Significant social events and increasing use of life-sustaining treatment: trend analysis using extracorporeal membrane oxygenation as an example

Table 2 Linear regression analyses for the trends of extracorporeal membrane oxygenation use over time

	Coefficients^afor incidence rate		Durbin-Watson tests for autocorrelation^d		Joinpoint regression analysis
	β (95% CI^c)	p value	Durbin-Watson Statistic	Status	Joinpoint
Taiwan’s CAIR^b	2.8058 (1.6925–3.9192)	<0.001	1.014	Indecision	One joinpoint at 2006
The World’s CAIR	0.0075 (0.0016–0.0135)	0.019	1.485	No autocorrelation	No joinpoint

^a“Coefficients” indicates changes in incidence rates per million people when the time changed by one year.
^bCAIR denotes for “crude annual incidence rate,” which means the annual of ECMO cases per million persons registered in National Health Insurance in the year.
^cCI denotes for “confidence interval”.
^dFor 5% significance points, levels below 0.879 indicate significant positive autocorrelation; levels between 0.879 and 1.320 indicate zone of indecision; levels between 1.320 and 2.680 indicate no autocorrelation; levels between 2.680 and 3.121 indicate zone of indecision; levels above 3.121 indicate significant negative autocorrelation.

ISSN: 1472-6939