Table 1 Quality appraisal of included qualitative papers and associated quantitative papers

Methods

Convenience sample: All patients with T2DM participating in an RCT (n = 89; 53 in the intervention arm, 36 in the control group)

Purposive sample: 30 patients with newly diagnosed T2DM participating in an RCT. Sampled to represent trial arm, recruitment site, and gender.

Purposive sample: 25 participants representing 4 groups depending on HbA1c control: ‘good’ (<7) or ‘poor’ (>9) (n = 13); or ‘improving’ or ‘deteriorating’ (n = 12)

Convenience sample: 40 patients with newly diagnosed T2DM.

Sample

Diabetes duration at baseline (mean)

6 years

6 months

6 years

6 months

Data collection

Cycles of semi-structured focus groups with all trial participants, pre- (n = 5) and post-intervention (n = 5). Group A at 6 and 12 months; Group B at 6 and 18 months; Group C at 12 months

Face to face interviews at 6 months (n = 30), and follow -up telephone interviews at 9 months (n = 29).( Trial paper [29] has 12mth data)

Semi structured face-to-face interviews at12 months (n = 25) ,and 24 months (n = 11).9 matched consultation sessions and telephone interviews at 36 months.

Semi structured face-to face interviews and fieldnotes at 0, 6, and 12 months (n = 40) and 48 months (n = 20).

Analysis

Constant comparison. Source, method and theoretical triangulation.

Constant comparison. Thematic analysis.

Constant comparison. Thematic analysis. Construction of extended case reports over time.

Grounded theory [38, 39]; Thematic analysis [40, 45]; Longitudinal [11];

Setting

Primary and secondary care, England

Primary care, South West England

Primary care, deprived area in North West England.

Primary and secondary care, Scotland

Trial design

89 patients, randomised control wait list design Group A were randomly allocated to the treatment initially (n = 30), whilst the Group B acted as the short-term control group (n = 23) These two groups were then combined to form the short term trial group.

593 patients randomly assigned in a 2:5:5 ratio. Control n = 99, Intensive Diet n = 249, Diet plus Activity n = 246

591 patients randomly allocated in a1:2 ratio. Control n = 197, Call-centre treatment support n = 394.

N/A. 40 patients with T2DM. Explorations of variance, location of care (12.5% primary care, 87.5% secondary care), diet, medication, class and gender.

Group C received the intervention at the end of the trial period (n = 36).

Patients randomized to ‘usual care’ received standard advice about diet from trial dieticians at their baseline visit, and were seen by a doctor blinded to treatment at baseline, six and twelve months [28].

Patients randomized to the ‘usual care’ group continued with conventional treatment based on local guidelines, which had been in place for over ten years, supported by a continuing education program among all primary care practices [34].

Intervention

8 week educational programme including: physical activity, exercise, relaxation and health topics.

Intensive diet (ID) or intensive diet plus activity (IDPA)

Tele-care phone support, titrated to HbA1c, to improve blood glucose control

Not applicable

Trial results

At 6 months, intervention associated with benefits in HbA1c levels (−0.1%), illness attitudes, and perceived treatment effectiveness, compared to controls. At 12 months, only illness attitudes and self-monitoring showed benefit [27].

At 6 and 12 months, glycaemic control had improved in the diet (−0.28%) and diet/activity groups (−0.33%), but worsened in the control group [29]

At 12 months, compared with the control group, HbA1c improved by 0.31% in the intervention group, and the improvement was significantly greater for those with a baseline HbA1c > 7% [30].

Not applicable

At 12 months, the control group saw an improvement in their understanding, expectation of disease continuation, and concern of their illness; while the intervention groups increased their understanding, became less concerned, felt more in control of their illness, were more satisfied with their diabetes treatment, and had higher self-reported health scores [30].

At 12 months, the intervention group continued to report high levels of satisfaction with their treatment [34]

At 36 months, there was a statistically significant reduction of HbA1c by 0.24% attributable to the intervention [36].