RCT of Screening Strategies Among Patients at High Risk for Developing HCC in a Safety-net Health System

RCT of Screening Strategies Among Patients at High Risk for Developing HCC in a Safety-net Health System

RCT of Strategies to Improve Screening Rates Among a Cohort of Cirrhotic Patients at High Risk for Developing HCC in a Safety-net Health System

Agency for Healthcare Research and Quality (AHRQ), Parkland Health and Hospital System, The University of Texas Health Science Center, Houston, Johns Hopkins Bloomberg School of Public Health

Hepatocellular carcinoma (HCC) is the 9th leading cause of cancer-related death in the US and one of the leading causes of death in patients with cirrhosis. Fewer than 1 in 5 high-risk patients undergo HCC screening, with lower rates in non-Caucasian and low socioeconomic status patients receiving care through safety-net health systems. Screening and follow-up failures lead to more advanced cancers, when curative therapies are not available and survival is significantly worse. Over 60% of HCC are diagnosed at advanced stages, due to poor recognition of high-risk patients, underuse of screening among these patients, and poor follow-up of abnormal screening tests. To address these barriers, the investigators propose to conduct a comparative effectiveness research randomized controlled trial of three screening strategies among a socioeconomically disadvantaged and racially diverse cohort of cirrhotic patients at high risk for developing HCC. Overall, 1800 patients attending Parkland, the Dallas safety-net health system, will be randomized to: Group 1: Usual care, with visit-based HCC screening per discretion of individual providers Group 2: Mailed HCC screening invitation outreach to eligible patients (low resource intensity) Group 3: Mailed HCC screening invitation outreach to eligible patients combined with centralized patient navigation to promote screening completion and follow-up (high resource intensity) Through three specific aims, this effectiveness research randomized controlled trial will: Aim 1: Engage stakeholders in design and implementation of HCC screening outreach interventions. Aim 2: Compare the clinical effectiveness and patient acceptability of the intervention strategies to increase completion of one-time and repeat HCC screening. Aim 3: Evaluate whether intervention effects are moderated by patient sex, race, ethnicity, English proficiency, and connectedness to primary care. The screening intervention strategies combine EMR-enabled case identification, system-level screening outreach, and patient navigation to improve identification of previously unrecognized cirrhotic patients, promote HCC screening completion, and facilitate follow-up of abnormal screening tests. This study will engage stakeholders throughout the research process, evaluate the effectiveness and acceptability of HCC screening strategies, and determine which patient subgroups benefit the most.

Carcinoma, Hepatocellular, Liver Neoplasms, Liver Disease, Early Detection of Cancer, Comparative Effectiveness Research

Individuals randomized to Group 1 will receive usual medical care and will not be directly contacted at any point of the trial. Study data and outcomes will be abstracted from the EMR.

Individuals randomized to Group 2 will receive: Mailed outreach invitation to complete HCC screening ultrasound and alpha-fetoprotein (AFP) blood test. "Live" phone calls 2 to 4 weeks after the mailed invitation to facilitate HCC screening completion. Up to three attempts will be made. All communications will use standard English or Spanish scripts. Centralized process to promote guideline-appropriate follow up testing with CT or MRI or referral to GI Clinic.

Individuals randomized to Group 3 will receive: Mailed outreach invitation to complete HCC screening ultrasound and alpha-fetoprotein (AFP) blood test. "Live" phone calls 2 to 4 weeks after the mailed invitation to facilitate HCC screening completion "and address patients' self-reported barriers to HCC screening (e.g., it does not apply to me). Up to three attempts will be made. All communications will use standard English or Spanish scripts. Centralized navigation to promote screening completion (i.e., appointment reminder phone calls from patient navigator) and guideline-appropriate follow up testing with CT or MRI or referral to GI Clinic.

Defined as the proportion of patients completing HCC screening every 6 months within 18 months of randomization.

Defined as the proportion of patients completing HCC screening every 7 months within 21 months of randomization.

Simple total program costs will be calculated for the intervention arms (Groups 2 and 3) and compared with a one-way ANOVA model.

Defined as the proportion of patients with a suspicious lesion (as any solid-appearing mass ≥1 cm in diameter not characterized as benign).

As the risk of HCC is significant in those with cirrhosis regardless of age, gender, or race, we will include adult patients with cirrhosis of all ages (> 21 years old), both sexes, and all races/ethnicities who speak English or Spanish. Our study leverages Parkland's electronic medical record (EMR) and uses a novel EMR-enabled case-finding algorithm to identify patients with known cirrhosis, using ICD-9 codes, as well as those with unrecognized but suspected cirrhosis, using a set of laboratory data. Patients with ICD-9 codes for cirrhosis or cirrhosis complication will be eligible for study enrollment if they meet the following criteria: One or more encounter with ICD-9 codes 456.0, 456.1, 456.2, 456.21, 567.23, 572.2, 572.3, and 572.4; OR, Two or more encounters with ICD-9 codes 571.2 and/or 571.5; OR, One encounter with ICD-9 codes 571.2 or 571.5 from a Parkland primary care clinic, GI, or women's health center. Patients with an AST to platelet ratio index (APRI) > 1.5 in combination with a platelet count < 300, and aspartate aminotransferase (AST) < 1,000 during study enrollment will be eligible. We will exclude patients who have known HCC or a suspicious appearing mass on imaging within six months prior to ascertainment of eligibility, as these patients require further diagnostic testing instead of routine screening. We will exclude patients with Child Pugh class C cirrhosis or other significant comorbid conditions with a life expectancy less than one year, (e.g., extrahepatic malignancy) because HCC screening is not recommended in these subgroups of patients. Inclusion Criteria: Parkland patients ≥ 21 years of age Diagnosis of cirrhosis or meets criteria for suspected cirrhosis ≥ 1 outpatient visit during 12 months prior to randomization Contact information on file English or Spanish speaking Exclusion Criteria: HCC or suspicious mass on imaging Any malignancy except malignant neoplasm of skin Metastatic solid tumor Palliative care referral Liver transplant Child Pugh C

Singal AG, Tiro JA, Murphy CC, Marrero JA, McCallister K, Fullington H, Mejias C, Waljee AK, Pechero Bishop W, Santini NO, Halm EA. Mailed Outreach Invitations Significantly Improve HCC Surveillance Rates in Patients With Cirrhosis: A Randomized Clinical Trial. Hepatology. 2019 Jan;69(1):121-130. doi: 10.1002/hep.30129. Epub 2018 Dec 14.

Singal AG, Tiro JA, Marrero JA, McCallister K, Mejias C, Adamson B, Bishop WP, Santini NO, Halm EA. Mailed Outreach Program Increases Ultrasound Screening of Patients With Cirrhosis for Hepatocellular Carcinoma. Gastroenterology. 2017 Feb;152(3):608-615.e4. doi: 10.1053/j.gastro.2016.10.042. Epub 2016 Nov 5.