Background
Gynaecological cancers are a significant burden in the United Kingdom (UK): they have a combined incidence second only to breast cancer [
1] and incidence is likely to rise as the population ages and becomes more overweight [
2]. Some symptoms of gynaecological cancers can be present long before diagnosis [
3], suggesting that earlier diagnosis could be made on the basis of these symptoms. However, some of these symptoms are relatively common and have a low positive predictive value [
3‐
5] with the result that they may be mistaken for symptoms of more benign conditions by patients and healthcare professionals alike [
6]. Although in 2015 the National Institute for Health and Care Excellence (NICE) lowered the threshold at which patients should be referred for cancer diagnostic tests [
7], if the patient herself does not recognise and act upon symptoms there is no opportunity for referral for investigations and early diagnosis.
Due to the progressive nature of cancer, much effort has been put into understanding what contributes to a shorter patient interval – the time taken for an individual to notice and interpret a new symptom as worthy of medical attention and to seek medical advice [
8]. Shorter patient intervals could lead to earlier-stage diagnoses and improved survival [
9,
10]. Studies of women’s levels of awareness of gynaecological cancer symptoms show that free recall of symptoms is low and, although prompted recognition is higher, most women still do not recognise all symptoms [
11‐
13]. Furthermore some symptoms (e.g., abdominal pain) are well-recognised by women as important, while others (e.g., feeling full quickly and difficulty eating) are less well-known as possible symptoms of ovarian cancer [
12]. When women experience these symptoms, they typically attribute them to benign causes rather than cancer [
14], and may not seek a medical opinion promptly.
UK studies have reported that participants who do not recognise a symptom as being a cancer alarm symptom are more likely to anticipate waiting longer to seek help than those who do [
15,
16]. In addition, when General Practitioners (GPs) were asked what they thought caused delay in presentation for gynaecological cancer symptoms they agreed that patients’ lack of knowledge was a substantial barrier [
17]. This is consistent with the finding that knowledge that a symptom could be caused by cancer acts as a prompt to help-seeking [
18]. However, there is a dearth of research on interventions to promote prompt help-seeking in this context [
19].
The provision of information is an ongoing challenge for healthcare providers. Completeness of information must be balanced with considerations of the target population’s health literacy and the importance of not inducing undue anxiety, which itself can be a barrier to help-seeking [
20]. To date, the use of focus groups and other qualitative methods to develop and refine written information has dominated this area, while experimental evaluation of the impact of information leaflets is rare. Qualitative evaluations of health information leaflets tend to report descriptive results about whether the leaflet was found to be acceptable and comprehensible by the patients [
21‐
23], and although this can be an important step in development, how effective written information is at changing attitudes or behaviours is still under-researched [
24].
This study therefore represents a crucial, but often-neglected stage of testing in leaflet development. It is an attempt to go beyond the focus groups and interviews [
17] that have already informed the initial development of this leaflet, and obtain a quantitative measure of its immediate impact before testing its effect on consultation rates in primary care. This approach is in line with the Complex Interventions Guidance from the Medical Research Council (MRC) which emphasises the need for a ‘development-evaluation-implementation’ process, and for
‘greater investment in developmental studies prior to large-scale evaluations’ [
25] (p33).
In the current study, we evaluated a leaflet designed for use in a primary care setting which aims to increase appropriate presentation of symptoms associated with gynaecological cancers. It does this by laying out the symptoms clearly, reassuring women about their potential concerns and including a checklist of symptoms that might empower them to take their concerns to a healthcare professional. We quantified the change, after reading the leaflet, in symptom knowledge, barriers to presentation and anxiety, and assessed whether these were related to anticipated prompt help-seeking.
Discussion
This quantitative approach to the evaluation of an information leaflet for gynaecological cancer demonstrated a short-term positive impact of the leaflet on symptom knowledge, perceived barriers and anxiety levels. It also reduced the median time the participants would wait before seeking help across all 12 symptoms. Furthermore, we have shown that this increase in symptom knowledge, and decrease in the number of barriers endorsed, were independently predictive of an increase in the number of symptoms for which these women would seek help promptly. Similar to previous research into information provision on testicular cancer [
33], the inclusion of reassuring messages and a symptom checklist which gives women a clear course of action facilitated the uptake of the information while avoiding increased anxiety.
Providing women with knowledge of gynaecological cancer symptoms while addressing barriers to help-seeking is important to improve rates of early diagnosis. This is particularly true because the anatomical location of these types of cancer could mean that embarrassment and unwillingness to seek help may present more significant barriers than for other cancers. This may disproportionately impact women from Black, Asian and Minority Ethnic backgrounds [
34], which may in part explain the persistent reticence shown by the non-white women in our study to anticipate seeking help. As education levels and health literacy were adjusted for, these cannot explain the large difference found for this group. Further research should investigate what might be added to or changed in such a leaflet to tackle the barriers to prompt help-seeking that are more pronounced in non-white women.
In addition to embarrassment, the non-specific nature of many symptoms of gynaecological cancers, for example bloating and change in bowel habits, has been found to be a barrier to prompt help seeking as women do not associate such symptoms with gynaecological cancers [
14], and are less concerned about symptoms that do not specifically involve the reproductive organs [
13]. This study showed that by simply highlighting these symptoms to women, they may be more willing to seek help when they notice these changes.
Our previous research with GPs showed that while they believed the leaflet was important, they were concerned about a potential increase in anxiety and consultation rates [
17]. We found that the leaflet was able to reduce the anxiety scores of our participants. This effect was however not associated with an increase in intentions to seek help promptly. While a decrease in anxiety score is a positive effect as previous research has linked high levels of fear and anxiety about cancer to delays in help-seeking [
35], the change in the mean anxiety score, although significant, was relatively small and as such may not have been big enough to affect help-seeking intentions. The ways of reducing anxiety and its effect on prompt help-seeking warrant further investigation.
The main strength of this study is that it presents a frequently-lacking quantitative evaluation of a low intensity intervention. Our study also benefitted from an ethnically diverse sample (25 % non-white compared to 14 % in the general population [
36]), with a wide age range, and adequate power to produce robust findings. Additionally, using a repeated measures design allowed us to measure changes over time, which showed large differences prospectively, albeit over a short time interval.
While strengthening the study and allowing us to assess changes over time, the test-retest design is also a potential weakness in that some of the effect seen on knowledge could have been the result of the women being more practised at the questions on the second completion of the questionnaire. A longer time before re-test was unfortunately not practical here, but despite this the results do provide a first, objective evaluation of the impact of the information provided. In addition, women may have been motivated to provide responses at T2 that were consistent with researchers’ expectations of change (social desirability bias), particularly in relation to speed of help-seeking. As we did not measure social desirability, we cannot rule this out as a possibility.
The large changes in knowledge and barriers led to very large changes in anticipated help-seeking, with evidence of a ‘dose response’ relationship: larger changes in knowledge and barriers leading to larger changes in promptness of help-seeking. This suggests that the more difference we can make to these factors, the more impact there will be on help-seeking. Therefore, although this study can only demonstrate the short term impact, the large effect found implies great potential to make a longer term difference.
This study sample over-represented higher educational levels compared to the general population (54 % had a higher degree compared to 30 % of the general population [
37]) and was recruited from a population with an interest in or connection to cancer because recruitment took place at Cancer Research UK events. Those with higher socio-economic status, implied by the higher education levels, are a group more likely to take part in surveys [
38], and so generalising must be done with caution. Despite this, there was a good spread of health literacy scores, which was comparable to that found in other studies [
39,
40]. Furthermore, there is no reason to think that the women came to these events with any more knowledge of gynaecological cancers than other members of the public, as in this sample knowing someone with cancer was not found to be associated with increased knowledge of these symptoms (mean symptom knowledge score at T1 was 6.19 in those with no cancer experience (
n = 42) and 6.62 in those with a friend or family member who had had cancer (
n = 429);
p = 0.42).
Although these data were collected in 2011, there is little reason to believe that knowledge of symptoms or other barriers have changed substantially in the intervening time. Recent studies continue to find that women frequently misattribute the symptoms of gynaecological cancer, and that their help-seeking is influenced by barriers such as the demands of everyday life, or the anxiety that surrounds cancer [
14,
41]. The importance of exploring and addressing these factors in help-seeking behaviour remains undiminished and we believe our findings make a significant contribution.
Finally, this stage of testing involved measuring intentions only, not actual behaviour, however it has been found that the correlation between intentions and behaviour in prospective studies is 0.53, which could be considered a large effect size based on Cohen’s interpretation of correlations [
42,
43]. This part of the development, therefore, constituted a proof of concept, to support the next stage of testing and development in a primary care setting, assessing the impact of the leaflet on consultation rates (particularly in ethnic minority and socioeconomically deprived groups), investigations ordered and referrals made.
Conclusions
When developing written information aimed at changing behaviour, it is essential to carry out thorough testing in the target population to demonstrate the impact of the intervention on behavioural intention, and to understand the mechanism for that impact. This study shows that useful information can be gathered through the under-used step of quantitative evaluation of a leaflet and as such indicates a method through which more effective patient information documents could be created. Improving rates of early diagnosis of cancer in the UK depends, in part, on ensuring the patient interval is kept as brief as possible. This study has shown that, as a minimum, intentions to seek help for gynaecological cancer symptoms can be improved through information leaflets which also address barriers to help-seeking and provide a tool with which the woman can approach her GP.
Efforts should now be focussed on trialling the leaflet in primary care settings and assessing its impact on help-seeking behaviours in the real world.
Ethics approval and consent to participate
We confirm all patient/personal identifiers have been removed or disguised so the patient/person(s) described are not identifiable and cannot be identified through the details of the story. All named authors have agreed to the submission and have contributed significantly to the manuscript, and it is not under consideration elsewhere. Ethical approval was granted by the UCL Research Ethics Committee (reference 1122/004), and we have observed appropriate ethical guidelines and legislation in conducting the study described in this paper.
Competing interests
The authors declare that they have no competing interests.
Authors’ contributions
AS, JWar and JWal conceived of and designed the study. AS, MM and EB carried out the data collection. AS, MM, JWal and CFS contributed to the data analysis. All the authors have been involved in the drafting of the manuscript, given final approval of the submitted version and agree to be accountable for all aspects of the work.