Mody 2008.

Methods	2‐arm RCT undertaken in India
Participants	48 participants (recruited from inpatient wards), 15 of whom were reported to have DM and a foot ulcer. Data for these 15 participants only were presented Inclusion criteria: people admitted to general surgery, physical medicine and rehabilitation wards and referred by the surgical consultants for care of an acute or chronic extremity, sacral or abdominal wound that could not be treated with primary closure Exclusion criteria: ischaemic wounds; or wounds: in anatomical locations where an adequate seal around the wound site could not be obtained; with exposed bowel or blood vessels; with necrotic tissue that could not be debrided; with communicating fistulae; with malignancy; with recent grafts; or presence of osteomyelitis; or receiving therapeutic anticoagulation Key baselines covariates (foot ulcers in people with DM only): Wound area (cm²): Group A: 25.7 (SD 9.7) Group B: 48.1 (SD 53.5) Wound duration (days): Group A: 8.5 (SD 8.3) Group B: 5.2 (SD 2.3)
Interventions	Group A (n = 6): locally constructed (homemade) device: a sterilised, porous packing material obtained from a local source was cut to fit the wound. A 14‐French suction catheter was tunnelled into the packing material, which then was placed into the wound cavity. A sterile adhesive plastic drape (Dermincise, Vygon, UK) was cut to overlap the surrounding skin and applied over the packing material, forming an airtight seal. Tubing was used to attach the free end of the suction catheter to a wall suction canister. The TNP timer was placed in circuit between the wall suction apparatus and the wall suction canister The TNP timer, constructed from local electronics, was designed to cycle wall suction intermittently using a simple timed switch and a system of valves. For the study protocol, the timer was set to cycle for 2 minutes on, followed by 5 minutes off. Wall suction pressure was set at 125 mmHg. In sensitive wounds, suction was reduced to a tolerable level (usually 50–100 mmHg) until it could be comfortably increased. For oedematous wounds, the suction was kept on a continuous setting until oedema had been reduced and an intermittent regimen could be followed. The dressing was changed every 2 days unless otherwise scheduled by the treating physician. Wounds were debrided as required to keep the wound bed free of necrotic tissue. Participants receiving NPWT who no longer required hospitalisations for their primary diagnosis, or could not afford to remain in the hospital, remained in the study with conventional wound dressings in the outpatient setting, but outcomes were analysed in the original treatment groups. Group B (n = 9): saline‐soaked gauze and dry pads used to cover the wound. Dressing changes typically performed twice daily; frequency adjusted according to the judgement of the treating physician. Wounds in both treatment groups were debrided before dressing application.
Outcomes	Primary review outcomes: number of wounds completely healed (satisfactory healing defined as complete wound closure by secondary intention or wound readiness for delayed primary closure as determined by the study investigator and treating surgeon) Secondary review outcomes: number of wounds closed or covered with surgery
Notes	Participants were followed until wound closure or being lost to follow‐up for a mean of 26.3 days (SD 18.5) in the control and 33.1 days (SD 37.3) in the treatment group.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Quote: "wounds that met inclusion and exclusion criteria were assessed for size (in a manner that allowed blinding) and then block‐randomized using a concealed computer‐generated table in a 1‐to‐2 ratio of TNP closure versus conventional wound dressing." Comment: adequate method
Allocation concealment (selection bias)	Unclear risk	Quote: "following enrolment, wound size was assessed using computer‐aided measurements of digital photographs and block‐randomized to the study arms using a concealed allocation table." Comment: unclear how allocation concealment was conducted
Blinding of participants and personnel (performance bias) All outcomes	Unclear risk	Not reported
Blinding of outcome assessment (detection bias) All outcomes	Unclear risk	Not reported
Incomplete outcome data (attrition bias) All outcomes	Low risk	Seems that participants were analysed in groups as randomised
Other bias	Low risk	No evidence of other risk of bias