+Model DIABET-661; No. of Pages 7
ARTICLE IN PRESS Available online at
ScienceDirect www.sciencedirect.com Diabetes & Metabolism xxx (2014) xxx–xxx
Original article
Screening for gestational diabetes in the Lombardy region: A population-based study F. Nicotra a , C. Molinari b , N. Dozio b , M.T. Castiglioni c , B. Ibrahim a , A. Zambon a,∗ , G. Corrao a , M. Scavini d a
Department of Statistics and Quantitative Methods, Division of Biostatistics, Epidemiology and Public Health, Laboratory of Healthcare Research and Pharmacoepidemiology, University of Milano-Bicocca, Via Bicocca degli Arcimboldi 8, 20126 Milan, Italy b Università Vita Salute San Raffaele, Department of Internal Medicine, Division of General Internal Medicine, Diabetes and Endocrine Diseases, IRCCS San Raffaele Scientific Institute, via Olgettina, 60, 20132 Milan, Italy c Department of Obstetrics and Gynecology, IRCCS San Raffaele Scientific Institute, via Olgettina, 60, 20132 Milan, Italy d Division of Immunology, Transplantation and Infectious Diseases, Diabetes Research Institute (DRI), IRCCS San Raffaele Scientific Institute, via Olgettina, 60, 20132 Milan, Italy Received 8 July 2014; received in revised form 24 September 2014; accepted 18 November 2014
Abstract Aim. – As the treatment of hyperglycaemia during pregnancy with diet or insulin reduces the risk of adverse maternal outcomes and perinatal complications, screening for gestational diabetes mellitus (GDM) is included, albeit to variable extents, in all guidelines of care for pregnant women. The aim of the present investigation was to estimate the proportion of pregnancies screened for GDM in Lombardy between 2007 and 2010, and to identify predictors of screening. Methods. – A retrospective cross-sectional study using regional healthcare utilization databases of Lombardy was conducted. The study included all residents of Lombardy without pregestational diabetes who delivered between 1 January 2007 and 31 December 2010. The proportion of pregnancies with at least one screening test for GDM was calculated, along with the odds ratios and 95% confidence intervals associated with selected covariates for GDM screening. Results. – Of the 362,818 pregnancies included in the sample, 30% were screened for GDM. The proportion of pregnancies screened increased slightly from 2007 (27%) to 2010 (33%) and with maternal age (from 28% among women < 25 years to 32% among those ≥ 35 years), and varied widely across local health management organizations (HMOs) of residence (range: 20% to 68%). Socioeconomic indicators (education, immigrant status), obstetric history and prepregnancy hypertension were independent predictors of GDM screening. Conclusion. – The study finding of a low rate of pregnant women screened for GDM among residents of Lombardy supports the need for programmes to improve training of healthcare professionals, to raise women’s awareness of GDM and to eliminate barriers to GDM screening. © 2014 Elsevier Masson SAS. All rights reserved. Keywords: Gestational diabetes mellitus; Oral glucose tolerance test; Screening; Predictors; Cross-sectional study; Healthcare utilization databases
1. Introduction Gestational diabetes mellitus (GDM) has been defined as “any degree of glucose intolerance with onset or first recognition
∗
Corresponding author. Tel.: +39 02 64485814; fax: +39 02 64485899. E-mail address: [email protected] (A. Zambon).
during pregnancy” [1]. GDM is associated with an increased risk of adverse pregnancy outcomes for both the mother and child (such as preeclampsia, prematurity, caesarean section, macrosomia and neonatal hypoglycaemia) [2–4]. GDM also affects women and their children well beyond delivery. Compared with women without a history of GDM, women with a pregnancy complicated with GDM have a sevenfold increased risk of developing type 2 diabetes (T2D) in the years following childbirth [5].
http://dx.doi.org/10.1016/j.diabet.2014.11.008 1262-3636/© 2014 Elsevier Masson SAS. All rights reserved.
Please cite this article in press as: Nicotra F, et al. Screening for gestational diabetes in the Lombardy region: A population-based study. Diabetes Metab (2014), http://dx.doi.org/10.1016/j.diabet.2014.11.008
+Model DIABET-661; No. of Pages 7 2
ARTICLE IN PRESS F. Nicotra et al. / Diabetes & Metabolism xxx (2014) xxx–xxx
Moreover, children of mothers with GDM have an increased risk of developing obesity in childhood and adolescence, as well as GDM and T2D later in life [1,2,4,6–8]. Identifying women with GDM is important, as treatment of hyperglycaemia during pregnancy with either diet or insulin greatly reduces the risk of serious perinatal complications [9] and, less consistently, maternal outcomes [9–12]. Screening is essential for diagnosing GDM, as hyperglycaemia is usually mild and non-symptomatic [10]. In Europe, GDM is most often reported as affecting 2–6% of pregnancies [13]. However, more extreme values (ranging from 1% to 28%) have been observed in specific countries, depending on their sociodemographic characteristics, prevalence of diabetes and screening policies [13]. In 2010, the International Association of Diabetes and Pregnancy Study Groups (IADPSG) suggested significant changes to diagnostic criteria that were likely to increase the proportion of pregnancies with GDM to 18–20% [1,14]. Even though no universal consensus on screening procedures or diagnostic criteria has yet been reached [2,15–17], in almost all Western countries, including Italy, national healthcare providers and professional associations recommend screening for GDM in either all pregnant women (universal screening) or only those at increased risk of GDM [11,13,18–22]. Yet, it is unknown as to what extent recommendations for GDM screening are implemented [13], given the very limited information available in the literature on the proportion of pregnancies screened for GDM at the population level [23]. Therefore, the present cross-sectional study was conducted to estimate the proportion of pregnancies screened for GDM in Lombardy between 2007 and 2010, and to identify screening predictors, with the use of regional healthcare utilization (HCU) databases.
• an archive of outpatient diagnostic imaging and laboratory tests provided to beneficiaries; • a database of certificates of care at delivery (CEDAP), including information on the pregnancy, delivery, newborns and parents.
2. Materials and methods
• copayment exemptions for diabetes (code 013.250) granted at any time prior to the estimated start of gestation; • previous hospital-discharge including a diagnosis of diabetes (ICD-9 code: 249* to 250*); • three prescriptions for drugs used to treat diabetes (ATC code: A10), with at least one of those for either insulin or sulphonylurea, in the year preceding gestation.
2.1. Data sources The data analyzed in this retrospective cross-sectional study were retrieved from the electronic HCU databases of Lombardy, the largest region of Italy with nine million residents, 16% of the Italian population. The Italian National Healthcare Service (NHS) [24] provides full coverage to all residents for general practitioner (GP) care and hospitalizations, and coverage with copayment for diagnostic procedures and laboratory tests, specialist care and drug prescriptions. Exemptions from copayment are granted based on age or income, or for selected diseases or conditions. The delivery of NHS services to its beneficiaries is tracked using a system of HCU databases that includes: • an archive of NHS beneficiaries (practically the entire resident population), and their demographic and administrative data; • a hospital-discharge database, covering all discharges from public and private hospitals in Lombardy; • a dispensed-drug database, containing information on the drugs dispensed through the NHS; • a database of exemptions from copayment with the date granted;
The CEDAP is a nationwide mandatory questionnaire completed by the midwife or physician attending the delivery. Twice a year, the Ministry of Health issues a report based on the analysis of CEDAP data to guide planning of maternal services. Because healthcare coverage in Italy is universal, these databases provide complete and comprehensive information on all diagnostic procedures and laboratory tests, specialist care and prescription drugs provided to the entire population, and constitute a unique source of data for population-based epidemiological studies [25–29]. For every NHS beneficiary, information from different databases can be linked together through a non-informative identifier. 2.2. Cross-sectional sample The present study source population included all women residing in Lombardy who were NHS beneficiaries during the period 2007–2010. All of these women’s deliveries were identified by linking the CEDAP and hospital-discharge databases (ICD-9 codes: 370* to 375*). Those deliveries that were either missing gestational age or had gestational durations < 24 weeks or > 43 weeks were excluded, extending by 1 week the duration of gestation defined by Italian law [30] to account for any imprecise reporting of pregnancy duration. In addition, to exclude deliveries of women with pregestational type 1 diabetes (T1D) or T2D, there was no consideration of deliveries with:
The remaining deliveries constituted the study sample. 2.3. Screening for GDM In the period covered by the study, screening for GDM was recommended for women at increased risk and consisted of a 50g glucose challenge, followed by a 100-g oral glucose tolerance test (OGTT) for those with a positive challenge test. Increased risk for GDM included obesity, family history of T2D in firstdegree relatives, history of glucose intolerance and macrosomia in previous pregnancies [31]. 2.4. Covariates For each delivery, demographic and clinical information was retrieved from various data sources. GDM screening was
Please cite this article in press as: Nicotra F, et al. Screening for gestational diabetes in the Lombardy region: A population-based study. Diabetes Metab (2014), http://dx.doi.org/10.1016/j.diabet.2014.11.008
+Model DIABET-661; No. of Pages 7
ARTICLE IN PRESS F. Nicotra et al. / Diabetes & Metabolism xxx (2014) xxx–xxx
identified through records of a glucose challenge test or an OGTT (code: 90.26.4 or 90.26.5 from the outpatient diagnostic imaging/laboratory test archive) performed at any time during pregnancy. Maternal age at delivery and identification of the local health management organization (HMO) of residence were retrieved from the NHS beneficiaries archive. The CEDAP database was used to retrieve information about the mother (place of birth, marital status, employment, education), and about the pregnancy and delivery (multiple births, use of assisted fertilization, number of previous spontaneous abortions, induced abortions, stillbirths and live births), and the outpatients dispensed-drug database to retrieve information on exposure to at least one antihypertensive drug in the year prior to the first GDM screening (for women who were screened) or before gestational week 24 (for women not screened). 2.5. Statistical analyses Point prevalence of GDM screening was estimated as the ratio between the total number of deliveries with GDM screening and total number of deliveries in the study sample. Also, the normal approximated 95% confidence intervals (CI) for this proportion were computed. Descriptive statistics were reported as frequencies and proportions for categorical variables, and as means ± standard deviation (SD) for continuous variables. The Chi2 test for trend was used to compare ordinal variables. A generalized estimating equation (GEE) model for binary outcomes, assuming binomial probability and logit link functions, was implemented [32–34] to estimate determinants of GDM screening. This model takes into account the fact that, within the study period, repeated deliveries from the same women are not independent of each other (within-subject correlations). The estimated effects were expressed as odds ratios (ORs) and their 95% CI. The model considered all covariates
3
jointly to account for any potential confounding effect of each covariate on the others. SAS version 9.3 statistical software was used for the analyses (SAS Institute, Cary, NC, USA). For all hypotheses tested, twotailed P values < 0.05 were considered significant. 3. Results 3.1. Cross-sectional sample During the period 2007–2010, the regional official statistics reported 390,221 deliveries [35]. By linking up the CEDAP and hospital-discharge databases, 367,466 deliveries (involving 334,068 women) were identified. From this initial sample, 4648 deliveries met criteria for exclusion. Thus, a final sample of 362,818 deliveries (for 330,264 women) was identified (Fig. 1). Average maternal age at delivery was 32 ± 5 years. Also, around one-fourth (27%) of the deliveries where the mother’s country of birth was known (95% of all deliveries) involved foreign-born women. 3.2. GDM screening and screening predictors The proportion of pregnancies in our cross-sectional sample screened for GDM with either a glucose challenge test or an OGTT was 30% (95% CI: 30–30; 110,503 pregnancies). Fig. 2 shows that the first test for GDM screening was performed, on average, at 25 ± 4 weeks of gestation. The first, second and third quartiles of GDM screening distribution were at weeks 23, 25 and 27, respectively. The proportion of pregnancies screened for GDM after 32 weeks was 4%. Table 1 presents the total number of deliveries by demographic, clinical and other characteristics, the relative proportion of deliveries with GDM screening and the OR estimates for GDM screening derived from the GEE model. The proportion of pregnancies screened increased slightly over time, from 27%
Fig. 1. Flow diagram of criteria for participation in this study to determine gestational diabetes mellitus screening in pregnant women in Lombardy.
Please cite this article in press as: Nicotra F, et al. Screening for gestational diabetes in the Lombardy region: A population-based study. Diabetes Metab (2014), http://dx.doi.org/10.1016/j.diabet.2014.11.008
+Model DIABET-661; No. of Pages 7 4
ARTICLE IN PRESS F. Nicotra et al. / Diabetes & Metabolism xxx (2014) xxx–xxx
Table 1 Proportions of gestational diabetes mellitus (GDM) screening according to deliveries and covariates, with estimates of association derived by the generalized estimating equation (GEE) model. Deliveries (n)
GDM screening (%)
Odds ratio
95% CI
88,062 92,220 92,371 90,165
27 29 32 33
1.00 1.10 1.21 1.29
(reference) (1.08–1.12) (1.18–1.23) (1.26–1.32)
Maternal age (years) < 25 25–34 ≥ 35
36,595 211,777 114,446
28 30 32
1.00 1.15 1.25
(reference) (1.12–1.18) (1.21–1.29)
Place of birth Italy Foreign-born
253,784 91,390
30 32
1.00 1.27
(reference) (1.24–1.30)
53,597 34,051 23,544 32,143 44,636 47,163 21,535 12,650 12,748 8004 13,860 17,175 6104 31,948 3660
24 34 28 35 22 20 51 20 43 47 35 24 31 41 68
1.00 1.73 1.23 1.77 0.97 0.95 3.49 0.79 2.52 3.20 1.76 1.16 1.56 2.29 8.17
(reference) (1.68–1.79) (1.18–1.27) (1.71–1.83) (0.94–1.00) (0.92–0.99) (3.37–3.62) (0.75–0.83) (2.41–2.63) (3.03–3.39) (1.69–1.84) (1.11–1.22) (1.47–1.66) (2.22–2.37) (7.54–8.85)
Marital status Unmarried Married
80,373 245,324
31 31
1.00 1.04
(reference) (1.02–1.06)
Employment Student Housewife Unemployed Employed
2667 84,311 13,082 238,581
29 31 32 31
1.00 1.00 1.08 1.08
(reference) (0.91–1.09) (0.98–1.19) (0.98–1.18)
Education Primary school/no school Secondary/high school Bachelor’s/master’s degree
7586 245,375 83,307
30 32 29
1.00 1.10 0.95
(reference) (1.04–1.16) (0.90–1.01)
Use of ART No Yes
355,461 4778
30 37
1.00 1.11
(reference) (1.04–1.18)
Multiple births No Yes
357,228 5590
30 42
1.00 1.65
(reference) (1.56–1.75)
Previous spontaneous abortions 0 1 2 ≥3
301,297 48,899 9731 2888
30 32 34 38
1.00 1.09 1.18 1.41
(reference) (1.07–1.11) (1.12–1.23) (1.30–1.53)
Previous induced abortions 0 ≥1
344,098 18,718
30 34
1.00 1.16
(reference) (1.12–1.20)
Previous stillbirths 0 ≥1
360,401 2414
30 38
1.00 1.42
(reference) (1.29–1.55)
Delivery year 2007 2008 2009 2010
Local HMO of residence Milano Milano 1 Milano 2 Milano 3 Bergamo Brescia Como Cremona Lecco Lodi Mantova Pavia Sondrio Varese Vallecamonica-Sebino
Please cite this article in press as: Nicotra F, et al. Screening for gestational diabetes in the Lombardy region: A population-based study. Diabetes Metab (2014), http://dx.doi.org/10.1016/j.diabet.2014.11.008
+Model
ARTICLE IN PRESS
DIABET-661; No. of Pages 7
F. Nicotra et al. / Diabetes & Metabolism xxx (2014) xxx–xxx
5
Table 1 (Continued) Deliveries (n)
GDM screening (%)
Odds ratio
95% CI
Previous live births 0 ≥1
199,828 162,988
31 30
1.00 0.95
(reference) (0.94–0.97)
Hypertension No Yes
359,657 3161
30 42
1.00 1.51
(reference) (1.40–1.64)
ART: assisted reproductive technology.
in 2007 to 33% in 2010 (P value for trend: < 0.0001), and with maternal age at delivery, from 28% in women aged < 25 years to 32% for women ≥ 35 years (P value for trend: < 0.0001). Foreign-born women had a 27% (95% CI: 24–30%) greater probability of being screened for GDM than women born in Italy. There was also a wide heterogeneity in the proportion of pregnancies screened for GDM among different HMOs of residence, from 20% in Brescia to 68% in Vallecamonica-Sebino. There was no association between the mother’s employment and probability of GDM screening. However, women with secondary/high school education had a 10% higher probability of being screened (95% CI: 4–16%) than those with only primary school or no school attendance. The mother’s clinical history also affected the probability of GDM screening. Compared with women who had never had a spontaneous abortion, those with one, two or three or more previous abortions were screened more frequently: 9% (95% CI: 7–11%), 18% (95% CI: 12–23%) and 41% (95% CI: 30–53%), respectively (P value for trend: < 0.0001). In addition, women who had used assisted reproductive technology had an 11% (95% CI: 4–18%) greater probability of being screened than women who did not. Similar findings were observed for women with multiple births, previous induced abortions, previous stillbirths and hypertension, whose probability of GDM screening increased by 65% (95% CI: 56–75%), 16% (95% CI: 12–20%), 42% (95% CI: 29–55%) and 51% (95% CI: 40–64%), respectively, compared with women with single births, no previous induced abortions, no previous stillbirths and no hypertension.
4. Discussion Very few epidemiological studies have documented adherence to GDM screening policies during pregnancy, although GDM is associated with an increased risk of adverse pregnancy outcomes, maternal T2D later in life and obesity, and later T2D in the child [3,7–9]. The present study is the first in Italy, and one of the few worldwide, to provide population-based estimates of GDM screening proportions and population-based predictors of GDM screening. From the HCU databases for Lombardy, which included more than 360,000 resident women who delivered during 2007–2010, it was estimated that only 30% of pregnancies were screened for GDM. The proportion of pregnancies screened increased slightly over the years covered by the study and with maternal age, and varied widely among HMOs. Socioeconomic indicators, obstetric history and maternal hypertension were independent predictors of GDM screening. The observed proportion of pregnancies screened in Lombardy was lower than reported in other countries; for example, proportions were 89% in Israel, where a universal screening policy is implemented, and 68% in the US, involving women who are beneficiaries of healthcare insurance aged > 25 years [23,36]. In Lombardy, the percentage of women screened for GDM among those aged ≥ 35 was 32% compared with 71% in the US study [36]. Also, of women included in the study screened for GDM, 33% were aged ≥ 35 years, 28% were immigrants and 2% used assisted reproductive technology whereas, in Israel, these percentages were 21%, 17% and 6%, respectively [23].
Recommended screening (24-28 weeks): 53.9%
Proportion of pregnancies screened for GDM (%)
16
14 12 10 Early screening (≤18 weeks): 5.7%
8
Late screening (≥32 weeks): 4.3%
6
4 2 0
0
2
4
6
8
10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 Gestational week
Fig. 2. Distribution of the first gestational diabetes mellitus (GDM) screening test by gestational week.
Please cite this article in press as: Nicotra F, et al. Screening for gestational diabetes in the Lombardy region: A population-based study. Diabetes Metab (2014), http://dx.doi.org/10.1016/j.diabet.2014.11.008
+Model DIABET-661; No. of Pages 7 6
ARTICLE IN PRESS F. Nicotra et al. / Diabetes & Metabolism xxx (2014) xxx–xxx
Although GDM screening rates increased from 2007 to 2010, they were still low, considering that the national guidelines at the time recommended GDM screening for women with risk factors, including maternal age > 25 years [31]. If, in the cross-sectional sample, only maternal age > 25 years was considered a risk factor for GDM, the proportion of pregnancies for which GDM screening was recommended would have been approximately 90% – three times what was actually observed. Many factors may have contributed to the low proportion of pregnancies screened for GDM in Lombardy. First, healthcare professionals may have been poorly compliant with guidelines, as suggested by the small increase in the proportion of pregnancies screened for GDM with increasing maternal age. GDM screening was only 32% among women aged ≥ 35 years, for whom GDM incidence is estimated to be > 15% [37]. This is especially relevant, as women in Western countries are now postponing their first pregnancy to their third decade of life [38]. Even among women whose pregnancy is likely to receive increased medical attention (users of assisted reproductive technology, twin pregnancies, women with several previous spontaneous abortions, still births and hypertension), the proportion of GDM screening was only slightly higher than among women with spontaneous singleton pregnancies, no obstetric history and no hypertension, thus, far from reflecting any appreciation of the increased GDM risk in these subgroups. This may be explained, at least in part, by the fact that these conditions are usually not listed as risk factors for GDM in screening guidelines and also because delivery by caesarean section is often planned for these women, thereby making macrosomia a lesser concern. Among physicians, conflicting and frequently revised guidelines, a perceived low value of treating hyperglycaemia during pregnancy despite the evidence accumulated over the past few years [9–12], concerns over excessive medicalization of pregnancy and economic pressure to contain healthcare expenditures may all have contributed to low adherence to GDM screening policies. On the other hand, pregnant women may be poorly compliant with GDM screening because of limited awareness of GDM-associated risks, and fear of the discomfort of being diagnosed with GDM and having to take insulin. Nevertheless, compliance during pregnancy in women is usually high, as reported for screening of transmittable diseases, and cessation of smoking and alcohol consumption [39,40] during pregnancy. Furthermore, in our study, 79% of the women screened for GDM had their test within 2 weeks of the recommended timing (24 to 28 weeks of gestation), supporting women’s compliance with GDM screening. Also, it cannot be excluded that some women may have elected for GDM screening outside of the NHS, in which case the HCU database would not have recorded the event. However, 95% of the women in the cross-sectional sample had blood glucose tests performed during pregnancy through the NHS, making it unlikely that GDM screening outside the NHS contributed significantly to the low screening proportion for GDM in Lombardy during 2007–2010. The wide variability in the proportion of pregnancies screened in different HMOs is more likely to reflect differences in attitude towards GDM screening by prescribers rather than due to any heterogeneity in the underlying GDM risk of the
residents. This is supported by the fact that the distribution of covariates included in our multivariate model was similar among the HMOs in Lombardy (data not shown). It is acknowledged that this study had limitations. First, HCU databases do not include information such as anthropometrics, family history of diabetes and obstetric history, which would have allowed a precise identification of the population of pregnant women at risk of GDM who should have undergone screening. Also, adding risk factors other than maternal age ≥ 25 years would have had only limited impact, given the small number of women aged < 25 years (10% of our study sample) in whom GDM screening should have been prescribed. Second, although Lombardy accounts for 16% of the entire Italian population, the study results may not be applicable to other Italian regions, as all regions are autonomous in implementing national clinical practice guidelines for their residents [24]. Third, the use of such a large sample of deliveries may have highlighted some associations of limited or no clinical relevance. The potential implications of the low proportion of women screened for GDM documented in Lombardy are twofold. First, women with GDM not diagnosed because of lack of screening are more likely to have a macrosomic child, thereby contributing to the proportion of caesarean sections in Lombardy which, although relatively low compared with other Italian regions, is still among the highest in Europe [41]. Second, women not identified as having GDM during pregnancy because of lack of screening miss out on interventional programmes for primary prevention of T2D after delivery [42]. In conclusion, the present findings support the need for programmes to improve training of healthcare professionals, to raise women’s awareness of GDM and to eliminate barriers to GDM screening. In addition, the present study also demonstrates that HCU databases are an invaluable, albeit as yet mostly untapped, source of information on clinical practices in Italy. Also, future analyses should provide insights on the impact of using a single 75-g OGTT to screen for GDM. Disclosure of interest The authors declare that they have no conflicts of interest concerning this article. References [1] Metzger BE, Gabbe SG, Persson B, Buchanan TA, Catalano PA, Damm P, et al. International association of diabetes and pregnancy study groups recommendations on the diagnosis and classification of hyperglycaemia in pregnancy. Diabetes Care 2010;33:676–82. [2] Mulla WR, Henry TQ, Homko CJ. Gestational diabetes screening after HAPO: has anything changed? Curr Diab Rep 2010;10:224–8. [3] Metzger BE, Lowe LP, Dyer AR, Trimble ER, Chaovarindr U, Coustan DR, et al. Hyperglycaemia and adverse pregnancy outcomes. N Engl J Med 2008;358:1991–2002. [4] Jovanovic L, Pettitt DJ. Gestational diabetes mellitus. JAMA 2001;286:2516–8. [5] Bellamy L, Casas JP, Hingorani AD, Williams D. Type 2 diabetes mellitus after gestational diabetes: a systematic review and meta-analysis. Lancet 2009;373:1773–9.
Please cite this article in press as: Nicotra F, et al. Screening for gestational diabetes in the Lombardy region: A population-based study. Diabetes Metab (2014), http://dx.doi.org/10.1016/j.diabet.2014.11.008
+Model DIABET-661; No. of Pages 7
ARTICLE IN PRESS F. Nicotra et al. / Diabetes & Metabolism xxx (2014) xxx–xxx
[6] Huidobro A, Prentice A, Fulford T, Parodi C, Rozowski J. Gestational diabetes, comparison of women diagnosed in second and third trimester of pregnancy with non GDM women: analysis of a cohort study. Rev Med Chil 2010;138:316–21. [7] Burguet A. Long-term outcome in children of mothers with gestational diabetes. Diabetes Metab 2010;36:682–94. [8] Wu CS, Nohr EA, Bech BH, Vestergaard M, Olsen J. Long-term health outcomes in children born to mothers with diabetes: a population-based cohort study. PLoS One 2012;7:e36727. [9] Crowther CA, Hiller JE, Moss JR, McPhee AJ, Jeffries WS, Robinson JS, et al. Effect of treatment of gestational diabetes mellitus on pregnancy outcomes. N Engl J Med 2005;352:2477–86. [10] van Leeuwen M, Louwerse MD, Opmeer BC, Limpens J, Serlie MJ, Reitsma JB, et al. Glucose challenge test for detecting gestational diabetes mellitus: a systematic review. BJOG 2012;119:393–401. [11] Landon MB, Spong CY, Thom E, Carpenter MW, Ramin SM, Casey B, et al. A multicenter, randomized trial of treatment for mild gestational diabetes. N Engl J Med 2009;361:1339–48. [12] Langer O, Yogev Y, Most O, Xenakis EM. Gestational diabetes: the consequences of not treating. Am J Obstet Gynecol 2005;192:989–97. [13] Buckley BS, Harreiter J, Damm P, Corcoy R, Chico A, Simmons D, et al. Gestational diabetes mellitus in Europe: prevalence, current screening practice and barriers to screening. A review. Diabet Med 2012;29:844–54. [14] Sacks DA, Hadden DR, Maresh M, Deerochanawong C, Dyer AR, Metzger BE, et al. Frequency of gestational diabetes mellitus at collaborating centers based on IADPSG consensus panel-recommended criteria: the Hyperglycaemia and Adverse Pregnancy Outcome (HAPO) Study. Diabetes Care 2012;35:526–8. [15] Reece EA, Leguizamón G, Wiznitzer A. Gestational diabetes: the need for a common ground. Lancet 2009;373:1789–97. [16] Tieu J, Middleton P, McPhee AJ, Crowther CA. Screening and subsequent management for gestational diabetes for improving maternal and infant health. Cochrane Database Syst Rev 2010;7:CD007222. [17] Holt RI, Coleman MA, McCance DR. The implications of the new International Association of Diabetes and Pregnancy StudyGroups (IADPSG) diagnostic criteria for gestational diabetes. Diabet Med 2011;28:382–5. [18] Hoffman L, Nolan C, Wilson JD, Oats JJ, Simmons D. Gestational diabetes mellitus: management guidelines. The Australasian Diabetes in Pregnancy Society. Med J Aust 1998;169:93–7. [19] American Diabetes Association Workshop. Conference on gestational diabetes: summary and recommendations. Diabetes Care 1980;3:499–501. [20] National Collaborating Centre for Women’s and Children’s Health (UK). Diabetes in Pregnancy: Management of diabetes and its complications from preconception to the postnatal period. London: RCOG Press; 2008. [21] Scottish Intercollegiate Guidelines Network (SIGN). Management of diabetes. A national clinical guideline. Edinburgh: SIGN; 2010. [22] Collège national des gynécologues et obstétriciens franc¸ais, Société francophone du diabète. Recommandations pour la pratique clinique. Le diabète gestationnel. J Gynecol Obstet Biol Reprod 2010;39:S1–342. [23] Sella T, Shalev V, Elchalal U, Chovel-Sella A, Chodick G. Screening for gestational diabetes in the 21st century: a population-based cohort study in Israel. J Matern Fetal Neonatal Med 2013;26:412–6. [24] Lo Scalzo A, Donatini A, Orzella L, Cicchetti A, Profili S, Maresso A. Italy. Health system review. Health Syst Transit 2009;11:1–243.
7
[25] Di Bari M, Balzi D, Roberts AT, Barchielli A, Fumagalli S, Ungar A, et al. Prognostic stratification of older persons based on simple administrative data: development and validation of the “Silver Code” to be used in emergency department triage. J Gerontol A Biol Sci Med Sci 2010;65:159–64. [26] Corrao G, Conti V, Merlino L, Catapano AL, Mancia G. Results of a retrospective database analysis of adherence to statin therapy and risk of nonfatal ischemic heart disease in daily clinical practice in Italy. Clin Ther 2010;32:300–10. [27] Corrao G, Zambon A, Bertù L, Botteri E, Leoni O, Contiero P. Lipid lowering drugs prescription and the risk of peripheral neuropathy: an exploratory case-control study using automated databases. J Epidemiol Community Health 2004;58:1047–51. [28] Gini R, Francesconi P, Mazzaglia G, Cricelli I, Pasqua A, Gallina P, et al. Chronic disease prevalence from Italian administrative databases in the VALORE project: a validation through comparison of population estimates with general practice databases and national survey. BMC Public Health 2013;13:15. [29] Degli Esposti L, Saragoni S, Batacchi P, Benemei S, Geppetti P, Sturani A, et al. Adherence to statin treatment and health outcomes in an Italian cohort of newly treated patients: results from an administrative database analysis. Clin Ther 2012;34:190–9. [30] Art. 232 of the Italian Civil Code. [31] Metzger BE, Coustan DR. Summary and recommendations of the Fourth International Workshop – Conference on Gestational Diabetes Mellitus. The Organizing Committee. Diabetes Care 1998;21(Suppl 2):B161–7. [32] Hanley JA, Negassa A, Edwardes MD, Forrester JE. Statistical analysis of correlated data using generalized estimating equations: an orientation. Am J Epidemiol 2003;157:364–75. [33] Tirosh D, Benshalom-Tirosh N, Novack L, Press F, Beer-Weisel R, Wiznitzer A, et al. Hypothyroidism and diabetes mellitus: a risky dual gestational endocrinopathy. Peer J 2013;1:e52. [34] Peeters GM, Burton NW, Brown WJ. Associations between sitting time and a range of symptoms in mid-age women. Prev Med 2013;56:135–41. [35] http://www.asr-lombardia.it/ASR/sanita-e-salute/condizioni-di-salute-/ lombardia-e- province/tavole/100405/ [36] Blatt AJ, Nakamoto JM, Kaufman HW. Gaps in diabetes screening during pregnancy and postpartum. Obstet Gynecol 2011;117:61–8. [37] Mayorga ME, Reifsnider OS, Neyens DM, Gebregziabher MG, Hunt KJ. Simulated estimates of pre-pregnancy and gestational diabetes mellitus in the US: 1980 to 2008. PLoS One 2013;8:e73437. [38] Matthews TJ, Hamilton BE. Delayed childbearing: more women are having their first child later in life. NCHS Data Brief 2009;21:1–8. [39] Frischknecht F, Sell W, Trummer I, Brühwiler H. Serological testing for infectious diseases in pregnant women: are the guidelines followed? Swiss Med Wkly 2011;140:w13138. [40] Yamada H, Tairaku S, Morioka I, Ebina Y, Sonoyama A, Tanimura K, et al. A nationwide survey of maternal screening for mother-to-child infections in Japan. Congenit Anom (Kyoto) 2014;54:100–3. [41] Narayanan RB, Girdhar A, Girdhar BK, Malaviya GN. Immunohistological analysis of nerve granulomas in neuritic leprosy. Int Arch Allergy Appl Immunol 1990;92:50–5. [42] Bentley-Lewis R, Levkoff S, Stuebe A, Seely EW. Gestational diabetes mellitus: postpartum opportunities for the diagnosis and prevention of type 2 diabetes mellitus. Nat Clin Pract Endocrinol Metab 2008;4:552–8.
Please cite this article in press as: Nicotra F, et al. Screening for gestational diabetes in the Lombardy region: A population-based study. Diabetes Metab (2014), http://dx.doi.org/10.1016/j.diabet.2014.11.008
ARTICLE IN PRESS Available online at
ScienceDirect www.sciencedirect.com Diabetes & Metabolism xxx (2014) xxx–xxx
Original article
Screening for gestational diabetes in the Lombardy region: A population-based study F. Nicotra a , C. Molinari b , N. Dozio b , M.T. Castiglioni c , B. Ibrahim a , A. Zambon a,∗ , G. Corrao a , M. Scavini d a
Department of Statistics and Quantitative Methods, Division of Biostatistics, Epidemiology and Public Health, Laboratory of Healthcare Research and Pharmacoepidemiology, University of Milano-Bicocca, Via Bicocca degli Arcimboldi 8, 20126 Milan, Italy b Università Vita Salute San Raffaele, Department of Internal Medicine, Division of General Internal Medicine, Diabetes and Endocrine Diseases, IRCCS San Raffaele Scientific Institute, via Olgettina, 60, 20132 Milan, Italy c Department of Obstetrics and Gynecology, IRCCS San Raffaele Scientific Institute, via Olgettina, 60, 20132 Milan, Italy d Division of Immunology, Transplantation and Infectious Diseases, Diabetes Research Institute (DRI), IRCCS San Raffaele Scientific Institute, via Olgettina, 60, 20132 Milan, Italy Received 8 July 2014; received in revised form 24 September 2014; accepted 18 November 2014
Abstract Aim. – As the treatment of hyperglycaemia during pregnancy with diet or insulin reduces the risk of adverse maternal outcomes and perinatal complications, screening for gestational diabetes mellitus (GDM) is included, albeit to variable extents, in all guidelines of care for pregnant women. The aim of the present investigation was to estimate the proportion of pregnancies screened for GDM in Lombardy between 2007 and 2010, and to identify predictors of screening. Methods. – A retrospective cross-sectional study using regional healthcare utilization databases of Lombardy was conducted. The study included all residents of Lombardy without pregestational diabetes who delivered between 1 January 2007 and 31 December 2010. The proportion of pregnancies with at least one screening test for GDM was calculated, along with the odds ratios and 95% confidence intervals associated with selected covariates for GDM screening. Results. – Of the 362,818 pregnancies included in the sample, 30% were screened for GDM. The proportion of pregnancies screened increased slightly from 2007 (27%) to 2010 (33%) and with maternal age (from 28% among women < 25 years to 32% among those ≥ 35 years), and varied widely across local health management organizations (HMOs) of residence (range: 20% to 68%). Socioeconomic indicators (education, immigrant status), obstetric history and prepregnancy hypertension were independent predictors of GDM screening. Conclusion. – The study finding of a low rate of pregnant women screened for GDM among residents of Lombardy supports the need for programmes to improve training of healthcare professionals, to raise women’s awareness of GDM and to eliminate barriers to GDM screening. © 2014 Elsevier Masson SAS. All rights reserved. Keywords: Gestational diabetes mellitus; Oral glucose tolerance test; Screening; Predictors; Cross-sectional study; Healthcare utilization databases
1. Introduction Gestational diabetes mellitus (GDM) has been defined as “any degree of glucose intolerance with onset or first recognition
∗
Corresponding author. Tel.: +39 02 64485814; fax: +39 02 64485899. E-mail address: [email protected] (A. Zambon).
during pregnancy” [1]. GDM is associated with an increased risk of adverse pregnancy outcomes for both the mother and child (such as preeclampsia, prematurity, caesarean section, macrosomia and neonatal hypoglycaemia) [2–4]. GDM also affects women and their children well beyond delivery. Compared with women without a history of GDM, women with a pregnancy complicated with GDM have a sevenfold increased risk of developing type 2 diabetes (T2D) in the years following childbirth [5].
http://dx.doi.org/10.1016/j.diabet.2014.11.008 1262-3636/© 2014 Elsevier Masson SAS. All rights reserved.
Please cite this article in press as: Nicotra F, et al. Screening for gestational diabetes in the Lombardy region: A population-based study. Diabetes Metab (2014), http://dx.doi.org/10.1016/j.diabet.2014.11.008
+Model DIABET-661; No. of Pages 7 2
ARTICLE IN PRESS F. Nicotra et al. / Diabetes & Metabolism xxx (2014) xxx–xxx
Moreover, children of mothers with GDM have an increased risk of developing obesity in childhood and adolescence, as well as GDM and T2D later in life [1,2,4,6–8]. Identifying women with GDM is important, as treatment of hyperglycaemia during pregnancy with either diet or insulin greatly reduces the risk of serious perinatal complications [9] and, less consistently, maternal outcomes [9–12]. Screening is essential for diagnosing GDM, as hyperglycaemia is usually mild and non-symptomatic [10]. In Europe, GDM is most often reported as affecting 2–6% of pregnancies [13]. However, more extreme values (ranging from 1% to 28%) have been observed in specific countries, depending on their sociodemographic characteristics, prevalence of diabetes and screening policies [13]. In 2010, the International Association of Diabetes and Pregnancy Study Groups (IADPSG) suggested significant changes to diagnostic criteria that were likely to increase the proportion of pregnancies with GDM to 18–20% [1,14]. Even though no universal consensus on screening procedures or diagnostic criteria has yet been reached [2,15–17], in almost all Western countries, including Italy, national healthcare providers and professional associations recommend screening for GDM in either all pregnant women (universal screening) or only those at increased risk of GDM [11,13,18–22]. Yet, it is unknown as to what extent recommendations for GDM screening are implemented [13], given the very limited information available in the literature on the proportion of pregnancies screened for GDM at the population level [23]. Therefore, the present cross-sectional study was conducted to estimate the proportion of pregnancies screened for GDM in Lombardy between 2007 and 2010, and to identify screening predictors, with the use of regional healthcare utilization (HCU) databases.
• an archive of outpatient diagnostic imaging and laboratory tests provided to beneficiaries; • a database of certificates of care at delivery (CEDAP), including information on the pregnancy, delivery, newborns and parents.
2. Materials and methods
• copayment exemptions for diabetes (code 013.250) granted at any time prior to the estimated start of gestation; • previous hospital-discharge including a diagnosis of diabetes (ICD-9 code: 249* to 250*); • three prescriptions for drugs used to treat diabetes (ATC code: A10), with at least one of those for either insulin or sulphonylurea, in the year preceding gestation.
2.1. Data sources The data analyzed in this retrospective cross-sectional study were retrieved from the electronic HCU databases of Lombardy, the largest region of Italy with nine million residents, 16% of the Italian population. The Italian National Healthcare Service (NHS) [24] provides full coverage to all residents for general practitioner (GP) care and hospitalizations, and coverage with copayment for diagnostic procedures and laboratory tests, specialist care and drug prescriptions. Exemptions from copayment are granted based on age or income, or for selected diseases or conditions. The delivery of NHS services to its beneficiaries is tracked using a system of HCU databases that includes: • an archive of NHS beneficiaries (practically the entire resident population), and their demographic and administrative data; • a hospital-discharge database, covering all discharges from public and private hospitals in Lombardy; • a dispensed-drug database, containing information on the drugs dispensed through the NHS; • a database of exemptions from copayment with the date granted;
The CEDAP is a nationwide mandatory questionnaire completed by the midwife or physician attending the delivery. Twice a year, the Ministry of Health issues a report based on the analysis of CEDAP data to guide planning of maternal services. Because healthcare coverage in Italy is universal, these databases provide complete and comprehensive information on all diagnostic procedures and laboratory tests, specialist care and prescription drugs provided to the entire population, and constitute a unique source of data for population-based epidemiological studies [25–29]. For every NHS beneficiary, information from different databases can be linked together through a non-informative identifier. 2.2. Cross-sectional sample The present study source population included all women residing in Lombardy who were NHS beneficiaries during the period 2007–2010. All of these women’s deliveries were identified by linking the CEDAP and hospital-discharge databases (ICD-9 codes: 370* to 375*). Those deliveries that were either missing gestational age or had gestational durations < 24 weeks or > 43 weeks were excluded, extending by 1 week the duration of gestation defined by Italian law [30] to account for any imprecise reporting of pregnancy duration. In addition, to exclude deliveries of women with pregestational type 1 diabetes (T1D) or T2D, there was no consideration of deliveries with:
The remaining deliveries constituted the study sample. 2.3. Screening for GDM In the period covered by the study, screening for GDM was recommended for women at increased risk and consisted of a 50g glucose challenge, followed by a 100-g oral glucose tolerance test (OGTT) for those with a positive challenge test. Increased risk for GDM included obesity, family history of T2D in firstdegree relatives, history of glucose intolerance and macrosomia in previous pregnancies [31]. 2.4. Covariates For each delivery, demographic and clinical information was retrieved from various data sources. GDM screening was
Please cite this article in press as: Nicotra F, et al. Screening for gestational diabetes in the Lombardy region: A population-based study. Diabetes Metab (2014), http://dx.doi.org/10.1016/j.diabet.2014.11.008
+Model DIABET-661; No. of Pages 7
ARTICLE IN PRESS F. Nicotra et al. / Diabetes & Metabolism xxx (2014) xxx–xxx
identified through records of a glucose challenge test or an OGTT (code: 90.26.4 or 90.26.5 from the outpatient diagnostic imaging/laboratory test archive) performed at any time during pregnancy. Maternal age at delivery and identification of the local health management organization (HMO) of residence were retrieved from the NHS beneficiaries archive. The CEDAP database was used to retrieve information about the mother (place of birth, marital status, employment, education), and about the pregnancy and delivery (multiple births, use of assisted fertilization, number of previous spontaneous abortions, induced abortions, stillbirths and live births), and the outpatients dispensed-drug database to retrieve information on exposure to at least one antihypertensive drug in the year prior to the first GDM screening (for women who were screened) or before gestational week 24 (for women not screened). 2.5. Statistical analyses Point prevalence of GDM screening was estimated as the ratio between the total number of deliveries with GDM screening and total number of deliveries in the study sample. Also, the normal approximated 95% confidence intervals (CI) for this proportion were computed. Descriptive statistics were reported as frequencies and proportions for categorical variables, and as means ± standard deviation (SD) for continuous variables. The Chi2 test for trend was used to compare ordinal variables. A generalized estimating equation (GEE) model for binary outcomes, assuming binomial probability and logit link functions, was implemented [32–34] to estimate determinants of GDM screening. This model takes into account the fact that, within the study period, repeated deliveries from the same women are not independent of each other (within-subject correlations). The estimated effects were expressed as odds ratios (ORs) and their 95% CI. The model considered all covariates
3
jointly to account for any potential confounding effect of each covariate on the others. SAS version 9.3 statistical software was used for the analyses (SAS Institute, Cary, NC, USA). For all hypotheses tested, twotailed P values < 0.05 were considered significant. 3. Results 3.1. Cross-sectional sample During the period 2007–2010, the regional official statistics reported 390,221 deliveries [35]. By linking up the CEDAP and hospital-discharge databases, 367,466 deliveries (involving 334,068 women) were identified. From this initial sample, 4648 deliveries met criteria for exclusion. Thus, a final sample of 362,818 deliveries (for 330,264 women) was identified (Fig. 1). Average maternal age at delivery was 32 ± 5 years. Also, around one-fourth (27%) of the deliveries where the mother’s country of birth was known (95% of all deliveries) involved foreign-born women. 3.2. GDM screening and screening predictors The proportion of pregnancies in our cross-sectional sample screened for GDM with either a glucose challenge test or an OGTT was 30% (95% CI: 30–30; 110,503 pregnancies). Fig. 2 shows that the first test for GDM screening was performed, on average, at 25 ± 4 weeks of gestation. The first, second and third quartiles of GDM screening distribution were at weeks 23, 25 and 27, respectively. The proportion of pregnancies screened for GDM after 32 weeks was 4%. Table 1 presents the total number of deliveries by demographic, clinical and other characteristics, the relative proportion of deliveries with GDM screening and the OR estimates for GDM screening derived from the GEE model. The proportion of pregnancies screened increased slightly over time, from 27%
Fig. 1. Flow diagram of criteria for participation in this study to determine gestational diabetes mellitus screening in pregnant women in Lombardy.
Please cite this article in press as: Nicotra F, et al. Screening for gestational diabetes in the Lombardy region: A population-based study. Diabetes Metab (2014), http://dx.doi.org/10.1016/j.diabet.2014.11.008
+Model DIABET-661; No. of Pages 7 4
ARTICLE IN PRESS F. Nicotra et al. / Diabetes & Metabolism xxx (2014) xxx–xxx
Table 1 Proportions of gestational diabetes mellitus (GDM) screening according to deliveries and covariates, with estimates of association derived by the generalized estimating equation (GEE) model. Deliveries (n)
GDM screening (%)
Odds ratio
95% CI
88,062 92,220 92,371 90,165
27 29 32 33
1.00 1.10 1.21 1.29
(reference) (1.08–1.12) (1.18–1.23) (1.26–1.32)
Maternal age (years) < 25 25–34 ≥ 35
36,595 211,777 114,446
28 30 32
1.00 1.15 1.25
(reference) (1.12–1.18) (1.21–1.29)
Place of birth Italy Foreign-born
253,784 91,390
30 32
1.00 1.27
(reference) (1.24–1.30)
53,597 34,051 23,544 32,143 44,636 47,163 21,535 12,650 12,748 8004 13,860 17,175 6104 31,948 3660
24 34 28 35 22 20 51 20 43 47 35 24 31 41 68
1.00 1.73 1.23 1.77 0.97 0.95 3.49 0.79 2.52 3.20 1.76 1.16 1.56 2.29 8.17
(reference) (1.68–1.79) (1.18–1.27) (1.71–1.83) (0.94–1.00) (0.92–0.99) (3.37–3.62) (0.75–0.83) (2.41–2.63) (3.03–3.39) (1.69–1.84) (1.11–1.22) (1.47–1.66) (2.22–2.37) (7.54–8.85)
Marital status Unmarried Married
80,373 245,324
31 31
1.00 1.04
(reference) (1.02–1.06)
Employment Student Housewife Unemployed Employed
2667 84,311 13,082 238,581
29 31 32 31
1.00 1.00 1.08 1.08
(reference) (0.91–1.09) (0.98–1.19) (0.98–1.18)
Education Primary school/no school Secondary/high school Bachelor’s/master’s degree
7586 245,375 83,307
30 32 29
1.00 1.10 0.95
(reference) (1.04–1.16) (0.90–1.01)
Use of ART No Yes
355,461 4778
30 37
1.00 1.11
(reference) (1.04–1.18)
Multiple births No Yes
357,228 5590
30 42
1.00 1.65
(reference) (1.56–1.75)
Previous spontaneous abortions 0 1 2 ≥3
301,297 48,899 9731 2888
30 32 34 38
1.00 1.09 1.18 1.41
(reference) (1.07–1.11) (1.12–1.23) (1.30–1.53)
Previous induced abortions 0 ≥1
344,098 18,718
30 34
1.00 1.16
(reference) (1.12–1.20)
Previous stillbirths 0 ≥1
360,401 2414
30 38
1.00 1.42
(reference) (1.29–1.55)
Delivery year 2007 2008 2009 2010
Local HMO of residence Milano Milano 1 Milano 2 Milano 3 Bergamo Brescia Como Cremona Lecco Lodi Mantova Pavia Sondrio Varese Vallecamonica-Sebino
Please cite this article in press as: Nicotra F, et al. Screening for gestational diabetes in the Lombardy region: A population-based study. Diabetes Metab (2014), http://dx.doi.org/10.1016/j.diabet.2014.11.008
+Model
ARTICLE IN PRESS
DIABET-661; No. of Pages 7
F. Nicotra et al. / Diabetes & Metabolism xxx (2014) xxx–xxx
5
Table 1 (Continued) Deliveries (n)
GDM screening (%)
Odds ratio
95% CI
Previous live births 0 ≥1
199,828 162,988
31 30
1.00 0.95
(reference) (0.94–0.97)
Hypertension No Yes
359,657 3161
30 42
1.00 1.51
(reference) (1.40–1.64)
ART: assisted reproductive technology.
in 2007 to 33% in 2010 (P value for trend: < 0.0001), and with maternal age at delivery, from 28% in women aged < 25 years to 32% for women ≥ 35 years (P value for trend: < 0.0001). Foreign-born women had a 27% (95% CI: 24–30%) greater probability of being screened for GDM than women born in Italy. There was also a wide heterogeneity in the proportion of pregnancies screened for GDM among different HMOs of residence, from 20% in Brescia to 68% in Vallecamonica-Sebino. There was no association between the mother’s employment and probability of GDM screening. However, women with secondary/high school education had a 10% higher probability of being screened (95% CI: 4–16%) than those with only primary school or no school attendance. The mother’s clinical history also affected the probability of GDM screening. Compared with women who had never had a spontaneous abortion, those with one, two or three or more previous abortions were screened more frequently: 9% (95% CI: 7–11%), 18% (95% CI: 12–23%) and 41% (95% CI: 30–53%), respectively (P value for trend: < 0.0001). In addition, women who had used assisted reproductive technology had an 11% (95% CI: 4–18%) greater probability of being screened than women who did not. Similar findings were observed for women with multiple births, previous induced abortions, previous stillbirths and hypertension, whose probability of GDM screening increased by 65% (95% CI: 56–75%), 16% (95% CI: 12–20%), 42% (95% CI: 29–55%) and 51% (95% CI: 40–64%), respectively, compared with women with single births, no previous induced abortions, no previous stillbirths and no hypertension.
4. Discussion Very few epidemiological studies have documented adherence to GDM screening policies during pregnancy, although GDM is associated with an increased risk of adverse pregnancy outcomes, maternal T2D later in life and obesity, and later T2D in the child [3,7–9]. The present study is the first in Italy, and one of the few worldwide, to provide population-based estimates of GDM screening proportions and population-based predictors of GDM screening. From the HCU databases for Lombardy, which included more than 360,000 resident women who delivered during 2007–2010, it was estimated that only 30% of pregnancies were screened for GDM. The proportion of pregnancies screened increased slightly over the years covered by the study and with maternal age, and varied widely among HMOs. Socioeconomic indicators, obstetric history and maternal hypertension were independent predictors of GDM screening. The observed proportion of pregnancies screened in Lombardy was lower than reported in other countries; for example, proportions were 89% in Israel, where a universal screening policy is implemented, and 68% in the US, involving women who are beneficiaries of healthcare insurance aged > 25 years [23,36]. In Lombardy, the percentage of women screened for GDM among those aged ≥ 35 was 32% compared with 71% in the US study [36]. Also, of women included in the study screened for GDM, 33% were aged ≥ 35 years, 28% were immigrants and 2% used assisted reproductive technology whereas, in Israel, these percentages were 21%, 17% and 6%, respectively [23].
Recommended screening (24-28 weeks): 53.9%
Proportion of pregnancies screened for GDM (%)
16
14 12 10 Early screening (≤18 weeks): 5.7%
8
Late screening (≥32 weeks): 4.3%
6
4 2 0
0
2
4
6
8
10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 Gestational week
Fig. 2. Distribution of the first gestational diabetes mellitus (GDM) screening test by gestational week.
Please cite this article in press as: Nicotra F, et al. Screening for gestational diabetes in the Lombardy region: A population-based study. Diabetes Metab (2014), http://dx.doi.org/10.1016/j.diabet.2014.11.008
+Model DIABET-661; No. of Pages 7 6
ARTICLE IN PRESS F. Nicotra et al. / Diabetes & Metabolism xxx (2014) xxx–xxx
Although GDM screening rates increased from 2007 to 2010, they were still low, considering that the national guidelines at the time recommended GDM screening for women with risk factors, including maternal age > 25 years [31]. If, in the cross-sectional sample, only maternal age > 25 years was considered a risk factor for GDM, the proportion of pregnancies for which GDM screening was recommended would have been approximately 90% – three times what was actually observed. Many factors may have contributed to the low proportion of pregnancies screened for GDM in Lombardy. First, healthcare professionals may have been poorly compliant with guidelines, as suggested by the small increase in the proportion of pregnancies screened for GDM with increasing maternal age. GDM screening was only 32% among women aged ≥ 35 years, for whom GDM incidence is estimated to be > 15% [37]. This is especially relevant, as women in Western countries are now postponing their first pregnancy to their third decade of life [38]. Even among women whose pregnancy is likely to receive increased medical attention (users of assisted reproductive technology, twin pregnancies, women with several previous spontaneous abortions, still births and hypertension), the proportion of GDM screening was only slightly higher than among women with spontaneous singleton pregnancies, no obstetric history and no hypertension, thus, far from reflecting any appreciation of the increased GDM risk in these subgroups. This may be explained, at least in part, by the fact that these conditions are usually not listed as risk factors for GDM in screening guidelines and also because delivery by caesarean section is often planned for these women, thereby making macrosomia a lesser concern. Among physicians, conflicting and frequently revised guidelines, a perceived low value of treating hyperglycaemia during pregnancy despite the evidence accumulated over the past few years [9–12], concerns over excessive medicalization of pregnancy and economic pressure to contain healthcare expenditures may all have contributed to low adherence to GDM screening policies. On the other hand, pregnant women may be poorly compliant with GDM screening because of limited awareness of GDM-associated risks, and fear of the discomfort of being diagnosed with GDM and having to take insulin. Nevertheless, compliance during pregnancy in women is usually high, as reported for screening of transmittable diseases, and cessation of smoking and alcohol consumption [39,40] during pregnancy. Furthermore, in our study, 79% of the women screened for GDM had their test within 2 weeks of the recommended timing (24 to 28 weeks of gestation), supporting women’s compliance with GDM screening. Also, it cannot be excluded that some women may have elected for GDM screening outside of the NHS, in which case the HCU database would not have recorded the event. However, 95% of the women in the cross-sectional sample had blood glucose tests performed during pregnancy through the NHS, making it unlikely that GDM screening outside the NHS contributed significantly to the low screening proportion for GDM in Lombardy during 2007–2010. The wide variability in the proportion of pregnancies screened in different HMOs is more likely to reflect differences in attitude towards GDM screening by prescribers rather than due to any heterogeneity in the underlying GDM risk of the
residents. This is supported by the fact that the distribution of covariates included in our multivariate model was similar among the HMOs in Lombardy (data not shown). It is acknowledged that this study had limitations. First, HCU databases do not include information such as anthropometrics, family history of diabetes and obstetric history, which would have allowed a precise identification of the population of pregnant women at risk of GDM who should have undergone screening. Also, adding risk factors other than maternal age ≥ 25 years would have had only limited impact, given the small number of women aged < 25 years (10% of our study sample) in whom GDM screening should have been prescribed. Second, although Lombardy accounts for 16% of the entire Italian population, the study results may not be applicable to other Italian regions, as all regions are autonomous in implementing national clinical practice guidelines for their residents [24]. Third, the use of such a large sample of deliveries may have highlighted some associations of limited or no clinical relevance. The potential implications of the low proportion of women screened for GDM documented in Lombardy are twofold. First, women with GDM not diagnosed because of lack of screening are more likely to have a macrosomic child, thereby contributing to the proportion of caesarean sections in Lombardy which, although relatively low compared with other Italian regions, is still among the highest in Europe [41]. Second, women not identified as having GDM during pregnancy because of lack of screening miss out on interventional programmes for primary prevention of T2D after delivery [42]. In conclusion, the present findings support the need for programmes to improve training of healthcare professionals, to raise women’s awareness of GDM and to eliminate barriers to GDM screening. In addition, the present study also demonstrates that HCU databases are an invaluable, albeit as yet mostly untapped, source of information on clinical practices in Italy. Also, future analyses should provide insights on the impact of using a single 75-g OGTT to screen for GDM. Disclosure of interest The authors declare that they have no conflicts of interest concerning this article. References [1] Metzger BE, Gabbe SG, Persson B, Buchanan TA, Catalano PA, Damm P, et al. International association of diabetes and pregnancy study groups recommendations on the diagnosis and classification of hyperglycaemia in pregnancy. Diabetes Care 2010;33:676–82. [2] Mulla WR, Henry TQ, Homko CJ. Gestational diabetes screening after HAPO: has anything changed? Curr Diab Rep 2010;10:224–8. [3] Metzger BE, Lowe LP, Dyer AR, Trimble ER, Chaovarindr U, Coustan DR, et al. Hyperglycaemia and adverse pregnancy outcomes. N Engl J Med 2008;358:1991–2002. [4] Jovanovic L, Pettitt DJ. Gestational diabetes mellitus. JAMA 2001;286:2516–8. [5] Bellamy L, Casas JP, Hingorani AD, Williams D. Type 2 diabetes mellitus after gestational diabetes: a systematic review and meta-analysis. Lancet 2009;373:1773–9.
Please cite this article in press as: Nicotra F, et al. Screening for gestational diabetes in the Lombardy region: A population-based study. Diabetes Metab (2014), http://dx.doi.org/10.1016/j.diabet.2014.11.008
+Model DIABET-661; No. of Pages 7
ARTICLE IN PRESS F. Nicotra et al. / Diabetes & Metabolism xxx (2014) xxx–xxx
[6] Huidobro A, Prentice A, Fulford T, Parodi C, Rozowski J. Gestational diabetes, comparison of women diagnosed in second and third trimester of pregnancy with non GDM women: analysis of a cohort study. Rev Med Chil 2010;138:316–21. [7] Burguet A. Long-term outcome in children of mothers with gestational diabetes. Diabetes Metab 2010;36:682–94. [8] Wu CS, Nohr EA, Bech BH, Vestergaard M, Olsen J. Long-term health outcomes in children born to mothers with diabetes: a population-based cohort study. PLoS One 2012;7:e36727. [9] Crowther CA, Hiller JE, Moss JR, McPhee AJ, Jeffries WS, Robinson JS, et al. Effect of treatment of gestational diabetes mellitus on pregnancy outcomes. N Engl J Med 2005;352:2477–86. [10] van Leeuwen M, Louwerse MD, Opmeer BC, Limpens J, Serlie MJ, Reitsma JB, et al. Glucose challenge test for detecting gestational diabetes mellitus: a systematic review. BJOG 2012;119:393–401. [11] Landon MB, Spong CY, Thom E, Carpenter MW, Ramin SM, Casey B, et al. A multicenter, randomized trial of treatment for mild gestational diabetes. N Engl J Med 2009;361:1339–48. [12] Langer O, Yogev Y, Most O, Xenakis EM. Gestational diabetes: the consequences of not treating. Am J Obstet Gynecol 2005;192:989–97. [13] Buckley BS, Harreiter J, Damm P, Corcoy R, Chico A, Simmons D, et al. Gestational diabetes mellitus in Europe: prevalence, current screening practice and barriers to screening. A review. Diabet Med 2012;29:844–54. [14] Sacks DA, Hadden DR, Maresh M, Deerochanawong C, Dyer AR, Metzger BE, et al. Frequency of gestational diabetes mellitus at collaborating centers based on IADPSG consensus panel-recommended criteria: the Hyperglycaemia and Adverse Pregnancy Outcome (HAPO) Study. Diabetes Care 2012;35:526–8. [15] Reece EA, Leguizamón G, Wiznitzer A. Gestational diabetes: the need for a common ground. Lancet 2009;373:1789–97. [16] Tieu J, Middleton P, McPhee AJ, Crowther CA. Screening and subsequent management for gestational diabetes for improving maternal and infant health. Cochrane Database Syst Rev 2010;7:CD007222. [17] Holt RI, Coleman MA, McCance DR. The implications of the new International Association of Diabetes and Pregnancy StudyGroups (IADPSG) diagnostic criteria for gestational diabetes. Diabet Med 2011;28:382–5. [18] Hoffman L, Nolan C, Wilson JD, Oats JJ, Simmons D. Gestational diabetes mellitus: management guidelines. The Australasian Diabetes in Pregnancy Society. Med J Aust 1998;169:93–7. [19] American Diabetes Association Workshop. Conference on gestational diabetes: summary and recommendations. Diabetes Care 1980;3:499–501. [20] National Collaborating Centre for Women’s and Children’s Health (UK). Diabetes in Pregnancy: Management of diabetes and its complications from preconception to the postnatal period. London: RCOG Press; 2008. [21] Scottish Intercollegiate Guidelines Network (SIGN). Management of diabetes. A national clinical guideline. Edinburgh: SIGN; 2010. [22] Collège national des gynécologues et obstétriciens franc¸ais, Société francophone du diabète. Recommandations pour la pratique clinique. Le diabète gestationnel. J Gynecol Obstet Biol Reprod 2010;39:S1–342. [23] Sella T, Shalev V, Elchalal U, Chovel-Sella A, Chodick G. Screening for gestational diabetes in the 21st century: a population-based cohort study in Israel. J Matern Fetal Neonatal Med 2013;26:412–6. [24] Lo Scalzo A, Donatini A, Orzella L, Cicchetti A, Profili S, Maresso A. Italy. Health system review. Health Syst Transit 2009;11:1–243.
7
[25] Di Bari M, Balzi D, Roberts AT, Barchielli A, Fumagalli S, Ungar A, et al. Prognostic stratification of older persons based on simple administrative data: development and validation of the “Silver Code” to be used in emergency department triage. J Gerontol A Biol Sci Med Sci 2010;65:159–64. [26] Corrao G, Conti V, Merlino L, Catapano AL, Mancia G. Results of a retrospective database analysis of adherence to statin therapy and risk of nonfatal ischemic heart disease in daily clinical practice in Italy. Clin Ther 2010;32:300–10. [27] Corrao G, Zambon A, Bertù L, Botteri E, Leoni O, Contiero P. Lipid lowering drugs prescription and the risk of peripheral neuropathy: an exploratory case-control study using automated databases. J Epidemiol Community Health 2004;58:1047–51. [28] Gini R, Francesconi P, Mazzaglia G, Cricelli I, Pasqua A, Gallina P, et al. Chronic disease prevalence from Italian administrative databases in the VALORE project: a validation through comparison of population estimates with general practice databases and national survey. BMC Public Health 2013;13:15. [29] Degli Esposti L, Saragoni S, Batacchi P, Benemei S, Geppetti P, Sturani A, et al. Adherence to statin treatment and health outcomes in an Italian cohort of newly treated patients: results from an administrative database analysis. Clin Ther 2012;34:190–9. [30] Art. 232 of the Italian Civil Code. [31] Metzger BE, Coustan DR. Summary and recommendations of the Fourth International Workshop – Conference on Gestational Diabetes Mellitus. The Organizing Committee. Diabetes Care 1998;21(Suppl 2):B161–7. [32] Hanley JA, Negassa A, Edwardes MD, Forrester JE. Statistical analysis of correlated data using generalized estimating equations: an orientation. Am J Epidemiol 2003;157:364–75. [33] Tirosh D, Benshalom-Tirosh N, Novack L, Press F, Beer-Weisel R, Wiznitzer A, et al. Hypothyroidism and diabetes mellitus: a risky dual gestational endocrinopathy. Peer J 2013;1:e52. [34] Peeters GM, Burton NW, Brown WJ. Associations between sitting time and a range of symptoms in mid-age women. Prev Med 2013;56:135–41. [35] http://www.asr-lombardia.it/ASR/sanita-e-salute/condizioni-di-salute-/ lombardia-e- province/tavole/100405/ [36] Blatt AJ, Nakamoto JM, Kaufman HW. Gaps in diabetes screening during pregnancy and postpartum. Obstet Gynecol 2011;117:61–8. [37] Mayorga ME, Reifsnider OS, Neyens DM, Gebregziabher MG, Hunt KJ. Simulated estimates of pre-pregnancy and gestational diabetes mellitus in the US: 1980 to 2008. PLoS One 2013;8:e73437. [38] Matthews TJ, Hamilton BE. Delayed childbearing: more women are having their first child later in life. NCHS Data Brief 2009;21:1–8. [39] Frischknecht F, Sell W, Trummer I, Brühwiler H. Serological testing for infectious diseases in pregnant women: are the guidelines followed? Swiss Med Wkly 2011;140:w13138. [40] Yamada H, Tairaku S, Morioka I, Ebina Y, Sonoyama A, Tanimura K, et al. A nationwide survey of maternal screening for mother-to-child infections in Japan. Congenit Anom (Kyoto) 2014;54:100–3. [41] Narayanan RB, Girdhar A, Girdhar BK, Malaviya GN. Immunohistological analysis of nerve granulomas in neuritic leprosy. Int Arch Allergy Appl Immunol 1990;92:50–5. [42] Bentley-Lewis R, Levkoff S, Stuebe A, Seely EW. Gestational diabetes mellitus: postpartum opportunities for the diagnosis and prevention of type 2 diabetes mellitus. Nat Clin Pract Endocrinol Metab 2008;4:552–8.
Please cite this article in press as: Nicotra F, et al. Screening for gestational diabetes in the Lombardy region: A population-based study. Diabetes Metab (2014), http://dx.doi.org/10.1016/j.diabet.2014.11.008
