Skip to main content

Table 7 Sensitivity analyses

From: Effects of the Informed Health Choices podcast on the ability of parents of primary school children in Uganda to assess the trustworthiness of claims about treatment effects: one-year follow up of a randomised trial

 

Control group

Podcast group

Adjusted odds ratioa

Adjusted differencea

One year after listening to the podcast

 Mean score

  Primary analysis

Mean score 52.6%

(SD 20.4%)

Mean score 58.9%

(SD 20.6%)

 

Mean difference: 6.7%

(95% CI 3.3% to 10.1%)

p < 0.0001

  Lee bounds

   

6.2% to 6.7%

(95% CI 1.8% to 9.3%)

Initially after listening to the podcast

 Mean score

  Primary analysis

Mean score 52.4%

(SD 17.6%)

Mean score 67.8%

(SD 19.6%)

 

Mean difference: 15.5%

(95% CI 12.5% to 18.6%)

  Excluding participants lost to 1-year follow up

Mean score 53.0%

(SD 17.9%)

Mean score 67.6%

(SD 19.7%)

 

Mean difference: 14.9%

(95% CI 11.7% to 18.1%)

 Passing score

  Primary analysis

37.7% of parents

(n = 103/ 273)

70.5% of parents

(n = 203/288)

4.2 (95% CI 2.9 to 6.0)

p < 0.0001

34.0% more parents

(95% CI 26.2% to 40.7%)

  Excluding participants lost to 1-year follow-up

39.6% of parents

(n = 101/ 255)

69.8% of parents

(n = 187/268)

3.8 (95% CI 2.6 to 5.5)

p < 0.0001

31.5% more parents

(95% CI 23.4% to 38.6%)

  1. a Adjusted for the stratification variables (education and child’s study group in the Informed Health Choices primary school trial). The odds ratios from the logistic regressions for passing scores have been converted to differences based on the intervention school proportions and the odds ratios calculated using the intervention schools as the reference (the inverse of the odds ratios shown here)