (Figure
Means with standard deviations for the Chronic Liver Disease Questionnaire domains.
This prospective cohort study showed that one in five outpatients with cirrhosis had a low health-related quality of life. The finding was surprising considering the high proportion of patients without signs of hepatic decompensation. The severity of the underlying liver disease predicted both poor HRQOL and mortality. The nutritional status estimated by the BMI was a predictor of HRQOL, independent of the presence of ascites or low albumin. Patients with a low BMI were more likely to have a low HRQOL.
The relatively small sample size is the main limitation of the present study and may explain why the Child-Pugh score, non-alcoholic etiology, and BMI were the only predictors of quality of life. However, our results concur with studies showing that the Child-Pugh score is associated with the quality of life. Our results concur with studies showing that the Child-Pugh score is associated with the quality of life [
Our results do not allow for a recommendation of the optimal BMI in chronic liver disease, neither does our study offer a causal explanation of the association between lower BMI and low HRQOL. A low BMI may cause low HRQOL, for example, due to decreased resilience. Alternatively, the body weight of patients with a low HRQOL may decrease, for example, due to inadequate food intake. However, as excess body weight has been associated with increased risk of HCC in chronic liver disease and disease progression in alcoholic liver disease, it can be speculated that BMI in patients with cirrhosis should not exceed 25 [
The CLDQ is developed and validated in cohorts of patients with all types of liver diseases [
In agreement with previous studies, 60% of included patients had covert HE [
Although most of the included patients had compensated liver disease at baseline, the prognosis was severe. Twenty-one percent of included patients died. Seven of nineteen deaths occurred in patients who were classed as Child-Pugh group A. Most of these patients had previous signs of decompensated liver disease with ascites, variceal bleeding, spontaneous bacterial peritonitis, or overt HE. Our results support the theory that decompensating events as well as Child-Pugh scores predict long-term prognosis in cirrhosis [
In conclusion, this study found that the prognosis of patients with cirrhosis is severe. The finding suggests that these patients should be followed at outpatient clinics as even patients with a low Child-Pugh class may benefit from regular visits. Additional studies are needed to identify the most efficient management strategies in order to improve the prognosis as well as the quality of life. Interventions directed against malnutrition may help achieve this goal.
Body mass index
Chronic Liver Disease Questionnaire
Continuous reaction time test
Hepatic encephalopathy
Health-related quality of life.
There is no conflict of interests.
Maja Thiele is the gurantator of the paper. Conceptualization: Lise Lotte Gluud, Ole Hamberg, Hans B. Timm. Recruitment: All authors. Data collection: Gro Askgaard, Hans B. Timm, Maja Thiele. Statistical analyses: Maja Thiele, Lise Lotte Gluud. Drafted the paper: Maja Thiele, Lise Lotte Gluud, Gro Askgaard. All authors have approved of the final version of the paper.
The authors would like to thank Mette Munk Lauridsen, M.D., for advice regarding the continuous reaction time test and the specialist nurses Kirsten Passow, Hanne Bennick, Birte Röttig, and Kirsten Larsen for their dedicated work, which facilitated the conduct of the present study. The study was supported by a working grant from the START Fund, Copenhagen University Hospital Gentofte. Results from the study were presented at the 15th ISHEN Symposium (International Society for Hepatic Encephalopathy and Nitrogen Metabolism) and at the 2012 annual meeting of the Danish Society of Gastroenterology and Hepatology.