Skip to main content
. 2019 Apr 19;2(4):e192662. doi: 10.1001/jamanetworkopen.2019.2662

Table. Trend Models for ED Visits and Hospitalizations.

Dependent Variable Model Slope 1 (pa) Join Point Slope 2 (pa) P Valueb
Trend models for % of ED visits by insurance type
% Private insurance visits Linear −0.8 (0.00003) NA NA .81
% Medicare visits 1 Join point 2.1 (0.11) 2008 0.06 (0.63) .02
% Medicaid visits Linear 0.9 (0.004) NA NA .18
% Uninsured visits 1 Join point −0.2 (0.11) 2013 −2.1 (0.003) .001
Trend models for No. of ED visits by insurance type
Total No. of ED visits Linear 2.3 (0.0005) NA NA .49
No. of private insurance visits Linear −0.4 (0.07) NA NA .30
No. of Medicare visits 1 Join point 3.4 (0.05) 2008 0.4 (0.02) .02
No. of Medicaid visits Linear 1.8 (0.0007) NA NA .24
No. of uninsured visits 1 Join point 0.04 (0.86) 2013 −2.6 (0.03) .005
Trend models for % of hospital discharges by insurance type
% Private insurance visits Linear −0.5 (0.00008) NA NA .30
% Medicare visits Linear 0.3 (0.0006) NA NA .56
% Medicaid visits Linear 0.4 (0.0003) NA NA .38
% Uninsured visits 1 Join point −0.03 (0.61) 2013 −0.6 (0.04) .02

Abbreviations: ED, emergency department; NA, not applicable.

a

The numbers in parentheses indicate the significance level of alternative hypothesis that the slope does not equal 0 vs null hypothesis that the slope is 0. Slopes indicate change per year in percentage points.

b

Significance of alternative hypothesis of 1 join point vs null hypothesis of 0 join points (straight line). Slopes indicate change per year in percentage points.