Main analysis
From baseline to 6 months - adjusting for control vars.
Analysis outcome soteria hospital online Paper
Dana Sinai
2024-05-29
## Stratified by group
## Overall Hospital
## n 214 66
## age (mean (SD)) 36.21 (14.12) 37.64 (15.81)
## gender = Men (%) 105 (49.1) 28 (42.4)
## in_a_relationsip = In a relationship (%) 60 (28.0) 15 (22.7)
## lives_alone = Yes (%) 164 (77.7) 54 (83.1)
## sourceincome = 1 (%) 76 (35.5) 34 (51.5)
## years_education = 12 years and over (%) 183 (85.5) 48 (72.7)
## mainhospitalizationreason = Other (%) 140 (65.4) 36 (54.5)
## maindiagnosis = Other (%) 111 (51.9) 27 (40.9)
## personalitycomorbidity = Yes (%) 57 (26.6) 18 (27.3)
## disability = Yes (%) 76 (35.5) 34 (51.5)
## Stratified by group
## Soteria Online p
## n 94 54
## age (mean (SD)) 33.11 (12.22) 39.85 (14.12) 0.012
## gender = Men (%) 46 (48.9) 31 (57.4) 0.263
## in_a_relationsip = In a relationship (%) 17 (18.1) 28 (51.9) <0.001
## lives_alone = Yes (%) 66 (70.2) 44 (84.6) 0.062
## sourceincome = 1 (%) 24 (25.5) 18 (33.3) 0.003
## years_education = 12 years and over (%) 87 (92.6) 48 (88.9) 0.002
## mainhospitalizationreason = Other (%) 59 (62.8) 45 (83.3) 0.003
## maindiagnosis = Other (%) 42 (44.7) 42 (77.8) <0.001
## personalitycomorbidity = Yes (%) 33 (35.1) 6 (11.1) 0.006
## disability = Yes (%) 24 (25.5) 18 (33.3) 0.003
## Stratified by group
## test
## n
## age (mean (SD))
## gender = Men (%)
## in_a_relationsip = In a relationship (%)
## lives_alone = Yes (%)
## sourceincome = 1 (%)
## years_education = 12 years and over (%)
## mainhospitalizationreason = Other (%)
## maindiagnosis = Other (%)
## personalitycomorbidity = Yes (%)
## disability = Yes (%)There are significant differences between groups on all variables except for gender, so all will be considered control vars.
In order to examine the differences in effectiveness across the three modules, multi-level modeling was employed (MLM), whereby Level-1 was the measurement level across different time points, and Level-2 was the patients’ level. Differences between the three modules were calculated as the interaction between group (Soteria, TAIHT, and psychiatric ward) and time (admission, discharge, 3-month follow-up, and 6-month follow-up) using direct contrasts.
Clinical and demographic variables that demonstrated significant differences between modules were adjusted in all models. When models did not converge, we considered a simpler model for evaluation
Baseline values on outcome variables
## Stratified by group
## Hospital Soteria Online p test
## n 66 94 54
## oq (mean (SD)) 92.21 (27.17) 90.12 (25.15) 84.83 (24.18) 0.275
## oqsd (mean (SD)) 54.65 (18.29) 54.88 (17.43) 53.02 (15.95) 0.809
## oqir (mean (SD)) 20.76 (8.04) 20.16 (6.76) 17.11 (7.03) 0.015
## oqsr (mean (SD)) 16.80 (6.69) 15.08 (6.55) 14.70 (6.37) 0.152
## hscl (mean (SD)) 2.87 (0.72) 2.86 (0.69) 2.70 (0.65) 0.299
## panss (mean (SD)) 12.58 (6.91) 11.47 (6.38) 8.27 (3.78) 0.001
## gaf (mean (SD)) 26.77 (10.96) 34.70 (12.88) 41.37 (7.95) <0.001
## mansa (mean (SD)) 3.66 (1.14) 3.78 (0.99) 3.81 (0.98) 0.703significant differences at baseline were observed for OQ-IR, PANSS, GAF.
OQ-45
## Number of non-NA observations of oq at every data collection point:
## # A tibble: 6 × 3
## time count_non_na percent_of_b
## <fct> <int> <dbl>
## 1 b 213 100
## 2 r 0 0
## 3 3m 0 0
## 4 6m 124 58.2
## 5 1y 157 73.7
## 6 1.5y 156 73.2## Model Summary:
## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: formula
## Data: data
##
## REML criterion at convergence: 3001
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -2.8596 -0.4700 0.1148 0.4901 2.1666
##
## Random effects:
## Groups Name Variance Std.Dev.
## id (Intercept) 285.3 16.89
## Residual 372.0 19.29
## Number of obs: 332, groups: id, 210
##
## Fixed effects:
## Estimate Std. Error df t value
## (Intercept) 85.51423 8.70057 208.03933 9.829
## groupSoteria -0.42618 4.39410 274.05880 -0.097
## groupOnline -9.98520 5.27276 271.40694 -1.894
## time6m -20.11774 4.21586 164.38753 -4.772
## age 0.06789 0.13468 208.05612 0.504
## in_a_relationsipIn a relationship 2.84199 4.50342 221.01544 0.631
## lives_aloneYes -0.11353 4.33104 203.02869 -0.026
## sourceincome1 1.97385 3.63533 208.66811 0.543
## years_education12 years and over -9.66159 4.83514 197.96520 -1.998
## mainhospitalizationreasonOther 5.87209 4.68262 200.82773 1.254
## maindiagnosisOther 9.88127 4.54052 201.44809 2.176
## personalitycomorbidityYes 8.01425 4.25922 207.37919 1.882
## groupSoteria:time6m 2.50865 5.39944 162.09424 0.465
## groupOnline:time6m 15.25965 6.51584 168.43293 2.342
## Pr(>|t|)
## (Intercept) < 2e-16 ***
## groupSoteria 0.9228
## groupOnline 0.0593 .
## time6m 4.01e-06 ***
## age 0.6147
## in_a_relationsipIn a relationship 0.5286
## lives_aloneYes 0.9791
## sourceincome1 0.5877
## years_education12 years and over 0.0471 *
## mainhospitalizationreasonOther 0.2113
## maindiagnosisOther 0.0307 *
## personalitycomorbidityYes 0.0613 .
## groupSoteria:time6m 0.6428
## groupOnline:time6m 0.0204 *
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## fit warnings:
## fixed-effect model matrix is rank deficient so dropping 1 column / coefficient
##
## Estimated Marginal Means (emmeans):
## group = Hospital:
## contrast estimate SE df t.ratio p.value
## b - 6m 20.12 4.23 153 4.761 <.0001
##
## group = Soteria:
## contrast estimate SE df t.ratio p.value
## b - 6m 17.61 3.38 148 5.210 <.0001
##
## group = Online:
## contrast estimate SE df t.ratio p.value
## b - 6m 4.86 4.98 161 0.976 0.3307
##
## Results are averaged over the levels of: in_a_relationsip, lives_alone, years_education, mainhospitalizationreason, maindiagnosis, personalitycomorbidity, disability
## Degrees-of-freedom method: kenward-roger
OQ had measurements only at baseline and at 6 months
OQ-SD
## Number of non-NA observations of oqsd at every data collection point:
## # A tibble: 6 × 3
## time count_non_na percent_of_b
## <fct> <int> <dbl>
## 1 b 213 100
## 2 r 0 0
## 3 3m 0 0
## 4 6m 124 58.2
## 5 1y 160 75.1
## 6 1.5y 160 75.1## Model Summary:
## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: formula
## Data: data
##
## REML criterion at convergence: 2739.6
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -2.65988 -0.50751 0.03844 0.59931 1.91444
##
## Random effects:
## Groups Name Variance Std.Dev.
## id (Intercept) 103.7 10.18
## Residual 178.3 13.35
## Number of obs: 332, groups: id, 210
##
## Fixed effects:
## Estimate Std. Error df t value
## (Intercept) 44.21477 5.61547 207.18245 7.874
## groupSoteria 1.26694 2.87288 281.89218 0.441
## groupOnline -4.38419 3.44543 278.91647 -1.272
## time6m -12.73405 2.89558 168.55033 -4.398
## age 0.09613 0.08692 206.48090 1.106
## in_a_relationsipIn a relationship 1.11492 2.91375 221.00034 0.383
## lives_aloneYes 0.50023 2.79236 200.92183 0.179
## sourceincome1 2.48438 2.34642 207.09286 1.059
## years_education12 years and over -3.00107 3.11416 195.22398 -0.964
## mainhospitalizationreasonOther 4.79301 3.01768 198.41122 1.588
## maindiagnosisOther 7.79047 2.92647 199.08310 2.662
## personalitycomorbidityYes 4.03006 2.74842 205.71691 1.466
## groupSoteria:time6m 0.87254 3.71054 165.93316 0.235
## groupOnline:time6m 7.27347 4.47104 173.15098 1.627
## Pr(>|t|)
## (Intercept) 1.90e-13 ***
## groupSoteria 0.6596
## groupOnline 0.2043
## time6m 1.93e-05 ***
## age 0.2700
## in_a_relationsipIn a relationship 0.7024
## lives_aloneYes 0.8580
## sourceincome1 0.2909
## years_education12 years and over 0.3364
## mainhospitalizationreasonOther 0.1138
## maindiagnosisOther 0.0084 **
## personalitycomorbidityYes 0.1441
## groupSoteria:time6m 0.8144
## groupOnline:time6m 0.1056
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## fit warnings:
## fixed-effect model matrix is rank deficient so dropping 1 column / coefficient
##
## Estimated Marginal Means (emmeans):
## group = Hospital:
## contrast estimate SE df t.ratio p.value
## b - 6m 12.73 2.90 158 4.387 <.0001
##
## group = Soteria:
## contrast estimate SE df t.ratio p.value
## b - 6m 11.86 2.32 152 5.102 <.0001
##
## group = Online:
## contrast estimate SE df t.ratio p.value
## b - 6m 5.46 3.41 167 1.599 0.1117
##
## Results are averaged over the levels of: in_a_relationsip, lives_alone, years_education, mainhospitalizationreason, maindiagnosis, personalitycomorbidity, disability
## Degrees-of-freedom method: kenward-roger
OQ-IR
## Number of non-NA observations of oqir at every data collection point:
## # A tibble: 6 × 3
## time count_non_na percent_of_b
## <fct> <int> <dbl>
## 1 b 213 100
## 2 r 0 0
## 3 3m 0 0
## 4 6m 124 58.2
## 5 1y 160 75.1
## 6 1.5y 160 75.1## Model Summary:
## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: formula
## Data: data
##
## REML criterion at convergence: 2201.8
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -2.42468 -0.51290 -0.04646 0.49537 2.45792
##
## Random effects:
## Groups Name Variance Std.Dev.
## id (Intercept) 31.26 5.591
## Residual 25.27 5.027
## Number of obs: 332, groups: id, 210
##
## Fixed effects:
## Estimate Std. Error df t value
## (Intercept) 24.55240 2.61456 205.43151 9.391
## groupSoteria -0.11486 1.29249 257.19461 -0.089
## groupOnline -3.07091 1.55231 255.13311 -1.978
## time6m -4.38744 1.11403 152.71724 -3.938
## age -0.02128 0.04046 206.24391 -0.526
## in_a_relationsipIn a relationship 0.10344 1.34726 216.56174 0.077
## lives_aloneYes -0.88441 1.30350 202.16217 -0.678
## sourceincome1 -0.31961 1.09199 206.82084 -0.293
## years_education12 years and over -4.37937 1.45773 198.18362 -3.004
## mainhospitalizationreasonOther -0.47465 1.41035 200.47425 -0.337
## maindiagnosisOther 1.84339 1.36726 200.99517 1.348
## personalitycomorbidityYes 1.86071 1.27997 205.71530 1.454
## groupSoteria:time6m 2.21587 1.42544 150.98945 1.555
## groupOnline:time6m 5.63728 1.72460 155.78576 3.269
## Pr(>|t|)
## (Intercept) < 2e-16 ***
## groupSoteria 0.929256
## groupOnline 0.048972 *
## time6m 0.000124 ***
## age 0.599561
## in_a_relationsipIn a relationship 0.938871
## lives_aloneYes 0.498235
## sourceincome1 0.770058
## years_education12 years and over 0.003006 **
## mainhospitalizationreasonOther 0.736813
## maindiagnosisOther 0.179099
## personalitycomorbidityYes 0.147551
## groupSoteria:time6m 0.122156
## groupOnline:time6m 0.001329 **
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## fit warnings:
## fixed-effect model matrix is rank deficient so dropping 1 column / coefficient
##
## Estimated Marginal Means (emmeans):
## group = Hospital:
## contrast estimate SE df t.ratio p.value
## b - 6m 4.39 1.116 145 3.931 0.0001
##
## group = Soteria:
## contrast estimate SE df t.ratio p.value
## b - 6m 2.17 0.891 141 2.438 0.0160
##
## group = Online:
## contrast estimate SE df t.ratio p.value
## b - 6m -1.25 1.319 151 -0.948 0.3449
##
## Results are averaged over the levels of: in_a_relationsip, lives_alone, years_education, mainhospitalizationreason, maindiagnosis, personalitycomorbidity, disability
## Degrees-of-freedom method: kenward-roger
OQ-SR
## Number of non-NA observations of oqsr at every data collection point:
## # A tibble: 6 × 3
## time count_non_na percent_of_b
## <fct> <int> <dbl>
## 1 b 213 100
## 2 r 0 0
## 3 3m 0 0
## 4 6m 124 58.2
## 5 1y 160 75.1
## 6 1.5y 160 75.1## Model Summary:
## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: formula
## Data: data
##
## REML criterion at convergence: 2105.5
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -2.31499 -0.58323 0.03547 0.61664 2.03900
##
## Random effects:
## Groups Name Variance Std.Dev.
## id (Intercept) 11.11 3.333
## Residual 26.48 5.146
## Number of obs: 332, groups: id, 210
##
## Fixed effects:
## Estimate Std. Error df t value
## (Intercept) 16.768098 2.015632 187.934107 8.319
## groupSoteria -1.591580 1.046779 283.294052 -1.520
## groupOnline -2.636882 1.254553 279.353801 -2.102
## time6m -2.947620 1.106251 153.787309 -2.665
## age -0.005221 0.031191 186.069745 -0.167
## in_a_relationsipIn a relationship 1.776457 1.048643 203.400338 1.694
## lives_aloneYes 0.067043 1.000942 179.648644 0.067
## sourceincome1 -0.165666 0.842135 186.693787 -0.197
## years_education12 years and over -2.173354 1.114969 173.005721 -1.949
## mainhospitalizationreasonOther 1.423263 1.081144 176.669991 1.316
## maindiagnosisOther 0.404299 1.048609 177.421239 0.386
## personalitycomorbidityYes 1.993668 0.986148 185.164871 2.022
## groupSoteria:time6m -0.514092 1.418482 150.800153 -0.362
## groupOnline:time6m 2.386005 1.706332 159.050698 1.398
## Pr(>|t|)
## (Intercept) 1.77e-14 ***
## groupSoteria 0.12951
## groupOnline 0.03646 *
## time6m 0.00853 **
## age 0.86725
## in_a_relationsipIn a relationship 0.09179 .
## lives_aloneYes 0.94667
## sourceincome1 0.84426
## years_education12 years and over 0.05288 .
## mainhospitalizationreasonOther 0.18973
## maindiagnosisOther 0.70029
## personalitycomorbidityYes 0.04465 *
## groupSoteria:time6m 0.71754
## groupOnline:time6m 0.16396
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## fit warnings:
## fixed-effect model matrix is rank deficient so dropping 1 column / coefficient
##
## Estimated Marginal Means (emmeans):
## group = Hospital:
## contrast estimate SE df t.ratio p.value
## b - 6m 2.948 1.11 164 2.658 0.0086
##
## group = Soteria:
## contrast estimate SE df t.ratio p.value
## b - 6m 3.462 0.89 156 3.891 0.0001
##
## group = Online:
## contrast estimate SE df t.ratio p.value
## b - 6m 0.562 1.30 173 0.431 0.6668
##
## Results are averaged over the levels of: in_a_relationsip, lives_alone, years_education, mainhospitalizationreason, maindiagnosis, personalitycomorbidity, disability
## Degrees-of-freedom method: kenward-roger
HSCL
## Number of non-NA observations of hscl at every data collection point:
## # A tibble: 6 × 3
## time count_non_na percent_of_b
## <fct> <int> <dbl>
## 1 b 212 100
## 2 r 154 72.6
## 3 3m 145 68.4
## 4 6m 0 0
## 5 1y 81 38.2
## 6 1.5y 72 34.0## Model Summary:
## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: formula
## Data: data
##
## REML criterion at convergence: 995.1
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -2.09850 -0.58369 0.00341 0.62235 2.41310
##
## Random effects:
## Groups Name Variance Std.Dev.
## id (Intercept) 0.1973 0.4441
## Residual 0.2545 0.5044
## Number of obs: 504, groups: id, 209
##
## Fixed effects:
## Estimate Std. Error df t value
## (Intercept) 2.354960 0.214454 208.859185 10.981
## groupSoteria 0.016624 0.114337 344.623906 0.145
## groupOnline -0.258689 0.136401 334.136446 -1.897
## timer -0.598092 0.098347 320.833102 -6.081
## time3m -0.574477 0.102985 326.720470 -5.578
## age 0.004587 0.003273 195.109754 1.402
## in_a_relationsipIn a relationship -0.034813 0.107915 197.130428 -0.323
## lives_aloneYes 0.028037 0.105010 190.365471 0.267
## sourceincome1 0.125992 0.087596 191.303205 1.438
## years_education12 years and over 0.005748 0.117904 190.844142 0.049
## mainhospitalizationreasonOther 0.238430 0.114170 189.504664 2.088
## maindiagnosisOther 0.197767 0.112058 194.076676 1.765
## personalitycomorbidityYes 0.154577 0.104153 196.181842 1.484
## groupSoteria:timer -0.078398 0.130016 322.851104 -0.603
## groupOnline:timer 0.185610 0.147504 319.263108 1.258
## groupSoteria:time3m -0.022510 0.133255 326.021435 -0.169
## groupOnline:time3m 0.233014 0.153809 324.674831 1.515
## Pr(>|t|)
## (Intercept) < 2e-16 ***
## groupSoteria 0.8845
## groupOnline 0.0588 .
## timer 3.38e-09 ***
## time3m 5.10e-08 ***
## age 0.1626
## in_a_relationsipIn a relationship 0.7473
## lives_aloneYes 0.7898
## sourceincome1 0.1520
## years_education12 years and over 0.9612
## mainhospitalizationreasonOther 0.0381 *
## maindiagnosisOther 0.0792 .
## personalitycomorbidityYes 0.1394
## groupSoteria:timer 0.5469
## groupOnline:timer 0.2092
## groupSoteria:time3m 0.8660
## groupOnline:time3m 0.1308
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## fit warnings:
## fixed-effect model matrix is rank deficient so dropping 1 column / coefficient
##
## Estimated Marginal Means (emmeans):
## group = Hospital:
## contrast estimate SE df t.ratio p.value
## b - r 0.5981 0.0984 321 6.078 <.0001
## b - 3m 0.5745 0.1031 327 5.574 <.0001
## r - 3m -0.0236 0.1072 306 -0.220 0.9736
##
## group = Soteria:
## contrast estimate SE df t.ratio p.value
## b - r 0.6765 0.0851 326 7.953 <.0001
## b - 3m 0.5970 0.0846 326 7.058 <.0001
## r - 3m -0.0795 0.0903 305 -0.881 0.6530
##
## group = Online:
## contrast estimate SE df t.ratio p.value
## b - r 0.4125 0.1100 319 3.751 0.0006
## b - 3m 0.3415 0.1143 324 2.987 0.0085
## r - 3m -0.0710 0.1207 312 -0.589 0.8264
##
## Results are averaged over the levels of: in_a_relationsip, lives_alone, years_education, mainhospitalizationreason, maindiagnosis, personalitycomorbidity, disability
## Degrees-of-freedom method: kenward-roger
## P value adjustment: tukey method for comparing a family of 3 estimates
PANSS
## Number of non-NA observations of panss at every data collection point:
## # A tibble: 6 × 3
## time count_non_na percent_of_b
## <fct> <int> <dbl>
## 1 b 210 100
## 2 r 184 87.6
## 3 3m 0 0
## 4 6m 0 0
## 5 1y 106 50.5
## 6 1.5y 76 36.2## Model Summary:
## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: formula
## Data: data
##
## REML criterion at convergence: 2186.7
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -2.3158 -0.4604 -0.0380 0.3200 3.4052
##
## Random effects:
## Groups Name Variance Std.Dev.
## id (Intercept) 16.173 4.022
## Residual 7.064 2.658
## Number of obs: 389, groups: id, 208
##
## Fixed effects:
## Estimate Std. Error df t value
## (Intercept) 15.754501 1.688134 198.363643 9.332
## groupSoteria -0.231393 0.831466 245.303444 -0.278
## groupOnline -1.883717 1.008633 239.566543 -1.868
## timer -1.489336 0.500700 179.846168 -2.975
## age 0.006463 0.026024 190.664654 0.248
## in_a_relationsipIn a relationship -1.232506 0.856310 191.067140 -1.439
## lives_aloneYes -0.179110 0.847251 189.754290 -0.211
## sourceincome1 1.354090 0.704014 191.547238 1.923
## years_education12 years and over -1.017246 0.945044 189.995895 -1.076
## mainhospitalizationreasonOther -2.679332 0.917666 190.927073 -2.920
## maindiagnosisOther -2.866705 0.892934 189.297192 -3.210
## personalitycomorbidityYes -1.028042 0.828836 191.050063 -1.240
## groupSoteria:timer -1.305116 0.650983 179.333267 -2.005
## groupOnline:timer 0.694121 0.744450 177.666957 0.932
## Pr(>|t|)
## (Intercept) < 2e-16 ***
## groupSoteria 0.78102
## groupOnline 0.06304 .
## timer 0.00334 **
## age 0.80413
## in_a_relationsipIn a relationship 0.15170
## lives_aloneYes 0.83280
## sourceincome1 0.05591 .
## years_education12 years and over 0.28311
## mainhospitalizationreasonOther 0.00392 **
## maindiagnosisOther 0.00156 **
## personalitycomorbidityYes 0.21637
## groupSoteria:timer 0.04648 *
## groupOnline:timer 0.35240
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## fit warnings:
## fixed-effect model matrix is rank deficient so dropping 1 column / coefficient
##
## Estimated Marginal Means (emmeans):
## group = Hospital:
## contrast estimate SE df t.ratio p.value
## b - r 1.489 0.501 186 2.973 0.0033
##
## group = Soteria:
## contrast estimate SE df t.ratio p.value
## b - r 2.794 0.416 185 6.717 <.0001
##
## group = Online:
## contrast estimate SE df t.ratio p.value
## b - r 0.795 0.551 182 1.443 0.1507
##
## Results are averaged over the levels of: in_a_relationsip, lives_alone, years_education, mainhospitalizationreason, maindiagnosis, personalitycomorbidity, disability
## Degrees-of-freedom method: kenward-roger
GAF
## Number of non-NA observations of gaf at every data collection point:
## # A tibble: 6 × 3
## time count_non_na percent_of_b
## <fct> <int> <dbl>
## 1 b 214 100
## 2 r 167 78.0
## 3 3m 142 66.4
## 4 6m 125 58.4
## 5 1y 98 45.8
## 6 1.5y 73 34.1## Model Summary:
## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: formula
## Data: data
##
## REML criterion at convergence: 4990.2
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -3.07011 -0.48510 -0.01978 0.53482 2.77883
##
## Random effects:
## Groups Name Variance Std.Dev.
## id (Intercept) 56.77 7.534
## Residual 122.22 11.055
## Number of obs: 639, groups: id, 211
##
## Fixed effects:
## Estimate Std. Error df t value
## (Intercept) 24.11296 3.91182 226.65956 6.164
## groupSoteria 5.33178 2.24934 461.04977 2.370
## groupOnline 11.65337 2.67426 443.59627 4.358
## timer 16.05573 2.09567 469.47908 7.661
## time3m 18.05843 2.27204 484.14546 7.948
## time6m 22.30202 2.31393 484.98025 9.638
## age 0.01440 0.05904 202.57684 0.244
## in_a_relationsipIn a relationship -1.11824 1.94519 202.85660 -0.575
## lives_aloneYes -0.22329 1.86985 189.48450 -0.119
## sourceincome1 -5.53157 1.57598 194.57888 -3.510
## years_education12 years and over 5.34652 2.11870 193.18248 2.523
## mainhospitalizationreasonOther 2.82003 2.02613 185.58308 1.392
## maindiagnosisOther 0.41947 1.97921 190.27815 0.212
## personalitycomorbidityYes -1.00776 1.85254 196.42571 -0.544
## groupSoteria:timer 0.65191 2.73681 471.33124 0.238
## groupOnline:timer -8.37335 3.16397 470.39104 -2.646
## groupSoteria:time3m 5.16530 2.93613 483.43557 1.759
## groupOnline:time3m -6.53888 3.33058 480.58235 -1.963
## groupSoteria:time6m 4.46669 2.99759 484.21874 1.490
## groupOnline:time6m -11.46177 3.59988 488.08692 -3.184
## Pr(>|t|)
## (Intercept) 3.21e-09 ***
## groupSoteria 0.018180 *
## groupOnline 1.64e-05 ***
## timer 1.06e-13 ***
## time3m 1.34e-14 ***
## time6m < 2e-16 ***
## age 0.807562
## in_a_relationsipIn a relationship 0.566013
## lives_aloneYes 0.905070
## sourceincome1 0.000557 ***
## years_education12 years and over 0.012425 *
## mainhospitalizationreasonOther 0.165640
## maindiagnosisOther 0.832383
## personalitycomorbidityYes 0.587068
## groupSoteria:timer 0.811830
## groupOnline:timer 0.008406 **
## groupSoteria:time3m 0.079173 .
## groupOnline:time3m 0.050189 .
## groupSoteria:time6m 0.136852
## groupOnline:time6m 0.001546 **
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## fit warnings:
## fixed-effect model matrix is rank deficient so dropping 1 column / coefficient
##
## Estimated Marginal Means (emmeans):
## group = Hospital:
## contrast estimate SE df t.ratio p.value
## b - r -16.056 2.10 464 -7.659 <.0001
## b - 3m -18.058 2.27 479 -7.943 <.0001
## b - 6m -22.302 2.32 480 -9.632 <.0001
## r - 3m -2.003 2.34 448 -0.855 0.8278
## r - 6m -6.246 2.38 451 -2.624 0.0443
## 3m - 6m -4.244 2.50 441 -1.698 0.3257
##
## group = Soteria:
## contrast estimate SE df t.ratio p.value
## b - r -16.708 1.76 469 -9.492 <.0001
## b - 3m -23.224 1.86 477 -12.485 <.0001
## b - 6m -26.769 1.90 479 -14.057 <.0001
## r - 3m -6.516 1.93 441 -3.382 0.0043
## r - 6m -10.061 1.97 444 -5.110 <.0001
## 3m - 6m -3.545 2.05 445 -1.731 0.3086
##
## group = Online:
## contrast estimate SE df t.ratio p.value
## b - r -7.682 2.37 466 -3.239 0.0070
## b - 3m -11.520 2.44 472 -4.727 <.0001
## b - 6m -10.840 2.76 486 -3.928 0.0006
## r - 3m -3.837 2.55 446 -1.505 0.4355
## r - 6m -3.158 2.86 461 -1.105 0.6865
## 3m - 6m 0.679 2.88 450 0.236 0.9954
##
## Results are averaged over the levels of: in_a_relationsip, lives_alone, years_education, mainhospitalizationreason, maindiagnosis, personalitycomorbidity, disability
## Degrees-of-freedom method: kenward-roger
## P value adjustment: tukey method for comparing a family of 4 estimates
MANSA
## Number of non-NA observations of mansa at every data collection point:
## # A tibble: 6 × 3
## time count_non_na percent_of_b
## <fct> <int> <dbl>
## 1 b 212 100
## 2 r 155 73.1
## 3 3m 147 69.3
## 4 6m 128 60.4
## 5 1y 81 38.2
## 6 1.5y 72 34.0## Model Summary:
## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: formula
## Data: data
##
## REML criterion at convergence: 1701.3
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -3.6762 -0.4872 -0.0294 0.4663 3.2471
##
## Random effects:
## Groups Name Variance Std.Dev.
## id (Intercept) 0.5709 0.7556
## Residual 0.5145 0.7173
## Number of obs: 633, groups: id, 209
##
## Fixed effects:
## Estimate Std. Error df t value
## (Intercept) 3.950623 0.336132 212.368076 11.753
## groupSoteria -0.008561 0.177466 346.632625 -0.048
## groupOnline 0.089355 0.211788 335.416803 0.422
## timer 0.313888 0.139425 451.943250 2.251
## time3m 0.459704 0.146992 456.872588 3.127
## time6m 0.403315 0.155279 458.463764 2.597
## age -0.004442 0.005148 201.469556 -0.863
## in_a_relationsipIn a relationship 0.215384 0.169233 202.949775 1.273
## lives_aloneYes 0.061275 0.164249 193.309133 0.373
## sourceincome1 -0.251590 0.137401 195.970753 -1.831
## years_education12 years and over 0.300670 0.184046 192.759537 1.634
## mainhospitalizationreasonOther -0.342947 0.178944 193.617936 -1.917
## maindiagnosisOther -0.119962 0.174945 195.380540 -0.686
## personalitycomorbidityYes -0.136532 0.163010 198.850267 -0.838
## groupSoteria:timer 0.029831 0.184812 453.100860 0.161
## groupOnline:timer -0.139530 0.209664 449.827373 -0.665
## groupSoteria:time3m -0.160138 0.189707 455.671011 -0.844
## groupOnline:time3m -0.300065 0.218212 454.257762 -1.375
## groupSoteria:time6m 0.042485 0.197880 456.960759 0.215
## groupOnline:time6m -0.143050 0.237002 457.718453 -0.604
## Pr(>|t|)
## (Intercept) < 2e-16 ***
## groupSoteria 0.96155
## groupOnline 0.67336
## timer 0.02485 *
## time3m 0.00188 **
## time6m 0.00970 **
## age 0.38928
## in_a_relationsipIn a relationship 0.20458
## lives_aloneYes 0.70951
## sourceincome1 0.06861 .
## years_education12 years and over 0.10396
## mainhospitalizationreasonOther 0.05677 .
## maindiagnosisOther 0.49371
## personalitycomorbidityYes 0.40328
## groupSoteria:timer 0.87184
## groupOnline:timer 0.50608
## groupSoteria:time3m 0.39904
## groupOnline:time3m 0.16978
## groupSoteria:time6m 0.83010
## groupOnline:time6m 0.54642
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## fit warnings:
## fixed-effect model matrix is rank deficient so dropping 1 column / coefficient
##
## Estimated Marginal Means (emmeans):
## group = Hospital:
## contrast estimate SE df t.ratio p.value
## b - r -0.3139 0.139 449 -2.250 0.1115
## b - 3m -0.4597 0.147 454 -3.126 0.0102
## b - 6m -0.4033 0.155 456 -2.596 0.0478
## r - 3m -0.1458 0.152 431 -0.959 0.7729
## r - 6m -0.0894 0.159 431 -0.563 0.9431
## 3m - 6m 0.0564 0.164 430 0.343 0.9861
##
## group = Soteria:
## contrast estimate SE df t.ratio p.value
## b - r -0.3437 0.121 452 -2.834 0.0247
## b - 3m -0.2996 0.120 451 -2.498 0.0616
## b - 6m -0.4458 0.123 452 -3.635 0.0018
## r - 3m 0.0442 0.128 428 0.346 0.9858
## r - 6m -0.1021 0.130 429 -0.785 0.8611
## 3m - 6m -0.1462 0.129 430 -1.131 0.6707
##
## group = Online:
## contrast estimate SE df t.ratio p.value
## b - r -0.1744 0.157 445 -1.113 0.6819
## b - 3m -0.1596 0.161 449 -0.989 0.7556
## b - 6m -0.2603 0.179 455 -1.453 0.4672
## r - 3m 0.0147 0.170 434 0.087 0.9998
## r - 6m -0.0859 0.187 442 -0.460 0.9677
## 3m - 6m -0.1006 0.188 435 -0.535 0.9505
##
## Results are averaged over the levels of: in_a_relationsip, lives_alone, years_education, mainhospitalizationreason, maindiagnosis, personalitycomorbidity, disability
## Degrees-of-freedom method: kenward-roger
## P value adjustment: tukey method for comparing a family of 4 estimates