Interventions to improve cervical cancer screening uptake amongst young women: A systematic review

Abstract

Objectives. In view of declining screening uptake in young women, this review aims to summarise the available evidence relating to interventions designed to increase cervical screening uptake amongst women aged ≤ 35 years.

Methods. Electronic databases were searched and further articles located by manual searches. Study designs employing a valid comparison group and including women aged ≤ 35 years published through 2012 were considered. Data was extracted on the uptake from either screening programme statistics or as reported by the study subjects. A narrative synthesis was undertaken for each category of interventions identified.

Results. Ninety-two records were screened with 36 articles retrieved for further assessment. Four studies met the inclusion criteria, two of which evaluated more than one intervention. One of the studies evaluated the use of a modified invitation letter and reported no significant increase in uptake compared to a standard invitation. Three studies investigated the use of a reminder letter, with two reporting a positive effect on screening uptake in women aged 24–34. Three studies were included which supported the use of physician and telephone reminders. One study on HPV self-sampling reported a positive effect when compared with a reminder letter.

Conclusions. There is a lack of randomised controlled trials designed to specifically address falling cervical screening uptake in amongst young women. Cervical screening programmes need to look beyond the use of invitation/reminders letters in this group of women to develop interventions which attempt to overcome as many barriers to uptake as possible.

Coverage rates for cervical screening have become noticeably lower in younger women as observed in several developed countries including UK and Sweden [1,2]. In UK, for example, during the period 2008/2009, only 61% of women aged 25–29 were recorded as having an adequate test within the last 3.5 years [3]. This was a slight increase from 59% in 2007/2008 [4]; however, the coverage rates for this age group are consistently lower than those of older women – 78% of women aged 45–49 had an adequate test recorded in the last 3.5 years in 2008/2009 [3]. A common underlying cause for this trend has yet to be identified [2]. A fall in coverage amongst young women is of particular concern as the greatest benefit of cervical screening is gained by women who are screened from a young age [5] and having previous tests recorded has been shown to make women more likely to attend for screening in the future [6].

This trend of falling participation in younger women must also be considered in relation to the increasing understanding of the link between human papilloma virus (HPV) and cervical cancer [7,8]. The introduction of HPV vaccines to the European market in 2006–2007 and consequent media attention on HPV and cervical cancer changes the situation with regard to young women in particular. The importance of HPV-vaccinated women continuing to participate in cervical screening has been emphasised since the vaccines only protect against HPV types 16/18 [9]. It is not yet clear how a better understanding of HPV and the HPV vaccines will affect screening attendance, however, there is a growing concern that vaccinated women might feel they do not need to be screened. In a 2011 qualitative study from the UK including teenage girls and their parents, there was a lack of understanding both among the teenagers and their parents that cervical screening would be required irrespective of vaccination status [10].

In light of the lower participation and the new situation induced by the HPV vaccination it is becoming increasingly important to investigate ways of reversing the trend of declining uptake and find solutions both for young women who have not been offered HPV vaccination and for future vaccinated cohorts who may be increasingly less likely to attend for cervical screening.

State of the current knowledge

The reasons why women of different ages do not attend for screening are well documented [6,11–16]. Barriers reported in studies with young women were, e.g. frequent changes of address and not registering with healthcare providers, competing time demands, low risk perception, inadequate knowledge about cervical cancer and prevention, fear of pain, discomfort and embarrassment and general procrastination [16–19]. In a Swedish focus group study from 2011, 30-year-old women stated that cancer was not a disease affecting young people, which made them not prioritise cervical screening [15]. In the UK, the NHS Cancer Screening Program annual review, women aged 25–34 are described to be more likely to be embarrassed about attending and to think the test might be uncomfortable or painful than older women (aged 35–45) [20]. Waller et al. described how practical barriers were raised more frequently by younger women, with older women appearing to have more negative attitudes to cervical screening [16].

Previous systematic reviews of interventions to increase cervical screening uptake across the target population as a whole have concluded that invitations [21,22] and educational interventions [22] appear to be effective. Another review suggests that community-based strategies may be successful if a combination of mass media campaigns and educational information targeted at individual women is employed [23]. It must, however, be noted that the previous reviews are not specific to younger women and it remains unclear whether such interventions will be effective in this age group. An increase in the number of tests recorded in 2008/2009 in UK, when an additional 500 000 screenings were recorded, was attributed to the death of a young media celebrity from cervical cancer [24]; an event which raised awareness of the disease nationally, however a media phenomenon like this is a rare occurrence and planned interventions need to be developed to ensure increased and sustained uptake [25].

This study is the result of a collaborative project between research groups in Manchester and Stockholm aiming to better understand ways to increase cervical screening participation in young women. The review was undertaken prior to conducting a cluster randomised trial to increase cervical screening uptake in women aged 25 in UK, who are receiving their first invitation from the NHS Cervical Screening Programme [26]. This systematic review aims to summarise the available evidence relating to interventions that improve screening uptake specifically in younger women, in a narrative synthesis.

Methods

The Database of Abstracts of Reviews of Effects (DARE), The Cochrane Library, Health Technology Assessments database, Bandolier, Medline, EMBASE and CINAHL were searched to retrieve relevant articles using search terms adapted from a previous Cochrane Review [27] (see Supplementary Appendix 1, available online at http://informahealthcare.com/doi/abs/10.3109/0284186X.2013.869618, for the search strategy). A search of the grey literature was also performed. All databases were searched up until the end of 2012. The reference lists of included studies were also manually searched to identify further relevant articles. Full text articles were only retrieved for assessment if they were written in English.

All study designs were considered for inclusion provided that they incorporated a valid comparison group, i.e. concurrent or consecutive cohorts. Studies employing a retrospective comparison were also considered suitable for inclusion. Studies which included women aged 35 years and under were eligible for inclusion, regardless of whether women were attending for first tests, routine recalls, early repeats or overdue tests. Thirty-five years and under was chosen as the age group of interest in line with previous reports of the lowest coverage rates being observed in this group [1–2,28]. Studies were excluded if it was not possible to extract results specific to women aged 35 and under from the published data.

All interventions designed to increase cervical screening uptake were considered, and for the purposes of the analysis were split into three categories based upon the types of interventions assessed in the articles that met the inclusion criteria: 1) modified invitation letters (is an invitation letter containing additional information); 2) issuing reminders of overdue tests (via letter or telephone call, or by a physician); and 3) HPV self-sampling, that is a sample collected for HPV testing by the women, instead of a physician. Uptake of cervical screening collected from routine screening programme statistics or self-reported by the study subjects was used as the outcome measure in all included studies.

If studies reported more than one intervention, data were extracted and included in all relevant categories. A narrative synthesis was undertaken for each category of intervention to compare the effects of each on cervical screening uptake [29]. Further quantitative and subsequent meta-analysis was not appropriate due to the limited number of studies, and the heterogeneous nature of their study designs and the interventions under evaluation.

Results

Figure 1 shows the number of studies that were identified through both electronic database searching and manual searches of references. Ninety-two records were screened with 36 full text articles subsequently retrieved for assessment. Four studies were included in the narrative synthesis with the remaining 32 excluded for varying reasons. The majority of studies (20/32; 63%) were excluded because, although young women comprised a proportion of the overall study population, data-specific to young women could not be extracted. In six of the papers (19%) insufficient demographic data were given to determine whether young women comprised a proportion of the study population. The remainder of the studies were excluded due to: 1) no young women included in the study; 2) inadequate clinical endpoints utilised; and 3) failure to use a control group.

Figure 1. PRISMA flow diagram.

Table I summarises the characteristics of the included studies. Three studies evaluated the effect of either an invitation to participate in screening or a reminder about overdue tests, whilst one evaluated HPV self-sampling. These four studies had randomised controlled designs, which should permit more reliable estimates of the effects of the interventions than comparative or consecutive cohort studies.

Table I. Characteristics of included studies.

Study	N (N aged 25–34)	Population	Study design	Interventions	Comparison	Location
Eaker et al., 2004 [30]	12 240 (2941 aged 24–29 received intervention)	Women aged 25–59	Randomised controlled trial	1. Modified invitation letter 2. Reminder letter after first intervention 3. Phone reminder after reminder letter	1. Standard invitation letter 2. No reminder letter 3. No phone reminder	Uppsala county, Sweden
Morrell et al., 2005 [32]	90 000 (34 397 aged 20–34)	Women aged 20–69	Randomised controlled trial	Reminder letter	No reminder letter	New South Wales, Australia
McDowell et al., 1989 [31]	2034	Women aged 18–35	Randomised controlled trial	1. Physician reminder 2. Letter reminder 3. Phone reminder	1. No physician reminder 2. No letter reminder 3. No phone reminder	Ottawa, Canada
Virtanen et al., 2011 [33]	8699 (1557 aged 30–34)	Women aged 30–60	Randomised controlled trial	HPV self-sampling	Reminder letter	Espoo, Finland

The effect of invitation letters on cervical screening uptake

Only one study [30] meeting the inclusion criteria was found which evaluated the effect of modified invitation letters on cervical screening uptake in young women (Table II). This Swedish study found no difference in the proportion of women attending for screening after receiving a modified invitation letter explaining the cervical screening test in more detail, compared to those who received an unmodified routine invitation letter. Overall, the study found a borderline increase in attendance amongst all women in the study after receiving the modified invitation letter, however this effect was not seen in younger women alone, with a difference of −2.0% (95% CI −5.0–1.0%) being observed. It is not clear why the intervention did not increase uptake in younger women. In the absence of a group who did not receive an invitation the effect of standard invitation letters compared to no invitation could not be evaluated.

Table II. The effect of invitation letters, reminder letter and HPV self-sampling on cervical screening uptake in young women.

Study	Comparison	Outcome measure	N	Results
Invitation letters
Eaker et al., 2004 [30]	Modified vs. standard invitation letter	Difference in proportion of women attending screening after intervention compared to reference class	2941 received intervention	2.0% less women attended for screening in the intervention group (95% CI 25.0–1.0)
Reminder letters
Eaker et al., 2004 [30]	Reminder letter vs. no reminder letter	Difference in proportion of women attending screening after intervention, compared to reference class	2271 women received intervention	7.9% more women attended (95% CI 5.4–10.4)
Morrell et al., 2005 [32]	Reminder letter vs. no reminder letter	Pap test rate ratio in women receiving letter compared with women not receiving the reminder letter	22 961 women received intervention	Age 20–24: pap test rate ratio 0.98 (95% CI 0.72–1.35); 4.02% vs. 4.09% Age 25–29: pap test rate ratio 1.48 (95% CI 1.20–1.82); 3.83% vs. 2.58% Age 30–34: pap test rate ratio 1.39 (95% CI 1.15–1.68); 3.60% vs. 2.59%
McDowell et al., 1989 [31]	Letter reminder vs. no reminder	Percentage of women screened compared to control group	293 women received intervention	25.9% (76/293) of women were screened compared to 13.7% (35/255) of controls
Eaker et al., 2004 [30]	Phone reminder vs. no phone reminder	Difference in proportion of women attending screening after intervention, compared to reference class	445 women received intervention	21.7% more women attended (95% CI 14.7–28.7)
McDowell et al., 1989 [31]	Phone reminder vs. no reminder	Percentage of women screened compared to control group	300 women received intervention	20.0% (60/300) of women were screened compared to 13.7% (35/255) of controls
McDowell et al., 1989 [31]	Physician reminder vs. no reminder	Percentage of women screened compared to control group	255 women received intervention	16.1% (41/255) of women were screened compared to 13.7% (35/255) of controls
HPV self-sampling
Virtanen et al., 2011 [33]	HPV self-sampling vs. reminder letter	Percentage of women screened compared to the control group	413 women were offered the self-sampling intervention	110/413 women (26.6%) were screened following the offer of HPV-self sampling, 246/1144 women (21.5%) were screened following a reminder letter. Relative risk = 1.23 (95% CI 1.01–1.48)

The effect of reminders on cervical screening uptake

Table II summarises the studies that evaluate the effect of reminders on cervical screening uptake in young women. The eligible studies report varying methods of reminding young women to attend for cervical screening – reminder letters, telephone reminders and physician reminders.

All three studies reported an increase in the number of women attending for screening after receiving a reminder letter compared with those not sent a reminder, however, in one study the number of women was relatively small (n = 293) [31] and the significance of the 12.2% increase (13.7% control group vs. 25.9% intervention group) is not reported. Another [32] reported no difference in the cervical screening test rate ratio amongst women aged 20–24 (test rate ratio 0.98; 95% CI 0.72–1.35), although in some places these women would be below the age threshold when cervical screening commences. The same study [32] did however, report a difference in women aged 25–29 (test rate ratio 1.48; 95% CI 1.20–1.82) and 30–34 (test rate ratio 1.39; 95% CI 1.15–1.68) whilst Eaker et al. [30] reported an increase of 7.9% amongst women aged 24–29. Despite heterogeneity in the reported outcome measures, and one study with a relatively small sample size [31], there is some evidence to suggest that reminder letters have a positive effect on screening uptake in young women.

Two studies were included which reported the effects of a telephone reminder from a female research assistant or practice nurse on cervical screening uptake in young women. Both studies reported an increase in the proportion of women attending for cervical screening after receiving a telephone reminder (6.3% and 21.7%); however, the significance of the increase in one of the studies [31] is not reported. Although both evaluations included a relatively small number of women, the reported positive effect suggests that telephone reminders may be a candidate intervention for further evaluation in screening programmes.

McDowell et al. [31] also reported the effect of computer generated reminders for physicians to prompt women about overdue cervical screening attendance amongst young women. The study reports an increase of 2.4% in the proportion of women attending for cervical screening after a physician reminder, however the significance of this increase is not known, and the sample size is relatively small.

The effect of HPV self-sampling on screening uptake

One study [33] eligible for inclusion evaluated the effect of HPV self-sampling on cervical screening uptake in young women. Although not specifically designed to investigate the impact on young women, women aged 30–34 comprised a sub-sample of the study population. HPV self-sampling appeared to be more effective than a reminder letter, with an increase in screening uptake of 5.1% reported amongst young women offered the HPV self-sampling intervention. Women who received the Delphi-screener self- sampling kit (lavage device) also received a brochure on cervical screening and HPV infections which were made available to women attending for routine cytology screening. Although the number of young women included in the study was relatively small, the positive effect suggests that this intervention may be worth further evaluation as an alternative option for non-responders in programmes already using invitation and reminder letters.

Discussion

This systematic review of intervention studies for improving cervical screening uptake in young women has found insufficient evidence to conclusively determine which interventions to increase cervical screening are effective amongst women aged ≤ 35. Only four studies met inclusion criteria with notably few high quality intervention studies focusing on young women, despite the importance of focusing on this group.

None of the included studies were designed specifically to measure a difference in cervical screening uptake amongst younger women, however, all did include some age-specific data which could be extracted for use in this synthesis. The conclusions drawn on the reported effects of the interventions cannot therefore be considered reliable for young women as the sample size of younger women for each intervention was often small. Despite these limitations in the available data, this systematic review can be used as a starting point for further research into this problematic area of public health practice. The review has also highlighted that, whilst there are a range of published studies reporting the effects of interventions designed to increase cervical screening uptake, very few studies have focused on young women. One possible reason interventions may not work in younger women might be that young women do not feel that cancer concerns them [15].

One intervention which exhibited a consistent increase in screening uptake was the use of reminder letters. In many cervical screening programmes, uptake amongst young women remains relatively low compared to other age groups despite the widespread use of standardised invitation and reminder letters. It is therefore apparent that the use of invitation and reminder letters is not the solution for all women and further interventions to improve screening uptake need to be developed in programmes already employing a systematic call and recall system. There is some evidence to suggest that physician reminders may be effective, and also may be readily implemented in many general practices.

The falling coverage amongst young women in organised screening programmes, where women are invited, is likely to be due to these women not engaging with the programme rather than being unaware that they are eligible to participate. For such programmes, the use of HPV self-sampling may be a novel intervention for targeting non-responders [33]. Despite the small number of women included in the HPV self-sampling study, this may be a successful strategy to overcome some of the barriers reported by young women [16–19] and could be a candidate intervention for randomised controlled trials in organised cervical screening programmes.

It was noted that this review did not find any evaluations of nurse or midwifery led counselling and health promotion education to help increase the attendance for cervical screening amongst young women, although a 2008 literature review by Gannon and Dowling [34] emphasised nurses as an important professional group to encourage women's attendance in cervical screening. Furthermore, there were no evaluations found of the effect of modifying cytology screening procedures on uptake in women aged ≤ 35, a category which could include interventions such as offering timed-appointments or offering a choice of male or female sample taker.

In order to fully understand ways to increase young women's participation in cervical screening, it is important to complement researchers’ perspectives with those of young women. However, after a perusal for qualitative studies, we have found that even among these studies there is a lack of age differentiation, although several studies do differentiate between cultural groups, i.e. Chinese women in Hong Kong, [35] Sikh and other Asian groups of women in Canada [36,37] as well as Afro-American women in the US [14]. Studies on minority populations in different settings often generate particularly specific knowledge, with unclear relevance for implementation in other settings. It was also apparent from this perusal of the qualitative literature that, as with quantitative studies, few papers in this area focus solely on young women. Moreover, only one study was found which explored facilitators for young women – a focus group study was performed with 30-year-old women in Sweden [38]. In the findings some women stated that information about the relationship between HPV and cervical cancer would be important in motivating them to participate in cervical screening [38]. However, more research is required to establish whether a better understanding of the link between HPV and cervical cancer among women will be an important factor in increasing participation.

Conclusion

The evidence from this review suggests that the use of letter or physician reminders and HPV self-sampling may have an impact on cervical screening uptake in young women. Organised cervical screening programmes with falling attendance rates despite the use of reminders may need to adopt a more pragmatic approach if they are to improve their falling attendance rates amongst young women. It is apparent that there is a need for novel interventions designed to address some of the reasons for non-attendance given by young women. Future evaluations in this area ought to employ pragmatic randomised controlled trials, with interventions embedded with routine screening programme practice wherever possible. Further research is also required to determine which facilitators to cervical screening are likely to be most relevant to young women. It is likely that there will be no single solution, and that a range of interventions is required within each programme in order to attempt to overcome as many barriers to uptake as possible.

Declaration of interest: The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.

Funding facilitating this review (CW salary) was provided through grants from the Swedish Council for Working Life and Social Research, and the Swedish Research Council. HK is Chair of the Advisory Committee in Cervical Screening (ACCS). The views expressed in this report are those of the author(s) and in no way represent views of the ACCS or the Department of Health.