| Characteristic | 3D Replica, N = 241 | Control, N = 221 |
|---|---|---|
| Sex | ||
| Female | 13 (54%) | 16 (73%) |
| Male | 11 (46%) | 6 (27%) |
| Age (surgery) | 16.00 (16.00, 17.00) | 17.50 (17.00, 19.00) |
| Total surgery time (min) | 45 (40, 51) | 60 (50, 70) |
| Extra- alveolar time (s) | 35 (30, 60) | 55 (30, 83) |
| Donor fitting times | ||
| 1 | 15 (63%) | 7 (32%) |
| 2 | 5 (21%) | 8 (36%) |
| 3 | 4 (17%) | 5 (23%) |
| 4 | 0 (0%) | 1 (4.5%) |
| 5 | 0 (0%) | 1 (4.5%) |
| Moores Stage | ||
| 3 | 9 (38%) | 8 (36%) |
| 4 | 12 (50%) | 6 (27%) |
| 5 | 3 (13%) | 8 (36%) |
| 1 n (%); Median (IQR) | ||
03 Clinical Results Miks
Packages
Packages for the modeling
Dataset
EDA
Add the clinical data 2
Join
EDA Descriptive Table
Table 1
Group variables
Cleaning
Mobility plot
Bleeding plot
Periodontal pocket plot
Regression models
Mobility
| Mobility | 3D Replica | Control |
|---|---|---|
| 0 - Normal | 84.7% | 68.2% |
| 1 - Abnormal | 15.3% | 31.8% |
| Total | 100.0% | 100.0% |
Univariate analysis
Pearson's Chi-squared test with Yates' continuity correction
data: table(Mobility, Group)
X-squared = 4.4013, df = 1, p-value = 0.03591Multivariate analysis
| Characteristic | OR1 | 95% CI1 | p-value |
|---|---|---|---|
| Group | 0.005 | ||
| 3D Replica | — | — | |
| Control | 3.83 | 1.47, 10.9 | |
| time | <0.001 | ||
| 3 m | — | — | |
| 6 m | 0.13 | 0.04, 0.36 | |
| 12 m | 0.06 | 0.01, 0.20 | |
| Sex | 0.27 | ||
| Female | — | — | |
| Male | 1.74 | 0.65, 4.69 | |
| 1 OR = Odds Ratio, CI = Confidence Interval | |||
“Bleeding on probing”
| Bleeding | 3D Replica | Control |
|---|---|---|
| 0 - Normal | 97.2% | 86.4% |
| 1 - Abnormal | 2.8% | 13.6% |
| Total | 100.0% | 100.0% |
Univariate analysis
Pearson's Chi-squared test with Yates' continuity correction
data: table(Bleeding, Group)
X-squared = 4.1536, df = 1, p-value = 0.04155Multivariate analysis
Call:
glm(formula = as.factor(Bleeding) ~ Group + time + Sex, family = binomial,
data = clinical_long)
Coefficients:
Estimate Std. Error z value Pr(>|z|)
(Intercept) -5.3124 1.1350 -4.680 2.86e-06 ***
GroupControl 2.2568 0.8454 2.670 0.00760 **
time6 m 0.3651 0.8598 0.425 0.67108
time12 m 0.3651 0.8598 0.425 0.67108
SexMale 2.1755 0.7470 2.912 0.00359 **
---
Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
(Dispersion parameter for binomial family taken to be 1)
Null deviance: 76.745 on 137 degrees of freedom
Residual deviance: 60.781 on 133 degrees of freedom
AIC: 70.781
Number of Fisher Scoring iterations: 6| 3D Replica | Control | |
|---|---|---|
| 0 - Normal | 97.2% | 86.4% |
| 1 - Abnormal | 2.8% | 13.6% |
| Total | 100.0% | 100.0% |
Univariate analysis
Pearson's Chi-squared test with Yates' continuity correction
data: table(Pocket, Group)
X-squared = 4.1536, df = 1, p-value = 0.04155Multivariate analysis
Call:
glm(formula = as.factor(Pocket) ~ Group + time + Sex, family = binomial,
data = clinical_long)
Coefficients:
Estimate Std. Error z value Pr(>|z|)
(Intercept) -5.3124 1.1350 -4.680 2.86e-06 ***
GroupControl 2.2568 0.8454 2.670 0.00760 **
time6 m 0.3651 0.8598 0.425 0.67108
time12 m 0.3651 0.8598 0.425 0.67108
SexMale 2.1755 0.7470 2.912 0.00359 **
---
Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
(Dispersion parameter for binomial family taken to be 1)
Null deviance: 76.745 on 137 degrees of freedom
Residual deviance: 60.781 on 133 degrees of freedom
AIC: 70.781
Number of Fisher Scoring iterations: 6Modelling
Mobility model
Linear mixed model fit by REML. t-tests use Satterthwaite's method [
lmerModLmerTest]
Formula:
Mobility ~ Group * time + `Age (surgery)` + Sex + `Total surgery time (min)` +
`Extra- alveolar time (s)` + `Moores Stage` + `Donor fitting times` +
(1 | id)
Data: clinical_long
REML criterion at convergence: 132.9
Scaled residuals:
Min 1Q Median 3Q Max
-1.64797 -0.74486 -0.00841 0.42711 2.27320
Random effects:
Groups Name Variance Std.Dev.
id (Intercept) 0.01223 0.1106
Residual 0.09386 0.3064
Number of obs: 138, groups: id, 46
Fixed effects:
Estimate Std. Error df t value Pr(>|t|)
(Intercept) 0.449475 0.345442 35.529098 1.301 0.201580
GroupControl 0.121499 0.106828 94.193338 1.137 0.258284
time6 m -0.375000 0.088442 87.999998 -4.240 5.50e-05 ***
time12 m -0.416667 0.088442 87.999998 -4.711 9.14e-06 ***
`Age (surgery)` 0.001424 0.022670 34.000000 0.063 0.950298
SexMale 0.059302 0.071419 34.000000 0.830 0.412142
`Total surgery time (min)` 0.004275 0.002148 34.000000 1.990 0.054665 .
`Extra- alveolar time (s)` -0.001300 0.001105 34.000000 -1.177 0.247511
`Moores Stage`4 -0.310120 0.081033 34.000000 -3.827 0.000530 ***
`Moores Stage`5 -0.389971 0.103990 34.000000 -3.750 0.000659 ***
`Donor fitting times`2 -0.008938 0.082902 34.000000 -0.108 0.914779
`Donor fitting times`3 -0.052546 0.112550 34.000000 -0.467 0.643570
`Donor fitting times`4 1.022147 0.293722 34.000000 3.480 0.001396 **
`Donor fitting times`5 -0.252938 0.305483 34.000000 -0.828 0.413448
GroupControl:time6 m 0.011364 0.127888 87.999998 0.089 0.929398
GroupControl:time12 m -0.037879 0.127888 87.999998 -0.296 0.767784
---
Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
Bleeding
Linear mixed model fit by REML. t-tests use Satterthwaite's method [
lmerModLmerTest]
Formula:
Bleeding ~ Group * time + `Age (surgery)` + Sex + `Total surgery time (min)` +
`Extra- alveolar time (s)` + `Moores Stage` + `Donor fitting times` +
(1 | id)
Data: clinical_long
REML criterion at convergence: -96.6
Scaled residuals:
Min 1Q Median 3Q Max
-7.1625 -0.1608 -0.0532 0.0618 4.0847
Random effects:
Groups Name Variance Std.Dev.
id (Intercept) 0.03597 0.18965
Residual 0.00726 0.08521
Number of obs: 138, groups: id, 46
Fixed effects:
Estimate Std. Error df t value Pr(>|t|)
(Intercept) -0.2297298 0.3211929 34.1332598 -0.715 0.47933
GroupControl 0.0729192 0.0753630 39.5737721 0.968 0.33913
time6 m 0.0416667 0.0245969 88.0000023 1.694 0.09381 .
time12 m 0.0416667 0.0245969 88.0000023 1.694 0.09381 .
`Age (surgery)` 0.0069052 0.0212915 33.9999907 0.324 0.74768
SexMale 0.1251724 0.0670775 33.9999938 1.866 0.07068 .
`Total surgery time (min)` 0.0048846 0.0020176 33.9999939 2.421 0.02097 *
`Extra- alveolar time (s)` -0.0001918 0.0010374 33.9999940 -0.185 0.85443
`Moores Stage`4 -0.1765400 0.0761072 33.9999935 -2.320 0.02650 *
`Moores Stage`5 -0.1841585 0.0976680 33.9999930 -1.886 0.06792 .
`Donor fitting times`2 -0.1392278 0.0778619 33.9999939 -1.788 0.08267 .
`Donor fitting times`3 -0.1129832 0.1057078 33.9999938 -1.069 0.29268
`Donor fitting times`4 0.9337009 0.2758663 33.9999939 3.385 0.00181 **
`Donor fitting times`5 -0.3075329 0.2869123 33.9999941 -1.072 0.29133
GroupControl:time6 m -0.0416667 0.0355671 88.0000023 -1.171 0.24456
GroupControl:time12 m -0.0416667 0.0355671 88.0000023 -1.171 0.24456
---
Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1We fitted a linear mixed model (estimated using REML and nloptwrap optimizer)
to predict Bleeding with Group, time, Age (surgery), Sex, Total surgery time
(min), Extra- alveolar time (s), Moores Stage and Donor fitting times (formula:
Bleeding ~ Group * time + `Age (surgery)` + Sex + `Total surgery time (min)` +
`Extra- alveolar time (s)` + `Moores Stage` + `Donor fitting times`). The model
included id as random effect (formula: ~1 | id). The model's total explanatory
power is substantial (conditional R2 = 0.91) and the part related to the fixed
effects alone (marginal R2) is of 0.48. The model's intercept, corresponding to
Group = 3D Replica, time = 3 m, Age (surgery) = 0, Sex = Female, Total surgery
time (min) = 0, Extra- alveolar time (s) = 0, Moores Stage = 3 and Donor
fitting times = 1, is at -0.23 (95% CI [-0.87, 0.41], t(120) = -0.72, p =
0.476). Within this model:
- The effect of Group [Control] is statistically non-significant and positive
(beta = 0.07, 95% CI [-0.08, 0.22], t(120) = 0.97, p = 0.335; Std. beta = 0.27,
95% CI [-0.28, 0.82])
- The effect of time [6 m] is statistically non-significant and positive (beta
= 0.04, 95% CI [-7.03e-03, 0.09], t(120) = 1.69, p = 0.093; Std. beta = 0.15,
95% CI [-0.03, 0.33])
- The effect of time [12 m] is statistically non-significant and positive (beta
= 0.04, 95% CI [-7.03e-03, 0.09], t(120) = 1.69, p = 0.093; Std. beta = 0.15,
95% CI [-0.03, 0.33])
- The effect of Age (surgery) is statistically non-significant and positive
(beta = 6.91e-03, 95% CI [-0.04, 0.05], t(120) = 0.32, p = 0.746; Std. beta =
0.05, 95% CI [-0.23, 0.33])
- The effect of Sex [Male] is statistically non-significant and positive (beta
= 0.13, 95% CI [-7.64e-03, 0.26], t(120) = 1.87, p = 0.064; Std. beta = 0.46,
95% CI [-0.03, 0.95])
- The effect of Total surgery time (min) is statistically significant and
positive (beta = 4.88e-03, 95% CI [8.90e-04, 8.88e-03], t(120) = 2.42, p =
0.017; Std. beta = 0.34, 95% CI [0.06, 0.61])
- The effect of Extra- alveolar time (s) is statistically non-significant and
negative (beta = -1.92e-04, 95% CI [-2.25e-03, 1.86e-03], t(120) = -0.18, p =
0.854; Std. beta = -0.03, 95% CI [-0.38, 0.31])
- The effect of Moores Stage [4] is statistically significant and negative
(beta = -0.18, 95% CI [-0.33, -0.03], t(120) = -2.32, p = 0.022; Std. beta =
-0.65, 95% CI [-1.20, -0.10])
- The effect of Moores Stage [5] is statistically non-significant and negative
(beta = -0.18, 95% CI [-0.38, 9.22e-03], t(120) = -1.89, p = 0.062; Std. beta =
-0.68, 95% CI [-1.39, 0.03])
- The effect of Donor fitting times [2] is statistically non-significant and
negative (beta = -0.14, 95% CI [-0.29, 0.01], t(120) = -1.79, p = 0.076; Std.
beta = -0.51, 95% CI [-1.08, 0.05])
- The effect of Donor fitting times [3] is statistically non-significant and
negative (beta = -0.11, 95% CI [-0.32, 0.10], t(120) = -1.07, p = 0.287; Std.
beta = -0.42, 95% CI [-1.19, 0.35])
- The effect of Donor fitting times [4] is statistically significant and
positive (beta = 0.93, 95% CI [0.39, 1.48], t(120) = 3.38, p < .001; Std. beta
= 3.43, 95% CI [1.43, 5.44])
- The effect of Donor fitting times [5] is statistically non-significant and
negative (beta = -0.31, 95% CI [-0.88, 0.26], t(120) = -1.07, p = 0.286; Std.
beta = -1.13, 95% CI [-3.22, 0.96])
- The effect of Group [Control] × time [6 m] is statistically non-significant
and negative (beta = -0.04, 95% CI [-0.11, 0.03], t(120) = -1.17, p = 0.244;
Std. beta = -0.15, 95% CI [-0.41, 0.11])
- The effect of Group [Control] × time [12 m] is statistically non-significant
and negative (beta = -0.04, 95% CI [-0.11, 0.03], t(120) = -1.17, p = 0.244;
Std. beta = -0.15, 95% CI [-0.41, 0.11])
Standardized parameters were obtained by fitting the model on a standardized
version of the dataset. 95% Confidence Intervals (CIs) and p-values were
computed using a Wald t-distribution approximation.
Linear mixed model fit by REML. t-tests use Satterthwaite's method [
lmerModLmerTest]
Formula:
Pocket ~ Group * time + `Age (surgery)` + Sex + `Total surgery time (min)` +
`Extra- alveolar time (s)` + `Moores Stage` + `Donor fitting times` +
(1 | id)
Data: clinical_long
REML criterion at convergence: -96.6
Scaled residuals:
Min 1Q Median 3Q Max
-7.1625 -0.1608 -0.0532 0.0618 4.0847
Random effects:
Groups Name Variance Std.Dev.
id (Intercept) 0.03597 0.18965
Residual 0.00726 0.08521
Number of obs: 138, groups: id, 46
Fixed effects:
Estimate Std. Error df t value Pr(>|t|)
(Intercept) -0.2297298 0.3211929 34.1332598 -0.715 0.47933
GroupControl 0.0729192 0.0753630 39.5737721 0.968 0.33913
time6 m 0.0416667 0.0245969 88.0000023 1.694 0.09381 .
time12 m 0.0416667 0.0245969 88.0000023 1.694 0.09381 .
`Age (surgery)` 0.0069052 0.0212915 33.9999907 0.324 0.74768
SexMale 0.1251724 0.0670775 33.9999938 1.866 0.07068 .
`Total surgery time (min)` 0.0048846 0.0020176 33.9999939 2.421 0.02097 *
`Extra- alveolar time (s)` -0.0001918 0.0010374 33.9999940 -0.185 0.85443
`Moores Stage`4 -0.1765400 0.0761072 33.9999935 -2.320 0.02650 *
`Moores Stage`5 -0.1841585 0.0976680 33.9999930 -1.886 0.06792 .
`Donor fitting times`2 -0.1392278 0.0778619 33.9999939 -1.788 0.08267 .
`Donor fitting times`3 -0.1129832 0.1057078 33.9999938 -1.069 0.29268
`Donor fitting times`4 0.9337009 0.2758663 33.9999939 3.385 0.00181 **
`Donor fitting times`5 -0.3075329 0.2869123 33.9999941 -1.072 0.29133
GroupControl:time6 m -0.0416667 0.0355671 88.0000023 -1.171 0.24456
GroupControl:time12 m -0.0416667 0.0355671 88.0000023 -1.171 0.24456
---
Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
All regression plots
All Regression tables
| Characteristic | Mobility | Bleeding on probing | |||||||
|---|---|---|---|---|---|---|---|---|---|
| Beta | 95% CI1 | p-value | Beta | 95% CI1 | p-value | Beta | 95% CI1 | p-value | |
| Group | |||||||||
| 3D Replica | — | — | — | — | — | — | |||
| Control | 0.12 | -0.09, 0.33 | 0.3 | 0.07 | -0.08, 0.23 | 0.3 | 0.07 | -0.08, 0.23 | 0.3 |
| time | |||||||||
| 3 m | — | — | — | — | — | — | |||
| 6 m | -0.38 | -0.55, -0.20 | <0.001 | 0.04 | -0.01, 0.09 | 0.094 | 0.04 | -0.01, 0.09 | 0.094 |
| 12 m | -0.42 | -0.59, -0.24 | <0.001 | 0.04 | -0.01, 0.09 | 0.094 | 0.04 | -0.01, 0.09 | 0.094 |
| Age (surgery) | 0.00 | -0.04, 0.05 | >0.9 | 0.01 | -0.04, 0.05 | 0.7 | 0.01 | -0.04, 0.05 | 0.7 |
| Sex | |||||||||
| Female | — | — | — | — | — | — | |||
| Male | 0.06 | -0.09, 0.20 | 0.4 | 0.13 | -0.01, 0.26 | 0.071 | 0.13 | -0.01, 0.26 | 0.071 |
| Total surgery time (min) | 0.00 | 0.00, 0.01 | 0.055 | 0.00 | 0.00, 0.01 | 0.021 | 0.00 | 0.00, 0.01 | 0.021 |
| Extra- alveolar time (s) | 0.00 | 0.00, 0.00 | 0.2 | 0.00 | 0.00, 0.00 | 0.9 | 0.00 | 0.00, 0.00 | 0.9 |
| Moores Stage | |||||||||
| 3 | — | — | — | — | — | — | |||
| 4 | -0.31 | -0.47, -0.15 | <0.001 | -0.18 | -0.33, -0.02 | 0.027 | -0.18 | -0.33, -0.02 | 0.027 |
| 5 | -0.39 | -0.60, -0.18 | <0.001 | -0.18 | -0.38, 0.01 | 0.068 | -0.18 | -0.38, 0.01 | 0.068 |
| Donor fitting times | |||||||||
| 1 | — | — | — | — | — | — | |||
| 2 | -0.01 | -0.18, 0.16 | >0.9 | -0.14 | -0.30, 0.02 | 0.083 | -0.14 | -0.30, 0.02 | 0.083 |
| 3 | -0.05 | -0.28, 0.18 | 0.6 | -0.11 | -0.33, 0.10 | 0.3 | -0.11 | -0.33, 0.10 | 0.3 |
| 4 | 1.0 | 0.43, 1.6 | 0.001 | 0.93 | 0.37, 1.5 | 0.002 | 0.93 | 0.37, 1.5 | 0.002 |
| 5 | -0.25 | -0.87, 0.37 | 0.4 | -0.31 | -0.89, 0.28 | 0.3 | -0.31 | -0.89, 0.28 | 0.3 |
| Group * time | |||||||||
| Control * 6 m | 0.01 | -0.24, 0.27 | >0.9 | -0.04 | -0.11, 0.03 | 0.2 | -0.04 | -0.11, 0.03 | 0.2 |
| Control * 12 m | -0.04 | -0.29, 0.22 | 0.8 | -0.04 | -0.11, 0.03 | 0.2 | -0.04 | -0.11, 0.03 | 0.2 |
| 1 CI = Confidence Interval | |||||||||
Tidymodels
Truth
Prediction 3D Replica Control
3D Replica 11 7
Control 4 13# A tibble: 2 × 3
.metric .estimator .estimate
<chr> <chr> <dbl>
1 accuracy binary 0.686
2 kap binary 0.374