Women’s beliefs about medication use during their pregnancy: a UK perspective

Clear guidelines exist for maternity care in the UK [1], however, the current advice regarding medicines use during pregnancy on the National Health Service (NHS) website is to “use as few over-the-counter medicines as possible” [2]. The guidance relating to the use of medicines, issued by the National Institute of Health and Care Excellence (NICE) is vague, suggesting simply that, if necessary, antacids should be used for heartburn and diet and lifestyle changes should be attempted for constipation [1]. Moreover, the package insert leaflets often provide general statements that use in pregnancy should only occur if the benefits outweigh the risks [3]. For both health care professionals and pregnant women themselves, lack of clear recommendations make decisions concerning medication use during pregnancy difficult.

There have been numerous studies investigating the prevalence of medicine use during pregnancy [4‐7]. Few studies, however, distinguishing between over-the-counter (OTC) and prescription medicines and report how different conditions in pregnancy are being treated. Currently, the majority of UK research focuses on the use of electronic database to characterise utilisation patterns or to study medication safety in pregnancy in a way that does not explore perceptions of pregnant women regarding taking medication [8‐10].

Analysis of international data indicates that women overestimate the risk of taking OTC and prescribed medicines during pregnancy [11]. The authors suggest this may have an impact on whether a woman wants to treat a particular condition during pregnancy. There is a lack of knowledge about the relationship between the woman’s risk perception and her actual medicines use during pregnancy in the UK.

This study is a sub-study of the Multinational Medication Use in Pregnancy Study which has been described in detail elsewhere [12]. Study respondents were recruited via advertisements on national websites commonly used by pregnant women or new mothers in the UK: www.​bounty.​com (a national charity involved in distributing support information to pregnant women) and Pregnancy Forum www.​pregnancyforum.​co.​uk (a national website for pregnant women) during the period 15th November 2011 to 15th January 2012. The websites were selected on the basis of the number of daily users and subscribers. Women who were interested in participating were directed to the study questionnaire platform by clicking on the study URL in the advertisement. Before the woman could access the on-line questionnaire, she was asked to read the study description. Thereafter, the question “Are you willing to participate in the study?” was posed. Informed consent was given by ticking the answer “yes”. The researchers had no access to these e-mail addresses or personally identifiable data either from the website company or the returned questionnaires. The website company had no knowledge of those women who had completed the questionnaire. Only women who were either pregnant or within 1 year of giving birth were eligible. If they did not meet these criteria they were thanked for their interest and exited from the study.

General information regarding the respondents demographics and activities undertaken during pregnancy e.g. smoking and drinking alcohol were obtained. Health literacy was measured by the Set of Brief Screening Questions (SBSQ) [13] and its use in this study is described in more detail in a previous publication [14]. The SBSQ scale was used as an ordinal variable in the analysis. The SBSQ sum score was dichotomized into low/medium (score 0–9) and high health literacy (score 10–12) due to numbers available for analysis.

Respondents were asked to detail which common conditions they had experienced during pregnancy from a list of common ailments experienced in pregnancy, namely: nausea, heartburn, constipation, common cold, urinary tract infections, pain in the neck or pelvic girdle, headache and sleeping problems. The questions were indication-for-use orientated. The medicines used to treat these common conditions were collected and coded using the World Health Organisation’s ATC indexing system [15]. With the exception of certain common medicines e.g. paracetamol, coding was performed to the fourth level i.e. therapeutic group. Respondents were also asked separately to detail specific uses of OTC medicines and prescription medications during pregnancy and were aided with an OTC example to aid recall.

General beliefs about medicines were obtained using the validated Beliefs about Medicines Questionnaire-General (BMQ-General) [16] with an additional four questions added to reflect the benefits of medicines as described by Horne et al. [17]. Respondents were therefore posed four statements relating to the general overuse of medicines, four relating to the harm associated with taking medicines and four relating to the benefits of medicines. Each item was scored using a 1–5 scale where 1 = strongly disagree and 5 = strongly agree. Responses to each group of four statements were group together to give a score ranging from 4 to 20. Higher scores indicate a higher level of agreement with the constructs represented by each of the sub-scales.

The extent to which respondents would avoid certain medicines was ascertained by asking them to detail if they had deliberately avoided any medicines during their pregnancy and, if so, which ones. The avoided medications were coded using the World Health Organisation’s ATC indexing system [15]. Risk perception was determined by asking respondents to rate their perception of risk of certain common medicines, foods and other substances on the foetus. Respondents were asked to tick a box if they perceived the substances to be not harmful (score = 0) to very harmful (score = 10). A full list can be found in a previous publication [18]. We analysed only the three medication groups that were specifically ask about both regarding use and risk perception; OTC anti-nauseants, paracetamol and antibiotics.

Descriptive statistics were used to calculate the prevalence of medication use during pregnancy and presented as percentages with 95 % CI. Chi-square tests and Fishers exact tests were performed to investigate the relationship between socio-demographic characteristics and medication use. Where data were normally distributed, means and standard deviations were used along with independent samples t tests for comparisons. This was applied when assessing the association between Beliefs about medication (BMQ) and medication use. Where data were not normally distributed or formed from an ordinal scale, medians and interquartiles (IQ) were used along with Mann–Whitney U tests for comparisons. This was applied when assessing the association between risk perception and medication use. The significance level was set at p < 0.05. The Statistical Package for Social Sciences (SPSS) v21 (IBM SPSS Statistics) was used for the analysis.

In order to assess the sample for generalisability, a comparison was made between the respondent demographics and data available for the national population at the time of the study [19‐22].

To our knowledge this is the first study specifically exploring the relationship between beliefs and medications use according to specific acute conditions in pregnancy in a UK population with several of our findings important for clinical practice. Firstly we show that a significant number of women are treating common conditions experienced during pregnancy with medication that is either prescribed or purchased over-the-counter. The most concerning figure in this data are the number of women reporting having a UTI and not receiving treatment. Secondly we show that medication use is associated with different socio-demographic factors, particularly in cases of heartburn and constipation. Thirdly, we illustrate the common medicines that are avoided during pregnancy and womens’ reasons for doing so. We go on to show that medicines in general are perceived to be beneficial and slightly overused and that beliefs and risk perception may play an important role in medication use among women with heartburn and UTIs. Finally, pregnant women used a variety of sources when searching for information about medicines, with 50 % choosing to use internet sources.

The disparity in beliefs about medicines and risk perception in respondents indicates a potential lack of awareness of the appropriateness of medicines, particularly in women experiencing heartburn and UTIs and not treating the condition. It is clear that further work should be undertaken to explore these risk perceptions of medicines to understand why women feel these medicines are not safe. This data shows that women viewed medicines as beneficial in general, with fewer suggesting that medicines are overused and harmful when compared to other studies in pregnant women [23].

In those respondents who did not take medicines to treat the UTI, the risk perception aligned well with that from the analysis of the overall international study [11], whereas women treating the condition were significantly below this international average. It may indicate that women need more information regarding the safety of medicines during pregnancy and what is appropriate for them to use, in order to encourage them to treat conditions effectively. Understanding women’s concerns is also essential to promote adherence to prescribed therapy during pregnancy.

The lack of medicines use in women experiencing nausea and constipation is perhaps explained by the relative ease with which these symptoms can be alleviated using diet and lifestyle measures [1, 2]. As this survey did not collect data on severity of conditions or non-medicinal treatments, we cannot evaluate the appropriateness of therapy. However, in those with a UTI, the number not receiving treatment is concerning and potentially links to the respondent’s perception of risk. If left untreated, UTIs can cause significant complications and as such established guidelines exist for the treatment of this acute condition in pregnant women [24].

It is clear from this survey that women have different views regarding their beliefs about medicines and their risk perception of certain medicines for common conditions. This may be having an impact on their willingness to take medicines for these conditions which could lead to a worsening of symptoms or, in the case of untreated UTIs, harm on the fetus. This is particularly visible with the OTC analgesic paracetamol, the use of which during pregnancy is deemed suitable [25]. Yet a large number of women perceived this as a risky medication and would avoid it during their pregnancy. More work needs to be done to explore these concerns with pregnant women at the outset in order to structure a patient-centred consultation around their beliefs and concerns regarding medicines.

Interestingly, the pharmacy is being used as a source of information for medicines equivalent to the number seeking advice from physicians. Currently, pharmacists are not included in the routine care of pregnant women; however, this data shows that they are being used as a source of advice. In previous commentary, it has been stated that pharmacists do not engage in providing information to pregnant women, instead preferring to refer to an appropriate physician, resulting in further delay and appointments for the woman [26]. Therefore, pharmacists should be made aware of these findings and receive adequate professional training in evidence-based treatments in order that they can structure their consultations around the medicines-related beliefs of the patient. There is also, no current guidance readily available for pharmacists and therefore this should be the focus of work in the future to ensure women receive the appropriate treatment at the right time and if necessary are referred to secondary care services. As discussed in the introduction current guidance is vague, with Summaries of Product Characteristics (SPCs) and national formularies often reliant on manufacturer’s warnings rather than evidence behind their use in practice. In the UK, the Teratology Information Service provides guidance on the use of medicines in pregnancy [27] however it is not clear to what extent this is accessed by healthcare professionals.

This was a large cross-sectional, web-based study conducted in the UK. The most important strength is that we were able to include the women’s perspectives in relation to their medication use in pregnancy. This permits comparison between attitudes and actual medication use during pregnancy. Moreover, the use of an anonymous on-line questionnaire may promote a more truthful answer to sensitive questions [28] and enabled us to reach a large proportion of the birthing population in the UK. Internet penetration rates in the UK are relatively high (85 % with access to internet in their homes and 70 % using the internet every day or almost every day) [29] making this method more applicable compared to countries with lower daily access to the internet. Also, several validated instruments were used including the BMQ and the SBSQ enabling comparison with other prior studies.

However, there are also several limitations that should be acknowledged. Firstly, by using a web-based data collection system we could not calculate a conventional response rate, and the possibility of a self-selection bias cannot be excluded. In order to reduce this risk and reach the widest possible segment of the target population, the invitation to participate in the study was sent to all women subscribed to one website and posted on a pregnancy forum, which were both selected according to the number of daily users/subscribers. Use of pregnancy forums was endorsed given their widespread use among the pregnant population, even of more socially disadvantaged status [30‐32]. Recent epidemiological studies indicate reasonable validity of web-based recruitment methods [28, 33]. Hence, the degree to which our findings can be extrapolated to the target population is based on the representativeness of the respondents to the general birthing populations in each country. In most questionnaire-based studies, the study participants are somewhat more resourceful than the general birthing populations. This might also be the case in our study where 7 % of the respondents reported smoking during pregnancy versus 13 % in the general UK birthing population (Table 1). Our findings may therefore be an underestimation of the true occurrence of medication use during pregnancy. Another limitation of this survey is the self-report nature of the responses. The recall of respondents who were towards the latter end of one year post birth may have been a problem in this survey and despite currently being pregnant those women who had not yet given birth would not have been able to provide a full account of their pregnancy. As previously tested for the entire Multinational Medication Use in Pregnancy Study dataset [12], the extent of OTC (8 % difference) and medication use for acute/short-term illnesses (4 % difference) reported by pregnant women was lower than that reported by new mothers. The survey also only asked about medicines-related treatments and did not explore diet and lifestyle modifications that women undertook to treat these common conditions. This would have been useful as most of these common conditions, in the early stages, can be treated with the appropriate changes to diet and lifestyle. Our results should be interpreted with these limitations in mind.

Beliefs and risk perception play an important role on medication taking patterns in pregnancy. Pregnant women should be encouraged to discuss their concerns about medicines taking with healthcare professionals in order to ensure they receive timely and effective treatment. However, more work needs to be done to determine if healthcare professionals have the tools available to them to allay concerns, assess severity of conditions and provide appropriate treatment advice for pregnant women.