Skip to main content

Table 4 Interrupted time series (ITS) segmented regression analysis: difference-in-difference (DID) estimates, comparing treatment (dual eligibles enrolled in D-SNP) vs. comparator (individuals 60–64 enrolled in Medicaid Managed Care) on changes in monthly utilization time trends associated with ACA DSNP Changes

From: The impact of Dual Eligible Special Need Plan regulations on healthcare utilization

 

Post Period (vs. Pre Period)

State

Outcome

DID Level1

P-Value

DID Slope2

P-Value

AZ

Any Hospitalization

− 0.19%

0.702

0.01%

0.710

Any ED Visit

−0.48%

0.461

−0.03%

0.545

NJ

Any Hospitalization

−3.37%

0.022*

−0.23%

0.014*

Any ED Visit

−2.87%

0.057

−0.18%

0.103

TN

Any Hospitalization

0.32%

0.565

−0.01%

0.903

Any ED Visit

−1.38%

0.156

0.05%

0.562

  1. Notes: Linear regression used for utilization outcomes. Sample is person-months from 2010 to 2015. *denotes significance at p < .05. Regression covariates of interest were study Group (treatment vs. comparator); a linear monthly time trend, indicators and splines for the post period (2012–2015); and the interactions between these variables and Group. Other covariates included sex, age, race, indicator for SSI status, and 18 comorbidity indicators. Repeated measures adjusted for using Huber-White robust standard errors, clustering at the person level
  2. 1. Difference between treatment vs. comparator in the discontinuity (change in level) for the post period, measured using the interaction between Group & an indicator variable for the post period
  3. 2. Difference between treatment vs. comparator in the change in slope for the post period, measured using the interaction between Group & a spline variable for the post period
Back to article page

BMC Health Services Research

ISSN: 1472-6963

Contact us