Malone 1989.

Methods	RCT
Participants	227 people with diabetes mellitus and foot infection, ulceration or prior amputation ‐ randomised. 203 people included: intervention 103 vs control 100. Baseline risk for foot ulceration: although described as 'not significant', prior vascular reconstruction higher in control and incidence of foot callus higher in intervention (P < 0.05). No significant differences in foot deformities, neuropathy, gangrene, prior amputation or ulcer and level of distal pulses Study setting: secondary outpatient care, podiatric or vascular surgery care in the USA Inclusion criteria: people with diabetes (unclear which type) with foot infection, ulceration or prior amputation referred for podiatry or vascular surgery
Interventions	Intervention group: 1‐hour group patient education with slides given by podiatrist and set of patient instructions. Content: slides of infected diabetic feet and amputated diabetic limbs, simple set of patient instructions for diabetic foot care Routine patient education. Content: routine diabetic teaching on diet, weight, exercise and medication Control group: Routine patient education. Content: routine diabetic teaching on diet, weight, exercise and medication Adherence: no data provided
Outcomes	Primary outcomes: ulcer incidence, incidence of infections, amputation rate Secondary outcomes: none
Duration and completion of follow‐up	Intervention mean 12 months, median 13.2 months (range 1 to 26 months) vs control mean 8 months, median 9.2 months (range 1 to 26 months); 182 people completed follow‐up: intervention 90 vs control 92
Types of assessment	No information provided
Notes	Unit of randomisation: individual people. Unit of analyses: separate limbs
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	High risk	Quote: "Patients were randomised into two groups based upon the odd or even last digit of their Social Security number"
Allocation concealment (selection bias)	High risk	Sequence generation was based upon the last digit of the persons social security number
Blinding (performance bias and detection bias) Blinding of outcome assessors	Unclear risk	No information provided
Incomplete outcome data (attrition bias) All outcomes	High risk	182 of 227 people completed follow‐up (intervention 90 vs control 92). Reasons for dropping out were not fitting the eligibility criteria (n = 24), death (n = 13) and unspecified (n = 8) No ITT analyses were undertaken
Selective reporting (reporting bias)	Low risk	No study protocol available, but the trial report listed the outcomes of interest in both the methods and the results section
Other bias	Unclear risk	Baseline risk for foot ulceration: although described as 'not significant', prior vascular reconstruction higher in control and incidence of foot callus higher in intervention (P < 0.05). No significant differences in foot deformities, neuropathy, gangrene, prior amputation or ulcer and level of distal pulses Co‐interventions were not described Adherence: no data provided