Comparative Effectiveness for Ventral Hernia Repairs among an Obese Patient Population

Chao Song, MPH¹, Emelline Liu, MSHS¹, Lizheng Shi, PhD², Daniel Marcus, MD³. ¹Intuitive Surgical, ²Tulane Univeristy, ³Marina Del Rey / Cedars Sinai Medical Center

Objectives: Obesity is prevalent among ventral hernia repair (VHR) patients and increases surgical complexity. When cases are more complex, different surgical approaches often have greater variation in clinical and economic outcomes. Therefore, comparison of VHR surgical approaches is particularly important in an obese patient population. This retrospective observational study examines the comparative effectiveness of 3 surgical approaches for VHRs: (1) robotic-assisted/RVHR, (2) laparoscopic/LVHR, and (3) open/OVHR in obese patient population.

Methods: Obese (BMI≥30) patients with ICD-9-CM codes (53.4x, 53.5x, 53.6x) were identified from Premier Perspective Database (Premier) from 09/2013-09/2015. Patients ≥18 years old having elective VHRs in an inpatient setting were included. Patients whose primary procedure were not VHRs or had gastrointestinal tract resections were excluded. Surgical approaches were compared through 1:1 Greedy propensity score matching. Confounding factors including patient socio-demographic and clinical characteristics, along with hospital level factors were applied to generate the propensity score. Clinical effectiveness, services utilizations, and costs were assessed. A range of complications during hospitalization were captured and were categorized by systems or organs. ICD-9-CM codes, along with billing record search were utilized to extract the surgical approaches and outcomes. All analyses were performed with SAS 9.4.

Results: The study included 6,642 inpatient elective obese VHRs performed, 4,354 (67%) of them were OVHR, 1,992(31%) LVHR, and 96 (2%) RVHR. Prior to matching, all 3 surgical cohort’s confounding factors were statistically different. Post-matching, all cohorts’ confounding factors were statistically similar. The 1:1 matching resulted in 2 matched patient cohorts: 96 OVHR: 96 RVHR, and 94 LVHR: 94 RVHR, respectively. Comparing these post-matched cohorts (See Table 1), the RVHR cohort was found to have a lower overall complication rate, blood transfusion rate, and PCA dose, along with shorter hospital length of stay, but with a longer OR time than that of the OVHR cohort. Additionally, the RVHR cohort was found to have lower conversion and blood transfusion rates but longer OR time when compared to LVHR. Overall, the RVHR cohort shows comparable costs to OVHR and LVHR cohorts.

Conclusion: Comparatively, a robotic-assisted surgical approach for elective ventral hernia repairs was found to have better clinical effectiveness and comparable costs to open and laparoscopic surgical approaches in obese inpatient population. Further investigation is needed to evaluate impacts in other complex patient populations.