Statistical analysis

STATA 18 was used to analyse the data. We firstly determined the proportion of all adults that self-reported previously being tested for HCV in the sero-survey dataset, and how long ago their last test occurred. We evaluated the level of concordance between self-reported test results and HCV-Ab test results from the sero-survey among those who reported their previous HCV test result. Binary logistic regression was then used to determine what factors are associated with self-reporting ever being tested for HCV to evaluate whether specific population sub-groups may have been missed by previous testing strategies. First, bivariable regression analysis was used to determine associations between ever being HCV tested and either socio-demographic or various risk factors linked to HCV infection. Following this, multivariable logistic regression was used, only including those factors that had an association in the bivariable analysis (p-value<0.25) based on the p-value cut-off point suggested in the literature that discusses the purposeful variable selection technique [23]. We reported the unadjusted Odds Ratios (OR) and adjusted Odds Ratios (aOR) with 95% confidence intervals (95%CI) and p-value.

Secondly, we identified all adults who have ever been eligible for HCV treatment in the sero-survey dataset, defined as either being currently infected (test HCV-Ab positive and HCV-RNA positive) or having previously been infected (test HCV-Ab positive but HCV-RNA negative) and reporting being treated in the survey. We then determined the HCV-CoC among this group. We estimated the percentages of these treatment-eligible adults who reported previously being tested for HCV, reported testing positive for their previous HCV test (i.e., are diagnosed), reported receiving HCV treatment following their previous positive test result, and the percentage of treated individuals that had an undetectable HCV-RNA test result in the survey (i.e., evidence of successful treatment). People that tested HCV-Ab positive but did not have an HCV-RNA test were not included in this analysis (Fig 1).

Download:

• PNG
  larger image
• TIFF
  original image

Fig 1. Hepatitis C virus cascade of care framework.

https://doi.org/10.1371/journal.pgph.0004706.g001

Lastly, we undertook a sensitivity analysis in which we restricted our analysis to individuals who had attended at least primary education to examine how education may affect our HCV cascade of care results (S1 Text).

Factors associated with reporting ever being Hepatitis C tested

In the bi-variable unadjusted analyses, variables associated with ever HCV testing at p < 0.25 included age, gender, educational status, history of being shaved by a traditional barber, ever had a tattoo/acupuncture, ever received a blood transfusion, ever received an invasive dental procedure, family history of hepatitis, ever received HBV vaccination, ever been hospitalised and receiving a therapeutic injection in the last year (Table 2). These were included in the multivariable model.

Download:

• PNG
  larger image
• TIFF
  original image

Table 2. Factors associated with self-reporting ever being tested for Hepatitis C among adults in the 2019-2020 survey.

https://doi.org/10.1371/journal.pgph.0004706.t002

The multivariable model suggested that males (aOR=1.5, 95%CI:1.0-2.1), adults with secondary or higher educational level (aOR=2.3, 95%CI: 1.6-3.2), adults aged 25–64 years old (aOR=5.1, 95%CI: 3.1-8.6) or 65 + years (aOR=2.6, 95%CI:1.2-5.5) were more likely to report ever being tested for HCV. Additionally, adults who had a family history of hepatitis (aOR=4.7, 95%CI:3.4-6.3), ever had tattoos/acupuncture (aOR=1.5; 95%CI:1.0-2.2), ever received blood transfusions (aOR=2.2, 95%CI:1.5-3.3), had been hospitalised in the last year (aOR=1.8, 95%CI:1.3-2.6), received therapeutic injections in the past year (aOR=1.5, 95%CI: 1.1-2.1), or ever received HBV vaccination (aOR=8.9, 95%CI:5.5-14.4) had a higher likelihood of reporting ever being HCV tested (Table 2).

HCV cascade of care among treatment-eligible adults

Overall, 397 adults (397/3684 or 10.8%, 95%CI: 9.8-11.8%) tested HCV-Ab-positive in the 2019–2020 survey, with 321 (321/397 or 80.9%) having a valid HCV-RNA test result in the survey dataset. There was over 84% concordance between antibody test results and self-reported HCV status (Table A in S1 Text). Of 321 adults with HCV-RNA test results, 232 were defined as treatment eligible, with 204 (204/232 or 87.9%) still having HCV infection (HCV-RNA positive) and 28 (28/232 or 12.1%) no longer having HCV infection (HCV-RNA negative) and reporting previous treatment. The remaining 89 (89/321 or 27.7% of those tested for HCV-RNA) that tested HCV-Ab positive and were HCV-RNA negative reported no previous treatment, and so were assumed to have spontaneously cleared their infection (Table B in S1 Text).

Of the treatment-eligible adults, 95 (95/232 or 40.9%, 95%CI: 34.6-47.6%) reported ever being tested for HCV, with 48 (48/95 or 50.5%) saying they were last tested 1–5 years ago and 16 (16/95 or 16.8%) tested in the last year. Of tested individuals, 87 (87/95 or 91.2%, 95% CI: 84.1- 96.3%) reported their last HCV test was positive, and 60 (60/87 or 69%, 95%CI: 58.1%-78.5%) reported receiving HCV treatment. This equates to 37.5% (87/232: 95%CI: 31.3-44.1%) of treatment eligible adults reporting that they have ever been diagnosed for HCV and 25.9% (60/232; 95%CI: 20.4-32.0%) of treatment eligible adults reporting they have ever received HCV treatment, with 28 (28/60 or 46.7%, 95%CI: 33.7-60.0%) of treated individuals having undetectable HCV-RNA levels (Fig 2).

Download:

• PNG
  larger image
• TIFF
  original image

Fig 2. Hepatitis C Cascade of Care among adults that have ever been eligible for HCV treatment in the 2019-2020 household sero-survey in Sindh province, Pakistan.

We define adults that have ever been eligible for HCV treatment as either being currently infected (test HCV-Ab positive and HCV-RNA positive) or having previously been infected (test HCV-Ab positive but HCV-RNA negative) and self-reporting ever being treated in the survey.

https://doi.org/10.1371/journal.pgph.0004706.g002

Strengths and limitations

This is the first study that we know to evaluate the cascade of care for HCV in Pakistan and the factors associated with HCV testing among adults (≥18 years). The study utilised a large community-based household survey that took a representative sample from urban and rural areas in Sindh Province, enhancing the generalisability of our findings.

Limitations mainly involve the reliance on self-reported data for determining history of HCV testing, diagnosis and treatment history. Although our findings are broadly similar when we consider literate participants (the HCV testing coverage was slightly higher as expected, 13.1% among individuals who attended at least primary education versus 9.3% overall. Regarding the HCV cascade of care, from 62 treatment-eligible individuals who attended at least primary education, 53.2% reported ever being tested vesrsus 41.2% overall, 93.9% reported testing positive versus 91.6% overall, and 74.2% reported receiving HCV treatment versus 69.0% overall), this might affect the accuracy of our estimates, particularly among those who were illiterate (50% of the sample). Importantly, of those that tested HCV-Ab negative, we found that 15.5% reported a previous positive HCV diagnosis, while 13.0% of those testing HCV-Ab positive reported a previous negative diagnosis. Some level of discordance should be expected due to some participants not fully understanding what HCV is and/or getting confused between testing for HCV and testing for other forms of hepatitis (e.g., hepatitis B) or other infections, especially if the test had been done years ago. This concern is supported by there being higher levels of discordant results among survey participants that were illiterate compared to those that have secondary or higher education (19.9% versus 8.1% overall discordance levels), highlighting the importance of improving health education in this group. We also have no information on the accuracy of the tests used among individuals self-reporting previously being HCV tested. The tests may have had poor sensitivity and specificity compared to the SD-Bioline test used in this survey [21,22,24], or the SD-Bioline test may have had some issues of accuracy in some survey settings, possibly due to such things as low temperature [22,25]. It is also possible that some of the observed discrepancies in individuals testing HCV-Ab negative are due to individuals losing their antibody response after undergoing treatment, as observed in Australia following the scale-up in DAA treatment [26]. In support of this, 22 of the 33 individuals with discordant results among adults testing Ab-negative in our survey reported ever having HCV treatment, and so sero-reversion could explain some of the discrepancies. Other studies of the general population [27–29] and individuals infected with HIV [30–32] have also observed this, with levels of discordance varying from 6.3% to 25.8%. It has been hypothesised that full or partial sero-reversion could occur due to loss of antigenic stimulation or low viraemia levels that are insufficient to maintain humoral immune responses [28,33], especially among immunocompromised [34] and treated individuals [31,32,35].

Conversely, for discrepancies among individuals testing Ab-positive, it is possible that the discrepancy could be due to individuals becoming infected since their last reported negative test. Reinfection could also partially explain the low proportion of testing RNA-negative (46.7%) among individuals who reported ever having treatment, although it could also be due to the low SVR from using old interferon-based treatments [12,36]. However, we were unable to confirm this due to a lack of data on treatment compliance or detailed treatment regimens among individuals who self-reported previous HCV treatment.

Lastly, the sampling unit for this survey was households. Because of this, the survey may not be representative of high-risk populations because it will under-sample high-risk populations such as homeless people, people who inject drugs (PWIDs), migrants, and prisoners [37].

Comparison with previous studies

The overall proportion of adults reporting ever being tested for HCV (9.3%) in this study is consistent with what was reported in a household serosurvey undertaken in Punjab in 2018 [38], which estimated that 8.0% of adults in Punjab had ever been diagnosed with Hepatitis B or C. In contrast, the 2017/18 Demographic Health Survey (DHS) for Pakistan suggested a higher rate of ever testing (25.4%) for HBV or HCV among reproductive-aged women [39]. Unfortunately, neither survey distinguished between Hepatitis B and C. The higher rate of testing in the Pakistan DHS may be due to it being undertaken among reproductive-aged women who are likely to have a higher HBV testing rate due to the Pakistan policy of HBV screening among pregnant women [10]. Otherwise, our levels of diagnosis agree with other estimates. We found that 37.5% of adults that have ever been eligible for treatment reported ever being diagnosed with HCV, while 28.9% of adults with chronic HCV infection (test RNA-positive) reported ever being diagnosed in 2019–2020. This is similar to a modelled estimate that 22% of persons living with HCV in Pakistan were diagnosed in 2020 [8].

Conclusion and implications

Our analysis suggests moderate levels of HCV diagnosis and treatment in Sindh province in 2019–2020. Levels of HCV diagnosis and treatment are likely to have improved since then considering the ongoing HCV testing and treatment initiatives being undertaken in Sindh [6], although COVID-19 posed significant challenges to viral hepatitis services in Pakistan [40,41]. Given the increasing prevalence of HCV exposure among adults in the Province [6], there is a need to maintain and scale-up existing HCV diagnosis and treatment services to reduce the expanding HCV epidemic and achieve the WHO targets of diagnosing 80% of HCV infections and treating 65% of HCV infections in Sindh province and Pakistan. Adults with risk factors linked to being HCV infected (e.g., having a family history of hepatitis, blood transfusion, tattoo/acupuncture, hospitalisation, or therapeutic injection history) were more likely to be tested for HCV, which suggests existing HCV testing strategies target individuals with higher levels of risk, as confirmed by the high level of diagnosis among those with chronic HCV infection. The HCV testing rate was higher among individuals who had received HBV vaccination. The provincial HCV guidelines for Punjab, Pakistan, recommend HBV screening and HBV vaccination (if negative) for individuals with HCV infection [42]. This recommendation is also practised in Sindh and is likely to explain the strong associations of HBV vaccination with HCV testing. Education is strongly linked to healthcare access and health literacy in Pakistan, with individuals with higher educational levels having better access to healthcare and improved health literacy [43], thereby facilitating HCV testing. The higher HCV testing coverage among older adults could be explained by previous risk-based screening, given the higher prevalence of HCV infection in this group, as well as the effect of age, which will increase the chance of HCV testing as people grow older.

Women, young adults and individuals with low levels of education had lower levels of HCV testing highlighting that future initiatives need to target these groups, especially illiterate individuals seeing they also have a higher prevalence of HCV infection [6]. Studies suggest that financial incentives could promote HCV testing, so this could be considered in these groups [44]. HCV testing should be improved among women to prevent HCV mother-to-child transmission, which is an important route of HCV transmission for children in Pakistan [45]. The lower HCV testing coverage among women could occur for two reasons. Firstly, women in Pakistan have restricted freedom, and so many women remain at home [46,47]. This could restrict them being tested in community settings, while being tested through outreach activities may also be restricted because it is culturally inappropriate for male health workers to enter the home [48], and so they would not be tested if a female health worker is not available. Secondly, they may refuse testing due to fear of stigma and possible abandonment or divorce if they were found to be infected [49,50]. Antenatal HCV screening could be a good strategy for enhancing HCV testing among women in Pakistan [51], which has high acceptability [52]. Through doing these strategies, Pakistan should be able to achieve HCV elimination as recently achieved by Egypt [53].