Skip to main content

Table 2 The Monte Carlo mean (standard deviation) of the final HbA1c outcomes (Y) and cost per subject and the probability of selecting each AI as optimal (poptAI) for SMART and RCT of sample size n = 100. The simulation size was 10,000. According to the data generation model of the simulation study, AI2 is the truly optimal AI

From: Sequential Multiple Assignment Randomized Trial (SMART) to identify optimal sequences of telemedicine interventions for improving initiation of insulin therapy: A simulation study

 

SMART

RCT

Overall

Y

8.23 (0.17)

8.23 (0.17)

Cost (USD)

343.32 (2.14)

343.22 (2.16)

By (embedded) AI

YAI1

8.12 (0.28)

8.12 (0.34)

YAI2

8.04 (0.26)

8.05 (0.33)

YAI3

8.39 (0.29)

8.39 (0.34)

YAI4

8.36 (0.28)

8.36 (0.34)

poptAI1

29.41%

33.77%

poptAI2

48.73%

43.38%

poptAI3

10.06%

10.43%

poptAI4

11.80%

12.42%