Carga de base

Carga de librerías

library(tidyverse)

## Warning: package 'ggplot2' was built under R version 4.4.3

## ── Attaching core tidyverse packages ──────────────────────── tidyverse 2.0.0 ──
## ✔ dplyr     1.1.4     ✔ readr     2.1.5
## ✔ forcats   1.0.0     ✔ stringr   1.5.1
## ✔ ggplot2   4.0.0     ✔ tibble    3.2.1
## ✔ lubridate 1.9.3     ✔ tidyr     1.3.1
## ✔ purrr     1.0.2     
## ── Conflicts ────────────────────────────────────────── tidyverse_conflicts() ──
## ✖ dplyr::filter() masks stats::filter()
## ✖ dplyr::lag()    masks stats::lag()
## ℹ Use the conflicted package (<http://conflicted.r-lib.org/>) to force all conflicts to become errors

library(knitr)
library(tableone)
library(lme4)

## Cargando paquete requerido: Matrix
## 
## Adjuntando el paquete: 'Matrix'
## 
## The following objects are masked from 'package:tidyr':
## 
##     expand, pack, unpack

library(sjPlot)

## 
## Adjuntando el paquete: 'sjPlot'
## 
## The following object is masked from 'package:ggplot2':
## 
##     set_theme

library(performance)

## Warning: package 'performance' was built under R version 4.4.3

library(lmerTest)

## Warning: package 'lmerTest' was built under R version 4.4.3

## 
## Adjuntando el paquete: 'lmerTest'
## 
## The following object is masked from 'package:lme4':
## 
##     lmer
## 
## The following object is masked from 'package:stats':
## 
##     step

#library(rworldmap) # no es necesario 
#library(countrycode) # no es necesario
library(dplyr)
library(performance)
library(DHARMa)

## Warning: package 'DHARMa' was built under R version 4.4.3

## This is DHARMa 0.4.7. For overview type '?DHARMa'. For recent changes, type news(package = 'DHARMa')

library(ggeffects)
library(naniar)

## Warning: package 'naniar' was built under R version 4.4.3

library(marginaleffects)

## Warning: package 'marginaleffects' was built under R version 4.4.3

library(data.table)

## Warning: package 'data.table' was built under R version 4.4.3

## 
## Adjuntando el paquete: 'data.table'
## 
## The following objects are masked from 'package:lubridate':
## 
##     hour, isoweek, mday, minute, month, quarter, second, wday, week,
##     yday, year
## 
## The following objects are masked from 'package:dplyr':
## 
##     between, first, last
## 
## The following object is masked from 'package:purrr':
## 
##     transpose

library(haven)
library(lme4)
library(geepack)

## Warning: package 'geepack' was built under R version 4.4.3

library(performance)
library(psych)

## 
## Adjuntando el paquete: 'psych'
## 
## The following objects are masked from 'package:ggplot2':
## 
##     %+%, alpha

library(sjstats)

## 
## Adjuntando el paquete: 'sjstats'
## 
## The following object is masked from 'package:psych':
## 
##     phi
## 
## The following objects are masked from 'package:performance':
## 
##     icc, r2

options(scipen=999, digits=2)

Datos faltantes

vis_miss(TES)

Eliminar datos faltantes de las columnas de caract del pie

TES <- TES %>%
  filter(!if_any(16:25, is.na))

Pasar variables nuéricas a factor

TES[sapply(TES, is.character)] <- lapply(TES[sapply(TES, is.character)], as.factor)
TES$Cicatrización <- as.factor(TES$Cicatrización)


TES <- TES %>%
  mutate(across(16:25, as.factor))
TES <- TES %>%
  mutate(across(5:13, as.factor))
TES <- TES %>%
  mutate(across(OM, as.factor))

Evaluar cuántos pacientes cicatrizaron

table1<-table(TES$Cicatrización)
table1

## 
##   0   1 
## 124 175

prop.table(table1)

## 
##    0    1 
## 0.41 0.59

Distribución de variables según outcomes

TES$cicat=TES$Cicatrización
#Isquemia
TES%>%
  dplyr::group_by(isq) %>%
  dplyr::summarise(tar = mean (as.numeric(cicat)-1)) %>%
  ggplot(aes(x = isq, y = tar)) +
  geom_point() +
  geom_smooth(method = "lm", col = "blue") +
  geom_text(aes(label = round(tar, 2)), vjust = -0.5)

## `geom_smooth()` using formula = 'y ~ x'

TES %>%
  dplyr::group_by(isq) %>%
  dplyr::summarise(tar = mean(as.numeric(cicat)-1)) %>%
  ggplot(aes(x = isq, y = tar)) +
  geom_col(fill = "steelblue") +
  geom_text(aes(label = round(tar, 2)), vjust = -0.3) +
  labs(x = "isquemia", y = "Proporción de cicatrizados") +
  theme_minimal()

#localizacion
TES%>%
  dplyr::group_by(loca) %>%
  dplyr::summarise(tar = mean (as.numeric(cicat)-1)) %>%
  ggplot(aes(x = loca, y = tar)) +
  geom_point() +
  geom_smooth(method = "lm", col = "blue") +
  geom_text(aes(label = round(tar, 2)), vjust = -0.5)

## `geom_smooth()` using formula = 'y ~ x'

TES %>%
  dplyr::group_by(loca) %>%
  dplyr::summarise(tar = mean(as.numeric(cicat)-1)) %>%
  ggplot(aes(x = loca, y = tar)) +
  geom_col(fill = "steelblue") +
  geom_text(aes(label = round(tar, 2)), vjust = -0.3) +
  labs(x = "localizacion", y = "Proporción de cicatrizados") +
  theme_minimal()

#topo
TES%>%
  dplyr::group_by(topo) %>%
  dplyr::summarise(tar = mean (as.numeric(cicat)-1)) %>%
  ggplot(aes(x = topo, y = tar)) +
  geom_point() +
  geom_smooth(method = "lm", col = "blue") +
  geom_text(aes(label = round(tar, 2)), vjust = -0.5)

## `geom_smooth()` using formula = 'y ~ x'

TES %>%
  dplyr::group_by(topo) %>%
  dplyr::summarise(tar = mean(as.numeric(cicat)-1)) %>%
  ggplot(aes(x = topo, y = tar)) +
  geom_col(fill = "steelblue") +
  geom_text(aes(label = round(tar, 2)), vjust = -0.3) +
  labs(x = "topografia", y = "Proporción de cicatrizados") +
  theme_minimal()

#numero
TES%>%
  dplyr::group_by(nume) %>%
  dplyr::summarise(tar = mean (as.numeric(cicat)-1)) %>%
  ggplot(aes(x = nume, y = tar)) +
  geom_point() +
  geom_smooth(method = "lm", col = "blue") +
  geom_text(aes(label = round(tar, 2)), vjust = -0.5)

## `geom_smooth()` using formula = 'y ~ x'

TES %>%
  dplyr::group_by(nume) %>%
  dplyr::summarise(tar = mean(as.numeric(cicat)-1)) %>%
  ggplot(aes(x = nume, y = tar)) +
  geom_col(fill = "steelblue") +
  geom_text(aes(label = round(tar, 2)), vjust = -0.3) +
  labs(x = "numero de zonas", y = "Proporción de cicatrizados") +
  theme_minimal()

#infeccion
TES%>%
  dplyr::group_by(infe) %>%
  dplyr::summarise(tar = mean (as.numeric(cicat)-1)) %>%
  ggplot(aes(x = infe, y = tar)) +
  geom_point() +
  geom_smooth(method = "lm", col = "blue") +
  geom_text(aes(label = round(tar, 2)), vjust = -0.5)

## `geom_smooth()` using formula = 'y ~ x'

TES %>%
  dplyr::group_by(infe) %>%
  dplyr::summarise(tar = mean(as.numeric(cicat)-1)) %>%
  ggplot(aes(x = infe, y = tar)) +
  geom_col(fill = "steelblue") +
  geom_text(aes(label = round(tar, 2)), vjust = -0.3) +
  labs(x = "infeccion", y = "Proporción de cicatrizados") +
  theme_minimal()

#edema
TES%>%
  dplyr::group_by(ede) %>%
  dplyr::summarise(tar = mean (as.numeric(cicat)-1)) %>%
  ggplot(aes(x = ede, y = tar)) +
  geom_point() +
  geom_smooth(method = "lm", col = "blue") +
  geom_text(aes(label = round(tar, 2)), vjust = -0.5)

## `geom_smooth()` using formula = 'y ~ x'

TES %>%
  dplyr::group_by(ede) %>%
  dplyr::summarise(tar = mean(as.numeric(cicat)-1)) %>%
  ggplot(aes(x = ede, y = tar)) +
  geom_col(fill = "steelblue") +
  geom_text(aes(label = round(tar, 2)), vjust = -0.3) +
  labs(x = "edema", y = "Proporción de cicatrizados") +
  theme_minimal()

#neuropatia
TES%>%
  dplyr::group_by(neur) %>%
  dplyr::summarise(tar = mean (as.numeric(cicat)-1)) %>%
  ggplot(aes(x = neur, y = tar)) +
  geom_point() +
  geom_smooth(method = "lm", col = "blue") +
  geom_text(aes(label = round(tar, 2)), vjust = -0.5)

## `geom_smooth()` using formula = 'y ~ x'

TES %>%
  dplyr::group_by(neur) %>%
  dplyr::summarise(tar = mean(as.numeric(cicat)-1)) %>%
  ggplot(aes(x = neur, y = tar)) +
  geom_col(fill = "steelblue") +
  geom_text(aes(label = round(tar, 2)), vjust = -0.3) +
  labs(x = "Neuropatia", y = "Proporción de cicatrizados") +
  theme_minimal()

#tisu profundidad
TES%>%
  dplyr::group_by(tisu) %>%
  dplyr::summarise(tar = mean (as.numeric(cicat)-1)) %>%
  ggplot(aes(x = tisu, y = tar)) +
  geom_point() +
  geom_smooth(method = "lm", col = "blue") +
  geom_text(aes(label = round(tar, 2)), vjust = -0.5)

## `geom_smooth()` using formula = 'y ~ x'

TES %>%
  dplyr::group_by(tisu) %>%
  dplyr::summarise(tar = mean(as.numeric(cicat)-1)) %>%
  ggplot(aes(x = tisu, y = tar)) +
  geom_col(fill = "steelblue") +
  geom_text(aes(label = round(tar, 2)), vjust = -0.3) +
  labs(x = "Profundidad", y = "Proporción de cicatrizados") +
  theme_minimal()

#area
TES%>%
  dplyr::group_by(area) %>%
  dplyr::summarise(tar = mean (as.numeric(cicat)-1)) %>%
  ggplot(aes(x = area, y = tar)) +
  geom_point() +
  geom_smooth(method = "lm", col = "blue") +
  geom_text(aes(label = round(tar, 2)), vjust = -0.5)

## `geom_smooth()` using formula = 'y ~ x'

TES %>%
  dplyr::group_by(area) %>%
  dplyr::summarise(tar = mean(as.numeric(cicat)-1)) %>%
  ggplot(aes(x = area, y = tar)) +
  geom_col(fill = "steelblue") +
  geom_text(aes(label = round(tar, 2)), vjust = -0.3) +
  labs(x = "Area", y = "Proporción de cicatrizados") +
  theme_minimal()

#cicat
TES%>%
  dplyr::group_by(cica) %>%
  dplyr::summarise(tar = mean (as.numeric(cicat)-1)) %>%
  ggplot(aes(x = cica, y = tar)) +
  geom_point() +
  geom_smooth(method = "lm", col = "blue") +
  geom_text(aes(label = round(tar, 2)), vjust = -0.5)

## `geom_smooth()` using formula = 'y ~ x'

TES %>%
  dplyr::group_by(cica) %>%
  dplyr::summarise(tar = mean(as.numeric(cicat)-1)) %>%
  ggplot(aes(x = cica, y = tar)) +
  geom_col(fill = "steelblue") +
  geom_text(aes(label = round(tar, 2)), vjust = -0.3) +
  labs(x = "Fase de cicatrizacion", y = "Proporción de cicatrizados") +
  theme_minimal()

table(TES$cicat)

## 
##   0   1 
## 124 175

Categorización de las variables

#ISQuEMIA
TES <- TES %>%
  mutate(isq_factor = factor(if_else(isq %in% c(2, 3), "SI", "NO/leve")),
         isq_factor = factor(isq_factor, levels = rev(levels(isq_factor))))

table(TES$isq)

## 
##   0   1   2   3 
## 157  56  50  36

table(TES$isq_factor)

## 
##      SI NO/leve 
##      86     213

#localizacion
TES <- TES %>%
  mutate(loca_factor = factor(if_else(loca %in% c(2, 3), "Met/Tar", "Fal")),
         loca_factor = factor(loca_factor, levels = rev(levels(loca_factor))))


table(TES$loca)

## 
##   1   2   3 
## 166  99  34

table(TES$loca_factor)

## 
## Met/Tar     Fal 
##     133     166

#topog
TES <- TES %>%
  mutate(topo_factor = factor(if_else(topo %in% c(1,2),  "dor/plan/lat", "Mas de dos")))
TES <- TES %>%
  mutate(topo_factor = factor(if_else(topo %in% c(1, 2), "dor/plan/lat", "Mas de dos")),
         topo_factor = factor(topo_factor, levels = rev(levels(topo_factor))))

table(TES$topo)

## 
##   1   2   3 
## 136  60 103

table(TES$topo_factor)

## 
##   Mas de dos dor/plan/lat 
##          103          196

#nume
TES <- TES %>%
  mutate(nume_factor = factor(if_else(nume %in% c(2, 3),  "Dos o mas", "Una")))

table(TES$nume)

## 
##   1   2   3 
## 211  59  29

table(TES$nume_factor)

## 
## Dos o mas       Una 
##        88       211

#infe
TES <- TES %>%
  mutate(infe_factor = case_when(
    infe == 0 ~ "No",
    infe %in% c(1, 2) ~ "Leve/Mod",
    infe == 3 ~ "Grave",
    TRUE ~ NA_character_     # conserva los NA originales
  )) %>%
  mutate(infe_factor = factor(infe_factor, levels = c("No", "Leve/Mod", "Grave")))
TES$infe_factor <- factor (TES$infe_factor, levels= c ("Grave","Leve/Mod","No"))

table(TES$infe)

## 
##   0   1   2   3 
##  67  43 152  37

table(TES$infe_factor)

## 
##    Grave Leve/Mod       No 
##       37      195       67

#edema
TES <- TES %>%
  mutate(ede_factor = case_when(
    ede == 2 ~ "Unilateral",
    ede %in% c(0, 1) ~ "Sin/local",
    ede == 3 ~ "Bilateral",
    TRUE ~ NA_character_     # conserva los NA originales
  )) %>%
  mutate(ede_factor = factor(ede_factor, levels = c("Sin/local", "Unilateral", "Bilateral")))
TES$ede_factor <- factor (TES$ede_factor, levels= c ("Bilateral","Unilateral","Sin/local"))
table(TES$ede)

## 
##  0  1  2  3 
## 92 98 89 20

table(TES$ede_factor)

## 
##  Bilateral Unilateral  Sin/local 
##         20         89        190

#neur
TES <- TES %>%
  mutate(neur_factor = factor(
    if_else(neur %in% c(1, 2, 3), "Leve/mod/charcot", "No"),
    levels = c("No", "Leve/mod/charcot")
  ))
table(TES$neur)

## 
##   0   1   2   3 
##  10  57 216  16

table(TES$neur_factor)

## 
##               No Leve/mod/charcot 
##               10              289

#tisu (prof)
TES <- TES %>%
  mutate(tisu_factor = factor(if_else(tisu %in% c(2, 3),  "parcial/total", "Piel")))
table(TES$tisu)

## 
##   1   2   3 
##  52  98 149

table(TES$tisu_factor)

## 
## parcial/total          Piel 
##           247            52

#area
TES$area_factor <- with(TES, ifelse(area == 1, "Menor a 10",
                             ifelse(area %in% c(1, 2), "10 a 40",
                             ifelse(area == 3, "Mas de 40", NA))))
TES$area_factor <- factor(TES$area_factor, levels = c("Menor a 10", "10 a 40", "Mas de 40"))

table(TES$area)

## 
##   1   2   3 
## 169 103  27

table(TES$area_factor)

## 
## Menor a 10    10 a 40  Mas de 40 
##        169        103         27

#cica(fase)
TES <- TES %>%
  mutate(cica_factor = factor(if_else(cica %in% c(2, 3),  "Granu/infla", "Epit")))
TES$cica_factor <- factor (TES$cica_factor, levels= c ("Granu/infla","Epit"))
table(TES$cica)

## 
##   1   2   3 
##  14  48 237

table(TES$cica_factor)

## 
## Granu/infla        Epit 
##         285          14

Tabla 1

#Creo variable muerte 1 si/no para que la tabla se vea mejor
TES$cicat1 <- factor(TES$cicat,
                       levels = c(0, 1),
                       labels = c("No cicatrizo", "Cicatrizo"))

catvars = c( "isq_factor", "loca_factor", "topo_factor", "nume_factor", "infe_factor", "ede_factor", "neur_factor", "tisu_factor", "area_factor", "cica_factor", "OM","tbq", "aamen", "aamay", "dial", "ICC", "HTA", "CV", "sexo1fem")


vars = c("isq_factor", "loca_factor", "topo_factor", "nume_factor", "infe_factor", "ede_factor", "neur_factor", "tisu_factor", "area_factor", "cica_factor", "OM", "sanelian","SINBAD","tbq", "aamen", "aamay", "dial", "ICC", "HTA", "CV","Edad", "sexo1fem" )


tabla_1a <- CreateTableOne(vars = vars, strata = "cicat1", factorVars = catvars, data = TES)

print(tabla_1a)

##                                     Stratified by cicat1
##                                      No cicatrizo  Cicatrizo     p      test
##   n                                    124           175                    
##   isq_factor = NO/leve (%)              70 (56.5)    143 (81.7)  <0.001     
##   loca_factor = Fal (%)                 59 (47.6)    107 (61.1)   0.027     
##   topo_factor = dor/plan/lat (%)        70 (56.5)    126 (72.0)   0.008     
##   nume_factor = Una (%)                 62 (50.0)    149 (85.1)  <0.001     
##   infe_factor (%)                                                 0.002     
##      Grave                              23 (18.5)     14 ( 8.0)             
##      Leve/Mod                           83 (66.9)    112 (64.0)             
##      No                                 18 (14.5)     49 (28.0)             
##   ede_factor (%)                                                 <0.001     
##      Bilateral                          14 (11.3)      6 ( 3.4)             
##      Unilateral                         46 (37.1)     43 (24.6)             
##      Sin/local                          64 (51.6)    126 (72.0)             
##   neur_factor = Leve/mod/charcot (%)   123 (99.2)    166 (94.9)   0.084     
##   tisu_factor = Piel (%)                 9 ( 7.3)     43 (24.6)  <0.001     
##   area_factor (%)                                                 0.002     
##      Menor a 10                         57 (46.0)    112 (64.0)             
##      10 a 40                            49 (39.5)     54 (30.9)             
##      Mas de 40                          18 (14.5)      9 ( 5.1)             
##   cica_factor = Epit (%)                 1 ( 0.8)     13 ( 7.4)   0.017     
##   OM = 1 (%)                            65 (52.4)     74 (42.3)   0.107     
##   sanelian (mean (SD))               18.59 (4.03)  15.49 (3.96)  <0.001     
##   SINBAD (mean (SD))                  4.40 (1.12)   3.62 (1.15)  <0.001     
##   tbq (%)                                                         0.414     
##      0                                   3 ( 2.4)     11 ( 6.3)             
##      1                                  29 (23.4)     34 (19.5)             
##      2                                  61 (49.2)     86 (49.4)             
##      3                                  31 (25.0)     43 (24.7)             
##   aamen = 1 (%)                         40 (32.3)     41 (23.4)   0.119     
##   aamay = 1 (%)                         11 ( 8.9)      6 ( 3.4)   0.080     
##   dial = 1 (%)                          13 (10.5)     13 ( 7.4)   0.474     
##   ICC = 1 (%)                           20 (16.1)     20 (11.4)   0.315     
##   HTA = 1 (%)                           95 (76.6)    120 (68.6)   0.163     
##   CV = 1 (%)                            30 (24.2)     43 (24.6)   1.000     
##   Edad (mean (SD))                   59.00 (12.98) 56.98 (11.36)  0.154     
##   sexo1fem = 1 (%)                      29 (23.4)     36 (20.6)   0.660

kableone(tabla_1a)

	No cicatrizo	Cicatrizo	p
n	124	175
isq_factor = NO/leve (%)	70 (56.5)	143 (81.7)	<0.001
loca_factor = Fal (%)	59 (47.6)	107 (61.1)	0.027
topo_factor = dor/plan/lat (%)	70 (56.5)	126 (72.0)	0.008
nume_factor = Una (%)	62 (50.0)	149 (85.1)	<0.001
infe_factor (%)			0.002
Grave	23 (18.5)	14 ( 8.0)
Leve/Mod	83 (66.9)	112 (64.0)
No	18 (14.5)	49 (28.0)
ede_factor (%)			<0.001
Bilateral	14 (11.3)	6 ( 3.4)
Unilateral	46 (37.1)	43 (24.6)
Sin/local	64 (51.6)	126 (72.0)
neur_factor = Leve/mod/charcot (%)	123 (99.2)	166 (94.9)	0.084
tisu_factor = Piel (%)	9 ( 7.3)	43 (24.6)	<0.001
area_factor (%)			0.002
Menor a 10	57 (46.0)	112 (64.0)
10 a 40	49 (39.5)	54 (30.9)
Mas de 40	18 (14.5)	9 ( 5.1)
cica_factor = Epit (%)	1 ( 0.8)	13 ( 7.4)	0.017
OM = 1 (%)	65 (52.4)	74 (42.3)	0.107
sanelian (mean (SD))	18.59 (4.03)	15.49 (3.96)	<0.001
SINBAD (mean (SD))	4.40 (1.12)	3.62 (1.15)	<0.001
tbq (%)			0.414
0	3 ( 2.4)	11 ( 6.3)
1	29 (23.4)	34 (19.5)
2	61 (49.2)	86 (49.4)
3	31 (25.0)	43 (24.7)
aamen = 1 (%)	40 (32.3)	41 (23.4)	0.119
aamay = 1 (%)	11 ( 8.9)	6 ( 3.4)	0.080
dial = 1 (%)	13 (10.5)	13 ( 7.4)	0.474
ICC = 1 (%)	20 (16.1)	20 (11.4)	0.315
HTA = 1 (%)	95 (76.6)	120 (68.6)	0.163
CV = 1 (%)	30 (24.2)	43 (24.6)	1.000
Edad (mean (SD))	59.00 (12.98)	56.98 (11.36)	0.154
sexo1fem = 1 (%)	29 (23.4)	36 (20.6)	0.660

table(TES$cicat1)

## 
## No cicatrizo    Cicatrizo 
##          124          175

UNIVARIADO

TES$cicat=TES$Cicatrización
modelo1 <- glm(cicat ~ loca_factor+ topo_factor+nume_factor+ isq_factor+ infe_factor+ ede_factor+ neur_factor+ tisu_factor+ area_factor+ cica_factor , family = "binomial", data = TES)
t1<-tab_model(modelo1, show.se = T, show.dev = T, show.loglik = T, show.intercept = F,show.stat = F, show.reflvl = T, digits = 2, digits.p = 3, show.r2 = F,collapse.ci = F, title = "Cicatrización")
knitr::asis_output(t1$knitr)

Cicatrización
	cicat
Predictors	Odds Ratios	std. Error	CI	p
Met/Tar	Reference
Fal	1.89	0.55	1.08 – 3.35	0.027
Mas de dos	Reference
dor/plan/lat	0.94	0.33	0.47 – 1.85	0.852
Dos o mas	Reference
Una	5.53	2.06	2.72 – 11.75	<0.001
SI	Reference
NO/leve	3.44	1.07	1.89 – 6.38	<0.001
Grave	Reference
Leve/Mod	2.31	1.07	0.94 – 5.85	0.070
No	2.56	1.53	0.80 – 8.41	0.115
Bilateral	Reference
Unilateral	1.98	1.19	0.63 – 6.76	0.256
Sin/local	2.45	1.40	0.82 – 7.96	0.116
No	Reference
Leve/mod/charcot	0.30	0.34	0.02 – 1.90	0.284
parcial/total	Reference
Piel	1.89	0.90	0.76 – 4.99	0.183
Menor a 10	Reference
10 a 40	1.75	0.64	0.87 – 3.66	0.127
Mas de 40	1.96	1.23	0.57 – 6.88	0.286
Granu/infla	Reference
Epit	4.18	4.60	0.68 – 80.84	0.194
Observations	299
Deviance	321.666
log-Likelihood	-160.833

modelo2 <- glm(cicat ~ loca_factor , family = "binomial", data = TES)
modelo3 <- glm(cicat ~  topo_factor , family = "binomial", data = TES)
modelo4 <- glm(cicat ~ nume_factor , family = "binomial", data = TES)
modelo5 <- glm(cicat ~  isq_factor , family = "binomial", data = TES)
modelo6<- glm(cicat ~  infe_factor, family = "binomial", data = TES)
modelo7<- glm(cicat ~  ede_factor , family = "binomial", data = TES)
modelo8<- glm(cicat ~  neur_factor , family = "binomial", data = TES)
modelo9<- glm(cicat ~  tisu_factor , family = "binomial", data = TES)
modelo10<- glm(cicat ~  area_factor , family = "binomial", data = TES)
modelo11<- glm(cicat ~  cica_factor, family = "binomial", data = TES)
modelo12<- glm(cicat ~  evol, family = "binomial", data = TES)

library(broom)

## 
## Adjuntando el paquete: 'broom'

## The following object is masked from 'package:sjstats':
## 
##     bootstrap

library(dplyr)

# Crear una lista con los modelos
modelos <- list(
  modelo2, modelo3, modelo4, modelo5,
  modelo6, modelo7, modelo8, modelo9,
  modelo10, modelo11, modelo12
)

# Convertir a tabla ordenada
tabla_coef <- lapply(modelos, tidy) %>%
  bind_rows(.id = "modelo")

# Ver la tabla
tabla_coef

## # A tibble: 25 × 6
##    modelo term                    estimate std.error statistic  p.value
##    <chr>  <chr>                      <dbl>     <dbl>     <dbl>    <dbl>
##  1 1      (Intercept)               0.0451     0.173     0.260 7.95e- 1
##  2 1      loca_factorFal            0.550      0.237     2.32  2.05e- 2
##  3 2      (Intercept)              -0.0972     0.197    -0.492 6.22e- 1
##  4 2      topo_factordor/plan/lat   0.685      0.247     2.77  5.61e- 3
##  5 3      (Intercept)              -0.869      0.234    -3.72  2.00e- 4
##  6 3      nume_factorUna            1.75       0.278     6.27  3.52e-10
##  7 4      (Intercept)              -0.523      0.223    -2.35  1.90e- 2
##  8 4      isq_factorNO/leve         1.24       0.267     4.64  3.43e- 6
##  9 5      (Intercept)              -0.496      0.339    -1.46  1.43e- 1
## 10 5      infe_factorLeve/Mod       0.796      0.369     2.16  3.08e- 2
## # ℹ 15 more rows

#ordenar p valor
tabla_coef_ordenada <- tabla_coef %>%
  arrange(p.value)
# Ver la tabla
tabla_coef_ordenada

## # A tibble: 25 × 6
##    modelo term                    estimate std.error statistic  p.value
##    <chr>  <chr>                      <dbl>     <dbl>     <dbl>    <dbl>
##  1 3      nume_factorUna             1.75      0.278      6.27 3.52e-10
##  2 4      isq_factorNO/leve          1.24      0.267      4.64 3.43e- 6
##  3 9      (Intercept)                0.675     0.163      4.15 3.30e- 5
##  4 3      (Intercept)               -0.869     0.234     -3.72 2.00e- 4
##  5 8      tisu_factorPiel            1.43      0.388      3.67 2.38e- 4
##  6 11     (Intercept)                0.511     0.141      3.63 2.82e- 4
##  7 5      infe_factorNo              1.50      0.437      3.43 6.07e- 4
##  8 9      area_factorMas de 40      -1.37      0.439     -3.11 1.84e- 3
##  9 6      ede_factorSin/local        1.52      0.512      2.98 2.88e- 3
## 10 2      topo_factordor/plan/lat    0.685     0.247      2.77 5.61e- 3
## # ℹ 15 more rows

evaluación del ICC Aquí va una explicación clara y directa de por qué pasa esto y cómo se interpreta:

📌 Por qué el ICC cambia entre el modelo nulo y el modelo mixto

1. Modelo nulo (solo intercepto) En el modelo nulo, el ICC refleja qué proporción de la variabilidad total del desenlace se debe a las diferencias entre grupos (por ejemplo, pacientes, hospitales, operadores, etc.). Si el ICC es 0.04, significa que solo el 4% de la variabilidad está explicada por el efecto aleatorio (las diferencias entre sujetos/grupos). 👉 Interpretación: El agrupamiento explica muy poco antes de ajustar por predictores.

2. Modelo mixto con variables fijas Cuando agregás variables explicativas (factores fijos), parte de la variabilidad dentro de los grupos (intra-grupo) se explica por esas covariables.

Eso puede provocar dos cosas:

La varianza residual disminuye porque explicás parte de la variación del outcome.
La varianza del efecto aleatorio puede subir o bajar según cómo las nuevas variables redistribuyan la variabilidad.

Entonces, el ICC del modelo ajustado es:

[ ICC = ]

Si el ICC sube de 0.04 a 0.11, significa que:

✔️ Al ajustar por variables, la proporción de variabilidad atribuida a diferencias entre grupos aumentó.

Esto puede ocurrir por varios motivos:

🔸 Las covariables explican parte de la variabilidad dentro de los grupos, reduciendo la varianza residual → el ICC sube.

🔸 Las covariables “desenmascaran” variabilidad entre grupos que en el modelo nulo estaba oculta.

🔸 Hay variables que actúan diferente entre grupos, aumentando la heterogeneidad entre ellos.

📌 Interpretación práctica

✔️ Modelo nulo: ICC = 0.04

Las diferencias entre los grupos explican muy poco del desenlace.

✔️ Modelo ajustado: ICC = 0.11

Una vez que tenés en cuenta factores como isquemia, infección, neuropatía, etc., los grupos (por ejemplo pacientes/operadores/hospitales, según tu modelo) pasan a explicar más variabilidad del outcome.

📌 Conclusión: ¿qué significa que suba el ICC?

➡️ El agrupamiento adquiere más importancia relativa después del ajuste. ➡️ Las variables fijas están explicando variabilidad dentro de los grupos, dejando más clara la contribución del efecto aleatorio. ➡️ No es un error y no significa que los grupos estén “más correlacionados”, sino que su aporte es mayor en relación al total reducido.

ICC

TES$cicat_num <- as.numeric(as.character(TES$cicat))

mod_aleatorio_NULO <- lmer(cicat_num ~ (1 | NOMBRE), REML = T, data = TES)


tab_model(mod_aleatorio_NULO)

	cicat_num
Predictors	Estimates	CI	p
(Intercept)	0.60	0.51 – 0.68	<0.001
Random Effects
σ²	0.24
τ₀₀ _NOMBRE	0.01
ICC	0.04
N _NOMBRE	15
Observations	299
Marginal R² / Conditional R²	0.000 / 0.037

performance::icc(mod_aleatorio_NULO)

## # Intraclass Correlation Coefficient
## 
##     Adjusted ICC: 0.037
##   Unadjusted ICC: 0.037

library(broom.mixed)

## Warning: package 'broom.mixed' was built under R version 4.4.3

tidy(mod_aleatorio_NULO)

## # A tibble: 3 × 8
##   effect   group    term            estimate std.error statistic    df   p.value
##   <chr>    <chr>    <chr>              <dbl>     <dbl>     <dbl> <dbl>     <dbl>
## 1 fixed    <NA>     (Intercept)       0.597     0.0418      14.3  5.63   1.24e-5
## 2 ran_pars NOMBRE   sd__(Intercept)   0.0961   NA           NA   NA     NA      
## 3 ran_pars Residual sd__Observation   0.488    NA           NA   NA     NA

MODELO MIXTO 1

mod_mixto <- glmer(cicat ~ nume_factor + (1 | NOMBRE),
                   family = binomial,
                   data = TES)

## boundary (singular) fit: see help('isSingular')

tab_model(mod_mixto)

	cicat
Predictors	Odds Ratios	CI	p
(Intercept)	0.42	0.27 – 0.66	<0.001
nume factor [Una]	5.73	3.32 – 9.89	<0.001
Random Effects
σ²	3.29
τ₀₀ _NOMBRE	0.00
N _NOMBRE	15
Observations	299
Marginal R² / Conditional R²	0.162 / NA

MODELO MIXTO 2

mod_mixto2 <- glmer(cicat ~ nume_factor + isq_factor + (1 | NOMBRE),
                   family = binomial,
                   data = TES)


tab_model(mod_mixto2)

	cicat
Predictors	Odds Ratios	CI	p
(Intercept)	0.17	0.09 – 0.34	<0.001
nume factor [Una]	5.79	3.24 – 10.33	<0.001
isq factor [NO/leve]	3.46	1.95 – 6.13	<0.001
Random Effects
σ²	3.29
τ₀₀ _NOMBRE	0.03
ICC	0.01
N _NOMBRE	15
Observations	299
Marginal R² / Conditional R²	0.242 / 0.248

MODELO MIXTO 3

mod_mixto3 <- glmer(cicat ~ nume_factor + isq_factor +infe_factor +(1 | NOMBRE),
                   family = binomial,
                   data = TES)


tab_model(mod_mixto3)

	cicat
Predictors	Odds Ratios	CI	p
(Intercept)	0.08	0.03 – 0.24	<0.001
nume factor [Una]	5.12	2.81 – 9.30	<0.001
isq factor [NO/leve]	3.73	2.06 – 6.74	<0.001
infe factor [Leve/Mod]	2.19	0.90 – 5.36	0.086
infe factor [No]	3.31	1.16 – 9.47	0.025
Random Effects
σ²	3.29
τ₀₀ _NOMBRE	0.13
ICC	0.04
N _NOMBRE	15
Observations	299
Marginal R² / Conditional R²	0.270 / 0.297

MODELO MIXTO 4

mod_mixto4 <- glmer(cicat ~ nume_factor + isq_factor + tisu_factor + infe_factor+(1 | NOMBRE),
                   family = binomial,
                   data = TES)


tab_model(mod_mixto4)

	cicat
Predictors	Odds Ratios	CI	p
(Intercept)	0.09	0.03 – 0.27	<0.001
nume factor [Una]	5.02	2.76 – 9.13	<0.001
isq factor [NO/leve]	3.36	1.84 – 6.12	<0.001
tisu factor [Piel]	2.26	0.91 – 5.61	0.079
infe factor [Leve/Mod]	1.96	0.82 – 4.71	0.130
infe factor [No]	2.21	0.73 – 6.71	0.161
Random Effects
σ²	3.29
τ₀₀ _NOMBRE	0.10
ICC	0.03
N _NOMBRE	15
Observations	299
Marginal R² / Conditional R²	0.285 / 0.306

MATRIZ DE CORRELACION PARA EVALUAR COLINEALIDAD TISU INFE. Impresiona confusión

library(dplyr)

TES %>%
  select(tisu_factor, infe_factor) %>%
  mutate(across(everything(), ~as.numeric(as.factor(.)))) %>%
  cor()

##             tisu_factor infe_factor
## tisu_factor        1.00        0.45
## infe_factor        0.45        1.00

#DA COlinealidad MODERADA

MODELO MIXTO 5

mod_mixto5 <- glmer(cicat ~ nume_factor + isq_factor + tisu_factor + infe_factor+ ede_factor+(1 | NOMBRE),
                   family = binomial,
                   data = TES)


tab_model(mod_mixto5)

	cicat
Predictors	Odds Ratios	CI	p
(Intercept)	0.04	0.01 – 0.18	<0.001
nume factor [Una]	4.83	2.59 – 9.01	<0.001
isq factor [NO/leve]	3.33	1.81 – 6.14	<0.001
tisu factor [Piel]	2.22	0.88 – 5.63	0.093
infe factor [Leve/Mod]	2.02	0.81 – 5.02	0.130
infe factor [No]	2.21	0.69 – 7.08	0.184
ede factor [Unilateral]	2.64	0.78 – 8.93	0.119
ede factor [Sin/local]	2.93	0.91 – 9.39	0.071
Random Effects
σ²	3.29
τ₀₀ _NOMBRE	0.17
ICC	0.05
N _NOMBRE	15
Observations	299
Marginal R² / Conditional R²	0.305 / 0.338

MODELO MIXTO 6

mod_mixto6 <- glmer(cicat ~ topo_factor + nume_factor + isq_factor + tisu_factor + infe_factor+ ede_factor+(1 | NOMBRE),
                   family = binomial,
                   data = TES)


tab_model(mod_mixto6)

	cicat
Predictors	Odds Ratios	CI	p
(Intercept)	0.04	0.01 – 0.20	<0.001
topo factor [dor/plan/lat]	0.73	0.37 – 1.45	0.369
nume factor [Una]	5.41	2.74 – 10.67	<0.001
isq factor [NO/leve]	3.33	1.81 – 6.15	<0.001
tisu factor [Piel]	2.28	0.90 – 5.78	0.084
infe factor [Leve/Mod]	2.02	0.81 – 5.06	0.132
infe factor [No]	2.34	0.72 – 7.64	0.158
ede factor [Unilateral]	2.54	0.75 – 8.58	0.134
ede factor [Sin/local]	3.00	0.94 – 9.62	0.064
Random Effects
σ²	3.29
τ₀₀ _NOMBRE	0.17
ICC	0.05
N _NOMBRE	15
Observations	299
Marginal R² / Conditional R²	0.307 / 0.341

MODELO MIXTO 7

mod_mixto7 <- glmer(cicat ~ topo_factor + nume_factor + isq_factor + tisu_factor + infe_factor+ ede_factor+ neur_factor + (1 | NOMBRE),
                   family = binomial,
                   data = TES)

## Warning in checkConv(attr(opt, "derivs"), opt$par, ctrl = control$checkConv, :
## Model failed to converge with max|grad| = 0.00328887 (tol = 0.002, component 1)

tab_model(mod_mixto7)

	cicat
Predictors	Odds Ratios	CI	p
(Intercept)	0.15	0.01 – 2.19	0.166
topo factor [dor/plan/lat]	0.73	0.37 – 1.46	0.377
nume factor [Una]	5.35	2.70 – 10.61	<0.001
isq factor [NO/leve]	3.40	1.83 – 6.32	<0.001
tisu factor [Piel]	2.07	0.80 – 5.33	0.133
infe factor [Leve/Mod]	2.03	0.80 – 5.11	0.134
infe factor [No]	2.37	0.72 – 7.83	0.157
ede factor [Unilateral]	2.52	0.75 – 8.55	0.137
ede factor [Sin/local]	3.00	0.93 – 9.63	0.065
neur factor [Leve/mod/charcot]	0.26	0.03 – 2.39	0.236
Random Effects
σ²	3.29
τ₀₀ _NOMBRE	0.21
ICC	0.06
N _NOMBRE	15
Observations	299
Marginal R² / Conditional R²	0.318 / 0.359

MODELO MIXTO 8

mod_mixto8 <- glmer(cicat ~ topo_factor + nume_factor + isq_factor + tisu_factor + infe_factor+ ede_factor+ neur_factor + (1 | NOMBRE),
                   family = binomial,
                   data = TES)

## Warning in checkConv(attr(opt, "derivs"), opt$par, ctrl = control$checkConv, :
## Model failed to converge with max|grad| = 0.00328887 (tol = 0.002, component 1)

tab_model(mod_mixto8)

	cicat
Predictors	Odds Ratios	CI	p
(Intercept)	0.15	0.01 – 2.19	0.166
topo factor [dor/plan/lat]	0.73	0.37 – 1.46	0.377
nume factor [Una]	5.35	2.70 – 10.61	<0.001
isq factor [NO/leve]	3.40	1.83 – 6.32	<0.001
tisu factor [Piel]	2.07	0.80 – 5.33	0.133
infe factor [Leve/Mod]	2.03	0.80 – 5.11	0.134
infe factor [No]	2.37	0.72 – 7.83	0.157
ede factor [Unilateral]	2.52	0.75 – 8.55	0.137
ede factor [Sin/local]	3.00	0.93 – 9.63	0.065
neur factor [Leve/mod/charcot]	0.26	0.03 – 2.39	0.236
Random Effects
σ²	3.29
τ₀₀ _NOMBRE	0.21
ICC	0.06
N _NOMBRE	15
Observations	299
Marginal R² / Conditional R²	0.318 / 0.359

MODELO MIXTO 9

mod_mixto9 <- glmer(cicat ~ topo_factor + nume_factor + isq_factor + tisu_factor + infe_factor+ ede_factor+ neur_factor + evol + (1 | NOMBRE),
                   family = binomial,
                   data = TES)

## Warning in checkConv(attr(opt, "derivs"), opt$par, ctrl = control$checkConv, :
## Model failed to converge with max|grad| = 0.00965086 (tol = 0.002, component 1)

tab_model(mod_mixto9)

	cicat
Predictors	Odds Ratios	CI	p
(Intercept)	0.16	0.01 – 2.45	0.187
topo factor [dor/plan/lat]	0.71	0.35 – 1.46	0.355
nume factor [Una]	5.56	2.75 – 11.22	<0.001
isq factor [NO/leve]	3.52	1.85 – 6.68	<0.001
tisu factor [Piel]	1.83	0.68 – 4.89	0.230
infe factor [Leve/Mod]	2.16	0.83 – 5.62	0.114
infe factor [No]	3.24	0.92 – 11.43	0.068
ede factor [Unilateral]	2.59	0.75 – 8.99	0.134
ede factor [Sin/local]	3.30	1.00 – 10.84	0.049
neur factor [Leve/mod/charcot]	0.27	0.03 – 2.59	0.259
evol	0.99	0.99 – 1.00	0.012
Random Effects
σ²	3.29
τ₀₀ _NOMBRE	0.42
ICC	0.11
N _NOMBRE	15
Observations	299
Marginal R² / Conditional R²	0.368 / 0.439

MODELO MIXTO 10

mod_mixto10 <- glmer(cicat ~ topo_factor + nume_factor + isq_factor + tisu_factor + infe_factor+ ede_factor+ neur_factor + loca_factor+ (1 | NOMBRE),
                   family = binomial,
                   data = TES)

## Warning in checkConv(attr(opt, "derivs"), opt$par, ctrl = control$checkConv, :
## Model failed to converge with max|grad| = 0.00346577 (tol = 0.002, component 1)

tab_model(mod_mixto10)

	cicat
Predictors	Odds Ratios	CI	p
(Intercept)	0.09	0.01 – 1.49	0.093
topo factor [dor/plan/lat]	0.81	0.40 – 1.64	0.561
nume factor [Una]	4.78	2.39 – 9.58	<0.001
isq factor [NO/leve]	3.69	1.96 – 6.97	<0.001
tisu factor [Piel]	2.08	0.80 – 5.39	0.132
infe factor [Leve/Mod]	2.23	0.88 – 5.70	0.093
infe factor [No]	2.76	0.82 – 9.30	0.101
ede factor [Unilateral]	2.86	0.83 – 9.91	0.097
ede factor [Sin/local]	3.14	0.97 – 10.16	0.056
neur factor [Leve/mod/charcot]	0.25	0.03 – 2.37	0.226
loca factor [Fal]	1.85	1.04 – 3.28	0.035
Random Effects
σ²	3.29
τ₀₀ _NOMBRE	0.24
ICC	0.07
N _NOMBRE	15
Observations	299
Marginal R² / Conditional R²	0.341 / 0.386

Modelo 11

mod_mixto11 <- glmer(cicat ~ topo_factor + nume_factor + isq_factor + tisu_factor + infe_factor+ ede_factor+ neur_factor + loca_factor+ evol+ (1 | NOMBRE),
                   family = binomial,
                   data = TES)

## Warning in checkConv(attr(opt, "derivs"), opt$par, ctrl = control$checkConv, :
## Model failed to converge with max|grad| = 0.00625735 (tol = 0.002, component 1)

tab_model(mod_mixto11)

	cicat
Predictors	Odds Ratios	CI	p
(Intercept)	0.11	0.01 – 1.78	0.120
topo factor [dor/plan/lat]	0.78	0.38 – 1.62	0.506
nume factor [Una]	5.06	2.49 – 10.31	<0.001
isq factor [NO/leve]	3.74	1.95 – 7.17	<0.001
tisu factor [Piel]	1.86	0.70 – 4.99	0.215
infe factor [Leve/Mod]	2.30	0.88 – 5.99	0.090
infe factor [No]	3.52	0.99 – 12.50	0.051
ede factor [Unilateral]	2.88	0.82 – 10.15	0.100
ede factor [Sin/local]	3.41	1.03 – 11.22	0.044
neur factor [Leve/mod/charcot]	0.26	0.03 – 2.55	0.247
loca factor [Fal]	1.67	0.93 – 3.00	0.088
evol	0.99	0.99 – 1.00	0.020
Random Effects
σ²	3.29
τ₀₀ _NOMBRE	0.42
ICC	0.11
N _NOMBRE	15
Observations	299
Marginal R² / Conditional R²	0.379 / 0.450

Modelo 12

mod_mixto12 <- glmer(cicat ~  nume_factor + isq_factor +  infe_factor+ ede_factor+ neur_factor +  evol+ (1 | NOMBRE),
                   family = binomial,
                   data = TES)


tab_model(mod_mixto12)

	cicat
Predictors	Odds Ratios	CI	p
(Intercept)	0.16	0.01 – 2.50	0.191
nume factor [Una]	4.93	2.59 – 9.41	<0.001
isq factor [NO/leve]	3.77	2.01 – 7.08	<0.001
infe factor [Leve/Mod]	2.28	0.88 – 5.92	0.090
infe factor [No]	3.90	1.21 – 12.57	0.023
ede factor [Unilateral]	2.64	0.76 – 9.20	0.127
ede factor [Sin/local]	3.31	1.00 – 10.95	0.050
neur factor [Leve/mod/charcot]	0.22	0.02 – 2.14	0.194
evol	0.99	0.99 – 1.00	0.010
Random Effects
σ²	3.29
τ₀₀ _NOMBRE	0.45
ICC	0.12
N _NOMBRE	15
Observations	299
Marginal R² / Conditional R²	0.359 / 0.436

modelo 13

##ELIJO ESTE COMO EFECTOS PRINCIPALES
mod_mixto13 <- glmer(cicat ~ isq_factor + infe_factor + loca_factor +  nume_factor+ ede_factor + evol +  (1 | NOMBRE),
                   family = binomial,
                   data = TES)

## Warning in checkConv(attr(opt, "derivs"), opt$par, ctrl = control$checkConv, :
## Model failed to converge with max|grad| = 0.00214939 (tol = 0.002, component 1)

tab_model(mod_mixto13)

	cicat
Predictors	Odds Ratios	CI	p
(Intercept)	0.02	0.00 – 0.14	<0.001
isq factor [NO/leve]	3.96	2.10 – 7.46	<0.001
infe factor [Leve/Mod]	2.46	0.94 – 6.39	0.066
infe factor [No]	4.47	1.38 – 14.42	0.012
loca factor [Fal]	1.68	0.95 – 2.99	0.076
nume factor [Una]	4.71	2.47 – 8.99	<0.001
ede factor [Unilateral]	2.93	0.83 – 10.29	0.094
ede factor [Sin/local]	3.49	1.05 – 11.52	0.041
evol	0.99	0.99 – 1.00	0.015
Random Effects
σ²	3.29
τ₀₀ _NOMBRE	0.39
ICC	0.11
N _NOMBRE	15
Observations	299
Marginal R² / Conditional R²	0.358 / 0.427

#modmixto13 con factor de confusion tisu ES ESTE EL MODELO!!!
mod_mixto13ti <- glmer(cicat ~ isq_factor + infe_factor + loca_factor +  nume_factor+ ede_factor + evol + tisu_factor+ (1 | NOMBRE),
                   family = binomial,
                   data = TES)

## Warning in checkConv(attr(opt, "derivs"), opt$par, ctrl = control$checkConv, :
## Model failed to converge with max|grad| = 0.00378723 (tol = 0.002, component 1)

tab_model(mod_mixto13ti)

	cicat
Predictors	Odds Ratios	CI	p
(Intercept)	0.03	0.00 – 0.14	<0.001
isq factor [NO/leve]	3.65	1.92 – 6.92	<0.001
infe factor [Leve/Mod]	2.26	0.87 – 5.82	0.092
infe factor [No]	3.26	0.95 – 11.20	0.061
loca factor [Fal]	1.70	0.95 – 3.02	0.072
nume factor [Una]	4.68	2.46 – 8.92	<0.001
ede factor [Unilateral]	3.02	0.86 – 10.59	0.085
ede factor [Sin/local]	3.38	1.03 – 11.12	0.045
evol	0.99	0.99 – 1.00	0.019
tisu factor [Piel]	2.00	0.76 – 5.26	0.160
Random Effects
σ²	3.29
τ₀₀ _NOMBRE	0.34
ICC	0.09
N _NOMBRE	15
Observations	299
Marginal R² / Conditional R²	0.370 / 0.429

modelo 14

mod_mixto14 <- glmer(cicat ~ topo_factor + nume_factor + isq_factor + infe_factor+ ede_factor+ neur_factor + loca_factor+ evol+ (1 | NOMBRE),
                   family = binomial,
                   data = TES)

## Warning in checkConv(attr(opt, "derivs"), opt$par, ctrl = control$checkConv, :
## Model failed to converge with max|grad| = 0.00561179 (tol = 0.002, component 1)

tab_model(mod_mixto14)

	cicat
Predictors	Odds Ratios	CI	p
(Intercept)	0.12	0.01 – 2.00	0.139
topo factor [dor/plan/lat]	0.79	0.38 – 1.65	0.532
nume factor [Una]	5.06	2.48 – 10.31	<0.001
isq factor [NO/leve]	4.04	2.12 – 7.68	<0.001
infe factor [Leve/Mod]	2.48	0.94 – 6.50	0.066
infe factor [No]	4.66	1.40 – 15.50	0.012
ede factor [Unilateral]	2.80	0.79 – 9.90	0.110
ede factor [Sin/local]	3.50	1.06 – 11.59	0.041
neur factor [Leve/mod/charcot]	0.22	0.02 – 2.14	0.192
loca factor [Fal]	1.65	0.92 – 2.97	0.093
evol	0.99	0.99 – 1.00	0.017
Random Effects
σ²	3.29
τ₀₀ _NOMBRE	0.48
ICC	0.13
N _NOMBRE	15
Observations	299
Marginal R² / Conditional R²	0.370 / 0.451

modelo 15

mod_mixto15 <- glmer(cicat ~ topo_factor + nume_factor + isq_factor + infe_factor+ ede_factor+ neur_factor + loca_factor+ evol+  (1 | NOMBRE),
                   family = binomial,
                   data = TES)

## Warning in checkConv(attr(opt, "derivs"), opt$par, ctrl = control$checkConv, :
## Model failed to converge with max|grad| = 0.00561179 (tol = 0.002, component 1)

tab_model(mod_mixto15)

	cicat
Predictors	Odds Ratios	CI	p
(Intercept)	0.12	0.01 – 2.00	0.139
topo factor [dor/plan/lat]	0.79	0.38 – 1.65	0.532
nume factor [Una]	5.06	2.48 – 10.31	<0.001
isq factor [NO/leve]	4.04	2.12 – 7.68	<0.001
infe factor [Leve/Mod]	2.48	0.94 – 6.50	0.066
infe factor [No]	4.66	1.40 – 15.50	0.012
ede factor [Unilateral]	2.80	0.79 – 9.90	0.110
ede factor [Sin/local]	3.50	1.06 – 11.59	0.041
neur factor [Leve/mod/charcot]	0.22	0.02 – 2.14	0.192
loca factor [Fal]	1.65	0.92 – 2.97	0.093
evol	0.99	0.99 – 1.00	0.017
Random Effects
σ²	3.29
τ₀₀ _NOMBRE	0.48
ICC	0.13
N _NOMBRE	15
Observations	299
Marginal R² / Conditional R²	0.370 / 0.451

#modelo 16
mod_mixto16c <- glmer(cicat ~ topo_factor + nume_factor + isq_factor + infe_factor+ ede_factor+ neur_factor + loca_factor+ evol+ tisu_factor+ (1 | NOMBRE),
                   family = binomial,
                   data = TES)

## Warning in checkConv(attr(opt, "derivs"), opt$par, ctrl = control$checkConv, :
## Model failed to converge with max|grad| = 0.00677638 (tol = 0.002, component 1)

tab_model(mod_mixto16c)

	cicat
Predictors	Odds Ratios	CI	p
(Intercept)	0.11	0.01 – 1.79	0.120
topo factor [dor/plan/lat]	0.78	0.38 – 1.62	0.506
nume factor [Una]	5.06	2.49 – 10.31	<0.001
isq factor [NO/leve]	3.74	1.95 – 7.17	<0.001
infe factor [Leve/Mod]	2.30	0.88 – 5.99	0.090
infe factor [No]	3.52	0.99 – 12.50	0.051
ede factor [Unilateral]	2.88	0.82 – 10.15	0.100
ede factor [Sin/local]	3.41	1.03 – 11.22	0.044
neur factor [Leve/mod/charcot]	0.26	0.03 – 2.55	0.247
loca factor [Fal]	1.67	0.93 – 3.00	0.088
evol	0.99	0.99 – 1.00	0.020
tisu factor [Piel]	1.86	0.70 – 4.99	0.215
Random Effects
σ²	3.29
τ₀₀ _NOMBRE	0.42
ICC	0.11
N _NOMBRE	15
Observations	299
Marginal R² / Conditional R²	0.379 / 0.450

Comparación del rendimiento de los modelos

AIC(mod_mixto,  mod_mixto2, mod_mixto3, mod_mixto4, mod_mixto5, mod_mixto6, mod_mixto7, mod_mixto8, mod_mixto9, mod_mixto10, mod_mixto11, mod_mixto12,mod_mixto13, mod_mixto14, mod_mixto15, mod_mixto16c,mod_mixto13ti)

##               df AIC
## mod_mixto      3 368
## mod_mixto2     4 351
## mod_mixto3     6 350
## mod_mixto4     7 349
## mod_mixto5     9 349
## mod_mixto6    10 350
## mod_mixto7    11 351
## mod_mixto8    11 351
## mod_mixto9    12 344
## mod_mixto10   12 348
## mod_mixto11   13 343
## mod_mixto12   10 342
## mod_mixto13   10 341
## mod_mixto14   12 342
## mod_mixto15   12 342
## mod_mixto16c  13 343
## mod_mixto13ti 11 341

anova(mod_mixto, mod_mixto2, mod_mixto3, mod_mixto4, mod_mixto5, mod_mixto6, mod_mixto7, mod_mixto8, mod_mixto9, mod_mixto10, mod_mixto11, mod_mixto12, mod_mixto13,mod_mixto14, mod_mixto16c, mod_mixto15,mod_mixto13ti, test="Chisq")

## Data: TES
## Models:
## mod_mixto: cicat ~ nume_factor + (1 | NOMBRE)
## mod_mixto2: cicat ~ nume_factor + isq_factor + (1 | NOMBRE)
## mod_mixto3: cicat ~ nume_factor + isq_factor + infe_factor + (1 | NOMBRE)
## mod_mixto4: cicat ~ nume_factor + isq_factor + tisu_factor + infe_factor + (1 | NOMBRE)
## mod_mixto5: cicat ~ nume_factor + isq_factor + tisu_factor + infe_factor + ede_factor + (1 | NOMBRE)
## mod_mixto6: cicat ~ topo_factor + nume_factor + isq_factor + tisu_factor + infe_factor + ede_factor + (1 | NOMBRE)
## mod_mixto12: cicat ~ nume_factor + isq_factor + infe_factor + ede_factor + neur_factor + evol + (1 | NOMBRE)
## mod_mixto13: cicat ~ isq_factor + infe_factor + loca_factor + nume_factor + ede_factor + evol + (1 | NOMBRE)
## mod_mixto7: cicat ~ topo_factor + nume_factor + isq_factor + tisu_factor + infe_factor + ede_factor + neur_factor + (1 | NOMBRE)
## mod_mixto8: cicat ~ topo_factor + nume_factor + isq_factor + tisu_factor + infe_factor + ede_factor + neur_factor + (1 | NOMBRE)
## mod_mixto13ti: cicat ~ isq_factor + infe_factor + loca_factor + nume_factor + ede_factor + evol + tisu_factor + (1 | NOMBRE)
## mod_mixto9: cicat ~ topo_factor + nume_factor + isq_factor + tisu_factor + infe_factor + ede_factor + neur_factor + evol + (1 | NOMBRE)
## mod_mixto10: cicat ~ topo_factor + nume_factor + isq_factor + tisu_factor + infe_factor + ede_factor + neur_factor + loca_factor + (1 | NOMBRE)
## mod_mixto14: cicat ~ topo_factor + nume_factor + isq_factor + infe_factor + ede_factor + neur_factor + loca_factor + evol + (1 | NOMBRE)
## mod_mixto15: cicat ~ topo_factor + nume_factor + isq_factor + infe_factor + ede_factor + neur_factor + loca_factor + evol + (1 | NOMBRE)
## mod_mixto11: cicat ~ topo_factor + nume_factor + isq_factor + tisu_factor + infe_factor + ede_factor + neur_factor + loca_factor + evol + (1 | NOMBRE)
## mod_mixto16c: cicat ~ topo_factor + nume_factor + isq_factor + infe_factor + ede_factor + neur_factor + loca_factor + evol + tisu_factor + (1 | NOMBRE)
##               npar AIC BIC logLik deviance Chisq Df Pr(>Chisq)    
## mod_mixto        3 368 379   -181      362                        
## mod_mixto2       4 351 366   -172      343 19.13  1   0.000012 ***
## mod_mixto3       6 350 372   -169      338  5.27  2      0.072 .  
## mod_mixto4       7 349 375   -167      335  3.25  1      0.071 .  
## mod_mixto5       9 349 383   -166      331  3.44  2      0.179    
## mod_mixto6      10 350 387   -165      330  0.82  1      0.365    
## mod_mixto12     10 342 379   -161      322  8.47  0               
## mod_mixto13     10 341 378   -160      321  0.98  0               
## mod_mixto7      11 351 391   -164      329  0.00  1      1.000    
## mod_mixto8      11 351 391   -164      329  0.00  0               
## mod_mixto13ti   11 341 382   -160      319  9.68  0               
## mod_mixto9      12 344 388   -160      320  0.00  1      1.000    
## mod_mixto10     12 348 393   -162      324  0.00  0               
## mod_mixto14     12 342 387   -159      318  5.76  0               
## mod_mixto15     12 342 387   -159      318  0.00  0               
## mod_mixto11     13 343 391   -158      317  1.58  1      0.209    
## mod_mixto16c    13 343 391   -158      317  0.00  0               
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

MODELO DE EFECTOS PRINCIPALES mod_mixto13.

ranef(mod_mixto13ti)$NOMBRE

##                                                                      (Intercept)
## ABUDARA                                                                    0.277
## Centro de rehabilitación, Obesidad y Diabetes. Formosa                     0.427
## CER                                                                        0.470
## Clínica Centro Médico                                                      0.433
## Hosp. Dr.  J.R.Vidal                                                      -0.266
## Hospital Centro de Salud Zenón J. Santillán                               -0.525
## Hospital de Clínicas                                                       0.095
## Hospital de Día de Pie Diabético. Polo Sanitario Malvinas Argentinas      -0.553
## Hospital Güemes                                                           -0.054
## Hospital Perrupato                                                         0.317
## Hospital Posadas                                                           0.322
## Hospital san Martin                                                       -0.380
## Hospital Sommer                                                           -0.436
## HRRG                                                                       0.238
## Provincial Rosario                                                        -0.458

efectos_centros <- ranef(mod_mixto13ti)$NOMBRE
efectos_centros_df <- data.frame(
  Centro = rownames(efectos_centros),
  Intercepto_aleatorio = efectos_centros[, "(Intercept)"]
)
head(efectos_centros_df)

##                                                   Centro Intercepto_aleatorio
## 1                                                ABUDARA                 0.28
## 2 Centro de rehabilitación, Obesidad y Diabetes. Formosa                 0.43
## 3                                                    CER                 0.47
## 4                                  Clínica Centro Médico                 0.43
## 5                                   Hosp. Dr.  J.R.Vidal                -0.27
## 6            Hospital Centro de Salud Zenón J. Santillán                -0.53

intercepto_fijo <- fixef(mod_mixto13ti)["(Intercept)"]

efectos_centros_df$Intercepto_total <- intercepto_fijo + efectos_centros_df$Intercepto_aleatorio
library(ggplot2)

ggplot(efectos_centros_df, aes(x = reorder(Centro, Intercepto_total), y = Intercepto_total)) +
  geom_point() +
  coord_flip() +
  labs(x = "Centro", y = "Intercepto total", title = "Variación del intercepto por centro") +
  theme_minimal()

efectos_centros_df$Probabilidad_basal <- plogis(efectos_centros_df$Intercepto_total)
# 1️⃣ Cargar paquete
library(lme4)
library(ggplot2)
library(dplyr)

# 2️⃣ Extraer efectos aleatorios del modelo
efectos_centros <- ranef(mod_mixto13ti, condVar = TRUE)$NOMBRE

# 3️⃣ Convertir a data.frame
efectos_centros_df <- data.frame(
  Centro = rownames(efectos_centros),
  Intercepto_centro = efectos_centros[ , "(Intercept)"]
)

# 4️⃣ Sumar el intercepto fijo global
intercepto_fijo <- fixef(mod_mixto13ti)[["(Intercept)"]]
efectos_centros_df <- efectos_centros_df %>%
  mutate(
    Intercepto_total = intercepto_fijo + Intercepto_centro,
    Probabilidad_basal = plogis(Intercepto_total) # convierte log-odds a probabilidad
  )

# 5️⃣ Graficar la probabilidad basal de cicatrización por centro
ggplot(efectos_centros_df, aes(x = reorder(Centro, Probabilidad_basal), y = Probabilidad_basal)) +
  geom_point(size = 3, color = "steelblue") +
  geom_hline(yintercept = mean(efectos_centros_df$Probabilidad_basal), 
             linetype = "dashed", color = "red") +
  coord_flip() +
  labs(
    x = "Centro",
    y = "Probabilidad basal de cicatrización (ajustada)",
    title = "Variación entre centros en la probabilidad basal de cicatrización"
  ) +
  theme_minimal(base_size = 13)

Claro, te lo interpreto paso por paso. Este gráfico muestra la variación entre centros en la probabilidad de cicatrización, estimada a partir del modelo mixto donde el centro es un efecto aleatorio.

🧩 ¿Qué está mostrando el gráfico?

Cada punto representa un centro Los centros están ordenados desde menor a mayor probabilidad de cicatrización predicha por el modelo mixto.CUANDO TODAS LAS VARIABLES SON CERO!!
El eje X es la probabilidad de cicatrización Valores alrededor de 0.015–0.040 (es decir, 1.5% a 4.0%).
La línea roja punteada Marca la probabilidad promedio o el intercepto predicho global del modelo mixto (el valor promedio de todos los centros).
Separación entre puntos = variabilidad entre centros Muestra cómo algunos centros tienen más probabilidad de cicatrizar (puntos a la derecha) y otros menos (puntos a la izquierda).

📌 Interpretación general

👉 Hay heterogeneidad real entre centros. Algunos centros tienen una probabilidad de cicatrización claramente menor que el promedio (izquierda de la línea roja), mientras que otros la tienen mayor (derecha de la línea roja).

Esto significa que:

El centro influye en la probabilidad de cicatrizar (hay variación institucional).
El modelo mixto detecta diferencias que no dependen solo de las características del paciente sino que provienen de diferencias entre centros (procesos, equipos, recursos, protocolos, etc.).

🔍 Interpretación más detallada

Centros con peor desempeño (probabilidad más baja)

Ejemplo: los que están cerca de 0.015–0.020 → Probabilidad de cicatrización más baja que la media → posible necesidad de revisar procesos.

Centros cercanos al promedio

Son los que están cerca de la línea roja. → No difieren sustancialmente del promedio nacional/global del modelo.

Centros con mejor desempeño (probabilidad más alta)

Cerca de 0.035–0.040 → Probabilidad de cicatrización claramente superior → podrían tener mejores prácticas o recursos.

📉 ¿Esto representa el ICC (variación entre centros)?

Indirectamente sí:

La dispersión de los puntos refleja la varianza del intercepto aleatorio del modelo.
Centros muy separados indican alta variabilidad entre centros, es decir un ICC mayor.

🧠 Conclusión clínica/metodológica

Este gráfico demuestra que:

El centro de atención importa para la probabilidad de cicatrización.
Hay centros con mejor o peor desempeño incluso después de ajustar por las covariables incluidas en tu modelo.

Podemos profundizar si querés:

Interpretar el ICC
Comparar este gráfico con un caterpillar plot clásico
Ver cómo mejorar la visualización
Revisar el modelo mixto que usaste

¿Querés que analicemos qué tan grande es la variabilidad entre centros o cómo reportarlo en tu artículo?

#vamos a hacer las probabilidades por centro
pred <- predict(mod_mixto13ti, type = "response", re.form = ~(1|NOMBRE))
prob_centro <- aggregate(pred ~ NOMBRE, data = TES, mean)
prob_centro

##                                                                  NOMBRE pred
## 1                                                               ABUDARA 0.62
## 2                Centro de rehabilitación, Obesidad y Diabetes. Formosa 0.68
## 3                                                                   CER 0.60
## 4                                                 Clínica Centro Médico 0.41
## 5                                                  Hosp. Dr.  J.R.Vidal 0.57
## 6                           Hospital Centro de Salud Zenón J. Santillán 0.32
## 7                                                  Hospital de Clínicas 0.73
## 8  Hospital de Día de Pie Diabético. Polo Sanitario Malvinas Argentinas 0.55
## 9                                                       Hospital Güemes 0.74
## 10                                                   Hospital Perrupato 0.80
## 11                                                     Hospital Posadas 0.57
## 12                                                  Hospital san Martin 0.51
## 13                                                      Hospital Sommer 0.61
## 14                                                                 HRRG 0.63
## 15                                                   Provincial Rosario 0.54

#Residuos vs predichos
library(DHARMa)

sim_res <- simulateResiduals(fittedModel = mod_mixto13ti, n = 1000)
plot(sim_res)

# Test formales
testUniformity(sim_res)

## 
##  Asymptotic one-sample Kolmogorov-Smirnov test
## 
## data:  simulationOutput$scaledResiduals
## D = 0.05, p-value = 0.4
## alternative hypothesis: two-sided

testDispersion(sim_res)

## 
##  DHARMa nonparametric dispersion test via sd of residuals fitted vs.
##  simulated
## 
## data:  simulationOutput
## dispersion = 1, p-value = 0.6
## alternative hypothesis: two.sided

testZeroInflation(sim_res)

## 
##  DHARMa zero-inflation test via comparison to expected zeros with
##  simulation under H0 = fitted model
## 
## data:  simulationOutput
## ratioObsSim = 1, p-value = 1
## alternative hypothesis: two.sided

library(performance)

# Colinealidad
check_collinearity(mod_mixto13ti)

## # Check for Multicollinearity
## 
## Low Correlation
## 
##         Term  VIF   VIF 95% CI adj. VIF Tolerance Tolerance 95% CI
##   isq_factor 1.10 [1.03, 1.37]     1.05      0.91     [0.73, 0.97]
##  infe_factor 1.44 [1.28, 1.70]     1.20      0.69     [0.59, 0.78]
##  loca_factor 1.07 [1.01, 1.42]     1.03      0.94     [0.70, 0.99]
##  nume_factor 1.09 [1.02, 1.37]     1.04      0.92     [0.73, 0.98]
##   ede_factor 1.26 [1.14, 1.49]     1.12      0.80     [0.67, 0.88]
##         evol 1.06 [1.01, 1.47]     1.03      0.95     [0.68, 0.99]
##  tisu_factor 1.26 [1.14, 1.49]     1.12      0.79     [0.67, 0.88]

# Bondad de ajuste (R² marginal y condicional)
r2_nakagawa(mod_mixto13ti)

## # R2 for Mixed Models
## 
##   Conditional R2: 0.429
##      Marginal R2: 0.370

# Varianza de efectos aleatorios
VarCorr(mod_mixto13ti)

##  Groups Name        Std.Dev.
##  NOMBRE (Intercept) 0.581

ICC

performance::icc(mod_mixto13ti)

## # Intraclass Correlation Coefficient
## 
##     Adjusted ICC: 0.093
##   Unadjusted ICC: 0.059

Calibracion

#HyL
library(ResourceSelection)

## Warning: package 'ResourceSelection' was built under R version 4.4.3

## ResourceSelection 0.3-6   2023-06-27

pred <- predict(mod_mixto13ti, type = "response")
hoslem.test(TES$cicat, pred, g = 10)

## Warning in Ops.factor(1, y): '-' no es significativo para factores

## 
##  Hosmer and Lemeshow goodness of fit (GOF) test
## 
## data:  TES$cicat, pred
## X-squared = NA, df = 8, p-value = NA

#Brier
library(DescTools)

## 
## Adjuntando el paquete: 'DescTools'

## The following objects are masked from 'package:psych':
## 
##     AUC, ICC, SD

## The following object is masked from 'package:data.table':
## 
##     %like%

pred <- predict(mod_mixto13ti, type = "response")
BrierScore <- BrierScore(TES$cicat, pred)

## Warning in Ops.factor(resp, (1 - pred)^2): '*' no es significativo para
## factores

## Warning in Ops.factor(1, resp): '-' no es significativo para factores

BrierScore

## [1] NA

Calibración:NO SE QUE ES ESTO EN MODELO MIXTO

TES$lp <- predict(mod_mixto13ti, type = "link")  # predicción en escala logit

mod_slope <- glm(cicat ~ lp, family = binomial, data = TES)
summary(mod_slope)

## 
## Call:
## glm(formula = cicat ~ lp, family = binomial, data = TES)
## 
## Coefficients:
##             Estimate Std. Error z value            Pr(>|z|)    
## (Intercept)  -0.0332     0.1486   -0.22                0.82    
## lp            1.0811     0.1334    8.10 0.00000000000000053 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## (Dispersion parameter for binomial family taken to be 1)
## 
##     Null deviance: 405.76  on 298  degrees of freedom
## Residual deviance: 302.75  on 297  degrees of freedom
## AIC: 306.8
## 
## Number of Fisher Scoring iterations: 4

auc

library(pROC)

## Warning: package 'pROC' was built under R version 4.4.3

## Type 'citation("pROC")' for a citation.

## 
## Adjuntando el paquete: 'pROC'

## The following objects are masked from 'package:stats':
## 
##     cov, smooth, var

pred <- predict(mod_mixto13ti, type = "response")
roc_obj <- roc(TES$cicat, pred)

## Setting levels: control = 0, case = 1

## Setting direction: controls < cases

auc(roc_obj)

## Area under the curve: 0.82

plot(roc_obj, print.auc = TRUE, col = "blue", main = "Curva ROC - Modelo mixto")

#modificación de efecto tabaquismo activo
TES <- TES %>%
  mutate(tbq_factor = factor(if_else(tbq %in% c(1),  "si", "no")))

table(TES$tbq_factor)

## 
##  no  si 
## 236  63

mod_mixto13timod <- glmer(cicat ~ isq_factor + infe_factor + loca_factor +  nume_factor+ ede_factor + evol + tisu_factor+ isq_factor:tbq_factor+ (1 | NOMBRE),
                   family = binomial,
                   data = TES)

## Warning in checkConv(attr(opt, "derivs"), opt$par, ctrl = control$checkConv, :
## Model failed to converge with max|grad| = 0.0241635 (tol = 0.002, component 1)

tab_model(mod_mixto13timod)

	cicat
Predictors	Odds Ratios	CI	p
(Intercept)	0.02	0.00 – 0.13	<0.001
isq factor [NO/leve]	4.42	2.12 – 9.22	<0.001
infe factor [Leve/Mod]	2.26	0.87 – 5.87	0.093
infe factor [No]	3.23	0.93 – 11.19	0.064
loca factor [Fal]	1.70	0.95 – 3.02	0.073
nume factor [Una]	4.77	2.50 – 9.12	<0.001
ede factor [Unilateral]	3.12	0.87 – 11.18	0.080
ede factor [Sin/local]	3.45	1.02 – 11.64	0.046
evol	0.99	0.99 – 1.00	0.015
tisu factor [Piel]	1.86	0.70 – 4.91	0.213
isq factor [SI] × tbq factorsi	1.34	0.41 – 4.37	0.630
isq factor [NO/leve] × tbq factorsi	0.57	0.25 – 1.28	0.174
Random Effects
σ²	3.29
τ₀₀ _NOMBRE	0.32
ICC	0.09
N _NOMBRE	15
Observations	299
Marginal R² / Conditional R²	0.379 / 0.434

AIC(mod_mixto13timod)

## [1] 343

#modificacion efecto tabaquismo si o no 
TES <- TES %>%
  mutate(tbq_factor1 = factor(if_else(tbq %in% c(1,2),  "si", "no")))

table(TES$tbq_factor1)

## 
##  no  si 
##  89 210

mod_mixto13timod2 <- glmer(cicat ~ isq_factor + infe_factor + loca_factor +  nume_factor+ ede_factor + evol + tisu_factor+ isq_factor:tbq_factor1+ (1 | NOMBRE),
                   family = binomial,
                   data = TES)

## Warning in checkConv(attr(opt, "derivs"), opt$par, ctrl = control$checkConv, :
## Model failed to converge with max|grad| = 0.0074089 (tol = 0.002, component 1)

tab_model(mod_mixto13timod2)

	cicat
Predictors	Odds Ratios	CI	p
(Intercept)	0.03	0.01 – 0.20	<0.001
isq factor [NO/leve]	3.05	0.99 – 9.41	0.052
infe factor [Leve/Mod]	2.28	0.89 – 5.86	0.087
infe factor [No]	3.24	0.94 – 11.12	0.062
loca factor [Fal]	1.71	0.96 – 3.04	0.068
nume factor [Una]	4.59	2.41 – 8.76	<0.001
ede factor [Unilateral]	3.19	0.90 – 11.33	0.072
ede factor [Sin/local]	3.57	1.07 – 11.92	0.038
evol	0.99	0.99 – 1.00	0.019
tisu factor [Piel]	1.99	0.75 – 5.26	0.166
isq factor [SI] × tbq factor1si	0.66	0.22 – 2.03	0.474
isq factor [NO/leve] × tbq factor1si	0.85	0.39 – 1.89	0.698
Random Effects
σ²	3.29
τ₀₀ _NOMBRE	0.29
ICC	0.08
N _NOMBRE	15
Observations	299
Marginal R² / Conditional R²	0.373 / 0.423

AIC(mod_mixto13timod2)

## [1] 344

DICOTOMIZO INFECCIÓN SI O NO A VER SI HAY ALGUN CAMBIO, PERO NO ES SIGNIFICATIVO TAMPODO

#que pasa si dicotomizo infección?Tampoco modifica nada. No es significativo
TES <- TES %>%
  mutate(infe_factor1 = factor(if_else(infe %in% c(1,2,3),  "si", "no")))
TES$infe_factor1 <- factor (TES$infe_factor1, levels= c ("si","no"))
table(TES$infe_factor1)

## 
##  si  no 
## 232  67

mod_mixto13ti1 <- glmer(cicat ~ isq_factor + infe_factor1 + loca_factor +  nume_factor+ ede_factor + evol + tisu_factor+ (1 | NOMBRE),
                   family = binomial,
                   data = TES)
tab_model(mod_mixto13ti1)

	cicat
Predictors	Odds Ratios	CI	p
(Intercept)	0.06	0.01 – 0.23	<0.001
isq factor [NO/leve]	3.38	1.81 – 6.31	<0.001
infe factor1 [no]	1.48	0.66 – 3.31	0.338
loca factor [Fal]	1.63	0.92 – 2.86	0.092
nume factor [Una]	4.83	2.55 – 9.13	<0.001
ede factor [Unilateral]	2.64	0.77 – 9.10	0.123
ede factor [Sin/local]	3.35	1.02 – 11.01	0.046
evol	0.99	0.99 – 1.00	0.017
tisu factor [Piel]	2.16	0.83 – 5.59	0.114
Random Effects
σ²	3.29
τ₀₀ _NOMBRE	0.23
ICC	0.07
N _NOMBRE	15
Observations	299
Marginal R² / Conditional R²	0.360 / 0.402

AIC(mod_mixto13ti1)

## [1] 342

#OJO AGREGANDO DIALSIS BAJA LA DEVIANCE Y EL AIC
mod_mixto13ti2 <- glmer(cicat ~ isq_factor + infe_factor + loca_factor +  nume_factor+ ede_factor + evol + tisu_factor+dial+ (1 | NOMBRE),
                   family = binomial,
                   data = TES)

## Warning in checkConv(attr(opt, "derivs"), opt$par, ctrl = control$checkConv, :
## Model failed to converge with max|grad| = 0.00791649 (tol = 0.002, component 1)

tab_model(mod_mixto13ti2)

	cicat
Predictors	Odds Ratios	CI	p
(Intercept)	0.03	0.00 – 0.15	<0.001
isq factor [NO/leve]	3.59	1.88 – 6.84	<0.001
infe factor [Leve/Mod]	2.25	0.87 – 5.81	0.094
infe factor [No]	3.33	0.96 – 11.49	0.057
loca factor [Fal]	1.69	0.95 – 3.01	0.074
nume factor [Una]	4.71	2.47 – 8.99	<0.001
ede factor [Unilateral]	2.99	0.85 – 10.54	0.088
ede factor [Sin/local]	3.39	1.02 – 11.21	0.046
evol	0.99	0.99 – 1.00	0.018
tisu factor [Piel]	1.95	0.74 – 5.15	0.177
dial [1]	0.74	0.26 – 2.10	0.575
Random Effects
σ²	3.29
τ₀₀ _NOMBRE	0.36
ICC	0.10
N _NOMBRE	15
Observations	299
Marginal R² / Conditional R²	0.372 / 0.434

AIC(mod_mixto13ti2)

## [1] 343

AIC(mod_mixto13ti)

## [1] 341

deviance(mod_mixto13ti2)

## [1] 302

deviance(mod_mixto13ti)

## [1] 303

#VOY A AGREGAR AAMAY
#OJO AGREGANDO DIALSIS BAJA LA DEVIANCE Y EL AIC
mod_mixto13ti3 <- glmer(cicat ~ isq_factor + infe_factor + loca_factor +  nume_factor+ ede_factor + evol + tisu_factor+dial+aamay+ (1 | NOMBRE),
                   family = binomial,
                   data = TES)

## Warning in checkConv(attr(opt, "derivs"), opt$par, ctrl = control$checkConv, :
## Model failed to converge with max|grad| = 0.00444906 (tol = 0.002, component 1)

tab_model(mod_mixto13ti3)

	cicat
Predictors	Odds Ratios	CI	p
(Intercept)	0.03	0.00 – 0.15	<0.001
isq factor [NO/leve]	3.32	1.73 – 6.36	<0.001
infe factor [Leve/Mod]	2.33	0.90 – 6.06	0.082
infe factor [No]	3.73	1.06 – 13.13	0.040
loca factor [Fal]	1.71	0.96 – 3.05	0.071
nume factor [Una]	4.54	2.38 – 8.67	<0.001
ede factor [Unilateral]	3.33	0.94 – 11.79	0.063
ede factor [Sin/local]	3.67	1.11 – 12.19	0.033
evol	0.99	0.99 – 1.00	0.012
tisu factor [Piel]	2.07	0.77 – 5.58	0.150
dial [1]	0.79	0.27 – 2.28	0.658
aamay [1]	0.32	0.10 – 1.07	0.065
Random Effects
σ²	3.29
τ₀₀ _NOMBRE	0.37
ICC	0.10
N _NOMBRE	15
Observations	299
Marginal R² / Conditional R²	0.386 / 0.447

AIC(mod_mixto13ti3)

## [1] 341

AIC(mod_mixto13ti2)

## [1] 343

AIC(mod_mixto13ti)

## [1] 341

deviance(mod_mixto13ti2)

## [1] 302

deviance(mod_mixto13ti)

## [1] 303

deviance(mod_mixto13ti3)

## [1] 299

library(pROC)

pred <- predict(mod_mixto13ti3, type = "response")
roc_obj <- roc(TES$cicat, pred)

## Setting levels: control = 0, case = 1

## Setting direction: controls < cases

auc(roc_obj)

## Area under the curve: 0.83

plot(roc_obj, print.auc = TRUE, col = "blue", main = "Curva ROC - Modelo mixto")

Tesis para Diego

2025-11-15