Methods

Quasi‐randomised controlled trial

Participants

276 pregnant women enrolled – 240 had GDM and 36 had type 1 diabetes.

Setting: 12 Italian diabetes clinics

Inclusion criteria: 240 women with GDM were included within a week of diagnosis of GDM (Carpenter and Coustan 1982 criteria); mean of 28 weeks’ gestation.

Exclusion criteria: not stated.

Interventions

Telemedicine (n = unclear; 88 followed up)

Women received standard care plus telemedicine. They were given training on the use of the equipment and were asked to submit their glycaemic data every week, and more often if necessary, and had a medical examination at the diabetes clinic once a month. Women were asked to report their glucose concentrations (as recorded by the glucometer) using an interfacing device that converted the values into audio tones which could be transmitted via a normal telephone receiver. Women dialled the Glucobeep server’s number and identified themselves by a code – the system received their glycaemic data; they could also record a voice message containing any details they deemed useful to help the physician interpret their glycaemic values. Physicians logged in to the server to download the women’s glucose values and any messages; they analysed the data and recorded prescriptions in a message on the server; women then called the server to hear the message containing any new prescriptions. Both women and physicians received a text message immediately when their messages were received by the other party.

Standard care (n = not clear; 115 followed up)

Women received standard care (see below for details).

All women

Women were given standard care according to the recommendations of the American Diabetes Association; women with GDM were placed on a diet and trained to monitor their blood glucose using a home monitor. Women were asked to measure their blood glucose 4 times per day. Insulin was provided when glucose exceeded 95 mg/dl (5.3 mmol/L) fasting, or 130 mg/dl (7.2 mmol/L) 1 hour after meals. Women had a medical examination every 2 weeks. All women could contact the physician whenever they wished.

Outcomes

Review outcomes reported: maternal morbidity (including gestational hypertension, pre‐eclampsia, eclampsia, hypoglycaemic episodes); caesarean section; use of additional pharmacotherapy (insulin therapy); glycaemic control (HbA1c in third trimester); maternal hypoglycaemia; gestational weight gain; adherence to intervention; quality of life; views of intervention; neonatal morbidity (including hypoglycaemic, hyperbilirubinaemia, respiratory distress syndrome, shoulder dystocia, malformations); macrosomia; gestational age at birth; birthweight; medical examinations and visits to diabetic clinic

Notes

Funding: not reported

Declarations of interest: not reported

Dates: not specified

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

High risk

Quote: "Women were sequentially assigned to the two groups: one patient was followed up using the telemedicine approach, and the next using the conventional approach (usual care)."

Allocation concealment (selection bias)

High risk

As above

Blinding of participants and personnel (performance bias)
All outcomes

High risk

Not feasible to blind participants and personnel

Blinding of outcome assessment (detection bias)
All outcomes

Unclear risk

No detail provided; unclear how lack of blinding would have affected outcomes

Incomplete outcome data (attrition bias)
All outcomes

High risk

270 pregnant women were enrolled (240 with GDM, 36 with type 1 diabetes); 203/240 women with GDM and 32/36 women with type 1 diabetes were analysed; the others were excluded as they did not complete the questionnaires at the end of the study. While the authors reported "The demographic, clinical and metabolic characteristics of the women excluded from the study were no different from those of the women who completed the study," there were unbalanced numbers of women with GDM included in the analysis (88 in intervention; 115 in control) indicating a higher rate of exclusion from the intervention group.

Selective reporting (reporting bias)

Unclear risk

No access to trial protocol to permit confident assessment of selective reporting. Some results (particularly surrounding adherence and views) reported incompletely in text.

Other bias

Unclear risk

Lack of methodological detail provided to determine risks of other bias.

Methods

Randomised controlled trial

Participants

66 women randomised

Setting: University of California, Irvine and Long Beach Memorial Medical Centre, USA

Inclusion criteria: women with GDM who required insulin at or before 30 weeks’ gestation, with a singleton fetus. Women with elevated fasting values at the time of a 3‐hour OGTT received insulin; for others, insulin therapy was initiated if weekly fasting and post‐prandial (1 hour after breakfast) values exceeded 105 mg/dL or 140 mg/dL, respectively.

Women were screened for GDM at 24 to 28 weeks' gestation with a 50 g 1‐hour OGCT; if the value was ≥ 140 mg/dL (7.8 mmol/L) but < 190 mg/dL* (10.6 mmol/L) a 3‐hour OGTT was performed, with GDM diagnosed with any 2 of the following abnormal values: fasting > 105 mg/dL (5.9 mmol/L); 1‐hour > 190 mg/dL (10.6 mmol/L); 2‐hour > 165 mg/dL (9.2 mmol/L); or 3‐hour > 145 mg/dL (8.1 mmol/L). (O'Sullivan and Mahan 1964 criteria; O'Sullivan 1964).
*Women with a value ≥ 190 mg/dL on the initial screening test were also diagnosed with GDM.

Exclusion criteria: women with a history of diabetes before pregnancy, with pre‐existing hypertension, renal disease or autoimmune disorders

Interventions

Postprandial monitoring plan (n = 33)

Women were required to undertake daily monitoring of blood glucose concentrations before breakfast (fasting), and 1 hour after each meal for the duration of the pregnancy.

preprandialmonitoring plan (n = 33)

Women were required to undertake daily monitoring of fasting, preprandial and bedtime capillary blood glucose concentrations for the duration of the pregnancy.

All women

Women were evaluated weekly by the perinatal diabetes team (obstetrician, dietitian, nurse educator, counsellor) unless pregnancy complications (including poor glycaemic control, preterm labour or hypertension) made hospitalisation necessary. Women had a diet prescribed with 30 kcal to 35 kcal per kg of ideal body weight, divided into 3 meals and 1 to 3 snacks (with 40% to 45% of the energy provided by carbohydrates); calorie intake and food choices were adjusted at the weekly visits according to weight gain and blood glucose. All women received split‐dose therapy, with short‐ and intermediate‐acting human insulin, adjusted to achieve fasting blood glucose of 60 mg/dL to 90 mg/dL (3.3 mmol/L to 5 mmol/L) and preprandial values of 60 mg/dL to 105 mg/dL (3.3 mmol/L to 5.9 mmol/L) or post‐prandial values < 140 mg/dL. Women used memory‐based reflectance glucometers to measure their blood glucose; adjustments to insulin doses were made if any of the values were consistently higher than the target concentrations (with efforts made to normalise fasting glucose first).

Outcomes

Review outcomes reported: hypertensive disorders of pregnancy (pre‐eclampsia); caesarean section; perineal trauma (3rd or 4th degree lacerations); gestational weight gain; adherence to intervention (compliance with schedule); use of pharmacotherapy (insulin dose); glycaemic control (change in HbA1c; hospitalisation for glycaemic control; success in glycaemic control); large‐for‐gestational age; stillbirth; gestational age at birth; Apgar score < 7 at 5 minutes; macrosomia; small‐for‐gestational age; birthweight; shoulder dystocia; nerve palsy (Erb’s palsy); bone fracture; hypoglycaemia (requiring glucagon or dextrose infusion); hyperbilirubinaemia

Notes

Funding: not reported

Declarations of interest: not reported

Dates: not specified

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Unclear risk

Quote: "permuted‐block randomization was used;" no further details provided.

Allocation concealment (selection bias)

Unclear risk

No further details provided.

Blinding of participants and personnel (performance bias)
All outcomes

High risk

Not feasible to blind participants and personnel.

Blinding of outcome assessment (detection bias)
All outcomes

High risk

As above; and described in Discussion as a "non‐blinded study".

Incomplete outcome data (attrition bias)
All outcomes

Low risk

No losses to follow up or missing data.

Selective reporting (reporting bias)

Unclear risk

Reported all outcomes as per manuscript methods; however no access to trial protocol/registration to assess selective reporting further.

Other bias

Low risk

No other obvious risk of bias identified.

Methods

Randomised controlled trial

Participants

50 women randomised.

Setting: 2 specialist antenatal diabetes clinics in North Ireland and the Republic of Ireland from January 2012 and May 2013

Inclusion criteria: women with GDM or impaired glucose tolerance following an OGTT (usually at week 24 to 28 weeks); with the ability to use the telemedicine equipment following training, sufficient communication abilities to be fully involved, and willingness to use one of the approved blood glucose meters for the duration of the study (including women who had a previous diagnosis of GDM or impaired glucose tolerance). For GDM or impaired glucose tolerance diagnosis, the report referenced the National Institute of Clinical Excellence: Diabetes in Pregnancy 2008 guidelines.

Exclusion criteria: prior diagnosis of type 1 or 2 diabetes; receipt of oral steroid therapy

Interventions

Telemedicine (n = 24)

Women received standard care plus telemedicine. Women were reviewed weekly using telemedicine (commercially available facilities from Northern Ireland). They were given a set of scales, a blood pressure monitor, blood glucose meter and telemedicine hub (with a small screen and 3 buttons to collect and transit data from the woman’s home). Once a week the hub would activate and remind the woman it was time for her telemedicine session; she would measure her weight and blood pressure and send 7‐day stored blood glucose values to the hub, and would answer 3 questions (Have you been taking your insulin? Have you had any hypoglycaemic episodes? Have you had any intercurrent illness?) using yes/no buttons; the information was transmitted to the central server where it could be accessed by the women’s healthcare practitioner. A website was available where women could review their data. 1 to 2 days after submission of the data, the healthcare practitioner reviewed the data; if there were any problems, the healthcare practitioner could contact the woman by telephone to discuss any changes needed, or arrange a visit.

Standard care (n = 26)

Women received standard care (see below for details).

All women

Standard care followed the National Institute for Health and Care Excellence (NICE) guidelines for the management of GDM or impaired glucose tolerance; women were asked to monitor their blood glucose 7 times per day (before and after each meal and before bed), and to attend a specialist diabetes clinic at least every 2 weeks, where weight and blood pressure were measured, urinalysis was performed and glycaemia was evaluated using self‐monitored blood glucose records and HbA1c.

Outcomes

Review outcomes reported: hypertensive disorders of pregnancy (pre‐eclampsia, pregnancy‐induced hypertension), caesarean section; induction of labour; use of additional pharmacotherapy (diabetes therapy at time of delivery); glycaemic control (HbA1c at 36 weeks; mean self‐monitored blood glucose); maternal hypoglycaemia; gestational weight gain (weight at 36 weeks); adherence (appointments attended; average daily self‐monitored blood glucose frequency (meter memory, diary)); average review length; stillbirth (intrauterine death); preterm birth; macrosomia; shoulder dystocia; respiratory distress syndrome (or transient tachypnoea of the newborn); hypoglycaemia (treated with dextrose); jaundice (no treatment needed; requiring phototherapy); malformations (diabetes‐related); neonatal intensive care unit admission; gestational age at birth; birthweight; head circumference; length

Notes

Funding: "This work constituted part of a PhD for J.E.G., which was funded by the Department for Employment and Learning for Northern Ireland. A small start‐up grant was provided for the study by Derry City Council. The telemedicine service was provided free of charge at one of the trial sites;" "There was no involvement of the funders of this research or the telemedicine service provider in study design, data collection, data analysis, and manuscript preparation or publication decisions."

Declarations of interest: "M.J.O’K. has received research funding from Nova Biomedical (a manufacturer of glucose meters). J.E.G., B.P.B., F.D., and V.E.C. declare no competing financial interests exist."

Dates: January 2012 and May 2013

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Low risk

Random allocation software was used to generate a stratified randomisation scheduled with blocks of 4‐6.

Allocation concealment (selection bias)

Low risk

An independent research secretary was used to allocate women to a group according to the schedule.

Blinding of participants and personnel (performance bias)
All outcomes

High risk

Quote: "Because of the nature of the intervention it was not possible to blind participants or HCPs."

Blinding of outcome assessment (detection bias)
All outcomes

Unclear risk

No details provided; unclear how lack of blinding would have affected outcomes.

Incomplete outcome data (attrition bias)
All outcomes

Unclear risk

Of the 24 women allocated to the intervention group, 3 were excluded from analyses (2 did not have a landline and mobile network coverage to allow data transfer; 1 withdrew); none of the 26 women in the control group were excluded from analyses. Questionnaires completed by 19/24 women in the intervention group and 22/26 in the control group; unclear impact in already small sample. For a number of outcomes, data are taken from the supplementary tables which indicated missing data for some women/babies, for different outcomes (i.e. separate Ns are reported for each outcome).

Selective reporting (reporting bias)

Unclear risk

No access to trial protocol to confidently assess selective reporting.

Other bias

Low risk

Very few baseline characteristics reported, though no clear differences; no other obvious sources of bias identified.

Methods

Randomised controlled trial

Participants

61 women randomised

Setting: The Diabetes‐in‐Pregnancy Program at Temple University Hospital, Philadelphia, USA and/or its satellite hospitals, from March 1998 to November 1999

Inclusion criteria: women with GDM ≤ 33 weeks' gestation and fasting blood glucose ≤ 95 mg/dL on OGTT. GDM criteria not defined.

Exclusion criteria: none stated

Interventions

Self‐monitoring (n = 31)

Women were taught to perform self‐monitoring of blood glucose using a reflectance meter with memory (One Touch Profile) and asked to measure the blood glucose 4 times per day (fasting and 1 hour after meals), for a total of 4 times per week.

Periodic monitoring (n = 30)

Women had their blood glucose concentrations measured (fasting and 1 hour after a meal) at each prenatal visit, or more frequently if clinically indicated.

All women

The diabetes and obstetric management protocol was identical for both groups, except for glucose surveillance. Individualised teaching/counselling regarding GDM and its management was provided to all women by the diabetes educator. All women were treated to attain the same metabolic goals: fasting blood glucose < 95 mg/dL, and 1‐hour post‐prandial glucose < 120 mg/dL; women who failed to meet the targets at least 90% of the time, were started on insulin.

Outcomes

Review outcomes reported: pre‐eclampsia; caesarean section; perinatal mortality; large‐for‐gestational age; placental abruption; postpartum haemorrhage; gestational weight gain; adherence to the intervention (Dietary Compliance Questionnaire); sense of well‐being and quality of life (Diabetes Empowerment Scale; Appraisal of Diabetes Scale); use of additional pharmacotherapy; glycaemic control (preprandial and post‐prandial glucose); stillbirth; neonatal mortality; gestational age at birth; birthweight; hypoglycaemia; hyperbilirubinaemia; number of antenatal visits or admissions (visits with diabetes team); neonatal intensive care unit admission; 'birth trauma'; 'respiratory complications'.

Notes

Funding: "This work was supported by a grant from the General Clinical Research Center branch of the National Center for Research Resources, Grant No. 2M01‐RR‐349" and "We would also like to thank LifeScan Inc, who generously donated the glucose meters for this study."

Declarations of interest: not reported

Dates: March 1998 to November 1999

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Unclear risk

Quote: "A randomized controlled trial was undertaken".

Allocation concealment (selection bias)

Unclear risk

As above; no further details provided.

Blinding of participants and personnel (performance bias)
All outcomes

High risk

Not feasible to blind participants and personnel.

Blinding of outcome assessment (detection bias)
All outcomes

Unclear risk

No detail provided; unclear how lack of blinding would have affected outcomes.

Incomplete outcome data (attrition bias)
All outcomes

Unclear risk

3 women in the periodic monitoring group were lost to follow‐up and were removed, as no outcome data were available; for the Diabetes Empowerment Scale, results were reported for 28/31 (90%) and 19/27 (70%) women in the 2 groups, indicating fewer women in the control group completed the questionnaire, but no reasons were given. It appears that all 31 and 27 women completed the other questionnaires.

Selective reporting (reporting bias)

Unclear risk

Not possible to assess confidently; no access to trial protocol. Insulin therapy reported only in Abstract, and not in Results text of manuscript.

Other bias

Low risk

No other obvious sources of bias identified.

Methods

Randomised controlled trial

Participants

63 women randomised

Setting: prenatal clinics at Temple University Hospital, Philadelphia, USA, or one of its satellites; from September 2004 to May 2006

Inclusion criteria: women with GDM (3‐hour OGTT using Carpenter and Coustan 1982 criteria), between 18 and 45 years, at 33 weeks' gestation or less

Exclusion criteria: women with prior history of glucose intolerance, or with multiple gestations

Interventions

Telemedicine (n = 34)

Women were asked to transmit information via a diabetes health network at least 3 times per week to their healthcare provider. Women without access to the internet received a refurbished computer with free telephone‐based web access to be used during the study; they received a 1‐hour training session by graduate/undergraduate students on how to use a computer, how to access websites, how to set up an email address and receive/send emails. The intervention used 'ITSMyHealthfile', a web‐based disease management interactive healthcare delivery system, with a secure internet server and database which allowed women to send blood glucose and other health data directly to their care provider (blood glucose; fetal movement counts; insulin doses; episodes of hypoglycaemia; ask questions/messages), and receive information and advice from healthcare providers. It required a log‐on ID and password.

Standard care (n = 29)

Women were asked to record information in a log‐book, which was reviewed by the medical team at prenatal visits.

All women

All received standard care in the ‘diabetes‐in‐pregnancy program’: were seen for clinical evaluation every 2 weeks until 36 weeks, after which they were seen weekly. Care was provided by a team of maternal‐fetal medicine specialists, residents, diabetes educator, and nutritionists. All received individualised dietary counselling and diabetes education and were instructed in glucose self‐monitoring with portable reflectance meters – women were asked to monitor 4 times per day (before breakfast; 2 hours after meals), and were treated to maintain: fasting glucose ≤ 95 mg/dL, and 2‐hour ≤ 120 mg/dL; women who failed to meet the targets > 90% of the time were started on glyburide or insulin. Women were also asked to perform fetal movement counting 3 times per day and record insulin doses and episodes of hypoglycaemia.

Outcomes

Review outcomes reported: hypertensive disorders of pregnancy (pre‐eclampsia/gestational hypertension); caesarean section; placental abruption; use of additional pharmacotherapy (glyburide; insulin); glycaemic control (fasting blood sugar; blood glucose 2 hours post breakfast, lunch, dinner; mean; HbA1c at birth); adherence to intervention (frequency of monitoring; appointment adherence); sense of well‐being and quality of life (maternal feelings of diabetes self‐efficacy); large for gestational age; perinatal mortality; neonatal mortality or morbidity composite ('composite outcome'); stillbirth; neonatal mortality; preterm birth; respiratory distress syndrome/respiratory complications; hypoglycaemia; hyperbilirubinaemia/jaundice; neonatal intensive care unit admission; gestational age at birth; birthweight

Notes

Funding: "This study was supported by grant RO3 NR008776‐01 from the National Institute of Nursing Research, National Institutes of Health".

Declarations of interest: not reported

Dates: September 2004 to May 2006

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Unclear risk

Quote: "Women were randomized into one of two groups."

Allocation concealment (selection bias)

Unclear risk

As above; no further details provided.

Blinding of participants and personnel (performance bias)
All outcomes

High risk

Not feasible to blind participants and personnel.

Blinding of outcome assessment (detection bias)
All outcomes

Unclear risk

No detail provided; unclear how lack of blinding would have affected outcomes.

Incomplete outcome data (attrition bias)
All outcomes

Low risk

29 women randomised to control group; 2 lost to follow‐up, 1 excluded (twin pregnancy), 1 did not meet criteria for GDM; therefore 25 included in analyses; 34 women randomised to intervention group, 2 formally withdrew; therefore 32 included in analyses.

Selective reporting (reporting bias)

Unclear risk

No access to trial protocol to enable confident assessment of selective reporting.

Other bias

Low risk

Comparable groups at baseline; no other obvious sources of bias apparent.

Methods

Randomised controlled trial

Participants

80 women randomised

Setting: prenatal clinics at Temple University Hospital, Philadelphia, and Diabetes Education Program at Tallahassee Memorial Hospital, Tallahassee, USA; from September 2007 to November 2009

Inclusion criteria: women with GDM (on 3‐hour OGTT using the Carpenter and Coustan 1982 criteria), between 18 and 45 years, at 33 weeks' gestation or less

Exclusion criteria: women with a prior history of glucose intolerance outside of pregnancy, or multiple gestations (twins, triplets)

Interventions

Telemedicine (n = 40)

Women were trained in the use of a computer and the internet, instructed on the details of the Internet program, and introduced to the website on a demonstration terminal at the clinic; automated telephone communication option and instructions were also offered for women with no Internet access. Women were provided with a password and log‐in name to gain access to the website, and an identification number to gain access to the telephone communication system. Women were prompted to input clinical data (blood glucose readings, changes in medications, episodes of hypoglycaemia) and identify the day or time. Women were provided with feedback, emotional support and reinforcement regarding self‐management with each transmission. Women received brief education message or suggestion each time they accessed the system. Women were asked to transmit information via phone/internet weekly to healthcare providers. The telemedicine system used was 'ITSMy Healthrecord', a web‐based, nurse‐co‐ordinated communication system, composed of a secure Internet server, a phone system and database, allowing data transfer (messages or clinical information) from patient to practice, and practice to patient in an asynchronous manner. Women dialled a toll‐free number or logged on through the internet using a unique log‐in and password – and could send messages or ask a question (45 seconds, or unlimited text input); nurses could then listen/review and respond (by typing a response that could be read or listened to); the system also provided reminders for women to record and transmit their data when required (3 phone call reminders, a day apart).

Standard care (n = 40)

Women were asked to record information in a logbook which was reviewed by the medical team at prenatal visits.

All women

Women were asked to monitor blood glucose daily (4 times per day: before breakfast and 2 hours after each meal). Women were treated to obtain metabolic goals: fasting glucose ≤ 95 mg/dL, 2‐hour: ≤ 120 mg/dL; women who did not meet these targets > 90% of the time were started on glyburide/insulin. Women were asked to perform fetal movement counting 3 times a day, and record insulin doses and episodes of hypoglycaemia. All women received standard care, and were seen for clinical evaluation every 2 weeks until 36 weeks, after which they were seen weekly (car provided by team: maternal‐fetal specialists, residents, diabetes educators, nutritionists); all women received individualised dietary counselling and diabetes education, and were instructed in glucose self‐monitoring.

Outcomes

Review outcomes reported: hypertensive disorders of pregnancy (pre‐eclampsia/gestational hypertension); caesarean section; use of additional pharmacotherapy (oral agents; insulin); glycaemic control (fasting blood sugar; blood glucose 2‐hour post breakfast, lunch, dinner; mean); adherence to intervention (frequency of monitoring); large‐for‐gestational age; perinatal mortality; stillbirth; neonatal mortality; preterm birth; respiratory distress syndrome/respiratory complications; hypoglycaemia; hyperbilirubinaemia/jaundice; neonatal intensive care unit admission; gestational age at birth; birthweight; length of stay in neonatal intensive care unit

Notes

Funding: "This study was supported by grant R21‐DK‐071694 from the National Institute of Diabetes and Digestive and Kidney Diseases, National Institutes of Health".

Declarations of interest: "C.J.H., L.D., K.R., W.M., D.M., and J.G. have nothing to disclose. W.P.S. has stock ownership in Insight Telehealth Systems. A.A.B. is a consultant for Insight Telehealth Systems".

Dates: September 2007 to November 2009

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Unclear risk

Quote: "Women were randomized into one of two groups."

Allocation concealment (selection bias)

Unclear risk

As above; no further details provided.

Blinding of participants and personnel (performance bias)
All outcomes

High risk

Not feasible to blind participants and personnel.

Blinding of outcome assessment (detection bias)
All outcomes

Unclear risk

No detail provided; unclear how lack of blinding would have affected outcomes.

Incomplete outcome data (attrition bias)
All outcomes

Low risk

3/40 women in the intervention group and 2/40 in the control group were lost to follow‐up; 1 woman in the intervention group experienced a fetal loss at 19 weeks.

Selective reporting (reporting bias)

Unclear risk

No access to trial protocol to enable confident assessment of selective reporting.

Other bias

Low risk

Comparable groups at baseline; no other obvious sources of bias apparent.

Methods

Randomised controlled trial

Participants

73 women randomised

Setting: Finland

Inclusion criteria: women with GDM, with singleton pregnancies. All women had a 2‐hour 75 g OGTT performed at 22 to 34 weeks' gestation as they belonged to a high‐risk group according to the evaluation system used in Finland (BMI > 25, > 40 years, previous child over 4500 g, glucosuria during pregnancy, weight gain during pregnancy > 20 kg, previous GDM or suspected fetal macrosomia). GDM was defined as having at least 2 abnormal high plasma glucose values out of 3 measurements in the 75 g OGTT (fasting > 5.1 mmol/L, 1‐hour > 10 mmol/L, 2‐hour > 8.7 mmol/L).

Exclusion criteria: not reported

Interventions

CGMS (n = 36)

Women were taught how to use the CGMS equipment (and at least 4 daily plasma glucose calibration values were introduced to the apparatus); the mean duration of the registration period was 47.4 (2.5) hours. Women were asked not to take a shower during the monitoring period.

Self‐monitoring (n = 37)

All women

All women came to the hospital for an interview and dietary counselling for low‐GI, low saturated fat eucaloric diet, and were taught how to measure plasma glucose, with an instruction to measure it 5 times per day (fasting plasma glucose, preprandial values and post‐prandial values 90 minutes after main meals), and asked to keep a dietary diary, and record exercise also. Treatment mode was determined within a week of starting monitoring; diet alone was considered insufficient if fasting plasma glucose was > 5.5 mmol/L twice, or was at least once > 5.5 mmol/L and post‐prandial value > 7.8 mmol/L or fasting plasma glucose ≤ 5.5 mmol/L but post‐prandial was at least twice > 7.8 mmol/L. Insulin was offered if at least 2 post‐prandial measurements were > 8.0 mmol/L; women with elevated fasting plasma glucose only were initially treated with metformin, and later combined with insulin if needed.

Outcomes

Review outcomes reported: hypertensive disorders of pregnancy (pre‐eclampsia, pregnancy‐induced hypertension); caesarean section; perineal trauma; use of additional pharmacotherapy (insulin; metformin); gestational weight gain (total); perinatal mortality; stillbirth; neonatal mortality; gestational age at birth; preterm birth; macrosomia; birthweight; hypoglycaemia; hyperbilirubinaemia (UV treated); neonatal intensive care unit admission; length of stay in neonatal intensive care unit.

Notes

Funding: "Turku University Central Hospital Research Fund, and The Foundation of Gynaecologists and Obstetricians in Finland supported this study".

Declarations of interest: not reported

Dates: not specified

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Unclear risk

Quote: "Patients were randomly allocated"; no further details provided.

Allocation concealment (selection bias)

Unclear risk

No details provided

Blinding of participants and personnel (performance bias)
All outcomes

High risk

Not feasible to blind participants and personnel.

Blinding of outcome assessment (detection bias)
All outcomes

Unclear risk

No details provided; unclear how lack of blinding would have affected outcomes.

Incomplete outcome data (attrition bias)
All outcomes

Unclear risk

It was not clear whether there were any women lost to follow‐up, or missing data for the outcomes reported, as this information was not provided.

Selective reporting (reporting bias)

High risk

Some results reported in text but not in tables, e.g. "There were no statistically significant differences between the two groups in… frequency of pre‐eclampsia, pregnancy‐induced hypertension, maternal lacerations… Apgar at… 10 min"; additional outcomes discussed in methods (e.g. shoulder dystocia, and PROM) not reported in results. Ns for hypoglycaemia and NICU admission unclear, as percentages indicate may not be 37 for control group.

Other bias

Unclear risk

Limited methodological details provided did not permit thorough assessment of other sources of bias.

Methods

Randomised controlled trial

Participants

Setting: the Endocrinology and Metabolism Clinic of a large Midwestern health system; presumed to be in USA

Inclusion criteria: pregnant women 18 years or older who sought treatment at the Endocrinology and Metabolism Clinic of a large Midwestern health system and were diagnosed with GDM. GDM diagnostic criteria not reported.

Exclusion criteria: women who had been diagnosed with type 1 or 2 diabetes prior to pregnancy, or who were unwilling/unable to follow the protocol

Interventions

Modem transmission (n = unclear; 72 across both groups; 18 analysed for satisfaction outcomes)

Women transmitted blood glucose data to the clinic via the Acculink Modem; women were trained on how to use the modem to transmit blood glucose data from the meter to the clinic.

Telephone transmission (n = 72 across both groups; 20 analysed for satisfaction outcomes)

Women transmitted blood glucose data to the clinic via telephone calls directly to clinic personnel (research assistant/registered nurse).

All women

In both groups, women self‐tested blood glucose with an Accu‐Chek Complete meter. The duration of participation was from the first clinic visit to birth (thus length of participation was dependent on gestational age at GDM diagnosis; the mean duration was 6 weeks, with a range of 1 week to 22 weeks). At their first clinic visit, healthcare providers trained women about self‐monitoring of blood glucose, nutrition, and overall GDM management through educational literature and videos; if necessary, women received instruction on insulin administration; all women were asked to record the blood glucose results in a diary; all women were instructed to test and record their results 5 times daily, in accordance with standard care (before breakfast; 1 hour after each meal; before bed). Standard care involved reporting blood glucose values daily for the first 2 weeks after the first visit, then weekly thereafter; clinic personnel reviewed the blood glucose data received and provided feedback and guidance to the women via the telephone. Subsequent clinic visits were scheduled for 1 week after the first, and then every 2 weeks until 1 month before birth, then weekly until birth.

Outcomes

Review outcome reported: views of the intervention (percentages were provided and used to calculate group numbers); gestational age at birth

Notes

Funding: "Roche Diagnostics provided the authors with a grant as well as the blood glucose monitor supplies to do this study"

Declarations of interest: not reported

Dates: not specified

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Unclear risk

Quote: "were randomized by block randomization into two groups".

Allocation concealment (selection bias)

Unclear risk

No further details provided

Blinding of participants and personnel (performance bias)
All outcomes

High risk

Not feasible to blind participants and personnel.

Blinding of outcome assessment (detection bias)
All outcomes

Unclear risk

No details provided; unclear how lack of blinding would have affected outcomes.

Incomplete outcome data (attrition bias)
All outcomes

High risk

Results reported incompletely (see below), therefore not possible to determine incomplete outcome data for some outcomes (mean telephone consultation time; length of clinic visit; weeks of gestation); participant satisfaction survey response rate was 52%.

Selective reporting (reporting bias)

High risk

No access to published trial protocol/trial registration; few outcomes reported (no clinical outcomes, except weeks of gestation) and focused on consultation time, and satisfaction only. A number of results reported as P values only.

Other bias

Unclear risk

No baseline characteristics were reported for the 2 groups (ethnicity and gestational age, and age across the population were reported).

Methods

Randomised controlled trial

Participants

100 women randomised

Setting: the Unit of Gestational Diabetes of the Hospital Clinico Universitario San Carlos of Madrid, Spain from June to December 2007

Inclusion criteria: women with GDM (Carpenter‐Coustan criteria) before 28 weeks

Exclusion criteria: not detailed

Interventions

Telemedicine (n = 50)

Women received a Glucometer (Accu‐Chek Compact Plus) with a cellular phone (which allowed the transmission of glucose values to the central database via a short message service SMS). Women were recommended to send their glucose values to the medical terminal once a week. An endocrinologist and diabetes nurse educator evaluated the data from a PC with Internet connection. Health professionals could then send text messages from their computer to the women, making recommendations for nutritional changes, or adjustments in insulin doses; women could send text messages with questions/answers.

Standard care (n = 50)

Women were followed according to standard protocol (face‐to‐face outpatient monitoring), and were given the opportunity to attend the outpatient clinic without prior appointment, and bring in their logbook when their glucose values were above the targets, of for any queries re: nutritional recommendations or insulin dose.

All women

Women were instructed by the nurse educator in nutritional habits and self‐monitoring of blood glucose and informed about the goals of glycaemic control: fasting blood glucose and preprandial blood glucose < 95 mg/dL; 1‐hour post‐prandial blood glucose < 120 mg/dL. At visit 1 (before 28 weeks' gestation), capillary blood glucose values were evaluated (6 measurements a day were recommended during the first week; if more than 4 of 5 fasting and pre‐meal glycaemic values were < 95 mg/dL in the first week, only 1‐hour post‐meal capillary blood glucose measurements were recommended daily or every other day until delivery). During the follow up, 4 face‐to‐face visits (once a month) were scheduled until birth; glucose values recorded by women in their log‐books were evaluated and episodes of mild/severe hypoglycaemia and insulin requirements registered.

Outcomes

Review outcomes reported: hypertensive disorders or pregnancy (pregnancy‐induced hypertension); caesarean section; placental abruption; use of additional pharmacotherapy (insulin treated); glycaemic control (HbA1c at visit 4 (39‐40 weeks); HbA1c < 5.8%); gestational weight gain; view of intervention; total contact hours per‐insulin treated woman; face‐to‐face visits; unscheduled visits; costs; large‐for‐gestational age; preterm birth; small‐for‐gestational age; shoulder dystocia; hypoglycaemia; hypocalcaemia; polycythaemia (poliglobulia); gestational age at birth; birthweight

Notes

Funding: "This work was supported by grants from Fundacion para Estudios Metabolicos".

Declarations of interest: "The authors declare that they have no conflicts of interest"

The 8 women most likely to require insulin after the evaluation of the first week were allocated to telemedicine group; and the other 92 were randomised to telemedicine or standard monitoring.

Dates: June to December 2007

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

High risk

Quote "were randomized;" see below

Allocation concealment (selection bias)

High risk

Not detailed; not achieved for 8 women most likely to require insulin who were allocated to telemedicine group.

Blinding of participants and personnel (performance bias)
All outcomes

High risk

Not feasible to blind participants and personnel.

Blinding of outcome assessment (detection bias)
All outcomes

Unclear risk

No detail provided; unclear how lack of blinding would have affected outcomes.

Incomplete outcome data (attrition bias)
All outcomes

Low risk

Of 100 women randomised, 97 (49 in the intervention group and 48 in the control group) completed the study.

Selective reporting (reporting bias)

Unclear risk

No access to trial protocol to enable confident assessment of selective reporting. Some data incompletely reported, e.g. "We did not detect differences in clinical and laboratory data during the follow up nor were differences in delivery and neonatal outcomes observed"; while the average total contact per insulin‐treated woman was reported, no variance was reported for this outcome; and the contact for all women was not clearly reported (only "There was a significant 62% reduction in outpatient clinic visits in women from the telemedicine group"). Methods detailed that loss of work days and the number of hospital admissions were regularly recorded, but no results were reported for these outcomes.

Other bias

Low risk

Comparable groups at baseline; no other obvious sources of bias apparent.

Methods

Randomised controlled trial

Participants

347 women randomised

Setting: Sainte‐Justine Hospital, Montreal, Canada, from 1 June 1993 to 31 May 1994

Inclusion criteria: women with GDM (diagnosed according to institution criteria: 1‐hour 50 g glucose screen ≥ 11.1 mmol/L (200 mg/dL)); or glucose screen between 8.9 (160 mg/dL) and 11.0 mmol/L (198 mg/dL) with 2 or more abnormal values on a 3‐hour 100 g OGTT before 26 weeks (fasting: 5.3 mmol/L (95 mg/dL), 1‐hour: 10 mmol/L (180 mg/dL); 2‐hour: 8.9 mmol/L (160 mg/dL); 3‐hour: 7.8 mmol/L (140 mg/dL)); or during and after 26 weeks (fasting: 5.6 mmol/L (101 mg/dL), 1‐hour: 11.1 mmol/L (200 mg/dL); 2‐hour: 9.2 mmol/L (166 mg/dL); 3‐hour: 8.3 mmol/L (149 mg/dL)).

Exclusion criteria: multiple pregnancy, fetus with congential malformation, current diet or insulin therapy, before 22 weeks or later than 38 weeks. Post‐randomisation exclusions: birth in another centre, birth within 2 weeks of randomisation, steroid therapy

Interventions

Self‐monitoring (home) (n = 172; 112 with breakfast result < 7.8 mmol/L; 60 with breakfast result ≥ 7.8 mmol/L)

Women took blood glucose measurements 3 times daily (1 hour after each meal) alternating with 4 times a day (before each meal and at bed time) using memory‐based reflectance meters, checked regularly.

Periodic monitoring (outpatient follow‐up) (n = 170; 115 with breakfast result < 7.8 mmol/L; 55 with breakfast result ≥ 7.8 mmol/L)

Women had outpatient clinic follow‐up, with capillary fasting and/or 1‐hour post‐breakfast glucose assessment by nurse educator every 2 weeks. Women were transferred to home monitoring as soon as 1 of their glucose assessments was higher than the goals.

All women

After an 8‐hour fast, women ingested standardised breakfast over 10 minutes, at 8am; 1 hour later, blood glucose was sampled: results were divided into those below and at or above 7.8 mmol/L (140 mg/dL).

Target glucose concentrations were 5.3 mmol/L (95 mg/dL) fasting; 5.6 mmol/L (101 mg/dL) before dinner and supper and 7.8 mmol/L (140 mg/dL) 1‐hour post‐prandial. Women were started on insulin therapy when 25% or more of their home blood glucose concentrations during the previous 2 weeks were higher than the previously mentioned goals; insulin was adjusted to reach the same target goals.

All women were prescribed a diet without refined carbohydrates, with 45% complex carbohydrate, 35% fat, 20% protein, divided into 3 meals and 4 snacks: 35 kcal/kg for non‐obese, 25 kcal/kg for obese women; all women were seen at 2‐week intervals.

Outcomes

Review outcomes reported: caesarean section; gestational weight gain; adherence (poor compliance); use of additional pharmacotherapy (insulin therapy); glycaemic control (preprandial glucose; post‐prandial glucose); perinatal mortality; large‐for‐gestational age; stillbirth; gestational age at birth; macrosomia; small‐for‐gestational age; birthweight; shoulder dystocia; hypoglycaemic; hyperbilirubinaemia; costs associated with the intervention

Notes

Funding: "This study was supported by a grant from Lilly Canada"

Declarations of interest: not reported

Dates: June 1993 to May 1994

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Low risk

Quote: "women were randomized with stratification according to their ethnic origin (white and others) to either outpatient clinic follow‐up or self‐monitoring of blood glucose with a meter (home monitoring). A computer‐generated table of random numbers with permuted‐block randomization was used".

Allocation concealment (selection bias)

Unclear risk

No details provided

Blinding of participants and personnel (performance bias)
All outcomes

High risk

Not feasible to blind participants and personnel.

Blinding of outcome assessment (detection bias)
All outcomes

Unclear risk

No detail provided; unclear how lack of blinding would have affected outcomes.

Incomplete outcome data (attrition bias)
All outcomes

Low risk

Of 347 women randomised, 5 were excluded (2 delivered preterm within 2 weeks of enrolment; 2 received long‐term high‐dose steroids for preterm labour; 1 delivered in another centre). Although 2.6% and 52.7% of women with a breakfast < 7.8 mmol/L and ≥ 7.8 mmol/L on clinic follow‐up were transferred to home monitoring respectively, "The data from the women included in the statistics were analysed according to their randomization".

Selective reporting (reporting bias)

Unclear risk

No access to trial protocol to enable confident assessment of selective reporting.

Other bias

Low risk

No other obvious sources of bias identified.

Methods

Randomised controlled trial

Participants

120 women randomised

Setting: Department of Endocrinology, Zhongda Hospital, the Affiliated Hospital of Southeast University, China from September 2011 to December 2012

Inclusion criteria: women between 24 and 28 weeks' gestation with a singleton pregnancy, GDM, and written informed consent.

All women underwent a 75 g OGTT at 24 to 28 weeks according to the criteria of the American Diabetes Association (ADA). Based on the 1‐step approached recommended by the World Health Organization, ADA and IADPSG, women were defined as having GDM if they had at least 1 abnormally high plasma glucose out of the 3 measurements: fasting > 92 mg/dL (5.1 mmol/L), 1‐hour > 180 mg/dL (10.0 mmol/L), or 2‐hour > 153 mg/dL (8.5 mmol/L).

Exclusion criteria: diagnosis of diabetes mellitus, previous treatment for GDM, presence of infection, or other severe metabolic, endocrine, medical or psychological co‐morbidities

Interventions

CGMS (n = 58*)

There were 2 groups: CGMS early subgroup (n = 30): during gestational weeks 24 to 28; and CGMS late subgroup (n = 30): during gestational weeks 28 to 36.

*There were 3 additional women who withdrew following allocation, it was unclear if they were allocated to the early or late subgroup.

The CGMS group was monitoring using CGMS (Gold Medtronic MiniMed, Northbridge, CA, USA), with the sensor inserted into the upper outer buttock of the subjects. While women wore the sensor, glucose concentrations (bedtime, and 1 hour before the beginning of each meal) were also monitored using the Accu‐Chek meters and were input into the CGMS as calibration 4 times per day.

Self‐monitoring (n = 62)

All women

Women were taught to perform self‐monitoring of blood glucose using Accu‐Chek Advantage meters (Roche Diagnostics, Manheim, Germany). Women were instructed to check their level 4 times per day (fasting, and 1 hour after the beginning of each meal) from the first visit at which they received the GDM diagnosis until birth (except during the period where the GGMS group used their device).

Women underwent dietary counselling for a eucaloric diet with a low GI and low saturated fat levels (50% to 60% energy from carbohydrates; 25% to 30% from fat; 15% to 20% from protein; with energy intake distributed as equally as possible throughout the day, based on recommendations of the China Diabetic Association). Moderate intensity exercise was encouraged; 35 to 45 minute sessions, three times a week. A physiotherapist motivated the women to continue or start exercising, and provided written instructions, with walking, swimming and cycling recommended as types of exercise. Insulin treatment was administered under conditions of 2 fasting blood glucose values > 105 mg/dL (5.8 mmol/L), 2 1‐hour post‐prandial levels > 155 mg/dL (8.6 mmol/L), a 2‐hour post‐prandial level > 130 mg/dL (7.2 mmol/L), or a fasting blood glucose > 90 mg/dL (5.5 mmol/L) with at least 2 post‐prandial values > 141 mg/dL (7.5 mmol/L) according to ADA guidelines. Women received NPH insulin as an immediate acting insulin with an initial dose of 0.2 units/kg; if fasting levels were high, treatment was given before bed‐time; if post‐prandial levels were high, regular insulin or short acting insulin was given before meals. Follow‐up meetings were every 2 to 4 weeks until 28 weeks, fortnightly until 32 weeks, and weekly thereafter.

Outcomes

Review outcomes reported: caesarean section; use of additional pharmacotherapy (insulin use; dose); glycaemic control (HbA1c); hypoglycaemia; gestational weight gain (and excessive; inadequate; appropriate gain); views of intervention; large‐for‐gestational age; perineal mortality; stillbirth; neonatal death; preterm birth; macrosomia; small‐for‐gestational age; hypoglycaemia; gestational age at birth; birthweight

Notes

Funding: "This work was partially supported by the following foundation: the Social Development Project of JiangSu Province (No. SBE201170735, Wang SH)".

Declarations of interest: "The authors declare no competing financial interests".

Dates: September 2011 to December 2012

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Low risk

Quote: "randomly allocated … by a computer generated random number table".

Allocation concealment (selection bias)

Unclear risk

No details provided

Blinding of participants and personnel (performance bias)
All outcomes

High risk

Not feasible to blind participants and personnel.

Blinding of outcome assessment (detection bias)
All outcomes

Unclear risk

No detail provided; unclear how lack of blinding would have affected outcomes.

Incomplete outcome data (attrition bias)
All outcomes

Low risk

120 women were randomised, and 106 were analysed; 58 were randomised to CGMS and 62 to self‐monitoring; overall 51/58 were analysed in the CGMS group and 55/62 in the self‐monitoring group. In the CGMS group, 2 withdrew, 1 had site discomfort, 3 were lost, and 1 discontinued; in the self‐monitoring group, 4 were lost, and 3 discontinued.

Selective reporting (reporting bias)

Unclear risk

No access to trial protocol to enable confident assessment of selective reporting.

Other bias

Low risk

No other obvious sources of bias identified.