Predictive Factors and Model Performance for Incisional Hernia Development After Liver Transplantation: Insights from a Single-Center Study
Ghazinoer Anna, Dhaene Sam, 2025
The research conducted in this thesis is dedicated to helping liver transplant patients avoid a common complication after surgery known as an incisional hernia. An incisional hernia occurs when internal organs or tissues push through a weakened spot in the abdominal wall where a surgical incision was made. This can lead to discomfort, additional surgeries, and a reduced quality of life for the patients.
Why are liver transplant patients at higher risk for developing incisional hernias?
First, liver transplants require large abdominal incisions, that weaken the abdominal wall. After the surgery, patients need to take medications called immunosuppressants to prevent their bodies from rejecting the new liver. While essential, these medications also reduce the body's ability to heal wounds and fight infections. If the surgical wound becomes infected, it can further slowdown healing and weaken the area around the incision. Additionally, many liver transplant patients have multiple health issues such as liver scarring (cirrhosis), poor nutrition, a history of alcohol abuse and high blood pressure. These conditions can impair the healing process, increasing the likelihood of an incisional hernia developing.
How can incisional hernias be prevented?
Surgeons can strengthen the abdominal wall during the initial surgery by placing a mesh — a flexible, sterile material — directly into the abdominal wall. This method, called Prophylactic Mesh Augmentation (PMA), provides extra support as the body heals. Additionally, taking steps to minimize surgical wound infections after the surgery can further reduce the chances of developing incisional hernias.
What are the findings of this master’s thesis research?
This thesis evaluated the predictive performance of the Penn Hernia Risk Calculator, the leading tool for predicting incisional hernias, using data from 308 liver transplant patients at the University Hospital of Ghent (UZ Gent). Although the calculator included a model for transplant surgeries, it did not perform well. To improve its predictive performance, the Penn model was revised to better suit liver transplant patients. The revised model identified seven key factors: high blood pressure, a history of herniation, being 60 years of age or older, limited physical functionality before surgery, polycystic liver disease, prior abdominal surgeries and prior liver transplantation. The revised model showed good predictive performance in liver transplant patients. Additionally, the revised model was compared to nine other machine learning methods, all using the selected variables. The comparison showed that all models could predict incisional hernias in liver transplant patients, but the sensitivity and specificity varied depending on the model used. The thesis concludes that it is essential to avoid missing patients who might develop hernias (false negatives) while also preventing incorrect predictions that could lead to unnecessary treatments (false positives). Although identifying more potential hernia cases can facilitate preventive measures like the prophylactic mesh, these treatments also carry potential complications, such as infections and chronic pain. Therefore, when selecting a prediction model for clinical use, it is essential to ensure a careful balance between sensitivity and specificity. This dual focus helps accurately identify patients at risk while avoiding unnecessary treatments.
What is the added value and impact of this master's thesis research?
Accurate prediction models offer potential significant clinical and societal advantages by enabling physicians to identify patients at heightened risk of developing incisional hernias post-surgery. This risk stratification could allow for the implementation of preventive measures in high-risk individuals, such as the use of PMA during the initial surgical procedure. Consequently, this approach could decrease the incidence of hernia-related complications, reduce patient discomfort, promote smoother recovery and minimize the need for reoperations. Beyond the benefits to individual patients, this research positively impacts societal outcomes. Reducing the prevalence of hernias following liver transplants diminishes the need for additional surgeries and other medical interventions, leading to substantial cost savings for healthcare systems. Therefore, preventing hernias enables healthcare facilities to allocate resources more efficiently, thereby enhancing their capacity to provide high-quality care to a larger patient population. Furthermore, the decrease in medical complications results in reduced absenteeism from work for patients, boosting overall economic productivity and societal well-being.
| Promotor | Frederik Berrevoet |
| Opleiding | Geneeskunde |
| Domein | Chirurgie |
| Kernwoorden | machine learning Prediction model Incisional Hernia Liver Transplantation Hernia Prevention Hernia Prediction Models |