Consequences of surveying folate awareness

Abstract

Objectives To assess women's concerns when interviewed about the association between folate and neural tube defects (NTDs) and to determine how this is affected by time, being folate aware, having seen folate promotional material or being pregnant.

Design As part of a community randomized trial outcomes evaluation, independent cross‐sectional follow‐up surveys were carried out in 1997 and 2000.

Setting and participants Six local government areas in the state of Victoria, Australia; 2431 women aged between 15 and 44 years.

Main variables studied Whether or not women knew of the association between folate and NTDs (i.e. were folate aware), whether or not women had been concerned by seeing folate/NTD information and if an interview about folate and NTDs had raised any concerns for them.

Results In the 1997 survey, 36% of women said that the interview had raised concerns and this decreased to 26% in 2000. Women who were folate aware were significantly less likely to have raised concerns than women who were not folate aware (OR_adj= 0.38, 95% CI 0.24–0.60). In general, women who had seen promotional material were less likely to feel concern about the interview than those who had not, although this varied with whether or not the promotional material had raised concerns. These effects were greater in women who were pregnant.

Conclusions Women had increased concerns having seen folate promotional material and after being interviewed about it. These results are consistent with the proposition that an initial emotional response to sensitive health information is part of an adaptive response appropriate to the process of health‐related behaviour change.

Introduction

The potential to cause an adverse emotional response when providing sensitive health promotion material and conducting interviews about such material is an issue which must be considered in both the intervention itself and in any subsequent evaluation. As part of a randomized community intervention trial to assess the impact of an intervention aimed at raising awareness of the association between folate and neural tube defects (NTDs), we were able to investigate women's responses to the interview process. The intervention provided information to women of childbearing age in Victoria, Australia, on the role of periconceptional folate for risk reduction of NTDs. A baseline interview was undertaken in December 1996, a post‐intervention interview 1 year later ¹ and a further follow‐up in April 2000.

The rationale for the study was derived from the substantial evidence of many studies, now incorporated into a Cochrane review, that the risk of NTDs can be reduced by about 70% with daily folic acid supplementation. ² The National Health and Medical Research Council of Australia has recommended that women planning a pregnancy or likely to become pregnant (without a family history of NTDs) take a 0.5‐mg folic acid tablet daily and be advised about a folate‐rich diet. ³ In Victoria, the overall prevalence of NTDs remained steady until 1998 and has declined since then, ⁴ although the birth prevalence had declined due to the termination of affected pregnancies detected by prenatal screening and diagnosis. In 1994–95 more than 50% of cases of both anencephaly and spina bifida were terminated prior to 20 weeks gestation. ^{5 , 6} As preconception counselling is not widely used, about 40–50% of pregnancies are unplanned ⁷ and there had been no folate promotional campaigns undertaken in Victoria at that time, it seemed appropriate that a campaign be directed to reach all women of childbearing age, rather than solely those contemplating a pregnancy.

This paper is on the specific part of the randomized community trial aiming:

• to establish whether or not women have raised concerns when interviewed about the association between folate and neural tube defects; and

• to determine how these concerns are affected by time, women's response to having seen folate promotional material, by being aware of the folate/NTD association or by being pregnant.

Methods

The detailed design for the community randomized trial has been presented elsewhere. ¹ However, in brief, four rural and two metropolitan (outer suburban) local government areas (LGAs), geographically isolated from each other, were selected and matched in pairs for similar demographics, i.e. age, household income, education, occupation and non‐English speaking background. The six LGAs represented approximately 10% of the population of Victoria.

A baseline knowledge assessment using a telephone questionnaire interview was conducted in November–December 1996. For surveys, based on appropriate postcodes for the LGAs, lists of telephone numbers were generated for each location using a current residential database of Marketing‐Pro Software (Windows version, Desktop Marketing Systems Pty Ltd, Kew, Victoria, Australia). Interviews were based on a random selection of these telephone numbers and conducted by trained female interviewers between 17.00 and 21.00 (Sunday to Friday inclusive), unless respondents requested another time. Telephone numbers were followed‐up with a minimum of five attempts, with at least four of these calls on different days and the time of follow‐up calls was also varied in an attempt to increase the opportunity for contact. Women aged 15–44 years, with a quota of 2% for ages 40–44 years, were selected and 200 women per location were sampled. The questionnaires were completed on computer scannable forms.

One LGA from each pair was then randomly selected for the public health promotion intervention that comprised a brief programme from late July to early October 1997 using printed material only, and no radio or television coverage. The printed material which aimed to increase the knowledge of folate and the prevention of NTDs in women of childbearing age, and was designed to be eye‐catching and have broad appeal, was in the form of posters, leaflets and information kits, with the posters distributed throughout the community, the leaflets largely in supermarkets and the information kit at pharmacies and maternal and child health centres.

A re‐evaluation of knowledge then took place in November–December 1997 and in March–April 2000, again by telephone interview. The assessments were cross‐sectional in that an independent random sample was selected at all times. This paper only reports data from the 1997 and 2000 surveys.

In response to our request for approval to conduct the study, the Institutional Ethics Committee of the funding body raised the possibility of the questionnaire eliciting an emotional response in participating women and so two questions were added to the questionnaire. The first of these questions probed for a level of emotional response and the second was an open‐ended question to describe this response. Women were asked `Has what we've been talking about raised any concerns for you?'. Responses were coded as yes, no or not sure. Respondents who reported concerns were asked to describe them and the text‐based responses were coded.

Also, women who reported having seen promotional material on folate were also asked `Overall, how would you rate this [folate promotional] material in influencing your concerns about the link between folate and spinal birth defects?'. Responses were coded on a 5‐point scale from large decrease through to large increase. Women were also asked if they were pregnant.

The outcome of interest in this paper is concern associated with the interview. Factors possibly influencing this, such as calendar time, folate awareness, having seen folate promotional material, the influence of that material, whether or not the woman was pregnant, age, whether or not the woman was in an intervention community, number of folate information sources seen, timing of first seeing information, and whether or not women reported they had changed their diet as a result of the information, were assessed. The data were analysed both univariately and multivariately, adjusted for the above factors as well as age group, metropolitan/rural status, education level and occupation.

Adjusted response percentages were calculated and the logistic regression analysis was carried out using STATA statistical package (Release 6.0, 1999: Stata Corporation, College Station, Texas, USA) that enabled appropriate adjustment for the population sampling fraction and the survey cluster sampling effect.

Results

The data for this paper are from independent follow‐up surveys conducted in 1997 and 2000 from 1204 and 1227 respondents, respectively, excluding those with missing age.

Between the two surveys, there were significant changes in the prevalence of women reporting having seen promotional material (from 24%_adj to 52%_adj) and in being aware of the folate/NTD association (from 17%_adj to 30%_adj).

Overall, 738 (31%_adj) stated that the interview had increased their concerns. In the 2000 survey, fewer women reported having concerns with the interview than in 1997 (26%_adj vs. 36%_adj, adjusted odds ratio (OR_adj) = 0.68, 95% confidence interval (CI) 0.45–1.03 P = 0.06) (1, 2Table 1). Women who were aware of the folate/NTD association were significantly less likely to say that their concern had been raised by the interview (16%_adj vs. 35%_adj, OR_adj = 0.38, CI 0.24–0.60, P = 0.007).

Table 1.

 Emotional response to survey interview

Table 2.

 Reasons given for increased concern during interview when discussing folate/NTD association

Women who remembered previously having seen promotional material were less likely to have raised concerns with the survey interview (24%_adj) compared with those who had not seen any material (35%_adj), although this varied according to whether or not women reported increased concern with seeing the promotional material and whether or not they were folate aware. In particular, women who were not folate aware and had increased concerns with folate promotional material (41%_adj) were apparently more likely to be concerned with the interview (OR_adj = 1.34, CI 0.96–1.89, P = 0.07), whereas those who were folate aware and also had increased concerns with folate promotional material (15%_adj) were significantly less likely to have concerns raised at interview (OR_adj = 0.42, CI 0.38–0.46, P < 0.001). Being pregnant was not associated with being more likely to have raised concerns with the survey interview (OR_adj = 0.94, CI 0.78–1.13, P = 0.4). Women who were aged 25 years or more were significantly less likely to have raised concerns with the interview compared with women less than 25 years (OR_adj =  0.44, CI 0.36–0.54, P = 0.001) and those from intervention communities were marginally more likely to have raised concerns at interview (OR_adj = 1.11, CI 1.01–1.21, P = 0.04).

The more sources of folate promotional material that women had seen, the less likely they were to have raised concerns with the survey interview (OR_trend =  0.89, CI 0.79–1.01, P = 0.06), particularly in the 2000 survey where the proportion concerned at interview decreased from 26%_adj in women who had seen only one source to 14%_adj in those who had seen three or more. Women were asked when they had first seen the folate material and in 1997 only 12%_adj of women had seen information more than 3 months before, in 2000 this had increased to 44%_adj. Compared with those who had not seen folate material, women who had seen information more than 3 months before the interview were less likely to be concerned at interview (OR = 0.51, CI 0.49–0.53, P < 0.001) although this significance was not maintained in the multivariate model. Those who had seen it more recently had a similar response to those who had not seen any information (OR_adj = 1.35, CI 0.86–2.12, P = 0.1).

Only a small proportion of the women who had seen folate promotional material said that they had recently changed their diet because of the material (178/946 = 18%_adj) and of these, 135 (72%_adj) were folate aware, 155 (84%_adj) had increased concerns with the folate promotional material and 42 (25%_adj) were concerned with the interview. Although changing diet was associated with concern with the promotional material (OR_adj = 2.51, CI 0.94–6.77, P = 0.06), it was not significantly associated with concern with the interview (OR_adj = 1.50, CI 0.73–3.13, P = 0.2).

There were similar effects in the small group of women who were pregnant (n = 134). Fewer were likely to be concerned about the interview in 2000 than in 1997 (40%_adj vs. 11%_adj, OR_adj = 0.15, CI 0.03–0.67, P = 0.03, sub‐group analysis, 1, 2Table 1 and those who were folate aware or had seen folate promotional material were less likely to have raised concerns with the survey interview (OR_adj = 0.45, CI 0.14–1.42, P = 0.1).

The coded responses describing respondent's concerns with the interview for the two surveys are given in 2, 3. In 1997, 26% described `not being previously aware', 17% described a `need for more information' and a further 14% reported `starting to think about it'. In 2000, `not being previously aware' and `need for more information' remained important categories, but `starting to think about it' declined significantly (to 5%, P < 0.001), whereas `thought should know more' increased from 7% to 15% (P = 0.001) and `had personal relevance' increased from 7% to 23% (P < 0.001).

Table 3.

Table 1 (Continued)

Women who declared they were concerned about the interview were offered the folate information kit; in 1997, 364/420 (86%) accepted this offer, compared with 173/318 (54%) in 2000 (P < 0.001).

Discussion

When undertaking research it is always important to consider the impact of the research on the participants. Interviewing is not a simple process and Oakley ⁸ has reported that three‐quarters of the women interviewed for a study said that the experience of being interviewed had affected their experience of becoming a mother. Interviewing is not without the potential to do harm. Studies of debriefing (which admittedly is not simply interviewing but a process which encourages emotional processing) have shown that the ventilation of feelings may not have positive consequences. ^{9 , 10} The ethics committee from whom we sought approval for our community trial was concerned that our interview process might adversely affect women, particularly pregnant women. Whilst surveying women about their understanding of the folate/NTD association and their recall of seeing folate promotional material we asked them about the emotional impact of both our interview and also of the promotional material. This has given us the challenging and unique opportunity to attempt to map the process by which women respond to sensitive issues and assimilate it. Our two surveys have enabled us to consider this process over time as well.

The findings of this study show that although an important proportion of the women interviewed indicated that, for them, the interview increased concern (31%), this was lessened by time, by being aware of the folate/NTD association, and by having seen the folate promotional material. Women who were more than 25 years of age or who had seen more than one source of promotional material were also less likely to be concerned with the interview.

We were surprised that about two‐thirds of the women who had seen material said it had increased their concerns. However, the response that this group had to interview was complex and depended on whether or not they were aware of the folate/NTD association. In the 1997 survey, women who reported that they had seen or heard any information were asked how they valued the information. More than half the women deemed the information relevant, and more than a third stated it was not personally relevant to them as they were not planning a pregnancy. Two‐thirds of these women thought it was relevant to other women.

In analysing these data, with the primary aim of interpreting women's emotional response to an interview about a sensitive topic, three aspects require consideration. These are the large changes in both (i) exposure to folate information and (ii) awareness of the folate/NTD association that have occurred between the two survey periods and (iii) the interaction between this folate awareness and the emotional reaction to the folate promotional material.

We believe the interview itself is another exposure to folate information and women's responses to it depend on how much prior information they have had and whether or not they have assimilated that information. This is an adaptive process, somewhat like an inoculation process: an initial and continuing reaction to a stimulus (folate information) until there is a protection (folate awareness) against it. Our combined survey data support this in the following way: women who have not seen any information are likely to have an emotional reaction to the interview and unlikely to be folate aware (35% responded this way). For others, seeing the information has no impact emotionally and the message is not remembered (25% responded to the interview in this way). Those who remembered seeing information, remembered responding to it emotionally but who have not yet assimilated it, again respond emotionally to the interview (41%). At some point there is a trigger sufficient to cause assimilation of the information, and there is still some emotional response (15%). Thereafter the information is assimilated, possibly acted upon and there is less likely to be an emotional response (10%). Over time, the proportion of women in each of these groups has changed, supporting the idea of a flow‐through effect as more women see folate material and assimilate it, more women remember responding emotionally to the folate information, but fewer respond emotionally to the interview. The impact of a campaign such as our intervention would be likely to send a ripple of emotional reaction through a community.

The hypothesis of an adaptive response is supported by the information we have about how many sources of folate information women remembered seeing and when they first remembered seeing it. The proportion reacting to the interview decreased with increasing sources but increased if the interview took place within 3 months of seeing the information.

Emotional `processes' are recognized as one category of the `processes of change' within the transtheoretical model of health‐related behaviour change, ¹¹ an example being responding emotionally when confronted with health warnings. ¹² Within the health belief model of health‐related behaviour change, ¹³ part of the adaptive process is realizing one's `perceived susceptibility to disease' and when this occurs, the emotional response evoked is likely to be high. In our study this would occur at the time of assimilating the folate knowledge, when in fact we saw the highest emotional response to the interview.

With this interpretation, the issue of whether or not the interview can be regarded as `doing harm', can be addressed. If one starts with the premise that this information is important to have in the community and to be assimilated, then by whatever method it is promoted it will produce an emotional reaction which is not a harmful thing per se and is likely to be transitory. The tenor of women's responses towards the interview was generally positive, in that they saw the importance of assessing women's knowledge relating to folate:

`More people should be aware, a lot of pregnant females don't know.'

Women's responses to the question on how their concerns were best described had a strong emphasis on the need for more information and cues for action:

`Will go and look up some information.'

`I will need to look into this before I have children.'

Between the two surveys the pattern of responses changed markedly, with increasing emphasis on personal relevance. This indicates that as folate information became more widely distributed through the community, more women acknowledged the importance of this information:

`Made me think about the issue.'

These changes are consistent with changing types of emotional responses as the information becomes assimilated with time.

During the course of the interview, information was indirectly provided to women who did not know of the folate/NTD association. Thus the interview had elements of a health promotion intervention, and women's emotional responses to the interview need to be considered in that light. The interviewers were trained to be empathetic but since the interview was structured they did not seek to interpret or evaluate women's descriptions of their concerns, they merely recorded them. However, women with raised concerns were given the opportunity after the interview to be sent folate promotional material and to be referred to a medical adviser, the Project Officer for the Consumer Information Group on Pregnancy and Childbirth or the Victorian Clinical Genetic Service. Although fewer than 10 women requested personal referral, generally for extra information, a majority of concerned respondents accepted the offer of folate information. This response indicates that women appropriately managed their concerns and were stimulated to seek information rather than being alarmed or depressed. Also, the fact that fewer requests were made after the second survey is consistent with a diminished level of concern over time as folate awareness increased.

Whether or not women actually change their behaviour as a result of the information and awareness of its implications is difficult to assess in these data because we do not know which women subsequently became or intended to become pregnant. We do know that those who had increased concern with folate material stated they were more likely to change their dietary behaviour (OR = 2.51) and were not significantly more likely to be concerned at interview.

Pregnant women were not more likely to be concerned by the interview (OR = 0.94) than non‐pregnant women. Not surprisingly, pregnant women who were folate aware were very significantly less likely to be concerned at the interview than pregnant women who were not folate aware (OR = 0.15). Certainly, women already pregnant or contemplating pregnancy need to be given opportunities to discuss issues associated with emotional responses that may be invoked when they see promotional material.

Despite the potential for folate promotional material to raise concern, we believe that the need to provide folate promotional material to women of child‐bearing age is paramount and that concerns raised are an important, transitory and an unavoidable part of an adaptive response process leading to improved knowledge and potentially facilitating health‐related behaviour change.

Acknowledgements

We wish to acknowledge the support of the Victorian Department of Human Services for funding this research project, the contribution of the women who consented to be interviewed for the study and Tracey Carpenter of the Centre for the Study of Mothers' and Children's Health for her help in preparing the manuscript.