---

PURPOSE: This investigation sought to determine: (1) how nursing patient-centeredness influenced overall Medicaid obstetrical patient satisfaction, their likelihood to recommend the hospital, and their ratings of care, (2) whether the effects were equal across three national random samples of patients, and (3) the variance of obstetrical patient satisfaction explained by nursing patient-centeredness.  METHODS: A two-factor multigroup structural equation modeling design was used, with cross-group constraints to test the equality of effects across groups.  The hypothesized model was also compared with a competing model.  RESULTS: The model fit well.  Nurse patient-centeredness significantly influenced patient satisfaction, explaining 67 percent of its variability. When patient-centeredness increased by one unit, overall satisfaction increased by .790 units (.820 standardized), the likelihood to recommend the hospital by .735 (.727 standardized) and ratings of care by .728 units (.762 standardized).  These effects held across three random samples of patients.  The model was sustained when compared with a competing model.  Patient-centered nursing behaviors like friendliness and courtesy, making sure patients are informed, a positive attitude toward requests, and responsiveness to special requests all increased satisfaction with nursing.  DISCUSSION: The study offers a paradigm for measuring and improving overall Medicaid obstetrical patient satisfaction, their ratings of care, and their likelihood to recommend the hospital. The results have implications for evidence-based outcome measurement, quality, satisfaction improvement interventions, and nursing education.