Diabetic foot ulcer (DFU) is one of the commonest complications of poorly controlled diabetes mellitus and a leading cause of non-traumatic lower-limb amputation; the lifetime risk among people with diabetes is estimated at 15–25% [1,2]. Ulcers arise from neuropathy, foot deformity, repeated pressure and vascular compromise. Deep tissue samples guide antimicrobial therapy better than superficial swabs [3], osteomyelitis should be excluded, for example with the probe-to-bone test [4,5], and infected ulcers need antibiotics [6–9]. For non-infected ulcers, care rests on debridement, pressure off-loading, glycaemic control and treatment of ischaemia, with hyperbaric oxygen considered when healing stalls [10,11]. Ulcers are graded by depth, infection and ischaemia using systems such as the Wagner and University of Texas classifications [12,13].

Local dressing selection remains contested. Trials and reviews of silver, alginate, collagen-alginate and vacuum-assisted dressings against gauze have reported variable benefits [14–20], and pooled analyses found that most advanced dressings were not clearly superior to simple contact dressings, although hydrogels may perform better [21–24]. Negative pressure wound therapy (NPWT), including instillation variants, has been associated with faster healing and fewer amputations, mostly in low-certainty evidence [25–31]. Alginate-based dressings, topical oxygen, NPWT with saline irrigation and overviews of NPWT have also been reported [32–37]. More recent approaches include topical esmolol, micropower vacuum dressings, silver foam and stem-cell-loaded gels, and network meta-analyses rank dressings differently depending on the comparisons available [38–45]. Cost and care setting also influence dressing choice [46,47].

Single-centre studies, mostly from India, report faster healing of chronic and diabetic ulcers with collagen dressings, and with topical sucralfate, than with conventional saline dressings [48–52]. Sucralfate, a sucrose octasulfate aluminium salt, is proposed to act as a protective barrier through electrostatic binding to wound proteins and to have antibacterial, angiogenic and granulation-promoting effects [52]. Papain gel is also used as a topical dressing. To our knowledge, none of the studies cited here compared all four dressings in one cohort. We therefore compared collagen, topical sucralfate, papain gel and conventional saline dressings in non-infectious DFUs over three weeks, using wound area, necrotic tissue and granulation tissue as outcomes.

Study design and setting This prospective, randomized, four-arm parallel-group comparative study was conducted over 18 months between 2020 and 2022 in the surgical outpatient department and inpatient wards of PES Institute Of Medical Sciences And Research, kuppam, Andhra pradesh. Informed consent was obtained from all participants. Participants Eligible patients were aged 18–75 years, had diabetes mellitus and a non-infectious lower-limb ulcer of more than 2 weeks duration and less than 10 × 10 cm in size, and consented to participate. Exclusion criteria were a pulseless limb, immunocompromised state, associated osteomyelitis, skin malignancy, cellulitis, diabetic ketoacidosis, exposed tendon or bone or Charcot joint, diabetic gangrene of a toe, and infected wounds with features of the systemic inflammatory response syndrome. Randomization Using a pre-tested, pre-designed proforma, patients were allocated by a computerized randomization chart to one of four groups of 25: conventional normal saline dressing, topical sucralfate dressing, collagen dressing or papain gel dressing. Interventions After clinical examination and relevant investigations, all ulcers were sharply debrided and cleaned with normal saline. Pressure was off-loaded from the affected area and no antibiotics were used in any group. Dressings were applied once daily and patients were followed daily for 3 weeks. The dressings were applied as follows. • Saline (conventional): a saline-soaked gauze piece was placed over the ulcer and covered with a pad and roller bandage. • Sucralfate: 1 g sucralfate tablet was crushed to powder and suspended in 5ml sterile normal saline. sterile gauze soaked in the suspension was placed over the wound at 20 mg/cm². • Collagen: collagen powder was sprinkled over the whole wound surface, which was then closed with sterile gauze and covered with a pad and roller bandage • Papain gel: papain gel was applied to the ulcer first and then covered with a dressing Ulcer photographs and culture and sensitivity were obtained before and after treatment; these data are not analysed in this report. Outcome measures Wound surface area (cm²) was calculated as length × width after debridement at baseline and at 3 weeks. Necrotic tissue covering the ulcer was scored from 1 to 6 (1 = 76–100%, 2 = 51–75%, 3 = 26–50%, 4 = 11–25%, 5 = 0–10%, 6 = no necrotic tissue), so higher scores indicate less necrosis. Granulation tissue was scored from 1 to 4 (1 = none, 2 = pink or dull, 3 = bright red filling 25–74% of the wound, 4 = bright red filling 75–100%). The main outcome was change in wound area over 3 weeks, with necrotic and granulation tissue scores as secondary outcomes. We also recorded age, sex, socioeconomic status, onset, site and progress of the ulcer, ulcer duration, duration and treatment of diabetes, other illnesses, and alcohol and tobacco use. Sample size Sample size was calculated as 4pq/d², with p = 40.87%, q = 100 − p and an absolute error d of 10%, giving 96, which was rounded to 100 (25 per group). The value of p was taken from a previous comparison of sucralfate and saline dressings [52]. Statistical analysis Data were entered in Excel and analysed with SPSS version 23. Proportions were compared with the chi-square or Fisher’s exact test and means with the unpaired t-test, as specified in the original analysis plan. A p value below 0.05 was considered significant. Pairwise comparisons between individual dressings and adjustment for multiple comparisons were not performed.

Participants

All 100 randomized patients (25 per group) completed 3 weeks of follow-up; none were lost. Mean age was 56.6 years (range 34–75) and 65 patients (65%) were men. As shown in Table 1, the groups were not significantly different in demographic, diabetes-related or ulcer characteristics, and all patients had debridement at baseline. Baseline mean wound area was largest in the collagen group (50.39 cm²) and smallest in the papain group (37.05 cm²), a difference that did not reach significance (p = 0.056).

Wound area

Mean wound area at 3 weeks was 35.85 cm² (collagen), 34.30 cm² (saline), 30.95 cm² (papain) and 30.25 cm² (sucralfate), without a significant difference between groups (p = 0.529; Table 2). The mean absolute reduction in area was greatest with collagen (14.54 cm²), followed by saline (10.07 cm²), sucralfate (9.99 cm²) and papain gel (6.10 cm²) (p = 0.001). Because baseline areas differed, we also calculated the relative reduction of each group’s mean area from the group means: 28.9% for collagen, 24.8% for sucralfate, 22.7% for saline and 16.5% for papain gel. This calculation was not tested statistically and, unlike the absolute reduction, places sucralfate marginally ahead of saline.

Necrotic tissue

Necrotic tissue scores were similar at baseline (p = 0.217). At 3 weeks, 21 of 25 patients (84%) in the collagen group had 10% or less necrotic coverage or none (scores 5–6), compared with 18 (72%) in the saline group, 14 (56%) in the sucralfate group and 9 (36%) in the papain group (p = 0.001; Table 3). Complete absence of necrotic tissue was seen only in 2 collagen-treated patients. In contrast, 16 patients (64%) in the papain group still had 11–25% necrotic coverage, compared with 4 (16%), 7 (28%) and 11 (44%) in the collagen, saline and sucralfate groups.

Granulation tissue

Granulation scores were similar at baseline (p = 0.841). At 3 weeks, the highest score (bright red granulation filling 75–100% of the wound) was reached by 10 patients (40%) with collagen, 9 (36%) with sucralfate, 8 (32%) with saline and none with papain gel; pink or dull granulation (score 2) persisted in 11 patients (44%) in the papain group, compared with 1 (4%) with collagen, 3 (12%) with saline and 1 (4%) with sucralfate (overall p = 0.005; Table 4). Because pairwise tests were not done, the overall difference appears to be driven mainly by the papain group.

Variables not analysed

Pain, wound discharge, complete healing, time to healing, adverse events and culture results were not analysed in this report.

Tables

Table 1. Baseline characteristics of the 100 patients by dressing group

All patients had sharp debridement at baseline. Treatment of diabetes did not differ between groups (p = 0.095). p values are from the chi-square or Fisher’s exact test (categorical) and unpaired t-test (continuous).

Table 2. Wound surface area at baseline and after 3 weeks of dressings

† Absolute reduction divided by the baseline group mean; this is not the mean of individual percentage changes.

Table 3. Necrotic tissue score at 3 weeks, n (%)

Overall p = 0.001 (chi-square or Fisher’s exact test). Higher scores indicate less necrotic tissue. No patient had a score of 1–3 at 3 weeks.

Table 4. Granulation tissue score at 3 weeks, n (%)

Overall p = 0.005 (chi-square or Fisher’s exact test). No patient had a score of 1 (no granulation) at 3 weeks.

In this four-arm comparison of dressings for non-infectious diabetic foot ulcers, collagen was associated with the most favourable necrotic tissue clearance, granulation and absolute area reduction after 3 weeks, and papain gel with the least favourable results on all three measures. Sucralfate and saline dressings were intermediate and not clearly distinguishable from each other. However, final wound area did not differ significantly between groups (p = 0.529), so the differences lie in the amount of change over 3 weeks rather than in the size of the wounds at the end, and they should be read as short-term findings. Two features of the data call for caution. First, baseline wound area was largest in the collagen group and smallest in the papain group (p = 0.056). An absolute reduction in cm² is easier to achieve from a larger wound, so the imbalance may have favoured collagen and disadvantaged papain gel. When reduction was expressed relative to each group’s baseline mean, the order was collagen (28.9%), sucralfate (24.8%), saline (22.7%) and papain gel (16.5%), so the relative position of sucralfate and saline depends on the metric used. We therefore cannot conclude that saline was more effective than sucralfate. Second, the differences in necrotic tissue and granulation scores were tested only as overall four-group comparisons. The granulation pattern in particular suggests that the overall difference is driven by the papain group, because the proportions reaching the highest score were similar in the collagen, sucralfate and saline groups (40%, 36% and 32%). Pairwise tests with adjustment for multiple comparisons are needed before collagen can be called superior to every comparator. Our collagen findings are consistent with single-centre studies in chronic leg and foot ulcers, which reported greater ulcer size reduction and a shorter hospital stay [48], faster healing than saline-moistened gauze [49], and a shorter time to 50% wound size reduction with less need for skin grafting [50]. In a randomized trial of 60 patients with DFUs, collagen granules were associated with a smaller mean wound area by week 2, fewer patients with slough or necrotic tissue, and no wound discharge by week 4, compared with a conventional dressing [51]. Those studies compared collagen with conventional dressings over short follow-up; our study adds a direct comparison with two other topical agents. A previous study reported a greater area reduction with sucralfate than with saline dressings (40.87% versus 15.62%) and complete healing in 2.68 versus 5.36 weeks [52]. We found no clear advantage of sucralfate over saline (absolute reduction 9.99 versus 10.07 cm²; relative reduction 24.8% versus 22.7%). Differences in ulcer size, preparation (here a suspension at 20 mg/cm²), follow-up and outcome definitions may contribute, but we cannot exclude that sucralfate has little effect in this population. The poorer results with papain gel need replication: none of the studies in our reference list evaluated papain gel, and the product, concentration and application method may matter. Reviews and network analyses of DFU dressings suggest that most advanced dressings are not clearly superior to basic contact dressings and that selection should also consider exudate control, comfort and cost [20–24,41,42,46]. Our results fit this picture, with modest differences between dressings and no difference in final wound area. We considered collagen dressings more costly than saline gauze, but costs were not measured, so no cost-effectiveness conclusion can be drawn from our data. Strengths and limitations Strengths include random allocation, standardized debridement and off-loading, daily supervised dressing, no loss to follow-up, and the simultaneous comparison of four dressings. Limitations are the single centre and small groups (25 each); the short 3-week follow-up without a complete-healing endpoint; wound area estimated as length × width and tissue assessed by ordinal visual scores, which are subjective, with assessor blinding not reported; absence of pairwise testing and multiplicity adjustment; the baseline difference in wound area; the lack of pain, discharge, safety, culture and cost outcomes; and uncertainty about the basis of the sample-size estimate. The cohort (mean age 56.6 years, maximum 75) was middle-aged rather than predominantly geriatric, and the findings may not apply to older patients, infected ulcers, ulcers of 10 × 10 cm or larger, ischemic limbs or ulcers with osteomyelitis, which were excluded.

In this single-centre randomized comparison over 3 weeks, collagen dressings were associated with the most favourable necrotic tissue clearance, granulation and absolute wound area reduction in non-infectious diabetic foot ulcers, papain gel with the least favourable, and topical sucralfate was not clearly different from saline. Because final wound areas were similar, pairwise tests were not performed and cost was not measured, these findings are hypothesis-generating. Larger, adequately concealed and blinded trials with complete-healing, time-to-healing, pain, safety and cost outcomes are needed.

Declaration of interests: No conflict of interest

Funding: Not required