Effect of cervical cancer education and provider recommendation for screening on screening rates: A systematic review and meta-analysis

Background Although cervical cancer is largely preventable through screening, detection and treatment of precancerous abnormalities, it remains one of the top causes of cancer-related morbidity and mortality globally. Objectives The objective of this systematic review is to understand the evidence of the effect of cervical cancer education compared to control conditions on cervical cancer screening rates in eligible women population at risk of cervical cancer. We also sought to understand the effect of provider recommendations for screening to eligible women on cervical cancer screening (CCS) rates compared to control conditions in eligible women population at risk of cervical cancer. Methods We used the PICO (Problem or Population, Interventions, Comparison and Outcome) framework as described in the Cochrane Collaboration Handbook to develop our search strategy. The details of our search strategy has been described in our systematic review protocol published in the International Prospective Register of systematic reviews (PROSPERO). The protocol registration number is CRD42016045605 available at: http://www.crd.york.ac.uk/prospero/display_record.asp?src=trip&ID=CRD42016045605. The search string was used in Pubmed, Embase, Cochrane Systematic Reviews and Cochrane CENTRAL register of controlled trials to retrieve study reports that were screened for inclusion in this review. Our data synthesis and reporting was guided by the Preferred Reporting Items for Systematic Reviews and Meta-analysis (PRISMA). We did a qualitative synthesis of evidence and, where appropriate, individual study effects were pooled in meta-analyses using RevMan 5.3 Review Manager. The Higgins I2 was used to assess for heterogeneity in studies pooled together for overall summary effects. We did assessment of risk of bias of individual studies included and assessed risk of publication bias across studies pooled together in meta-analysis by Funnel plot. Results Out of 3072 study reports screened, 28 articles were found to be eligible for inclusion in qualitative synthesis (5 of which were included in meta-analysis of educational interventions and 8 combined in meta-analysis of HPV self-sampling interventions), while 45 were excluded for various reasons. The use of theory-based educational interventions significantly increased CCS rates by more than double (OR, 2.46, 95% CI: 1.88, 3.21). Additionally, offering women the option of self-sampling for Human Papillomavirus (HPV) testing increased CCS rates by nearly 2-fold (OR = 1.71, 95% CI: 1.32, 2.22). We also found that invitation letters alone (or with a follow up phone contact), making an appointment, and sending reminders to patients who are due or overdue for screening had a significant effect on improving participation and CCS rates in populations at risk. Conclusion Our findings supports the implementation of theory-based cervical cancer educational interventions to increase women’s participation in cervical cancer screening programs, particularly when targeting communities with low literacy levels. Additionally, cervical cancer screening programs should consider the option of offering women the opportunity for self-sample collection particularly when such women have not responded to previous screening invitation or reminder letters for Pap smear collection as a method of screening.


Objectives
The objective of this systematic review is to understand the evidence of the effect of cervical cancer education compared to control conditions on cervical cancer screening rates in eligible women population at risk of cervical cancer. We also sought to understand the effect of provider recommendations for screening to eligible women on cervical cancer screening (CCS) rates compared to control conditions in eligible women population at risk of cervical cancer.

Methods
We used the PICO (Problem or Population, Interventions, Comparison and Outcome) framework as described in the Cochrane Collaboration Handbook to develop our search strategy. The details of our search strategy has been described in our systematic review PLOS  Introduction Globally, 485,000 new cases of cervical cancer and 236,000 deaths due to cervical cancer occurred in 2013, ranking cervical cancer among the top 10 cancers in incidence and mortality globally. [1] The age-standardized incidence rate (ASIR) for cervical cancer is much lower in developed nations at 5.0 per 100,000 compared to developing nations at 8.0 per 100,000. [1] Similarly, the age-standardized death rate (ASDR) for cervical cancer is lower in developed nations at 2.2 per 100,000 compared with developing nations at 4.3 per 100,000. [1] In fact, surveillance data on worldwide cancer survival shows wide variation between nations, and these data have been used as a metric of the effectiveness of health systems in cancer prevention, control and treatment. [2] For instance, a systematic analysis of breast and cervical cancer in 187 countries between 1980 and 2010 found that developed countries with comprehensive cancer screening programs have recorded sustained declines in cervical cancer incidence and mortality while many developing countries in sub-Saharan Africa have experienced upsurges in new cases [3]. Even though there are ongoing efforts to increase human papillomavirus (HPV) vaccinations for primary cervical cancer prevention, early detection of precancerous cervical lesions through screening remains a critical health care service intervention for reducing cervical cancer incidence and mortality particularly in low-resource settings where HPV vaccination coverage is poor. [4] In comparison to developing countries with poor vaccination coverage and lack of organized cervical cancer screening programs, developed countries with well-organized cervical cancer screening programs have gained significant reduction in cervical cancer incidence and mortality. [2,[5][6][7][8][9] Indeed, since the introduction of the Papanicolaou smear cytology testing in the 1950s and 1960s, cervical cancer incidence and mortality have declined in the United States with organized cervical cancer screening programs and screening rates of 83%. [10][11][12] However, Cervical cancer remains a huge burden in developing countries where cervical cancer screening rates are currently low, ranging between 6-8% [13,14] These disparities in screening rates and HPV vaccination coverage might explain the differences in incidence and mortality associated with cervical cancer in different regions around the world. The epidemiologic link between high-risk human papillomavirus types and cervical cancer have led to the development of novel screening modalities such as testing for high-risk human papilloma virus (HPV testing) screening recommended by the World Health Organization (WHO) and the European Guidelines for Quality Assurance for Cervical Cancer Screening. [5,15] Human papillomavirus testing has proven effective in detection of precancerous cervical lesions particularly in population-based cervical screening programs. [4,[16][17][18][19][20] Although the recommended screening modalities for cervical cancer have contributed to a significant reduction in cervical cancer incidence and mortality due to cervical cancer, the benefits of cervical cancer screening are yet to be fully realized in countries with poorly organized screening programs for women at risk. It is also noteworthy that even in countries with organized screening services, these benefits are not maximized in underserved, uninsured and under-represented populations due to factors such as cost, access problems, anxiety, discomfort with the screening procedure, and fear of cancer or poor health literacy, all of which contribute to poor outcomes for cervical cancer. [21][22][23][24][25] Building health care systems that can address multiple factors simultaneously would improve cervical screening rates and overall outcomes for cervical cancer in populations at risk for this preventable cancer.
Previous reviews [11,26,27] on interventions to increase delivery and uptake of cervical screening have documented the effectiveness of provider reminders and invitation letters on uptake of cervical cancer screening. One of these reviews [11] focused on a range of interventions including invitations, reminders, education, message framing, counseling, risk factor assessment, procedures and economic factors. They found a significant positive effect of invitation letters on uptake of cervical screening. The review also found limited evidence to support educational interventions, but unclear on what format of educational intervention is most effective. Therefore the goal of this systematic review was to better understand the current evidence on the effect of cervical cancer education as an intervention to improve cervical cancer screening rates in women who are eligible for cervical cancer screening. We also sought to review the evidence of the effectiveness of provider recommendations for cervical cancer screening on screening rates in women at risk for cervical cancer.

Methods
Types of studies considered: In this review we considered randomized control trials, cluster randomized control trials and quasi-experimental designs of relevant interventions to increase cervical cancer screening in women at risk of cervical cancer. We included studies published through August 2016. There was no restriction on language, region, or country of study. The review protocol was published in the International Prospective Register of Systematic Reviews (PROSPERO) with registration number CRD42016045605, which is available at http://www. crd.york.ac.uk/prospero/display_record.asp?src=trip&ID=CRD42016045605.
Types of Participants: All women eligible for participation in a cervical cancer screening program, including women with no prior screening for cervical cancer and women due or overdue for screening visits in various settings.
Types of interventions: In this review, we focused on 2 main types of interventions used to improve cervical cancer screening rates: 1. Cervical Cancer Education. We included studies on any educational interventions aimed at increasing the participants' knowledge about cervical cancer (causes, importance of screening, how screening is done and where to have screening done, including interpretation and treatment of abnormal screening tests). Educational interventions that are theorybased were considered. We also included non-theory-based education interventions such as didactic health talks. These educational interventions could be mediated through videos, use of culturally sensitive educational materials, letters with fact sheets on cervical cancer and screening, cervical cancer screening brochures, and call or text-message mediated education. We examined the effect of these interventions singly or in combinations in various settings where the interventions were implemented.

Provider Recommendation.
We included studies on interventions initiated by health care providers/health facility or screening programs aimed at encouraging eligible women to accept screening or to comply with screening guidelines set by the screening program. These interventions include provider initiated screening during opportunistic encounters with eligible women in a health facility setting, invitation letters from a health facility/ screening program to eligible women with no prior screening or due for screening. We also included interventions such as reminder letters, phone calls, direct mailing of individualized letters or text-messaging to eligible women with screening past due. We also included interventions such as options for self-sample collections for HPV testing.
Comparison: Control conditions or routine standard screening practice in the setting.

Primary outcomes
The primary outcome measure of effectiveness was the proportion of eligible women exposed to the intervention or control who completed cervical cancer screening during the trial. In other words, cervical cancer screening rate was defined as the number of eligible women exposed to an intervention or control condition who had a screening during the intervention divided by the total number of women exposed.

Conceptual model for improving cervical cancer screening
The conceptual model guiding this review is adapted from the social ecological model (SEM) of health promotion proposed by the Centers for Disease Control and Prevention (CDC) for implementation of the National Breast and Cervical Cancer Early Detection Program (NBCCEDP).
[28] This conceptual model emphasizes the interplay of individual, interpersonal, organizational, community, and policy-level interventions in increasing breast and cervical cancer screening in at risk population. The aspects of this model most relevant to this review include: individual, interpersonal, organizational and the community bands of the SEM. Each of these bands are briefly described below: Individual: represented by the innermost band of the SEM rainbow refers to eligible women who need cervical cancer screening and will benefit from education on knowledge of cervical cancer risk, benefits of screening and how to access screening services.
Interpersonal: this band surround the individual band of the SEM and represents cervical cancer prevention activities implemented at the interpersonal level intended to facilitate individual behavior change by affecting social and cultural norms and overcoming individual-level barriers. In this review, health care providers, community health workers or promotoras, and patient navigators represents potential sources of interpersonal messages and support. Some of the relevant interventions appropriate for this level include: providers making screening recommendations to their patients, sending reminders about need for screening and patient navigators helping with logistical support and removing other barriers to screening.
Organizational: this band surrounds the interpersonal band of the SEM and represents screening activities initiated at the organizational levels (screening health facility or screening program). One of the activities at this level relevant to this review is the use of client and provider reminder systems to encourage recommendation and use of cervical cancer screening services.
Community: use of peer-educators and culturally-sensitive communication and education materials to encourage participation in cervical cancer screening activities.

Data synthesis
The synthesis and reporting of our findings was guided by the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-analysis) statement. [30] In this review, we did a qualitative synthesis of studies for which statistical pooling was not appropriate. Qualitative synthesis entailed a brief narrative of the types of intervention, setting, country, eligible study population, main outcomes and a summary of the intervention effects and confidence intervals for each study report. Where feasible, statistical pooling of the effects of individual studies was done with meta-analysis using the RevMan 5.3 Review Manager software. The Higgins I 2 statistic was used to assess for heterogeneity in studies pooled. Relevant forest plots were generated by RevMan 5.3 for graphic display of the individual study effects and the overall summary effect of the interventions on cervical cancer screening rates. We used odds ratio and random effects models to generate all statistical estimates of the individual and combined study effects of interventions in meta-analyses. Publication bias was assessed using funnel plots generated by RevMan 5.3. The details of the items reported in this review are included in the PRISMA 2009 checklist in S3 Appendix.

Risk of bias assessment and quality grading of studies included
The risk of bias for each study was assessed either as low, unclear, or high risk for each of the following criteria: selection bias, performance bias, detection bias, attrition bias and reporting bias as described in the Cochrane Handbook.
[29] The assessment of the quality of included studies was based on the criteria of risk of bias, inconsistency, indirectness, imprecision and reporting bias as described in the GRADE Quality Assessment Checklist. [31]

Results
Our search yielded 4371 published articles (2101 in Pubmed, 1931 in Embase, 116 in Cochrane Systematic Reviews and 223 in Cochrane CENTRAL register of controlled trials). After removing duplicate publications, we had 3072 study reports for screening. After screening study titles/abstracts we found 73 potentially relevant articles for full-text review and consideration for inclusion, and 2999 were discarded because they did not meet the criteria for further review of full-text. After completing full-text review, 28 articles were found to be eligible for inclusion in qualitative synthesis, 5 of which were included in meta-analysis of educational interventions and 8 combined in meta-analyses of HPV self-sampling interventions, while 45 were excluded for various reasons (Fig 1).
For the two questions covered in this review, we included 28 studies (26 RCTs and 2 quasiexperimental design) involving a total of 241,219 participants from 15 countries (Australia, Belgium, Canada, Finland, France, Germany, Italy, Japan, Kenya, Malaysia, Mexico, Sweden, Taiwan, Thailand, and USA) on 5 continents (Africa, Asia, Australia, Europe and North America).
Seven of these papers [32-38] were included in assessing the effectiveness of cervical cancer education on cervical cancer screening rates. Twenty-one [39-59] were eligible for inclusion in assessing the effectiveness of various aspects of provider screening recommendations on cervical cancer screening rates. The study reports on provider recommendations assessed interventions such as phone call reminders, invitation letters, reminder letters, appointment letters, and self-sampling for HPV testing. Cervical cancer education, provider recommendation and cervical cancer screening rates What is the effect of cervical cancer education on cervical cancer screening rates?
To address the question of the effect of cervical cancer education on cervical cancer screening rates, our search strategy yielded seven (six RCTs and one community-based participatory RCT) studies. Two studies [33, 37] were excluded from statistical pooling of the overall effect because of variations in methodology that contributed to substantial heterogeneity. The other five studies [32,[34][35][36]38] were pooled together in meta-analysis involving a total of 797 women who were exposed to cervical cancer education and 812 women in the comparison group. Our meta-analysis results presented in Fig 2 found evidence of an increase in cervical cancer screening rates in women exposed to the intervention compared to the controls. The pooled summary effect of the interventions included was two and a half times higher in comparison to the control (OR = 2.46; 95% CI: 1.88, 3.21).
What is the effect of provider recommendation for screening on cervical cancer screening rates?
For the question regarding the extent to which provider recommendations for cervical cancer screening increases screening rates, our search found 21 studies  in which there were 19 RCTs and two quasi-experimental studies. There were subtle differences in implementation of the various interventions, such as combining invitation letters with phone call reminders and some educational messages; appointment letters and reminder letters with educational messages. These differences limited statistical pooling of the individual effects of these interventions in meta-analyses. However, we found a trend toward positive effects of the various provider-based interventions on cervical cancer screening rates.
First, we found 4 RCTs that assessed the effectiveness of phone call contact and other outreach modalities to increase CCS rates in women who were either due or overdue for Pap test screening in various settings. [39-41, 55] Only one of these trials [40] found no significant difference in Pap smear screening uptake in women who received a telephone call reminder compared to a mail letter reminder (6.5% vs 5.8%) among women who initially did not respond to an invitation for a Pap smear screening. The other three RCTs showed consistent evidence of a significant increase in CCS rates among women who were exposed to the telephone outreach/ recall/reminder group compared to other outreach modalities or usual care. [39, 41] The CCS rates were 34.4% in the phone contact group compared with 18.8% in the usual group, with significantly higher odds of women returning for screening when contacted by a direct phone compared to a personal letter (OR, 2.38, 95% CI: 1.56, 3.62).
[39] Similarly, the CCS rates among women who received a phone call reminder for not having a Pap test in the previous 3 years was 41.4% compared with 10.0% in the usual care group.
[41] Also, an RCT testing the real-world effectiveness of various outreach modalities found CCS rates in the control group were 21.4% vs 24.5%, 25.5%, 29.2%, and 36.1% respectively, in the letter, email, telephone and multimodal outreach groups.
[55] Compared to women who received usual care, those in the multimodal (AOR 2.3, 95% CI: 1.4, 3.6) and telephone (AOR 1.7, 95% CI: 1.1, 2.8) groups were more likely to receive a Pap test during the follow-up period. In addition, the telephone and multimodal interventions significantly reduced median time to Pap screening. [55] The second group of provider interventions that are potentially useful for cervical cancer screening policy decision making on are related to either invitation or reminder letter/message to eligible women for screening. Our search found 6 RCTs [42, 44-47, 56] and 1 quasi-RCT [43] that reported the effectiveness of invitation and reminder interventions on CCS rates in various settings. We found a consistently positive effect of various modes of invitation and reminder systems on CCS rates. One of these trials reported participation of 5.9% in women who received an invitation letter to screen, which was significantly higher than the 3.1% CCS rate in the control group.
[42] After adjusting for other variables, women who were sent an invitation letter were significantly more likely to have had a Pap test within 6 months of the intervention than women in the control group (OR 2.6; 95% CI: 2.09-3.35). Another study investigated different models of invitation on CCS rate in a randomized population-based cohort in Germany and found significant differences in the proportion of women who received either invitation letter or invitation letter and information brochure compared to women who did not receive an invitation (91.8% versus 85.3%, p value <0.001; adjusted OR 2.69, 95% CI: 2.15, 3.37).
[56] The effect of these invitation letters was more profound in women who were older, had lower education and migrant women. [42, 56] Other trials also found a significantly higher net gain in screening rates (OR = 1.19; 95% CI: 1.14, 1.24) [43] when invitation letters were sent to women who have not had Pap smear screening in the past 30 months, particularly among older women.
[44] Reminder letters given to patients and creating a reminder system for physicians significantly increase cervical screening rates more in women who have not had a previous Pap screening test (OR = 1.39; 95% CI: 1.02, 1.89).
[46] Although one of the trials [45] did not find a significant difference in cervical screening rates in women sent a reminder letter after an initial invitation letter compared to women with no reminder letter (10.7% vs 6.3%), most of the studies found evidence of significant effects of reminders delivered to women through various modalities as a strategy to improve cervical cancer screening rates. Furthermore, one study [46] noted that once a primary care visit takes place, the behavior of the primary care provider with respect to recommending a screening test becomes an important determinant of cervical cancer screening use by eligible patients. Additionally, a trial among under-screened women randomized into a reminder letter group versus a no-letter group found a letter/no-letter Pap test rate ratio of 1.53; 95% CI: The third group of provider interventions potentially useful for policy decision making in our review were those in which appointment letters stating the screening visit dates were sent to eligible women compared to women with no appointment letters (44.7% vs 25.8% screening uptake, respectively) [48]; and provider recommendations offering to screen eligible women when they present in urgent care settings compared to referral to a gynecology clinic for screening (84.7% vs 29.0% screening uptake, respectively). [49] The fourth group of provider interventions potentially useful for policy decisions in cervical cancer screening programs are those offering eligible women the option for HPV self-sampling. Our search found eight trials that reported the effectiveness of these interventions in increasing CCS rates in various settings. [50-54, 57-59] The individual effects of these trials involving 22,256 women who were offered the option for HPV self-sampling as an intervention, and 18,312 women in the comparison group on CCS rate were pooled in meta-analysis. We found an overall summary effect of almost a twofold higher likelihood of having a CCS in women exposed to the intervention compared to the comparison, OR = 1.71, 95% CI: 1.32, 2.22 (Fig 3). The funnel plot in Fig 4 did not suggest evidence of publication bias in the studies included in this meta-analysis.

Risk of bias assessment and quality grade of studies included
Except for six of the included studies [32, 34, 36, 42, 43, 48] judged to have high-risk of bias and graded as low quality, the studies included in this review and meta-analysis were judged to have low-risk of bias with moderate to high quality grade. The details of characteristics of each study, risk of bias assessment for each study included, and reference list of studies included and excluded is available in S2 Appendix.
The summary of the studies included in this review is presented in Table 1.
The following studies  were excluded at the full-text review stage for specific reasons. The reasons for exclusion have been summarized in S1 Appendix in the section on characteristics of excluded studies.

Discussion
The principal findings of this review are that theory-based educational interventions and use of culturally sensitive languages in communities with low participation rates for cervical cancer screening are effective interventions that significantly improve cervical cancer screening rates. The pooled effects of five studies (see Fig 2) on cervical cancer educational interventions showed an overall effect of 2.5 times higher likelihood for women in the intervention groups to have a CCS compared to women in the comparison groups. We also found that invitation letters to women either alone or with a follow up telephone reminder significantly increased CCS rates in various screening populations. Additionally, we found that offering options for self-sample collection for HPV testing increased the likelihood of women completing a CCS by almost two-fold compared to women who received a reminder invitation for Pap test screening, particularly among unscreened and under-screened women and among non-compliant women who have not responded to prior invitations for Pap smear screening. [50,52] Cervical cancer education One of the effective, theory-based educational interventions within the studies reviewed was guided by the social cognitive framework. This theory posits that knowledge of health risks and benefits creates the precondition for change and if people lack knowledge about how their lifestyle habits affect their health, they have few reasons to put themselves through the travail of changing those detrimental habits. [100,101] Additionally, the Health Behavior Framework, which emphasizes that individual and health care system factors and environmental and personal barriers jointly determine health behaviors, was used in designing an educational intervention to increase cervical cancer screening rates among Samoan women.
[36] These theorybased educational interventions are particularly relevant for developing communities with low literacy levels as was demonstrated in the intervention communities of the studies in this review. Our findings showed a consistent positive effect of the use of theory-based, culturally and linguistically-sensitive, community-participatory modeled educational interventions. These interventions increased women awareness, knowledge of cervical cancer, importance of screening, and offered barrier counseling and guidance with scheduling screening appointments thereby increasing the overall likelihood of eligible women to have Pap smear screening.
[32, 34-36, 38] Based on the quality assessment of these trials, we have confidence in the findings and recommend that educational interventions to increase participation of women in cervical screening programs should be based on theory and use of culturally sensitive language tailored to specific communities. Delivering didactic health talks could increase women's awareness and knowledge of cervical cancer, but does not necessarily translate to increased cervical screening rates, as found in one of the trials in rural Kenya.
[33] Invitation letter, appointment letter, and phone calls Our findings suggest that strategies utilizing a combination of invitation letters, including an information pamphlet on cervical cancer and Pap test and additional telephone reminders with a short description of the importance of the Pap smear test, demonstrated a positive effect on cervical screening rates.
[41, 44] The critical role of a reminder phone call compared to invitation letter alone was demonstrated in one trial, which found a significant effect on screening rates in women due for a follow up Pap smear.
[39] Indeed, a prior trial on the effectiveness of a call/recall system in improving compliance with cervical cancer screening found that a letter of invitation alone was not enough to encourage women who have never or have infrequently undergone a Pap test to come for cervical cancer screening, and more aggressive follow up efforts with phone reminders and offering screening on specific appointment dates might be required to improve screening rates in such populations.
[45] However, the application of these findings will depend on the setting. For example, screening programs targeting hard-to-reach women in rural areas with poorly organized postal systems may find the use of a telephone strategy more feasible than a mailed invitation letter. Sending invitation letters may be more applicable in settings with well-organized postal systems, as supported by the trial done in Manitoba, Canada where invitation letters were sent to unscreened women using a forward sortation area and postal codes for the community.
[42] Although, there was a significant increase in screening rates in the communities targeted with the invitation letters compared to the control community, the authors cautioned that literacy could be a potential limitation on the effectiveness of letters [42], perhaps supporting the strategy of adding a phone call contact.
[39] A phone call has the advantage of providing direct communication with the participants, and this could help in building confidence and motivation for the screening test. The phone call also serves as a reminder strategy for women who have not initially responded to a screening invitation letter.
[40] Personal contact through a phone call might be important, especially for women who feel anxious about the examination or the Pap smear. Also, the possibility to have the Pap smear taken by the person to whom the women talked may further increase motivation for screening.
[41] We also suggest for further study to explore how the use of social media such as Twitter and FaceBook may improve delivery of educational messages and women participation in cervical cancer screening.

Self-sampling on screening rates
Our findings that offering the option of self-sampling for HPV DNA testing increases CCS suggests that if women have the required information on HPV testing, educational guides on how it is done, and are offered the option to self-collect vaginal samples for the HPV test, cervical cancer screening programs could significantly improve women participation and screening rates. Self-sampling helps remove potential barriers for women participating in screening programs, such as fear of discomfort during pelvic examination and concerns with privacy. Indeed, the findings in one of the studies suggests that in a population of eligible women who have not attended a primary screening invitation, self-sampling rather than a reminder invitation letter could potentially increase cervical cancer screening rates. [52]

Strengths and limitations of this review
The main strength of this review is the comprehensive search of the literature with involvement of a research librarian (L.O.) who ensured access to full-texts of all study reports we screened for eligibility and inclusion in the review. Additionally, our review was guided by a published systematic review protocol. Our major limitation is that we did not collect secondary outcome data on the cost of cervical cancer screening tests, health insurance coverage and how these variables contributed to the screening rates in women of various socio-economic status, age, and geographic settings. These factors should be considered in future reviews.

Comparability of our review findings with others
Our findings are consistent with the Cochrane review reported by Everret, et al [11] which found that invitation letters are effective interventions that increase the uptake of cervical cancer screening in women. In addition, our review demonstrated that a telephone reminder after an initial invitation letter had a substantial effect on cervical cancer screening rates. Our findings also provided conclusive evidence on the effectiveness of theory-based cervical cancer education at increasing cervical screening rates. In the previous review [11], though there was limited evidence of the effect of educational interventions on uptake of screening, it wasn't clear which format of education is most effective. [11] Our systematic review and meta-analysis showed that theory-based, culturally and linguistically-sensitive educational interventions administered by lay health advisors consistently demonstrated significant positive improvements in cervical cancer screening rates. Recent reviews by Cam, et al [102,103] found that group education involving presentations from physicians, lay-health advisors, or cancer survivors, and reducing structural barriers such as providing sign-ups for screening appointments at events, or providing transportation were evidence-based strategies that promote cancer screenings. We did not find any prior systematic reviews on the effectiveness of self-sampling collection in promoting cervical screening rates. Our review however, showed a consistently significant positive impact of this intervention at increasing cervical screening rates, particularly in women who had initially not responded to a Pap smear screening invitation.

Conclusions, implications for policy and future research
Our findings contribute to the literature supporting the implementation of theory-based cervical cancer educational interventions to increase women's participation in cervical cancer screening programs, particularly by targeting communities with low literacy levels. Indeed, a review of factors influencing cancer screening practices of underserved women [104] found intrinsic motivators for screening related to beliefs and perceptions of vulnerability, such as ignoring cervical cancer screening when no symptoms were present, believing that not knowing if one had cervical cancer was better, and thinking that only women who engage in sexual risk-taking behaviors need to obtain Pap smear testing. [104] Theory-based guided cervical cancer educational interventions such as social cognitive theory and the health belief framework target these constructs and help communities and women to make positive health decisions and take action toward acceptance and completion of screening activities. Provider recommendation interventions, such as invitation letters with follow up phone call reminders, are efforts worth investing in to achieve a significant improvement in screening rates. Implementation of novel sample collection methods such as self-sampling by women and creating reminder mechanisms for providers to initiate testing during opportunistic encounters in the health care setting may yield additional gains in screening rates. This evidence should be utilized to develop specific resource-setting guidelines for increasing CCS rates in developed and developing countries. For instance, utilizing theory-based cervical cancer education with culturally-sensitive language by lay health workers may yield better screening participation in underdeveloped settings with low literacy levels. Also, utilizing various provider recommendations should be guided by the unique characteristics of the population targeted as discussed earlier.
One area that merits further research is to conduct randomized control trials to better understand the independent effect of provider recommendation intervention variables such as invitation letters, phone calls, appointment letters, reminder letters, and self-sample collection methods on cervical cancer screening rates after adjusting for the effect of education. Most of the studies included in this review did not tease out the direct and indirect effect of education, making it difficult to understand whether or not provider recommendation interventions had their effect mediated through knowledge or education, and what the size and strength of these effects were with or without education as a factor. Conducting further studies with robust statistical modeling such as mediation and moderation regression analyses are also a future area worth considering. Additionally, the use of mobile communication technologies to deliver culturally-and linguistically-sensitive cervical cancer education and understanding the settings where these may work best are potential areas for future research.