Predicting Transvaginal Surgical Mesh Exposure Outcomes Using an Integrated Dataset of Blood Cytokine Levels and
Mihyun Lim Waugh1, Tyler Mills2, Nicholas Boltin1
1Department of Biomedical Engineering, University of South Carolina, 301 Main St, Rm 2C02, Columbia, SC, 29208-4101, United States, 1 8646336181.
Background:
Transvaginal insertion of polypropylene mesh was extensively used in surgical procedures to treat pelvic organ prolapse (POP) due to its cost-efficiency and durability. However, studies have reported a high rate of complications, including mesh exposure through the vaginal wall. Developing predictive models via supervised machine learning holds promise in identifying risk factors associated with such complications, thereby facilitating better informed surgical decisions. Previous studies have demonstrated the efficacy of anticipating medical outcomes by employing supervised machine learning approaches that integrate patient health care data with laboratory findings. However, such an approach has not been adopted within the realm of POP mesh surgery.
Objective:
We examined the efficacy of supervised machine learning to predict mesh exposure following transvaginal POP surgery using 3 different datasets: (1) patient medical record data, (2) biomaterial-induced blood cytokine levels, and (3) the integration of both.
Methods:
Blood samples and medical record data were collected from 20 female patients who had prior surgical intervention for POP using transvaginal polypropylene mesh. Of these subjects, 10 had experienced mesh exposure through the vaginal wall following surgery, and 10 had not. Standardized medical record data, including vital signs, previous diagnoses, and social history, were acquired from patient records. In addition, cytokine levels in patient blood samples incubated with sterile polypropylene mesh were measured via multiplex assay. Datasets were created with patient medical record data alone, blood cytokine levels alone, and the integration of both data. The data were split into 70% and 30% for training and testing sets, respectively, for machine learning models that predicted the presence or absence of postsurgical mesh exposure.
Results:
Upon training the models with patient medical record data, systolic blood pressure, pulse pressure, and a history of alcohol usage emerged as the most significant factors for predicting mesh exposure. Conversely, when the models were trained solely on blood cytokine levels, interleukin (IL)-1β and IL-12 p40 stood out as the most influential cytokines in predicting mesh exposure. Using the combined dataset, new factors emerged as the primary predictors of mesh exposure: IL-8, tumor necrosis factor-α, and the presence of hemorrhoids. Remarkably, models trained on the integrated dataset demonstrated superior predictive capabilities with a prediction accuracy as high as 94%, surpassing the predictive performance of individual datasets.
Conclusions:
Supervised machine learning models demonstrated improved prediction accuracy when trained using a composite dataset that combined patient medical record data and biomaterial-induced blood cytokine levels, surpassing the performance of models trained with either dataset in isolation. This result underscores the advantage of integrating health care data with blood biomarkers, presenting a promising avenue for predicting surgical outcomes in not only POP mesh procedures but also other surgeries involving biomaterials. Such an approach has the potential to enhance informed decision-making for both patients and surgeons, ultimately elevating the standard of patient care.
