Plus-Hex CLINICAL dressings as a control, which were replaced with semiocclusive dressings once the wounds granulated. In both groups, the treatments were covered with standardised bandages from digit to elbow, reducing external variability and contributing to a robust methodology.
Wounds were assessed at pre-established intervals via examination of photographs, with time to granulation, percentage of epithelialisation and wound contraction calculated. To eliminate observer bias and further strengthen the methodology, images were randomised and the assessor was blinded. Wound necrosis, haemorrhage and neutrophilic infiltration were calculated from examination of 2 mm biopsies taken at wound origination and four further intervals throughout the study. However, this intervention introduces potential confounding effects, through the disruption of the wound bed. Further to this, Demaria et al. [ 11 ] identified a lack of consistency in biopsy depth and angle across wounds, introducing further confounding variables. They suggested that larger, 4 mm biopsies may have increased consistency, although they did not investigate this further. The methodology used in the study may have been further compromised by the implementation of subjective assessment classification, for example, haemorrhage was classified as 0 = none, 1 = mild, 2 = moderate or 3 = severe. It is not specified whether the same board-certified veterinary pathologist evaluated every sample, potentially introducing perception-based inconsistency in grading.
Results showed that wounds treated with NPWT achieved granulation significantly faster( median time: 3 days NPWT vs 7 days control) and developed significantly smoother granulation tissue by day 14. However, the NPWT group presented significantly higher bacterial loads at day 7, and slower wound contraction. While the authors specify that a value of p < 0.05 indicated significance, individual p-values were not provided, limiting the transparency and interpretability of the results.
Overall, as the first canine study investigating the effect of NPWT, Demaria et al. [ 11 ] provide weak to moderate evidence for its efficacy. Their findings suggest that although canine NPWT may replicate some of its purported benefits in humans, it may not offer comparable infection control. Limitations, particularly related to biopsy execution and assessment subjectivity, constrain the generalisability of any drawn conclusions, underscoring the need for further, rigorously controlled primary research.
The research of Demaria et al. [ 11 ] was repeated and further developed by Stanley et al. [ 12 ], who also incorporated skin grafting. Wounds of 4 cm × 1.5 cm were created on the bilateral antebrachia of five purpose-bred female beagles, and excised skin sections were transposed to the opposite wound bed to create full-thickness skin grafts( FTSGs). Wounds were randomly assigned to either the NPWT group or the control group, which received bolster dressings. Bolster dressings are conventionally used in FTSGs [ 13, 14 ], and thus served as a suitable control. Additionally, randomised group assignment occurred after wound creation, which mitigated potential surgical bias.
Wounds were biopsied at regular intervals and assessed through blinded image analysis. Although blinding reduces observer bias, some subjective assessment points, for example,‘ moist’ versus‘ wet’ and‘ mild’ versus‘ moderate’, were used, and the authors do not specify whether the same assessor analysed the images. Consequently, some of the limitations present in Demaria et al. [ 11 ] are also seen in Stanley et al. [ 12 ]. However, Stanley et al. [ 12 ] directly addressed the previous study ' s concerns regarding biopsy consistency, by using a 4 mm biopsy punch at pre-established sites. This methodological refinement reduces sample variability, strengthening the study ' s overall reliability.
Stanley et al. [ 12 ] reported that wounds within the NPWT group achieved granulation significantly faster than control wounds( 2 days vs 7 days, p = 0.04). Furthermore, the control group exhibited significantly more necrosis on days 2, 4, 7 and 10( p = 0.01). In contrast to Demaria et al. [ 11 ], no positive bacterial cultures were identified in either group by Stanley et al. [ 12 ], reflecting the use of meticulous aseptic preparation. Aseptic wound creation limits the generalisability of the findings to traumatic wounds, although comparisons can still be made to highquality surgical wounds or well-prepared skin grafts. Furthermore, the absence of positive bacterial culture in the study by Stanley et al. [ 12 ] draws attention to the relatively high proportion of bacterial cultures isolated by Demaria et al. [ 11 ], suggesting potential inconsistencies in aseptic technique or wound management in the study of Demaria et al. [ 11 ] given the similar methodologies of the two studies.
Both studies share certain limitations, including small sample sizes and the use of subjective assessments. However, they also demonstrate rigorous overall methodologies, with Stanley et al. [ 12 ] directly building on earlier research to standardise key aspects of their approach, strengthening their reported evidence of the effects of NPWTs in terms of decreasing time to granulation and reduction of necrosis. Together, these preclinical studies provide a meaningful foundation of evidence for the use of NPWT in canine patients, justifying the advancement towards clinical-stage research.
Clinical research
Pitt and Stanley [ 15 ] reported the first larger-scale clinical study investigating the use of NPWT in canine patients. Forty-five patients were recruited, in which 53 open wounds were identified for study. Wounds were included if they were classified as full-thickness wounds, suitable for NPWT system integration, that were not
Volume 41( 1) • February 2026
47