Background

Diabetes mellitus is the most common medical conditions complicating pregnancy which has been reported in several countries including in the United Kingdom, United States of America and Australia [1,2,3]. Reports from these countries are showing increasing prevalence of pregnancies affected by diabetes throughout the years, in line with the increasing trend of diabetes prevalence among the general population globally [4, 5]. Diabetes in pregnancy is either pre-existing diabetes, where the diagnosis is prior to conception, or gestational diabetes, where diabetes is first detected during pregnancy. It has been established that pregnancies with diabetes mellitus are associated with adverse outcomes both to the mothers and babies compared to pregnancies without diabetes mellitus, including increased risk of pre-eclampsia, congenital malformations, spontaneous miscarriage, premature birth, foetal growth restriction, stillbirth, neonatal hypoglycaemia and birth trauma due to foetal macrosomia [6]. Pregnant women with pre-existing diabetes-related micro-vasculopathy are also at higher risk of disease progression [7]. Comparable perinatal outcome has been seen between pregnancies with and without diabetes mellitus when the mothers achieved satisfactory disease control [8]. Therefore, postponing pregnancy in suboptimized conditions and stabilisation of diabetic control among women with pregnancy intention should be prioritized. Contraception advice is one of the preconception care components that should be provided to all women with diabetes mellitus within the reproductive age group [9,10,11,12] especially when a significant proportion of pregnancies around the world were unintended [13]. The higher percentage of unintended pregnancies among women with diabetes is also worrying as it represented a missed opportunity to provide preconception care which includes provision of counselling regarding the risks associated with diabetes mellitus during pregnancy, medication review to avoid possible teratogenic agents, folic acid supplementation, diet and weight loss advice, as well as screening for other diabetes-related complications [14, 15]. Extensive literatures are available on determinants of family planning behaviours among women of general population but in-depth evidence for women with specific condition such as diabetes mellitus are still minimal. There is no previous systematic review on family planning among women with diabetes mellitus. Women with risk factors such as diabetes mellitus may have different aims for family planning compared to women in the general population which are mainly to provide spacing and limit the number of children. This warrants for a review of all the evidence available related to family planning behaviours among women with diabetes mellitus to identify knowledge gaps with the ultimate aim of preventing unintended pregnancies among women with diabetes mellitus which is associated with higher risk of morbidity and mortality.

This scoping review is mainly aimed to determine the scope and map the types of evidence available related to family planning behaviours among women with diabetes mellitus and its determinants. This review also hopes to identify and analyse knowledge gaps in the family planning behaviours of women with diabetes mellitus in preventing unintended pregnancies which is associated with higher risk of morbidity and mortality.

Materials and methods

This scoping review was guided by methodological framework by Arksey and O’Malleys [16] guided by Prisma-ScR checklist as outlined by Tricco [17] which includes identifying the research question, identifying the relevant studies, study selection, charting the data, and collating, summarising and reporting the results.

Scoping review methodology was selected to reflect our broad objectives to map the available evidence related to family planning behaviours among women with diabetes mellitus and eventually to identify and analyse the gaps in knowledge surrounding this topic. To identify the research question, Population-Concept-Context (PCC) framework (Table 1) as recommended by Joanna Briggs Institute for Scoping Reviews was used to outline the key elements of the review [18] which led to our main question ‘What is the available evidence surrounding family planning usage or practice among women with diabetes mellitus’. After the initial literature search, a specific sub-question was identified, which is to determine the factors influencing the usage.

Table 1 PCC Framework
Full size table

Protocol and registration

Our protocol is drafted based on Preferred Reporting Items for Systematic reviews and Meta-Analyses extension for Scoping Reviews (PRISMA-ScR) [17], which was revised by the research team. The final protocol is registered with the Open Science Framework (https://osf.io/kv9hu/).

Eligibility criteria

To answer our research questions and achieve the objectives of the review, studies mentioning family planning or contraceptive usage among women with diabetes mellitus were included in the review. All studies including quantitative, qualitative and mixed-methods studies published in peer-reviewed journal were included to consider contextual factors influencing family planning behaviours. Studies from 2000 until February 2022 were selected as the wider duration hoped to allow any changes in trends of emerging evidences surrounding this topic. We excluded reviews and limited our studies to empirical research to focus on new knowledge and prevent redundancy. Studies were excluded if they do not contain evidence of family planning behaviours among women with diabetes mellitus and if they were not available in English (Table 2).

Table 2 Selection criteria for studies to be included in the review
Full size table

Information sources and search strategy

The first phase of the review involved identifying keywords and filters to be used in database search. Articles were identified using the following search terms: “family planning”, “contraceptive” and “diabetes mellitus”. Boolean terms such as “AND” and “OR” were used to separate keywords.(Additional File 1 for search strategy) Articles or relevant documents were identified by searching databases including PubMed, EBSCOhost, OVID and Web of Science Core Collection. Two different platforms searching MEDLINE database (i.e.; EBSCOhost and OVID) were used to ensure the breadth of the search as there might be variations in indexing that influence the search yield. Search was carried out according to PRISMA guidelines. The keywords searched is included in Table 3 below. Search was limited for specific field (title and/or abstracts) to capture only highly relevant articles. Articles were also searched by analyzing the reference lists of all the identified studies for additional articles or grey literature.

Table 3 Keyword searches
Full size table

Study selection process

The search results were imported into reference manager (Mendeley) where they were screened for duplicates. Three reviewers independently evaluated the titles, abstracts and then full text of all articles identified based on the eligibility criteria described above to minimise reporting bias as recommended by Joanna Briggs’s Institute [19]. Disagreements or inconsistencies among reviewers were resolved by discussion with other reviewers. Since the aim of the review is to study the family planning behaviour of women with diabetes mellitus, we excluded clinical trials or papers that focused on clinical efficacy or safety of contraceptive methods.

Data charting process

Information from the selected studies were extracted into data-charting form in Microsoft Excel which contains the following variables; authors, year of publication, title of study, country of study, objectives of study, study methods and design, study settings, sample size, types of study participants and outcomes or factors studied in the study. Data were charted independently by each reviewer simultaneously followed by discussions to determine consistency of data extraction to answer our objectives and research questions. Data-charting forms were updated continuously in an iterative process.

Data items

Data extracted includes articles’ characteristics (e.g., country of the study, study design, study population, study settings), types of information available on the main outcome (e.g., types of family planning methods, effectiveness of methods used, methods’ duration of action) and levels or categories of determinants measured (e.g., sociodemographic factors, clinical factors, interpersonal factors, institutional factors, community factors, policy factors). This categorization is based on Socioecological Model which was frequently adopted in previous systematic reviews describing family planning usage among general population [20,21,22,23,24]

Synthesis of results

Once the data extraction was completed, content analysis of the extracted data was conducted. Studies were grouped based on study designs, types of study participants and outcomes measured. Quantitative and qualitative information were analyzed separately at first. Quantitative findings were collated based on association of explanatory variables or determinants in different studies. The explanatory variables were summarized according to its statistical significance whether positively or negatively associated with family planning behaviour. Qualitative findings were organized into main themes and further analyzed to identify contextual factors that influenced family planning behaviour among women with diabetes mellitus.

Results

Selection of sources of evidence

A total of 1,447 articles were identified through our database search. After exclusion of duplicates, screening of abstracts and assessment of eligibility criteria, 35 articles that met the eligibility criteria were included in the review (Fig. 1). Most articles were excluded during the initial screening because they were reviews on contraceptive recommendations for women with diabetes mellitus and studies were focused on the effect of hormonal contraceptives on diabetes mellitus and metabolic profiles. Eight articles were excluded after assessing the full texts due to following reasons:

  1. 1.

    Studies did not report family planning behaviours as the outcome [25,26,27,28]

  2. 2.

    Studies did not report family planning practice among women with diabetes mellitus separately from other groups [29, 30]

Fig. 1
figure 1

PRISMA flow diagram [31, 32]

Full size image

Characteristics of sources evidence

The majority of the studies (n = 20) [25, 33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51] were conducted in the United States. The rest of the studies were conducted in Malaysia (n = 3) [52,53,54], United Kingdom (n = 4) [55,56,57,58], Iran (n = 2) [59, 60], Italy (n = 1) [61], Australia (n = 1) [62], France (n = 1) [63], South Africa (n = 1) [64], Ethiopia (n = 1) [65] and Jamaica (n = 1) [66]. Table 4 provides summary of the studies included in the review. Only one study was a qualitative study [52] and one study utilised mixed-methods approach [58]. The studies were spread out from 2003 until January 2022. Six studies utilized secondary data from pre-existing survey database [25, 37, 44, 46, 51] where three of them were nationwide survey [25, 44, 51], two were state-wide [45, 46] and one was combined from multiple states survey [37]. Five studies used insurance claim records [34, 38, 42, 47, 48], one study used a general practice database [57] and one study used a tertiary centre database [39]. Other studies were based on primary data collection. Sample size of the studies ranged from 12 to 7.5 million where the bigger sample sizes were usually from nationwide data that also included women in the general population in the analysis.

Table 4 Summary of characteristics of the studies
Full size table

Study populations varied where most of the studies (n = 27) did not specify the types of diabetes mellitus while six studies were specifically among women with Type 1 diabetes mellitus in specialist centre [10, 35, 39, 40, 49, 50] and three studies among Type 2 diabetes mellitus [43, 52, 60]. Nine studies reported diabetes mellitus with other selected medical conditions [37, 38, 45, 46, 51, 54, 60, 65, 66] and another six studies include all women within reproductive age group in the analysis [25, 34, 37, 42, 47, 48]. Two studies targeted specifically postpartum women with diabetes mellitus [33, 44]. Four studies were specifically among adolescents with diabetes mellitus [35, 36, 40, 49]. The age ranges to define ‘reproductive age’ in the studies varied, with a study used a limited range of 24 to 32 years old [25] and other studies with wider ranges i.e., 15 to 44 years old, 18 to 44 years old and 15 to 45 years old.

One study made comparison of the proportion of family planning usage based on disease status; diagnosed diabetes, undiagnosed diabetes, uncontrolled diabetes and no diabetes [25]. Six studies have explored pregnancy intention [33, 41, 45, 51, 53, 56] but only one study descriptively investigated the relationship between pregnancy intention in relation to contraceptive use of the study participants [56]. Most of the studies described the usage of family planning practice by types of methods (n = 26) [10, 25, 33,34,35, 38, 40,41,42, 44, 47,48,49,50,51, 54, 56,57,58,59,60,61,62, 64,65,66] and eight studies categorized the methods according to level of effectiveness [25, 41, 46, 48, 50, 53, 56, 62]. Only one study studied specifically the use of long-acting reversible contraception (LARC) among women with diabetes mellitus [39]. Five studies described the barriers of family planning usage or reasons of contraceptive non-use [33, 41, 52, 58, 65]; two of them through qualitative exploration [52, 58].

Results of individual sources of evidence

Main findings of the studies including the proportion of women using contraceptive methods and factors influencing family planning behaviours are summarised in Table 5.

Table 5 Summary of factors influencing family planning behaviours
Full size table

Measures of family planning usage among women with diabetes mellitus

Majority of the studies defined family planning practice as binary outcome whether any contraceptive methods were used or not used. In relation to temporal measure, most of the studies described family planning usage as ‘currently using’, but three studies specify the duration of usage within the last one to three years [38, 62, 64]. One study categorized family planning behaviours based on a composite score which take into consideration contraception use, receipt of preconception counselling and initiation of discussion with healthcare professionals [50].

Eight studies explored and classified the usage of contraception based on its effectiveness or efficacy as summarized in Table 4. Phillips-Bell et al. [46] found that women without diabetes were more likely to use more effective methods compared to women with diabetes which was parallel with findings from Britton et al. [25] and Schwarz et al. [48] which reported that women with diabetes mellitus was less likely to practice highly effective methods. Meanwhile, Morris & Tepper [44] reported that women with diabetes were more likely to use effective and long-acting methods. Leow et al. [53] found that only small proportion of women with diabetes were using highly effective methods and it did not correspond with their risk perception.

Prevalence and likelihood of family planning usage among women with diabetes mellitus

Among the studies that compared the likelihood of family planning usage among women with diabetes mellitus compared to women without diabetes, six studies reported that women with diabetes mellitus were less likely to use family planning methods [25, 38, 45, 48, 54, 57]. Only one study found that the likelihood of practicing family planning methods among women with diabetes mellitus were not significantly different compared to women without diabetes [37].

The prevalence of family planning methods used by women with diabetes mellitus varied across the reviewed literature ranging from 4.8 to 89.8% among the studied population.

Reasons for contraceptive non-use among women with diabetes mellitus

In studies that provided additional information on the reasons of contraceptive non-use, the most commonly described reasons were the misconceptions on the safety of contraceptive usage with the presence of diabetes mellitus, fear of side effects, and perceptions of reduced fertility with diabetes mellitus. Lack of preconception or contraceptive counselling provision during visits for diabetes care were also commonly reported reasons [37, 41, 65]. These reasons were elicited from quantitative studies using self-administered questionnaires. One of the studies utilized free-text field for this measure [33] while two other studies did not specify how reasons for non-use were addressed in the survey [41, 67].

Factors influencing family planning usage

The explanatory variables or determinants of family planning usage were grouped according to different levels (i.e., individual levels, interpersonal levels, community levels, institutional levels and policy levels) that were commonly described in family planning usage among general population. The findings are summarized in Table 4. Only five studies included beyond individual-level factors. One study addressed differences in family planning practice among women with diabetes mellitus in different regions [61] and three studies included types of healthcare facilities and service providers which were considered as institutional-level factors [39, 54, 61]. A qualitative study reported interpersonal-level factor as described below.

Individual-level factors

Sociodemographic characteristics that were frequently described to influence family planning behaviour in the studies among women with diabetes mellitus include age, ethnicity, marital status and education attainment. Age was shown to have negative association with family planning practice, where the younger age group were more likely to use contraception while only Morris et al. [44] reported otherwise. Horwitz [42] and Manaf [54] found no significant association between ethnicity and family planning non-users while three other reports found significant association between family planning usage and specific ethnic groups [25, 44, 51]. Education attainment of the women was another commonly reported factors that were positively associated with family planning usage and the findings were quite consistent [25, 33, 37, 44, 54, 60, 61]. However, when specifically studied for sterilisation, women with lower education attainment were more likely to have the procedure performed [44]. Women with middle- and high-level household income were more likely to practice family planning [65]. Along the same line, qualitative exploration also found that family planning usage were influenced by social implications of pregnancy including financial burden and career disruption [52].

Significant reproductive history that predicts family planning usage among women with diabetes mellitus includes parity and history of previous adverse pregnancy outcome. Mixed findings were reported in relation to parity and its association with family planning usage [25, 54, 60, 64, 65]. Qualitative exploration among women with diabetes mellitus reported that their experience in previous pregnancy influenced their decision to practice family planning methods where they were more inclined to practice when they had ‘previous difficult labour’ [52].

Women with diabetes mellitus who previously received contraceptive counselling were significantly more likely to practice family planning [39, 45, 65]. Knowledge on risk of congenital malformations were reported to be positively associated with family planning usage [63]. A descriptive study explored knowledge and usage of family planning methods based on its effectiveness. Women with diabetes mellitus in that study were found to have inadequate knowledge where their pregnancy intention and risk perception did not correspond to the usage of effective contraception methods [53]. Positive perception towards contraception were found to be positively associated with its usage [25, 54].

Association of disease control with family planning usage were explored in two studies. Mekonnen, Woldeyohannes & Yigzaw [67] reported that women with controlled condition were four-times more likely to practice contraception, while Britton et al. reported that nearly half of women with suboptimal HbA1c practiced less effective contraceptive methods [25]. Four other studies also grouped women with diabetes mellitus according to their disease control but their association with family planning behaviours was not explored.

Interpersonal-level factors

Only one of the study explored this interpersonal aspects where in-depth interviews among women with diabetes mellitus reported that opinions of significant others including friends, respected older female relatives, health care personnel, or religious leaders influenced their decision in family planning [52]. Desired family size was also influenced by partners and directly influenced the usage of family planning methods.

Institutional-level factors

A study compared the types of facility where the women received care by either in health clinics or in hospitals [54]. This study found that contraceptive non-use was significantly more common among women who received care in the health clinics as compared to those who received treatment in the hospital. Another study compared the types of healthcare providers who provided the care to the women with diabetes mellitus [39]. Not surprisingly women who were seen by gynaecologist or fetomaternal specialists were more likely to be counselled on the use of family planning. A study in Italy described the proportion of oral contraceptive pills prescribed by different specialists where majority were prescribed by gynaecologists [61]. One study categorized access to healthcare and studied its relationship with family planning usage found that it was associated with the usage of effective contraception [25].

Community-level factors

Only one study addressed community-level factor by comparing oral contraceptive usage among women with diabetes mellitus from different regions in Italy where they reported significant regional difference in usage prevalence [61].

Discussion

Summary of evidence

This scoping review identified 35 studies addressing family planning practice among women with diabetes mellitus published between 2003 and January 2022 which revealed family planning behaviours that vary throughout the studies. This review also revealed inconclusive estimates on the proportion of women with diabetes mellitus who were using family planning methods with a huge range of prevalence reported. This may be attributed to the widely-diverse study populations and settings. Study populations included in this review ranged from all women of reproductive age based on a nationwide survey to a specific group of adolescents with a specific type of diabetes mellitus in a tertiary setting. The types of diabetes mellitus may have a different impact on family planning behaviours [41]. Variation in the age group of study participants also portrays the complexity in determining the group of women at risk of pregnancy even though WHO has defined reproductive age group as 15 to 49 years old [68]. Approach to prevent unintended pregnancies also should be tailored to different age group of girls or women.

As almost half of the studies were carried out among women with diabetes mellitus in specialist centres, knowledge on family planning among women in general practice or primary care was still limited. Majority of women with uncomplicated diabetes mellitus received care in the primary care setting. Manaf [54] reported that women who received their treatment in health clinics were less likely to use contraception than women who received their treatment in hospital specialist clinics. This is possibly due to different disease profile and level of diabetic complications amongst the women who received specialist care in hospitals.

The majority of the studies had focused on descriptive characteristics of family planning behaviours where there is knowledge gap remains particularly with regards to associations of family planning usage and its influencing factors. Exploration of factors influencing family planning behaviours among women with diabetes mellitus were also very limited to sociodemographic and individual-level characteristics. There were very limited studies that acknowledged the determinants beyond individual levels which represents a gap in the knowledge. This is because studies among general population have established multilevel influences in determining family planning usage [69]. The single qualitative study included in this review can be considered as an important initial exploration as it revealed that interpersonal factors play a big role in the decision to practice family planning, but the findings should be interpreted with caution as the study is done in a single state in Malaysia with different ethnic proportion from the rest of the country [52]. This highlighted the areas with insufficient knowledge that limits our understanding on the factors that influence family planning behaviours among this specific population.

Designs of the studies included in the review may also give rise to inaccuracy of family planning behaviour measurement. As most of the studies were cross-sectional, the information on family planning behaviours were only measured based on their current usage of the contraceptive methods. However, three studies explored the usage or prescription of family planning over a duration of time. Nine of the studies were done retrospectively by reviewing medical databases and insurance claim records which may underestimated family planning usage when documentation omission or errors frequently occurred [39]. Utilisation of this secondary data enabled analysis of larger number of women. However, the family planning usage may be underestimated as it only captured prescription-based methods and methods covered by the insurance plan [34, 38]. Prescription of contraception by the healthcare providers also did not provide information whether the women actually used the contraception and reliance on insurance claims record had limited information on non-prescription methods [38].

Preventing unintended pregnancy among women with poorly controlled diabetes mellitus or with complications is the main strategy to prevent morbidity and mortality, but only four studies included information on disease control in their study [25, 39, 50, 61]. This may represent inadequate attention given to the main objective of pregnancy planning among high-risk women. Other relevant clinically-relevant evidence in women with diabetes mellitus especially when conditions were not optimized is the effectiveness of family planning methods used, but limited number of studies explored family planning usage among this population according to their effectiveness. As the compliance with usage is another important factor in preventing unintended pregnancy, the duration of usage and continuation rate of the methods were valuable but were not measured in any of the studies. Only one study reported compliance with methods among the participants where half of the women who were on oral contraceptives claimed that they regularly missed the pills [41]. Most studies were based on self-reported data on the status of disease during survey. Therefore, reporting bias may have been introduced [25, 37, 45, 46, 51].

Only four studies reported family planning usage as the outcome after an interventional study [36, 43, 50, 55]. This indicates that there is inadequate evidence on effective strategies to improve family planning uptake among women with diabetes mellitus.

Future studies with more clinical context in relation to family planning behaviours and diabetes mellitus including types of diabetes, disease control, presence of diabetes complications would be beneficial to target high-risk women in prevention of unintended pregnancies. Knowledge gaps at system-level limit the availability of evidence-based strategies that can be utilized by health practitioner or policymakers. System-level evidence is also needed to address further targeted innovations in family planning service provision to women with diabetes mellitus.

Strengths and limitations

This is the first published scoping review to explore the evidence surrounding family planning behaviours among specific population of women with diabetes mellitus. This review offers a comprehensive overview of the available evidence on family planning behaviours of women with diabetes mellitus which may contribute to the improvement of family planning or preconception service provision to improve the outcome of mothers and their children.

The limitations of this scoping review are also acknowledged. Stakeholder consultation has been suggested as one of the distinctive components of a scoping review that will give additional insight from the stakeholders’ perspectives. However, this was not carried out in our scoping review. Future studies that incorporate this valuable element is recommended to add methodological rigor and enhance the applicability of the review [16]. Apart from that, studies published in other languages were not included which may have contributed to more culture-specific predictors of family planning behaviours. However, the studies included in the review were from various regions and these studies could have attempted to represent studies in the region. This review also does not incorporate the critical appraisal of the evidences (Additional file 1).

Conclusion

The evidence on family planning behaviours among women with diabetes mellitus was limited to the sociodemographic factors. Future studies with more clinical and contextual factors are needed to guide the strengthening of family planning services for high-risk women specifically those with diabetes mellitus and limited knowledge on clinical and contextual factors that influence the behaviour.