Initiating Colorectal Cancer Screening in Unscreened Individuals 45 to 54

AIM 1: Identify system-level determinants of CRC screening in individuals age 45 to 54 using mixed methods.

A. Overview: The objective of this aim is to identify systems-level determinants of CRC screening in unscreened individuals age 45 to 54 at CSHC using the CFIR framework. We will work with Dr. Rachel Hindin, our clinical champion, to: (1) conduct a validated organizational survey with staff and providers and connect to patient-level EHR data to quantitatively assess systems-level determinants of screening; (2) conduct in-depth interviews with patient-facing CHC staff and providers; and (3) conduct in-depth interviews with patients.

B. Approach: B.1. Organizational Survey: Participants and recruitment: The organizational survey will be conducted at CSHC and 9 other CHCs, which will allow us to gather generalizable information about systems-level determinants of screening (n(CHCs) = 10; n(survey participants) = 50; n=5 staff and providers directly involved in CRC screening per CHC). Eligible staff at CHCs will be identified by our staff contact at each CHC.

Procedures: We will use a validated survey previously developed by ISCCCE, to collect data on systems-level factors that contribute to CRC screening. The survey collects data on CFIR inner setting constructs at the CHC including human resources practices, learning climate, available resources, patient needs and resources, linguistic services, implementation demands, and leadership engagement. Survey data will be collected via RedCap. Electronic informed consent will be obtained prior to each survey. Written or phone surveys (written or verbal consent) will be offered to those unable to provide electronic responses. We will connect survey data with patient DRVS data and fit models to quantitatively determine independent systems-level determinants of CRC screening. We will collect data across 10 ISCCCE partner sites which will also allow us to identify sites within the Mass League with similar contexts for testing in my future R01 and to maximize generalizability.

Data Analysis: Characteristics will be measured using a 5-point Likert scale with a "1" rating meaning strongly disagree and "5" rating meaning "strongly agree." Scores across responses will be collected as mean and standard deviations, and an aggregate score will be created by averaging survey responses pertaining to each characteristic. Generalized linear mixed effect models will be used to examine the outcome of CRC screening (yes/no) as function of patient demographics (first level) and inner setting survey construct scores (second level). Findings from the organizational survey will be used to adapt staff and patient interview guides to allow us to dive deeper into findings from the survey. Prior to the start of analysis, I will complete coursework in multilevel modeling under the guidance of Dr. Chan.