Cervical cancer screening invitations in low and middle income countries: Evidence from Armenia

https://doi.org/10.1016/j.socscimed.2021.113739Get rights and content

Highlights

  • Ninety percent of cervical cancer deaths occur in Low- and Middle-Income Countries (LMICs).

  • We run a large-scale RCT in Armenia, a LMIC.

  • We test the effect of low-cost screening program invitation strategies on take-up.

  • Invitation letters enhance screening take-up, especially when followed by reminders.

  • We find no evidence that different letter and reminder frames affect participation.

Abstract

Roughly 90 percent of cervical cancer deaths occur in low- and middle-income countries (LMICs), where the lack of adequate infrastructures hampers screening, while informational, cultural, and socio-economic barriers limit participation in the few programs that do exist. We conducted a field experiment with the Armenian cervical cancer screening program to determine whether, despite these barriers, the simple, economical invitation strategies adopted in high-income countries could enhance screening take-up in LMICs. We find that letters of invitation increase screening take-up, especially when there are follow-up reminders. Different ways of framing messages appear to have no impact. Finally, women in rural areas are more likely to respond to invitation by letter, helping to narrow the urban-rural screening gap.

Introduction

Cervical cancer is the fourth most frequent cancer in women, accounting for roughly 10% percent of all female cancers diagnosed in 2018, with a mortality rate of 8.2/100,000 women at risk (Globocan, 2018). Besides morbidity and mortality, cervical cancer has significant adverse social and economic effects on patients, their families, and society as a whole, mostly involving the high cost of treatment and diminished productivity (Campos et al., 2017; Ekwueme et al., 2008; Oliva et al., 2005).

Owing to its specific mono-causal genesis, which requires considerable persistence of human papillomavirus (HPV) infection of the uterine cervix, cervical cancer is one of the most preventable neoplastic diseases. Thanks to systematic population-based screening programs (either via a cytology-based approach or via HPV testing), the high-income countries (HICs) have achieved a demonstrated, substantial decline in cervical cancer incidence and mortality in the last few decades (Arbyn et al., 2009; Lăără et al., 1987; Peto et al., 2004). For instance, recent European evidence suggests that women who participate in organized screening may have cervical cancer mortality ranging from 41% to 92% lower than non-participants (Jansen et al., 2020).

Some 90% of cervical cancer deaths now come in low- and middle-income countries (LMICs). This unhappy outcome has been attributed mostly to lack of infrastructures and scarce healthcare resources for national screening programs (Gakidou et al., 2008; O'Donovan et al., 2019; Rao, 2012; Sankaranarayanan et al., 2001). Even when such programs are in place, participation is often much lower than in HICs, the discrepancy depending in part on socio-economic and cultural barriers (Hull et al., 2020; Denny et al., 2006). Thus, the LMIC governments that do initiate population-based cervical cancer screening programs face an important policy challenge to increase their take-up rates, which requires a variety of interventions.

HICs have adopted a number of communication and invitation strategies to boost cancer screening take-up. For instance, there is robust evidence that invitation letters and reminders substantially increase women's participation in cervical cancer screening (Decker et al., 2013; Eaker et al., 2004; Radde et al., 2016; Tavasoli et al., 2016). Invitation letters increase recipients' awareness of the programs in place, underscore their eligibility, emphasize the importance of participation and make it easier for recipients to evaluate the costs and benefits (BIT, 2014; Van Gestel et al., 2017 and the references therein). Moreover, reminders help recipients to overcome any problems of self-control, memory, and attention, getting them to go for screening once their awareness of the program is enhanced through the invitation letters (e.g., Altmann and Traxler, 2014; Sunstein, 2014; Macharia et al., 1992; Milkman et al., 2011).

Turning to the LMICs, there is evidence that here too invitation messages and reminders can foster good health behaviors in various contexts, including child survival and development (Higgs et al., 2014; Bright et al., 2018), access to primary health care (Lewin et al., 2008; Joshi et al., 2014), attendance of psychiatric out-patient services (Rajasuriya et al., 2010), vaccination and adherence to malaria treatment guidelines (Kurumop et al., 2013), and prenatal care attendance (Watterson et al., 2020). As for cervical cancer, observational studies have confirmed the greater effectiveness of organized population-based screening programs with invitations sent to the entire target population over opportunistic participation (Sepúlveda and Prado, 2005; Bleggi Torres et al., 2003; Sivaram et al., 2018). However, despite the documented benefits of invitation letters and reminders in stimulating participation in screening programs in HICs, there is practically no evidence based on randomized controlled trials conducted in LMICs (Rees et al., 2018).

There are good reasons to believe that the efficacy of invitation letters and reminders for cervical cancer screening may be considerably weaker in LMICs than HICs. Cultural concerns – shame (since the virus is transmitted mostly through sexual contact), stigmatization, religious taboos, superstition and the like (Ndikom and Ofi, 2012; Vu et al., 2018) – may undermine participation irrespective of invitation strategy. Screening may also be discouraged by lack of spousal support, distrust of the healthcare system for the treatment of serious illnesses, and fear of out-of-pocket expenses for treatment of the cancer (Ebu et al., 2015; Islam et al., 2017; Mabele et al., 2018; Vu et al., 2018).

This study reports the results of a randomized controlled trial carried out in May–July 2019 in the Republic of Armenia to evaluate the effectiveness of a set of low-cost invitation interventions in fostering women's participation in a national cervical cancer screening program. The Armenian healthcare system shares some key characteristics with low-income countries in general: a low level of public funding, poor financial protection, and high out-of-pocket costs that increase the likelihood of people with serious illnesses falling into poverty (Lavado et al., 2018). Cervical cancer turns out to be a major health and social burden in Armenia (Bruni et al., 2019). It is the second most common cancer among Armenian women aged 15 to 44, and the country has the second highest age-standardized incidence in Western Asia (8.4 per 100,000 women, lower only than the 9.8 per 100,000 in Georgia). The trial was conducted in the region of Shirak, the country's poorest, with a large share of the population below the national poverty line. These structural and economic features suggest that the results of the study may be applicable to a good many LMICs around the world.

The randomized controlled trial assesses the effects of three interventions with respect to the status-quo – an opportunistic screening regime where women aged 30 to 60 can show up voluntarily at their general practitioner (GP) for free screening once every three years:

  • i.

    Invitation letters to women's homes to invite them to screen at a given date;

  • ii.

    Reinforcing the invitation with follow-up reminders in proximity of the screening date;

  • iii.

    Ad-hoc manipulations of the framing of invitation letters and reminders.

As regards the manipulations, the health persuasion literature in psychology (Rothman and Salovey, 1997) offers the hypothesis, based on prospect theory (Tversky and Kahneman, 1981), that gain-framed messages are more effective in promoting prevention behavior (such as vaccination), while loss-framed messages are better at promoting detection behavior (such as screening). This prediction has gained some empirical support in the context of disease detection behaviors (O'Keefe and Jensen, 2009). A randomized controlled trial in Italy (Bertoni et al., 2020) shows that letters giving loss-framed information on the risks of non-participation in screening enhance take-up. This evidence suggests that negatively framed letters and reminders should outperform neutral ones in enhancing participation.

Further, it is amply documented that many individuals have altruistic preferences and are willing to sacrifice their own welfare to benefit others (Becker, 1976; Simon, 1993). Consequently, a woman receiving the invitation may go for screening in the interest of her family members. So an other-regarding frame manipulation may also activate guilt aversion, whereby people seek to meet the expectations of others (Battigalli and Dufwenberg, 2007; Charness and Dufwenberg, 2006). Both of these considerations imply that other-regarding invitation letters and reminders may be more effective than neutral ones in stimulating participation.

Based on these behavioral insights, our experiment adopts three frame types for the invitation letters and reminders:

  • i.

    A “neutral” frame, giving general information about the program and briefly stating that screening can be beneficial.

  • ii.

    A “negative” frame, offering salient information on the possible adverse consequences of forgoing the check-up.

  • iii.

    An “other-regarding” frame, which stresses the importance of the check-up for the sake of the woman's family members.

We find that the invitation letters substantially increase the program take-up. Moreover, the invitations are considerably more effective when reinforced by follow-up reminders. Our data on letter and reminder delivery also show that roughly 4 out of 10 invitations failed to get delivered; so better mail delivery may be expected to produce greater treatment effects. Finally, we found no evidence that the frame of the letter or the reminder has any effect on participation. We also estimate heterogeneous effects, finding that letter recipients in rural areas increased participation more than those in urban areas. That is, the program helped to narrow the urban-rural screening gap that characterized the control group.

Our findings demonstrated that simple, inexpensive changes to the organization of cancer screening program invitations can increase screening in LMICs significantly, with appreciable benefits for the reduction of cervical cancer mortality.

The rest of the paper is structured as follows. The institutional setting is described in Section 2 and the experimental design and procedures in Section 3. Section 4 introduces the data and Section 5 illustrates the empirical methods. Section 6 presents the results and Section 7 concludes.

Section snippets

The institutional setting

The Republic of Armenia is an upper middle-income country in transition in the South Caucasus, with per capita GDP of $4238 in 2018 (World Bank, 2019), 23.5% of the population below the national poverty line (Armstat, 2019), and the share of government expenditure on public health lower than the world average (Lavado et al., 2018).

In the Soviet era Armenia had a highly centralized healthcare system, the entire society having access to free medical assistance irrespective of social status,

Experimental design and procedures

We designed the experiment to test the effects of three strategies for increasing participation with respect to the status quo of opportunistic screening and soft advocacy campaigns.

The first strategy involves letters inviting a random sample of women targeted by the program to visit their GP for the Pap smear at closed-date, pre-booked appointments. Each woman receiving the invitation was given a one-week slot to visit the GP at any time she chose. Subjects were allocated randomly across

The data

We use three data sources. First, population-level registers of the target population were provided by the MoH. These population registers are based on the records of primary healthcare units. They are complete and of high quality, as primary healthcare is provided nationwide and the databases are updated regularly. As noted, the vast majority of the urban and rural population is registered with primary healthcare units, so the population registers have the most accurate and up-to-date data in

Empirical methodology

We compare Intention-To-Treat (ITT) effects on screening take-up of the various invitation strategies by estimating the following OLS linear regression:Screenedij=αj+β0Lij+β1Lij×NegLAbsRij+β2Lij×OthLAbsRij+γ0Lij×Rij+γ1Lij×Rij×NeuLNegRij+γ2Lij×Rij×NeuLOthRij++γ3Lij×Rij×NegLNegRij+γ4Lij×Rij×OthLOthRij++εij

In Equation (1), i stands for individual patient and j for GP. The dependent variable Screenedij is a dummy equal to 1 if individual i underwent screening and 0 otherwise. The explanatory

Main results

Fig. 2 reports the percentages of screened subjects by treatment group with 95 percent confidence intervals. Participation is considerably higher in each treatment group than in the control. Moreover, groups to which letters and reminders were sent show higher take-up rates than those with letters only. Finally, there are no sharp differences in participation either between treatment groups that received differently framed invitation letters only or between those that received differently

Conclusion

Promoting screening participation is a key factor in fighting cervical cancer. This holds true especially in low- and middle-income countries, where mortality from this neoplastic disease continues to be excessively high relative to high-income countries. Despite the potential benefits, large-scale screening programs are still quite scarce in LMICs, while participation in those that do exist remains very low. Accordingly, “novel methods for improving take-up and implementation of cervical

Author credit statement

The authors contributed equally to all the stages of the project.

Acknowledgements

We thank to the Ministry of Health of the Republic of Armenia and the Office of the Deputy Prime Minister for allowing us to run the experiment within the national cervical cancer screening program in the region of Shirak, the Republic of Armenia. We are also grateful to Health Project Implementation Unit of the Republic of Armenia for the valuable support throughout the whole course of the implementation of the experiment. The experiment would not have been possible without close collaboration

References (60)

  • S.M. Tavasoli et al.

    Impact of invitation and reminder letters on cervical cancer screening participation rates in an organized screening program

    Prev. Med.

    (2016)
  • M. Vu et al.

    Cervical cancer worldwide

    Curr. Probl. Canc.

    (2018)
  • Armstat

    Armenia demographic and health survey 2010. Calverton, MA, Armstat

  • Armstat

    Social snapshot and poverty in Armenia

  • P. Battigalli et al.

    Guilt in games

    Am. Econ. Rev.

    (2007)
  • G.S. Becker

    Altruism, egoism, and genetic fitness: economics and sociobiology

    J. Econ. Lit.

    (1976)
  • M. Bertoni et al.

    The good outcome of bad news. A field experiment on formatting breast cancer screening invitation letters

    Am. J. Health Econ.

    (2020)
  • EAST: Four Simple Ways to Apply Behavioural Insights

    (2014)
  • L.F. Bleggi Torres et al.

    Cervical cancer screening program of Paraná: cost‐effective model in a developing country

    Diagn. Cytopathol.

    (2003)
  • T. Bright et al.

    A systematic review of strategies to increase access to health services among children in low and middle income countries

    BMC Health Serv. Res.

    (2017)
  • L. Bruni et al.

    Human papillomavirus and related diseases in Armenia

  • N.G. Campos et al.

    The health and economic impact of scaling cervical cancer prevention in 50 low‐and lower‐middle‐income countries

    Int. J. Gynecol. Obstet.

    (2017)
  • G. Charness et al.

    Promises and partnership

    Econometrica

    (2006)
  • B. Chumworathayi et al.

    Can an appointment-letter intervention increase pap smear screening in Samliem, Khon Kaen, Thailand?

    Asian Pac. J. Cancer Prev. APJCP

    (2007)
  • K.M. Decker et al.

    Evaluating the effectiveness of cervical cancer screening invitation letters

    J. Wom. Health

    (2013)
  • S. Eaker et al.

    A large population-based randomized controlled trial to increase attendance at screening for cervical cancer

    Canc. Epidemiol. Prevent. Biomark.

    (2004)
  • N.I. Ebu et al.

    Knowledge, practice, and barriers toward cervical cancer screening in Elmina, Southern Ghana

    Int. J. Wom. Health

    (2015)
  • D.U. Ekwueme et al.

    Years of potential life lost and productivity costs because of cancer mortality and for specific cancer sites where human papillomavirus may be a risk factor for carcinogenesis — United States, 2003

    Cancer

    (2008)
  • J. Farrington et al.

    Prevention and Control of Noncommunicable Diseases in Armenia

    (2019)
  • E. Gakidou et al.

    Coverage of cervical cancer screening in 57 countries: low average levels and large inequalities

    PLoS Med.

    (2008)
  • Cited by (0)

    View full text