All individuals with decompensated liver disease must be assessed and managed in specialist centres. Typical clinical presentations of liver decompensation include variceal haemorrhage, ascites, spontaneous bacterial peritonitis, hepatorenal syndrome, hepatic encephalopathy, hepatopulmonary syndrome and jaundice. All predict a poor prognosis. Multiple scoring systems have been proposed to predict prognosis for people with chronic liver disease, the most well known being the Child–Pugh score (based on degree of ascites, encephalopathy, serum bilirubin level, albumin level and INR) and the MELD score (based on serum bilirubin level, creatinine level and INR) (Supplementary Table 2). These scoring systems have clinical utility for predicting short-term mortality and for prioritising individuals on liver transplant waiting lists.
Liver transplantation provides excellent outcomes for patients with decompensated cirrhosis or early-stage HCC. People who are not referred until they have severe liver failure may not be suitable for transplantation, so early referral is advisable. Consider referring people to a transplant team if they have refractory ascites, an episode of spontaneous bacterial peritonitis or hepatorenal syndrome, recurrent or chronic hepatic encephalopathy, small HCCs or significant malnutrition. Additionally, people should be referred to a transplant team if they are eligible for liver transplantation and have a Child–Pugh score ≥ B7 or MELD score ≥ 13.
Contraindications to liver transplantation may include advanced HCC, extrahepatic malignancy, uncontrolled extrahepatic infection, active alcohol or substance misuse, significant coronary or cerebrovascular disease or inadequate social support. For more information about liver transplantation, see the Transplantation Society of Australia and New Zealand guidelines. 
In people with decompensated liver disease, the goal of therapy is SVR, with the aim of improving liver function. The first regimen to be specifically listed on the PBS for treatment of decompensated liver disease was sofosbuvir plus velpatasvir plus ribavirin. The eligibility criteria for other DAA regimens that are PBS-listed for the treatment of HCV do not distinguish between people with compensated versus decompensated liver disease, with the exception of regimens that include a protease inhibitor in the setting of hepatic decompensation (glecaprevir plus pibrentasvir is contraindicated and sofosbuvir plus velpatasvir plus voxilaprevir is not recommended for people with Child–Pugh B or C disease) (Table 5).
The efficacy of several DAA regimens in people with decompensated liver disease has been formally evaluated in clinical trials. [70-76]
Data from the ASTRAL-4 study support the combination of sofosbuvir plus velpatasvir plus ribavirin for 12 weeks as a first-line treatment for patients with HCV and decompensated liver disease. In this study, 267 patients with Gt 1, 2, 3, 4 or 6 HCV and decompensated cirrhosis (90% Child–Pugh class B or C) were randomly assigned to treatment with sofosbuvir plus velpatasvir for 12 weeks, or sofosbuvir plus velpatasvir plus ribavirin (daily, according to body weight: < 75 kg, 1000 mg; ≥75 kg, 1200 mg) for 12 weeks, or sofosbuvir plus velpatasvir for 24 weeks. SVR was 94% in people treated with sofosbuvir plus velpatasvir plus ribavirin for 12 weeks, versus 83% with sofosbuvir plus velpatasvir for 12 weeks, versus 86% with sofosbuvir plus velpatasvir for 24 weeks. Post-treatment virological relapse was observed in 2% of the 12-week group receiving sofosbuvir plus velpatasvir plus ribavirin, compared with 12% and 9%, respectively, in the groups that did not receive ribavirin. Although the ASTRAL-4 study was not powered to generate statistical significance, the data suggest that sofosbuvir plus velpatasvir plus ribavirin for 12 weeks is the optimal regimen for patients who will tolerate ribavirin. For patients in whom there is a concern about ribavirin intolerance, we recommend a starting dose of 600 mg daily, or treatment for 24 weeks without ribavirin. Important exclusion criteria for the ASTRAL-4 study included Child–Pugh score > C9, haemoglobin level < 100 g/L, platelet count ≤ 30 000/mm3, bilirubin level > 85.5 µmol/L and creatinine clearance < 50 mL/min.
There are very limited clinical data available to support treatment recommendations for patients with Gt 2, 4, 5 or 6 HCV infection and decompensated liver disease, which are based on expert opinion. As for patients with Gt 1 or 3 HCV, we recommend treatment with sofosbuvir plus velpatasvir plus ribavirin for 12 weeks.
People with decompensated liver disease should not be treated with regimens that include the HCV protease inhibitors glecaprevir (contraindicated in Child–Pugh B or C disease) or voxilaprevir (not recommended in Child–Pugh B or C disease), as there is a risk of causing further deterioration in liver function.
Early data based on short-term follow-up indicate that SVR may lead to improvement of liver function in some, but not all, people. The severity of baseline liver disease appears to determine the likelihood of clinical improvement. The presence of ascites or encephalopathy, serum albumin level < 35 g/L, serum ALT level < 60 U/L and body mass index > 25 kg/m2 are all associated with an increased risk of not achieving a reduction in Child–Pugh score to class A.  Three distinct groups are emerging:
i) people with a MELD score < 15 and Child–Pugh score B;
ii) those with a MELD score of 15–20 or Child–Pugh C cirrhosis; and
iii) those with a MELD score > 20.
People with a MELD score < 15 and Child–Pugh B cirrhosis are most likely to benefit from HCV cure and should start treatment immediately. In people with a MELD score of 15–20, or Child–Pugh C cirrhosis, liver function may improve with achievement of SVR, and some people may even be delisted for liver transplantation. However, predictive factors are yet to be determined and it must be noted that improvement in MELD score may result in prolonging the waiting time for transplantation in those who do not improve sufficiently to be delisted. Until predictive factors can be identified, it appears reasonable to treat and closely monitor the progress of patients on the liver transplant waiting list with MELD scores of 15–20. Longer term assessment of clinical outcomes after SVR in this population are needed to determine the impact on liver synthetic function, portal hypertension and HCC risk.
People with a MELD score > 20 are unlikely to benefit sufficiently from SVR to be delisted. [75,79] Antiviral therapy may be started with the intent of suppression and prevention of post-transplant HCV recurrence (see Preventing recurrent HCV after transplantation: treatment of people on the transplant waiting list). Alternatively, these individuals may be best served with HCV treatment after transplantation.
DAA therapy after liver transplantation results in higher SVR rates than in the pre-transplant population with decompensated liver disease (see Treatment of HCV and decompensated liver disease after transplantation), which minimises the risk of selecting for drug-resistant variants. Finally, among people who are not candidates for liver transplantation, it is reasonable to consider DAA therapy regardless of MELD score.
Note that ribavirin can cause adverse events, including anaemia, rash, cough, dyspnoea, insomnia and anxiety. Anaemia is more common in patients with decompensated liver disease, and it is recommended that ribavirin be started at a low dose of 600 mg daily for these patients. Ribavirin is renally excreted, and dose adjustment is required according to eGFR (see Special populations: treatment of HCV in people with renal impairment). Patients with renal impairment have increased risk of anaemia during ribavirin therapy. Monitoring of haemoglobin levels is recommended every 2-4 weeks during ribavirin therapy in people with decompensated liver disease.
As ribavirin is teratogenic, both women and men should be counselled about the risks of pregnancy and advised that two forms of contraception are recommended while taking ribavirin and for 6 months after treatment.