Population-Based Patient-Centric Care: Comprehensive Preventive Cancer Screening Using Health IT (TopCare)

Although there is considerable evidence that current health IT can improve certain elements of care, the most effective and efficient implementation of health IT systems for primary care population management are not currently known. Indeed, while many systems currently take a "case-management" approach to identify and address clinical care issues for high risk patients, no systems to our knowledge apply a risk-based approach that accounts both for adverse clinical outcome risk (e.g. breast cancer in a woman who has not had indicated screening for 4 years) and for clinical process risk (e.g. the likelihood that a specific patient will ignore a reminder letter and would therefore benefit from direct phone or in person contact). The investigators propose to directly test the hypothesis that implementing a health IT platform that 1) provides novel risk-based decision support using data derived from the electronic health record (EHR) and 2) leverages each clinician's unique knowledge of his or her patient panel will result in more effective and more efficient population-based primary care. The investigators will test this hypothesis in a practice-randomized clinical trial of preventive cancer screening within our primary care Practice-Based Research Network (PBRN).

In prior NIH-funded research, the investigators have demonstrated the efficacy of an IT-based population management system to improve breast cancer screening (NCI R21 CA121908). The investigators will expand our current IT platform from this single function (breast cancer screening) to a package of cancer prevention actions (breast, cervical, and colorectal cancer screening) and examine the added benefit of population-level preventive cancer care that is directed by specific clinician knowledge of individual patient needs. Moreover, rather than compare our system to currently sub-optimal "usual care" practice, our goal is to test whether the impact of our intervention exceeds the current state-of-the-art of IT-based population management. Therefore, control group practices will receive augmented standard care defined as a population-level reminder system with automated patient contacts. In augmented standard care control practices, the investigators will implement a system that includes: 1) a population-based perspective to identify all eligible patients overdue for screening, 2) an automated, centralized process to contact selected patients by letter, 3) a result management system that automatically tracks test scheduling and completion, 4) a web-based, easily accessible tool allowing practice personnel to contact patients not completing testing, and 5) use of patient navigators for high risk patients not responding to initial outreach. In the control arm, the process of escalating the reminder intervention from a letter, to contact by phone call, to a patient navigator, will occur in a standard algorithmic fashion without provider input. While not yet the standard of care nationwide, prior studies have proven the efficacy of such an approach. In intervention practices, the investigators will enhance augmented standard care by implementing a novel system that will enable physicians and clinical population managers to individualize care for each patient in their panel using tools to classify and organize patients by their clinical attributes. The investigators hypothesize that this personalized identification of patients by both their clinical outcome and clinical process risk status will improve the efficacy and efficiency of resource allocation decisions. The key additions to the health IT system for intervention practices will be: 1) a clinical systems IT platform to organize and present clinical data for each clinician's patient panel, 2) an accessible Web-based tool allowing clinicians (physicians and clinical population managers) to view, organize, and investigate their patient panels, and 3) a simple process where the clinician can make a tailored screening decision and designate the method of clinical intervention based upon the patient's risk profile.

medical informatics, health IT, primary care, screening, cervical cancer, colorectal cancer, breast cancer, mammography, colonoscopy, pap smear

The TOP-CARE intervention will be based on a medical informatics platform that: Identifies all patients eligible for any of the three cancer screening programs Links patients with a specific clinician Offers a visit-independent method for clinicians to review panels of their eligible patients For patients due for one or more cancer screenings, clinicians will access a web-based informatics tool to: Screen their panel based upon risk Defer patients, document exclusions, and update the EHR Order a screening test with patient information material based upon the patient's risk profile and automatically initiate the process of: Informing the patient by letter of the need to schedule a test, educating the patient with respect to the benefits of cancer screening, and properly documenting the transaction in the patient's EHR, or Referral to a patient navigator for patients most likely to benefit from this more intensive approach

In augmented standard care control practices, we will implement a system that includes: 1) a population-based perspective to identify all eligible patients overdue for screening, 2) an automated, centralized process to contact selected patients by letter, 3) a result management system that automatically tracks test scheduling and completion, 4) a web-based, easily accessible tool allowing practice personnel to contact patients not completing testing, and 5) use of patient navigators for high risk patients not responding to initial outreach. In the control arm, the process of escalating the reminder intervention from a letter, to contact by phone call, to a patient navigator, will occur in a standard algorithmic fashion without provider input.

We will be evaluating the system in practices where providers are involved in determining to contact a patient, including the most appropriate outreach method, compared with practices where the system is fully automated without provider input.

Average cancer screening test completion rate over the 1-year follow-up period for each eligible patient in all eligible cancers (breast, cervical, colorectal)

Percentage of patients completing all eligible cancer screening modalities at 1-year follow-up will be compared between study arms

Eligible population: prevalent (overdue at baseline) or incident (become overdue during study) cancer screening rates will be compared between study arms. Patient-physician linkage status: Analyses will be stratified to compare PCP-linked patients between study arms and practice-linked patients between study arms. Patient risk status: Patients at high risk for not completing screening will be compared between study arms.

Percentage of providers (PCPs and clinical population managers) in intervention practices using TOP-CARE system and percentage of eligible patients in clinician panel screened Percentage of providers (practice personnel and patient navigators) in intervention and control practices using TOP-CARE system and percentage of eligible patients assigned to provider who were contacted Percentage of eligible patients in clinician (PCP and clinical population manager) panel deferred and reasons selected

Inclusion Criteria: Breast cancer: Women 42-74 years old Cervical cancer: Women 21-65 years old Colorectal cancer: Women and men 52-75 years old Exclusion Criteria: Breast cancer: History of bilateral mastectomy in their EHR Cervical cancer: History of total hysterectomy in their EHR Colorectal cancer: History of total colectomy in their EHR

Atlas SJ, Zai AH, Ashburner JM, Chang Y, Percac-Lima S, Levy DE, Chueh HC, Grant RW. Non-visit-based cancer screening using a novel population management system. J Am Board Fam Med. 2014 Jul-Aug;27(4):474-85. doi: 10.3122/jabfm.2014.04.130319.

Zai AH, Kim S, Kamis A, Hung K, Ronquillo JG, Chueh HC, Atlas SJ. Applying operations research to optimize a novel population management system for cancer screening. J Am Med Inform Assoc. 2014 Feb;21(e1):e129-35. doi: 10.1136/amiajnl-2013-001681. Epub 2013 Sep 16.