[Summary text]
HbA1c >= 1 år - kritisk
QoL – længste follow-up
Følgende outcomes er vurderet vigtige:
HbA1c< 1 år
Komplikationer >= 1 år
Hjertekarsygdom >= 1 år
Selvrapporteret helbred - længste follow-up
Self-efficacy - længste follow-up
Depression - længste follow-up
Frafaldsrate - efter endt forløb
Study design: Randomized controlled trial
Study grouping: Parallel group
Open Label:
Cluster RCT: YES
Baseline Characteristics
intervention
control
Included criteria: The inclusion criteria for study subjects were: enrolledin the diabetes clinic and have a diagnosis for more than 3months with type 2 diabetes; more than 18 years of age;able to consent to participate on their own behalf; have noobvious confusion or psychiatric illness; and able to speak,read and write Chinese.
Excluded criteria: Individuals were excluded fromparticipating if they demonstrated: difficulty communicatingin Chinese or were too ill to participate because of aterminal illness or hemodialysis.
Intervention Characteristics
intervention
control
Continuous:
Dichotomous:
Adverse Events:
Sponsorship source: This research was supported by grants fromthe Pingtung Christian Hospital (PS-97002).
Country: Taiwan
Setting:
Comments:
Authors name: Shu Ming Chen
Institution: School of Nursing, Fooyin University, Kaohsiung, Taiwan
Email: ft036@mail.fy.edu.tw
Address: Faculty of nursing, Fooyin University, 151 Chin-Hsueh Rd., Ta-Liao, Hsiang, Kaohsiung, Hsion 83102, Taiwan, ROC.
Identification:
Participants:
Study design:
Baseline characteristics:
Intervention characteristics:
Pretreatment:
Solveig Jansen The participants in both group had the same chronic oral precriptions during this 3 month period
Continuous outcomes:
Solveig Jansen There were no significant differences between the withdrawal participants and those who continued with respect to baseline characteristics, HbA1c, self-manegement, and psychological outcomes. total randomized 250, 125 intervention and 125 control group. 214 people completed the outcome data collection. 104 (83%) intervention and 110 (85%) control group.
Elisabeth Ginnerup-Nielsen Euroqol er i dette tilfælde "WHOQOL"Citat: Quality of life increased steadily in the experimentalgroup (t = -4.49, p < 0.01) but not in control group(t = -0.93, p = 0.35) from baseline to 3 month follow-up.Because the interaction of group and baseline quality of lifewas significant (F = 6.49, p = 0.012),The diabetes management self-efficacy increased steadilyin the experimental group (t = -6.40, p < 0.01) but notin control group (t = -1.94, p = 0.054) from baseline to3 month follow-up. Because the interaction of group andbaseline self-efficacy was significant (F = 6.76, p = 0.010),Johnson–Neyman methods rather than ANCOVA was usedto explore the intervention effect on diabetes managementself-efficacy. The result showed the motivational interviewdid improve participants significantly in diabetes managementself-efficacy among diabetes people with baselinevalue less than 174.57 (possible score range: 0–200)
Dichotomous outcomes:
Adverse outcomes:
Study design: Randomized controlled trial
Study grouping: Parallel group
Open Label:
Cluster RCT: YES
Baseline Characteristics
intervention
control
Included criteria: Not described. Randomization at GP level
Excluded criteria: Not described. Randomization at GP level
Intervention Characteristics
intervention
control
Continuous:
Dichotomous:
Sponsorship source: The Belgian Natinal Institute for Health and Disability Insurance (NIHDI)
Country: Belgium
Setting:
Comments:
Authors name: Geert Goderis
Institution: Department of General Practice, Katholieke Universiteit Leuven, Belgium
Email: geert.goderis@skynet.be, geert.goderis@med.kuleuven.be
Address: Katholieke Universiteit Leuven, Department of General Practice (Academisch Centrum voor Huisartsgeneeskunde),Kapucijnenvoer 33/J Bus 7001, 3000 Leuven, Belgium
Identification:
Participants:
Study design:
Baseline characteristics:
Intervention characteristics:
Pretreatment:
Continuous outcomes:
Dichotomous outcomes:
Adverse outcomes:
Solveig Jansen Number of events are not directly mention. Reason for patient lost to follow up are for example: died, moving to a nursing home.
Study design: Randomized controlled trial
Study grouping: Parallel group
Open Label:
Cluster RCT:
Baseline Characteristics
intervention
control
Included criteria: type 2 diabetes 50-90 yrs, HbA1c >= 7.5% prior to study
Excluded criteria: dementia, creatinin above 200 micromol/l
Intervention Characteristics
intervention
control
Continuous:
Dichotomous:
Sponsorship source: EPIC study sponsors: Agency for Healthcare and quality, Centres for research and education on therapeutics, Clinical Scientist Development Award from the Doris Duke Charitable Foundation, National Institute of Aging and the Houston Health Service Research and DEvelopment Center of Excellence
Country: USA
Setting: Medical Center
Comments:
Authors name: Naik
Institution: Houston Health and Development Center of Excellence
Email: anaik@bcm.edu
Address: 2002 Holcombe, Houston, TX 77030, USA
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
control
Included criteria: type 2 diabetes, age 30-70, HbA1c >= 7.5%
Excluded criteria: major complications, pregnancy, psychiatric disorders, mental retardation or other problem affecting the completion of questionnaires
Intervention Characteristics
intervention
control
Continuous:
Dichotomous:
Sponsorship source: none indicated
Country: USA
Setting: Baystate Medical Center
Comments:
Authors name: Garry Welch
Institution: Department of Behavioreal Medicine Research, Baystate Medical Center.
Email: garry.welch@baystatehealth.org, garry.welch@bbs.org
Address: Baystate Medical Center., 140 High Street, Room 2104, Springfield,MA 01105, United States
Identification:
Participants:
Study design:
Baseline characteristics:
Intervention characteristics:
Pretreatment:
Continuous outcomes:
Dichotomous outcomes:
Adverse outcomes:
Wrong patient population
method paper
Wrong comparator
Wrong comparator
Wrong comparator
Wrong comparator
Wrong comparator
Wrong patient population
Wrong patient population
Selection bias (biased allocation to interventions) due to inadequate generation of a randomised sequence
Quote: "125 were randomly assigned to the experimental group, and 125 were randomized to the control group."
Comment: unclear how random
Quote: "Subjects were com- puter randomized lO one of three inter- vention groups."
Quote: "After the initial recruitment phase, a researcher blinded to the study design used computer-generated numbers to randomly assign practices."
Quote: "participants were then randomly assigned (using a random numbers table)"
Comment: block randomization of 10
Comment: randomized, method not indicated
Comment: method not indicated
Comment: randomization done within races, method not indicated
Selection bias (biased allocation to interventions) due to inadequate concealment of allocations prior to assignment
Comment: Not described
Not described
Comment: Not described
Quote: "assessment, participants were then randomly assigned (using a random numbers table) to one"
Comment: difficult to conceal allocation with a table. Might have been done though
Comment: assignment was blinded using sequencially numbered and Sealed evelopes
Comment: Not described
Comment: not indicated
Quote: "using a sequentially num- bered, closed-envelope procedure."
Performance bias due to knowledge of the allocated interventions by participants and personnel during the study
Comment: neither the nurses, researcher nor the patients could be blinded, due to intervention assigmentNo blinding and selfreported outcome
Comment: Objective outcomesnot indicated, probably not blinded
Comment: HbA1c not likely to be changed by this
Quote: "All patients were blinded to the study design, but physicians were not, as they were involved in the execution of the programs."
Comment: Probably not but for oubjective outcome not relevant
Comment: Selfreported outcome - no blinding. Only affects the important outcome though
Comment: unblinded but objective outcomes
Comment: Neither physicians nor patient could be blinded to the intervention assigmment
Comment: Probably not but outcome is objective
Detection bias due to knowledge of the allocated interventions by outcome assessors
Quote: "The research assistant involved in the recruitment, baseline and outcomes measures data collection was not informed of a participant’s group allocation. To ensure that the research protocol was followed precisely a research assistant with a clinical background was employed for baseline and post intervention data collection. Neither the research assistant nor the clinical diabetes educators knew the group allocations of participants."
Comment: Selfreported and objective outcome not likely to be changed by blinding
Not so relevant for outcome
Comment: objective outcomes.
Quote: "All patients were blinded to the study design, but physicians were not, as they were involved in the execution of the programs."
Quote: "Psychosocial assessments and a blood draw for HbA 1 c were administered during the lst hr (and again at 3-month follow-up) by clinic and research personnel blind to group assignment."
Comment: outcome either selfreported or objective
Comment: assessment were conducted by trained technicians blind to Group assignment
Comment: biochemical parametres was maesured in random samples, mediators of possible change in HbA1c were self-reported
Comment: Outcome assesors not involved in patients care
Attrition bias due to amount, nature or handling of incomplete outcome data
Quote: "The participants in both group had the same chronic oral prescriptions during this 3 month period. There were no significant differences between the withdrawal parti- cipants and those who continued with respect to baseline characteristics, HbA1c, self-management, and psychologi- cal outcomes."
Comment: 8/26 og 11/28 dropped out. No intention to treat
Comment: A little skewed dropout, but not too large.. intention to treat analysis has been done
Complete blood data were therefore available for 73 of the participants (90%). The main outcome analysis was done on an intent-to-treat basis, and missing follow-up values were assumed not to have changed."
Comment: attrition was described, equal drop-out rates was seen
Comment: Relatively large dropout and unclear why only 16 and not 22 were used. Originally 22 randomized..
Comment: large but equal drop-out rate (32.8-43.1%). Mean change in HbA1c was not significantly different comparing completers and drop-outs
Comment: intervention group was significantly more compliant to group sessions16/108 and 6/109. A little skewed dropout. Intention to treat should have been done!
Reporting bias due to selective outcome reporting
Comment: No trial protocol. Outcomes are unclearly reported.
Comment: outcome probably assessed but no trial protocol
Quote: "registered as # NTR1369,"
Comment: Outcomes from protocol reported
No trial protocol
Comment: this is discussed
Comment: no trial protocol but also only a protocol study. No reason to be selective in reporting.
Comment: few studies in the litteratur show different results, selective outcome reporting therefore nor likely
Quote: "NCT00007800, clinicaltrials.gov."
Comment: No outcomes provided in protocol?
Bias due to problems not covered elsewhere in the table
Quote: "A total of 142 physicians from 90 practices agreed to participate and were randomized (Fig. 1). Within the first 30 days, 22 physicians dropped out (UQIP = 18, AQIP = 4, p &lt; 0.001). Thus, only 120 physicians (36% of those available, 67 AQIP vs. 53 UQIP) registered baseline data. During the study"
Quote: "There were several weaknesses of the present study. First, considerable effort was required to motivate GPs for the project, resulting in a global enhancement of diabetes awareness throughout the region. Secondly, despite the motivation of the GPs, we observed a significant difference in initial drop-out rate between the UQIP and AQIP-assigned GPs, which may indicate that being assigned to AQIP motivated more physicians to pay closer attention to T2DM care. However, probably the most motivated GPs remained in the UQIP arm, with possible effects on the final outcomes. Thirdly, some of the additional interventions of the AQIP arm were only used by a small number of participants. This is particularly true for the health psychologist counseling, which was only used by 18 patients. Finally, the follow-up period may have been too short to find a significant difference between AQIP and UQIP. Therefore, it will be interesting to organize a post-intervention follow-up of the included patients."
Comment: older male veterans. No oother sources of bias
Comment: different rate of excercise, recorded calories and glucose monitoring
Comment: Different educational status at baseline was discussed as possible bias
Comment: Selected group (overweight women), but this will affect indirectness.