MATERNAL-FETAL MEDICINE

Gestational diabetes in a tertiary care hospital: implicationsof applying the IADPSG criteria

Mukesh M. Agarwal • Gurdeep S. Dhatt •

Yusra Othman

Received: 2 January 2012 / Accepted: 27 March 2012 / Published online: 7 April 2012

� Springer-Verlag 2012

Abstract

Background The American Diabetes Association has

endorsed the International Association of Diabetes and

Pregnancy Groups (IADPSG) recommendation that every

pregnant woman should undergo the 75 g oral glucose

tolerance test (OGTT) to screen for gestational diabetes

mellitus (GDM).

Purpose To find the cost and workload implications of

switching from the current two-step screening of GDM to

the one-step IADPSG approach.

Methods The cost (US $) and laboratory workload units

(WLU) were calculated for three possible strategies: (1)

50 g glucose screen, if positive, followed by the 100 g

OGTT; (2) universal 75 g OGTT; and (3) screening with

the initial fasting plasma glucose of the OGTT.

Results For the 1,101 pregnant women screened in

1 year, the cost of the three strategies was $ 31,985, $

55,250 and $ 35,875, respectively; the laboratory burden

was 28,975 WLU, 18,662 WLU and 12,215 WLU,

respectively.

Conclusions Switching to the one-step, strategy 2 (IAD-

PSG) would increase the cost by 42 % but decrease the

laboratory workload by 36 % compared to the two-step,

strategy 1. However, an initial screen by the fasting plasma

glucose of the OGTT is the ideal strategy, both in terms of

cost and laboratory workload.

Keywords Gestational diabetes � IADPSG � Cost �Screening

Introduction

One of the major problems of gestational diabetes mellitus

(GDM) is the array of algorithms available for screening

and diagnosis. Often, the Obstetric, Health and Endocrine

associations within the same country support markedly

dissimilar schemes for GDM [1] leading to major incon-

sistencies in the approach to GDM. In order to unify the

multiple algorithms, the International Association of Dia-

betes and Pregnancy Study Groups (IADPSG) proposed a

single approach for GDM worldwide [2]. In January 2011,

this strategy was endorsed by the American Diabetes

Association (ADA) [3] and has the potential to be accepted

by most major diabetes and health organizations

worldwide.

The unified IADPSG approach involves using the 75-g

oral glucose tolerance test (OGTT) with diagnostic

thresholds based on the prodigious Hyperglycemia and

Adverse Pregnancy Outcome (HAPO) study [4]; however,

each pregnant woman would need to undergo the OGTT

without any initial screening. This approach, though vali-

dated by extensive research and endorsed by experts, is

extremely demanding. The OGTT is a costly and cum-

bersome test for the patient and the health provider. Before

January 2011, the ADA [5] recommended that low-risk

populations for GDM to use clinical or laboratory screen-

ing methods to avoid the OGTT; the use of universal

OGTT, i.e., every pregnant woman undergo an OGTT, was

limited to high-risk populations. One such high-prevalence

country is the United Arab Emirates (UAE), a multi-ethnic,

multiracial society with the prevalence of Type 2 DM

M. M. Agarwal (&)

Department of Pathology, Faculty of Medicine, UAE University,

P.O. Box 17666, Al Ain, United Arab Emirates

e-mail: magarwal7@gmail.com

G. S. Dhatt � Y. Othman

Department of Pathology, Tawam Hospital, Al Ain,

United Arab Emirates

123

Arch Gynecol Obstet (2012) 286:373–378

DOI 10.1007/s00404-012-2324-4

(approximately 20.1 %), which is the second highest in the

world [6]. The prevalence of GDM in the UAE varies from

7.9 to 24.9 %, depending on the criteria used for the

diagnosis [7]; however, the prevalence increases to 37.7 %

by the IADPSG criteria [8].

Since 2005, our tertiary care facility in the UAE has

screened every pregnant woman with the 50-g glucose

challenge test (GCT) followed by the confirmatory 100-g

OGTT; this approach was a practical compromise despite

the ADA guidelines for high-risk populations. As the ADA

no longer recommends the two-step screening, it was

decided to institute the new IADPSG guidelines using the

one-step 75-g OGTT. This study was done to audit the

current practice and address the implications of switching

to screening all women with the OGTT.

Patients and methods

Subjects

The patients for this prospective study were the pregnant

women attending routine antenatal clinics at the Tawam

Hospital, Al Ain, United Arab Emirates; this hospital is

under the aegis of Johns Hopkins Medicine, USA.

Approximately 1,500 women are delivered annually at this

hospital. This hospital provides routine antenatal care

through a group of well-trained obstetricians while an

endocrinologist and a nutritionist oversee the care for

women with GDM. Currently, as pointed out earlier, the

hospital follows the two-step screening approach to GDM

(50-g GCT screening followed by the diagnostic 100-g

OGTT), which was the former recommendation of the

ADA and is still endorsed by the American College of

Obstetricians and Gynecologists (ACOG).

During a 12-month study period (1 July 2009–30 June

2010), 1,101 women underwent the 50-g GCT. According to

the hospital guidelines, all women with GCT C 7.8 mmol/l

(140 mg/dl) are scheduled to undergo an OGTT within the

same week. The GCT was planned at 24–28 weeks gesta-

tion; however, screening for GDM was done earlier for

clinical reasons (GDM in an earlier pregnancy or a strong

family history of diabetes) or later, if the pregnant woman

presented late for her first booking. Overall, during the study

period, 1,101 women underwent the GCT; 55 women did not

return for their scheduled OGTT; 10 women vomited and

could not complete the OGTT. All 1,101 women undergoing

the GCT were included in the study to determine the cost and

laboratory workload of switching to the universal 75-g

OGTT.

No informed consent was obtained from the patients, as

this study was a laboratory-based audit.

Blood collection and measurements of glucose

Usually, the GCT was done on the day of the clinic. A

venous blood sample was drawn from the non-fasting

pregnant woman 1 h after 50-g oral glucose (Trutol, 10

fluid ounces (296 ml) dextrose beverage, Nerl diagnostics,

Rhode Island, USA).

For the OGTT, the pregnant woman arrived at the

phlebotomy following an 8–10 h fast. The standard pro-

tocol for the OGTT was used; venous blood samples were

collected for fasting and 1, 2 and 3-h post oral 100-g glu-

cose (Trutol, 10 fluid ounces (296 ml) dextrose beverage,

Nerl Diagnostics, Rhode Island, USA).

The plasma glucose was estimated by the glucose oxi-

dase method (DxC-800 Synchron Clinical System, Beck-

man-Coulter Instruments, Brea, CA, USA). The overall

coefficient of variation (CV) of day-to-day imprecision for

glucose at 8.0 mmol/l (144 mg/dl) was 1.4 %. The hospital

laboratory subscribes to an overseas external quality con-

trol from the College of American Pathologists (CAP). The

average absolute deviation of glucose laboratory results

from the target mean during the study period was 2.3 %.

Hence, the laboratory met the analytical standards for

glucose [9]. The hospital uses the criteria of the ADA for

the diagnosis of GDM as applied to the 100-g OGTT.

GDM diagnostic criteria of the ADA [5] [i.e., if 2 or more

plasma venous glucose values, out of the 4 glucose values

of the OGTT (at 0, 1, 2 and 3 h), are equal to or more than

mmol/l (mg/dl) at 0 h, 5.3 (95.0); 1 h, 10.0 (180); 2 h, 8.6

(155); 3 h, 7.8 (140)]. Women with a GCT C 10.0 mmol/l

(180 mg/dl), were labeled as GDM, as per hospital policy;

these recommendations are a modification of the Canadian

Diabetes Association, which advocates that a GCT C 10.3

is confirmatory for GDM [10].

Data analysis

Data were logged into a computer database and analyzed

using IBM SPSS version 19 for Windows (IBM North

America, New York, United States). The distribution of

continuous variables was tested for normality by the

Shapiro–Wilk test. Non-parametric tests were used, as the

data were not normally distributed. Continuous variables

were compared by the Mann–Whitney test. The Chi square

test or the Fisher exact test was used to compare categorical

variables. The level of significance was 0.05.

The calculation of the costs (in US $) of the tests

involved (based on the insurance billing) was as follows: $

15.2, fasting plasma glucose; $ 18.6, 50-g GCT; $ 50, 75 or

100-g OGTT.

The standards for management information systems in

Canadian Health Service Organizations (MIS standards)

374 Arch Gynecol Obstet (2012) 286:373–378

123

were used to calculate the laboratory workload units

(WLU) [11]. Briefly, the WLU is established for all the

common laboratory tests. One WLU is equivalent to 1 min

of time; the number of WLUs of a test reflects the actual

hands-on time (in minutes) that the test requires to be

completed. WLUs are based on time studies and are peri-

odically updated to ensure they remain accurate and cur-

rent. Based on these standards, each glucose analysis by

our methodology was 5.25 WLU, which includes the fol-

lowing: (1) venous puncture, 1.7 WLU; (2) receipt in

laboratory, 0.6 WLU; (3) processing, 2.9 WLU, and (4)

storing sample, 0.05 WLU. For the OGTT/50-g GCT, the

patient would have to be explained the procedure, which

carries an additional 1.2 WLU (besides the WLU for the

glucose analysis described earlier).

Thus, WLU for all the tests were as follows: 50-g

OGTT, 6.45 WLU; FPG, 5.25 WLU; 75-g OGTT, 16.95

WLU; 100-g OGTT, 22.2 WLU; 75-g OGTT without the

FPG, 11.7 WLU.

The cost and laboratory workload were calculated for

three possible strategies: (1) 50-g GCT screen with 100-g

OGTT for positive screen, (2) universal 75-g OGTT; and

(3) initial FPG screen followed by the 75-g OGTT in

women with FPG between 4.4 mmol/l (79 mg/dl) and

5.0 mmol/l (90 mg/dl).

The third strategy was based on our earlier recommen-

dation for resource strapped countries [8, 12]. It has been

supported by independent reviewers (http://plus.mcmaster.

ca/evidenceupdates/Default.aspx). In this algorithm, the

FPG is used to limit the number of OGTTs needed for

the diagnosis of GDM using a two-threshold rule-in and

rule-out method. Two (instead of one) cut-off values are

used for a screening test (FPG, in this case). The higher

cut-off, which has an inherently increased specificity, is

used to rule in the disease (GDM); the lower cut-off

with its innately increased sensitivity is used to rule out

the disease. Subjects who have FPG values, in between

these two selected thresholds, are considered indetermi-

nate and would need the OGTT. The pregnant woman,

after her fasting glucose is drawn, should wait for the

FPG result, which the laboratory performs on a stat

basis. Based on the result, the decision to stop or pro-

ceed with the OGTT can be made. Thus, initial testing

by FPG can significantly decrease the number of the

OGTTs needed for the diagnosis of GDM. The thresh-

olds [5.1 mmol/l (92 mg/dl) and 4.4 mmol/l (79 mg/dl)]

chosen are based on data from our population and data

from the HAPO study [8, 12]. In this population, both

the specificity and positive predictive value of the FPG

for GDM diagnosis at 5.1 mmol/l (92 mg/dl) were

100 %. The sensitivity and negative predictive value of

FPG at 4.4 mmol/l (79 mg/dl) was, respectively, 95.4

and 91.9 % [8].

Results

Results of the glucose challenge test and GDM

prevalence

As shown in Fig. 1, in the total group of 1,101 subjects, 305

(27.7 %) women had a GCT C 7.8 mmol/l (140 mg/dl). Of

these 305 women, only 214 (70.2 %) women underwent the

OGTT. Of the 91(29.8 %) women not undergoing the

OGTT, 55 women refused the OGTT and did not come for

their scheduled appointment; 10 women vomited and could

not complete the OGTT; 26 women [GCT [ 10.0 mmol/l

(180 mg/dl)] were not worked up further with the OGTT,

since they were labeled as GDM. Additionally, 9 women

underwent the OGTT though their GCT was \7.8 mmol/l

(140 mg/dl) due to a strong clinical history suggestive of

GDM. Of the 223 women who underwent an OGTT, 72 had

GDM by the old ADA criteria for 100-g OGTT; an addi-

tional 26 women were labeled as GDM as their

GCT C 10 mmol/l (180 mg/dl). Thus, the prevalence of

GDM (excluding the 65 women who could not undergo the

OGTT) was 9.5 %. Assuming the sensitivity of the GCT to

be 80.0 % [5], the final GDM prevalence was calculated to

be 12.0 %.

Maternal age and gestational age

The median maternal age was 29 years (range 17–46). The

median gestational age (at time of OGTT) was 26.0 weeks

(range 5–39). The women with GDM were older with

higher fasting, 1, 2 and 3-h plasma glucose (P \ 0.001).

Ethnicity

The major ethnic groups of the study population were 1040

(94.5 %) Arabs; 31 (2.8 %) South Asians, i.e., nationals of

the Indian subcontinent (India, Pakistan, Bangladesh, and

Sri Lanka) and 26 (2.4 %) ‘Other’ nationalities from

multiple diverse countries like Canada, Philippines, UK,

and Nigeria were also represented. The ethnicity of 4

(0.4 %) women was unavailable.

The Arabs were segregated into groups as follows, n (%

of total cohort): 815 (74.0 %), UAE Arabs; 134 (12.2 %),

Asian Arabs from Oman, Kuwait, Saudi Arabia; 68

(6.2 %), Chami Arabs from Iraq, Lebanon, Syria and Pal-

estine; 18 (1.6 %), North African Arabs from Egypt,

Morocco, and Libya; 5 (0.5 %), East African Arabs from

Somalia and Sudan.

Costs

Thus, the yearly cost of each of the three strategies would

be as follows (Table 1): Strategy 1 50-g GCT screen (1,101

Arch Gynecol Obstet (2012) 286:373–378 375

123

women) with 100-g OGTT for positive screen (305

women): $ 31,985; Strategy 2 75-g OGTT only (1,101

women): $ 55,250; Strategy 3 FPG screen (1,101 women)

and 75-g OGTT (550 women): $ 35,875.

Laboratory WLU

Strategy 1 50-g GCT screen (1,101 women) with 100-g

OGTT for positive screen (305 women): 28,975 WLU.

Strategy 2 75-g OGTT only (1,101 women): 18,662 WLU.

Strategy 3 FPG screen (551 women) and 75-g OGTT

(550 women): 12,215 WLU.

Discussion

Before January 2011, for high-risk populations, the ADA

recommended that all women should undergo the OGTT

without the initial screening by the GCT. This is because it

is deemed that too many women in a high-prevalence

population would test positive on screening (by the GCT)

and would need to be further tested by the OGTT making

the screening unnecessary. There is limited evidence sub-

stantiating that in high-risk populations universal screening

using a one-step OGTT (without any screening) is better

[13]. However, this is not borne out in our high-risk pop-

ulation. In the current study, only 27.7 % women tested

positive; thus, the OGTT would not be needed in the

remaining two-third women.

However, in this population, the reasons to opt for the

universal 75-OGTT (i.e., IADPSG approach) are many.

There were many problems using the two-step approach.

i.e., the 50-g GCT followed by the 100-g OGTT. Of the 305

women with the positive GCT [i.e., GCT C 7.8 mmol/l

(140 mg/dl)], 55 (18 %) women did not report for their

OGTT as scheduled. The potential reason for this refusal

may be their negative experience with the GCT. In that

case, it would appear that the universal screening with the

OGTT in this population is preferable.

Strategy 1: the original two-step approach

Cost wise, it was the cheapest approach being 42 % less

than the IADPSG approach. However, this does not

translate into less workload for the laboratory. In fact, this

strategy the workload was 36 % more than if the 75-g

Women undergoing 50-gGCT

(n=1101)

GCT ≥ 7.8 mmol/l (140 mg/dl)

(n= 305)

GCT <7.8 mmol/l (140 mg/dl)

(n= 796)

Strong clinical history; underwent OGTT

(n=9)

Underwent OGTT(n= 214)

Did not show up for appointment (n= 55)

Vomited (n= 10)

GDMGCT ≥ 10 mmol/l (180 mg/dl)

(n= 26)

GDM(n=3)

GDM(n=69)

Fig. 1 Consort flow chart

Table 1 Cost and laboratory workload

Strategy Cost per year,

US $

WLU per

year

Universal 50-g GCT screen

with 100-g OGTT

31,985 28,975

Universal 75-g OGTT 55,250 18,662

FPG screen and 75-g OGTT 35,875 12,215

GCT glucose challenge test, FPG fasting plasma glucose, OGTT oral

glucose tolerance test, WLU workload units

376 Arch Gynecol Obstet (2012) 286:373–378

123

OGTT was done in all the women. Also, with the GCT

sensitivity for GDM at 80 % (i.e., 20 % GDM women

would be missed), coupled with the number of women

refusing the OGTT would point to the fact this approach

should have never been used in this population.

Strategy 2: the new IADPSG approach

The latest recommend approach by the IADPSG was the

costliest of the three strategies. The laboratory workload

would be intermediate, in between the two other approa-

ches. A major limitation of our setting is the limited

phlebotomy space. For the GCT, women come to the

phlebotomy during the whole day so there is no over-

crowding. However, with the IADPSG approach, all

women would come in the morning about the same time

after an overnight fast. This would lead to severe crowding

but this problem can be overcome.

Strategy 3: using the initial FPG to decide the need

for the OGTT

In this population, we have shown that the initial FPG (of

the OGTT) can avoid about 50 % of the OGTTs [8]. This

approach results in the least laboratory workload among all

the three strategies. After considering the costs and the

workload of the laboratory, this is the best approach. Fur-

thermore, in our hospital all women have to wait for the

FPG result to rule out overt diabetes mellitus; the OGTT is

not done if the FPG C 7.0 mmol/l (126 mg/dl).

There are many studies that have looked at costs of

different strategies for GDM [13–16]. However, now that

we have a new, widely touted strategy, i.e., IADPSG

approach, it is imperative that hospitals look at the cost

implications before applying this scheme to their setting.

In our hospital, the cost of the 75 and 100 g OGTT were

the same, while in Canada they are different with the

latter being more expensive. Moreover, this Canadian

study also considered indirect costs to the patient: cost of

travel for the additional OGTT and the time lost for travel

and waiting for the OGTT by considering the ‘‘hourly

wage’’ [12]. We have considered costs from the patient’s

perspective and not from the standpoint of the health

provider, i.e., if the patient pays more, it would mean

more cost for the patient though the hospital would get

more revenue.

In summary, this study highlights the cost and work-

load implications of switching to the new IADPSG cri-

teria. These IADPSG recommendations are costlier, but

they are based on sound scientific data. Furthermore, for

the additional reasons outlined, they are ideal for our

population. An obvious limitation of our cost-benefit

analysis is that the data presented is applicable only to our

population, which is at a high-risk for GDM. The cost and

workload of the laboratory would be different for different

populations. Thus, it is imperative that every hospital

work out these costs for switching to the new IADPSG

strategy in their population. Then, they can evaluate the

cost-benefit ratio of the IADPSG approach similar to this

study. Only long-term follow-up would confirm if pre-

ventive measures in the additional women identified with

GDM (by the IADPSG criteria) will help to forestall Type

2 DM, which may be the ultimate benefit for identifying

women with GDM.

Conflict of interest Nothing to declare.

References

1. Agarwal MM (2010) Evolution of screening and diagnostic cri-

teria for GDM worldwide. In: Kim C, Ferrara A (eds) Gestational

diabetes during and after pregnancy. Dordrecht, New York,

pp 35–49

2. International Association of Diabetes and Pregnancy Study

Groups (2010) Recommendations on the diagnosis and classifi-

cation of hyperglycemia in pregnancy. Diabetes Care 34:e102

3. Diagnosis and Classification of Diabetes Mellitus (2011) Position

statement American Diabetes Association. Diabetes Care

33(Suppl 1):S62–S69

4. HAPO Study Cooperative Research Group (2008) Hyperglyce-

mia and adverse pregnancy outcomes. N Engl J Med

358:1991–2002

5. Gestational Diabetes Mellitus (2004) Position statement Ameri-

can Diabetes Association. Diabetes Care 27(Suppl 1):S88–S90

6. International Diabetes Federation. Diabetes e-Atlas, 4th edn,

Brussels, 2009. http://www.eatlas.idf.org. Accessed 21 Dec 2011

7. Agarwal MM, Dhatt GS, Punnose J, Koster G (2005) Gestational

diabetes: dilemma caused by multiple international diagnostic

criteria. Diabet Med 22:1731–1736

8. Agarwal MM, Dhatt GS, Shah SM (2010) Gestational diabetes:

simplifying the IADPSG diagnostic algorithm using fasting

plasma glucose. Diabetes Care 33:2018–2020

9. Sacks DB, Bruns DE, Goldstein DE, Maclaren NK, McDonald

JM, Parrott M (2002) Guidelines and recommendations for lab-

oratory analysis in the diagnosis and management of diabetes

mellitus. Clin Chem 48:436–472

10. Canadian Diabetes Association (2008) Clinical practice guide-

lines for the prevention and management of diabetes in Canada.

Clinical Practice Guidelines Expert Committee. Can J Diabetes

32(Suppl 1):S1–S201

11. Canadian Institute for Health Information. http://www.cihi.ca/

CIHI-ext-portal/internet/en/Document/standards?and?data?sub

mission/standards/mis?standards/MIS_FAQ. Accessed 21 Mar

2012

12. Agarwal MM, Weigl B, Hod M (2011) Gestational diabetes

screening: the low-cost algorithm. Int J Gynecol Obstet

115(Suppl 1):S30–S33

13. Meltzer S, Snyder J, Penrod J, Nudi M, Morin L (2010) Gesta-

tional diabetes mellitus screening and diagnosis: a prospective

randomised controlled trial comparing costs of one-step and

two-step methods. BJOG 117:407–415

14. Nicholson WK, Fleisher LA, Fox HE, Powe NR (2005) Screening

for gestational diabetes mellitus: a decision and cost-effective-

ness analysis of four screening strategies. Diabetes Care

28:1482–1484

Arch Gynecol Obstet (2012) 286:373–378 377

123

15. Lavin JP Jr, Lavin B, O’Donnell N (2001) A comparison of costs

associated with screening for gestational diabetes with two-tiered

and one-tiered testing protocols. Am J Obstet Gynecol

184:363–367

16. Poncet B, Touzet S, Rocher L, Berland M, Orgiazzi J, Colin C

(2002) Cost-effectiveness analysis of gestational diabetes melli-

tus screening in France. Eur J Obstet Gynecol Reprod Biol

103:122–129

378 Arch Gynecol Obstet (2012) 286:373–378

123