Studies. The literature search identified a total of 407 unique references; all were exported to Endnote, along with 3 additional articles identified from other sources. Of these 410 articles, 127 were identified as duplicates, resulting 283 abstracts and titles. These were further screened for eligibility, generating 80 potentially relevant abstracts that were retrieved and considered for inclusion in the final review. Of these, 8 (6 RCTs13,38-42 and 2 nonrandomized CCTs43,44) met the inclusion criteria and were subjected to full-data extraction and quantitative analysis. Figure 1 provides a flow chart of the process and results for screening eligibility and study selection.
All 8 studies described the study target population as persons with SCIs who had at least 1 PrU. The number of patients per study ranged from 7 to 150, with the number of ulcers studied ranging from 7 to 192. Four (4) studies (N = 415 ulcers) measured mean daily percentage change in ulcer size,41-44 1 study (N = 192 ulcers) measured the mean weekly percentage change in ulcer size,13 and 3 studies (N = 58 ulcers) measured mean percentage change in ulcer size across the entire study period.38-40 Six (6) out of 8 trials (N = 509 ulcers) compared PC ES to sham/no ES treatment,13,38-41,44 1 trial (N = 50 ulcers) compared constant current ES to sham ES treatment,42 and 1 (N = 150 ulcers) compared PC ES to constant current ES therapy or sham ES.43 For stimulus pulse settings, stimulation varied from 40 Hz to 100 Hz in frequency and from 50 V to 150 V (or 4 mA to 45 mA) in intensity.
In terms of electrodes placement, in 3 studies38-40 (N = 58 ulcers) the electrodes were overlaid on wound bed, in 4 studies (N = 557 ulcers) the electrodes were applied to intact skin around the wound,13,41,43,44 and in 1 (N = 50 ulcers) ES treatment compared electrodes overlaid on the wound bed to electrodes placed on intact skin around the ulcers.42 Details of sample characteristics are shown in Table 1A and 1B.
Methodological quality. One (1) of the 6 RCTs described an appropriate method to generate the randomization sequence,40 2 were double-blinded and described the method of double-blinding,39,42 3 adequately described allocation concealment,38-40 and 2 used ITT to analyze the data.40,41 Two (2) RCTs were considered to be of reasonably good methodological quality according to the Jadad score along with 2 other items; hence, they were classified as low risk of bias trials.39,40 The remaining 4 RCTs were considered to exhibit moderate risk of bias.13,38,41,42 Two (2) nonrandomized CCTs were assessed for their reporting quality using the Down and Black tool, scoring 13 and 8, respectively, out of a total achievable score of 28 and subsequently considered relative high risk of bias trials.43,44
Data pooling and meta-analysis.
Effectiveness of ES. ES effectiveness was assessed according to mean weekly healing rate (ie, average percentage change per week in ulcer size), the number of ulcers healed, and the incidence of ulcers worsening (defined as wound size increased during study period), comparing different ES settings and/or how ES performed versus sham or no treatment.
With regard to overall healing rate by ES versus sham/no ES, the pooled analyses of the 7 relevant trials showed people receiving ES treatment in addition to conventional wound treatment (which involved cleansing, dressing, nutrition, and debridement as necessary) reporting a higher weekly healing rate than those without ES treatment by 4.97% (MD 4.97, 95% CI 1.97-7.98, P = 0.001) (see Figure 2). However, heterogeneity among the studies was substantial (I2 = 83%, P <0.00001). A subgroup analysis of RCTs considered to be of good methodological quality showed a trend toward a higher weekly healing rate in people treated with ES than people without ES treatment, but the pooled effect was not significant (P = 0.07).
When healing rate by PC ES versus sham/no ES was assessed, the pooled analysis of the 6 trials13,38-41,43 showed a significantly higher weekly healing rate in people who were treated with PC ES than those without ES treatment (MD 6.27, 95% CI 2.77-9.78, P = 0.0005, I2 = 74%( (see Figure 3).
When healing rate by constant DC versus sham/no ES was compared, the pooled analysis of the 2 trials in which constant current ES was applied showed a nonsignificantly higher weekly healing rate in people treated with constant current than those without ES (MD 4.50, 95% CI 1.19-10.18, P = 0.12, I2 = 0% (see Figure 4).
With regard to healing rate by PC versus DC ES treatment, 1 CCT43 compared PC ES treatment versus DC for PrU healing. A higher weekly healing rate was achieved by PC ES: 5.43% to 4.40% per day versus 3.11% to 3.83% per day, respectively (P = 0.03).
Healing rate by electrode placement also was assessed (ie, active electrode overlaid the wound bed versus sham/no ES). A meta-analysis of the 4 trials38-40,42 that applied the active electrodes directly on the wound found a significantly higher rate with ES irrespective of current type than without ES (MD 9.01, 95% CI 2.02-16.00, P = 0.01, I2 = 90%) (see Figure 5).
When electrode placement on intact skin was analyzed, the pooled analysis of the healing rate for the 4 trials13,41-43 where both electrodes were placed on intact skin/the edge of wound versus sham/no ES showed a significantly higher weekly healing rate in people who received ES than those who received no ES (MD 7.71, 95% CI 1.59-13.83; P = 0.01, I2 = 79%) (see Figure 6).
The healing rate when the active electrode overlaid the wound bed was compared to electrodes placed on intact skin/the edge of wound. One (1) study42 compared the effect of ES delivered by applying the electrodes either directly on the wound bed or on the edge/intact skin around the ulcer versus sham ES treatment. In group 1, the positive electrode overlaid the ulcer and 4 negative electrodes were laid around the ulcer; in group 2, 2 electrodes were laid on intact skin at the ulcer edge across the wound; and in group 3, 2 electrodes were laid on intact skin at the ulcer edge across the wound without ES delivery. The authors reported electrodes that overlaid wound bed achieved a higher healing rate than electrodes placed on intact skin around the ulcer or sham ES group (1.6%/day overlaid versus 4.8±1.5%/day versus 4.2±11%/day).
Two (2) trials13,40 addressed the number of ulcers completely healed by ES versus sham/no ES. Both trials reported a higher number of ulcers completely healed in the ES treatment group (52% and 37%, respectively13,40) compared to sham/no ES treatment (24% and 28%, respectively13,40). Pooled analysis of these 2 trials showed significantly higher numbers of ulcers (N = 226 ulcers) healed with ES treatment (OR 2.68, 95% CI 1.17–6.14, P = 0.02, I2 = 0%) (see Figure 7).
Two (2) studies40,41 reported the incidence of ulcers that worsened during the study period compared to sham/no treatment. Both trials reported lower number of ulcers worsened in individuals who had ES treatment (N = 143 ulcers). However, the pooled analysis of these 2 trials showed the difference was nonsignificant (OR 0.38, 95% CI 0.12–1.24, P = 0.11, I2 = 18%) (see Figure 8).
Adverse events. Only 1 of the 8 studies40 reported adverse events. The authors indicated some patients experienced minor adverse reactions related to ES, which included red, raised, itchy skin beneath the large dispersive electrode. One (1) patient had a persistent (>24 hour) redness or burn under the active electrode, presumably from too high a stimulus intensity, which resolved within 48 hours.