6.1. Screening for alcohol use and counselling to reduce moderate and high levels of alcohol intake
Existing recommendation from 2014
An alcohol intake assessment is recommended for all persons with HCV infection followed by the offer of a behavioural alcohol reduction intervention for persons with moderate-to-high alcohol intake.
Strong recommendation, moderate quality of evidence
Note: The WHO Alcohol, Smoking and Substance Involvement Screening Test (ASSIST) (146) screening questionnaire can be used to quantify the level of alcohol intake as low, moderate or high, based on the responses to eight screening questions that assess the frequency of use and presence of alcohol-associated problems.
6.1.1. Background
In many persons with chronic HCV infection, decades can pass between the time they acquire infection and when they develop fibrosis and cirrhosis. During that time, there are health conditions and behaviours that can accelerate the progression of liver damage, including alcohol consumption and obesity. The 2014 Guidelines Development Group assessed various interventions that slow the rate of liver damage among persons with HCV infection and decided to evaluate interventions to reduce alcohol intake because alcohol consumption is common, has been shown to accelerate the progression of liver disease among people with HCV infection (147) and it was felt that persons with HCV infection would be amenable to such measures. Reducing the use of cannabis in persons with HCV infection was discussed by the 2014 Guidelines Development Group but was not considered as part of a systematic review process due to a paucity of data and conflicting reports on any association with progression of liver disease (148).
A heavy intake of alcohol, of between 210 and 560 g/week (a glass of wine or can of beer contains 10–14 g alcohol), doubles the risk of cirrhosis, and even moderate alcohol consumption can be detrimental (149). The purpose of the systematic review was to investigate the effectiveness of behavioural interventions to reduce alcohol intake among people with HCV infection, in terms of HCV infection treatment outcomes, progression of liver disease and quality of life.
Alcohol use in persons with HCV varies considerably in different geographical regions and in different risk groups. Many countries have no published prevalence rates of alcohol use in persons with HCV infection. Some countries, such as Egypt and Saudi Arabia, report extremely low or negligible alcohol use in persons with HCV infection (149, 150). Considerably higher alcohol use is found in other countries, especially among PWID and prisoners. In China, the majority of PWID in one region was found to use alcohol regularly prior to starting injecting drug use (151). In one study from Russia, 26–30% of PWID drank moderate-to-heavy amounts of alcohol (152). In Brazil, HCV-infected youth offenders had high rates of alcohol use (153) and in a study among Nigerian prisoners, 59% with HCV infection also drank alcohol (154). Alcohol intake has also been found to be high in other groups of persons with HCV infection; 37% of male and 9% of female commercial plasma donors infected with HCV in Guan, China were found to drink >40 g of alcohol per day (155). In view of these figures, the 2014 Guidelines Development Group considered that even in countries where alcohol intake is low among the general population, alcohol reduction advice might have an impact.
6.1.2. Evidence
A systematic review was conducted of studies examining a brief behavioural alcohol reduction intervention versus no behavioural intervention for persons with HCV infection. The outcomes considered were reduction or cessation of alcohol intake, SVR, liver fibrosis, decompensated liver cirrhosis, HCC, quality of life and mortality (web Appendix 3, 2014).
Five trials were identified that met the PICO criteria for assessment; two RCTs (156, 157) and three cohort studies (158–160). These studies evaluated different interventions and used different measures of alcohol intake. The interventions that were evaluated included four sessions of motivational enhancement therapy, six two-hour group counselling sessions, 24-week integrated alcohol reduction and health-promotion counselling, and two studies with a single “brief” counselling session. These studies provided some evidence that alcohol reduction interventions can reduce alcohol consumption among people with moderate-to-high alcohol intake living with chronic HCV infection. However, the evidence was graded as being of moderate quality because of considerable heterogeneity in the intervention and comparison groups, and measures of alcohol intake across these studies.
There are more studies evaluating brief alcohol reduction counselling among HCV-uninfected persons. A Cochrane review conducted by Kaner et al. (161) found that among 5860 hazardous or dependent drinkers followed in 22 studies, screening for HCV followed by a brief intervention (compared with no intervention) significantly reduced mean weekly alcohol consumption of 313 g per week by 38 g per week. Klimas et al. (162) investigated the efficacy of psychosocial interventions for drinkers who concurrently used illicit drugs. Among 594 participants across four studies, alcohol-focused interventions resulted in significant reductions in alcohol consumption at 3 months (RR 0.32) and 9 months (RR 0.16) compared to treatment as usual. The quality of the evidence overall was considered to be moderate as there was variability in the type of interventions. Although these studies were conducted among persons without HCV infection, the 2014 Guidelines Development Group felt that the benefits demonstrated in these studies would apply to persons with HCV infection. One limitation is that most of the studies included in these reviews were from North America and Europe; thus, it is uncertain how generalizable they are to other parts of the world.
6.1.3. Rationale for the recommendation
In summary, the 2014 Guidelines Development Group concluded that there was evidence of moderate quality that alcohol reduction interventions would reduce alcohol consumption among persons with chronic HCV infection who consume moderate-to-large amounts of alcohol. Although there are no data on whether longer-term important outcomes, including treatment response, morbidity, mortality and quality of life, are affected by alcohol reduction interventions, the opinion of the Group was that these outcomes are likely to be improved. The 2014 Guidelines Development Group also felt that this intervention would be acceptable to key stakeholders.
Balance of benefits and harms
The evidence in favour of an alcohol reduction intervention was considered to be of moderate quality and the likelihood of undesirable effects minimal. However, the relevance of this advice is likely to be context specific and countries with low alcohol use may not wish to commit as much time and resources to carrying out alcohol reduction interventions as other countries.
Values and preferences
An intervention delivered in the context of a liver health assessment was felt to be acceptable to persons with HCV infection, assuming that confidentiality was maintained. Regarding equity, members of the 2014 Guidelines Development Group felt that alcohol use should not preclude treatment for HCV.
Resource considerations
The principal costs of implementing a brief alcohol reduction intervention were considered to be related to the training of clinicians and counsellors, and the additional time required to deliver counselling. Nevertheless, a brief 5–10 minute alcohol reduction intervention was considered to be unlikely to substantially increase costs and would be likely to be feasible to implement in most health-care settings.
6.1.4. Implementation
An important challenge to implementing a brief alcohol reduction intervention is deciding on which approach to consider. The 2014 Guidelines Development Group proposed that the WHO ASSIST (146) would be an appropriate framework to design alcohol screening and reduction interventions because it is evidence based, proposes a standardized approach, and is aimed at the primary health-care level. The ASSIST package includes tools for carrying out an assessment of the level of intake of alcohol and other substances, and instructions on implementing a brief counselling intervention.
The elements of the ASSIST approach are outlined in and include the administration of a questionnaire regarding the use of alcohol and other substances, classification of the level of consumption and, if needed, alcohol-reduction counselling or referral.
ASSIST – The Alcohol, Smoking and Substance Involvement Screening Test (146).
This approach is more fully described in the WHO Mental Health Gap Action Programme (mhGAP) guidelines for mental health, neurological and substance use disorders in non-specialized settings in LMIC (163).
6.1.5. Research questions
Additional research is required to fully assess the impact of a brief behavioural intervention such as the ASSIST intervention on other outcomes, including morbidity, mortality and quality of life, particularly in different geographical settings. Measuring alcohol consumption is complex and different instruments are used across studies, making comparisons and synthesis of the evidence difficult. Future research should consider using validated and standardized tools for measuring alcohol consumption where possible. Operational research is needed to evaluate approaches that integrate alcohol screening and counselling in different geographical settings.
6.2. Assessing the degree of liver fibrosis and cirrhosis
Existing recommendation from 2014
In resource-limited settings, it is suggested that aminotransferase/platelet ratio index (APRI) or FIB-4 be used for the assessment of hepatic fibrosis rather than other non-invasive tests that require more resources such as elastography or FibroTest.
Conditional recommendation, low quality of evidence
Note: This recommendation was formulated assuming that liver biopsy was not a feasible option. FibroScan®, which is more accurate than APRI and FIB-4, may be preferable in settings where the equipment is available and the cost of the test is not a barrier to testing.
6.2.1. Background
Assessing the degree of liver fibrosis is an important step in the clinical management of persons with HCV infection. Although HCV treatment should be considered for all persons with HCV infection, as indicated in section 10.1, persons with cirrhosis should be prioritized for treatment because they are at increased risk of HCC and death due to liver failure. Furthermore, the selection of treatment regimens can depend on the presence or absence of cirrhosis. Thus, the 2014 Guidelines Development Group felt it important to identify low-cost, effective methods of assessing the degree of fibrosis, which would be widely available in LMIC.
Liver biopsy is considered the gold standard method for fibrosis assessment, but it is not widely used in low-income countries because of its high cost, invasiveness, patient discomfort, risk of complications, as well as the need for expert histological interpretation. Several liver biopsy-scoring systems have been developed, of which the METAVIR system is most widely used ().
METAVIR liver biopsy scoring system (164).
A variety of non-invasive fibrosis tests based on blood indices and imaging modalities are now available, which may be more suitable for LMIC (). These include serum tests such as the aminotransferase/platelet ratio index (APRI), FIB-4 scores, which measure indirect markers of fibrosis such as ALT, aspartate aminotransferase (AST) and platelet count (Fig. 6.1); tests that should be available at all clinics treating patients with HCV infection. Other serum tests such as FibroTest measure direct markers of fibrosis such as haptoglobin. These tests are patented, must be performed in laboratories that meet certain quality standards, and are thus more expensive and less readily available. Not all of these tests can assess all stages of fibrosis as well as cirrhosis. For example, FIB-4 was evaluated only for the diagnosis of significant fibrosis (METAVIR stage ≥F2), while APRI was validated for the diagnosis of both significant fibrosis and cirrhosis. More recently, new techniques have been developed that are based on ultrasound technology and assess the degree of fibrosis and cirrhosis by measuring liver stiffness. Of these, transient elastography, which is performed with FibroScan® (Echosens, Paris) has been the most widely evaluated. Characteristics that limit the use of transient elastography include the high cost of the equipment, the need for regular recalibration, trained operators and the lack of validated cut-off values for specific fibrosis stages.
Selected non-invasive tests to assess liver fibrosis (75).
APRI and FIB-4 formulas. ALT: alanine aminotransferase; AST: aspartate aminotransferase; IU: international unit; ULN: upper limit of normal
6.2.2. Evidence
The PICO question for this recommendation was based on two assumptions. First, that liver biopsy would not be available for the reasons listed above, and second, that all sites would have access to the laboratory tests needed to calculate APRI and FIB-4 indices. Thus, the results of the systematic reviews were analysed to assess the benefit of more complex and expensive tests (e.g. FibroTest or FibroScan®) compared with APRI and FIB-4 (web Appendix 3, 2014). A systematic review was conducted to evaluate the diagnostic accuracy of non-invasive fibrosis assessment tests in adult patients with chronic HCV infection. The systematic review included full papers and abstracts, without language restrictions, which (i) evaluated non-invasive tests for the staging of liver fibrosis using liver biopsy as the reference standard, (ii) reported on the data necessary to calculate the true-positive, false-positive, true-negative and false-negative diagnostic results of the non-invasive tests based on a defined index test cut-off point, and (iii) had a maximum of six months of elapsed time between the liver biopsy and the index test. For data synthesis and analysis, the histological scores used in individual studies were transformed to the METAVIR staging system. Significant fibrosis (METAVIR stage ≥F2) and cirrhosis (F4) were assessed as outcome variables. Overall, the quality of evidence was found to be low, primarily because of potential bias due to the absence of predetermined index test cut-offs for diagnosing specific fibrosis stages, and low or unreported quality of liver biopsy samples. Summary sensitivity and specificity results and relevant confidence intervals are available in web Appendix 3, 2014.
Non-invasive tests provide a numerical value, while histological staging of liver biopsies yields descriptive, semi-quantitative categories. For the non-invasive tests, thresholds exist that correlate with specific histological stages and, in the cases of APRI and FIB-4, these cut-offs have been validated. APRI and FIB-4 have two cut-off values for diagnosing specific fibrosis stages, as the use of a single cut-off would result in suboptimal sensitivity and specificity: a high cut-off with high specificity (i.e. fewer false-positive results) and a low cut-off with high sensitivity (i.e. fewer false-negative results). A staging strategy that uses a combination of these two values uses the low cut-off to rule out the presence of a particular stage of fibrosis and the high cut-off to confirm that the patient has fibrosis that is greater than or equal to a particular stage (e.g. ≥F2). However, a number of patients will fall in the indeterminate range of test results (i.e. their score will be between the low and the high cut-off) and such patients will need either alternative testing or future retesting. Transient elastography uses a single cut-off; however, there are no uniformly established and validated cut-offs for specific fibrosis stages. Therefore, reported sensitivities and specificities of FibroScan® are probably overestimated. The established high and low cut-off values of the APRI and FIB-4 tests along with a range of the most commonly reported cut-offs of FibroScan® for diagnosing ≥F2 stage fibrosis and cirrhosis are presented in . The summary sensitivity and specificity of these tests and FibroScan® for the detection of significant fibrosis (≥F2 stage) and cirrhosis (F4 stage) are listed in .
Low and high cut-off values for the detection of significant fibrosis and cirrhosis.
Summary of sensitivity and specificity of APRI, FIB-4 and FibroScan® for the detection of advanced fibrosis and cirrhosis (all values are percentages).
Having established the sensitivity and specificity of the non-invasive tests compared with liver biopsy as the reference test (), the 2014 Guidelines Development Group considered the comparative performance of the non-invasive tests. For this analysis, APRI and FibroScan® were selected to illustrate clinical trade-offs, as these tests can assess both F2 and F4 cut-offs (i.e. F0–1 vs F2–4; and F0–3 vs F4).
A strategy that uses a combination of the high and low cut-off values was assessed. Using this strategy, patients with values above the APRI high cut-off value would be prioritized for treatment as they have a high probability (94%) of having F4 cirrhosis. For patients with an APRI score below the low cut-off value, treatment could be deferred as they have a very low probability (18%) of having advanced fibrosis (F2 fibrosis or higher) and could thus be reassured and reassessed periodically. Those patients with APRI values between the low and high cut-off values could either be retested every one or two years or, if resources are available, could be treated.
A number of caveats were considered. First, the APRI scoring system may be less reliable in persons with HIV due to the possibility of thrombocytopenia associated with HIV infection rather than cirrhosis. However, HIV-related thrombocytopenia would result in a higher APRI score, and thus earlier treatment. Although this was not assessed in the current analysis, a meta-analysis showed that the diagnostic accuracy of APRI did not significantly differ between HCV-monoinfected and HCV/HIV-coinfected patients (170). Theoretically, the FIB-4 test could also be affected by thrombocytopenia but this scoring system was first evaluated in patients with HIV and was found to perform well (171). Transient elastography values may be artificially increased by a number of factors, including acute liver inflammation, liver congestion (e.g. cardiac failure), a recent meal, amyloidosis and cholestasis. Moreover, the lack of validated cut-offs for the diagnosis of specific stages of fibrosis could hinder the interpretation of the test results.
6.2.3. Rationale for the recommendation
The use of non-invasive monitoring was considered by the 2014 Guidelines Development Group to be preferable to invasive testing, particularly in LMIC, as liver biopsy is an expensive and invasive procedure associated with patient discomfort, a small risk of serious bleeding and requires specialist histological examination for accurate staging. On the basis of the results of the systematic review discussed above, the Group considered that APRI, FIB-4 and transient elastography were the most useful tests for assessing the stage of liver disease. The advantage of APRI as compared with FIB-4 is that it is validated for the diagnosis of F4 fibrosis, and would thus be useful for identifying persons at greatest risk of morbidity who, therefore, could be prioritized for treatment. It was also recommended that persons who tested negative for significant fibrosis and/or cirrhosis could be retested periodically, and could thus be treated if their APRI or FIB-4 indices increased.
Balance of benefits and harms
The principal undesirable outcomes of this recommendation would be due to treatment decisions based on either a false-positive or false-negative APRI or FIB-4 test result. A false-positive test result would lead to a patient being potentially treated earlier than necessary, which would expose him or her to the risk of harm from drug-related side-effects and would also increase resource use. A false-negative result would mean that a person who needs treatment would not receive it, resulting in the possibility that the person would develop cirrhosis or HCC that could potentially have been prevented by treatment for HCV. Despite this, the potential increase in treatment availability resulting from increased access to low-cost, non-invasive monitoring and reduced risk of adverse events from liver biopsy was felt to outweigh the potential harms of false-positive and false-negative case identification.
Values and preferences
APRI and FIB-4 tests require only phlebotomy; thus, the 2014 Guidelines Development Group felt that these tests would be acceptable to patients. Similarly, transient elastography is non-invasive and thus would probably be acceptable.
Resource considerations
The lower cost of the serum-based non-invasive tests was the most important factor that drove the recommendation. The blood tests that are needed to calculate APRI and FIB-4 scores are inexpensive and would be available at health facilities providing treatment for HCV infection, as they are also needed to monitor patients before and after the commencement of treatment. In contrast, the cost of acquiring, running and maintaining a transient elastography machine such as the FibroScan® is very high. The cost of a fixed machine is US$ 100 000 and for a portable one it is US$ 30 000. The cost of yearly maintenance is US$ 4700. For these reasons, the use of transient elastography was considered to be not feasible in most LMIC.
6.2.4. Implementation considerations
The calculation of the APRI score should be easy to implement as it relies on tests that are available in most clinics. Evaluation of the results is more challenging because of the need to assess two cut-off values. However, the above-mentioned strategy provides an approach that should be feasible and will allow clinicians to decide who should be treated. As persons with advanced fibrosis and cirrhosis (METAVIR F3 and F4 stages) are at the highest risk of dying from complications of HCV infection, they need to be prioritized for treatment. If resources allow, treatment of persons with less advanced stages of cirrhosis could be considered.
