[Summary text]
HbA1c >= 1 år - kritisk
Livskvalitet (QoL) – længste follow-up - kritisk (NB! SF-36 scores for fysisk score og mental score)
Følgende secondære outcomes er vurderet vigtige
BMI >= 1 år
Vægt >= 1 år
HbA1c< 1 år
LDL =< 1 år
Antidiabetisk medicin (tabletbehandling (antal); Insulin (antal))
Antidiabetisk medicin < 1 år (antal præparater, dosis)
komplikationer >= 1 år
Hjertekarsygdom >= 1 år
Frafaldsrate - efter endt forløb
Study design: Randomized controlled trial
Study grouping: Parallel group
Open Label:
Cluster RCT:
Baseline Characteristics
intervention
usual care
Included criteria: receiving dietary or oral anti-diabetes treatment; <=75 years of age; HbA1c value ranging from 6.5 to 10%; diabetes duration of at least 1 year;considered by the physician and the diabetes specialist nursesat each primary care centre to be able to participate in aneducation group together with other people with diabetes; andable to understand the Swedish language.
Excluded criteria: known alcohol abuse; known mental disability; presenceof serious disease (stroke, late stage of cancer); patientswho had previously participated in group education
Intervention Characteristics
intervention
usual care
Continuous:
Dichotomous:
Adverse Events:
Sponsorship source: This study was supported by grants from the SwedishDiabetes Association.
Country: Sweden
Setting:
Comments:
Authors name: Eva Thors Adolfsson
Institution: Department of Medical Sciences, Faculty of Medicine, Uppsala University, Uppsala, Sweden
Email: eva.thors-adolfsson@ltv.se
Address: Herrgärdets Vårrdcentral, Karlsgatan 17A, SE-722 14 Västerås, Sweden
Identification:
Participants:
Study design:
Baseline characteristics:
Intervention characteristics:
Pretreatment:
Continuous outcomes:
Tina Povlsen QOL (Satisfaction with daily life, difference). This is measured in inter quartile range: Intervention group (n = 42): 2.5 (3.9, 7.5)Control group (n = 46): 0.0 (6.0, 7.5)
Dichotomous outcomes:
Adverse outcomes:
Study design: Randomized controlled trial
Study grouping: Parallel group
Open Label:
Cluster RCT:
Baseline Characteristics
intervention
usual care
Included criteria: All patients had type 2 diabetesdiagnosed for at least 1 year. Theywere undergoing at least annual medical checks related totheir diabetes.
Excluded criteria: Persistent defaulters, or those with significantalcohol or drug addiction problems were excluded, as werethose with significant disability which would interfere withthe educational process (e.g. blindness or deafness).
Intervention Characteristics
intervention
usual care
Continuous:
Dichotomous:
Adverse Events:
Sponsorship source: We are grateful to Diabetes UK who funded this research
Country: UK
Setting:
Comments:
Authors name: Helen Cooper
Institution: Faculty of Health & Social Care, Department of Community & Child Health, University of Chester,Chester United Kingdom
Email: g.gill@liv.ac.uk
Address:
Identification:
Participants:
Study design:
Baseline characteristics:
Intervention characteristics:
Pretreatment:
Continuous outcomes:
Elisabeth Ginnerup-Nielsen Der er opgivet BMI i studiet men den eneste usikkerhed i forbindelse med værdien er et "NS"
Dichotomous outcomes:
Adverse outcomes:
Study design:
Study grouping: Parallel group
Open Label:
Cluster RCT:
Baseline Characteristics
intervention
usual care
Included criteria:
Excluded criteria:
Intervention Characteristics
intervention
usual care
Continuous:
Dichotomous:
Adverse Events:
Sponsorship source:
Country: UK
Setting:
Comments:
Authors name:
Institution:
Email: trudi.deakin@nhs.net
Address:
Identification:
Participants:
Study design:
Baseline characteristics:
Intervention characteristics:
Pretreatment:
Continuous outcomes:
Mette Sonne QoL data from study noted as QoL VAS in table
Dichotomous outcomes:
Ole Snorgaard 16% of patient in intervention reduced medication compared to 1% in controls. 21% increased medication in intervention against 46 in control
Mette Sonne Only medicine increase/decrease or if unchanged is mentioned. No information regarding what type of medicaion.
Adverse outcomes:
Study design:
Study grouping: Parallel group
Open Label:
Cluster RCT:
Baseline Characteristics
intervention
usual care
Included criteria: patients with diabetes, age18 or over, were identified by ICD 9 encounter codes(two or more visits for diabetes within the last year).
Excluded criteria: Patients unable to speak English and residents ofnursing homes were not eligible for participation inthe study
Intervention Characteristics
intervention
usual care
Continuous:
Dichotomous:
Adverse Events:
Sponsorship source: Not described
Country: US
Setting:
Comments:
Authors name: Robert A. Gabbay
Institution: The Penn State College of Medicine, Penn State Diabetes Center, Department of Endocrinology, Diabetes and Metabolism, 500 University Drive, HO44, Hershey, PA 17033, USA
Email: rgabbay@psu.edu
Address: Endocrinology Department, Reading Hospital, Reading, 19601, PA, USA
Identification:
Participants:
Study design:
Baseline characteristics:
Intervention characteristics:
Pretreatment:
Continuous outcomes:
Tina Povlsen Weigt is measured in pund:Intervention:6 mounth: 205 (47) 12 mount 207 (47)Control: 6 mounth: 201 (46) 12 mount 202 (47)
Elisabeth Ginnerup-Nielsen Vægt (LBS) changed to kg LDL mg/dl changed to mmol/l via http://www.endmemo.com/medical/unitconvert/Low-density_lipoprotein_cholesterol.php
Dichotomous outcomes:
Adverse outcomes:
Study design: Randomized controlled trial
Study grouping: Parallel group
Open Label:
Cluster RCT:
Baseline Characteristics
intervention
usual care
Included criteria: patients under age 76 years and with HbA1c≥7.0%, oral medication was optimized [8]. After this optimization,76 patients had HbA1c≥7.0% while taking the maximumfeasible dosages of two different oral hypoglycaemic agents,mostly sulphonylurea and metformin. These patients were eligiblefor the present study.
Excluded criteria: severecomorbidity (defined as having an illness that surpasses the impactof diabetes); insufficient understanding of spoken Dutch tofollow instructions; or requirement for insulin therapy in theshort term on account of severe hyperglycaemic symptoms.
Intervention Characteristics
intervention
usual care
Continuous:
Dichotomous:
Adverse Events:
Sponsorship source: This study was supportedby an unrestricted research grant from Novo NordiskPharma.
Country: The Netherlands
Setting:
Comments:
Authors name: Alex N. Goudswaard
Institution: Julius Centre for Health Sciences and Primary Care
Email: lex@goudswaard.cx
Address: Julius Centre for Health Sciences andPrimary Care, PO Box 80560, 3508 AB Utrecht, the Netherlands.
Identification:
Participants:
Study design:
Baseline characteristics:
Intervention characteristics:
Pretreatment:
Continuous outcomes:
Elisabeth Ginnerup-Nielsen > 1 år svarer her til 18 mdr.
Dichotomous outcomes:
Elisabeth Ginnerup-Nielsen
Adverse outcomes:
Study design: Randomized controlled trial
Study grouping: Parallel group
Open Label: YES
Cluster RCT:
Baseline Characteristics
intervention
usual care
Included criteria: BMI >27 kg m-2) 30-75 years old. Random blood glucose over 11 mmol 1 -1 in symptomatic patients or a blood glucose over 11mmol 1-1 2 h after a 75 g oral glucose load in those who were asymptomatic.
Excluded criteria: Ketonuria, Diabetes diagnosis made as an inpatient.Use of any type of glucose lowering drug
Intervention Characteristics
intervention
usual care
Continuous:
Dichotomous:
Adverse Events:
Sponsorship source: Boehringer Corporation, Lewes, UK and British Diabetic Association
Country: UK
Setting: Hospital, out patient clinic
Comments:
Authors name: S. R. Heller, P. Clarke, H. and Daly, I. et. al.
Institution: Diabeles Unit, Depart menl ol Medicine, Universily Hospilal, Nollingham, UK
Email:
Address: Heller, Diabetes Unit, Department of Medicine, C Floor, South Block, University Hospital, Queen's Medical Centre,Nottingham, NG7 2UH, UK.
Identification:
Participants:
Study design:
Baseline characteristics:
Intervention characteristics:
Pretreatment:
Continuous outcomes:
Tina Povlsen HbA1 was rapported in median and inter quartile range.Intervention:6 mounth: 7.5 (7.0-8.1) 12 mounth: 9.0(8.2-9.8)n: 36Control:6 mounth: 9.5 (8.7-10.4) 12 mounth: 9.9(8.9-10.9)n: 39
Dichotomous outcomes:
Mette Sonne The groups studied were on diet alone, no glucose lowering drugs allowed
Adverse outcomes:
Study design: Randomized controlled trial
Study grouping: Parallel group
Open Label:
Cluster RCT: YES
Baseline Characteristics
intervention
usual care
Included criteria: All of their patients (aged 40–80 years) who, according to computerized patientrecords, had been diagnosed with type 2 diabetesduring the previous 2 years were identified (n = 257).
Excluded criteria: a psychiatric diagnosis, dementia, alcoholism or otherdrug addiction; a severe somatic or disabling diagnosis(e.g. final-stage cancer, reduced perception such asdeafness or blindness); and inability to speak Swedish.
Intervention Characteristics
intervention
usual care
Continuous:
Dichotomous:
Adverse Events:
Sponsorship source: This study was funded by the Swedish DiabetesAssociation, the County Council of Va¨sterbotten, andthe Medical Faculty of Umea° University, Umea°,Sweden.
Country: Sweden
Setting:
Comments:
Authors name: Åsa Hörrnsten
Institution: Department of Nursing, Umeå University, Umeå 90187, Sweden
Email: asa.hornsten@nurs.umu.se
Address:
Identification:
Participants:
Study design:
Baseline characteristics:
Intervention characteristics:
Pretreatment:
Continuous outcomes:
Elisabeth Ginnerup-Nielsen longst f.u. 5 years
Tina Povlsen Longest follow-up is 5 years
Dichotomous outcomes:
Elisabeth Ginnerup-Nielsen Frafald er frafald ved længste follow - up (5 years)
Adverse outcomes:
Study design: Randomized controlled trial
Study grouping: Parallel group
Open Label:
Cluster RCT:
Baseline Characteristics
Structured health education programme by a trained diabetic education nurse. The nurse concentrated on the major CVD risk factors
Usual medical care
Included criteria: The inclusion criterion included poor glycaemic control[glycated haemoglobin (HbA1c)≥8–11%] and age range 35 to70 years. Patients were recruited irrespective of their anti-diabetic,anti-hypertensive or lipid-lowering medication regimens.
Excluded criteria: Not described
Intervention Characteristics
Structured health education programme by a trained diabetic education nurse. The nurse concentrated on the major CVD risk factors
Usual medical care
Continuous:
Dichotomous:
Adverse Events:
Sponsorship source: Not described
Country: Kina
Setting:
Comments:
Authors name: G. T. C. Ko
Institution: Department of Medicine, AH Nethersole Hospital, Hong KOng
Email: gtc_ko@hotmail.com
Address: Gary T. C. Ko, Department of Medicine, AH Nethersole Hospital, 11, Chuen On Road, Tai Po, NT, Hong Kong
Identification:
Participants:
Study design:
Baseline characteristics:
Intervention characteristics:
Pretreatment:
Continuous outcomes:
Dichotomous outcomes:
Adverse outcomes:
Study design: Randomized controlled trial
Study grouping: Parallel group
Open Label:
Cluster RCT: YES
Baseline Characteristics
intervention
usual care
Included criteria: Diabetes conformend by lab data (high fasting glucose, high random glucose, HbA1c > 7%) or the use of diabetes medications
Excluded criteria: Not mentioned
Intervention Characteristics
intervention
usual care
Continuous:
Dichotomous:
Adverse Events:
Sponsorship source:
Country: US
Setting: general, family, and internal medicine practices
Comments:
Authors name: Piatt, Trevor and Emerson et.al.
Institution: Department of Epidemiology and Diabetes Center and Medical Center,University of Pittsburgh, Pittsburg,
Email: piattg@upmc.edu
Address: Kaufmann Building, Suite 601, 3471 Fifth Ave., Pittsburgh, PA 15213.
Identification:
Participants:
Study design:
Baseline characteristics:
Intervention characteristics:
Pretreatment:
Continuous outcomes:
Mette Sonne Diabetes empowerment scale score noted as SF-36 mental health
Dichotomous outcomes:
Adverse outcomes:
Study design: Randomized controlled trial
Study grouping: Parallel group
Open Label:
Cluster RCT:
Baseline Characteristics
intervention
usual care
Included criteria: Two inclusion criteria for registeringpersons applying for participation were used: participantshad to be diagnosed with Type 2 diabetes and, if treatedwith insulin, only for 2 years or less.
Excluded criteria: The exclusion of persons with long-term insulin treatmentwas determined based on reports from the study circleleaders who felt that dietary and exercise interventions didnot lead to immediately demonstrable effects for this groupof participants.
Intervention Characteristics
intervention
usual care
Continuous:
Dichotomous:
Adverse Events:
Sponsorship source: This study was supported by the Swedish Foundationfor Health-care Sciences and Allergy Research Grant No.V2000 225, the National Corporation of Swedish Pharmacies,and Uppsala University. Funding for the first author,Anna Sarkadi, have been thankfully received from the Knutand Alice Wallenberg Foundation in Stockholm, Sweden,grant nr. KAW 2001.0303.
Country: Sweden
Setting:
Comments:
Authors name: Anna Sarkadi
Institution: Department of Public Health and Caring Sciences, Uppsala University, Uppsala Science Park, Uppsala SE-751 85, Sweden
Email: Anna.Sarkadi@pubcare.uu.se
Address:
Identification:
Participants:
Study design:
Baseline characteristics:
Intervention characteristics:
Pretreatment:
Continuous outcomes:
Dichotomous outcomes:
Adverse outcomes:
Study design: Randomized controlled trial
Study grouping: Parallel group
Open Label:
Cluster RCT:
Baseline Characteristics
intervention
usual care
Included criteria: We enrolled adults between ages 20 and 75 years who wereoutpatients seen atDepartment of Diabetes and Metabolism, theUniversity of Tokyo Hospital, and who were diagnosed withtype 2 diabetes, and who had HbA1C values between 6.5% and8.5% on an average in three tests assessed within recent 3months, and who could not use insulin.
Excluded criteria: Subjects were requiredto have neither serious ongoing illness nor cognitive disorder
Intervention Characteristics
intervention
usual care
Continuous:
Dichotomous:
Adverse Events:
Sponsorship source: This study was supported by theMinistry of Health, Labour, and Welfare ScientificResearch Grants in FY2003, and by Japanese NursingAssociation in FY2004.
Country: Japan
Setting:
Comments:
Authors name: Taiga Shibayama
Institution: Institute of Nursing Science, Graduate School of Comprehensive Human Sciences, University of Tsukuba
Email: taiga@md.tsukuba.ac.jp
Address: Institute of Nursing Science, Graduate School of Comprehensive Human Sciences, University of Tsukuba,1-1-1 Tennoudai, Tsukuba-shi, Ibaraki 305-8575, Japan
Identification:
Participants:
Study design:
Baseline characteristics:
Intervention characteristics:
Pretreatment:
Continuous outcomes:
Tina Povlsen
Elisabeth Ginnerup-Nielsen 95% CI i Role Physical er ændret. I teksten står den som - 6 (13;1) den er ændret til -6 (-13; 1)
Dichotomous outcomes:
Adverse outcomes:
Wrong intervention
followup study
Wrong intervention
Selection bias (biased allocation to interventions) due to inadequate generation of a randomised sequence
Quote: "A research assistant at the Centre for Clinical Research performed the randomization in blocks of 4 (two assignments to intervention and control groups, respectively) at each of the 7 centres. Only the research assistant was aware of the blocks,"
Comment: Unclear if randomisation was done randomly
Comment: Not described
Quote: "A total of 332 patients were randomized (by method of odd and even numbers)"
Quote: "Randomization was done by a telephone call to an independent trial centre, which used a computer-generated random assignment with blocks of eight at a time."
Quote: "Eligible patients were randomized to receive usual clinic care or to attend group education classes"
Comment: unclear how random
Comment: Randomization was done by coin-tossing s. 1275. Should be fine but The greater number of men and smokers in the intervention Group may be due to `suboptimal` randomization s. 1278
Quote: "prac- tices were randomized into one of three study groups (Fig. 1). An initial block ran- domization procedure was undertaken, with practice size"
Quote: "For those participants eligible for randomisation, the in- formed consent sheet and the questionnaire were put into an unmarked envelope, one for each participant. The identical envelopes were then put into a box. Each time 20 complete sets of participant items were collected, randomisation was performed. An assistant mixed the envelopes in the box, took them out one at a time, and randomly placed them into two piles. A"
Comment: Not described
Selection bias (biased allocation to interventions) due to inadequate concealment of allocations prior to assignment
Quote: "Only the research assistant was aware of the blocks, which were contained in a set of sequentially numbered opaque sealed envelopes. These"
Comment: Not described
Comment: Not described
Quote: "Randomization was done by a telephone call to an independent trial centre,"
Comment: no mention in text
Comment: Not described
Comment: Not described
Comment: no mention of concealment of sequence
Quote: "third person, acting as a witness, pointed out which pile should be allocated to the intervention group and which"
Quote: "pile to the control group. Each participant was then assigned a code, beginning with a different letter for the intervention (I) and control (C) groups."
Comment: Not described
Performance bias due to knowledge of the allocated interventions by participants and personnel during the study
Comment: Not described. probably not done but for objective outcomes this is probably not important
Comment: Probably not blinded but outcome objective
Comment: Probably not blinded but outcomes are objective
Quote: "Moreover, since this was an unblinded study and randomization was done on a patient level, in the control group patients as well as doctors have become increasingly aware of the issue of tighter control and started acting upon it. This may have diminished the effect of the intervention."
Comment: It is noteworthy that glycaemic control in the control groupgradually improved during the study (Fig. 2). This might be aregression-to-the-mean effect [15]. Moreover, since this wasan unblinded study and randomization was done on a patientlevel, in the control group patients as well as doctors havebecome increasingly aware of the issue of tighter control andstarted acting upon it. (Interessant-normalt ville jeg sige at det ingen betydning havde for resultatet når outcomes er objektive)
Comment: no mention i the text but the subjects could possibly not be blinded.
Comment: Probably not blinded but outcome is objective
Comment: Probably No blinding of participants or personnel not but outcomes are objective
Comment: not blinded
Comment: Not described, problably not done, however outcome is not likely to be influenced by this
Comment: probably none - QOL scores subjective
Detection bias due to knowledge of the allocated interventions by outcome assessors
Comment: Not described. probably not done but for objective outcomes this is probably not important
Comment: Not describedProbably not blinded but outcome objective
Comment: Not described. probably not done but for objective outcomes this is probably not important
Quote: "Moreover, since this was an unblinded study and randomization was done on a patient level, in the control group patients as well as doctors have become increasingly aware of the issue of tighter control and started acting upon it. This may have diminished the effect of the intervention. Further"
Comment: Not described, problably no blinding, however problably no influences in outcome
Comment: Not descibed. Outcome objective
Comment: Some blindingand outcomes are objective
Quote: "Each participant’s cardiovascular status and any CVD risk factors were recorded by the educator (at the end of the study in the controls). The physicians were blind to which group the patients belonged to."
Comment: Some blinding
Comment: Not described, problably not done, however will problably not influence outcome.
Comment: Not described, problably not done, however outcome is not likely to be influenced by this
Quote: "Physicians did not know which patients served as control subjects for this study."
Comment: QOL scores selfreported and other outcomes objective
Attrition bias due to amount, nature or handling of incomplete outcome data
Comment: Intention to treat and per protocol analyses have beenperformed. Data are presented per protocol. When the resultsof intention to treat analyses differed from those of the perprotocol analyses it is presented in the text. (And relatively small dropout)
Quote: "A total of 11 failed to complete the programme (12%). This included 5 deaths, and the remaining 6 were not able to complete because of time constraints. Of the remaining, 76% attended at least 7 of the 8 educational sessions, with the commonest reason for non-attendance being illness."
Comment: Relatively small equal dropout
Comment: Dropout not described no intention to treat.
Quote: "Analyses were based on intention to treat, with the last value carried forward for missing data. Ineligible patients mistakenly randomized, and patients who withdrew before the start of the intervention, were excluded from analysis [10]."
Comment: drop outs and reason for drop out mentioned. Only completers included in the analysis. No intention to treat.
Comment: 4/44 dropouts in intervention and 1/60 in the control group. 5 years f.u. 5 and 10 respectively. Ikke så stort dropout, men de burde have lavet intention to treat
Comment: No intention to treat - apparantly no dropout but even in China this seems strange!
Comment: Numbers in each group mentioned compared with total number of randomized. Reasons for drop out mentioned. No mention of how/ or if drop outs were included in the analysis eg. intention to treat. Attrition is mentioned with a greater than 75 % attending more than 75% of the sessions.
Comment: No intention to treat analysisRelatively small dropout but itt should have been done
Reporting bias due to selective outcome reporting
Comment: No trial protocol. Relevant outcome seems assessed.
Comment: No trial protocol. Strange that QOL scores are only for the interevtnion group. Also unclearly reported!
Comment: No trial protocol
Comment: No available study protocol, however all outcomes are rapported
Comment: No trial protocol
Comment: No study proctocol available, however all outcome are presented.
Comment: No study protocol. Very hard to read the numbers in the graph. There is uncertainty in reading the. numbers
Comment: No trial protocol
Bias due to problems not covered elsewhere in the table
Comment: The study appears to be free of other sources of bias.
Quote: "This selection procedure introduced a systematic bias in the sample, pre- sumably resulting in persons motivated to improve diabetes self-management. This observation also provides a probable explanation for the rather low mean initial HbA 1c of partic- ipants, with 52% under the WHO target value of 6.5% [20], which can be compared with a proportion of ∼40% in a national sample of 10,000 persons [21]. On the other hand, randomisation occurred after the recruitment of participants so the bias, if any, was equally present for both intervention and control groups."