Skip to main content

Table 3 Intent-to-treat (ITT) analysis: primary and secondary outcomes between the personalised and non-personalised groups

From: The effect of mobile personalised texting versus non-personalised texting on the caries risk of underprivileged adults: a randomised control trial

ITT analysis

Group

  

Personalised (n = 56)

Non-personalised (n = 55)

 

p value

Primary outcome (‘chance of avoiding new cavities’)

Mean difference (95% CI)

 

 Baseline

  mean (SD)

39.3 (20.2)

36.5 (23.4)

  

  median (min. to max.)

37.5 (6 to 81)

31.0 (3 to 94)

  

 Follow-up

0.7 (−5.5, 6.9)

p = 0.820

  mean (SD)

46.2 (19.6)

42.8 (22.0)

  median (min. to max.)

44.5 (8 to 83)

41.0 (9 to 93)

Secondary outcome (number (%) of participants with Score 0, 1)

Odds ratio (95% CI)

 

‘Diet frequency’

 Baseline

39 (69.6)

36 (65.5)

  

 Follow-up

47 (83.9)

43 (78.2)

0.8 (0.3, 2.3)

p = 0.663

‘Diet contents’

 Baseline

27 (48.2)

30 (54.5)

  

 Follow-up

27 (48.2)

30 (54.5)

1.0 (0.4, 2.6)

p = 0.945

‘Plaque amount’

 Baseline

25 (44.6)

19 (34.5)

  

 Follow-up

31 (55.4)

33 (60.0)

1.7 (0.7, 3.9)

p = 0.247

‘Mutans streptococci’

 Baseline

34 (60.7)

24 (43.6)

  

 Follow-up

36 (64.3)

31 (56.4)

1.1 (0.4, 2.6)

p = 0.917

‘Fluoride programme’a

 Baseline

55 (98.2)

51 (92.7)

  

 Follow-up

56 (100.0)

54 (98.2)

 

p = 0.941

‘Saliva secretion’

 Baseline

45 (80.4)

40 (72.7)

  

 Follow-up

51 (91.1)

41 (74.5)

0.3 (0.1, 0.9)

p = 0.036*

‘Saliva buffer capacity’

 Baseline

54 (96.4)

51 (92.7)

  

 Follow-up

45 (80.4)

40 (72.7)

0.8 (0.3, 2.1)

p = 0.653

  1. *p < 0.05
  2. ITT intent-to-treat, SD Standard deviation, CI Confidential interval
  3. aModel fit was questionable – odds ratio estimates unreliable
  4. The primary outcome is a comparison of ‘chance of avoiding new cavities’ calculated by the Cariogram. The secondary outcome measures are the seven biological risk parameters out of the ten risk parameters in the Cariogram. Scores 0 and 1, and Scores 2 and 3 (if any) are combined as ‘lower score’ and ‘higher score’, respectively. The table indicates number (%) of participants with ‘lower score’