2025_Daniels_ortho

Author

Sergio Uribe

CREATED

April 10, 2025

UPDATED

May 4, 2025

1 Packages

2 DATA PREPARATION

[1] "/home/sergiouribe/Insync/sergio.uribe@gmail.com/Google Drive/Research Drive/2025_Daniels Belajev_Institute_Orthodontics_Jakobsone/data 25.03.2025"

2.1 Name changes:

2.2 Data cleaning

Convert labelled values:

2.2.1 Transform variables: Descriptive

(Why still unknown is shown???)

ahh, is a missing value!

2.2.2 Transform variables: Doctor perception

Fits_6mo and Fits_12mo (merge rare levels into “Poor fit”

Patient_missed_6mo and _12mo

Parents_involved_6mo and _12mo

Shows_interest_6mo and _12mo

Poor_hygiene_6mo and _12mo

Uses_rubberbands_6mo and _12mo

Shows_dissatisfaction_6mo and _12mo

3 MODELLING

4 TABLE 1

Characteristic N = 531
Gender
    Vīriešu 16 (30%)
    Sieviešu 37 (70%)
Age 34 (25, 39)
Occupation_type
    Student 6 (11%)
    Employed 41 (77%)
    Unemployed 3 (5.7%)
    Other/Unknown 3 (5.7%)
Education_level
    Primary education 1 (1.9%)
    Secondary education 11 (21%)
    Vocational secondary 6 (11%)
    Higher education 35 (66%)
Household_size
    With parents 8 (15%)
    With partner 27 (51%)
    Alone 12 (23%)
    Other 6 (11%)
Doctor_communication
    Somewhat agree 19 (36%)
    Strongly agree 31 (58%)
    Other opinion 3 (5.7%)
Family_informed
    Strongly agree 49 (92%)
    Other opinion 4 (7.5%)
Family_pressure
    Strongly disagree 36 (68%)
    Somewhat disagree 6 (11%)
    Neutral 8 (15%)
    Other opinion 3 (5.7%)
Doctor_pressure
    Strongly disagree 25 (47%)
    Somewhat disagree 12 (23%)
    Neutral 10 (19%)
    Other opinion 6 (11%)
Financial_tracking
    Strongly agree 19 (36%)
    Somewhat agree 18 (34%)
    Neutral 11 (21%)
    Other opinion 5 (9.4%)
Extra_budget_money
    Somewhat agree 19 (36%)
    Other opinion 34 (64%)
Psych_emotional_symptoms
    No 49 (92%)
    Yes or skipped 4 (7.5%)
BFI_Conscientiousness 55 (50, 61)
BFI_Fake 11.00 (8.00, 13.00)
BFI_Neuroticism 41 (34, 46)
1 n (%); Median (Q1, Q3)

Save as Table 1

4.1 Motivation

4.1.1 Table data doctor

Characteristic N = 531
Aligns_12mo 39 (74%)
Fits_6mo
    Perfect fit 33 (62%)
    Slight gap (1–2 teeth) 19 (36%)
    Poor fit 1 (1.9%)
Fits_12mo
    Perfect fit 29 (55%)
    Slight gap (1–2 teeth) 14 (26%)
    Poor fit 10 (19%)
Patient_missed_6mo
    Never 41 (77%)
    Missed appointments 12 (23%)
Patient_missed_12mo
    Never 38 (72%)
    Missed appointments 15 (28%)
Parents_involved_6mo
    Never 46 (87%)
    Sometimes involved 7 (13%)
Parents_involved_12mo
    Never 47 (89%)
    Sometimes involved 6 (11%)
Shows_interest_6mo
    Always 17 (32%)
    Often 28 (53%)
    Sometimes 8 (15%)
Shows_interest_12mo
    Always 16 (30%)
    Often 27 (51%)
    Sometimes 10 (19%)
Poor_hygiene_6mo
    Never 43 (81%)
    Sometimes 10 (19%)
Poor_hygiene_12mo
    Never 37 (70%)
    Sometimes 16 (30%)
Uses_rubberbands_6mo
    Not prescribed 29 (55%)
    Uses regularly 21 (40%)
    Rarely/never 3 (5.7%)
Uses_rubberbands_12mo
    Not prescribed 30 (57%)
    Uses regularly 17 (32%)
    Rarely/never 6 (11%)
Shows_dissatisfaction_6mo
    Never 44 (83%)
    Sometimes 9 (17%)
Shows_dissatisfaction_12mo
    Never 40 (75%)
    Sometimes 13 (25%)
1 n (%)

4.2 Change in position

5 MODELLING

Now the question is: what explain the difference between the expected and the actual movement?

Neuroticism and conscientiousness personality traits were assessed with the validated Big Five Personality Inventory (BFI)

BFI_Conscientiousness
BFI_Fake
BFI_Neuroticism

Patient cooperation was assessed by evaluating clinical fitting of aligners to dental arches and comparison of the planned and achieved upper 1st premolar expansion (%).

Fits_6mo
Fits_12mo
Patient_missed_6mo:Shows_dissatisfaction_12mo

Index of Orthodontic Treatment Needed

the higher the worst the orthodontic status of the patient

IOTN_AH
IOTN_DH

and the target variable is the difference between the

Expected_position_mm

and the

Real_position_mm  

Model the difference between expected and actual movement

5.1 Change from the initial to the final position

5.2 Preparing variables

Ensure age and gender will be included

5.3 Model 1: Lineal Regression

5.3.1 For the movement

Characteristic Beta 95% CI p-value
Age -0.01 -0.03, 0.02 0.6
Sex


    Vīriešu
    Sieviešu 0.11 -0.28, 0.51 0.6
Patient_missed_6mo


    Never
    Missed appointments 0.38 -0.11, 0.87 0.13
Shows_dissatisfaction_12mo


    Never
    Sometimes 0.38 -0.03, 0.79 0.069
IOTN_AH 0.00 -0.16, 0.17 >0.9
IOTN_DH 0.01 -0.26, 0.28 >0.9
BFI_Conscientiousness 0.01 -0.02, 0.03 0.5
BFI_Neuroticism -0.01 -0.03, 0.01 0.3
Abbreviation: CI = Confidence Interval

5.3.2 Explore BFI_Conscientiousness & BFI_Neuroticism

There are some issues:

  • Observed vs Model-Predicted Density: check, maybe robust regression, transforming the outcome, or checking outliers?

  • check cases 24, 30, 38, 47)

  • Consider robust regression or bootstrapping if inference is sensitive.

5.3.3 Check outliers

5.4 Multinomial models

5.4.1 Model for Fits 6 and

# weights:  30 (18 variable)
initial  value 58.226451 
iter  10 value 27.608110
iter  20 value 22.612271
iter  30 value 22.304646
iter  40 value 22.264025
final  value 22.263629 
converged
Characteristic log(OR) 95% CI p-value
Slight gap (1–2 teeth)
Age -0.05 -0.14, 0.03 0.2
Sex


    Vīriešu
    Sieviešu -0.22 -2.0, 1.6 0.8
Patient_missed_6mo


    Never
    Missed appointments -2.1 -4.6, 0.44 0.11
Shows_dissatisfaction_12mo


    Never
    Sometimes 1.1 -0.69, 3.0 0.2
IOTN_AH 0.51 -0.26, 1.3 0.2
IOTN_DH -0.24 -1.4, 0.95 0.7
BFI_Conscientiousness 0.05 -0.04, 0.15 0.3
BFI_Neuroticism 0.22 0.09, 0.36 0.001
Poor fit
Age -1.7 -3.2, -0.26 0.021
Sex


    Vīriešu
    Sieviešu 21 21, 21 <0.001
Patient_missed_6mo


    Never
    Missed appointments 16 16, 16 <0.001
Shows_dissatisfaction_12mo


    Never
    Sometimes -3.8 -3.8, -3.8 <0.001
IOTN_AH -0.13 -0.35, 0.10 0.3
IOTN_DH 9.9 9.7, 10 <0.001
BFI_Conscientiousness -0.89 -4.7, 2.9 0.6
BFI_Neuroticism 0.52 -3.5, 4.6 0.8
Abbreviations: CI = Confidence Interval, OR = Odds Ratio

Outcome

Characteristic

OR

95% CI

p-value

Slight gap (1–2 teeth)

Age

0.95

0.87, 1.03

0.2

Sex

Vīriešu

Sieviešu

0.80

0.14, 4.71

0.8

Patient_missed_6mo

Never

Missed appointments

0.13

0.01, 1.55

0.11

Shows_dissatisfaction_12mo

Never

Sometimes

3.15

0.50, 19.7

0.2

IOTN_AH

1.67

0.77, 3.62

0.2

IOTN_DH

0.79

0.24, 2.58

0.7

BFI_Conscientiousness

1.06

0.96, 1.17

0.3

BFI_Neuroticism

1.25

1.09, 1.43

0.001

Poor fit

Age

0.18

0.04, 0.77

0.021

Sex

Vīriešu

Sieviešu

1,596,196,257

1,476,191,068, 1,725,957,126

<0.001

Patient_missed_6mo

Never

Missed appointments

12,208,005

11,717,317, 12,719,242

<0.001

Shows_dissatisfaction_12mo

Never

Sometimes

0.02

0.02, 0.02

<0.001

IOTN_AH

0.88

0.70, 1.11

0.3

IOTN_DH

19,221

15,791, 23,395

<0.001

BFI_Conscientiousness

0.41

0.01, 18.4

0.6

BFI_Neuroticism

1.68

0.03, 97.2

0.8

Abbreviations: CI = Confidence Interval, OR = Odds Ratio

5.4.2 Model for Fit 12 months

# weights:  30 (18 variable)
initial  value 58.226451 
iter  10 value 33.433381
iter  20 value 24.267881
iter  30 value 20.780780
iter  40 value 15.727085
iter  50 value 15.633035
iter  60 value 15.554048
iter  70 value 15.530776
iter  80 value 15.523711
iter  90 value 15.521127
iter 100 value 15.519470
final  value 15.519470 
stopped after 100 iterations
Characteristic log(OR) 95% CI p-value
Slight gap (1–2 teeth)
Age -0.04 -0.14, 0.06 0.4
Sex


    Vīriešu
    Sieviešu -0.39 -2.9, 2.1 0.8
Patient_missed_6mo


    Never
    Missed appointments 0.93 -1.9, 3.7 0.5
Shows_dissatisfaction_12mo


    Never
    Sometimes -1.5 -3.8, 0.89 0.2
IOTN_AH 0.28 -0.71, 1.3 0.6
IOTN_DH -0.61 -2.4, 1.2 0.5
BFI_Conscientiousness -0.09 -0.21, 0.03 0.15
BFI_Neuroticism 0.25 0.03, 0.47 0.028
Poor fit
Age 0.70 -0.21, 1.6 0.13
Sex


    Vīriešu
    Sieviešu 25 -13, 64 0.2
Patient_missed_6mo


    Never
    Missed appointments -55 -81, -30 <0.001
Shows_dissatisfaction_12mo


    Never
    Sometimes 17 -14, 48 0.3
IOTN_AH 14 -5.3, 33 0.2
IOTN_DH 57 33, 81 <0.001
BFI_Conscientiousness 7.4 5.7, 9.0 <0.001
BFI_Neuroticism 14 13, 14 <0.001
Abbreviations: CI = Confidence Interval, OR = Odds Ratio

5.5 Model 2 Robust regression

Characteristic Beta 95% CI
Age 0.00 -0.02, 0.02
Sex

    Vīriešu
    Sieviešu 0.04 -0.30, 0.39
BFI_Conscientiousness 0.00 -0.02, 0.02
BFI_Neuroticism -0.01 -0.02, 0.01
Patient_missed_6mo

    Never
    Missed appointments 0.23 -0.20, 0.65
Shows_dissatisfaction_12mo

    Never
    Sometimes 0.31 -0.05, 0.67
IOTN_AH 0.02 -0.12, 0.17
IOTN_DH -0.01 -0.24, 0.23
Abbreviation: CI = Confidence Interval

5.6 Tidymodel

Tidy Output of Ordinal Regression Model (Sorted by p-value)
term estimate std.error statistic p.value
(Intercept) -0.425 0.084 -5.071 0.000
Shows_dissatisfaction_12mo_Sometimes 0.165 0.089 1.861 0.069
Patient_missed_6mo_Missed.appointments 0.161 0.103 1.562 0.125
BFI_Neuroticism -0.102 0.107 -0.959 0.343
BFI_Conscientiousness 0.063 0.102 0.612 0.544
Sex_Sieviešu 0.053 0.091 0.583 0.563
Age -0.050 0.092 -0.539 0.592
IOTN_DH 0.010 0.122 0.084 0.934
IOTN_AH 0.001 0.124 0.008 0.994

Presentation: https://docs.google.com/presentation/d/1VqYmsV5PSVvMKO3lzGamVfkDsTMCwQqNK47oTP_s-lw/edit#slide=id.p10

5.7 Effect on subjective measurements

5.7.1 Fits 6 months

Characteristic OR 95% CI
BFI_Neuroticism 1.17 1.08, 1.30
Gender

    Vīriešu
    Sieviešu 1.22 0.27, 6.16
Abbreviations: CI = Confidence Interval, OR = Odds Ratio