A Cost-Effectiveness Analysis of thePenn Colorectal Cancer ScreeningNavigation ProgramLeonard Davis Institute of Health EconomicsSummer Undergraduate Minority Research ProgramAugust 15, 2019Khalid El-JackBoston UniversityMentor: Carmen Guerra, M.D., M.S.C.E.

Outlinew Discuss the disparities in colorectal cancer screeningsbetween populationsw Define patient navigation and its effectivenessw Describe Penn Colorectal Cancer Screening NavigationProgramw Determine cost-effectiveness of the programw Elaborate on lessons learned2

Colorectal Cancer (CRC) Backgroundw Third most commonlydiagnosed cancer in the U.S.(ACS 2011)w Usually develops slowly over10-15 yearsw Screening is universallyrecommended by guidelinesfor all age 50 (ACS 2018).w Racial/ethnic minorities, nonEnglish speakers and low incomeindividuals have lower rates ofscreening and present with moreadvanced stage cancer.Source:

Cancer Disparities in African AmericanswAfrican Americans are more likely tobe diagnosed and die from colorectalcancer than whitesw 19% higher mortality rateSource: American Cancer Society. Cancer Facts and Figures for African Americans2013-14. Atlanta: American Cancer Society, 2019-214

Stage and mortality are closely linkedwAt 5 years, survival forwCRC: Stage I – 90% vs. Stage IV – 10%wThus, early detection of cancer throughscreening is critical to improving survivalwYet screening is underutilizedwPA CRCS rate: 66.8%Source: BRFSS 20145

Why were you unable to complete your previousCRCS appointment?w “I don’t have medical insurance.”w “I was a little scared.”w “I don’t have a family history, so I don’t feel it is highlyurgent.”w “There is a possibility that they will puncture myintestine.”w “I just never did it.”6

What is a Patient Navigator?w An individual that guides patients through procedures ortreatmentsw Meets with patients directly and/or communicates viatelephonew Identifies and addresses barriers to carew Mitigates patient concernsw Goal: to ensure completion of screening, timelydiagnosis, and treatment7

History of Patient Navigationw Started in HarlemHospital in New YorkCity in 1990 by Dr.Harold Freemanw Aimed at helpingpoor Hispanic andAfrican-AmericanpatientsDr. Harold Freemanw Targeted increasingbreast cancersurvival rates8

Navigation Project Aimsw Clinical Goal Launch and evaluate a Patient Navigation Programat the University of Pennsylvania Health System(UPHS) specifically related to Colorectal CancerScreening (CRCS) of West Philadelphia Patientsw Research Goals Better understand the barriers to CRCS in the WestPhiladelphia population Improve CRCS rates in West Philadelphia, especiallyamong African- Americans Determine patients’ levels of satisfaction with thenavigation program9

Project Populationw Inclusion Criteria West Philadelphiaresident Over 50 years old Doctor had to orderCRCS for patient Appointment was neverscheduled or patientfailed to show up forhis/her appointmentSource:

West Philadelphia CRCS Patient Navigation Programw Hired an MA to serve as patient navigatorw Training at the Harold Freeman PatientNavigation Institute, Bronx, NYw Used grant funding for programimplementation, patient care costs, MA salaryand trainingw Created low literacy version of prepinstructions and video

West Philadelphia Colorectal Cancer Screening Navigation Program Results 90% of patients screened were African American Completed 763 Screening Colonoscopies 327 colonoscopies resulted in at least oneadenomatous polyp (42.9%) Detection of 5 cancer casesScreening colonoscopy results(n 763)Normal/no pathology or hyperplasticpolyp(s)353 (46.3%)At least one adenomatous polyp327 (42.9%)Adenocarcinoma5 (0.7%)Repeat16 (2%)Other30 (4%)Pending scheduling32 (4%)CRC StageNStage IStage IIStage IIIStage IVTotal10315

Patient SatisfactionPatient Satisfaction (n 180)Overall, I am satisfied with the navigationservices I received from the navigatorStrongly agreeAgreeNeither Agree or disagreeDisagreeStrongly disagree168 (93.3)11 (6.1)01 (0.6)0

Conclusions about Navigation Programw A patient navigation program for CRCS forUPHS patients who are residents of WestPhiladelphia and have not previously beenable to complete screening colonoscopy isw Feasiblew Acceptablew Effectivew Associated with high patient satisfactionw Reduced colonoscopy no showsw Builds Trust1

SUMR Project Specific Aimsw To determine if the navigation program iscost-effective relative to a 2010 controlgroupw Hypothesis: A CRCS navigation programfocused on the West Philadelphiapopulation is Cost-Effective1

Colonoscopy Data Collection and Categorizationw To gather colonoscopy data, I manually abstracted datafrom a cohort of patient colonoscopy and pathology resultsin EPIC who meet the inclusion criteriaw I then categorized the outcomes as normal, benign, orabnormal based on EPIC Reportw Notable outcome was that the navigated group had a higheradenoma detection rate (38%) vs. a control group (27%)NavigatedGroup (2012)Outcomes of Completed SCs“Normal” SCBenign Pathologyi“Abnormal” Pathology-Adenoma Detection Rateii-Cancer Incidencei.ii.(n 134)41%19%40%38%2%Includes colonic mucosal polyps and hyperplastic polypsIncludes serrated adenomas and tubular adenomasControl Group(2010)(n 366)48%23%28%27%1%1

Costs of Navigation Programw Largest driver of costs is labor cost of patient navigator( 64,531)w Overall navigation program costs in 2012 totaled 76,666w The average total cost of each completed screeningcolonoscopy for a navigated patient was 703.36w These costs are not reimbursed by insurers despite proveneffectiveness of patient navigationOutputsPer Completed SC-Navigated patients (n 109)-All patients (n 132)i.Average TotalCosti (USD) 703.36 580.80Calculated based on 2012 program operating costs (including start-up costs).1

Cost-Effectiveness AnalysisTable 5: Cost-Effectiveness Analysis, Calendar Year 2012w Incremental Revenuerange comes fromMedicare data atUPHSw Using a 2010 controlgroup, we projectedan additional 30patients would bescreened if navigatedIncremental Revenue,RangeIncremental Cost PerCompleted SCPatients ScheduledPatients Screened% Patients Screened 703.3664436656.8%13810979%2012 Projections, UsualCare% Patients Screened (Proj.)Patients Screened (Proj.)w This resulted in a netbenefit range of( 282.65) to 138.06per navigated patientscreenedControl Group Received Navigation(2010)(2012) 1,500.00 to- 3,000.00Additional PatientsScreenedTotal Additional Revenue,RangeTotal Additional CostNet Benefit, RangeNet Benefit, Range PerPatient Screened (n 109)56.8%7830.6 45,857.14 to 90,428.57 76,666.24- 30,809.10 to 15,048.05- 282.65 to 138.061

Project Conclusionsw In higher revenue ranges, the CRCS navigationprogram is cost effectivew Some limitations to consider are the use of asingle hospital network and potential presence forself-selection biasw Dissemination of project data must be presentedto insurance companies to influence policychange for navigation reimbursement1

Lessons Learnedw Abstraction of data from EMRw Categorization of colonoscopies based on EMR resultsw The importance of patience and flexibility in researchw How to lead a project as a co-investigatorw The community impact of dedicated physician-researchers20

Acknowledgementsw Joanne Levy, MBA, MCPw Carmen Guerra, M.D.,M.S.C.E.w SUMR staff and otherscholars21

A Cost-Effectiveness Analysis of the Penn Colorectal Cancer Screening Navigation Program Leonard Davis Institute ofHealthEconomics Summer Undergraduate Minority ResearchProgram August 15,2019 Khalid El-Jack Boston University Mentor: Carmen Guerra, M.D.,M.S.C.E. Outline wDiscuss the dispar