- Email: [email protected]
Fasting plasma glucose to avoid a full OGTT in the diagnosis of gestational diabetes
diabetes research and clinical practice 105 (2014) 322–326
Contents available at ScienceDirect
Diabetes Research and Clinical Practice jou rnal hom ep ag e: w ww.e l s e v i er . c om/ loca te / d i ab r es
Fasting plasma glucose to avoid a full OGTT in the diagnosis of gestational diabetes J. Trujillo a,*, A. Vigo a, A. Reichelt b, B.B. Duncan a, M.I. Schmidt a a
Post Graduate Studies Program in Epidemiology, School of Medicine, Rio Grande do Sul Federal University, Rua Ramiro Barcelos 2600, sala 414, 90035-003, Porto Alegre, RS, Brazil b Hospital de Clı´nicas de Porto Alegre, Porto Alegre, Brazil
article info
abstract
Article history:
Aims: To evaluate the performance of fasting plasma glucose (FPG) in determining the need
Received 24 March 2014
for a full oral glucose tolerance test (OGTT) to diagnose gestational diabetes (GDM) by the
Received in revised form
International Association of Diabetes and Pregnancy Study Groups (IADPSG) criteria.
31 May 2014
Methods: A multicenter cohort study of 4926 pregnant women 20 years or older consecu-
Accepted 2 June 2014
tively enrolled in prenatal care clinics of the Brazilian National Health Service from 1991 to
Available online 28 June 2014
1995. All women underwent a single 2 h 75 g OGTT by weeks 24–28 of pregnancy and were
Keywords:
Results: A FPG cut-off value of 80 mg/dl indicated that only 38.7% of all women needed to
followed to detect adverse pregnancy outcomes. Fasting plasma glucose
undergo a complete OGTT, while detecting 96.9% of all GDM cases. When the 85 mg/dl cut-
Diagnosis
off was used, the corresponding percentages were 18.7% and 92.5%, respectively. The
Gestational diabetes
fraction of women labeled with GDM who had adverse pregnancy outcomes was nearly identical when using FPG strategies and universal full testing. Conclusions: Using a FPG cut-off to diagnose GDM and to determine the need for post-load OGTT measurements is a valid strategy to diagnose GDM by IADPSG criteria. This approach may improve feasibility of applying IADPSG diagnostic criteria by reducing costs and increasing convenience. # 2014 Elsevier Ireland Ltd. All rights reserved.
1.
Introduction
Gestational diabetes mellitus (GDM) has become an important public health problem owing to its growing prevalence and its association with adverse pregnancy outcomes and type 2 diabetes mellitus later in life. However, controversies on screening and diagnostic strategies still remain. In 2010, the International Association of Diabetes and Pregnancy Study Groups (IADPSG) recommended applying a diagnostic 2 h 75 g OGTT to all women reaching 24th to 28th gestational weeks and not diagnosed as having overt diabetes * Corresponding author. Tel.: +55 51 33085347. E-mail address: [email protected] (J. Trujillo). http://dx.doi.org/10.1016/j.diabres.2014.06.001 0168-8227/# 2014 Elsevier Ireland Ltd. All rights reserved.
or GDM earlier in the pregnancy [1]. This criterion has been endorsed by the American Diabetes Association (ADA) and by the World Health Organization (WHO) for the diagnosis of gestational diabetes [2,3]. However, its poor cost-effectiveness has limited its implementation [4], so that new strategies are being considered to improve this setting. In Brazil, a fasting plasma glucose (FPG) test has been largely used as a part of a two-step screening approach, with cut-offs between 85 mg/dl (4.7 mmol/l) and 90 mg/dl (5 mmol/ l) to define a positive screening test, and thus the need for a diagnostic test (2 h 75 g OGTT) [5]. These cut-offs have been evaluated against the 1999 WHO criteria and the previous ADA
323
diabetes research and clinical practice 105 (2014) 322–326
criteria, [6,7] but to our knowledge, have been little evaluated against the IADPSG criteria [8,9]. Neither of those previous studies evaluated the ability of such a strategy to identify women who suffered adverse pregnancy outcomes during follow-up. The Brazilian Study of Gestational Diabetes (Estudo Brasileiro de Diabetes Gestacional, or EBDG) [10] an observational study which tested and followed a large number of pregnant women during a period in which diagnosis of gestational diabetes was not standardized and treatment for those with lesser than diabetes hyperglycemia was uncommon, offers a good opportunity for further evaluation. The purpose of this study was to evaluate a strategy based on a FPG cut-off to avoid the necessity of obtaining 1 h and 2 h OGTT measurements in the diagnosis of GDM based on the IADPSG criteria.
2.
Materials and methods
Analyses were based on the EBDG, a cohort of 5564 pregnant women consecutively enrolled at gestational weeks 20th– 28th, and receiving prenatal care in the National Health Service in six state capitals of Brazil from May 1991 to August 1995 [10]. Ethics committees of each center approved the study protocol, and all patients gave informed consent to participate. Eligibility criteria included having no history of diabetes outside pregnancy and being 20 years or older. At enrolment, all women answered a structured questionnaire and were scheduled for a single 2 h 75 g anhydrous glucose OGTT according to standardized WHO procedures (plasma glucose was measured in FPG, after load in 1 h and 2 h) [11]. The blood samples were collected in fluoride tubes and kept them at 48 C until centrifugation, up to 2 h later. Plasma glucose was measured by glucose oxidase method [12] in local laboratories previously certified by the Study’s quality control committee. External quality controls with three different glucose concentrations were used for certification and monitoring. A coefficient of variation 5% at any point was a reason to suspend glucose determinations. Body mass index (BMI) was calculated as weight (kg)/height (cm)2 based on anthropometric measures at enrolment obtained in duplicate following a standard protocol. GDM was defined according to the IADPSG diagnostic criteria, which require only one abnormal value out of the three cut-off points: fasting 92 mg/dl (5.1 mmol/l); 1 h 180 mg/dl (10.0 mmol/l); or 2 h 153 mg/dl (8.5 mmol/l) [1]. Women were followed until delivery, data on outcomes being collected from chart review using a structured protocol. The composite adverse pregnancy outcome consisted of at least one of the following three outcomes: large for gestational age, preeclampsia or perinatal death. LGA was defined as a birth weight at or above the gestational age-specific (by week) 90th percentile for the study sample, as previously described [10]. Preeclampsia (or eclampsia) was defined as either chronic or incident hypertension associated with either proteinuria or convulsions after 20 weeks of gestation, in accordance with the recommendations of the National High Blood Pressure Education Program Working Group [13]. Perinatal deaths were ascertained as a fetal loss weighting more than 1 kg or with
estimated gestational age >28th weeks; or an early neonatal death (up to 7 days). Data were available for 4926 (90%) women, after excluding 566 women who did not perform the OGTT or had incomplete values for the OGTT, 21 women reaching criteria for diabetes (fasting plasma glucose level 126 mg/dl and/or 2 h plasma glucose level 200 mg/dl), 2 women who received insulin treatment, and 49 women with multiple pregnancies. Missing values for the different outcomes led to slight variation in the total sample size of specific analyses. Analyses of adverse pregnancy outcomes, due to losses during follow-up, were performed on data for 4160 women. Data are described as absolute and relative frequencies (%) for categorical variables, and as mean and standard deviation (SD) for continuous ones. The performance of FPG in detecting GDM as diagnosed by the IADPSG criteria and in predicting the composite outcome of adverse pregnancy events was evaluated using the area under the receiver operating characteristic (ROC) curve (AUC). We calculated sensitivity, specificity, and positive and negative predictive values for FPG threshold values considered relevant as well as the percentage of women with a positive result for each threshold [14]. We employed the FPG which was part of the previously described OGTT in all analyses. The software package Statistical Analysis System (version 9.2; SAS Institute, Cary, NC) was used for all analyses.
3.
Results
Main characteristics of pregnant women enrolled in the EBDG study are presented in Table 1. Means (SD) of plasma glucose for samples obtained during the OGTT were 81.5 mg/dl (4.5 mmol/l) for fasting, 121.2 mg/dl (6.7 mmol/l) for 1 h, and 103.2 mg/dl (5.7 mmol/l) for 2 h values. Prevalence of GDM was
Table 1 – Characteristics of the 4926 women participating in the Brazilian Gestational Diabetes Study. Clinical characteristics Maternal age (years) BMI at the enrolment (kg/m2) Gestational age at delivery (weeks) Ethnicity (%) White Black Mixed Other Education (%) <8 years 8–11 years >11 years Parity (%) 0 1 2 3 Family history of diabetes (%) Cesarean delivery (%)
Mean (SD) or N (%) 27.8 (5.4) 26.0 (4.0) 38.4 (2.4) 2206 673 2026 20
(44.8) (13.7) (41.1) (0.4)
2159 (43.9) 2275 (46.3) 482 (9.8) 1350 1472 829 741 685 1624
(30.7) (33.5) (19.0) (16.8) (14.8) (37.8)
Number of women for each characteristic varies slightly due to missing values.
324
diabetes research and clinical practice 105 (2014) 322–326
various FPG cut-off points are presented in Table 2. Values between 80 and 92 mg/dl classified 54.3–15.6% of the women as having a positive test and were thus considered as potentially relevant cut-off points. Sensitivity decreased and specificity increased as FPG cut-off points were raised, being 96.9% and 55.0% for the cut-off point of 80 mg/dl and 86.8% and 100% for the cut-off point of 92 mg/dl. The diagnostic properties of FPG for the detection of the composite adverse pregnancy outcome were less favorable. Sensitivities ranged from 60.2% to 19.7%, consistent with the low AUC of FPG in the prediction of these outcomes. Overall, the most relevant cut-offs appeared to be 80 and 85 mg/dl. Table 3 evaluates different cut-offs for detecting GDM using this approach of avoiding 1 h and 2 h measures based on the FPG value when compared to universal full OGTT testing as recommended by the IADPSG. When testing initially with just a FPG, values 92 mg/dl label women as GDM. Cut-off values within the range of 80 and 85 mg/dl would classify 38.7%– 18.7% of all women as needing post-load OGTT measurements. The percent of all IADPSG GDM cases detected with this approach was 96.9% when using the cut-off point of 80 mg/dl, and 92.5% when using the cut-off point of 85 mg/dl. The prevalence of detected GDM by the FPG strategy based on these two cut-offs was 17.4% and 16.6%, respectively, compared to 18% with the universal full OGTT. The percent of LGA births and preeclampsia cases predicted with this FPG strategy (sensitivity) was very similar to that predicted by the universal OGTT. The 80 mg/dl cut-off identified the same number of cases who were diagnosed as having preeclampsia during the pregnancy as universal full OGTT testing, and only 0.4% fewer cases of LGA births.
Table 2 – Performance of differing fasting plasma glucose (FPG) values for the detection of gestational diabetes (GDM) based on IADPSGa criteria and a composite measure of adverse pregnancy outcomes. Performance measures
Cut-off points of FPG (mg/dl)
Positive test (%) Se (%) To detect GDM To predict adverse outcomes Sp (%) To detect GDM To predict adverse outcomes PPV (%) To detect GDM To predict adverse outcomes NPV (%) To detect GDM To predict adverse outcomes
80
85
90
92
54.3
34.3
19.8
15.6
96.9 60.2
92.5 41.5
88.3 26.1
86.8 19.7
55.0 46.3
78.4 66.5
95.1 81.0
100.0 85.0
32.0 19.0
48.3 20.6
79.8 22.4
100.0 21.5
98.8 84.7
97.9 84.4
97.4 84.0
97.2 83.5
Se, sensitivity; Sp, specificity; PPV, positive predictive value; NPV, negative predictive value. a International Association of Diabetes and Pregnancy Study Groups.
18.0% (95% CI 16.9–19.0) when defined by the full IADPSG criteria. There were 767 (15.6%) women with FPG 92 mg/dl. We ascertained 721 (17.3%) composite adverse pregnancy outcomes, 481 (11.8%) LGA newborns, 113 (2.7%) perinatal deaths and 149 (3.6%) cases of preeclampsia. The capacity of FPG values to detect GDM defined by the IADPSG diagnostic criteria evaluated by the AUC was 0.960 (95% CI 0.952–0.969). However, the AUC for the capacity of FPG values to detect the composite adverse pregnancy outcome was only 0.549 (95% CI 0.526–0.573). For comparison, the AUC for the fasting, 1 h and 2 h glycemic levels together in the detection of this composite outcome was 0.582 (95% CI 0.559–0.604), and similarly for only 1 h and 2 h levels together (AUC = 0.581, 95% CI 0.558–0.604). Percent of women with a positive test, sensitivity, specificity, and positive and negative predictive values for
4.
Discussion
Our findings indicate that FPG, when used not just as part of the IADPSG diagnostic algorithm, but also as a means of avoiding 1 h and 2 h post-load measures, was an excellent strategy to classify pregnant women into three groups – those already meeting the IADPSG criteria; those not yet meeting them but at a sufficiently high risk so as to merit post-load OGTT measures; and those at sufficiently low risk not to require any further testing. This FPG strategy correctly
Table 3 – Outcome of strategies using FPG to direct further testing in the diagnosis of gestational diabetes (GDM) based on IADPSGa criteria. Strategy
Initial FPG results Cut-off (mg/dl)
FPG, then 1 h and 2 h Universal full OGTT
Screen +
GDM
OGTT Need OGTT
Pregnancy outcomes
GDM
LGA
Preeclampsia
N
(%)
N
(%)
N
(%)
N
(%)
N
(%)
N
(%)
N
(%)
80
2675
54.3
767
15.6
1908
38.7
89
4.7
856
96.9
112
23.3
31
20.8
85
1691
34.3
767
15.6
924 4926
18.7 100.0
50 883
5.4 18.0
817 883
92.5 100.0
105 114
21.8 23.7
30 31
20.1 20.8
LGA, large for gestational age; OGTT, oral glucose tolerance test. International Association of Diabetes and Pregnancy Study Groups.
a
Total GDM
diabetes research and clinical practice 105 (2014) 322–326
identified as many as 97% of all GDM cases while avoiding a full OGTT in well over half of all women. Further, the probability of not having GDM given a FPG value below the cut-offs of 80 mg/dl and 85 mg/dl was very high, around 98%. Additionally, and importantly, the fraction of women who would go on to have an adverse pregnancy outcome predicted by this strategy was quite similar to that obtained with universal OGTT testing. This latter finding, to our knowledge, has not been previously reported. Over the years, FPG has been widely applied as a screening test for GDM, given its advantages of being less expensive, more reproducible, well accepted and easier to administer universally. Furthermore, properties of FPG here described are very similar to those described for the traditionally adopted 50 g OGTT in recent reviews [15,16]. Based on the previous ADA GDM criteria, FPG values between 85 and 89 mg/dl (4.7–4.9 mmol/l) presented sensitivity of about 80% and specificity of about 70% [7,17]. Based on the 1999 WHO GDM criteria, sensitivity and specificity were 69% and 68%, respectively, for a FPG cut-off point of 85 mg/dl [6]. Since the evaluations utilized FPG obtained from the same test which diagnosed GDM, these figures overestimate the diagnostic properties if a two-step process is to be used, particularly those based on the previous ADA criteria, which employed a lower value of FPG in GDM diagnosis than the 1999 WHO criteria. Of note, a FPG performed in early pregnancy was not shown to be a good predictor of GDM when detected later in pregnancy [18]. Few studies have been performed investigating FPG as a means of orienting further testing when GDM is based on IADPSG criteria. Agarwal et al. [8] found an AUC of 0.907 (95% CI 0.899–0.914) for FPG as a test to direct further testing for the detection of GDM. The authors recommended a cut-off point of 4.4 mmol/l (80 mg/dl), which provided a sensitivity of 95.4% and specificity of 32.0%. Another recent study also suggested the cut-off value of 4.4 mmol/l, finding a sensitivity of 87.8% and specificity of 45.8%, although the AUC found for FPG was lower, 0.836 (95% CI 0.829–0.843) [9]. However, neither of these studies evaluated the performance of FPG, when used for this purpose, in predicting pregnancy outcomes, as we have done. Even though maternal and fetal outcomes could have been improved by treatment, it should be emphasized that our study occurred in a time when lifestyle interventions were not routinely recommended. Also, diagnostic cut-off values for GDM were higher and not consensual and we excluded from the analyses the few women who received insulin treatment. A feasible FPG strategy would be as follows. A FPG 92 mg/ dl would label women as GDM. Cut-off points of FPG between 80 and 85 mg/dl could then be chosen to indicate the need for 1 h and 2 h post load testing for those with FPG <92 mg/dl, according to resources available in specific settings. Using a cut-off of 80 mg/dl, more than one-third of all women would be classified as requiring the OGTT, in a process which detects 96.9% of IADPSG defined GDM cases. Using a cut-off of 85 mg/ dl, less than 20% of all women would require OGTT testing, in a process which detects 92.5% of these GDM cases. The choice of the threshold is dependent on the local priorities and the resources available for GDM screening. When to perform post-load testing in those for which the FPG does not complete diagnostic classification merits discussion. Ideally, it should be performed during the same
325
clinic visit. However, as has been noted [19], this is not always possible. If the OGTT were performed on a subsequent visit, the diagnostic criteria shown here will somewhat overestimate reality, given that plasma glucose level varies from day to day. We know of no evaluation of this short term variability in pregnancy. However, it is not high in the general population [20], and thus should not alter the diagnostic properties of the criteria greatly. The extrapolation of these results to other populations merits discussion, given the wide variability in the percentage of women diagnosed with GDM exclusively by FPG [21–23]. Although reasons for this variability are not clear, ethnicity and varying rates of obesity may play a role [24]. Incomplete fasting could be another reason. In this regard, our results are more directly applicable to populations with a high fraction of GDM women being diagnosed with FPG. However, studies performed in settings with a lower fraction of GDM diagnosed by FPG [8,9] also show excellent results for this strategy. When incomplete fasting is likely, repeat testing of FPG for women with an initial FPG 92 mg/dl is indicated. A possible additional strategy, which would not require rapid laboratory turnaround, is to test all women initially with a FPG and, on a second visit, to repeat the FPG for confirmation when the initial value was 92 mg/dl and to perform an OGTT when the initial value was less than 92 mg/dl but above the cut-off point of 80 (or 85) mg/dl. An additional potential limitation of our study is the possibility of selection bias due to missing information or incomplete follow-up. However, the OGTT was performed in over 90% of the women enrolled; outcomes were ascertained for at least 74% of all enrolled women and for 80% of those who underwent a 2 h 75 g OGTT. In conclusion, a strategy in which FPG is used to detect cases of GDM and to determine the extent of further testing appears to be a valid approach to identify women with GDM by IADPSG criteria. This strategy would improve the feasibility of applying IADPSG diagnostic criteria by reducing costs and improving the convenience of GDM detection in pregnancy.
Conflict of interest The authors report no conflict of interest.
Acknowledgments We are grateful to the women who participated at this study and the members of the Brazilian Gestational Diabetes Study group. The study was supported in part by the Brazilian Ministry of Health, the Pan-American Health Organization (PAHO), Awards for Groups of Excellence (PRONEX No. 661041/1998-0) of the Brazilian National Council for Technologic and Scientific Development (CNPq), the Foundation for the Support of Research of the State of Rio Grande do Sul (FAPERGS), Bristol-Meyers Squibb Foundation, Becton Dickinson, Bayer of Brasil and Biobra´s. JT received a doctoral fellowship from Capes/CNPq-IEL National- Brazil (No. 013/2008).
326
diabetes research and clinical practice 105 (2014) 322–326
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 hyperglycemia in pregnancy. Diabetes Care 2010;33:676–82. [2] American Diabetes Association. American Diabetes Association Expert Committee. Diagnosis and classification of diabetes mellitus. Diabetes Care 2011;34(Suppl 1):S62–9. [3] WHO Guideline Development Group. WHO – Diagnostic criteria and classification of hyperglycaemia first detected in pregnancy. WHO; 2013. [4] Werner EF, Pettker CM, Zuckerwise L, Reel M, Funai EF, Henderson J, et al. Screening for gestational diabetes mellitus: are the criteria proposed by the International Association of the Diabetes and Pregnancy Study Groups cost-effective? Diabetes Care 2012;35:529–35. [5] Reichelt AJ, Oppermann MLR, Schmidt MI. Guidelines of the 2nd Meeting of The Diabetes and Pregnancy Task Force. Arq Bras Endocrinol Metabol 2002;46:574–81. [6] Reichelt AJ, Spichler ER, Branchtein L, Nucci LB, Franco LJ, Schmidt MI. Fasting plasma glucose is a useful test for the detection of gestational diabetes. Brazilian Study of Gestational Diabetes (EBDG) Working Group. Diabetes Care 1998;21:1246–9. [7] Agarwal MM, Dhatt GS, Punnose J. Gestational diabetes: utility of fasting plasma glucose as a screening test depends on the diagnostic criteria. Diabet Med J Br Diabet Assoc 2006;23:1319–26. [8] Agarwal MM, Dhatt GS, Shah SM. Gestational diabetes mellitus: simplifying the International Association of Diabetes and Pregnancy diagnostic algorithm using fasting plasma glucose. Diabetes Care 2010;33:2018–20. [9] Zhu W, Fan L, Yang H, Kong L, Su S, Wang Z, et al. Fasting plasma glucose at 24–28 weeks to screen for gestational diabetes mellitus: new evidence from China. Diabetes Care 2013;36:2038–40. [10] Schmidt MI, Duncan BB, Reichelt AJ, Branchtein L, Matos MC, Costa e Forti A, et al. Gestational diabetes mellitus diagnosed with a 2-h 75-g oral glucose tolerance test and adverse pregnancy outcomes. Diabetes Care 2001;24:1151–5. [11] World Health Organization. WHO Consultation: Definition, Diagnosis and Classification of Diabetes Mellitus and its Complications: Report of a WHO Consultation. Part 1: Diagnosis and Classification of Diabetes Mellitus.(WHO/ NCD/NCS/99.2) Geneva: World Health Org.; 1999.
[12] Trinder P. Determination of blood glucose using an oxidase-peroxidase system with a non-carcinogenic chromogen. J Clin Pathol 1969;22:158–61. [13] Working Group Report. National High Blood Pressure Education Program Working Group Report on High Blood Pressure in Pregnancy. Am J Obstet Gynecol 1990;163:1691–712. [14] Fletcher RH, Fletcher SW. Clinical epidemiology: the essentials. 4th ed. Lippincott Williams & Wilkins; 2005. [15] Prutsky GJ, Domecq JP, Sundaresh V, Elraiyah T, Nabhan M, Prokop LJ, et al. Screening for gestational diabetes: a systematic review and meta-analysis. J Clin Endocrinol Metab 2013;98:4311–8. [16] Donovan L, Hartling L, Muise M, Guthrie A, Vandermeer B, Dryden DM. Screening tests for gestational diabetes: a systematic review for the U.S. Preventive Services Task Force. Ann Intern Med 2013;159:115–22. [17] Perucchini D, Fischer U, Spinas GA, Huch R, Huch A, Lehmann R. Using fasting plasma glucose concentrations to screen for gestational diabetes mellitus: prospective population based study. Br Med J 1999;319:812–5. [18] Sacks DA, Chen W, Wolde-Tsadik G, Buchanan TA. Fasting plasma glucose test at the first prenatal visit as a screen for gestational diabetes. Obstet Gynecol 2003;101:1197–203. [19] Agarwal MM, Dhatt GS. Comment on zhu et al. Fasting plasma glucose at 24–28 weeks to screen for gestational diabetes mellitus: new evidence from China. Diabetes Care 2013;36:2038–40 [Diabetes Care 2014;37:e33]. [20] Eckfeldt JH, Chambless LE, Shen YL. Short-term, withinperson variability in clinical chemistry test results. Experience from the Atherosclerosis Risk in Communities Study. Arch Pathol Lab Med 1994;118:496–500. [21] 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 Hyperglycemia and Adverse Pregnancy Outcome (HAPO) Study. Diabetes Care 2012;35:526–8. [22] Black MH, Sacks DA, Xiang AH, Lawrence JM. Clinical outcomes of pregnancies complicated by mild gestational diabetes mellitus differ by combinations of abnormal oral glucose tolerance test values. Diabetes Care 2010;33:2524–30. [23] Moses RG, Morris GJ, Petocz P, San Gil F, Garg D. The impact of potential new diagnostic criteria on the prevalence of gestational diabetes mellitus in Australia. Med J Aust 2011;194:338–40. [24] Wong VW. Gestational diabetes mellitus in five ethnic groups: a comparison of their clinical characteristics. Diabet Med J Br Diabet Assoc 2012;29:366–71.
Contents available at ScienceDirect
Diabetes Research and Clinical Practice jou rnal hom ep ag e: w ww.e l s e v i er . c om/ loca te / d i ab r es
Fasting plasma glucose to avoid a full OGTT in the diagnosis of gestational diabetes J. Trujillo a,*, A. Vigo a, A. Reichelt b, B.B. Duncan a, M.I. Schmidt a a
Post Graduate Studies Program in Epidemiology, School of Medicine, Rio Grande do Sul Federal University, Rua Ramiro Barcelos 2600, sala 414, 90035-003, Porto Alegre, RS, Brazil b Hospital de Clı´nicas de Porto Alegre, Porto Alegre, Brazil
article info
abstract
Article history:
Aims: To evaluate the performance of fasting plasma glucose (FPG) in determining the need
Received 24 March 2014
for a full oral glucose tolerance test (OGTT) to diagnose gestational diabetes (GDM) by the
Received in revised form
International Association of Diabetes and Pregnancy Study Groups (IADPSG) criteria.
31 May 2014
Methods: A multicenter cohort study of 4926 pregnant women 20 years or older consecu-
Accepted 2 June 2014
tively enrolled in prenatal care clinics of the Brazilian National Health Service from 1991 to
Available online 28 June 2014
1995. All women underwent a single 2 h 75 g OGTT by weeks 24–28 of pregnancy and were
Keywords:
Results: A FPG cut-off value of 80 mg/dl indicated that only 38.7% of all women needed to
followed to detect adverse pregnancy outcomes. Fasting plasma glucose
undergo a complete OGTT, while detecting 96.9% of all GDM cases. When the 85 mg/dl cut-
Diagnosis
off was used, the corresponding percentages were 18.7% and 92.5%, respectively. The
Gestational diabetes
fraction of women labeled with GDM who had adverse pregnancy outcomes was nearly identical when using FPG strategies and universal full testing. Conclusions: Using a FPG cut-off to diagnose GDM and to determine the need for post-load OGTT measurements is a valid strategy to diagnose GDM by IADPSG criteria. This approach may improve feasibility of applying IADPSG diagnostic criteria by reducing costs and increasing convenience. # 2014 Elsevier Ireland Ltd. All rights reserved.
1.
Introduction
Gestational diabetes mellitus (GDM) has become an important public health problem owing to its growing prevalence and its association with adverse pregnancy outcomes and type 2 diabetes mellitus later in life. However, controversies on screening and diagnostic strategies still remain. In 2010, the International Association of Diabetes and Pregnancy Study Groups (IADPSG) recommended applying a diagnostic 2 h 75 g OGTT to all women reaching 24th to 28th gestational weeks and not diagnosed as having overt diabetes * Corresponding author. Tel.: +55 51 33085347. E-mail address: [email protected] (J. Trujillo). http://dx.doi.org/10.1016/j.diabres.2014.06.001 0168-8227/# 2014 Elsevier Ireland Ltd. All rights reserved.
or GDM earlier in the pregnancy [1]. This criterion has been endorsed by the American Diabetes Association (ADA) and by the World Health Organization (WHO) for the diagnosis of gestational diabetes [2,3]. However, its poor cost-effectiveness has limited its implementation [4], so that new strategies are being considered to improve this setting. In Brazil, a fasting plasma glucose (FPG) test has been largely used as a part of a two-step screening approach, with cut-offs between 85 mg/dl (4.7 mmol/l) and 90 mg/dl (5 mmol/ l) to define a positive screening test, and thus the need for a diagnostic test (2 h 75 g OGTT) [5]. These cut-offs have been evaluated against the 1999 WHO criteria and the previous ADA
323
diabetes research and clinical practice 105 (2014) 322–326
criteria, [6,7] but to our knowledge, have been little evaluated against the IADPSG criteria [8,9]. Neither of those previous studies evaluated the ability of such a strategy to identify women who suffered adverse pregnancy outcomes during follow-up. The Brazilian Study of Gestational Diabetes (Estudo Brasileiro de Diabetes Gestacional, or EBDG) [10] an observational study which tested and followed a large number of pregnant women during a period in which diagnosis of gestational diabetes was not standardized and treatment for those with lesser than diabetes hyperglycemia was uncommon, offers a good opportunity for further evaluation. The purpose of this study was to evaluate a strategy based on a FPG cut-off to avoid the necessity of obtaining 1 h and 2 h OGTT measurements in the diagnosis of GDM based on the IADPSG criteria.
2.
Materials and methods
Analyses were based on the EBDG, a cohort of 5564 pregnant women consecutively enrolled at gestational weeks 20th– 28th, and receiving prenatal care in the National Health Service in six state capitals of Brazil from May 1991 to August 1995 [10]. Ethics committees of each center approved the study protocol, and all patients gave informed consent to participate. Eligibility criteria included having no history of diabetes outside pregnancy and being 20 years or older. At enrolment, all women answered a structured questionnaire and were scheduled for a single 2 h 75 g anhydrous glucose OGTT according to standardized WHO procedures (plasma glucose was measured in FPG, after load in 1 h and 2 h) [11]. The blood samples were collected in fluoride tubes and kept them at 48 C until centrifugation, up to 2 h later. Plasma glucose was measured by glucose oxidase method [12] in local laboratories previously certified by the Study’s quality control committee. External quality controls with three different glucose concentrations were used for certification and monitoring. A coefficient of variation 5% at any point was a reason to suspend glucose determinations. Body mass index (BMI) was calculated as weight (kg)/height (cm)2 based on anthropometric measures at enrolment obtained in duplicate following a standard protocol. GDM was defined according to the IADPSG diagnostic criteria, which require only one abnormal value out of the three cut-off points: fasting 92 mg/dl (5.1 mmol/l); 1 h 180 mg/dl (10.0 mmol/l); or 2 h 153 mg/dl (8.5 mmol/l) [1]. Women were followed until delivery, data on outcomes being collected from chart review using a structured protocol. The composite adverse pregnancy outcome consisted of at least one of the following three outcomes: large for gestational age, preeclampsia or perinatal death. LGA was defined as a birth weight at or above the gestational age-specific (by week) 90th percentile for the study sample, as previously described [10]. Preeclampsia (or eclampsia) was defined as either chronic or incident hypertension associated with either proteinuria or convulsions after 20 weeks of gestation, in accordance with the recommendations of the National High Blood Pressure Education Program Working Group [13]. Perinatal deaths were ascertained as a fetal loss weighting more than 1 kg or with
estimated gestational age >28th weeks; or an early neonatal death (up to 7 days). Data were available for 4926 (90%) women, after excluding 566 women who did not perform the OGTT or had incomplete values for the OGTT, 21 women reaching criteria for diabetes (fasting plasma glucose level 126 mg/dl and/or 2 h plasma glucose level 200 mg/dl), 2 women who received insulin treatment, and 49 women with multiple pregnancies. Missing values for the different outcomes led to slight variation in the total sample size of specific analyses. Analyses of adverse pregnancy outcomes, due to losses during follow-up, were performed on data for 4160 women. Data are described as absolute and relative frequencies (%) for categorical variables, and as mean and standard deviation (SD) for continuous ones. The performance of FPG in detecting GDM as diagnosed by the IADPSG criteria and in predicting the composite outcome of adverse pregnancy events was evaluated using the area under the receiver operating characteristic (ROC) curve (AUC). We calculated sensitivity, specificity, and positive and negative predictive values for FPG threshold values considered relevant as well as the percentage of women with a positive result for each threshold [14]. We employed the FPG which was part of the previously described OGTT in all analyses. The software package Statistical Analysis System (version 9.2; SAS Institute, Cary, NC) was used for all analyses.
3.
Results
Main characteristics of pregnant women enrolled in the EBDG study are presented in Table 1. Means (SD) of plasma glucose for samples obtained during the OGTT were 81.5 mg/dl (4.5 mmol/l) for fasting, 121.2 mg/dl (6.7 mmol/l) for 1 h, and 103.2 mg/dl (5.7 mmol/l) for 2 h values. Prevalence of GDM was
Table 1 – Characteristics of the 4926 women participating in the Brazilian Gestational Diabetes Study. Clinical characteristics Maternal age (years) BMI at the enrolment (kg/m2) Gestational age at delivery (weeks) Ethnicity (%) White Black Mixed Other Education (%) <8 years 8–11 years >11 years Parity (%) 0 1 2 3 Family history of diabetes (%) Cesarean delivery (%)
Mean (SD) or N (%) 27.8 (5.4) 26.0 (4.0) 38.4 (2.4) 2206 673 2026 20
(44.8) (13.7) (41.1) (0.4)
2159 (43.9) 2275 (46.3) 482 (9.8) 1350 1472 829 741 685 1624
(30.7) (33.5) (19.0) (16.8) (14.8) (37.8)
Number of women for each characteristic varies slightly due to missing values.
324
diabetes research and clinical practice 105 (2014) 322–326
various FPG cut-off points are presented in Table 2. Values between 80 and 92 mg/dl classified 54.3–15.6% of the women as having a positive test and were thus considered as potentially relevant cut-off points. Sensitivity decreased and specificity increased as FPG cut-off points were raised, being 96.9% and 55.0% for the cut-off point of 80 mg/dl and 86.8% and 100% for the cut-off point of 92 mg/dl. The diagnostic properties of FPG for the detection of the composite adverse pregnancy outcome were less favorable. Sensitivities ranged from 60.2% to 19.7%, consistent with the low AUC of FPG in the prediction of these outcomes. Overall, the most relevant cut-offs appeared to be 80 and 85 mg/dl. Table 3 evaluates different cut-offs for detecting GDM using this approach of avoiding 1 h and 2 h measures based on the FPG value when compared to universal full OGTT testing as recommended by the IADPSG. When testing initially with just a FPG, values 92 mg/dl label women as GDM. Cut-off values within the range of 80 and 85 mg/dl would classify 38.7%– 18.7% of all women as needing post-load OGTT measurements. The percent of all IADPSG GDM cases detected with this approach was 96.9% when using the cut-off point of 80 mg/dl, and 92.5% when using the cut-off point of 85 mg/dl. The prevalence of detected GDM by the FPG strategy based on these two cut-offs was 17.4% and 16.6%, respectively, compared to 18% with the universal full OGTT. The percent of LGA births and preeclampsia cases predicted with this FPG strategy (sensitivity) was very similar to that predicted by the universal OGTT. The 80 mg/dl cut-off identified the same number of cases who were diagnosed as having preeclampsia during the pregnancy as universal full OGTT testing, and only 0.4% fewer cases of LGA births.
Table 2 – Performance of differing fasting plasma glucose (FPG) values for the detection of gestational diabetes (GDM) based on IADPSGa criteria and a composite measure of adverse pregnancy outcomes. Performance measures
Cut-off points of FPG (mg/dl)
Positive test (%) Se (%) To detect GDM To predict adverse outcomes Sp (%) To detect GDM To predict adverse outcomes PPV (%) To detect GDM To predict adverse outcomes NPV (%) To detect GDM To predict adverse outcomes
80
85
90
92
54.3
34.3
19.8
15.6
96.9 60.2
92.5 41.5
88.3 26.1
86.8 19.7
55.0 46.3
78.4 66.5
95.1 81.0
100.0 85.0
32.0 19.0
48.3 20.6
79.8 22.4
100.0 21.5
98.8 84.7
97.9 84.4
97.4 84.0
97.2 83.5
Se, sensitivity; Sp, specificity; PPV, positive predictive value; NPV, negative predictive value. a International Association of Diabetes and Pregnancy Study Groups.
18.0% (95% CI 16.9–19.0) when defined by the full IADPSG criteria. There were 767 (15.6%) women with FPG 92 mg/dl. We ascertained 721 (17.3%) composite adverse pregnancy outcomes, 481 (11.8%) LGA newborns, 113 (2.7%) perinatal deaths and 149 (3.6%) cases of preeclampsia. The capacity of FPG values to detect GDM defined by the IADPSG diagnostic criteria evaluated by the AUC was 0.960 (95% CI 0.952–0.969). However, the AUC for the capacity of FPG values to detect the composite adverse pregnancy outcome was only 0.549 (95% CI 0.526–0.573). For comparison, the AUC for the fasting, 1 h and 2 h glycemic levels together in the detection of this composite outcome was 0.582 (95% CI 0.559–0.604), and similarly for only 1 h and 2 h levels together (AUC = 0.581, 95% CI 0.558–0.604). Percent of women with a positive test, sensitivity, specificity, and positive and negative predictive values for
4.
Discussion
Our findings indicate that FPG, when used not just as part of the IADPSG diagnostic algorithm, but also as a means of avoiding 1 h and 2 h post-load measures, was an excellent strategy to classify pregnant women into three groups – those already meeting the IADPSG criteria; those not yet meeting them but at a sufficiently high risk so as to merit post-load OGTT measures; and those at sufficiently low risk not to require any further testing. This FPG strategy correctly
Table 3 – Outcome of strategies using FPG to direct further testing in the diagnosis of gestational diabetes (GDM) based on IADPSGa criteria. Strategy
Initial FPG results Cut-off (mg/dl)
FPG, then 1 h and 2 h Universal full OGTT
Screen +
GDM
OGTT Need OGTT
Pregnancy outcomes
GDM
LGA
Preeclampsia
N
(%)
N
(%)
N
(%)
N
(%)
N
(%)
N
(%)
N
(%)
80
2675
54.3
767
15.6
1908
38.7
89
4.7
856
96.9
112
23.3
31
20.8
85
1691
34.3
767
15.6
924 4926
18.7 100.0
50 883
5.4 18.0
817 883
92.5 100.0
105 114
21.8 23.7
30 31
20.1 20.8
LGA, large for gestational age; OGTT, oral glucose tolerance test. International Association of Diabetes and Pregnancy Study Groups.
a
Total GDM
diabetes research and clinical practice 105 (2014) 322–326
identified as many as 97% of all GDM cases while avoiding a full OGTT in well over half of all women. Further, the probability of not having GDM given a FPG value below the cut-offs of 80 mg/dl and 85 mg/dl was very high, around 98%. Additionally, and importantly, the fraction of women who would go on to have an adverse pregnancy outcome predicted by this strategy was quite similar to that obtained with universal OGTT testing. This latter finding, to our knowledge, has not been previously reported. Over the years, FPG has been widely applied as a screening test for GDM, given its advantages of being less expensive, more reproducible, well accepted and easier to administer universally. Furthermore, properties of FPG here described are very similar to those described for the traditionally adopted 50 g OGTT in recent reviews [15,16]. Based on the previous ADA GDM criteria, FPG values between 85 and 89 mg/dl (4.7–4.9 mmol/l) presented sensitivity of about 80% and specificity of about 70% [7,17]. Based on the 1999 WHO GDM criteria, sensitivity and specificity were 69% and 68%, respectively, for a FPG cut-off point of 85 mg/dl [6]. Since the evaluations utilized FPG obtained from the same test which diagnosed GDM, these figures overestimate the diagnostic properties if a two-step process is to be used, particularly those based on the previous ADA criteria, which employed a lower value of FPG in GDM diagnosis than the 1999 WHO criteria. Of note, a FPG performed in early pregnancy was not shown to be a good predictor of GDM when detected later in pregnancy [18]. Few studies have been performed investigating FPG as a means of orienting further testing when GDM is based on IADPSG criteria. Agarwal et al. [8] found an AUC of 0.907 (95% CI 0.899–0.914) for FPG as a test to direct further testing for the detection of GDM. The authors recommended a cut-off point of 4.4 mmol/l (80 mg/dl), which provided a sensitivity of 95.4% and specificity of 32.0%. Another recent study also suggested the cut-off value of 4.4 mmol/l, finding a sensitivity of 87.8% and specificity of 45.8%, although the AUC found for FPG was lower, 0.836 (95% CI 0.829–0.843) [9]. However, neither of these studies evaluated the performance of FPG, when used for this purpose, in predicting pregnancy outcomes, as we have done. Even though maternal and fetal outcomes could have been improved by treatment, it should be emphasized that our study occurred in a time when lifestyle interventions were not routinely recommended. Also, diagnostic cut-off values for GDM were higher and not consensual and we excluded from the analyses the few women who received insulin treatment. A feasible FPG strategy would be as follows. A FPG 92 mg/ dl would label women as GDM. Cut-off points of FPG between 80 and 85 mg/dl could then be chosen to indicate the need for 1 h and 2 h post load testing for those with FPG <92 mg/dl, according to resources available in specific settings. Using a cut-off of 80 mg/dl, more than one-third of all women would be classified as requiring the OGTT, in a process which detects 96.9% of IADPSG defined GDM cases. Using a cut-off of 85 mg/ dl, less than 20% of all women would require OGTT testing, in a process which detects 92.5% of these GDM cases. The choice of the threshold is dependent on the local priorities and the resources available for GDM screening. When to perform post-load testing in those for which the FPG does not complete diagnostic classification merits discussion. Ideally, it should be performed during the same
325
clinic visit. However, as has been noted [19], this is not always possible. If the OGTT were performed on a subsequent visit, the diagnostic criteria shown here will somewhat overestimate reality, given that plasma glucose level varies from day to day. We know of no evaluation of this short term variability in pregnancy. However, it is not high in the general population [20], and thus should not alter the diagnostic properties of the criteria greatly. The extrapolation of these results to other populations merits discussion, given the wide variability in the percentage of women diagnosed with GDM exclusively by FPG [21–23]. Although reasons for this variability are not clear, ethnicity and varying rates of obesity may play a role [24]. Incomplete fasting could be another reason. In this regard, our results are more directly applicable to populations with a high fraction of GDM women being diagnosed with FPG. However, studies performed in settings with a lower fraction of GDM diagnosed by FPG [8,9] also show excellent results for this strategy. When incomplete fasting is likely, repeat testing of FPG for women with an initial FPG 92 mg/dl is indicated. A possible additional strategy, which would not require rapid laboratory turnaround, is to test all women initially with a FPG and, on a second visit, to repeat the FPG for confirmation when the initial value was 92 mg/dl and to perform an OGTT when the initial value was less than 92 mg/dl but above the cut-off point of 80 (or 85) mg/dl. An additional potential limitation of our study is the possibility of selection bias due to missing information or incomplete follow-up. However, the OGTT was performed in over 90% of the women enrolled; outcomes were ascertained for at least 74% of all enrolled women and for 80% of those who underwent a 2 h 75 g OGTT. In conclusion, a strategy in which FPG is used to detect cases of GDM and to determine the extent of further testing appears to be a valid approach to identify women with GDM by IADPSG criteria. This strategy would improve the feasibility of applying IADPSG diagnostic criteria by reducing costs and improving the convenience of GDM detection in pregnancy.
Conflict of interest The authors report no conflict of interest.
Acknowledgments We are grateful to the women who participated at this study and the members of the Brazilian Gestational Diabetes Study group. The study was supported in part by the Brazilian Ministry of Health, the Pan-American Health Organization (PAHO), Awards for Groups of Excellence (PRONEX No. 661041/1998-0) of the Brazilian National Council for Technologic and Scientific Development (CNPq), the Foundation for the Support of Research of the State of Rio Grande do Sul (FAPERGS), Bristol-Meyers Squibb Foundation, Becton Dickinson, Bayer of Brasil and Biobra´s. JT received a doctoral fellowship from Capes/CNPq-IEL National- Brazil (No. 013/2008).
326
diabetes research and clinical practice 105 (2014) 322–326
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 hyperglycemia in pregnancy. Diabetes Care 2010;33:676–82. [2] American Diabetes Association. American Diabetes Association Expert Committee. Diagnosis and classification of diabetes mellitus. Diabetes Care 2011;34(Suppl 1):S62–9. [3] WHO Guideline Development Group. WHO – Diagnostic criteria and classification of hyperglycaemia first detected in pregnancy. WHO; 2013. [4] Werner EF, Pettker CM, Zuckerwise L, Reel M, Funai EF, Henderson J, et al. Screening for gestational diabetes mellitus: are the criteria proposed by the International Association of the Diabetes and Pregnancy Study Groups cost-effective? Diabetes Care 2012;35:529–35. [5] Reichelt AJ, Oppermann MLR, Schmidt MI. Guidelines of the 2nd Meeting of The Diabetes and Pregnancy Task Force. Arq Bras Endocrinol Metabol 2002;46:574–81. [6] Reichelt AJ, Spichler ER, Branchtein L, Nucci LB, Franco LJ, Schmidt MI. Fasting plasma glucose is a useful test for the detection of gestational diabetes. Brazilian Study of Gestational Diabetes (EBDG) Working Group. Diabetes Care 1998;21:1246–9. [7] Agarwal MM, Dhatt GS, Punnose J. Gestational diabetes: utility of fasting plasma glucose as a screening test depends on the diagnostic criteria. Diabet Med J Br Diabet Assoc 2006;23:1319–26. [8] Agarwal MM, Dhatt GS, Shah SM. Gestational diabetes mellitus: simplifying the International Association of Diabetes and Pregnancy diagnostic algorithm using fasting plasma glucose. Diabetes Care 2010;33:2018–20. [9] Zhu W, Fan L, Yang H, Kong L, Su S, Wang Z, et al. Fasting plasma glucose at 24–28 weeks to screen for gestational diabetes mellitus: new evidence from China. Diabetes Care 2013;36:2038–40. [10] Schmidt MI, Duncan BB, Reichelt AJ, Branchtein L, Matos MC, Costa e Forti A, et al. Gestational diabetes mellitus diagnosed with a 2-h 75-g oral glucose tolerance test and adverse pregnancy outcomes. Diabetes Care 2001;24:1151–5. [11] World Health Organization. WHO Consultation: Definition, Diagnosis and Classification of Diabetes Mellitus and its Complications: Report of a WHO Consultation. Part 1: Diagnosis and Classification of Diabetes Mellitus.(WHO/ NCD/NCS/99.2) Geneva: World Health Org.; 1999.
[12] Trinder P. Determination of blood glucose using an oxidase-peroxidase system with a non-carcinogenic chromogen. J Clin Pathol 1969;22:158–61. [13] Working Group Report. National High Blood Pressure Education Program Working Group Report on High Blood Pressure in Pregnancy. Am J Obstet Gynecol 1990;163:1691–712. [14] Fletcher RH, Fletcher SW. Clinical epidemiology: the essentials. 4th ed. Lippincott Williams & Wilkins; 2005. [15] Prutsky GJ, Domecq JP, Sundaresh V, Elraiyah T, Nabhan M, Prokop LJ, et al. Screening for gestational diabetes: a systematic review and meta-analysis. J Clin Endocrinol Metab 2013;98:4311–8. [16] Donovan L, Hartling L, Muise M, Guthrie A, Vandermeer B, Dryden DM. Screening tests for gestational diabetes: a systematic review for the U.S. Preventive Services Task Force. Ann Intern Med 2013;159:115–22. [17] Perucchini D, Fischer U, Spinas GA, Huch R, Huch A, Lehmann R. Using fasting plasma glucose concentrations to screen for gestational diabetes mellitus: prospective population based study. Br Med J 1999;319:812–5. [18] Sacks DA, Chen W, Wolde-Tsadik G, Buchanan TA. Fasting plasma glucose test at the first prenatal visit as a screen for gestational diabetes. Obstet Gynecol 2003;101:1197–203. [19] Agarwal MM, Dhatt GS. Comment on zhu et al. Fasting plasma glucose at 24–28 weeks to screen for gestational diabetes mellitus: new evidence from China. Diabetes Care 2013;36:2038–40 [Diabetes Care 2014;37:e33]. [20] Eckfeldt JH, Chambless LE, Shen YL. Short-term, withinperson variability in clinical chemistry test results. Experience from the Atherosclerosis Risk in Communities Study. Arch Pathol Lab Med 1994;118:496–500. [21] 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 Hyperglycemia and Adverse Pregnancy Outcome (HAPO) Study. Diabetes Care 2012;35:526–8. [22] Black MH, Sacks DA, Xiang AH, Lawrence JM. Clinical outcomes of pregnancies complicated by mild gestational diabetes mellitus differ by combinations of abnormal oral glucose tolerance test values. Diabetes Care 2010;33:2524–30. [23] Moses RG, Morris GJ, Petocz P, San Gil F, Garg D. The impact of potential new diagnostic criteria on the prevalence of gestational diabetes mellitus in Australia. Med J Aust 2011;194:338–40. [24] Wong VW. Gestational diabetes mellitus in five ethnic groups: a comparison of their clinical characteristics. Diabet Med J Br Diabet Assoc 2012;29:366–71.