Harvard Pilgrim Health Care

• Most experienced managed care organization in New England; first members enrolled in 1969
• Not-for-profit health plan providing a variety of insurance plan options and self-funding arrangements
• Over 1 million members in MA, NH and ME
• Network of over 135 hospitals and 28,000 doctors/clinicians
• Rated #1 health plan in the United States by NCQA and U.S. News and World Report for the past 3 years

Problem Statement

• Colorectal Cancer (CRC) is the 3rd leading cause of cancer death in U.S. and a leading cause of cancer morbidity among all races/ethnicities
• CRC screening is an effective tool to identify CRC, which is highly curable when detected early
• Removal of precancerous polyps contributed to 10-yr decline in incidence rate of CRC among both men and women
• CRC death rates also declining, likely due to detection at earlier stage
• Despite information on screening and recommendations from experts, many HPHC members not getting screened
• Baseline CRC screening “defect rate” of 34% meant opportunity for improvement among all Harvard Pilgrim members.
• Disparities in CRC screening among Hispanic members and those living in communities with lower education & income levels

Annual Report to the Nation on the Status of Cancer, 1975-2004 appearing in the November 15, 2007 issue of Cancer.

CRC Screening Initiative

• Goals
• Improve Harvard Pilgrim’s Colorectal Cancer (CRC) screening rate overall
• Reduce the observed disparity in CRC screening among Hispanic/Latino members
• Address language and literacy issues that contribute to disparities
• Identify barriers to CRC screening among member sub-groups

Project Team

• Lydia Bernstein, MPH
• Kathryn Coltin, MPH
• Arthur Ensroth, MPH
• Carla Rosenkrans, MPH, RN
• Risa Shames, MPH

CRC Screening Interventions

• Direct-to-member, population based outreach, based on age criteria, beginning in 2004
• High-risk component targeted to Hispanic/Latino members, identified using geocoding and surname coding, added in 2005
• Use IVR to outreach to members
• Personalized interactive voice response (IVR) phone technology using voice only (no touch pad)
• Caller ID boxes display Harvard Pilgrim name/phone
• Provide information about the importance of CRC screening, including culturally appropriate motivational statements for members flagged as Hispanic or Latino
• Collect information on member behavior, including barriers to care
• Direct patients to their PCP for test selection
• Supports Provider-directed initiatives to improve CRC screening
• Defect lists, P4P, Performance by R/E, Honor Roll, Quality Grants

Why IVR calls?

• Rationale
• IVR provides an opportunity to question as well as educate members
• Members may elect to call a toll-free number to hear the information at a time that is more convenient for them.
• Spoken messages may achieve better comprehension by individuals with a low literacy level. Members can ask to have statements and questions repeated as often as necessary.
• Advances in IVR technology enable calls in Spanish language.
• Previous IVR initiatives showed a large percentage of members were reached and stayed on the phone to hear and respond to the messages. (Similar data are not available for mailed outreach.)
• Computer generated messages may be perceived as less threatening than a personal discussion.
• Information reported by patients through IVR is as reliable as that obtained through structured clinical interviews.

Piette JD. Interactive voice response systems in the diagnosis and management of chronic disease. Am J Managed Care. 2000:6:817-827.

History of CRC Screening Interventions


Group & PCP Performance Reporting Tool


Evaluation Methods

• Past success measured by:
• Improvement in overall CRC screening rate relative to baseline and national 90th percentile
• Reduction in disparity observed for Latino members relative to benchmark population and national 90th percentile
• Selected IVR metrics
• Quantitative data from call tracking records (e.g., # members participating in calls, brochure requests)
• Qualitative data from recorded calls (i.e., verbatim comments)
• Future success measured by:
• Percent of all members and Hispanic/Latino members who were reached/not reached via IVR, who remained a Harvard Pilgrim member and received CRC screening within 12 months of their outreach call
• Target population was not engaged in the evaluation

Impact

• Ongoing initiatives resulted in consistent, incremental improvements in overall CRC screening rates since 2003
• Harvard Pilgrim 2005 performance rate increased 3.7 percentage points over 2003 baseline
• Hispanic/Latino members are still less likely to be screened for CRC than benchmark population, but their screening rate improved and their baseline disparity was reduced
• Hispanic/Latino members had 4.7 percentage point improvement in screening rate in 2005 compared to 2003
• Hispanic/Latino members CRC screening rate surpassed the HEDIS national 90th percentile in 2005