Table 1 One-quarter-ahead (and two quarter-ahead) prediction accuracy of different models across states, 2005–2015.

From: Internet search and medicaid prescription drug data as predictors of opioid emergency department visits

 

Google search data

SDUD Rx utilization data

Both Google and drug data

 

R2

RMSE

R2

RMSE

R2

RMSE

CA

0.8736 (0.7844)

638.6979 (834.2924)

0.9062 (0.8282)

550.1519 (744.5870)

0.9594 (0.8969)

361.9566 (576.9079)

FL

0.8059 (0.7200)

97.0510 (116.5707)

0.8489 (0.7370)

85.6269 (112.9681)

0.9271 (0.8375)

59.4883 (88.7939)

GA

0.8697 (0.6800)

16.4743 (25.8185)

0.8409 (0.7387)

18.2066 (23.3316)

0.9127 (0.8278)

13.4825 (18.9402)

IN

0.856 (0.8127)

58.8990 (67.1762)

0.8851 (0.7547)

52.6304 (76.8859)

0.9405 (0.8503)

37.8547 (60.0672)

MA

0.8522 (0.7560)

219.4779 (282.0361)

0.8885 (0.7022)

190.6938 (311.6135)

0.9287 (0.8157)

152.4159 (245.1396)

MN

0.899 (0.7672)

51.6804 (78.4571)

0.8208 (0.7865)

68.8408 (75.1380)

0.9602 (0.8420)

32.4432 (64.6257)

MO

0.8559 (0.7295)

51.0897 (69.9864)

0.8812 (0.7628)

46.3891 (65.5383)

0.9271 (0.8207)

36.338 (56.9789)

NJ

0.8668 (0.7651)

254.2480 (337.6893)

0.9013 (0.7835)

218.8394 (324.1491)

0.9555 (0.8599)

147.0135 (260.8051)

NY

0.7541 (0.7249)

430.4102 (455.1811)

0.8900 (0.7231)

308.49 (489.5146776)

0.9462 (0.8077)

215.769 (407.9654628)

TN

0.8566 (0.7458)

60.3684 (80.3787)

0.8946 (0.7872)

51.7539 (73.5375)

0.9471 (0.8345)

36.6575 (64.8513)

WI

0.8005 (0.7520)

153.2267 (170.8507)

0.8993 (0.7620)

108.8656 (167.3505)

0.9554 (0.8218)

72.4859 (144.8327)