Baker 1996.

Methods	Design: randomised controlled study
Participants	Health condition: elderly people with spinal cord injury Sample size: Randomisation: at participant level (more than 1 ulcer for some participants)§ Randomised (participants, n; ulcers, n): 80; 185 Experimental 1: 20; not reported Experimental 2: 21; not reported Experimental 3: 20; not reported Control: 19; not reported Analysed (participants, n; ulcers, n): not reported; not reported* Experimental 1: not reported; not reported Experimental 2: not reported; not reported Experimental 3: not reported; not reported Control: not reported; not reported Setting, country: inpatient and outpatients of a Medical Centre (1 site), USA Inclusion criteria: participants with spinal cord injury medically cleared by a physician Exclusion criteria: not reported Characteristics of pressure ulcer: Experimental 1 Ulcer duration**, mean (SD): 183 days (42) Location of ulcer* (n): sacral (24), ischial tuberosity (20), thigh (10), foot (9) and others (3) Ulcer stage: not reported Experimental 2 Ulcer duration**, mean (SD; range): 231 days (38; 2‐1095) Location of ulcer* (n): sacral (19), ischial tuberosity (18), thigh (13), foot (5) and others (3) Ulcer stage: not reported Experimental 3 Ulcer duration**, mean (SD; range): 154 days (39; 5‐961) Location of ulcer* (n): sacral (10), ischial tuberosity (12), thigh (11), foot (3) and others (6) Ulcer stage: not reported Control Ulcer duration**, mean (SD; range): 86 days (24; 5‐415) Location of ulcer* (n): sacral (9), ischial tuberosity (10), thigh (4) and foot (2) Ulcer stage: not reported Mean age (SD; range): Experimental 1: 34 years (2; 19‐64) Experimental 2: 40 years (2; 21‐64) Experimental 3: 36 years (2; 17‐64) Control: 33 years (4; 19‐76) Gender: Experimental 1: 85% male Experimental 2: 76% male Experimental 3: 85% male Control: 84% male §It is not clearly stated whether participants or pressure ulcers were randomised but the number of pressure ulcers is greater than the number of participants. We have assumed that randomisation was at the participant level. *Participants in the control group were moved across to the experimental group but outcome data prior to switching groups are not provided.
Interventions	Total groups in this study: four Experimental 1: asymmetric biphasic stimulation plus standard therapy Duration: 30 minutes per session; 3 sessions per day; 5 days per week; until ulcer healed Electrode placement: both over the area around the ulcer Device, manufacturer: UltraStim, Henley International Texas USA Intensity of ES: below a visible minimum muscle contraction Frequency: 50 Hz Type of current: pulsed Polarity: negative electrode proximal to the ulcer; positive electrode distal to the ulcer Experimental 2: symmetric biphasic stimulation plus standard therapy Same settings and duration as Experimental 1 but phase duration was set as 300 μs Experimental 3: microcurrent stimulation plus standard therapy Same settings and duration as Experimental 1 except phase duration and frequency was set at 10 μs and 1 Hz, respectively. Control: sham microcurrent stimulation plus standard therapy Same settings and duration as Experimental 3 but without current and administered for 28 days or until the ulcer was healed. Standard therapy for all groups This included the sulfadiazine cream, occlusive dressing, wet‐to‐dry dressing with saline solution and/or dry dressing.
Outcomes	Outcomes included in this review: none Not useable data**: presented as [name of outcome in review]‐ [name of outcome in study] Proportion of pressure ulcers healed ‐ pressure ulcers healed (expressed as numbers) Surface area of pressure ulcers ‐ wound area (expressed as cm2) Rate of pressure ulcer healing ‐ healing rates (expressed as percentage per week) Timing of outcome measures: until healed
Notes	Withdrawals, (n; reason)**: Experimental 1: 7 ulcers; reasons not specified Experimental 2: 8 ulcers; reasons not specified Experimental 3: 6 ulcers; reasons not specified Control: 6 ulcers; reasons not specified Funding source: supported by the National Institute on Disability Research and Rehabilitation, Department of Education, USA Trial registration or published protocol: no information about trial registration or published protocol provided **These data are not reliable as participants from the control group were reassigned to Experimental 1 or Experimental 2 after 28 days leading to duplication of data.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Quote: "Patients were randomly assigned..." p22 Comment: insufficient detail reported
Allocation concealment (selection bias)	Unclear risk	Comment: insufficient detail reported
Blinding of participants (performance bias) All outcomes	Low risk	Quote: "The control group received the same stimulation procedures as the MC [microcurrent] treatment group, but special leads were used to interrupt the passage of current so the patient received no electrical stimulation" p23 Comment: control participants received sham treatment
Blinding of personnel (performance bias) All outcomes	High risk	Quote: "Only the therapist doing daily stimulation treatment……knew the group assignment of each subject." p23 Comment: personnel were not blinded
Blinding of outcome assessment (detection bias) All outcomes	High risk	Quote: "Only the therapist doing …… weekly tracings knew the group assignment of each subject." p23 Comment: assessor was not blinded
Incomplete outcome data (attrition bias) All outcomes	Unclear risk	Quote: Table 4‐ 27/192 withdrew; p25 Comment: insufficient detail reported
Selective reporting (reporting bias)	High risk	Comment: insufficient detail reported to include outcome data in meta‐analysis
Other bias	High risk	Comment: Participants in the control group were moved across to the experimental group but outcome data prior to switching groups are not provided. Participants in the experimental groups stopped therapy if the ulcer healed, the doctor decided to intervene or the patient withdrew. Many wounds for a single patient, but it was not clear how this was dealt with in the analysis. Participants finished the intervention but were then re‐entered into the trial if they developed another wound. Trial was sponsored by third parties who were likely to have a vested interest in a positive finding. The unit of randomisation was the participant, but some participants had more than one pressure ulcer and data were analysed by pressure ulcers with no account for non‐independence of data.