Skip to main content
. 2017 Sep 5;177(9):1334–1342. doi: 10.1001/jamainternmed.2017.2775

Table 4. Mean Outcomes and Adjusted Difference-in-Differences Estimates for Hospitalizations, Emergency Department Visits, and Medicare Spending for All Medicare Patients and High-Risk Medicare Patients.

Patients Intervention Group Comparison Group Adjusted Difference-in-Differences (90% CI)a P Value
All Medicare FFS Patients
Medicare patients per quarter, mean No. 34 364 68 854 NA NA
Hospitalizations, rate per 1000 patients per quarter
All-cause
Preintervention year 79.4 78.6 NA NA
Intervention mo
1-12 75.4 71.9 1.4 (−2.4 to 5.3) .54
13-30 71.5 68.9 1.4 (−2.1 to 5.0) .51
Ambulatory care-sensitive
Preintervention year 13.0 12.8 NA NA
Intervention mo
1-12 11.5 11.3 −0.1 (−1.4 to 1.2) .89
13-24b 11.6 11.0 0.4 (−0.9 to 1.7) .61
Outpatient ED visits, rate per 1000 patients/quarter
Preintervention year 79.2 80.8 NA NA
Intervention mo
1-12 81.2 80.6 2.2 (−1.5 to 6.0) .33
13-30 83.0 87.8 −2.5 (−6.2 to 1.1) .26
Medicare Part A and B spending, $/patient/per mo
Preintervention year 989 976 NA NA
Intervention mo
1-12 987 949 12 (−27 to 52) .61
13-30 1006 987 −1 (−40 to 39) .98
High-Risk Medicare FFS Patients
Medicare patients per quarter, mean No. 10 036 20 839 NA NA
Hospitalizations, rate per 1000 patients/quarter
All-cause
Preintervention year 162.3 158.0 NA NA
Intervention mo
1-12 149.2 140.7 2.8 (−7.1 to 12.7) .64
13-30 142.0 137.2 0.2 (−9.4 to 9.7) .98
Ambulatory care-sensitive
Preintervention year 31.8 31.1 NA NA
Intervention mo
1-12 27.9 26.7 0.6 (−3.1 to 4.2) .80
13-24b 28.7 27.5 0.9 (−2.9 to 4.7) .92
Outpatient ED visits, rate per 1000 patients/quarter
Preintervention year 133.7 137.5 NA NA
Intervention mo
1-12 133.8 129.9 8.3 (−0.9 to 17.5) .14
13-30 137.7 147.9 −4.6 (−14.4 to 5.1) .44
Medicare Part A and B spending, $/patient/mo
Preintervention year 1851 1822 NA NA
Intervention mo
1-12 1789 1728 10 (−89 to 110) .87
13-30 1808 1773 −11 (−115 to 92) .86

Abbreviations: ED, emergency department; FFS, fee-for-service; NA, not applicable.

a

The regressions adjusted for the same variables indicated in Table 3.

b

To ensure comparability of measure definition throughout the study period, we did not construct rates of ambulatory care-sensitive conditions after October 2015, when Medicare claims began using the 10th edition (instead of the 9th) revision of International Classification of Diseases (ICD). The other outcomes do not rely on ICD codes and so could be constructed through the end of the program.