How Effective are Preconception Care Services for Women with Type 1 or Type 2 Diabetes: A Systematic Review

Aim

The aim of this systematic review is to establish how effective preconception care services are for women with type 1 or type2 diabetes. The effectiveness of these services can be influenced by many external factors such as social, economic, psychological and cultural factors.

The objectives

• To review the current evidence base concerning preconception care.
• To identify effective approaches to reduce complications during pregnancy
• To identify the social, cultural and economic factors that affect women in seeking preconception care services.
• To address the effectiveness of current preconception care services and how they can be improved. Methodology This review was conducted using the Cochrane handbook of systematic reviews [12] in collaboration with the Prisma guidelines [13]. The purpose of this review was to evaluate the literature around preconception care and diabetes in accordance with the aims and objectives stated below.

Eligibility Criteria

Included Studies: Randomised control, prospective cohort and population based studies were used in these reviews which were conducted within the last ten years. All studies included were in English language, due to time constraints and difficulty in interpreting them; studies from other languages were excluded. By only including English studies it ensures that there are no misinterpretations of the study’s results.

Included Participants: Participants with pre-existing type one or type two diabetes and planning pregnancy or currently pregnant will be included in this review. Participants with ‘gestational diabetes’ will be excluded. Due to the nature of the review only women were included however there were no other restrictions to participants. Included Interventions: The intervention being evaluated was preconception care, therefore the included studies must have had an adequate amount of preconception care. ‘Preconception care’ includes any type of health intervention in order for the participants to have a healthy and risk free pregnancy. Studies which include women who are not using these services will be excluded. Included Control Groups: Only studies which included control groups were included in the study, this was to ensure a clear comparison can be made on the effectiveness of preconception care. Included Outcome Measures: This study required to include information of participants Hba1c levels, number of congenital anomalies or stillbirths and the number of women taking folic acid both with and without intervention.

Search Strategy for Included Studies: In order to ensure that there is minimal database bias, the following databases will be included in the search; CINAHL, central database of Cochrane collaboration, Embase and OVID. The databases will be accessed through the University of Nottingham login. The method of hand searching and foot chasing will also be used as this may help to review references of published studies related to the outcome measure.

Table 2: Table categorizing the number of articles gained from each identified database and the use of Boolean operators to combine each search term (studies included from January 2006 – January 2016). The literature search was carried out in order of figure 1 where titles and abstracts will be screened for appropriateness and counter checked by a second individual (project supervisor). Those studies which did not appear to be relevant were excluded. The second stage involved all studies which meet the inclusion criteria, full texts were sought to make a final selection. Studies which did not meet the inclusion criteria were to be excluded from the review. Quality Appraisal: A high quality systematic review provides a comprehensive summary of research-based knowledge on a topic, taking into consideration their quality and validity (DiCenso, Guyatt and Ciliska, 2005). This systematic review will be looking at primary studies of quantitative background. Therefore, emphasis will be placed upon internal and external validity. Critical Appraisal Tool for Included Studies: Due to the heterogeneity of studies, the Cochrane tool of assessing risk of bias was used. The tool involves assessing each study with either a ‘high, low or unclear’ risk of bias. The tool allows researchers to interpret which studies are of high quality (with lower risks being better quality). Data Extraction and Synthesis: The findings of the review are presented using three main themes; attendance to preconception care, intake of folic acid and glycemic control up until the end of the first trimester of pregnancy. A table format has been used to present the findings of each study to allow comparisons to be made. Ethical Considerations: Although this study did not require ethical approval, each included study was assessed for appropriate ethical consent.

Database Search Results

There appeared to be very little research surrounding diabetes and preconception care. However, the initial database search revealed (CINAHL – 245, EMBASE – 48, MEDLINE – 277) studies. Further to this, 728 studies were excluded due to failing to meet the inclusion criteria. Rejected studies included participants, which did not have type 1 or 2 diabetes mellitus, studies which were more than 10 years old and studies which were duplicates. A total of 6 full text articles were taken for further assessment. A thorough search of grey literature and reference list provided 14 additional articles, providing 22 full text articles for examination. The flow diagram in Figure 1 illustrates the study selection/rejection process. Of the 22 full text articles, 6 met the inclusion criteria and were included in this review.

Click to enlarge

Number of congenital

Number of neural tube defect rates Number of Anomalies/ malformations.

stillbirths.

effect $$ \begin{array}{c c c c c c c} \mathrm {P=} & 0 & 1 0 & \mathrm {P=0. 0 2} & 1 & 6 \\ < 0. 0 0 0 1 & & & & & \end{array} $$

P=0.7 NR NR

NR Table 3: Table to show a summary of the main findings from the studies (1).

A total of 6 studies were included in this review, a summary of each study can be found in tables 3, 4 and 5. The included studies either focused on maternal and foetal outcomes and participant demographics. Each study focused on at least one of the above key themes. However as previously stated, preconception care in diabetes is an area of research which lacks in depth findings based on the primary literature search.

form of a sufficient (with a value around 7mmol %) or insufficient (significantly over 7mmol %) value) [1]. Tripathi [14] explored 588 women with diabetes and found preconception counseling was significantly associated with better glycemic control. There were 240 women who received preconception care from which 51 of these women had an HbA1c <7% and 109 women had an HbA1c >7%. In comparison to this, 297 women did not receive preconception care from which 29 of these had an HbA1c of <7% and 117 who had an HbA1c >7%. However, data regarding the HbA1c level was missing for 276 women accounting for 47% of the sample of the study. Owens [18] examined 272 women and found that glycemic control improved for women with both type 1 and 2 diabetes mellitus. There were 61% of these women who had type 1 diabetes and 39% who had type 2 diabetes. The overall initial HbA1c value decreased from 7.3 to 6.9% and the number of women achieving HbA1c under <7% during their initial assessment increased significantly. Women with type 1 diabetes during the period of 2005-2007 achieved a mean HbA1c of 7.9% and 7.6% during 2008-2010. Women with type 2 diabetes mellitus during the period 2005- 2007 7.7% and 7.1 during 2008-2010. Similarly, Callec, Galet, Morel et al. [16] carried out an observational study using a total of 97 women with type 2 diabetes mellitus. The mean HbA1c during the first trimester of pregnancy was 7.23% and at the end of the pregnancy was 6.07%. There were up to 25% of women, who did not receive preconception care and had an HbA1c level of up to 8%. It was found that women who received preconception care had lower HbA1c levels in later part of their pregnancy. Temple, Aldridge and Murphy [17] used a prospective observational cohort study of 290 women with type 1 diabetes. Two women declined to be part of the study. A total of 110 women received pre-pregnancy care at a clinic. At an initial pregnancy clinic visit 82.7% of women with pre-pregnancy care had an HbA1c <7.5% and 62.7% had an HbA1c of 7%. In comparison to this 50% of women who did not receive pre- pregnancy care had an HbA1c of <7.5% and 35.4% <7%. It was found that women who attended pre-pregnancy care programs had a mean HbA1c of 5.9% in comparison to 6.6% to women that did not attend pre- pregnancy care. However there was no evidence of significant difference in the incidence of hypoglycemia between women who did and did not receive preconception care. Murphy, Roland, Skinner, et al. [19] used a prospective cohort study in 680 women with type 1 and type 2 diabetes. A total of 181 (27%) women attended preconception care and 499 (73%) women did not. It showed 53% of the intervention group receiving preconception care had maintained HbA1c levels below 7%. For women with type 1 diabetes their glycemic control was better in comparison to type 2 diabetes. A prospective population based study by Holmes, Hamill, Alderdice, et al. [15] included 249 women of which 135 were post intervention and 114 pre-intervention. There were 183 women who had type1 diabetes and 66 with type 2 diabetes. In addition to this 37% of the pre- intervention group had an HbA1c <7% compared to 39% in the post intervention group. The study showed the post intervention group to have a lower HbA1c at the first clinic visit (6.9% vs 7.8%) in comparison to the group with no intervention. Although despite this, the overall difference in the average HbA1c level in the first trimester was not as significant (7.4% vs 7.3%). Theme 2-Intake of Folic Acid: Four studies found an increase in women taking folic acid after preconception care. One study did not investigate folic acid. Owens, et al. studied 272 pregnancies during a 6-year period. Pregnancy outcomes were compared from 2005-2007 to 2008-2010. There were 272 pregnancies which were examined (61% type 1 diabetic and 39% type 2 diabetic). During the first period of the study 41% women received folic acid and during the second 62% received folic acid. There was an overall non-significant increase in the number of women receiving folic acid during pregnancy. Similarly, Temple, et al. [17] showed that women who were not attending pre-pregnancy care were less likely to be taking folic acid (although data was not provided). All women in the study were advised to take folic acid supplementation before or as soon as they discontinued contraception. Women who had not received any preconception care were unlikely to have been receiving folic acid supplementation. Tripathi, et al. [14] studied 588 women and 448 (76%) had type 1 diabetes. There were 297 women of the 588 who did not receive preconception care. There were also 297 women who did not receive folic acid preconception. In comparison to this, preconception care was significantly associated with folic acid being taken within 3 months’ preconception. There were 134 (68.4%) women who were receiving preconception care that took folic acid within 3 months of preconception in comparison to 62 (31.6%) women who were not receiving preconception care. Murphy, et al. [19] found that 157/178 (88.2%) women attending preconception care took folic acid during the preconception period in comparison to 112/420 (26.7%) women who did not attend preconception care. However, Holmes, et al. [15] found that 47/111 (42.3) women pre-intervention were taking 5mg folic acid preconception in comparison to 65/134 (48.5) women who were post intervention.

Theme 3-Attendance to Preconception Care: All studies included elements of preconception care including advice on intake of folic acid, glycemic control and preconception counseling. Each study then measured congenital malformation rates in order to make a sound comparison to evaluate the effectiveness of the preconception care provided. Callec, et al. [16] found that preconception care was effective in 4 out of 97 pregnancies. A total of 97 women with type 2 diabetes were included in the study and compared to the general population. There were 6 congenital malformations which was 3 times higher than the general population. Although despite this, no congenital malformations were found in fetuses of planned pregnancies. Overall preconception care was effective in only 4 out of 97 pregnancies. Temple, et al. [17] studied 290 pregnancies in a prospective observational cohort study between 1991- 1996 and 1997-2002. There were 110 (38%) women who attended pre-pregnancy care. Women attending the care were seen 1-3 months by the consultant physician and also the diabetes specialist nurse. During the two study periods the mean age of women attending preconception care was 29.4 vs 28.6 years old. There was also an association between accessing preconception care and a reduced risk of spontaneous abortion. However, data on socioeconomic status of participants was not collected and this was recognized as a potential factor in women seeking preconception care. Although overall the study showed a positive relationship between preconception care and maternal/fetal outcomes. Although a prospective evaluation by Owens, et al. [18] found that the number of women attending preconception care increased in the period of 6 years, Tripathi, et al. [14] found a positive relationship between preconception care and pregnancy outcomes. However, half of the participants in the study did not receive preconception care. In addition to this, the study also recognized that positive outcomes of pregnancy may not be entirely linked to women receiving preconception care as women who actively seek help are likely to be more cautious of their health. Murphy, et al. [19] found that there was a clear difference in pregnancy outcomes in women attending preconception care services in comparison to those who did not. The rates of malformations, stillbirths and perinatal mortality occurred in no more than 2 pregnancies. This number was more than double in women who did not receive in preconception care. Holmes, et al. [15] decided to provide preconception care through an educational DVD via primary care and secondary care practitioners. The rate of congenital malformation and awareness of how to maintain a healthy pregnancy were higher in the post DVD group in comparison to pre DVD.

Study Characteristics

The characteristics of each study have been discussed in relation to outcomes measured, interventions, participants, methods/design and their quality.

Participants

The nature of this study meant only women and babies were included. The number of participants ranged from 100 to 680. Many of these were recruited from pre-pregnancy/perinatal care clinics, regional maternity survey centres as well as maternity departments in hospitals. The highest number of participants came from the pre-pregnancy clinics and the lowest came from the hospitals. The participants were recruited using an inclusion/exclusion criterion which involved the type of diabetes they were diagnosed with and whether they were receiving preconception care. The criteria did not state a specific age however it just stated that women should be of child bearing age with pre-gestational diabetes.

Interventions

The included studies used interventions which explored medical and social aspects of preconception care concerning women and the fetus. Many of the interventions took place over a prolonged period which varied from one year to eleven years. All of the studies involved women who were receiving or not receiving preconception care and their pregnancy was evaluated using the main themes; HbA1c level in the first trimester, attendance to preconception care and intake of folic acid.

Outcome Measures

The primary outcome of the systematic review was to evaluate the effectiveness of preconception care in women with type 1 and/or type2 diabetes. The key themes gained from the outcomes as a whole were glycaemic control and the number of women who maintained an acceptable HbA1c level (<7%) within 3 months of the first trimester. It also included the number of women who attended preconception care services, the number of women taking folic acid and the social demographics of the participants. These outcomes described were measured through surveys, audits and observations in the included studies. It is important to appreciate that these outcome measures will be used to evaluate how effective preconception care is in reducing maternal and foetal mortality and morbidity.

Risk of Bias

The matrix illustrated in table 6 shows the overview of the risk of bias in the studies. The table helps to reflect the rationale for the risk of bias in further depth.

Key: UR = unclear risk LR= low risk HR=high risk

The Intake of Folic Acid

The intake of 5mg folic acid preconception and during the first 12 weeks of pregnancy has been linked to a decreased risk of congenital malformations. Three studies show an improvement in the number of women who took folic acid if they had received preconception care. Owens, et al. [18] found a low improvement rate in women taking folic acid after preconception care (41% in the preconception care group vs 62% in the no preconception care group). Two studies did not investigate folic acid [16, 17]. Murphy, et al. [19] researched 10 different regional preconception care centres (including women with type 1 and type 2 diabetes) and found the highest number (88.2%) of women taking folic acid. One advantage of this prospective cohort study is that the use of 10 regional centres reduced the risk of selection bias. Holmes et al. [15] demonstrated the second highest rate of women taking folic acid preconception and during pregnancy (81% of the intervention group). One of the strengths of this study is that it took place in a primary care setting and therefore it may have been easily accessible for women. The use of an electronic resource helps to educate a larger number of participants than just clinics alone. In addition, patients can access information in a time and place that is convenient to them. Similarly, Tripathi, et al. [14] found the third highest number of women taking folic acid (68% of the intervention group). This study compared data from many different maternity centres in one region and therefore also targeted a wider range of participants. It is important to note that although a multitude of maternity centres were compared, each centre may have differing protocols on their delivery of pre- conception care; ultimately, this can impact the generalisability of the findings. Owens, et al. found the lowest number of women taking the supplement (41% from the preconception care group vs 62% in the group not receiving any intervention). One reason for the reduced uptake of women taking folic acid in this study compared to other studies may be due to the fact that women were taking the lower dose at 500 micrograms. A dose at this strength is available without a prescription and therefore if an adequate medical history is not taken by a clinician, then this information may not be documented in the medical records. Therefore, this difference in result does not necessarily mean that women were not taking the supplement at all. However, many of the patients in these studies were self- reporting whether they had taken the supplement and therefore an element of reported and recall bias must be considered. Subsequently, this method leaves space for uncertainty and error in interpreting the findings. It is important to mention that some patients may not be taking the non-prescription strength folic acid, but they may feel obliged to state that they are. This may be due to an expectation or some form of direct or indirect influence from the clinician. The patient may not be taking the folic acid for a whole host of reasons; to include the pharmacological side effects, any preconceptions they already have about the drug or they simple choose not to take it. It is also important to note that even if the medical records state that a patient is taking folic acid, it does not necessarily mean that the patient is concordant with the medication regime. Therefore, there is no certainty that patients are taking folic acid regularly, which could lead to inaccurate conclusions being drawn about the effectiveness of it. As stated earlier, the aim of folic acid supplementation is to reduce the risks of neural tube defects. However only one out of the six included studies [17] state that there were 6 neural tube defects in the babies whose mothers received preconception care; it did not state the figure of neural tube defects in women who had not received preconception care. Therefore this made it difficult to analyse the relationship between preconception care and neural tube defects. Neural tube defects are known as one of the most common malformations in neonates which occur in the early weeks of gestation [21]. They can significantly decrease the quality of life for a baby. The incidence varies from 12 per 1000 pregnancies in Ireland and Wales and Indian Sikhs in Egypt [21]. Diabetes in pregnant women is known to increase the incidence of neural tube defects [1]. One method to promote the importance of taking folic acid this could be to educate all women of child bearing age on the importance of taking the supplement when planning and during pregnancy. Local campaigns can also be implemented to raise awareness in women regarding the impacts of neural tube defects on babies. In addition, education about this can begin in science teaching in secondary schools. Service users can also be involved in teaching healthcare professionals on how to work with women at a higher risk of delivering a baby with neural tube defects. A study by McGarry and Thom [20] showed student nurses had a positive experience of service user involvement in teaching. It also emphasised the value of practising person centred care to promote good health.

The rate of attendance to preconception care clinics

The number of women attending preconception varied amongst all six studies. Many of the women attending were of Caucasian, Irish or British ethnicity as shown by Holmes, et al., Murphy, et al. and Tripathi, et al. [14, 15, 19]. There was an exception of three studies which did not state the ethnic background of the women attending preconception care clinics [16, 17, 18]. It has been demonstrated that Caucasian, white British and Irish population attending preconception care represented the highest numbers in the studies compared to any other ethnic group [14, 15, 19]. Therefore, it could be argued as to whether other ethnicities have a lower rate of attendance to preconception care clinics due to socio- economic and cultural barriers. One of the causes behind these barriers is low health literacy. Health literacy refers to a person’s ability to access and understand information regarding their health [23]. Low health literacy can be influenced by factors such as socio-economic position, culture and language and the educational status of a patient [23]. A low level of health literacy may result in patients being unable to manage their health effectively or access healthcare services. The rapid estimate of adult literacy in medicine (REALM) is used to assess a patient’s [24] health literacy status. However this tool fails to consider patient attitude, values and beliefs [21]. These findings raise the question of whether care is being individualised and person centred for women with specific social and cultural backgrounds. Therefore, future research in developing this assessment tool further may prove valuable. Holmes, et al. [15] used a DVD to educate women on maintaining a healthy pregnancy. The study had data missing regarding many pregnancy outcomes such as congenital anomalies, stillbirth and neural tube defect rates. However it proves to be useful in providing insight in promoting health. Therefore one of the alternatives to address health inequalities may be to collaborate multidisciplinary care clinics alongside complimentary resources (which are available in all languages). Shared decision making may also prove useful in preconception care [25]. This will help women make more informed decisions about the care and ultimately allow patient autonomy. Murphy, et al. [19] found that the women not attending preconception care were the most socially deprived. Each participant was categorised in quantiles ranging from 1-4 (4 being the most deprived). One of the strengths of this study is that it included both type 1 and type 2 diabetic women, making it generalisable to all women with diabetes. Moreover, it also implemented the study over 10 different regional centres helping to reduce selection bias. Callec, et al. [16] found that preconception care was only effective in 4 out of 97 pregnancies in women with type 2 diabetes. This indicates the need for preconception care to be individualised. Although Temple, et al. [17] found there was no difference in social demographics between women accessing care earlier and later. They identified there was a strong need to encourage women to attend preconception care earlier. Owens, et al. [18] further supported this by stating that earlier glycaemic control helped to reduce the risk of delivering an infant with macrosomia, risk of miscarriage and foetal malformations. This raises the importance of women understanding the value of attending preconception clinics earlier. The use of technology in healthcare may prove to be useful in reaching out to socially disadvantaged women. Although this is not to say that current assisting resources such as patient information leaflets provided by doctors and specialist nurses [26] should be neglected. One of the limitations of the study included that it did not clearly state whether type 1 or type 2

diabetic women benefited from structured preconception counselling. Further to this Tripathi, et al. [14] stated that preconception care may not be the sole factor in improving pregnancy outcomes. It was demonstrated that patient motivation could also be a factor in particularly for type 2 diabetic women. Therefore, this helps to reinforce the idea of whether healthcare providers (particularly primary care services) need to alter preconception care services to attract type 2 diabetic women and alter their way of thinking. It also highlights the importance of gathering information such as hidden agendas carefully from patients; this can be through the Cambridge-Calgary consultation skills model whereby patient ideas, concerns and expectations are gathered [27]. This model also focuses strongly on nonverbal communication and being aware of recognising, acknowledging and validating patient concerns [27]. Gross and Kinnison [28] cited the protection motivation theory by Rogers (1975) which helps to identify potential reasons why women do not access services to help their health. It shows how fear may influence women in changing health behaviours. The model summarises the concept of efficacy of health behaviour (response efficacy) and the amount of confidence a women may have to carry out an action (self-efficacy) against the risks and benefits of it [28]. The concept of self-efficacy is a key component of this model as it helps to reflect motivation in changing health behaviours. For instance, low self-efficacy would help to reflect why women do not access healthcare services. For instance acknowledging their pregnancy could have risks may result in them experiencing stress and anxiety. As a result this may discourage women in accessing preconception services. However the national service framework for diabetes set by the Department of Health (2001) [29] states that the aim of maternity care is to ensure all pregnant women have a positive experience of their pregnancy that promotes physical and psychological health. It also acknowledges that some women experiencing constant ‘medicalised’ care may feel less encouraged about their pregnancy.