Insurance policies and perceived quality of primary care among privately insured patients: do features of managed care widen the racial, ethnic, and language-based gaps?

CONTEXT

Little is known about whether some features of managed care widen disparities in patients' evaluations of primary care.

OBJECTIVES

We investigated whether the magnitudes of racial and ethnic/language-based differences in patients' evaluations of the quality of primary care vary by capitation and gatekeeping.

DESIGN

We used a telephone survey of a representative sample of the US noninstitutionalized population, Community Tracking Study Household Survey 1998-1999, and Followback Survey of respondents' insurance administrators.

SETTING AND PARTICIPANTS

Our sample was privately insured adults who saw a physician at least once during the year preceding the interview and whose last visit was to a primary care physician.

MAIN OUTCOME MEASURES

We measured patients' evaluations of (1) how well the physician listened, (2) how well the physician explained, and (3) how thorough and careful the physician was during the last visit.

RESULTS

Significant white-minority differences emerge more often in plans using capitation or gatekeeping than in other plans. The gaps in patients' evaluations of their primary care providers' (PCP) explanations and thoroughness between whites and Hispanics interviewed in English are larger when the PCP is capitated than when the PCP is not capitated. The gap in the evaluations of their PCP's explanations by whites and Hispanics interviewed in English is larger in plans that require referrals for specialist visits than in other plans. The magnitude of racial and ethnic/language-based gaps for Hispanics interviewed in Spanish, blacks, and Native American/Asian/Pacific Islanders do not differ by capitation and gatekeeping.

CONCLUSION

English-speaking Hispanics' perceptions of the quality of primary care may be more dissimilar from whites' when capitation or gatekeeping are used than when these policies are not used.