Four guidelines were identified in PubMed, four in the Guideline Clearing House and seven in Google. Guidelines were excluded for different reasons; the guideline of the International Diabetes Center [
15] was excluded because this guideline was similar to the guideline of the American Diabetes Association [
16,
17]. Other guidelines were excluded because they did not concern Western populations [
18], did not focus on diabetes in pregnancy [
19,
20] or because of publication date [
21,
22].
Quality assessment
There was a high level of agreement between the two independent reviewers. In general, the quality of most guidelines was questionable when using the AGREE instrument. The domains with the highest scores were applicability, clarity and presentation and scope and purpose. The domains editorial independence, stakeholder involvement and rigour of development had the lowest scores. The domain scores of each guideline are shown in Table
2.
Table 2
Domain Scores of the guidelines (percentages)
Scope and purpose | 67 | 56 | 39 | 33 | 28 | 61 | 67 | 94 |
Stakeholder involvement | 17 | 8 | 42 | 4 | 14 | 21 | 8 | 100 |
Rigour of development | 48 | 29 | 38 | 12 | 31 | 31 | 38 | 93 |
Clarity and presentation | 67 | 71 | 75 | 50 | 83 | 63 | 71 | 100 |
Applicability | 44 | 56 | 56 | 33 | 39 | 56 | 94 | 89 |
Editorial independence | 0 | 0 | 42 | 0 | 0 | 8 | 50 | 50 |
The majority of the guidelines described their scope and purpose quite specific. Four guidelines adequately described both objectives and patient groups [
16,
17,
29‐
31]. In the other guidelines description of objectives was incomplete [
23‐
28]. Moreover, specific clinical questions were only provided by the NICE guideline [
31].
Regarding stakeholder involvement, seven guidelines scored below 50%. Patients’ view and preferences were often not included in the development of the guideline. Moreover, the guidelines had not been piloted among target users, with the exception of the NICE guideline [
31].
Seven guidelines had a low score on rigour of development. This was mainly because most guidelines did not describe explicitly how they had identified, selected and summarized the available evidence. Also information on updating was not reported adequately in most guidelines; only the NICE guideline [
31] provided a procedure for updating.
In contrast, except for the guideline from the BCRCP [
26,
27], all guidelines provided information on health benefits, side effects or adverse effects. In addition, most guidelines provided an explicit link between recommendations and evidence by references or a summary of the evidence. Only the CCGC guideline [
28] did not provide adequate information on the relation between evidence and recommendation.
Regarding clarity and presentation, seven guidelines scored over 50%. Key recommendations were easily identifiable and different options for management were given. It should be noted that only the CCGC guideline [
28] and NICE guideline [
31] were supported with tools for application, such as a quick reference guide.
On applicability, five guidelines scored over 50%. All guidelines presented key review criteria for monitoring and/or audit purposes. Considerations on costs or additional resources were described by four guidelines [
16,
17,
24,
25,
30,
31]. The guidelines of the DSOG [
29], IDF [
30] and NICE [
31] also provided information on potential organizational barriers.
Regarding editorial independence, all eight guidelines scored at or below 50%. Five guidelines did not report anything on both independency of funding body and conflicts of interest [
16,
17,
23‐
25,
28,
29]. The other three guidelines only fulfilled one of the two items [
26,
27,
30,
31].
Comparison of recommendations on management of diabetes during pregnancy
The recommendations in the guidelines can be divided into four domains, namely screening for GDM, glycemic control, prenatal care and labour and preconception counselling for women with DM1 or 2. The similarities and differences between the different guidelines on each domain will be discussed.
The first domain is screening for GDM. The recommendations in this domain were inconsistent. Three guidelines recommended that all pregnant women should be screened [
24,
25,
28,
30]. On the other hand, four guidelines recommended that only women with risk factors should be screened [
16,
17,
26,
27,
29,
31]. However, some guidelines refined their recommendation. The ADIPS guideline [
24,
25] stated that selective screening should be used when resources are limited. Furthermore, the IDF guideline [
30] stated that when the effectiveness of selective screening is shown, this should be recommended.
Not all guidelines were explicit about what screening criteria to use and at what time the screening should take place. Regarding the screening criteria, two guidelines stated that the Coustan & Carpenter criteria should be used [
26‐
28] and one guideline [
24,
25] recommended modified WHO-criteria. Concerning the time of screening, the BCRCP guideline [
26,
27] stated that it is not important to screen early while the CCGC guideline [
28] distinguished between women with and without risk factors. Women with risk factors should be screened as early in pregnancy as possible. For women without risk factors, early screening is not necessary.
The second domain is glycemic control. All guidelines made similar recommendations such as, among others;
Use a multidisciplinary approach that is adjusted to the individual.
Primary strategy is focused on nutrition and physical activity.
When glucose levels remain too high, medication therapy should be started.
Insulin is the primary choice; the use of hypoglycaemic agents is discouraged.
Blood glucose should be self-monitored regularly and the importance of maintaining normal glucose levels should be emphasized.
The third domain is prenatal care and labour. Two recommendations were similar for most guidelines, namely; 1) extra fetal surveillance is not necessary unless there are complications and 2) during labour, it is important to maintain normal glucose values. Therefore, glucose values should be monitored regularly.
Some guidelines made additional recommendations. For example, the ADIPS guideline [
24,
25] stated that after delivery, women with pre-existing DM should be monitored closely to find a new balance. Furthermore, some guidelines [
16,
17,
23,
29‐
31] stated that during prenatal care there should be searched for possible complications.
However, there was one important difference between the guidelines. Four guidelines recommended that delivery should not take place before full term unless there are complications [
23‐
28]. In contrast to this, three guidelines stated that delivery should be induced after approximately 38 weeks of gestation [
16,
17,
29,
31]. The IDF guideline [
30] did not make recommendations on this topic.
The last domain is preconception counselling for women with pre-existing DM. The recommendations in this domain were similar for most guidelines. The first recommendation was that all women with DM1 or 2 in reproductive age should receive counselling. This counselling should include information about the risks in pregnancies complicated by diabetes. Also the importance of normal glucose values before conception should be stressed [
16,
17,
23‐
31]. Other recommendations in this domain discussed the use of medication. Medication that is used before conception, for example hypoglycaemic agents, should be evaluated to determine if it is safe during pregnancy [
16,
17,
23‐
31]. In addition, it must be checked if there are any complications, such as retinopathy, present before conception [
16,
17,
23,
29].