D.J. Vitello2, M. Paukner1, B. Hasjim1, A. Jain3, A. Duarte1,5, L. Kulik1,6, L.B. VanWagner7, F. Obradovic3, J.E. Obayemi1,5, F. Crippa3, P. Barrios-Martinez1,5, E. Koep8, T. Banea1, O. Dentici1, L. Zhao1,5, C. Manski3, D.P. Ladner1,5 1Northwestern University, Transplant Outcomees Research Collaboreative, Chicago, IL, USA 2Feinberg School Of Medicine – Northwestern University, Department Of Surgery, Chicago, IL, USA 3Northwestern University, Department Of Economics, Evanston, IL, USA 4London School of Economics, London, LONDON, United Kingdom 5Northwestern Medicine, Division Of Transplantation, Chicago, IL, USA 6Northwestern Medicine, Division Of Gastroenterology And Hepatology, Chicago, IL, USA 7University Of Texas Southwestern Medical Center, Division Of Digestive And Liver Diseases, Dallas, TX, USA 8UnitedHealth Group, Center For Healthcare Research, Chicago, IL, USA
Introduction:
National guidelines recommend hepatocellular carcinoma (HCC) surveillance every 4-8 months for patients with cirrhosis as it is cost effective and results in earlier intervention leading to longer overall survival. Adherence is estimated to be extremely low. This study aimed to determine the rate of HCC surveillance adherence and identify predictors of improved adherence.
Methods:
Adults with cirrhosis and at least 1 year of follow up were identified using validated ICD-9/-10 codes within a claims database of a large national insurer between 2011-2021. The primary outcome of interest was proportion of time covered by HCC surveillance exams. Patients were observed from the time of cirrhosis identification until disenrollment or the development of HCC. Surveillance exams were identified using CPT codes for serum alpha fetoprotein level with an abdominal ultrasound (AFP with US) or contrast-enhanced CT or MRI as previously published. Patients were considered adherent for 8 months following any surveillance exam. A multivariable generalized linear mixed-effects model was used to identify predictors of HCC surveillance adherence. The model adjusted for baseline demographics produced adjusted odds ratios (aORs) with 95% confidence intervals (CIs).
Results:
270,686 patients were identified. Key demographic information and longitudinal HCC surveillance adherence are presented in Figure 1. Within the 12 months following cirrhosis identification, 26,825 (9.9%) had an AFP with US, 24,406 (9.0%) had an MRI, and 14,635 (5.4%) had a CT. Only 97,102 (35.9%) patients ever received at least any one surveillance exam over the entire course of follow up. Figure 1 shows the rate of adherence by length of follow up. Having follow up with a gastroenterologist (aOR 7.1 CI 7.0-7.2), HCV cirrhosis compared to MASLD (aOR 3.0 CI 2.9-3.1), and decompensation (aOR 1.4 CI 1.3-1.5) were significant predictors of higher HCC surveillance adherence.
Conclusion:
In this study of a large, national, insured cohort of patients with cirrhosis, HCC surveillance adherence was estimated to be low (Figure 1). Patients who followed with a gastroenterologist were significantly more likely to be guideline adherent. Further involvement of gastroenterologists and heightened incentives to adhere to national guidelines are needed to maximize the population-level benefits of HCC surveillance.