BSH Cancer Screening, Help-Seeking and Prevention Journal Club

About this Journal Club

Comments 1 to 8 of 8

  • BSH Cancer Screening, Help-Seeking and Prevention Journal Club2017 Nov 08 06:57 a.m. 2 of 2 people found this helpful

    The BSH Cancer Screening, Help-Seeking and Prevention Journal Club read with great interest this paper, which we feel provides a useful addition to the literature. In this paper, the authors present a framework for the complex relations between cognition and affect in predicting health behaviour. The authors describe the model that is currently used most often in health behaviour research, which looks at the independent effects of cognition and effect on health behaviour (the “main effects approach”). They then convincingly argue that this model does not consider possible interconnections of thought and feeling and describe three alternative models that may be better at explaining subsequent health behaviour: a mediation model (either an affect-preceding-cognition model or a cognition-preceding-affect model), a moderation model, and a contextualised effects model, which are helpfully explained graphically in the Figure on page 3 of the paper. The authors also provide empirical evidence from the health behaviour literature to support each of these models, and argue that these complex relations should be routinely examined in health behaviour research.

    Our group recognised the importance and relevance of this topic to much of our work on the determinants and possible points for intervention in cancer screening, help-seeking, and cancer prevention. The authors provide a good reminder of how theory can be improved to better understand behaviour, which is relevant to our work on, for example, cancer screening as a teachable moment or cancer worry as determinant of screening uptake, but is also applicable to a wide range of other health behaviours such as smoking, exercise, and healthy eating.

    However, the paper also raised some questions in our group. First, it was unclear to us which model would be applicable under what circumstances. For example, is this a function of the behaviour or of the affective state that is under study, or could this also be construed as characteristic of an individual or group of individuals? For example, for some people, perhaps those who are more organised or conscientious, the cognition-preceding-affect model may better predict subsequent behaviour, while for others, perhaps those who are struggling to cope due to life difficulties or mental health issues, the affect-preceding-cognition model may better predict behaviour. The authors acknowledge that the models they present are “not mutually exclusive” (p.4), and so “multiple types of relations could be involved in determining engagement in a particular health behaviour” (p.4), but this does not provide much guidance on how these models might guide our formulation of hypotheses to be tested in a particular study. Relatedly, it is unclear how these models can (or should) be applied to existing health behaviour models, and to what extent they require an overhaul of these existing models. Our group noted that the inclusion of these complex relationships could water down existing theoretical models, especially if the specific relationship cannot be identified a priori on theoretical grounds but is a function of the behaviour, affective state, individual, or group under study. From the empirical examples that the authors provide throughout, it is unclear whether the complex relations in those studies were pre-specified based on theoretical grounds, or merely exploratory in nature. In their Discussion on p.11, the authors seem to acknowledge that inclusion of the often-neglected interconnections between cognition and affect will require an exploratory, theory-building approach. Our group would have found it helpful if the authors had provided more practical advice on how to formulate a priori hypotheses about these complex relations, and perhaps some worked examples.

    Other questions that were raised by our group are of a more pragmatic nature, such as what the implications of the inclusion of complex relations between cognition and affect would have on study sample size and power (especially if not pre-defined a priori but tested post hoc). A related concern was how to practically take the ideas presented by the authors forward, given that mediation and moderation analyses may require a slightly different skill set, and one that many social scientists may not be very familiar with.

    Overall, however, the group felt that these concerns -especially those of a more practical nature- do not override the importance of taking forward the excellent ideas presented in this paper, which could herald a new era for health behaviour research, both in terms of theory and practice.

  • BSH Cancer Screening, Help-Seeking and Prevention Journal Club2016 Nov 17 11:16 a.m. 3 of 3 people found this helpful

    The HBRC discussed this paper during the journal club held on November 15th 2016. This paper fits well with research conducted within our group on early diagnosis and symptomatic presentation. We considered this paper to be a useful addition to the literature and the paper raises some interesting findings which could be investigated further.

    The study examined the factors associated with a colorectal cancer (CRC) diagnosis and time to diagnosis (TDI). These factors included symptoms, clinical factors and sociodemographic factors. Due to the important role early diagnosis plays in survival from CRC, it is pertinent to investigate at which point diagnosis may be delayed. Early diagnosis of CRC can be problematic due to many of the symptoms being able to be associated with other health problems or being benign. The authors acknowledge that most cases of CRC present symptomatically.

    The group was interested in the finding that less specific symptoms such as indigestion or abdominal pain were associated with shorter patient intervals and that specific classic symptoms, such as rectal bleeding were associated with shorter health system intervals (HSI). So what patients might perceive to be alarm symptoms differs from perceptions of healthcare professionals. It was also highlighted that there was a discrepancy in the patient interval found in this study with a previous study, with this study showing 35 days as the median patient interval, compared to a primary care audit conducted by Lyratzopoulos and colleagues (2015) which showed a patient interval of 19 days. It was also interesting that family history of cancer was associated with a longer HSI, given that family history is a risk factor for cancer.

    The main advantage of this study is the prospective design, with the recruitment of patients prior to their diagnosis. Patients reported their symptoms and so provided insight into what they experienced, but the group did acknowledge that this was retrospective as symptoms were those experienced before they presented at the GP, with these being up to 2 years before diagnosis. The group felt the authors’ use of multiple regression models was a benefit to the study, allowing an investigation into time-constant and duration-varying effects, as in line with previous research, it was shown that rectal bleeding becomes normalised over time.

    We discussed limitations of the study and recognised that the authors did not acknowledge the Be Clear on Cancer Awareness Campaigns which took place during the data collection (Jan-March 2011, Jan-March 2012, Aug-Sept 2012) and could have had an impact by shortening patient interval and increasing referral rates. We also discussed that there could be an inherent bias in GPs and that HSI could be due to this bias of GP’s wanting to reassure patients that their symptom is likely to be the sign of something other than cancer. This could also help explain the longer time to diagnosis and HSI in those with depression and anxiety, as GP’s may feel the need to over reassure these patients, recognising that they are already anxious. However, when symptoms have been shown to be a ‘false alarm’, overreassurance and undersupport from healthcare professionals has been shown to lead patients to interpret subsequent symptoms as benign and express concern about appearing hypochondriacal (Renzi, Whitaker and Wardle, 2015). It may also be due to healthcare professionals attributing symptoms to some of the side effects related to medication for depression and anxiety such as diarrhoea, vomiting, and constipation. The authors suggest also that healthcare professionals might not take these patients physical symptoms seriously. There was also a small number of CRC patients given the amount of patients approached, with the authors recognising the study is underpowered. There may also have been an overestimate of the number of bowel symptoms in non-cancer patients, which was recognised by the authors. It was also unclear that the authors had also conducted univariate analyses and that these were included in the supplementary material until they were mentioned at the end of the results.

    There may also be differences in TDI depending on the type of referral e.g. two week wait, safety netting, and the group would have liked some more information about this. The group would also have liked to see some discussion about the median HSI being longer (58 days) than the 31 days currently recommended for diagnosis from the day of referral and the new target for 2020 of 28 days from referral to diagnosis. It would have also been useful to have some information about how many consultations patients had before being referred, as the authors state in the introduction that 1/3 of CRC patients have three or more consultations with the GP before a referral is made. It would also have been informative for the data on how long participants took to return their questionnaire, with the authors stating that most were completed within 2 weeks, but that some were within 3 months.

    It would be interesting to look further into the factors affecting patients presenting to their GP straight away with symptoms and those which delay. We discussed possible explanations being personality, extreme individual differences in how symptoms are perceived as serious or not and external factors such as being too busy. It would also be interesting to consider whether these symptoms were mentioned by patients as an afterthought at the end of a consultation about something else, or whether this was the symptom that patients primarily presented to the doctor with.

    In conclusion, the HBRC group read the article with great interest and would encourage further studies in this area.

    Conflicts of interest: We report no conflict of interests and note that the comments produced by the group are collective and not the opinion of any one individual.

    References

    1) Lyratzopoulos G, Saunders CL, Abel GA, McPhail S, Neal RD, Wardle J, Rubin GP (2015) The relative length of the patient and the primary care interval in patients with 28 common and rarer cancers. Br J Cancer 112(Suppl 1): S35–S40.

    2) Renzi C, Whitaker KL, Wardle J. (2015) Over-reassurance and undersupport after a 'false alarm': a systematic review of the impact on subsequent cancer symptom attribution and help seeking. BMJ Open. 5(2):e007002. doi: 10.1136/bmjopen-2014-007002.

  • BSH Cancer Screening, Help-Seeking and Prevention Journal Club2016 Jan 19 04:43 a.m. 3 of 3 people found this helpful

    The HBRC discussed this paper during the journal club held on January 12th 2016. We read the paper with great interest as the topic is particularly relevant for our group. We also considered it as a very timely and useful contribution to the discussions taking place in the wider screening community on issues of overdiagnosis/overtreatment and informed decision making.

    The study examined how using different terms for describing ductal carcinoma in situ (DCIS) might affect women’s concern and treatment preferences. This is important because, as stated by the authors, about 20% of screening detected breast cancers are DCIS. Overdiagnosis and overtreatment can be an important problem for DCIS and it has been suggested that a possible approach to reduce this problem may be to change the name and use alternative terms avoiding the word cancer and carcinoma.

    The group agreed that the authors should be acknowledged for their study which used a randomised design within a broader community survey. This allowed to include women with different socio-demographic characteristics and different prior screening experiences. However, as already pointed out by the authors, the group felt that the women’s responses to hypothetical scenarios might not necessarily reflect real life situations. For example, in a real life clinical context, after receiving a DCIS diagnosis, treatment options would be discussed with a clinician and the information and advice provided by the doctor would be important for helping the patient to make the best treatment decision. This was obviously not possible within the study context. Moreover, as also pointed out by the authors, the group felt that patients’ answers might have been strongly influenced by the statement ‘if research shows that watchful waiting is a safe and effective option’. This does not reflect the information that patients would get in a real clinical setting, considering that evidence is still lacking. Nevertheless, we agreed that it is extremely interesting to examine the effects of different terms on understanding, psychological outcomes and treatment preferences. The group would encourage more work in this area.

    We also discussed advantages and disadvantages of the cross-over design. It was highlighted that once a person has been presented with a specific scenario which includes the word cancer, it is unlikely that responses to scenarios/terms presented subsequently might not be influenced by the initial exposure. Therefore it was felt that the between group comparison was more relevant than the before-after comparison within participants.

    The group would have liked to have more information on how the terms used in the two study arms were selected and whether other alternative terms were excluded based on pilot testing. Future studies could explore different terminology. Furthermore, the relatively small sample size limited the possibility of examining whether results might be different for people with previous screening experience or belonging to specific age-groups. Larger studies and also research exploring the effect of different terminology in the context of other diseases (e.g. pulmonary lesions following imaging, diabetes/pre-diabetes) on understanding, patient concerns and treatment decisions would be welcome.

    In conclusion, the HBRC group read the article with great interest and would encourage further studies in this area.

    Conflicts of interest: We report no conflict of interests and note that the comments produced by the group are collective and not the opinion of any one individual.

  • BSH Cancer Screening, Help-Seeking and Prevention Journal Club2016 Jan 11 10:14 a.m. 3 of 3 people found this helpful

    Relevant to our current research goals, the HBRC journal club was interested to review this paper with particular interest in adding to our understanding of the role of descriptive norms in cancer risk-reducing behaviours. The paper demonstrates a strong rationale for investigating the role of descriptive norms in established risk-reducing behaviours, and in uncertain and relatively unknown behaviours. Furthermore, it offers an additional explanation to the reasons why descriptive norm literature is usually inconsistent, or every so often insignificant for health-behaviour change.

    The study employed a cross-sectional design that included an opportunistic sample aged 18 to 95, and reported the use of a theoretical framework, known as the Integrative Model of Behavioural Prediction, which strengthened the methodology of the paper. Although the authors acknowledged their decision for defining uncertainty as a limitation, they attributed the uncertainty of risk-reducing behaviours using scientific knowledge rather than public perceptions of the uncertainty of the behaviours. However, we felt that the study might have benefited from the inclusion of other elements such as beliefs about causes of cancer, scientific literacy, or specifically knowledge, and controlling for age to evaluate further associations with descriptive norms and scientific uncertainty.

    The journal club also discussed the possibility of controlling for individuals’ current health behaviours as another factor that could influence descriptive norms. We also deliberated on whether the established risk-reducing behaviours are also linked to other chronic illnesses, therefore; there may be differences in motivations to take up those behaviours.

    Overall, the HBRC journal club enjoyed reading this article and feel that the paper is an important example of using integrated theoretical models which could improve the way we implement theoretically grounded research.

    Conflicts of interest. We report no conflict of interests and note that the comments produced by the group are collective and not the opinion of any one individual.

  • BSH Cancer Screening, Help-Seeking and Prevention Journal Club2015 Sep 17 04:39 a.m.

    The HBRC Journal Club reviewed this paper, which investigates the effects of a decision aid administered in primary care on men’s intention to undergo prostate cancer screening. The authors conducted a randomised trial in a large number of general practices in France, in a mix of rural and urban areas. The topic intersects with several challenging and controversial areas of public health policy.

    The authors report that the intervention (consisting of a double-sided sheet of A4 describing various aspects of PSA testing) reduced the proportion of recruited men who intended to be screened for prostate cancer. However, although the study is stated to be a pragmatic trial, it is unclear how the intervention and associated findings apply to standard practice. Specifically, men appear to have been provided with the decision aid despite having given no indication that they were interested in PSA screening or concerned about prostate cancer, whereas these are the scenarios in which a decision aid is more likely to be used (notwithstanding possible differences between general practice in the UK and France). It is likely that men who spontaneously raise the issue of PSA screening with their general practitioner have preconceptions about the harms and benefits that are different to men who are not yet aware of it, which would lead them to respond to the decision aid quite differently. Moreover, the control condition, in which participants received only a questionnaire about prostate screening, may have had the paradoxical effect of increasing intentions among men who had not previously been considering prostate screening by raising the option without giving any indication of the associated harms.

    We also discussed characteristics of the decision aid itself. Inevitably, designing information for lay people involves a balance between competing priorities: it needs to be concise enough to be manageable within a given context (general practice, in this case) but comprehensive enough to convey all the desired detail. It also needs to be comprehensible to the target audience, most of whom will not be familiar with medical terminology or statistics. These competing trade-offs are made clear in this study, in which it appears that brevity has been prioritised over completeness. For example, we noted that there is no explicit information about the (limited) benefits of PSA screening on the first page of the decision aid, nor is there an estimate of the risk of false positives and overdiagnoses within the icon array. One might take the principled view that it is important to provide this information in order for men to be able to make an informed choice. Alternatively, one might take a pragmatic view that such information would burden participants with unnecessary detail. However, what constitutes the optimal method is subject to debate and is often a matter of personal opinion.

    A strength of the design was the follow-up question on the reasons underpinning participants’ screening intentions since this highlighted some of the specific viewpoints that were altered by the intervention. However, the study was limited by omitting a measure of knowledge, which is a necessary component of informed decision-making. For example, the study could have assessed whether the intervention was successful in terms of making participants aware of the possibility of overdiagnosis and lack of evidence to support a reduction in all-cause mortality.

    The HBRC Journal Club found this to be an interesting study, which benefited from its large sample size, randomised design and ‘intention reasons’ as a secondary outcome measure. However, it was limited by unclear applicability to standard practice and lack of additional secondary outcomes. In particular, this study illustrates the challenges of generating universally accepted information about screening.

    Conflicts of interest. We report no conflict of interests and note that the comments produced by the group are collective and not the opinion of any one individual.

  • BSH Cancer Screening, Help-Seeking and Prevention Journal Club2015 Jul 22 06:54 a.m.edited 3 of 3 people found this helpful

    The HBRC journal club read with interest this paper that uses an innovative qualitative method to examine the barriers to colorectal cancer (CRC) screening experienced by people from medically under-served areas in the US. The paper offers an account of how macro level factors, such as the socio-political context, combine with individual level factors to determine uptake of CRC screening.

    The authors used a participatory research method and PhotoVoice technique to engage participants in the research project, which took place over several months and included training sessions, group meetings, and individual meetings. We felt that one of the strengths of this approach was that it allowed rapport to be built over time, and that participants could express themselves in ways other than by just verbalising their thoughts and feelings. The photos taken by participants served as the “jumping off point” for the group discussions, which we felt helped obtain a richer picture of what people have to go through in order to obtain screening. Using each participant’s photos as a catalyst for the discussions may have encouraged a more evenly distributed participation across participants, because each participant was expected to contribute to the meetings. Although many of the barriers mentioned in this paper focused on the costs of colonoscopy screening and may therefore not necessarily generalise to healthcare contexts that use different screening methods or are free at the point of delivery, such as in the UK, some of the opportunity costs mentioned by participants did resonate with the journal club and may be more widely applicable. For example, those undergoing flexible sigmoidoscopy screening in the UK may also be faced with the opportunity cost of having to take (potentially unpaid) time off work.

    We felt the authors could have commented more on the possible limitations of their study. The chosen study method required a great deal of commitment and active participation from participants. Although the researchers note that the Photovoice method may address shortcomings of other qualitative work in this field, which “may simply generate ‘impression management discourses’”, we felt that the method does not necessarily circumvent these discourses, but may just elicit more “crafted” image discourses. In addition, the high level of commitment expected from participants in the current study may mean that only those who were most highly motivated and health conscious self-selected to participate, which may limit the generalisability of the study findings. The authors chose to focus on the obstacles experienced by those who had been screened, but those may not necessarily be the same as the obstacles faced by people who have not screened. The HBRC journal club would encourage further study in samples of participants who are not engaged with screening. We also felt that the results were presented in a way that suggested a broad consensus among participants and wondered whether there were any dissenting voices? Some of us felt that a greater emphasis could have been placed on the experience of facilitators to screening uptake, although some felt that this was adequately addressed by discussing the importance of social support around the time of screening. Finally, we felt that the paper formed a good illustration of some of the barriers and facilitators to CRC screening that we already know about, but some of us felt that it did not offer many new insights, and wondered whether the wealth of data collected in the study could have generated any novel insights? We recognise that journal requirements may only allow for a limited discussion of all findings in a study, and we would argue that online supplements could make a valuable contribution to convey the results of this type of study, for example, as a photo gallery with accompanying quotations.

    In sum, the HBRC journal club really enjoyed reading this paper and feel that the Photovoice method is a great method for doing explorative research in certain groups, because of the high level of active involvement of participants in the research project, and the meaning and value added by their pictures which may complement verbal explanations of behaviour. This study shows that decisions about screening are not just up to the individual, but are made in a socio-political context that can help or hinder individuals to obtain screening, and has important implications for policy making in preventive healthcare.

    Conflicts of interest. We report no conflict of interests and note that the comments produced by the group are collective and not the opinion of any one individual.

  • BSH Cancer Screening, Help-Seeking and Prevention Journal Club2015 Jun 05 05:49 a.m. 4 of 4 people found this helpful

    This article was selected for the HBRC journal club held on June 2nd, 2015. The topic of ‘don’t know responses’ in perceived risk items is of particular relevance to our group and raised several useful discussion points. We thank the authors for taking this work forward, and encourage further study in the area.

    The general consensus of the journal club was that the use of two samples was a strength of the research. However, we noted differences in the availability of perceived risk items, the use of different predictors between the datasets, and perhaps most importantly, the availability of explicit ‘don’t know’ response options in one survey but not the other. This made comparisons between the samples and interpretation of the findings difficult. The authors should be credited with making use of available data, but it seems clear that their interesting findings warrant further investigation using study designs that explicitly set out to investigate this phenomenon. Following this, the journal club discussed the research team’s recent research endeavours in this area, presented at the Society of Behavioral Medicine Conference.1

    Conversation turned to the meaning of a ‘don’t know’ response. We felt this was important because observational studies do not allow us to disentangle true ‘don’t know’ responses from those who lack the motivation to complete all survey items. Hay and colleagues addressed this limitation by excluding participants with multiple ‘don’t know’ responses elsewhere in the questionnaire; a decision we agreed with. Suggestions for future research investigating the meaning of a ‘don’t know’ response included using semi-structured interviews purposively sampling participants who selected a ‘don’t know’ response, or verbal protocol (think-aloud) studies encouraging subjects to vocalize their thoughts when completing perceived risk items. While these methods have limitations of their own, the consensus was that they would be worthwhile approaches. We also recommend that future research should include a range of perceived risk items, including ‘feelings of vulnerability’ also known as ‘risk as feelings’; an item not available to Hay and colleagues.

    In sum, the HBRC journal club enjoyed reading this article, and encourages others to take note of their findings. These data have clear implications for scientists who use survey methodology, as well as those making policy level decisions based on their findings.

    References

    1. Kiviniemi M. T., Ellis E. M., Orom H, Waters E. A., Hay J. (2015) Providing a “don’t know” response option changes population perceived risk estimates. Annals of Behavioral Medicine. 49 (Suppl1): S1-S258, C120
  • BSH Cancer Screening, Help-Seeking and Prevention Journal Club2015 Apr 24 10:46 a.m. 3 of 3 people found this helpful

    The HBRC journal club read Scherer et al’s paper with interest. While flu vaccination is not the focus of our work, using metaphors as a manipulation to increase the likelihood of a behaviour and the methods used to test the efficacy of doing so, resonated with our research team. Most of the group were unfamiliar with the metaphor literature and found the introduction to provide a useful summary of the field. The authors present a discussion of the role of risk perceptions and affect, but a more detailed discussion of how the two interact and their complexities might have provided a truer representation of the field.

    The group liked that the authors attempt to measure whether participants were likely to move beyond intention (measuring participants’ desire to receive a reminder email to get a flu vaccination), without having to measure behaviour (which is difficult to do objectively). However, contrary to the authors, the finding that individuals who occasionally get a flu vaccine were more likely to request an email reminder was unsurprising to us because individuals who always get a flu vaccination seemingly do not need reminding. A further strength of the paper is that non-emotive metaphors were considered (the flu as a weed), as this helped dispel the suggestion that the effect was due to vividity or violence (as might be the case with metaphors such as ‘beast’ and ‘riot’). The group wondered if a virus could also be considered to be a metaphor, given its use in computing. Additionally, it may have been of interest if the vaccination itself was also part of the metaphor, for example the flu virus being described as a ‘weed’ and the flu vaccine as ‘weed killer’. In Hauser DJ, 2015 using congruent metaphors to describe an illness and the measure to prevent the illness increased behavioural intentions compared to just using a metaphor to describe the illness.

    While metaphors may increase the vividness of the flu and encourage individuals to engage, the group were concerned about how use of metaphors to manipulate behaviour may not be congruent with informed decision making, instead being considered coercive. We disagreed with the authors suggestion that metaphors might have a use in decision aids, which we believe have a role in helping individuals make informed decisions and not swaying their opinion. We suggested that metaphors might be useful when the aim is to increase individuals’ understanding, rather than increasing intentions to engage in a behaviour. It may also be important to consider the unintended consequences of using metaphors, for example in the cancer field (the focus of our work), describing cancer as a battle may lead to suggestions that people who do not survive the disease did not fight hard enough. However, we acknowledge that metaphors are used ubiquitously in the media, which is difficult to control.

    The group felt that flu vaccination was a complex example to choose to conduct these studies. Flu is a fairly common illness, which might result in participants having an accurate estimate of their risk of contracting the illness or how serious it is. This might explain why the manipulation did not affect the mediators in the main analysis. Individuals are likely to have existing beliefs about vaccination (for example, beliefs about side effects, effectiveness) and the benefits of vaccination might not always be obvious (the individual does not contract the flu - a non event and herd immunity benefiting the population). It would have been helpful to have been informed about the flu vaccination recommendations in the USA where the study was conducted. Cultural differences in how the flu is appraised, treated and prevented between countries might alter the effect of metaphors. For example, in the group’s opinion, the metaphor ‘beast’ was an exaggerated conceptualisation of the flu and felt removed from the actual consequences of the virus.

    Study 3 was perceived by the group to be the most useful study of the paper as it used a validated measure of affect and a larger sample than Study 1. The ecological validity of studies 2 and 3 could have been increased if Study 1 had been a ‘think aloud’ study, whereby the authors could have gained a justification for the mediators of flu vaccination that were tested. The group thought that a strength of the studies is that the authors did consider the possible mediators of any effect of the metaphor. Other mediators that could have been considered include attitudes and a measurement of arousal (engagement). A ‘think aloud’ study might also help to ensure that metaphors are used appropriately, for example in Hauser DJ, 2015, use of an enemy metaphor reduced intentions for self-limiting cancer prevention behaviours (such as stopping smoking or limiting alcohol intake). The group also wondered how novel the ‘novel’ metaphors used in the studies actually were, something that pilot work could have investigated.

    The group thought it was interesting that the findings were not wholly consistent across the studies reported and considered that this could have been a product of differences in the sample. The authors do not describe whether participants were randomised to each condition, and no baseline measurements were reported, both of these factors leave readers unclear about whether the sample was similar across each of the studies. It would also have been helpful to have a justification for the sample size chosen for each of the studies.

    Scherer et al. present a novel paper, which has provided readers with examples of how to measure the impact of using metaphors to increase intentions to receive a flu vaccination. Future work in this field should consider conducting pilot work to ensure the topic, manipulation and measures used are relevant to the population of interest. We caution readers to consider the unintended consequences of using metaphors to manipulate behaviour, including concern about the ethical implications this might have.

    Conflicts of interest We report no conflict of interests and note that the comments produced by the group are collective and not the opinion of any one individual.

Supplemental Content

Authors of Comments

Find comments contributed by an author

Find publications with comments in PubMed

Include "has_user_comments[filter]" with the query or use the "Reader comments" filter in PubMed's side bar. Try it now

More tips for using PubMed Commons