310 likes | 484 Views
Randomized Comparison of Organized FIT Invitation, Organized Colonoscopy Invitation, and Usual Care for Colorectal Cancer Screening Among the Underserved . Authors
E N D
Randomized Comparison of Organized FIT Invitation, Organized Colonoscopy Invitation, and Usual Care for Colorectal Cancer Screening Among the Underserved Authors Samir Gupta, Marcia Hammons, Luisa Valdez, Elizabeth Carter, Mark Koch, Liyue Tong, ChulAhn, Don C. Rockey, JasminTiro, Ethan A. Halm, CeletteSuggSkinner Partnering Institutions John Peter Smith Health System, Harold C. Simmons Cancer Center, Moncrief Cancer Institute, UT Southwestern Medical Center Grant Support Cancer Prevention and Research Institute of Texas
Colorectal Cancer (CRC) is an Important Public Health Problem • 2ndleading cause of cancer death nationwide • Screening can reduce CRC mortality • Fecal occult blood testing, sigmoidoscopy, and colonoscopy • Mandel N Engl J Med. 2000 Nov 30;343(22):1603-7; KahiClinGastroenterolHepatol. 2009 Jul;7(7):770-5; Brenner J NatlCancerInst. 2010 Jan 20;102(2):89-95; Brenner J ClinOncol. 2011 Oct 1;29(28):3761-7; AtkinLancet. 2010 May 8;375(9726):1624-33; Scholefield Gut. 2002 Jun;50(6):840-4; KronborgLancet. 1996 Nov 30;348(9040):1467-71; Mandel J NatlCancerInst. 1999 Mar 3;91(5):434-7; Baxter Ann Intern Med. 2009 Jan 6;150(1):1-8; ManserGastrointestEndosc. 2012 Apr 11. [Epubaheadofprint].
Screening Participation is Substantial, but Suboptimal • Screening has been promoted in the US for over 15 years, and steady gains have been realized • National screening rate is >55% • However, not all populations have benefited from these gains • Uninsured • Pre-Medicare age • Medicaid • Minorities MMWR MorbMortalWklyRep. 2012 Jan 27;61(3):41-5. Klabunde Cancer Epidemiol Biomarkers Prev. 2011 Aug;20(8):1611-21
Two Key Challenges to Improving Screening for the Underserved • Identifying the unscreened • Determining which test or tests to offer
Challenge 1: Identifying the Unscreened • In the US, most screening is primary care visit-based • Uninsured/underserved have limited access • No visit, no identification of need, no screening offer • Recent NIH State of the Science Conference on Enhancing CRC Screening emphasized need to develop methods to identify unscreened underserved/uninsured individuals SteinwachsAnn Intern Med. 2010 May 18;152(10):663-7.
Challenge 1: Identifying the Unscreened • Potential solution is to leverage relationships safety-net systems have with the underserved • Care for uninsured, Medicaid, and minority groups • Have readily available administrative claims data that can be used to: • Measure and track screening rates • Individually identify the unscreened for interventions to boost screening
Challenge 1: Identifying the Unscreened • We tested and validated this approach at John Peter Smith Health System, the safety net health system serving Fort Worth and Tarrant County, Texas and found: • Screening rate far below the national average: 22% • 16,000 unscreened patients could be individually identified • Positioned us to test interventions to boost screening Gupta S et al. Cancer Epidemiol Biomarkers Prev. 2009 Sep;18(9):2373-9; Gupta et al. Am J Med Sciat press; Marquez E, Gupta S, CryerB. ClinGastroenterolHepatol. 2011 Feb;9(2):106-9.
Challenge 2: Determining Test Type to Use • Could recommend a “colonoscopy first” strategy for all underserved patients • Expensive, infrastructure required substantial • Does not take into account potential for test-specific differences in participation • Fecal immunochemical testing (FIT), CT colonography, and colonoscopy may have different rates of participation • Understanding test-specific differences is critical • Test specific participation rates may be more important that test-specific sensitivity for CRC • Possible that “Best test is the one that gets done” ZauberAnn Intern Med. 2008 Nov 4;149(9):659-69; Gupta Ann Intern Med. 2009 Mar 3;150(5):359; Marquez E, Gupta S, CryerB. ClinGastroenterolHepatol. 2011 Feb;9(2):106-9; Gupta Lancet Oncol. 2012 Mar;13(3):e90.
The Two Challenges Offered an Opportunity to Increase CRC Screening at a Safety Net • Baseline screening rates were far below national average, at just 22% • Local data could be leveraged to individually identify the unscreened for interventions to boost screening • Uncertainty regarding best test or tests to offer • Compelled us to develop an intervention that could: • Boost screening, addressing barriers such as infrequent access to care • Determine which test would result in the highest screening rate for the population: FIT vs. colonoscopy Gupta S et al. Cancer Epidemiol Biomarkers Prev. 2009 Sep;18(9):2373-9; Marquez E, Gupta S, CryerB. ClinGastroenterolHepatol. 2011 Feb;9(2):106-9.
Aims Among uninsured patients, not up-to-date with screening, to: • Determine if a organized outreach program boosts screening compared to usual care • Determine if organized outreach offering a fecal immunochemical test is more effective at boosting screening participation compared to organized outreach offering free colonoscopy
Methods - Design Randomized controlled trial • Usual Care • Organized outreach invitation to either FIT or colonoscopy • Mailed invitation, with information on screening • English/Spanish, low literacy • FIT kit or phone number to schedule colonoscopy • FIT one sample • Telephone reminders (automated and live) • Assistance with test completion and guideline appropriate follow up • Clarified FIT process, colonoscopy scheduling, prep, and follow up
Methods - Design Inclusion Criteria • Age 54 to 64 • >1 primary care visit in last year • Uninsured, but enrolled in medical assistance program Exclusion Criteria • Up to date with CRC screening, based on: • FOBT within 1 year, sigmoidoscopy or barium enema within 5 years, colonoscopy within 10 years • Prior CRC, inflammatory bowel disease, or polyps • Missing address/phone number
Design - Analysis • Primary outcome was screening participation, one year after randomization • Intension to screen analysis • Secondary outcomes include: • Rate of lesion detection/patient invited • Costs • Sample size/power • Based on maximizing screening delivery given local colonoscopy capacity • Planned to assign n=480 to colonoscopy, n=1600 to FIT, and > n=1600 to usual care • >90% power to detect differences of >10%, alpha=0.025
A Waiver of Informed Consent was Obtained • Interventions an adjunct to, rather than a replacement for usual care • Enhances interpretation and generalization of results • Requiring consent would have enrolled patients predisposed to complete screening • Reflects “real world” response to interventions • Fits with concept of comparative effectiveness trials • Approved by UT Southwestern and JPS Institutional Review Boards ClinicalTrials.govID# NCT01191411
Results Assessed for eligibility (n=12,295) • Excluded (n=6,301) • Screening up-to-date (n=1,573) • No recent primary care visit (n=1,217) • Prior polyps, IBD, or CRC (n=1,905) • Missing address/phone number (n=112) • Age (n=836) • 1° language not English/Spanish (n=658) Randomized (n=5994) Colonoscopy n=480 FIT n=1600 Usual Medical Care n=3914
Comparison of Usual Care & Organized Outreach Usual Care (n=204/3914) Organized Outreach (n=665/2080) p<0.0001
Comparison of Usual Care, Organized Colonoscopy, & Organized FIT* Usual Care (n=204/3914) Organized Colonoscopy (n=87/480) Organized FIT (n=578/1600) *p<0.0001 all comparisons
Results – Neoplasia Detected • Neoplasia detection rate appears higher for colonoscopy, but: • FIT achieved similar results with a much lower colonoscopy rate • Results reflect one time screening
Conclusions • Organized outreach is promising for boosting screening for large populations • May be particularly effective for underserved populations, such as the uninsured and minorities • Screening participation rates may be highly test- specific • Differences may be large enough to overcome differences in test-specific sensitivity for neoplasia • Overall, rapid improvement in screening rates is achievable for the underserved • We screened 665 patients, detected 21 patients with advanced polyps, and detected 4 patients with CRC
Acknowledgements Grant Support/Other Support • Cancer Prevention and Research Institute Grant PP100039 (Gupta, Project Director) • Becky Garcia, Ramona Magid • Polymedco Corporation • MoncriefCancer Institute • Harold C. Simmons Cancer Center JPS Partners • Administration: Robert Earley, Gary Floyd, David Salsberry, Bobby Miller, Sue Crabtree, Patty Angell, Stacy Boatman, Teresa Carver, Jay Haynes • Clinical Lab: Lonnie Dear, Donna Flowers, Mark Sackovich, Melissa Ruperto, Stephanie Zitrick • Endoscopy Lab: Donna Franklin, Rohan Clarke, ShilpaMadadi, Sangameshwar Reddy • Family Medicine and GI Clinics: Maria Asprilla, Rachel Stewart, Alba Perez Smith, Kathy Cabello • IRB: Josephine Fowler, KarshenaValsin, Danielle Ramirez, ElieChoufani, Hao Wong Outreach Staff • Marcia Hammons • Luisa Valdez Data Management • Dawn Houser • Adam Loewen • Wes Senter Collaborators • Elizabeth Carter • Mark Koch • Keith Argenbright • CeletteSugg Skinner • Don C. Rockey • Ethan A. Halm • JasminTiro • Michael Kashner • ChulAhn • LiyueTong
Best Test is the One that Gets Done: Thought Exercise • Colonoscopy is 95% sensitive for CRC (one-time) • Fecal Immunochemical Testing is 70% sensitive for CRC (one-time) • 20,000 individuals to screen • 1% have CRC (n=200) • Should we implement population screening with colonoscopy or FIT? Parekh M Ail PharmTher. 2008 Feb; 27: 697-712. Woolf SH Ann Fam Med. 2005 Nov-Dec;3(6):545-52.
At Equivalent Participation, Colonoscopy Detects the Most CRCs 95% Sensitivity Individuals with CRC Detected CRC Screening Test Participation Rate
With Substantially Higher Participation, a FIT Strategy Might Have Similar--Even Superior--CRC Detection Individuals with CRC Detected CRC Screening Test Participation Rate
With Substantially Higher Participation, a FIT Strategy Might Have Similar--Even Superior--CRC Detection Individuals with CRC Detected CRC Screening Test Participation Rate