Cost-utility analysis of primary HPV testing through home-based self-sampling in comparison to visual inspection using acetic acid for cervical cancer screening in East district, Sikkim, India, 2023

Introduction

Cervical cancer ranks as the second most prevalent cancer among women in India. Globocan estimates for 2020 showed 123,907 incident cases, and 77,348 cases succumbed to this disease [1]. In 2016, the Government of India, through the Ministry of Health and Family Welfare, implemented an operational framework with the goal of facilitating the screening and early detection of three preventable cancers: oral, breast, and cervical cancer [2]. Although visual inspection using the acetic acid (VIA) screening method adopted in the program for cervical cancer is evidence-based, there are concerns about the feasibility of its implementation in the field and beneficiary satisfaction [3,4]. Further, Human Papilloma Virus (HPV) testing exhibits higher sensitivity (98%) and specificity (90.6%) than VIA [sensitivity 31.6% and specificity 87.5%] [5].

In accordance with World Health Organization’s (WHO) recommendation, high-income countries have transitioned from cytology-based screening to primary HPV testing [6]. Research has demonstrated that self-collected samples for HPV testing by women yield comparable results to clinician-collected cervical samples [7,8]. Nevertheless, the implementation of HPV programs in low- and middle-income countries is hindered by the upfront financial burden associated with the adoption of this approach despite its evident advantages over VIA. A systematic review on the cost-effectiveness of all cervical cancer screening methods in low- and middle-income countries revealed that HPV testing through self-sampling is the most cost-effective to screen cervical cancer [9]. In the context of India, a single modeling study has examined the cost-effectiveness of three cervical cancer screening strategies: VIA, Pap smear, and HPV DNA [10]. However, to obtain more precise and trustworthy insights, empirical data is required. By utilizing real-world data, we can gather more reasonable information on the actual costs, effectiveness, and outcomes of cervical cancer prevention strategies when implemented within the Indian healthcare system. Such empirical studies are crucial for enhancing the applicability and validity of findings, empowering policymakers to make well-informed decisions regarding resource allocation for cervical cancer prevention programs in India.

Sikkim, a northeastern state in India, has piloted primary HPV screening through self-sampling in the East district to explore the challenges in implementation and planning to scale up to all districts based on the results and availability of funds for the same. In order to consider HPV testing as the primary screening of cervical cancer in the State, a detailed economic analysis using contextual data is needed. An essential factor in evaluating the validity of a new screening program is its ability to demonstrate economic efficiency while also surpassing the physical, psychological, and societal harms stemming from the associated testing and screening procedures. Thus, the current study was planned to assess the cost-utility of home-based HPV self-sampling compared to VIA for cervical cancer screening among women over 30 years in the East district of Sikkim state.

Methods

Study design and setting

We did a cross-sectional study in the East district, Sikkim, India, in which cost-related data was collected retrospectively, while the data for utility measures was collected prospectively. Fig 1 illustrates the study workflow. It includes estimates on the cost of population-based cervical cancer screening through VIA (facility-based) under the National Program for Prevention and Control of Non-Communicable Diseases (NP-NCD) program and primary HPV screening through self-sampling (home-based) from a disaggregated societal perspective. This perspective considers costs estimated at both the health provider/system level and the individual level, specifically focusing on the direct expenses borne by women undergoing screening in the form of out-of-pocket payments.

• Download figure
• Open in new tab

Fig 1: Illustration of the Study Work Flow

Under the NP-NCD, VIA is performed on the eligible population (women aged 30 years and older) by the Community Health Officer (CHO) at Health and Wellness Centres (HWCs). If they are found to be VIA positive, women are referred to PHC for confirmation by Primary Health Centre (PHC) Medical Officer. In the primary HPV screening group, Accredited Social Health Activists (ASHAs) provide the sampling kits to women when they make home visits for either filling Community Based Assessment Checklist (CBAC) or Reproductive and Child Health (RCH) program-related activities. Community Health Workers in India are referred to as ASHAs. They are equipped with Information Education and Communication (IEC) materials, such as charts and videos, which demonstrate the correct procedure for collecting vaginal samples. The woman then collects the sample herself and hands over the sampler kit back to ASHA. The sample is transported to the HPV lab at the district hospital from PHC in an ambulance, which does a weekly visit to District Hospital for transportation of other contingencies like vaccines and medicines.

The detailed flow of the screening process in both groups is provided in Fig A of the supplementary file. The activities in both groups are the same once the woman is referred to PHC after screening. Hence, the unit cost is calculated only up to the screening process. Participants’ eligibility criteria for calculating the utility values were as follows; women aged 30 years or older who underwent screening using VIA at HWCs by CHOs, and women aged 30 years or above who underwent screening using HPV self-sampling at home were included in the study. Women who declined to provide consent were not included in the study. The reference period for costing was from April 2021 to March 2022.

Considering the mean EuroQol-5 dimension score (SD) of women after cervical cancer screening by VIA to be 0.89 (0.19) and the mean EQ-5D-5L score (SD) of women after HPV self-sampling to be 0.98 (0.19) [11] and power as 85%, alpha as 5% (two-sided), we calculated a sample size of 81 women in each group. Adjusting for a 15% attrition rate, the sample size was calculated to be 95 women in VIA and HPV groups. East district of Sikkim consists of eight PHCs (Two urban and six rural) and forty-three health and wellness centres (HWCs). HPV self-sampling (home-based) is implemented by the state govt in the East district in three PHC catchment areas (two urban and one rural). These HWCs selected for HPV screening do not perform VIA screening anymore. Out of the remaining thirty-two HWCs in the district, the best performing eleven HWCs in terms of number of women screened by VIA were selected.

Results

The baseline and screening attributes of women (95 in each group) interviewed in the VIA and HPV screening group are provided in Table 1. Both the HPV group and the VIA group were comparable in terms of age distributions, marital status, educational status, parity and age at first intercourse. However, there was a significant disparity in residence patterns, with VIA consisting of a higher proportion (99%) of participants residing in rural areas, while HPV predominantly comprised individuals living in urban regions (84%). There was a notable distinction in occupational profiles that emerged, with a greater proportion of homemakers (70%) observed in the VIA group, while the HPV testing group exhibited a more diverse range of occupations.

After screening, it was observed that the VIA group reported a higher prevalence of symptoms (74%) compared to the HPV group (9%). There was a significant difference among the groups regarding time spent on screening. The out-of-pocket expenditure (OOPE) of women in the VIA group ranged from 0 to ₹ 6000 (US$ 0 to 72.1) with a median of ₹120 (IQR ₹ 60-200) (US$ 1.4, IQR), which. Time to reach the facility and OOPE did not apply to the HPV group since the screening was conducted in their homes.

The total annual expenditure on screening women for cervical cancer through VIA and HPV DNA tests amounted to ₹5,79,059 (US$ 6955.4) and ₹28,16,019 (US$ 33825), respectively. The details of the annual costs for both screening methods are provided in Table 2. The distribution of the total annual costs for VIA and HPV screening revealed that training (33%) and consumables (30%) comprised the largest proportion of overall expenses in the VIA group. Notably, in the HPV group, approximately 40% (₹11,29,706, US$ 13569.6) of the total cost was allocated to the purchase of consumables, followed by human resources, which accounted for 30% (₹ 8,55,848 US$ 10280.1) of the expenditure.

During the study period, 392 women underwent VIA screening across eleven HWCs. These HWCs covered a population of 6,031 eligible women, resulting in a screening coverage of 6.5%. In the HPV group, 2,216 women were screened out of the 6,331 eligible women, yielding a screening coverage of 35%. The unit cost per woman screened by VIA and HPV was ₹ 1,597 (US$ 19.2) and ₹ 1,271 (US$ 15.3), respectively.

The comparison of EQ-5D-5L index scores of screened women in VIA and HPV groups is tabulated in Table 3. The mean EQ-5D-5L score was significantly higher in the HPV group (1.00±0) than in the VIA group (0.9903±0.02) (p<0.0001). On the eighth day of screening, the EQ-5D-5L scores remained significantly higher in the HPV group (1.000±0) than in the VIA group (0.9972±0.0118) (p=0.02). The overall health score on the day of screening was significantly higher in the HPV group (88.18±5.44) compared to the VIA group (84.4±6.41), (p<0.0001).

None of the women in the HPV group complained of any deviation from normal in the five dimensions of the EQ-5D-5L. However, in the VIA group, 17% and 7% experienced pain and anxiety, respectively, on day one; on day eight, 3% reported anxiety. (Supplementary Table A)

For a duration of 8 days, the QALD was found to be 7.977 for the VIA group and 8 for the HPV group. Hence the ICER was ₹-14,459 (US$ 173.7). This is much below the WTP threshold for eight days, i.e., ₹ 4,193 (US$ 50.4).

The sensitivity analysis results for changes in the number of women screened using the VIA method is shown in Fig. 2A. As the number of screened women decreases, the ICER increases. When the number of screened women exceeds 549, HPV self-testing ceases to be more cost-effective than VIA. This implies that a screening coverage increase of 40% is required for VIA to be considered cost-effective than HPV self-testing.

• Download figure
• Open in new tab

Fig 2A: Sensitivity analysis: Change in ICER value with variation in the number of women screened by VIA

In contrast, Fig. 2B demonstrates the impact of changes in the number of women screened by the HPV self-testing method on the ICER. As the number of women screened by HPV self-testing decreases, the ICER increases. Specifically, if the screened numbers fall below 1,665, HPV self-testing ceases to be cost-effective. However, even with a 25% lowering in screened numbers, the HPV self-testing method remains cost-effective.

• Download figure
• Open in new tab

Fig 2B: Sensitivity analysis: Change in ICER value with variation in the number of women screened by HPV

The sensitivity analysis results for VIA screening (Fig. 3A) highlighted that training expenses and consumables had the highest impact on the ICER. Conversely, for HPV screening, consumables and human resources were found to have the greatest impact on the ICER (Fig 3B). ICER was below the WTP threshold for all the listed scenarios listed. (Supplementary Table B)

• Download figure
• Open in new tab

Fig 3A: Sensitivity analysis; Key driving factors of ICER with 10% variation in item cost

• Download figure
• Open in new tab

Fig 3B: Sensitivity analysis; Key driving factors of ICER with 25% variation in item cost

Discussion

The findings of our study indicate that primary HPV screening through self-sampling is a cost-saving approach for cervical cancer screening, compared to the current recommendation of VIA for population-based screening in India. While the total cost of HPV screening may appear higher, the unit cost per test is substantially lower than VIA, primarily due to the higher screening coverage achieved with HPV screening. HPV self-sampling has been widely recognized for its ability to increase participation in cervical cancer screening as it overcomes various barriers; self-sampling effectively encourages more individuals to participate in regular screening programs, especially in low and middle-income countries [14,15]. Existing literature indicates that adopting self-sampling approaches can enhance screening participation among women and potentially lead to cost reductions [16]. According to a meta-analysis of 29 randomized controlled trials, the utilization of cervical cancer screening services was found to be twice as likely among women who underwent self-sampling compared to those who followed standard-of-care screening practices [17]. Women exhibit a preference for self-collection of samples within the convenience of their homes, as opposed to visiting healthcare facilities for screening [18]. It was found that there was a high level of compliance among women who opted for this self-collection method, with fewer operational challenges encountered during its implementation. Self-sampling presents a distinct opportunity to reach populations that are typically difficult to access, making it one of the most viable strategies for post-pandemic screening[19]. Self-sampling for HPV holds the promise of diminishing disparities in cervical screening participation within specific demographics and extending access to underserved and previously unscreened women [20,21].

Although VIA is known as a low-cost test and is recommended in low- and middle-income countries, the implementation of this subjective test is a major challenge. An Indian randomized controlled trial demonstrated that a single HPV test can effectively decrease the incidence of advanced cervical cancer and associated mortality. In contrast, VIA screening did not exhibit similar benefits in terms of cancer prevention and mortality reduction [22]. The significant hurdle to the widespread implementation of the more precise and efficient HPV screening method arises from the substantial upfront expenses required to establish screening programs. Although cost plays a vital role, it should not be the exclusive factor guiding decision-making. It is imperative to consider the well-being of women, along with their overall quality of life and satisfaction [23].

Our study findings also show significant differences in quality of life between women screened by HPV self-sampling who reported higher utility scores and a better quality of life compared to those who underwent VIA. The higher utility scores indicate that women screened through HPV self-sampling experienced better health-related quality of life, potentially due to reduced anxiety, pain and discomfort associated with the self-sampling approach. These results highlight the potential benefits of implementing HPV self-sampling as a preferred screening method, providing women with a more positive screening experience and ultimately contributing to improved health outcomes.

Previous studies have evaluated the cost-effectiveness of HPV self-sampling and VIA for cervical cancer screening [10, 24]. However, unlike these model-based studies, our study uniquely utilizes empirical data, marking the first instance of such an approach in India. An earlier model-based study in India that assessed the cost effectiveness of three screening strategies ie. Papanicolaou test, VIA and clinician based primary HPV testing for cervical cancer screening concluded VIA done every 5 years to be cost effective approach [10]. However, the study measured the health-related quality of life of cervical cancer patients rather than screened women to calculate the utility values. This approach failed to capture the immediate change in quality of life associated with the screening procedure itself. In addition, the study did not take into account the building cost of VIA screening as they reported data from camp-based approach.

Another study that evaluated cost effectiveness of same three screening strategies within a large cluster randomized trial from rural India (Legood R et al., 2005). The study dismissed HPV to be a cost-effective method for cervical cancer screening. It is important to note that this study accounted for all the costs in 2002, during which HPV testing was relatively new and prices were high, likely due to reduced competition in the market. Moreover, the study was conducted in a controlled settings of the trial that would vary from ‘real life’ situations. The ASPIRE randomized controlled trial included a cost-effectiveness study, which determined that community-based self-collected HPV testing is a cost-effective screening strategy [25].

Our study had few limitations. Building costs are typically higher in urban areas compared to rural areas. Differential implementation of HPV testing in urban areas and VIA testing in rural areas could have led to an overestimation of costs associated with the HPV testing strategy and an underestimation of costs associated with the VIA strategy. It is crucial to consider this limitation when interpreting the cost comparisons between the two screening methods. Another limitation of our study originates from the exclusive implementation of HPV screening within the East district of Sikkim [26]. The adoption of HPV screening specifically within urban area in the district has resulted in the inclusion of women who underwent VIA screening solely from rural PHC catchment regions. This specific sampling approach introduces a potential bias in the recruitment of participants, which may in turn limit the generalizability of the study findings.

In this study we have leveraged empirical data to evaluate the cost-utility of population-based self-collected HPV testing that is performed at women’s homes. The findings of this study provide a robust support for advocating the adoption and implementation of HPV self-sampling in India. The results indicate that the benefits obtained outweigh the associated costs, establishing it as a cost-effective approach within the provided threshold.. Further research following a similar framework in diverse settings across India is recommended. These studies would contribute to a more comprehensive understanding of the feasibility of HPV self-sampling as a primary screening method in various regional contexts.

Supplimentary file

1. Table A: EQ-5D-5L frequencies and proportions reported by dimensions and level

2. Table B: Comparison of ICER and unit cost for different cervical cancer screening scenarios

3. Fig A: Flow of screening process in HPV and VIA group