(WHS-I.07) STRUCTURAL EQUATION MODEL TO QUANTIFY IMPORTANCE OF PATIENT FACTORS AND WOUND MICROBIOME ON HEALING
Wednesday, May 15, 2024
1:45 PM – 4:00 PM East Coast USA Time
Background: The contribution of microbiome and host factors to driving chronicity and rate of healing in wounds is widely appreciated. However, there is currently little ability to account for the many variables and dynamics influencing differences in healing. Here, with the goal of developing a predictive framework, a novel structural equation modelling (SEM) approach was employed to model the chronic wound environment in relation to healing.
Methods: The dataset consisted of 565 chronic wound microbiomes detected at initial patient visit using 16S sequencing paired with patient medical information. A novel pre-modeling parcel optimization routine was developed to transform the microbiome species table into two latent variables that related to healing time either positively or negatively. These latent constructs in addition to specific species correlating with healing, and patient/wound data were evaluated for model fit in the SEM using backward selection and delta chi-squared testing.
Results: A microbiome latent construct significantly associated with improved healing was validated, and the final SEM included this latent construct plus three species associated with diminished healing (Anaerococcus vaginalis, Finegoldia magna, Pseudomonas aeruginosa), as well as smoking, wound volume, slough, exudate, edema, percent granulation, and wound type. This model explained 49% of variation in healing time with the microbiome contributing the largest proportion of variance explained. Percent granulation and wound volume accounted for the second and third most variance explained. Species that formed the latent construct tended to correlate with each other less than did the remaining species (p < 0.001), potentially reflecting that species associated with faster healing act individually rather than synergistically. Also, species that formed the latent construct that was associated with faster healing also included species that are routinely treated with biofilm-based wound care.
Conclusions: This study provides a novel pre-modeling approach allowing the integration of microbiome data into SEM. The final model validated the importance of many variables on differences in healing time, with wound microbiome species being the most important. The importance of microbes in this model advocates for the efficacy of guiding treatment based on results of DNA sequencing-based microbiota profiling.