Advancing Cervical Cancer Prevention in India: Implementation Science Priorities

Cervical cancer is the leading cause of cancer mortality in India, accounting for 17% of all cancer deaths among women aged 30 to 69 years. We conducted a review of the cervical cancer prevention research literature and programmatic experiences in India to summarize the current state of knowledge and practices and recommend research priorities to address the gap in services. Research and programs in India have demonstrated the feasibility and acceptability of cervical cancer prevention efforts; however, additional evidence generated through implementation science research is needed to ensure that cervical cancer prevention efforts have the desired impact and are cost-effective.

Abstract

Cervical cancer is the leading cause of cancer mortality in India, accounting for 17% of all cancer deaths among women aged 30 to 69 years. At current incidence rates, the annual burden of new cases in India is projected to increase to 225,000 by 2025, but there are few large-scale, organized cervical cancer prevention programs in the country. We conducted a review of the cervical cancer prevention research literature and programmatic experiences in India to summarize the current state of knowledge and practices and recommend research priorities to address the gap in services. We found that research and programs in India have demonstrated the feasibility and acceptability of cervical cancer prevention efforts and that screening strategies requiring minimal additional human resources and laboratory infrastructure can reduce morbidity and mortality. However, additional evidence generated through implementation science research is needed to ensure that cervical cancer prevention efforts have the desired impact and are cost-effective. Specifically, implementation science research is needed to understand individual- and community-level barriers to screening and diagnostic and treatment services; to improve health care worker performance; to strengthen links among screening, diagnosis, and treatment; and to determine optimal program design, outcomes, and costs. With a quarter of the global burden of cervical cancer in India, there is no better time than now to translate research findings to practice. Implementation science can help ensure that investments in cervical cancer prevention and control result in the greatest impact.

Implications for Practice:

Considerable research has been conducted on the prevention of cervical cancer in India. The majority of studies have focused on the feasibility, acceptability, and impact of secondary prevention of cancer through screening, early detection, and treatment. Despite this evidence, there have been few government-led public health programs to prevent and control cervical cancer. The primary goals of this review are to summarize the lessons learned from cervical cancer prevention research and pilot programs in India and to identify research priorities to facilitate the translation of existing knowledge into policies and programs that advance cervical cancer prevention.

Introduction

Cervical cancer is the most common cancer among women aged 15 years or older in India. Cervical cancer will occur in approximately 1 in 53 Indian women during their lifetime compared with 1 in 100 women in more developed regions of the world [1]. In 2010, there were nearly 74,000 new cases of cervical cancer in India. With 38% of cases occurring among women of reproductive age (15–49 years), the adverse social and economic impact of cervical cancer on families and communities is considerable [1].

Between 1980 and 2010, little progress was made in reducing cervical cancer mortality in India: 37 women died for every 100 new cases of cervical cancer in 1980 compared with 32 for every 100 new cases in 2010 [1]. High mortality rates are largely the result of nearly 70% of cervical cancer cases in India being diagnosed at an advanced stage (stage III or IV) [2]. Fewer than a third of Indian women diagnosed with stage III cervical cancer survive the first five years after their diagnosis, and the 5-year survival rate drops to nearly 6% among women diagnosed with stage IV disease [2].

In contrast to the United States and other high-income countries, where cervical cancer screening is offered as part of routine primary care, few large-scale screening programs exist in India [3]. Moreover, although primary prevention through human papilloma virus (HPV) vaccination is gaining acceptance in high-income countries and has been endorsed by the World Health Organization (WHO), vaccine awareness, access, and use are very low [4, 5]. However, cervical cancer prevention efforts appear to be gathering momentum. Coinciding with the United Nations High Level Summit on Non-Communicable Diseases in 2011, India's national government launched a program to address chronic and noncommunicable diseases (NCDs) that includes screening and treatment of cervical cancer. Simultaneously, several state governments decided to do a pilot test of their own NCD prevention efforts. In this context, we conducted a review of the cervical cancer prevention research literature and programmatic experiences to summarize the current state of knowledge and practice and recommend research priorities to facilitate the translation of existing knowledge into efficient, effective, and equitable public health action.

Methods

This paper is based on information gathered from a review of English-language peer-reviewed publications and gray literature (including unpublished program reports, white papers, and conference presentations) on cervical cancer prevention in India and interviews with individuals involved in the design and implementation of government-led cervical cancer prevention efforts in the southern state of Tamil Nadu.

The published literature was identified using ISI Web of Knowledge, PubMed, and Google Scholar using a broad time frame (1990 to the present). A combination of the following search terms was used: India, cervical cancer, screening, early detection, HPV, HPV vaccination, and visual inspection. The gray literature was identified using keyword search terms in Google's search engine, reviewing references of published papers, and searching Internet-based document repositories. Papers and reports were reviewed and key information regarding methods and findings was abstracted and organized thematically.

To better understand the translation of knowledge into practice, we conducted a case study of cervical cancer prevention efforts in the southern state of Tamil Nadu. Data sources included program reports and presentations, interviews, and field visits. Nine individuals (public health officials and cancer prevention experts) who were involved in program design and implementation were interviewed to ascertain factors that shaped program design, challenges faced in implementation, and strategies implemented to overcome those challenges. Field visits were made to a tertiary hospital where training in the diagnosis and treatment of cervical cancer is conducted for health care providers and to a district hospital and primary health center implementing cervical cancer screening, treatment, and referrals. Interviews and field notes were summarized thematically.

Research on Primary and Secondary Prevention of Cervical Cancer in India

Our review of the literature on cervical cancer prevention in India identified 44 peer-reviewed articles focused on the following themes: HPV prevalence and vaccination; performance of screening tools; and feasibility, acceptability, and effectiveness of screening and treatment interventions. Below we summarize the main findings by theme.

Primary Prevention Through HPV Vaccination

Available data indicate that HPV vaccination is a highly promising cervical cancer prevention method when used in conjunction with screening and treatment [6]. Approximately 70% of cervical cancers in India are caused by HPV types 16 and 18, which are targeted by the vaccine [7, 8]. However, there has been little empirical research on the feasibility, acceptability, and effectiveness of HPV vaccination in India. A qualitative study in the states of Gujarat and Andhra Pradesh identified three potential mechanisms for vaccine delivery: the existing national immunization program, adolescent health or cancer control services, and school- and community-focused campaigns [4, 9]. The study noted that successful introduction of the vaccine would depend on a communication strategy focused on raising community awareness, engagement with community gatekeepers, and advocacy to build political support for prevention. It also revealed that systems-level challenges associated with staffing, capacity, and coordination would have to be addressed [4]. Other research in southern India found that although parents held positive attitudes about vaccination, they were not in support of vaccinating pre-adolescent girls and felt that vaccination against HPV was more appropriate once girls had attained puberty [5].

Public provision of HPV vaccination has been debated in the mainstream press and in academic journals [4, 10–18]. Women's health activists and researchers in India have questioned the appropriateness of the HPV vaccination in light of limited data on safety and efficacy, the high cost, and the uncertain cost-effectiveness of the vaccine. They have argued that the government should focus on promoting awareness and ensuring access to comprehensive reproductive and sexual health services, including cervical cancer screening and treatment [14]. Others have challenged investments in HPV vaccination by disputing the data on India's cervical cancer burden, arguing that incidence rates are in fact on the decline in the country [15]. The deaths of four adolescent girls enrolled in a government-approved demonstration project in Gujarat and Andhra Pradesh escalated the controversy. Although an expert panel later concluded that the deaths were unrelated to the vaccine [9], the investigation revealed ethical violations related to the process of informed consent [11].

HPV vaccine debates in India raise several issues for consideration in future vaccination efforts. First is the need for systems to monitor, report, and respond to postvaccination adverse events. The Gujarat and Andhra Pradesh study found that although guidelines were available for postimmunization adverse event management, often they were not followed [4]. Additionally, the widely differing interpretations of the data on HPV vaccine safety among health activists, clinicians, public health practitioners, and journalists suggests that a more robust communication strategy is needed to disseminate evidence in support of vaccination. Research should identify effective ways to communicate information regarding the vaccines, including a focus on adolescent girls, partial efficacy, and the need for cervical cancer screening in adulthood.

Finally, cost and cost-effectiveness remain important considerations for India's HPV vaccination policy. A mathematic modeling study informed by empirical data from India found that pre-adolescent vaccination for HPV types 16 and 18 can reduce lifetime cervical cancer risk by 44%, assuming 70% vaccine coverage, and is more effective than screening alone, regardless of screening test, screening frequency, and target age groups [19]. Combining vaccination with screening was cost-effective according to WHO benchmarks for developing countries at a vaccine cost of US$2 per dose or less.

A mathematic modeling study informed by empirical data from India found that pre-adolescent vaccination for HPV types 16 and 18 can reduce lifetime cervical cancer risk by 44%, assuming 70% vaccine coverage, and is more effective than screening alone, regardless of screening test, screening frequency, and target age groups. Combining vaccination with screening was cost-effective according to WHO benchmarks for developing countries at a vaccine cost of US$2 per dose or less.

The Indian market price of the vaccine is approximately US$40 per dose. In 2013, the GAVI Alliance announced that it would offer the HPV vaccine at US$4.50 per dose to countries with a gross national income per capita below or equal to US$1520 [20]. The estimated total cost of vaccination per girl, taking into account three doses and vaccine delivery costs, is US$20.70 [21]. Eligible countries (of which there are 57, including India) can apply for support to introduce the vaccine nationwide or to conduct pilot projects that will enable them to prepare for national implementation. In light of the global evidence on HPV vaccination, pilot testing the vaccine and gathering empirical data on vaccine delivery costs and cost-effectiveness can significantly advance cervical cancer prevention efforts in India.

Secondary Prevention Through Screening

Secondary prevention through screening and treatment of precancerous and early stage cancerous cervical lesions can prevent disease progression and reduce subsequent morbidity and mortality. Substantial declines in cervical cancer incidence and mortality in high-income countries such as the United States have been attributed to screening. However, screening approaches used in high-income countries (namely, the annual Papanicolaou smear test) may not be appropriate in low- and middle-income countries where establishing laboratory infrastructure, training personnel such as cytotechnicians and pathologists, and implementing continuous quality assurance procedures have proven difficult [22]. Consequently, research has focused on evaluating visual inspection-based methods that use existing (or minimal additional) human resources and require less training and fewer clinic visits (Table 1).

Table 1.

Overview of primary screening tools for cervical cancer

^aAdapted from [55, 56].

Abbreviations: CIN2+, cervical intraepithelial neoplasia grade 2 or higher; HPV, human papilloma virus; VIA, visual inspection with acetic acid; VILI, visual inspection with Lugol's iodine.

Accuracy of Screening Tests

Over the past decade, 14 studies have been conducted in India to evaluate the accuracy and effectiveness of cervical cancer screening [23–34]. Studies have compared the specificity and sensitivity of visual inspection with acetic acid (VIA) or Lugol's iodine (VILI) and the Papanicolaou test and HPV DNA testing. These approaches have been implemented by health workers such as auxiliary nurse midwives (Table 2). The majority of these studies were cross-sectional in design and involved either community- or hospital-based samples [23–32, 35]; two studies were randomized controlled trials [33, 36–38]. To determine test sensitivity and specificity, most studies performed colposcopy for all participants and biopsy of abnormalities if indicated; in some, the colposcopist was blinded to the initial screening test results [24, 26, 27].

Table 2.

Health care workers trained to perform cervical cancer screening and treatment in studies in India

Abbreviations: HPV, human papilloma virus; IARC, International Agency for Research on Cancer; LEEP, loop electrosurgical excision procedure; NS, not specified; TMC, Tata Memorial Center, Mumbai; VIA, visual inspection with acetic acid; VILI, visual inspection with Lugol's iodine.

Salient findings of Indian studies on screening test performance are summarized in Table 3. Overall, research in India suggests that VIA and VILI have comparable sensitivity and specificity to Papanicolauo and HPV DNA test-based screening (Table 3). Studies have found that the sensitivity to detect cervical intraepithelial neoplasia (CIN) grade 2 or higher (CIN2+) is similar for VIA (range 64.5%–89.5%) and VILI (range 64.5%–100%) compared with Papanicolaou test (range 52.6%–62.3%). Research has found test specificity of VIA (76.4%–84.2%) to be lower than that of VILI (85.4%–93.4%) and HPV testing (80.7%–81.3%). Cytology testing has the highest specificity (76.1%–99.1%) [23–32, 35].

Table 3.

Summary of Indian studies on accuracy of cervical cancer screening tests

Abbreviations: ASCUS, atypical squamous cells of undetermined origin; CIN, cervical intraepithelial neoplasia; HSIL, high-grade squamous intraepithelial lesion; IARC, International Agency for Research on Cancer; LSIL, low-grade squamous intraepithelial lesion; RCT, randomized controlled trial; VILI, visual inspection with Lugol's iodine.

Studies have indicated that visual inspection-based approaches have lower test sensitivity when used with older women because of the migration of the transformation zone into the endocervical canal [37, 39]. Consequently, different screening strategies may be needed depending on the age group of the target population.

An advantage of visual inspection-based approaches is that the immediate availability of screening results provides the opportunity to conduct a biopsy or offer treatment at the same visit (a screen-and-treat approach), reducing the likelihood of loss to follow-up. That said, there is a risk for overtreatment in this context given that the specificity of visual inspection-based approaches is lower than that of the Papanicolaou test or HPV DNA testing. Although there is no clear evidence of harm arising from overtreatment [40, 41], some researchers contend that using the screen-and-treat approach entails risks that may only become apparent with longer follow-up [42]. The feasibility and effectiveness of the screen-and-treat method should be examined further in a patient program.

In conclusion, decisions regarding which screening strategy to use will largely depend on context, including the availability of infrastructure and human and financial resources. Studies on screening methods in India indicate that visual inspection-based approaches are feasible and reasonably accurate when compared with methods such as the Papanicolaou test and HPV DNA testing. However, training and quality assurance are especially important to ensuring acceptable test performance levels when using visual inspection. Thus, a key challenge is to create a system that is capable of effectively deploying the selected screening strategy. Establishing such a system and initiating screening can lay the foundation for the introduction of more optimal primary screening tests in the future.

Screening Coverage and Treatment Linkings

The effectiveness of screening is determined not only by the accuracy of the screening test, but also by screening coverage and linkages to diagnostic and treatment services. Several community-based studies have tracked participation and retention rates by first enumerating all eligible women in the target communities [25, 33, 36] or using census lists to determine the number of women eligible [29]. Even in the context of well-resourced studies, achieving high levels of screening participation and adherence to diagnostic and treatment recommendations—without which it is difficult to justify screening—has been a challenge.

Studies that successfully recruited and retained participants in screening and treatment engaged with women, husbands, and family members; offered appointments and screening cards to women; and involved local government health workers and community leaders in mobilization efforts (Table 4) [25, 33, 36]. In Dindigul district, Tamil Nadu, accessibility was further enhanced by offering screening services at primary health centers, municipal offices, schools, women's clubs, and homes [33]. The study screened 63% of eligible women.

Table 4.

Strategies used to promote screening in India

Sources: [25, 33, 48, 57, 58].

The importance of community engagement is underscored by the Community Access to Cervical Health study (CATCH), which despite using a fairly comprehensive approach involving community liaisons to promote screening, house-to-house visits, and transportation to a tertiary screening center, had among the lowest participation rates of published Indian studies: nearly three in five eligible women (59.4%) refused to undergo screening [29]. Reasons for refusal included lack of symptoms indicating a health problem; fear of the tests, the pelvic examination, and a cancer diagnosis; and community gossip and misconceptions.

Similar reasons for screening refusal were reported by Basu et al. [43] in their Kolkata study. Additionally, they found that some women were unable to attend screening clinics because of household responsibilities, family problems and illnesses, and refusal of husbands or other relatives to grant permission. A few women who made it to the clinic left prior to being screened because of apprehension about the process (after seeing the screening instruments), the long waiting times, and the presence of male doctors.

Facilitators of and barriers to patient adherence to diagnostic and treatment recommendations were generally similar to those for screening. Loss to follow-up tended to be lower in the case of more advanced disease, but did not vary consistently by location of diagnostic and treatment service provision [29, 33, 34, 36]. In the Dindigul study, trained nurses used a screen-and-treat approach for most precancerous lesions and referred women requiring loop electrosurgical excision procedure or cold knife conization or with invasive cancers to a tertiary center [33]. The study documented high loss to follow-up among women who did not choose to undergo screening and treatment at the same visit: 53% of women who were eligible for the screen-and-treat approach but decided to first consult their husband or other family members did not return for treatment. Diagnostic and treatment completion rates among women referred to a tertiary center were 80% and 75%, respectively [33]. In the Osmanabad, Maharashtra, trial, 85% of those referred to a central clinic for diagnostic and treatment procedures adhered to recommendations [36]. The CATCH study reported the lowest acceptance of diagnostic and treatment procedures, with 34% of women refusing colposcopy and biopsy after a positive screening test [29]. The relative role of factors such as cancer stigma, accessibility of services, and family support in shaping diagnostic and treatment adherence needs to be explored further.

The translation of evidence on cervical cancer screening and treatment approaches into effective prevention programs depends on a sound understanding of factors operating at the level of the individual, community, and health system that influence women's access to information/education, screening, diagnosis, and treatment. Further investigation of these factors and evaluation of strategies to improve service uptake can facilitate future program planning and implementation.

Quality of Service Provision

To date, researchers have assessed the quality of cervical cancer prevention services primarily in terms of screening test performance, and quality assurance has focused on training of health care providers (Table 5). Although studies have found it feasible to train frontline health workers such as auxiliary nurse midwives and staff nurses to implement visual inspection-based screening and treatment of precancerous lesions using cryotherapy, concerns have been raised about the quality and consistency of service provision [36, 44]. Auxiliary nurse midwives in the Osmanabad study received three weeks of training on cervical sample collection, VIA, and cryotherapy, but poor concordance on VIA results between the master trainer and the nurses led to two rounds of refresher training [36].

Table 5.

Quality assurance methods

Prescreening and postscreening procedures, including informed consent for screening and postscreening counseling and monitoring patient follow-up rates for diagnostic confirmation and treatment, are important components of quality assurance that have not been discussed extensively in the research literature. It is unclear whether studies used electronic- or paper-based health information systems to monitor participant retention and follow-up, whether any specific actions were taken to minimize loss to follow-up, and which actions were most effective in promoting patient retention. Identifying best practices in quality assurance is a high priority.

Impact of Screening on Morbidity and Mortality

Three large-scale randomized controlled trials in India have found that screening can reduce cervical cancer morbidity and mortality [34, 37, 38]. In Dindigul, a single round of VIA-based screening led to a 25% reduction in cervical cancer incidence and a 35% reduction in mortality over seven years of follow-up [37]. Similar results (31% reduction in mortality) were reported recently by a trial that offered multiple rounds of VIA-based screening in Mumbai, Maharashtra [34]. A third randomized controlled trial in Osmanabad found that a single round of HPV testing led to reductions in mortality, whereas VIA and cytology did not [38]. Although screening uptake and treatment adherence did not differ between the study arms, questions have been raised about whether differences in diagnostic and treatment procedures across study arms may explain the observed outcomes [45].

Overall, data from India indicate that VIA, when offered under the controlled conditions of a trial with systems for following-up with women whose screening is positive and high adherence to diagnostic and treatment recommendations, leads to significant declines in cervical cancer mortality within a decade of intervention initiation, lending further support to the rationale for investment in secondary prevention. That said, the effectiveness of screening in less controlled programs and programs with potentially fewer resources remains unclear.

In Dindigul, a single round of VIA-based screening led to a 25% reduction in cervical cancer incidence and a 35% reduction in mortality over seven years of follow-up. Similar results (31% reduction in mortality) were reported recently by a trial that offered multiple rounds of VIA-based screening in Mumbai, Maharashtra.

Cost Effectiveness of Screening

Modeling studies using primary and secondary research data from India indicate that cost-effectiveness of vaccination is influenced by vaccine cost, efficacy, coverage, and duration of protection, whereas cost-effectiveness of screening depends on the linkages between screening and treatment, number of clinic visits, and laboratory infrastructure required [19, 46]. Strategies that targeted women in their mid-30s were the most cost-effective (compared with those that screened women less than 30 years old or more than 45 years old).

The Osmanabad randomized controlled trial, which estimated screening costs and cost-effectiveness, found that program implementation expenses (e.g., for monitoring patient recruitment and follow-up) were substantial—as high as one fifth of the total costs in the VIA arm [47]. VIA was the least expensive option, detecting 7.5 CIN2/3+ cases per 1,000 eligible women and costing $522 per case detected (compared with no screening), whereas cytology detected 10 cases per 1,000 and cost $659 per case. HPV DNA testing was nearly twice as expensive as cytology and detected fewer cases. Another modeling study recommended screening adult women two to three times per lifetime in addition to administering pre-adolescent vaccination, and found that this combination (assuming 70% coverage for both strategies) yielded a 56% to 63% reduction in cancer incidence [19].

Program data are needed to better understand cost and cost-effectiveness of available prevention strategies. Monitoring and evaluation of pilot programs in different settings (e.g., rural vs. urban) can yield important insights about the choice of prevention strategies. Finally, analyses should also account for treatment costs and longer-term benefits in terms of life years saved.

Translation of Research Into Practice

Tamil Nadu State's cervical cancer screening and treatment program designed and implemented by the World Bank-supported Tamil Nadu Health Systems Project (TNHSP) is among the few examples of large-scale government-led prevention efforts in India. In 2007, Tamil Nadu launched a pilot cervical cancer screening program in two predominantly rural districts, Theni and Thanjavur, to identify challenges and potential solutions prior to a state-wide scaling up of the program [48]. The goals of the cervical cancer program were to raise community awareness, promote early detection by offering routine VIA/VILI-based screening to 30- to 60-year-old women seeking public health services, and offer appropriate referrals and treatment (Table 6).

Table 6.

Tamil Nadu Health Systems Project cervical cancer program by level of care

Source: Tamil Nadu Health Systems Project.

Abbreviations: VIA, visual inspection with acetic acid; VILI, visual inspection with Lugol's iodine.

The pilot program, led by a senior bureaucrat and designed by technical experts, received administrative support from the government of Tamil Nadu, political backing, and the necessary financial resources. A team of public health physicians who were tasked with implementation coordination and oversight and staff nurses who were trained to provide NCD prevention services including VIA/VILI, and data management personnel were hired. Consultations were held with experts in NCD screening, diagnosis, and treatment to finalize clinical protocols and operations manuals, and a detailed capacity-building process was developed to train and equip staff to implement the program. In addition, the program commissioned an external monitoring and evaluation and cost analysis of the program.

Between 2007 and 2010, the program screened nearly 500,000 women, which translates to about 74% of the eligible population [48]. However, VIA/VILI positivity was lower than reported in the literature: 5.4% in Thanjavur and 2.6% in Theni compared with approximately 15% to 20% in published Indian studies [35, 39, 48]. Follow-up rates were also relatively low, with approximately half of women with a positive screen undergoing diagnostic investigations. The program detected relatively few precancerous lesions (103 CIN2/3 lesions) compared with invasive cancers (887), although this may be in part because the population was largely unscreened. Notably, the overall proportion of women treated was low; only 13% of women needing treatment received it through the program [48].

A number of implementation challenges are likely to have led to these program outcomes (Table 6). Efforts to mobilize women to undergo screening were constrained by a limited evidence base to guide information, education, and communication efforts and outreach, which was initially carried out by contractual workers hired by nongovernmental organizations, was inconsistent and often of poor quality. The high coverage achieved by the program despite these limitations was likely the result of Tamil Nadu's strong primary health care system combined with the concerted efforts of the staff of the pilot program. However, that a large proportion of women did not return for diagnostic and treatment services suggests the existence of other individual-, community- and health system-level barriers such as lack of transportation or resources to travel to higher-level facilities for diagnostic confirmation and treatment, lack of familial support, cancer-related stigma, and inadequate referral systems. TNHSP subsequently hired a communications agency to conduct formative research and develop an information, education, and communication plan. However, administrative issues prevented the deployment of the plan within the pilot implementation period, and the effectiveness of the information, education, and communication strategy has not been evaluated.

Program outcomes such as relatively low VIA/VILI positivity, high loss to follow-up, and low levels of treatment provision were likely the result of several factors within the health system. Ensuring adequate, consistent, and high-quality training is one factor. Staff nurses received two days of training on VIA/VILI, which may not have been sufficient. In fact, similarly low VIA/VILI positivity was reported in Bangladesh, highlighting the need for refresher training [49]. Moreover, training quality could have been improved by greater standardization of the number of VIA/VILI tests done, the number of positive cases observed, and structured post-training assessments of knowledge and skills. Anecdotal evidence also suggests that refresher training may have been needed to increase physician confidence to conduct cryotherapy, conization, and loop electrosurgical excision procedure and to promote adherence to program protocols.

Several additional human resource-related issues may have also influenced the level and consistency of program implementation. Frequent changes in personnel at the local, district, and state levels because of transfers and staff attrition meant that program-related knowledge and experiences had to be continuously reestablished. In addition, accuracy of data reporting may have been reduced because nurses felt pressured by the pilot program's target-oriented approach; each center had fixed targets for the number of women they had to screen each month regardless of the sociodemographic profile of the communities they served. In many health centers, the newly hired nurses were tasked with provision of routine health care services, despite having been recruited specifically for the NCD prevention program. This situation was exacerbated because existing health department staff were reluctant to provide NCD-related services. Finally, difficulties in ensuring the availability of sufficient numbers of pathologists and obstetrician/gynecologists may have contributed to delays in reporting of biopsy results and treatment provision. These challenges were further compounded by infrastructural and logistic issues such as lack of availability of drugs and reagents (e.g., drugs to treat reproductive tract infections) and infrastructural deficiencies (e.g., lack of electricity, which made visualization of the cervix difficult).

Perhaps the most important challenges faced by the TNHSP pilot program had to do with the establishment of mechanisms to facilitate patient follow-up and quality assurance. Although screening was promoted at all primary care facilities, diagnostic confirmation and treatment were offered only at select secondary and tertiary facilities where related infrastructure and human resources were already available. In this scenario, communication and coordination of referrals between levels of the health care system were especially difficult. Moreover, the reliance on “village link volunteers” to facilitate patient follow-up may have been ineffective because these individuals were hired by local nongovernmental organizations and did not have sufficient familiarity with the health system. The volunteers had a poor rapport with existing health department staff such as the village health nurses, and they were offered only an honorarium (which was not always paid on time by their nongovernmental organization employer). Inaccurate and inconsistent data reporting such as incomplete or false addresses and lack of clarity regarding the definition of clinical terms and outcomes also hampered follow-up of patients and program monitoring, preventing timely responses to program deficiencies.

Similar to early stage screening programs in other low- and middle-income countries, the TNHSP pilot emphasized strengthening the health care infrastructure and training health care workers to deliver cancer prevention services [49]. Establishing a comprehensive quality assurance program, identifying strategies to promote communication and coordination between different levels of the health care system, and strengthening systems to facilitate identification and monitoring of cases for follow-up are issues that need to be addressed in the next phase of the program [50–52].

Recognition of these challenges has led to a range of responses in the design of the statewide scale up of the program (Table 7). These include the provision of refresher training for all health care workers; the introduction of an electronic health information system, with patient records accessible at different levels of the health care system; and the use of novel strategies to improve patient follow-up, such as cell phone reminder calls. Ongoing analyses of data captured by this system are shared by state-level program officials with district-level staff through monthly videoconferences and with community-level staff (mainly nurses) through weekly meetings at the district hospital. A variety of quality assurance methods, including facility checklists, exit interviews, and structured observations, are being used.

Table 7.

Lessons learned from implementation of a pilot cervical cancer screening program in Tamil Nadu

Source: Tamil Nadu Health Systems Project, personal communication.

Abbreviation: NCD, noncommunicable disease.

The TNHSP pilot program has demonstrated the feasibility and acceptability of introducing cervical cancer prevention into the Indian public health system when there is adequate and consistent political and administrative support. It has also highlighted several implementation challenges related to service uptake and program delivery. Although few conclusions can be drawn about the outcomes and impacts of the pilot program given the lack of evaluation data, these pilot experiences and systematic documentation and analysis of the scale-up program can provide important insights for future prevention efforts in the country. Experiences in countries such as Zambia [51] and El Salvador [52] have revealed that focused efforts to improve quality of service provision, including follow-up rates, entail a strong commitment to quality improvement, dedicated resources, and a defined methodology. Based on these criteria, TNHSP is well positioned to successfully improve cervical cancer prevention across the state of Tamil Nadu.

Conclusion

Implementation Science Priorities

The availability of primary and secondary prevention tools has accelerated global efforts to prevent and control cervical cancer. In India, qualitative research has identified factors that influence the feasibility and acceptability of HPV vaccination. Cross-sectional studies and randomized controlled trials have shown that visual inspection-based screening approaches can achieve sensitivity and specificity comparable to that of cytology-based screening, that this screening can be implemented by frontline health workers, and that it can reduce cervical cancer incidence and mortality. The TNHSP pilot program has demonstrated the feasibility of applying research insights to public health policy and practice. However, our review of research and program experiences has revealed a number of critical gaps in the translation of the available evidence for public health action. Implementation science, which may be defined as research that supports the uptake, implementation, and sustained use of evidence-based interventions in routine practice and program settings, can help bridge these gaps [53].

Implementation science can provide evidence on program effectiveness, efficiency, and sustainability [54], and involves the use of a wide range of methods, data sources, and study designs. For example, implementation science research can entail the use of data from program monitoring and evaluation systems to better understand factors that explain the gap between desired and observed program outputs and outcomes. It can involve experimental and quasiexperimental approaches to assess the comparative advantages and cost-effectiveness of different intervention packages. Alternatively, it can use qualitative and quantitative methods to identify ways to enhance service uptake, improve implementation quality and efficiency, and ensure that a continuum of services is being delivered.

Below, we put forth implementation science priorities for advancing cervical cancer prevention and control in India:

• Understanding individual and community-level barriers to uptake of screening, diagnostic, and treatment services can lead to the identification and testing of strategies to increase program coverage and improve outcomes. For example, research is needed to determine how best to communicate the benefits and limitations of available prevention and control strategies to key stakeholders such as women and their husbands and to explore the role of cultural beliefs, stigma, gender inequities, and other factors in shaping health care decisions. Available evidence suggests that women who are older and have fewer socioeconomic resources are less likely to undergo cervical cancer screening and to have poorer outcomes. Program monitoring data can be analyzed to assess whether specific groups of women are underserved and why, leading to appropriate programmatic responses.

• Improving health care worker performance by identifying effective methods for training, supporting, and supervising providers is a priority. This includes the use of quality improvement tools such as “plan, do, study, act” cycles, telemedicine, simulation-based training, and financial and nonfinancial incentives.

• Strengthening links among screening, diagnosis, and treatment is critical to program success. The TNHSP program integrated cervical cancer prevention into a broader initiative focused on NCDs, with screening taking place at the primary health care level and diagnostic confirmation and treatment taking place at the secondary and tertiary levels. Although screening coverage was high, patient follow-up rates for diagnostic confirmation and treatment were poor. Implementation science research can be used to identify and evaluate strategies to increase patient retention and improve communication and coordination across levels of the health care system. Potential strategies include the provision of transportation, employment of patient navigators, and/or the use of mobile communication technologies.

• Determining optimal program design, outcomes, and costs is important given the number of competing public health priorities facing the Indian government, including the growing burden of NCDs such as diabetes, hypertension, and cancer. In their 2013 guidance note, WHO emphasized the importance of taking a comprehensive approach to cervical cancer prevention and control by implementing vaccination programs as well as screening and treatment in conjunction with education and social mobilization at appropriate points across the life course [21]. Research is needed to identify the optimal way to implement cervical cancer prevention in India, taking into account cost, human resources, and infrastructural elements. For example, should cervical cancer screening be packaged along with NCD prevention or family planning services? What are the best entry points for HPV vaccination? Data on cost and cost-effectiveness of primary and secondary prevention strategies are also important to ensure that limited resources are used most strategically. Key stakeholders such as program managers and policy makers should be engaged in this process to ensure policy relevance.

With a quarter of the global burden of cervical cancer in India, there is no better time than now to translate research findings into practice. Implementation science can help ensure that investments in cervical cancer prevention and control result in the greatest impact.

Author Contributions

Conception/Design: Suneeta Krishnan, Emily Madsen, Deborah Porterfield, Beena Varghese

Provision of study material or patients: Suneeta Krishnan

Collection and/or assembly of data: Suneeta Krishnan, Emily Madsen

Data analysis and interpretation: Suneeta Krishnan, Emily Madsen, Deborah Porterfield, Beena Varghese

Manuscript writing: Suneeta Krishnan, Emily Madsen, Deborah Porterfield, Beena Varghese

Final approval of manuscript: Suneeta Krishnan, Emily Madsen, Deborah Porterfield, Beena Varghese

Disclosures

The authors indicated no financial relationships.