Using health-related quality of life assessments to evaluate care support within Medicaid.

2008 
CareOregon, an Oregon-based not-for-profit Medicaid health plan, successfully piloted a "CareSupport" model that identifies high-risk members and clinically stratifies them for intervention. Internal analyses indicate that CareSupport lowers utilization and cost; CareOregon, however, has lacked patient-reported outcome data on the health-related quality of life (HRQL) of CareSupport participants. Between September 2005 and November 2006, we conducted a pilot study in which the Health Utilities Index Mark 3 (HUI3), a generic preference-based measure of health status and HRQL, was integrated into CareOregon's existing screening algorithm for possible admission into CareSupport. We obtained baseline data on 616 CareSupport candidates and 4-month HUI3 follow-up data on 143 candidates (104 CareSupport, 39 non-CareSupport). On a 0.00 (dead)-to- 1.00 (perfect health) scale, the mean overall baseline HUI3 score for CareSupport patients was 0.18 (0.20 for non-CareSupport patients), comparable to baseline means reported elsewhere for much older patients immediately after suffering serious acute medical events, such as stroke or hip fracture. A 0.05 mean 4-month improvement in overall HRQL among CareSupport enrollees relative to non-CareSupport enrollees was clinically important but not statistically significant. A 0.10 improvement in HUI3 emotion was both statistically significant and clinically important. Study results provide good preliminary evidence of the value of patient-reported outcomes in clarifying individual illness burden and assessing intervention effectiveness. Keywords: care management; utility; outcomes; feasibility Started in 1994, CareOregon is a not-for-profit Medicaid health plan in Oregon serving nearly 90,000 low-income enrollees who do not have employment-based health insurance. Since 2004, CareOregon has focused on improving the outcomes of its highest-risk members (i.e., those with multiple comorbidities, complex psychosocial and environmental risk factors, and consequently unstable or "brittle" health status). It has successfully piloted a care support model (CareSupport) that identifies highrisk members in advance of a health care crisis and clinically stratifies them for the intervention program. CareSupport builds on the Chronic Care Model (Wagner et al., 2001) of evidence-based, proactive, and maximally informed care to explicitly account for the complexity of multiple chronic conditions and provider types experienced by high-risk patients. Internal analyses of administrative data indicate that CareSupport lowers utilization and CareOregon's costs. From 2004 to 2005, CareSupport participants reduced the hospitalization rate by 43%, compared with growth of nearly 11% for members without CareSupport. Short-term (i.e., less than 1 month) CareSupport participants reduced hospitalizations by 17% over the same period. CareOregon officials estimate that fewer hospitalizations saved over $20,000 per high-risk member during that period. They further estimate that overall return on investment in CareSupport over this period was $5.9 million. Although CareOregon has captured cost data, it has lacked patient-reported outcome data on the health status and healthrelated quality of life (HRQL) of CareSupport participants. This article reports the results of a pilot study in which the Health Utilities Index Mark 3 (HUI3), a generic preference-based measure of health status and HRQL (Furlong, Feeny, Torrance, & Barr, 2001; Horsman, Furlong, Feeny, & Torrance, 2003), was integrated into CareOregon's existing screening algorithm for possible entry into CareSupport. CareOregon's Quality Improvement Management Committee and Kaiser Permanente Northwest's Institutional Review Board approved this study. STUDY SETTING The CareOregon network includes 950 primary care providers, 3,000 specialists, 33 hospitals contracted statewide, and 14 public health departments. Approximately 85% of CareOregon's membership lives in the Portland metropolitan area. …
    • Correction
    • Source
    • Cite
    • Save
    • Machine Reading By IdeaReader
    0
    References
    2
    Citations
    NaN
    KQI
    []