Strategies for the prevention of perinatal hepatitis B transmission in a marginalized population on the Thailand-Myanmar border: a cost-effectiveness analysis

An estimated 240 million individuals are chronically infected with hepatitis B virus (HBV) worldwide [1, 2]. An estimated 686,000 people die globally due to complications of hepatitis B, including liver cirrhosis and hepatocellular carcinoma, which cause approximately 90% of deaths [3]. Transmission occurs through unprotected sex and body fluids [4], including the birth process, which leads to the majority of perinatal infections. Individuals who are HBV e-antigen positive (HBeAg+) are at the highest risk of transmitting to others. Of those who acquire the infection perinatally, 65% become chronic carriers of the disease and therefore are at a higher risk of morbidity and mortality as compared to 28% in those born to HBeAg- mothers [5]. An estimated 5–15% of perinatal infections occur in-utero, which vaccination at birth does not prevent [6]; however, an estimated 90% of perinatal infections could be averted through infant HBV vaccination alongside administration of HBV immunoglobulin (HBIG) to the infant [7‐12]. Due to problems of production, storage (cold chain) and cost, most low resource settings do not have access to HBIG. The reduction of mother to child transmission is a core intervention in the World Health Organisation (WHO) global health sector strategy on viral hepatitis, which proposes expansion of coverage from current its level of 38% to 90% by 2030 [13].

The HBV vaccine is the only routine childhood vaccine where prompt administration after birth (within 24 h) is required to prevent transmission [14‐16]. A modelling study [17] found that providing the first dose of HBV vaccine at birth could prevent an additional 16% of deaths from the disease, and a more recent study estimated that 210 million chronic infections had been prevented by 2015 through the vaccination of infants and neonates [18]. Universal HBV vaccination with a three to four dose schedule has a protective efficacy of 70–85% [19], which increases up to 95% when administered with HBIG [20]. HBV vaccination has been shown to be cost effective [21‐24], and nearly all countries have adopted the WHO recommendation of universal HBV vaccination without screening. In high resource settings, universal vaccination alongside antenatal maternal screening with HBIG administered in hospital to newborns of those who screen HBV surface antigen positive (HBsAg+) is considered to be a cost-effective strategy [7, 25, 26]. These screening strategies are often not taken up by low income countries due to the high costs [1]; yet it is often these low income countries that carry higher burdens of the disease. A lack of information exists on the cost-effectiveness of this strategy in resource limited settings, where late antenatal care is more common and the proportion of births at home is higher.

The Asia-Pacific region disproportionately shares the burden of HBV, with 75% of chronic HBV carriers in the world found in Asia [27]. Vaccine programmes in Lao PDR, Vietnam and Cambodia have successfully lowered the prevalence of chronic HBV carriage to less than 2% by 2012 [28]. Myanmar, however, has continued to have areas of high HBV prevalence, especially in rural and border regions of Myanmar where a prevalence of 8.3% was recently reported [29]. This is widely attributed to the high percentage HBeAg + women of reproductive age in this region.

The aim of this study was to evaluate the cost-effectiveness of options for the prevention of the perinatal HBV transmission in a marginalized (refugee and migrant) population on the Thailand-Myanmar border.

Table 3 presents the results for the base case analysis. The cohort costs ranged from US$21,673.15 for the vaccine only strategy to US$47,477.10 for the HBIG after RDT strategy; these are equivalent to US$4.33 to US$9.50 per woman presenting at the clinic. The vaccine only had the lowest effectiveness with 64 perinatal infections in the hypothetical cohort of 5000 newborns. The number of infections was reduced to 41 with the HBIG after confirmatory test and 28 with the HBIG after RDT strategy. The HBIG after confirmatory test strategy were removed due to extended dominance by the HBIG after RDT strategy, which had an ICER of US$716.78 per infection averted, which was below the willingness to pay threshold of US$1200. The PSA produced a 95% CrI of US$343.00–2159.90 for the ICER. Figure 1 shows the cost-effectiveness plane.

Figure 2 shows the results of the one way sensitivity analysis on the ICER for HBIG after RDT strategy as compared to the vaccine only strategy. The three parameters that had the largest impact on the ICER of the HBIG after RDT strategy were all related to transmission: the probability of HBV perinatal infection for both HBeAg + and HBeAg- mothers when newborns were given vaccine and the probability of HBV perinatal infection for HBeAg + mothers if newborns were given vaccine and HBIG. A 50% change in the cost of HBIG also caused the ICER to change by nearly 40% in both directions with a low cost causing a reduction in the ICER. When the high value for the probability of HBV perinatal infection for HBeAg + mothers when newborns were given the vaccine and HBIG was raised to 43%, the HBIG after RDT strategy had an ICER of US$1462, which was the only time the strategy rose above the threshold of US$1200. For all other parameter values evaluated in the one-way sensitivity analysis, the HBIG after RDT strategy remained below US$1200.

The HBIG after confirmatory test was only non-dominated for two parameter values. The first was when a low cost for the confirmatory test (US$8.93) was used, which resulted in an ICER of US$636.14 for the HBIG after confirmatory test and an ICER of US$846.16 for the HBIG after RDT strategy. The second parameter value was the high HBIG cost (US$64.50) was used. This resulted in ICERs of US$925.90 and US$1097.90 for the HBIG after confirmatory test and the HBIG after RDT strategy, respectively. Figure 3 shows the cost-effectiveness acceptability curve for the comparison between HBIG after RDT and vaccine only. At a willingness to pay threshold of US$1200 per perinatal infection averted, the HBIG after RDT strategy had an 87% likelihood of being cost-effective. If the willingness to pay threshold was lowered to US$600 per infection averted, the likelihood of the HBIG after RDT strategy being cost-effective dropped to 32%.

Our results indicate that two cost-effective strategies are available for the prevention of perinatal transmission of HBV in a population on the Thailand-Myanmar border. With an ICER of US$716.78, the HBIG after RDT strategy is cost-effective at the one GDP per capita threshold of US$1200 per hepatitis B infection averted, which would result in even greater benefits in terms of disability-adjusted life years averted. In such a resource-constrained setting where a dose of HBIG can be equivalent to one month’s salary for the average family, this intervention may still not be feasible. Assuming that funding is available, a switch from the current practice of HBIG after confirmatory test to the HBIG after RDT strategy, infections could be reduced by nearly a third for a relatively small increase in the overall programme costs. Importantly, the HBIG after RDT strategy was consistently a cost-effective option in the one-way sensitivity analysis. Our results are similar to a study conducted in Taiwan, where maternal screening for HBsAg and HBIG given to all infants of HBV positive mothers along with universal vaccination was also found to be cost effective at all levels of endemicity [7].

Two cost parameters, the low value for the confirmatory test and the high value for the HBIG, had a significant impact on the results of the one-way sensitivity analysis, shifting the HBIG after confirmatory test to become a cost-effective strategy. As confirmatory tests require facilities found at a hospital, it is unlikely to drop in price by 50% to the low value of US$8.93 that was used in this analysis. Ideally, a new rapid diagnostic test would have confirmatory capability and could make the HBIG after confirmatory test a more viable option, as HBIG costs are also very unlikely to come down in price.

Few economic evaluations of strategies involving HBV screening in low resource settings have been published; these studies mainly focus on universal vaccination, which is consistently cost-effective with the exception of very low endemicity settings [21, 24, 49‐52]. In most high resource settings, a low prevalence of HBV has produced mixed cost-effectiveness results [53‐55]. Many earlier studies have found that universal maternal screening programmes are not cost effective compared to screening only those who are at high risk; however, more recent studies have found universal antenatal screening for HBV to be cost effective [56, 57]. In almost all high resource settings, HBIG is offered to infants born to known HBsAg + mothers. The costs of HBIG are generally covered by long-sighted policies for HBV control. Recent publications suggest that offering HBIG only to those who are willing to pay for it would result in zero uptake in marginalized populations existing on subsistence daily wages due to the high cost [58].

There is evidence from previous settings of high endemicity that HBV transmission can be reduced and elimination of the disease could be possible [59]. The use of HBIG could accelerate this process and this study has shown it is a cost effective strategy to use in a low resource setting. As settings achieve lower endemicity, the cost-effectiveness of interventions to reduce mother to child transmission of HBV will decrease [7, 50], which may make it even less viable to continue with the HBIG after RDT strategy when this money could be directed towards other healthcare needs.

The RDT in use at the site carries a false positive rate of 3.1% (95%CI 1.7–5.4) [29], which raises the possibility of a child receiving HBIG when they do not need it if the HBIG after RDT strategy were adopted. Common mild adverse effects of HBIG include pyrexia, malaise, drowsiness and urticaria with rare reports of serious adverse effects including anaphylaxis [60]. The HBIG after confirmatory test option eliminates this problem, but we were unable to find estimates in the literature for the rate of serious events of HBIG administration in neonates without HBV, making it difficult to directly include in the model. In addition, when the specificity of the RDT was reduced to 95%, which is the low end of the CI recently reported by Banks et al. for this population [29], the ICER rose from US$716.78 to US$828.97, decreasing the cost-effectiveness of the HBIG after RDT strategy. Poor diagnostic accuracy has been reported in the literature so finding a locally-distributed test with better specificity may be advantageous [61]. RDT tests for HBV have significantly raised the safety of blood for transfusion in resource limited settings when used to exclude potentially positive cases from donating blood [61]. Pragmatically the same RDT used to screen blood donors was used to screen pregnant women at the point of care in this setting. While tests with better diagnostic accuracy are being developed, their feasibility in limited-resource settings must be considered [61]. A more specific RDT would increase the cost-effectiveness of the HBIG after RDT strategy, but the increased cost associated with such an RDT may offset this or cause it to be less cost-effective than reported here.

Our study has several limitations. Firstly, long-term effects including morbidity, mortality and costs and effects of HBV infection are not evaluated in our analysis. Also, not all of those who are infected with HBV will suffer long term damage from the disease. Of those infected at birth, 15–25% will acquire cirrhosis of the liver or develop hepatic carcinoma [4, 62]. In clinics such as SMRU, and most limited-resource settings where HBV and associated diseases cannot be treated due to resource constraints, the health gains associated with more thorough neonatal intervention are considerable [63].

Another limitation is the exclusion of data on multiple births, stillbirths, miscarriages and women who delivered at other clinics or hospitals. Approximately 1% of women in this population have multiple births. These could impact the model results; however, due to the robustness of the model in relation to transmission rate changes, it is unlikely to alter the cost effectiveness of any of the strategies. Other important data limitations relate to the uncertainties surrounding the parameters of the model. In particular, a great deal of uncertainty surrounds the transmission probabilities, especially for those who do not receive vaccination or HBIG. For transmission rates for HBsAg carriers and HBeAg carriers who did not receive a vaccine or vaccine with HBIG, the model was robust to variation in the sensitivity analysis. The transmission rates for HBsAg carriers and HBeAg carriers who were vaccinated with or without HBIG, however, had a sizeable impact on the ICER.

Finally, the HBV transmission rates after vaccination and after vaccination and HBIG were based on studies where participants had received all three vaccines as this is the recommendation of the Thai government. Yet in this population, as in most low resource settings, low rates of full vaccination coverage exist. While in Mae La refugee camp the full course coverage was as high as 98%, it may be much lower in migrant settings where access or migration due to work result in reduced completion rates [64]. The effect of the vaccination and HBIG on vertical transmission of HBV will most likely be provided by the dose given at birth and therefore poor uptake of the second and third vaccine would have a low impact on the results of this study [17]. Newer treatment modalities, such as the use of tenofovir to reduce vertical transmission of HBV in resource limited settings, should be considered due to their improved efficacy and potential cost benefits [65]. Tenofovir would be ideal in middle to high HBV endemic settings with high rates of homebirth. The high price of vaccines [66] alongside country-wide HBIG shortages experienced in this setting in 2016 make exploring the option of Tenofovir even more attractive.

This study presents the first economic evaluation of strategies for the prevention of vertical transmission in a marginalized population of HBV on the Thailand-Myanmar border population. The results demonstrate that the current clinical strategy of HBIG after confirmatory test was not cost effective when compared to HBIG after RDT. Barriers to implementing HBIG after RDT strategy in limited-resource settings remain including: quality of the HBV RDT (risking administration of HBIG to infants without risk of HBV), HBIG supply problems, and home births. This study adds to the very limited body of literature on the cost effectiveness of HBV prevention strategies for vertical transmission in low resource settings. The use of HBIG after RDT could be considered in low resource settings, particularly those with high HBV prevalence; however, the need for more cost-effective options for low resource settings is urgent.