Advertisement
Journal Home
Search for

Volume 56, Issue 1, Pages 88-93 (January 2003)


View previous. 12 of 14 View next.

Long-term survival and cause-specific mortality in patients with cirrhosis of the liver: a nationwide cohort study in Denmark

Henrik Toft SørensenCorresponding Author Informationabemail address, Ane Marie Thulstrupab, Lene Mellemkjarc, Peter Jepsena, Erik Christensend, Jørgen H Olsenc, Hendrik Vilstrupa

Received 22 March 2002; received in revised form 4 September 2002; accepted 18 September 2002.

Abstract 

Mortality from cirrhosis of the liver has been examined in few long-term follow-up studies. In the Danish National Registry of Patients, 1982–1989, we identified a cohort of 10,154 patients with liver cirrhosis and divided them according to the etiology of their liver disease. Causes of death were identified in the Danish Death Registry, 1982–1993. We estimated relative survival and standardized mortality ratios by comparing with the mortality in the general population. The 10-year relative survival was worse in patients with alcoholic cirrhosis (34%) or nonspecified cirrhosis (32%) than in patients with primary biliary cirrhosis (58%) or chronic hepatitis (66%). The standardized mortality ratio for all causes of death combined was 12-fold increased, 5-fold excluding cirrhosis-related causes. Mortality in all disease categories was increased, even in those not traditionally related to cirrhosis. In conclusion, patients with cirrhosis of the liver face reduced life expectancy due to several causes of death.

Article Outline

Abstract

1. Introduction

2. Materials and methods

2.1. Statistical analysis

3. Results

3.1. Survival

3.2. Causes of death

4. Discussion

Acknowledgment

References

Copyright

1. Introduction 

return to Article Outline

Liver cirrhosis and its complications are major clinical problems that carry a considerable risk of disability and death. There is no curative treatment for liver cirrhosis, so prevention and treatment of complications such as variceal hemorrhage, ascites, and encephalopathy are keystones for the clinical care of these patients. Because the liver plays an important role for many vital functions (eg, immunity, metabolism, and coagulation), the range of cirrhotic complications is broad. Consequently, understanding the disease process, making appropriate risk stratification, and implementing tertiary prevention requires valid and precise understanding of outcomes such as mortality patterns. The short-term prognosis of liver cirrhosis is relatively well described within the framework of randomized controlled trials, but population-based reports of long-term follow-up in causes of death are few and limited in size 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13.

To clarify the clinical course of patients with liver cirrhosis, we determined the survival and causes of death of a large cohort of such patients, using population-based data from the Danish National Registry of Patients (NRP) and the Danish Death Registry.

2. Materials and methods 

return to Article Outline

The study cohort was identified in the NRP, which contains information on all hospital admissions in Denmark (population 5.2 million) since 1977 [14]. Each admission record includes a date of admission, a date of discharge, up to 20 discharge diagnoses, and a civil registry number assigned to all Danes at birth, incorporating date of birth and sex. Diagnoses were classified according to the Danish version of the International Classification of Diseases, 8th edition (ICD-8) during the study period [14]. Patients were eligible for the study if they had been discharged with a diagnosis of alcoholic cirrhosis (ICD-8 = 571.09), primary biliary cirrhosis (571.90), nonspecified cirrhosis (571.92), chronic hepatitis (571.93), or “other types of cirrhosis, alcoholism not indicated” (571.99). Including all cases thus identified would have led to the inclusion of prevalent and incident cases because the patients could have been hospitalized with liver cirrhosis before the registry was established in 1977 [15]. To focus on incident cases, we restricted the cohort to patients whose first discharge diagnosis of liver cirrhosis was between 1982 and 1989. Patients younger than 15 years of age were excluded because most liver diseases in childhood are due to congenital malformations [16].

The cohort of patients with cirrhosis was considered as a whole and also as four sub-cohorts. The sub-cohorts were defined according to ICD-8 discharge diagnosis codes for alcoholic cirrhosis (571.09), primary biliary cirrhosis (571.90), chronic hepatitis (571.93), and nonspecified cirrhosis (571.92) plus cirrhosis, alcoholism not indicated (571.99) considered together as nonspecified cirrhosis. From 1977 to 1989, 12% of the cohort was diagnosed with more than one type of cirrhosis. These patients were assigned to a sub-cohort based on the following hierarchy: (1) alcoholic cirrhosis, (2) primary biliary cirrhosis, (3) chronic hepatitis, and (4) nonspecified cirrhosis. In addition, cirrhotic patients with a diagnosis of alcoholism (ICD-8 = 303) at any time from 1977 to 1989 were categorized as having alcoholic liver cirrhosis, regardless of the type of cirrhosis registered. On this background, we transferred 16% from the cohorts of primary biliary cirrhosis, chronic hepatitis, and nonspecified cirrhosis to the sub-cohort of alcoholic liver cirrhosis.

We obtained information on the vital status of cohort members from the Danish Civil Registration System, which, since 1968, has kept electronic records of all changes in vital status, such as emigration and death. Patients who died during the first hospital admission were included in the study. The cohort from the NRP was linked to the National Registry of Death (NRD) by the civil registry number. The NRD contains information on all death certificates since 1943, also coded in ICD-8 during the study period. We used the underlying cause of death reported on the death certificate in the analysis. A grouping into 50 categories based on ICD-8 as defined by the Danish National Board of Health [17] was applied in the main analyses. The different groups of death causes were merged into larger groups in the analysis of the sub-cohorts of liver cirrhosis. Cirrhosis-related death consisted of liver cirrhosis (ICD-8, 571), esophageal varices (ICD-8, 456), and hepatic coma (ICD-8, 473).

2.1. Statistical analysis 

The number of person-years under observation was calculated from the date of first discharge with a diagnosis of liver cirrhosis until death, emigration, or the end of follow-up on 31 December 1993, whichever occurred first.

The relative survival was calculated as the ratio of the observed survival proportion to that expected for persons of the same age and sex in the general population. The expected survival proportions after every full year of follow-up time were computed from national mortality rates and were subsequently used to calculate a continuous expected rate [18]. We reported the 10-year relative survival because only a few patients were followed for all 12 years.

National sex-specific mortality rates in 5-year age groups and 5-year calendar periods were multiplied by person-years under observation to yield the number of deaths that would be expected if patients with liver cirrhosis had the same risk of death as the general population. The standardized mortality ratio (SMR) is the ratio of observed to expected number of deaths, based on the assumption that the observed number of cases in a specific category follows a Poisson distribution. We reported the SMR after 12 years of follow-up, and this was also divided into a 1-year SMR and a 2- to 12-year SMR. Confidence intervals (95% CI) for SMRs were calculated from an accurate asymptomatic approximation; exact limits were used when the observed number was less than 10 [19].

3. Results 

return to Article Outline

We identified 10,154 patients with liver cirrhosis during the study period; 6374 (63%) were men, and 3780 (37%) were women. The mean age at entry into the study cohort was 56.9 years. The median and mean follow-up were 2.8 years and 3.7 years, respectively; the total number of person-years at risk was 37.382 years. Alcoholic cirrhosis was the largest of the four sub-cohorts (61% of all patients), whereas primary biliary cirrhosis accounted for only 2%. The sub-cohorts of primary biliary cirrhosis and chronic hepatitis had the longest mean follow-up period. In the alcoholic cirrhosis group, 71% of the patients were men, whereas 80% of patients with primary biliary cirrhosis were women (Table 1).

Table 1.

Characteristics of 10,154 patients with different types of liver cirrhosis in Denmark, 1982–1989.

Type of liver cirrhosis
Patient characteristicAlcoholicPBCChronic hepatitisNonspecifiedTotal
Number of patients6139223992280010,154
Proportion of total, %60.52.29.827.6100
Male:female ratio4353:178645:178455:5371521:12796374:3780
Mean age at entry into the study, y56.360.152.259.656.9
Mean follow-up time, y3.85.35.82.53.7
Total number of person-years at risk23,51811865780689837,382
Number of deaths in the follow-up period423010440422416979

Abbreviation: PBC = primary biliary cirrhosis.

3.1. Survival 

Of the 10,154 patients, 3863 (38%) died within the first year of follow-up; 56% died of a cirrhosis-related disease (eg, liver cirrhosis, esophageal varices, or hepatic coma), and 44% died of other causes. Beyond the first year of follow-up, other causes of death became relatively more frequent than death from liver cirrhosis (Fig. 1). During the 12-year follow-up period, 6979 (69%) died; 51% died from a cirrhosis-related disease, and 49% died from other causes.


View full-size image.

Fig. 1. Deaths in patients (n = 6979) with liver cirrhosis, grouped in three follow-up periods according to cirrhosis-related death and other causes of death.


The 10-year relative survival was 0.34 (95% CI 0.32–0.36) for alcoholic cirrhosis and 0.32 (95% CI 0.28–0.35) for nonspecified cirrhosis. Primary biliary cirrhosis and chronic hepatitis had a much better prognosis: a 10-year relative survival of 0.59 (95% CI 0.47–0.71) and 0.66 (95% CI 0.61–0.71), respectively (Figure 2).


View full-size image.

Fig. 2. Relative survival curves for 10,154 patients with liver cirrhosis, grouped in four sub-cohorts during a 10-year follow-up period.


The poorest prognosis was found in patients aged 60 years or more at diagnosis, with a relative 1-year survival rate of 0.67 (95% CI 0.66–0.69) and a 10-year relative survival rate of 0.25 (95% CI 0.22–0.28).

3.2. Causes of death 

The mortality in cirrhotic patients for noncirrhotic causes of death was increased 5-fold compared with the general population. Mortality was increased for all disease categories, including for accidents and suicides. There was a 5- to 22-fold increased risk of dying from various infectious diseases and an 8- to 14-fold increased risk of dying from disorders of the digestive system. Nonetheless, ischemic heart diseases contributed the largest excess of deaths (Table 2). There were 398 deaths from gastrointestinal cancer, of which 202 were deaths from primary liver cancer. The overall risk of mortality from gastrointestinal cancer was more than 8-fold increased (SMR 8.3, 95% CI 7.5–9.2) compared with the general population. The increase in overall mortality was most pronounced during the first year of follow-up.

Table 2.

Observed (OBS) and expected (EXP) number of deaths and cause-specific standardized mortality ratio (SMR) for all patients with liver cirrhosis during the 12-year, 1-year, 2- to 12-year follow-up*

*

EXP and SMR for cirrhosis-related death are not reported because a liver cirrhosis-related death requires a diagnosis of liver cirrhosis.

Mortality varied according to the underlying liver disease but was increased for all types of cirrhosis. There was a 5-fold increased mortality in alcoholic cirrhosis and a 7-fold increased mortality in nonalcoholic cirrhosis after subtracting the cirrhosis-related deaths; mortality was increased 2- to 3-fold in primary biliary cirrhosis and chronic hepatitis. The risk of dying from infections was most pronounced for nonspecified cirrhosis, chronic hepatitis, and alcoholic cirrhosis, but heart disease mortality was increased for all four sub-cohorts. A significantly increased risk of death from suicide was found among patients with alcoholic and nonspecified cirrhosis (Table 3).

Table 3.

Observed (OBS) and expected (EXP) number of deaths and cause-specific standardized mortality ratio (SMR) for categories of causes of death according to type of liver cirrhosis during the 12 year follow-up period

Alcoholic cirrhosisPrimary biliary cirrhosisChronic hepatitisNonspecified cirrhosis
Causes of deathOBSSMR95% CIOBSSMR95% CIOBSSMR95% CIOBSSMR95% CI
All causes423012.311.9–12.71043.83.1–4.74044.44.0–4.9224111.110.7–11.6
All causes without cirrhosis-related disease16755.04.8–5.2592.21.7–2.82903.22.9–3.614197.16.8–7.5
Heart disease3013.12.7–3.4192.41.5–3.8421.61.2–2.24396.76.1–7.4
Vascular disease1263.53.0–4.282.10.9–4.1231.91.2–2.81485.14.3–6.0
Infections678.46.5–10.411.10.1–6.23912.68.9–17.2598.36.3–10.7
Respiratory disease985.44.4–6.521.50.2–5.6112.41.2–4.4706.95.4–8.8
Digestive system disease8314.611.6–18.135.01.0–14.6157.94.4–13.06716.012.4–20.3
Endocrine disorders436.14.5–8.30 105.72.7–10.25214.410.8–18.9
Renal disease1813.88.2–21.9110.00.3–55.737.51.6–21.91111.05.5–19.7
All cancers5654.93.9–4.7192.71.6–4.2773.02.4–3.84188.88.0–9.7
Mental disorders11029.524.2–35.50 54.91.6–11.5135.93.2–10.1
Other diseases3898.48.2–9.891.70.8–9.5977.86.2–9.52718.17.2–9.2
Accidents10210.58.6–12.822.80.3–10.3258.05.2–11.9376.24.3–8.5
Suicide575.14.2–7.10 72.10.8–4.293.11.4–5.9

4. Discussion 

return to Article Outline

This analysis of more than 10,000 patients with liver cirrhosis and complete nationwide long-term follow-up showed that patients with liver cirrhosis have a high mortality, even for causes of death not usually associated with chronic liver disease. Survival was particularly poor within the first year after the initial hospitalization for cirrhosis and in patients with alcoholic and nonspecified cirrhosis.

Our study has important strengths and limitations. We used nationwide, population-based registries from a uniformly organized health care system with complete long-term follow-up data. This large population-based study was well defined and had a considerable advantage in statistical precision over previous long-term follow-up studies [5]. Furthermore, unlike most previous investigators, we were able to adjust for the expected survival in the general population [5]. The limitations of the study included lack of clinical details in the registries, the relatively broad classification of cirrhosis, and lack of data on lifestyle factors that may have contributed to the poor prognosis. This lack of clinical details prevented us from further subdivision of the sub-cohorts.

An additional limitation of our investigation is the use of administrative discharge data. It is well known that diagnoses at discharge are not entirely accurate [20], and liver cirrhosis may have been misclassified in 5% to 10% of the cases listed in the NRP [21]. This lack of specificity may have led us to underestimate the mortality associated with liver cirrhosis because most of the misclassified cases were likely to have had a more benign liver disease than cirrhosis. It is well known that the diagnosis on the death certificate might not be entirely correct [22]. It is thus possible that doctors who filled in the certificates might be biased because they know that the patients have liver cirrhosis. Therefore, liver cirrhosis as cause of death has possibly been overestimated compared with other causes.

An important clinical finding of our study was that the mortality of patients with liver cirrhosis was very high in the first year of follow-up, after which it declined but remained substantially higher than that of the general population. This was true for deaths from cirrhosis-related causes and for death from other causes. This demonstrates that patients with liver cirrhosis are seriously ill at the time of diagnosis and that immediate intensive care is crucial. Death from variceal bleeding is likely to be responsible for a large part of the increased immediate mortality from cirrhosis-related causes, and Sharara et al have proposed a classification of liver cirrhosis according to the presence of varices [23]. Concerning death from other causes, the increased immediate mortality shows the liver's importance for normal function of other organs and the immune system.

Our findings were consistent with Saunder's 20-year-old study of 512 patients in Birmingham, UK [5]. It showed a 1-year survival rate of 40% to 80% depending on the type of cirrhosis, patients with cryptogenic cirrhosis having the worst prognosis and those with chronic active hepatitis having the best [5]. Our data are not completely new due to lack of updating of the Danish Registry of Death. There are some indications that during the last years the outcome of esophageal varices may have been slightly improved 24, 25.

It is well known that the type of liver cirrhosis affects short-term prognosis [26]. This is not only because of differences in the clinical course, but also because of differences in lifestyle and risk factors for liver cirrhosis. Although we do not have the data to examine this, it is likely that some of the excess mortality, particularly the pattern of excess mortality from cardiovascular and respiratory diseases, diabetes, accidents, and suicide, is explained by smoking, alcohol, poor nutrition, and psychiatric diseases. The limitations of our data and the observational nature of our study prevent us from giving detailed guidelines to the clinical care of patients with liver cirrhosis. Probably a multi-disciplinary approach is needed. However, few data exist on the benefit of preventive care of patients with liver cirrhosis [27]. Screening for varices and treatment with nonselective beta-blockers reduce the risk of first variceal bleeding [25]. Liver patients are at increased risk of liver and other types of cancer [28]. However, screening for hepatocellular cancer with AFP and ultrasound and prophylaxis against spontaneous bacterial peritonitis remain controversial 29, 30.

In conclusion, our data showed that the prognosis of patients with liver cirrhosis is poor. There was a substantial excess mortality in all types of cirrhosis due to the underlying liver disease and lifestyle factors. Intensive multi-organ care for patients with newly diagnosed liver cirrhosis is crucial.

Acknowledgements 

return to Article Outline

This study received financial support from the Western Danish Research Forum for Health Sciences (Vestdansk Forskningsforum).

References 

return to Article Outline

1. 1 Sørensen TI, Orholm M, Bentsen KD, et al.  Prospective evaluation of alcohol abuse and alcoholic liver injury in men as predictors of development of cirrhosis. Lancet. 1984;2:241–244. MEDLINE

2. 2 Corrao G, Ferrari P, Zambon A, et al.  Trends of liver cirrhosis mortality in Europe, 1970–1989 (age-period- cohort analysis and changing alcohol consumption). Int J Epidemiol. 1997;26:100–109. MEDLINE | CrossRef

3. 3 D'Amico G, Morabito A, Pagliaro L, et al.  Survival and prognostic indicators in compensated and decompensated cirrhosis. Dig Dis Sci. 1986;31:468–475. MEDLINE | CrossRef

4. 4 Gines P, Quintero E, Arroyo V, et al.  Compensated cirrhosis (natural history and prognostic factors). Hepatology. 1987;7:122–128. MEDLINE | CrossRef

5. 5 Saunders JB, Walters JR, Davies AP, et al.  A 20-year prospective study of cirrhosis. Br Med J. 1981;282:263–266.

6. 6 van Dam GM, Gips CH. Primary biliary cirrhosis in The Netherlands (an analysis of associated diseases, cardiovascular risk, and malignancies on the basis of mortality figures). Scand J Gastroenterol. 1997;32:77–83. MEDLINE | CrossRef

7. 7 Bouchier IA, Hislop WS, Prescott RJ. A prospective study of alcoholic liver disease and mortality. J Hepatol. 1992;16:290–297. MEDLINE | CrossRef

8. 8 Schlichting P, Christensen E, Fauerholdt L, et al.  Main causes of death in cirrhosis. Scand J Gastroenterol. 1983;18:881–888. MEDLINE

9. 9 Nevens F, Broeckaert L, Rutgeerts P, et al.  The long-term morbidity and mortality rate in a cohort of patients with liver cirrhosis and oesophageal varices. Hepatogastroenterology. 1995;42:979–984. MEDLINE

10. 10 Schlichting P, Christensen E, Andersen PK, et al.  Prognostic factors in cirrhosis identified by Cox's regression model. Hepatology. 1983;3:889–895. MEDLINE

11. 11 Propst A, Propst T, Zangerl G, et al.  Prognosis and life expectancy in chronic liver disease. Dig Dis Sci. 1995;40:1805–1815. MEDLINE | CrossRef

12. 12 Murtaugh PA, Dickson ER, van Dam GM, et al.  Primary biliary cirrhosis (prediction of short-term survival based on repeated patient visits). Hepatology. 1994;20:126–134. MEDLINE | CrossRef

13. 13 Powell LW, Mortimer R, Harris OD. Cirrhosis of the liver (a comparative study of the four major aetiological groups). Med J Aust. 1971;1:941–950.

14. 14 Andersen TF, Madsen M, Jørgensen J, et al.  The Danish National Hospital Register (a valuable source of data for modern health sciences). Dan Med Bull. 1999;46:263–268. MEDLINE

15. 15 Almdal TP, Sørensen TI. Incidence of parenchymal liver diseases in Denmark, 1981 to 1985 (analysis of hospitalization registry data. The Danish Association for the Study of the Liver). Hepatology. 1991;13:650–655. MEDLINE | CrossRef

16. 16 Bircher J, Benhamou P, McIntyre N, et al.  Oxford Textbook of Clinical Hepatology. 2nd edition. Oxford: Oxford University Press;; 1999;.

17. 17 Juel K, Helweg-Larsen K. The Danish registers of causes of death. Dan Med Bull. 1999;46:354–357. MEDLINE

18. 18 Andersen PK, Væth M. Statistisk analyse af overlevelsesdata ved lægevidenskabelige undersøgelser. 3rd edition. Copenhagen, Denmark: Fadls Forlag;; 1993;.

19. 19 Rothman KJ, Boice JD. Epidemiologic analyses with a programmable calculator. Washington DC: US Government Printing Office;; 1979;.

20. 20 Sørensen HT. Regional administrative health registers as a resource in clinical epidemiology. Int J Risk Saf Med. 1997;10:1–22.

21. 21 Vestberg K, Thulstrup AM, Sørensen HT, et al.  Data quality of administratively collected hospital discharge data for liver cirrhosis epidemiology. J Med Syst. 1997;21:11–20. MEDLINE | CrossRef

22. 22 Mabeck CE, Wichmann B. Causes of death and death certificates (an evaluation of the diagnosis in 373 death certificates). Ugeskr Laeger. 1980;142:257–261. MEDLINE

23. 23 Sharara AI, Rockey DC. Gastroesophageal variceal hemorrhage. N Engl J Med. 2001;345:669–681. MEDLINE | CrossRef

24. 24 Kleber G, Ansari H, Sauerbruch T. Prophylaxis of first variceal bleeding. Baillieres Clin Gastroenterol. 1992;6:563–580. MEDLINE | CrossRef

25. 25 De Franchis R. Prediction of first variceal hemorrhage in patients with cirrhosis of the liver and esophageal varices. N Engl J Med. 1988;319:983–989. MEDLINE

26. 26 Christensen E. Prognostic models in chronic liver disease (validity, usefulness and future role). J Hepatol. 1997;26:1414–1424. Full-Text PDF (1199 KB) | CrossRef

27. 27 Riley TR, Smith JP. Preventive care in chronic liver disease. J Gen Intern Med. 1999;14:699–704. MEDLINE | CrossRef

28. 28 Sørensen HT, Friis S, Olsen JH, et al.  Risk of liver and other types of cancer in patients with cirrhosis (a nationwide cohort study in Denmark). Hepatology. 1998;28:921–925. MEDLINE | CrossRef

29. 29 Maringhini A, Cottone M, Sciarrino E, et al.  Ultrasonography and alpha-fetoprotein in diagnosis of hepatocellular carcinoma in cirrhosis. Dig Dis Sci. 1988;33:47–51. MEDLINE | CrossRef

30. 30 Gines P, Rimola A, Planas R, et al.  Norfloxacin prevents spontaneous bacterial peritonitis recurrence in cirrhosis (results of a double-blind, placebo-controlled trial). Hepatology. 1990;12:716–724. MEDLINE | CrossRef

a Department of Medicine V (Hepatology and Gastroenterology), Aarhus University Hospital, Nørrebrogade 44, DK-8000 Aarhus C, Denmark

b Department of Clinical Epidemiology, Aarhus University Hospital, Vennelyst Boulevard 6, and Aalborg Hospital, Stengade 10, Aarhus, Denmark

c The Danish Cancer Society, Institute of Cancer Epidemiology, Strandboulevarden 49, Copenhagen, Denmark

d Department of Internal Medicine, Bispebjerg University Hospital, Bispebjerg Bakke 23, Copenhagen, Denmark

Corresponding Author InformationCorresponding author. Department of Clinical Epidemiology, Aarhus University Hospital, Vennelyst Boulevard 6, Building 260, DK-8000 Aarhus C, Denmark. Tel.: + 45 89 42 60 77; fax: + 45 89 42 61 10.

PII: S0895-4356(02)00531-0


View previous. 12 of 14 View next.