Financial Integration's Impact On Care Delivery And Payment Reforms: A Survey Of Hospitals And Physician Practices

Abstract

Health systems continue to grow in size. Financial integration-the ownership of hospitals or physician practices-often has anticompetitive effects that contribute to the higher prices for health care seen in the US. To determine whether the potential harms of financial integration are counterbalanced by improvements in quality, we surveyed nationally representative samples of hospitals (n = 739) and physician practices (n = 2,189), stratified according to whether they were independent or were owned by complex systems, simple systems, or medical groups. The surveys included nine scales measuring the level of adoption of diverse, quality-focused care delivery and payment reforms. Scores varied widely across hospitals and practices, but little of this variation was explained by ownership status. Quality scores favored financially integrated systems for four of nine hospital measures and one of nine practice measures, but in no case favored complex systems. Greater financial integration was generally not associated with better quality.

Keywords: Financial Integration; Government programs and policies; Hospital quality; Hospitals; Integrated Delivery Systems; Medical research; Organization of care; Payment; Physician practices; Primary care; Quality of care; Systems of care; consolidation; health policy; quality improvement.

Exhibit 3. Comparison of Care Delivery and Payment Scores for Hospitals in Different Ownership Categories

Source / Note: Source: Authors’ analysis of NSHOS Hospital Survey Responses. Note: Exhibit 3 shows estimated scores, adjusted as described in Methods. Vertical lines represent estimated scores; horizontal lines represent 95% confidence intervals for the mean outcome with end-points the 2.5’th and 97.5’th confidence interval limits. Asterisks are used to report tests of significance for each measure on two comparisons: (1) a test of whether independent hospitals differ significantly from all hospitals owned by systems of some type; if significant the asterisk is put next to the label “independent hospitals”; (2) a test of whether there are significant differences across hospitals owned by systems; if significant, the asterisk is placed next to the label “simple systems”. The following convention is used: * p < 0.10 ** p < 0.05 *** p < 0.01 **** p < 0.001

Exhibit 4. Comparison of Care Delivery and Payment Scores for Practices in Different Ownership Categories

Source / Note: Source: Authors’ analysis of NSHOS Practice Survey Responses. Note: Exhibit 4 shows estimated scores, adjusted as described in Methods. Vertical lines represent estimated scores; horizontal lines represent 95% confidence intervals for the mean outcome with end-points the 2.5’th and 97.5’th confidence interval limits. Asterisks are used to report tests of significance for each measure on two comparisons: (1) a test of whether independent practices differ significantly from all practices owned by systems of some type; if significant the asterisk is put next to the label “independent practices”; (2) a three-way test of whether there are significant differences among practices owned by different categories of systems; if significant, the asterisk is placed next to the label “medical groups”. It is possible for individual differences between medical groups and complex systems or between medical groups and simple systems to be statistically significant when the three-way comparison is not significant and vice-versa. The statistical significance of the individual contrasts between medical groups and complex systems and between medical groups and simple systems are given in the tables of the fitted hierarchical regression models in Appendix Table 3.2.b. Although several of these pairwise comparisons were significant, all of these favored medical groups over simple and/or complex systems. The following convention is used: * p < 0.10 ** p < 0.05 *** p < 0.01 **** p < 0.001