GATclaselatente

Analisis asociaciones - perfiles GAT

Tranformacion variables a dicotomicas

library(readxl)
GAT <- read_excel("GAT.xlsx") # cargue base de datos

library(dplyr)
GAT <- GAT %>% filter(EstadoCancer != "No sabe") #se excluyen los registros que no saben su estadio de cancer

GAT$Nestadocancer <- ifelse(GAT$EstadoCancer %in% c("Con metástasis a otros órganos"), "Metastasico", 
                     ifelse(GAT$EstadoCancer %in% c("Localizado", "Con ganglios comprometidos, pero sin metástasis a otros órganos"), "No metastasico", GAT$EstadoCancer)) # categoriza metastasico y no metastasico

GAT <- GAT %>% 
  mutate(SQUIM = ifelse(AQuimioTerSiste == "Checked" | BQuimioTerSiste == "Checked" & ARadio != "Checked" & BRadio != "Checked", "Si", "No")) %>% 
  mutate(SRAD = ifelse(ARadio == "Checked" | BRadio == "Checked"  & AQuimioTerSiste != "Checked" & BQuimioTerSiste != "Checked", "Si", "No")) #solo quimio, #solo radio


GAT <- GAT %>% mutate(Edadcat = cut(Edad, breaks = c(-Inf, 66, Inf), right = F, labels = c("< 65", ">=65"))) #edad


GAT$Nocup <- ifelse(GAT$Ocupacion %in% c("Cesante", "Hogar", "Pensionado"), "-Ocup",
                    ifelse(GAT$Ocupacion %in% c("Empleado", "Independiente", "Estudiante"), "+Ocup", GAT$Ocupacion)) #ocupacion


GAT$Nrel <- ifelse(GAT$Religion %in% c("Católica"), "Católica",
            ifelse(GAT$Religion %in% c("Cristiana", "Judía", "No profesa religión", "Otra"), "Otra", GAT$Religion)) #religion


GAT$CIE <- ifelse(substr(GAT$TipoCancer, 1, 1) == "D", "D", substr(GAT$TipoCancer, 1, 3))
GAT$Tipotumor <- ifelse(GAT$CIE == "D", "hematologico", "tumor solido") #tipo de tumor


#IDENTIFICANDO EL TIPO DE MAC 

GAT <- GAT %>% 
  mutate(cuerpo = ifelse(Masaje == "Si" | Reflexología == "Si" | Quiropraxia == "Si", 1, 0)) %>%
  mutate(cuerpo = tidyr::replace_na(cuerpo, 0)) #Cuerpo


GAT <- GAT %>% 
  mutate(energetica = ifelse(Bioenergética == "Si" | `Terapia electromagnética` == "Si" | Acupuntura == "Si" | TerapiaNeural == "Si", 1, 0)) %>%
  mutate(energetica = tidyr::replace_na(energetica, 0)) #energetica


GAT <- GAT %>% 
  mutate(mente = ifelse(Meditacion == "Checked" | Yoga == "Checked" | Taichi == "Checked" | Yorae == "Checked" |OtroMente == "Checked", 1, 0)) %>%
  mutate(mente = tidyr::replace_na(mente, 0)) #mente


GAT <- GAT %>% 
  mutate(Hierbas = tidyr::replace_na(Hierbas, "No")) %>% 
  mutate(MAC_Animal = tidyr::replace_na(OrigenAnimal, "No")) %>% 
  mutate(Nutricionales = ifelse(Dieta_def == "Si" | Alimentos == "Si", "Si", "No")) %>% 
  mutate(Nutricionales = tidyr::replace_na(Nutricionales, "No")) %>% 
  mutate(whole = ifelse(Homeopatia == "Si" | Tradicionales == "Si" | cuerpo == 1 | energetica == 1 | mente == 1, "Si", "No")) %>% 
  mutate(whole = tidyr::replace_na(whole, "No")) %>% 
  mutate(Vitaminas = tidyr::replace_na(Vitaminas, "No"))

## Intencion de uso
GAT <- GAT %>% 
  mutate(OTRA = ifelse(Recomendamedico == "Checked" | Recomendaotro == "Checked" | todaslasOpciones == "Checked" | Masnatural == "Checked" | NingunaRazon == "Checked" | OtraSiUsa == "Checked", 2, 1)) %>% 
  mutate(CURATIVA = ifelse(Curar == "Checked" | futurascompli == "Checked", 2, 1)) %>% 
  mutate(PALIATIVA = ifelse(ContrarestEfectos == "Checked" | Animo == "Checked",2, 1)) %>% 
  mutate(AMBAS = ifelse(CURATIVA == 2 & PALIATIVA == 2 & OTRA == 1, 2, 1))


## Fuete de informacion
GAT <- GAT %>% 
  mutate(INFOnm = ifelse(FamiliaAmigos == "Checked" | Pacientes == "Checked" | Medios == "Checked" | RedesSociales== "Checked" | Mercados == "Checked" | Farmaceuta == "Checked" | OtroEntero == "Checked", 2, 1))

GAT <- GAT %>% 
  mutate(INFOm = ifelse(Medico == "Checked" | MedicoAlterna == "Checked", 2,1))

#EFECTOS
GAT<- GAT %>% 
  mutate(EFECTS = ifelse(vomito == "Checked" | diarrea == "Checked" | Estrenimiento == "Checked" | Dolor == "Checked" | PerdidaApetito == "Checked" | Adormecimiento == "Checked" | OtroEfectoSec == "Checked", 2,1))


#afiliacion
GAT$Nafiliacion <- ifelse(GAT$Afiliacion %in% c("Subsidiado", "No asegurado"), "Subsidiado-No asegurado",
              ifelse(GAT$Afiliacion %in% c("Contributivo", "Plan complementario", "Otro"), "Otro",GAT$Afiliacion))


GAT <- GAT %>%
  mutate(nestadocivil= recode(EstadoCivil,
                                 "Soltero" = "Sin conyuge",
                                 "Divorciado" = "Sin conyuge",
                                 "Viudo" = "Sin conyuge",
                                 "Casado" = "Con conyuge",
                                 "Unión libre" = "Con conyuge"))


GAT$Estrato <- as.character(GAT$Estrato)#socioeconomic status
GAT$Estrato[GAT$Estrato== 0] <- "Bajo"
GAT$Estrato[GAT$Estrato== 1] <- "Bajo"
GAT$Estrato[GAT$Estrato== 2] <- "Bajo"
GAT$Estrato[GAT$Estrato== 3] <- "Alto"
GAT$Estrato[GAT$Estrato== 4] <- "Alto"
GAT$Estrato[GAT$Estrato== 5] <- "Alto"
GAT$Estrato[GAT$Estrato== 6] <- "Alto"
GAT$Estrato <- as.factor(GAT$Estrato)

GAT <- GAT %>%
  mutate(nestadocivil= recode(EstadoCivil,
                              "Soltero" = "Sin conyuge",
                              "Divorciado" = "Sin conyuge",
                              "Viudo" = "Sin conyuge",
                              "Casado" = "Con conyuge",
                              "Unión libre" = "Con conyuge"))

GAT$ResidenciaTipo <- as.factor(GAT$ResidenciaTipo)

Analisis bivariado, variable respuesta: USO MAC

MAC

	level	Overall	No	Si	p	test
n		2830	1365	1465
Sexo….	Femenino	1786 (63.1)	758 (55.5)	1028 (70.2)	<0.001
X	Masculino	1044 (36.9)	607 (44.5)	437 (29.8)
Nestadocancer….	Metastasico	826 (29.2)	401 (29.4)	425 (29.0)	0.862
X.1	No metastasico	2004 (70.8)	964 (70.6)	1040 (71.0)
SQUIM….	No	192 ( 6.8)	101 ( 7.4)	91 ( 6.2)	0.238
X.2	Si	2638 (93.2)	1264 (92.6)	1374 (93.8)
SRAD….	No	1846 (65.2)	897 (65.7)	949 (64.8)	0.629
X.3	Si	984 (34.8)	468 (34.3)	516 (35.2)
Edadcat….	< 65	1844 (65.2)	796 (58.3)	1048 (71.5)	<0.001
X.4	>=65	986 (34.8)	569 (41.7)	417 (28.5)
Nocup….	-Ocup	1945 (68.7)	974 (71.4)	971 (66.3)	0.004
X.5	+Ocup	885 (31.3)	391 (28.6)	494 (33.7)
Nrel….	Católica	2214 (78.2)	1104 (80.9)	1110 (75.8)	0.001
X.6	Otra	616 (21.8)	261 (19.1)	355 (24.2)
Tipotumor….	hematologico	80 ( 2.8)	36 ( 2.6)	44 ( 3.0)	0.636
X.7	tumor solido	2750 (97.2)	1329 (97.4)	1421 (97.0)
Nafiliacion….	Otro	1700 (60.1)	841 (61.6)	859 (58.6)	0.115
X.8	Subsidiado-No asegurado	1130 (39.9)	524 (38.4)	606 (41.4)
ResidenciaTipo….	Rural	642 (22.7)	303 (22.2)	339 (23.1)	0.580
X.9	Urbano	2188 (77.3)	1062 (77.8)	1126 (76.9)
Estrato….	Alto	981 (34.7)	459 (33.6)	522 (35.6)	0.280
X.10	Bajo	1849 (65.3)	906 (66.4)	943 (64.4)

Modelo multivariado

Se incluyen variables con valores p menos o iguales a 0.2

1. Sexo
2. Edad categorizada
3. Ocupación
4. Religión
5. Afiliacion

#modelo completo 

GAT$Sexo <- relevel(factor(GAT$Sexo), ref = "Masculino")
GAT$Edadcat <- relevel(factor(GAT$Edadcat), ref = ">=65")

library(dplyr)
GAT<- GAT %>%
  mutate(MAC = factor(ifelse(AlternativaActual == "Si", 1, 0)))


m1 <- glm (MAC ~ Sexo + Edadcat + Nrel + Nocup + Nafiliacion
             ,family = "binomial", data= GAT) #construido en orden de la significancia
summary(m1)

## 
## Call:
## glm(formula = MAC ~ Sexo + Edadcat + Nrel + Nocup + Nafiliacion, 
##     family = "binomial", data = GAT)
## 
## Coefficients:
##                                    Estimate Std. Error z value Pr(>|z|)    
## (Intercept)                        -0.68533    0.08615  -7.955 1.80e-15 ***
## SexoFemenino                        0.57627    0.08367   6.888 5.68e-12 ***
## Edadcat< 65                         0.37718    0.08731   4.320 1.56e-05 ***
## NrelOtra                            0.19404    0.09428   2.058  0.03957 *  
## Nocup+Ocup                          0.23321    0.08968   2.600  0.00931 ** 
## NafiliacionSubsidiado-No asegurado  0.08037    0.07965   1.009  0.31297    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## (Dispersion parameter for binomial family taken to be 1)
## 
##     Null deviance: 3919.7  on 2829  degrees of freedom
## Residual deviance: 3809.3  on 2824  degrees of freedom
## AIC: 3821.3
## 
## Number of Fisher Scoring iterations: 4

Modelo reducido

m2 <- glm (MAC ~ Sexo + Edadcat + Nrel + Nocup
             ,family = "binomial", data= GAT) #construido en orden de la significancia
summary(m2)

## 
## Call:
## glm(formula = MAC ~ Sexo + Edadcat + Nrel + Nocup, family = "binomial", 
##     data = GAT)
## 
## Coefficients:
##              Estimate Std. Error z value Pr(>|z|)    
## (Intercept)  -0.65637    0.08111  -8.092 5.87e-16 ***
## SexoFemenino  0.57703    0.08365   6.898 5.26e-12 ***
## Edadcat< 65   0.38600    0.08686   4.444 8.84e-06 ***
## NrelOtra      0.20014    0.09408   2.127   0.0334 *  
## Nocup+Ocup    0.21942    0.08865   2.475   0.0133 *  
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## (Dispersion parameter for binomial family taken to be 1)
## 
##     Null deviance: 3919.7  on 2829  degrees of freedom
## Residual deviance: 3810.3  on 2825  degrees of freedom
## AIC: 3820.3
## 
## Number of Fisher Scoring iterations: 4

ORs

library(gtsummary)
tbl_regression(m2, exponentiate = TRUE, add_estimate_to_reference_rows=TRUE)  %>% add_global_p()

Characteristic	OR1	95% CI1	p-value
Sexo			<0.001
Masculino	1.00	—
Femenino	1.78	1.51, 2.10
Edadcat			<0.001
>=65	1.00	—
< 65	1.47	1.24, 1.74
Nrel			0.033
Católica	1.00	—
Otra	1.22	1.02, 1.47
Nocup			0.013
-Ocup	1.00	—
+Ocup	1.25	1.05, 1.48
1 OR = Odds Ratio, CI = Confidence Interval

Comparando los dos modelos

library(lmtest)
lrtest(m1, m2)

## Likelihood ratio test
## 
## Model 1: MAC ~ Sexo + Edadcat + Nrel + Nocup + Nafiliacion
## Model 2: MAC ~ Sexo + Edadcat + Nrel + Nocup
##   #Df  LogLik Df  Chisq Pr(>Chisq)
## 1   6 -1904.6                     
## 2   5 -1905.1 -1 1.0182     0.3129

El modelo más simple es suficiente para explicar la variabilidad de los datos

library(performance)
model_performance(m2)

## # Indices of model performance
## 
## AIC      |     AICc |      BIC | Tjur's R2 |  RMSE | Sigma | Log_loss | Score_log |   PCP
## -----------------------------------------------------------------------------------------
## 3820.274 | 3820.295 | 3850.014 |     0.038 | 0.490 | 1.000 |    0.673 |      -Inf | 0.520

Prueba bondad de ajuste Hosmer Lemeshow

library(ResourceSelection)
hoslem.test(m2$y, fitted(m2), g=10)

## 
##  Hosmer and Lemeshow goodness of fit (GOF) test
## 
## data:  m2$y, fitted(m2)
## X-squared = 1.5952, df = 6, p-value = 0.9529

Buen ajuste a los datos

Analisis de clase latente

# filtrar uso MAC  y seleccion variables analisis

library(LCAvarsel)

LCA_data <- GAT %>%
  filter(AlternativaActual == "Si") %>%
  dplyr::select(ID,CURATIVA,PALIATIVA,OTRA)

#Se revisa que variables pueden ser las mas informativas

#library(poLCA)
#LCA_data2 <- LCA_data[,-1]
#LCA_data2 <- as.data.frame(LCA_data2)
#LCAvarsel(LCA_data2, verbose = FALSE)

set.seed(2025)
f <- cbind(PALIATIVA, CURATIVA, OTRA)~ 1
lCA4 <- poLCA(f, data = LCA_data, nclass=3,nrep= 15, graphs = F)

## Model 1: llik = -2655.705 ... best llik = -2655.705
## Model 2: llik = -2655.705 ... best llik = -2655.705
## Model 3: llik = -2655.705 ... best llik = -2655.705
## Model 4: llik = -2655.705 ... best llik = -2655.705
## Model 5: llik = -2655.705 ... best llik = -2655.705
## Model 6: llik = -2655.705 ... best llik = -2655.705
## Model 7: llik = -2655.705 ... best llik = -2655.705
## Model 8: llik = -2655.705 ... best llik = -2655.705
## Model 9: llik = -2655.705 ... best llik = -2655.705
## Model 10: llik = -2655.705 ... best llik = -2655.705
## Model 11: llik = -2655.705 ... best llik = -2655.705
## Model 12: llik = -2655.705 ... best llik = -2655.705
## Model 13: llik = -2655.705 ... best llik = -2655.705
## Model 14: llik = -2655.705 ... best llik = -2655.705
## Model 15: llik = -2655.705 ... best llik = -2655.705
## Conditional item response (column) probabilities,
##  by outcome variable, for each class (row) 
##  
## $PALIATIVA
##            Pr(1)  Pr(2)
## class 1:  0.0000 1.0000
## class 2:  0.4618 0.5382
## class 3:  0.7924 0.2076
## 
## $CURATIVA
##            Pr(1)  Pr(2)
## class 1:  0.9348 0.0652
## class 2:  0.0072 0.9928
## class 3:  0.8525 0.1475
## 
## $OTRA
##            Pr(1)  Pr(2)
## class 1:  0.8530 0.1470
## class 2:  0.6876 0.3124
## class 3:  0.0000 1.0000
## 
## Estimated class population shares 
##  0.3602 0.2967 0.3431 
##  
## Predicted class memberships (by modal posterior prob.) 
##  0.2881 0.3693 0.3427 
##  
## ========================================================= 
## Fit for 3 latent classes: 
## ========================================================= 
## number of observations: 1465 
## number of estimated parameters: 11 
## residual degrees of freedom: -4 
## maximum log-likelihood: -2655.705 
##  
## AIC(3): 5333.41
## BIC(3): 5391.596
## G^2(3): 9.405014e-10 (Likelihood ratio/deviance statistic) 
## X^2(3): 9.399236e-10 (Chi-square goodness of fit) 
##  
## ALERT: negative degrees of freedom; respecify model 
## 

prop.table(table(lCA4$predclass))*100

## 
##        1        2        3 
## 28.80546 36.92833 34.26621

Se unen los data frame.

LCA_data$class <-  lCA4$predclass
GAT_MAC <- GAT %>%
  filter(AlternativaActual == "Si")
GAT_MAC <- merge(GAT_MAC, LCA_data[, c("ID", "class")], by = "ID", all.x = TRUE)

GAT_MAC$class <- factor(GAT_MAC$class, 
                     levels = c(1, 2, 3),
                     labels = c("R.Paliativa", "R.Curativa", "R.Otra"))
GAT_MAC$INFOm <- factor(GAT_MAC$INFOm, 
                     levels = c(1, 2),
                     labels = c("No", "Si"))

GAT_MAC$INFOnm <- factor(GAT_MAC$INFOnm, 
                     levels = c(1, 2),
                     labels = c("No", "Si"))

catVars2 <- c("Sexo","Nestadocancer", "SQUIM", "SRAD", "Edadcat", "Nocup", "Nrel", "Tipotumor", "Nafiliacion", "ResidenciaTipo", "Estrato", "INFOnm", "INFOm", "EFECTS", "MAC_Animal", "whole", "Hierbas", "Nutricionales", "Vitaminas")


tab2 <- CreateCatTable(vars = catVars2, strata = "class", data = GAT_MAC, includeNA = F, test = T, addOverall = T )

table2 <- as.data.frame(print(tab2, showAllLevels= TRUE, printToggle = FALSE))
library(openxlsx)
write.xlsx(table2, "tablaclase.xlsx", rowNames = TRUE)

kable(table2, format = "html", caption = "MAC") %>%
  kable_styling(bootstrap_options = c("striped", "hover", "condensed"), 
                full_width = F, 
                position = "center") %>%
  column_spec(1, bold = T, color = "white", background = "#D7261E") %>%
  column_spec(2, border_left = T, background = "#F3E6E3")

MAC

	level	Overall	R.Paliativa	R.Curativa	R.Otra	p	test
n		1465	422	541	502
Sexo….	Masculino	437 (29.8)	134 (31.8)	153 (28.3)	150 (29.9)	0.505
X	Femenino	1028 (70.2)	288 (68.2)	388 (71.7)	352 (70.1)
Nestadocancer….	Metastasico	425 (29.0)	120 (28.4)	163 (30.1)	142 (28.3)	0.769
X.1	No metastasico	1040 (71.0)	302 (71.6)	378 (69.9)	360 (71.7)
SQUIM….	No	91 ( 6.2)	12 ( 2.8)	25 ( 4.6)	54 (10.8)	<0.001
X.2	Si	1374 (93.8)	410 (97.2)	516 (95.4)	448 (89.2)
SRAD….	No	949 (64.8)	285 (67.5)	355 (65.6)	309 (61.6)	0.145
X.3	Si	516 (35.2)	137 (32.5)	186 (34.4)	193 (38.4)
Edadcat….	>=65	417 (28.5)	128 (30.3)	132 (24.4)	157 (31.3)	0.029
X.4	< 65	1048 (71.5)	294 (69.7)	409 (75.6)	345 (68.7)
Nocup….	-Ocup	971 (66.3)	299 (70.9)	349 (64.5)	323 (64.3)	0.062
X.5	+Ocup	494 (33.7)	123 (29.1)	192 (35.5)	179 (35.7)
Nrel….	Católica	1110 (75.8)	309 (73.2)	404 (74.7)	397 (79.1)	0.089
X.6	Otra	355 (24.2)	113 (26.8)	137 (25.3)	105 (20.9)
Tipotumor….	hematologico	44 ( 3.0)	15 ( 3.6)	16 ( 3.0)	13 ( 2.6)	0.691
X.7	tumor solido	1421 (97.0)	407 (96.4)	525 (97.0)	489 (97.4)
Nafiliacion….	Otro	859 (58.6)	266 (63.0)	289 (53.4)	304 (60.6)	0.006
X.8	Subsidiado-No asegurado	606 (41.4)	156 (37.0)	252 (46.6)	198 (39.4)
ResidenciaTipo….	Rural	339 (23.1)	78 (18.5)	158 (29.2)	103 (20.5)	<0.001
X.9	Urbano	1126 (76.9)	344 (81.5)	383 (70.8)	399 (79.5)
Estrato….	Alto	522 (35.6)	163 (38.6)	143 (26.4)	216 (43.0)	<0.001
X.10	Bajo	943 (64.4)	259 (61.4)	398 (73.6)	286 (57.0)
INFOnm….	No	91 ( 6.2)	23 ( 5.5)	20 ( 3.7)	48 ( 9.6)	<0.001
X.11	Si	1374 (93.8)	399 (94.5)	521 (96.3)	454 (90.4)
INFOm….	No	1254 (85.6)	373 (88.4)	467 (86.3)	414 (82.5)	0.032
X.12	Si	211 (14.4)	49 (11.6)	74 (13.7)	88 (17.5)
EFECTS….	1	343 (23.4)	95 (22.5)	127 (23.5)	121 (24.1)	0.850
X.13	2	1122 (76.6)	327 (77.5)	414 (76.5)	381 (75.9)
MAC_Animal….	No	1144 (78.1)	363 (86.0)	378 (69.9)	403 (80.3)	<0.001
X.14	Si	321 (21.9)	59 (14.0)	163 (30.1)	99 (19.7)
whole….	No	1070 (73.0)	319 (75.6)	392 (72.5)	359 (71.5)	0.353
X.15	Si	395 (27.0)	103 (24.4)	149 (27.5)	143 (28.5)
Hierbas….	No	635 (43.3)	177 (41.9)	205 (37.9)	253 (50.4)	<0.001
X.16	Si	830 (56.7)	245 (58.1)	336 (62.1)	249 (49.6)
Nutricionales….	No	615 (42.0)	206 (48.8)	211 (39.0)	198 (39.4)	0.003
X.17	Si	850 (58.0)	216 (51.2)	330 (61.0)	304 (60.6)
Vitaminas….	No	686 (46.8)	223 (52.8)	245 (45.3)	218 (43.4)	0.011
X.18	Si	779 (53.2)	199 (47.2)	296 (54.7)	284 (56.6)

Regresion multinominal clases

#instalar paquetes
#install.packages("sjmisc")
#install.packages("sjPlot")
#install.packages("nnet")
#install.packages("wakefield")

library(sjmisc)
library(sjPlot)
library(nnet)
library(wakefield)

frq(GAT_MAC, class)

## class <categorical> 
## # total N=1465 valid N=1465 mean=2.05 sd=0.79
## 
## Value       |   N | Raw % | Valid % | Cum. %
## --------------------------------------------
## R.Paliativa | 422 | 28.81 |   28.81 |  28.81
## R.Curativa  | 541 | 36.93 |   36.93 |  65.73
## R.Otra      | 502 | 34.27 |   34.27 | 100.00
## <NA>        |   0 |  0.00 |    <NA> |   <NA>

Se cruzan algunas variables

sjt.xtab(GAT_MAC$class, GAT_MAC$Edadcat, show.col.prc = TRUE)

class	Edadcat		Total
	=65	< 65
R.Paliativa	128 30.7 %	294 28.1 %	422 28.8 %
R.Curativa	132 31.7 %	409 39 %	541 36.9 %
R.Otra	157 37.6 %	345 32.9 %	502 34.3 %
Total	417 100 %	1048 100 %	1465 100 %
χ2=7.061 · df=2 · Cramer’s V=0.069 · p=0.029

GAT_MAC$class.f <- GAT_MAC$class
modelo <- multinom(GAT_MAC$class.f ~ SQUIM + SRAD + Edadcat + Nocup+
                     Nrel + Nafiliacion + ResidenciaTipo + Estrato + INFOnm + 
                     INFOm + MAC_Animal + Hierbas + Nutricionales + Vitaminas,
                   data = GAT_MAC)

## # weights:  48 (30 variable)
## initial  value 1609.467003 
## iter  10 value 1539.372774
## iter  20 value 1519.427932
## iter  30 value 1518.535064
## final  value 1518.513933 
## converged

Se estiman los valores p

z_values <- summary(modelo)$coefficients / summary(modelo)$standard.errors
p_values <- 2 * (1 - pnorm(abs(z_values)))  # Prueba z para cada coeficiente
p_values

##            (Intercept)      SQUIMSi    SRADSi Edadcat< 65 Nocup+Ocup   NrelOtra
## R.Curativa 0.220977970 7.938691e-02 0.5055719   0.2783110 0.01725939 0.55638199
## R.Otra     0.001622557 3.001007e-05 0.2642931   0.5926915 0.02812648 0.03068386
##            NafiliacionSubsidiado-No asegurado ResidenciaTipoUrbano EstratoBajo
## R.Curativa                         0.29911634          0.005414589  0.01440886
## R.Otra                             0.06212114          0.267311294  0.15009476
##              INFOnmSi    INFOmSi MAC_AnimalSi  HierbasSi NutricionalesSi
## R.Curativa 0.06654108 0.03049394 2.241753e-06 0.24531901     0.057403838
## R.Otra     0.22941719 0.45120512 2.959605e-02 0.03701061     0.008841626
##            VitaminasSi
## R.Curativa  0.01123795
## R.Otra      0.01201105

Se excluyen la variable SRAD + Edadcat + Nafiliacion

modelo2 <- multinom(GAT_MAC$class.f ~ SQUIM + Nocup+
                     ResidenciaTipo + Estrato + INFOm + MAC_Animal + Hierbas + Nutricionales + Vitaminas + INFOnm + Nrel,data = GAT_MAC)

## # weights:  39 (24 variable)
## initial  value 1609.467003 
## iter  10 value 1542.881392
## iter  20 value 1523.383037
## final  value 1522.331586 
## converged

Valores p

z_values2 <- summary(modelo2)$coefficients / summary(modelo2)$standard.errors
p_values2 <- 2 * (1 - pnorm(abs(z_values2)))  # Prueba z para cada coeficiente
p_values2

##             (Intercept)      SQUIMSi  Nocup+Ocup ResidenciaTipoUrbano
## R.Curativa 0.3450121693 6.959754e-02 0.008575184          0.003512065
## R.Otra     0.0003476365 8.810118e-06 0.058523754          0.182801258
##            EstratoBajo    INFOmSi MAC_AnimalSi  HierbasSi NutricionalesSi
## R.Curativa 0.002055084 0.03439783 1.207936e-06 0.22235853     0.047449810
## R.Otra     0.387282854 0.51659812 1.803399e-02 0.05415065     0.009693912
##            VitaminasSi   INFOnmSi  NrelOtra
## R.Curativa  0.01415932 0.06213981 0.6330594
## R.Otra      0.01378125 0.21814267 0.0284339

library(lmtest)
lrtest(modelo, modelo2)

## Likelihood ratio test
## 
## Model 1: GAT_MAC$class.f ~ SQUIM + SRAD + Edadcat + Nocup + Nrel + Nafiliacion + 
##     ResidenciaTipo + Estrato + INFOnm + INFOm + MAC_Animal + 
##     Hierbas + Nutricionales + Vitaminas
## Model 2: GAT_MAC$class.f ~ SQUIM + Nocup + ResidenciaTipo + Estrato + 
##     INFOm + MAC_Animal + Hierbas + Nutricionales + Vitaminas + 
##     INFOnm + Nrel
##   #Df  LogLik Df  Chisq Pr(>Chisq)
## 1  30 -1518.5                     
## 2  24 -1522.3 -6 7.6353     0.2661

Se ejecuta un 3 modelo

modelo3 <- multinom(GAT_MAC$class.f ~ SQUIM + Nocup+
                     ResidenciaTipo + Estrato + INFOm + MAC_Animal + Hierbas + Nutricionales + Vitaminas  + Nrel,data = GAT_MAC)

## # weights:  36 (22 variable)
## initial  value 1609.467003 
## iter  10 value 1541.891150
## iter  20 value 1528.894735
## final  value 1528.504841 
## converged

z_values3 <- summary(modelo3)$coefficients / summary(modelo3)$standard.errors
p_values3 <- 2 * (1 - pnorm(abs(z_values3)))  # Prueba z para cada coeficiente
p_values3

##             (Intercept)      SQUIMSi Nocup+Ocup ResidenciaTipoUrbano
## R.Curativa 0.6407256039 6.496077e-02 0.01063005          0.003396393
## R.Otra     0.0001677208 8.799787e-06 0.05015167          0.181745990
##            EstratoBajo    INFOmSi MAC_AnimalSi  HierbasSi NutricionalesSi
## R.Curativa 0.002255399 0.16815946 1.101865e-06 0.19276121      0.03553969
## R.Otra     0.399057258 0.04635155 1.995607e-02 0.04651231      0.01190475
##            VitaminasSi  NrelOtra
## R.Curativa 0.008464415 0.5604208
## R.Otra     0.018996664 0.0332499

library(lmtest)
lrtest(modelo2, modelo3)

## Likelihood ratio test
## 
## Model 1: GAT_MAC$class.f ~ SQUIM + Nocup + ResidenciaTipo + Estrato + 
##     INFOm + MAC_Animal + Hierbas + Nutricionales + Vitaminas + 
##     INFOnm + Nrel
## Model 2: GAT_MAC$class.f ~ SQUIM + Nocup + ResidenciaTipo + Estrato + 
##     INFOm + MAC_Animal + Hierbas + Nutricionales + Vitaminas + 
##     Nrel
##   #Df  LogLik Df  Chisq Pr(>Chisq)   
## 1  24 -1522.3                        
## 2  22 -1528.5 -2 12.347   0.002084 **
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

modelo3$AIC

## [1] 3101.01

modelo2$AIC

## [1] 3092.663

Optar por stepwise

library(MASS)
modelo_stepwise <- stepAIC(modelo, direction = "backward", trace = TRUE)

## Start:  AIC=3097.03
## GAT_MAC$class.f ~ SQUIM + SRAD + Edadcat + Nocup + Nrel + Nafiliacion + 
##     ResidenciaTipo + Estrato + INFOnm + INFOm + MAC_Animal + 
##     Hierbas + Nutricionales + Vitaminas
## 
## # weights:  45 (28 variable)
## initial  value 1609.467003 
## iter  10 value 1544.649536
## iter  20 value 1531.074440
## iter  30 value 1529.981843
## final  value 1529.969372 
## converged
## # weights:  45 (28 variable)
## initial  value 1609.467003 
## iter  10 value 1538.420096
## iter  20 value 1519.690787
## iter  30 value 1519.151072
## final  value 1519.140097 
## converged
## # weights:  45 (28 variable)
## initial  value 1609.467003 
## iter  10 value 1539.485537
## iter  20 value 1520.814312
## iter  30 value 1520.011100
## final  value 1519.994459 
## converged
## # weights:  45 (28 variable)
## initial  value 1609.467003 
## iter  10 value 1539.935140
## iter  20 value 1522.554353
## iter  30 value 1521.937316
## final  value 1521.925548 
## converged
## # weights:  45 (28 variable)
## initial  value 1609.467003 
## iter  10 value 1541.248268
## iter  20 value 1522.334361
## iter  30 value 1521.113285
## final  value 1521.095738 
## converged
## # weights:  45 (28 variable)
## initial  value 1609.467003 
## iter  10 value 1545.833526
## iter  20 value 1522.304184
## iter  30 value 1520.272132
## final  value 1520.261467 
## converged
## # weights:  45 (28 variable)
## initial  value 1609.467003 
## iter  10 value 1536.456611
## iter  20 value 1523.104072
## iter  30 value 1522.639016
## final  value 1522.634624 
## converged
## # weights:  45 (28 variable)
## initial  value 1609.467003 
## iter  10 value 1541.658738
## iter  20 value 1527.228352
## iter  30 value 1526.716231
## final  value 1526.711280 
## converged
## # weights:  45 (28 variable)
## initial  value 1609.467003 
## iter  10 value 1540.457358
## iter  20 value 1525.183444
## iter  30 value 1524.482687
## final  value 1524.459087 
## converged
## # weights:  45 (28 variable)
## initial  value 1609.467003 
## iter  10 value 1543.556086
## iter  20 value 1522.761941
## iter  30 value 1521.172940
## final  value 1521.159242 
## converged
## # weights:  45 (28 variable)
## initial  value 1609.467003 
## iter  10 value 1547.413235
## iter  20 value 1531.932476
## iter  30 value 1530.745368
## final  value 1530.718617 
## converged
## # weights:  45 (28 variable)
## initial  value 1609.467003 
## iter  10 value 1544.138500
## iter  20 value 1525.316449
## iter  30 value 1524.630639
## final  value 1524.627719 
## converged
## # weights:  45 (28 variable)
## initial  value 1609.467003 
## iter  10 value 1542.070486
## iter  20 value 1522.954221
## iter  30 value 1522.124956
## final  value 1522.118789 
## converged
## # weights:  45 (28 variable)
## initial  value 1609.467003 
## iter  10 value 1545.556992
## iter  20 value 1523.528639
## iter  30 value 1522.638854
## final  value 1522.635532 
## converged
##                  Df    AIC
## - SRAD            2 3094.3
## - Edadcat         2 3096.0
## - Nafiliacion     2 3096.5
## <none>              3097.0
## - Nrel            2 3098.2
## - INFOm           2 3098.3
## - Nocup           2 3099.8
## - Nutricionales   2 3100.2
## - ResidenciaTipo  2 3101.3
## - Vitaminas       2 3101.3
## - INFOnm          2 3104.9
## - Hierbas         2 3105.3
## - Estrato         2 3109.4
## - SQUIM           2 3115.9
## - MAC_Animal      2 3117.4
## # weights:  45 (28 variable)
## initial  value 1609.467003 
## iter  10 value 1538.420096
## iter  20 value 1519.690787
## iter  30 value 1519.151072
## final  value 1519.140097 
## converged
## 
## Step:  AIC=3094.28
## GAT_MAC$class.f ~ SQUIM + Edadcat + Nocup + Nrel + Nafiliacion + 
##     ResidenciaTipo + Estrato + INFOnm + INFOm + MAC_Animal + 
##     Hierbas + Nutricionales + Vitaminas
## 
## # weights:  42 (26 variable)
## initial  value 1609.467003 
## iter  10 value 1549.078634
## iter  20 value 1532.461601
## iter  30 value 1531.627386
## final  value 1531.626730 
## converged
## # weights:  42 (26 variable)
## initial  value 1609.467003 
## iter  10 value 1538.078918
## iter  20 value 1521.057332
## iter  30 value 1520.648161
## final  value 1520.645832 
## converged
## # weights:  42 (26 variable)
## initial  value 1609.467003 
## iter  10 value 1539.488944
## iter  20 value 1523.433861
## iter  30 value 1522.557718
## final  value 1522.541605 
## converged
## # weights:  42 (26 variable)
## initial  value 1609.467003 
## iter  10 value 1540.598582
## iter  20 value 1522.540459
## iter  30 value 1521.902261
## final  value 1521.898822 
## converged
## # weights:  42 (26 variable)
## initial  value 1609.467003 
## iter  10 value 1544.015754
## iter  20 value 1521.879792
## iter  30 value 1520.879055
## final  value 1520.876026 
## converged
## # weights:  42 (26 variable)
## initial  value 1609.467003 
## iter  10 value 1534.279879
## iter  20 value 1524.066732
## iter  30 value 1523.258641
## final  value 1523.258592 
## converged
## # weights:  42 (26 variable)
## initial  value 1609.467003 
## iter  10 value 1543.174608
## iter  20 value 1528.093237
## iter  30 value 1527.348950
## final  value 1527.342068 
## converged
## # weights:  42 (26 variable)
## initial  value 1609.467003 
## iter  10 value 1544.085953
## iter  20 value 1525.922518
## iter  30 value 1525.114599
## final  value 1525.112146 
## converged
## # weights:  42 (26 variable)
## initial  value 1609.467003 
## iter  10 value 1542.680386
## iter  20 value 1522.631135
## iter  30 value 1521.771375
## final  value 1521.766656 
## converged
## # weights:  42 (26 variable)
## initial  value 1609.467003 
## iter  10 value 1550.201300
## iter  20 value 1532.652568
## iter  30 value 1531.402986
## final  value 1531.401486 
## converged
## # weights:  42 (26 variable)
## initial  value 1609.467003 
## iter  10 value 1543.935051
## iter  20 value 1526.594539
## iter  30 value 1525.220014
## final  value 1525.214201 
## converged
## # weights:  42 (26 variable)
## initial  value 1609.467003 
## iter  10 value 1540.726397
## iter  20 value 1523.160770
## iter  30 value 1522.602693
## final  value 1522.599954 
## converged
## # weights:  42 (26 variable)
## initial  value 1609.467003 
## iter  10 value 1541.137472
## iter  20 value 1524.038770
## iter  30 value 1523.327745
## final  value 1523.325329 
## converged
##                  Df    AIC
## - Edadcat         2 3093.3
## - Nafiliacion     2 3093.8
## <none>              3094.3
## - INFOm           2 3095.5
## - Nrel            2 3095.8
## - Nocup           2 3097.1
## - Nutricionales   2 3097.2
## - ResidenciaTipo  2 3098.5
## - Vitaminas       2 3098.7
## - INFOnm          2 3102.2
## - Hierbas         2 3102.4
## - Estrato         2 3106.7
## - MAC_Animal      2 3114.8
## - SQUIM           2 3115.2
## # weights:  42 (26 variable)
## initial  value 1609.467003 
## iter  10 value 1538.078918
## iter  20 value 1521.057332
## iter  30 value 1520.648161
## final  value 1520.645832 
## converged
## 
## Step:  AIC=3093.29
## GAT_MAC$class.f ~ SQUIM + Nocup + Nrel + Nafiliacion + ResidenciaTipo + 
##     Estrato + INFOnm + INFOm + MAC_Animal + Hierbas + Nutricionales + 
##     Vitaminas
## 
## # weights:  39 (24 variable)
## initial  value 1609.467003 
## iter  10 value 1545.549057
## iter  20 value 1533.997145
## iter  30 value 1533.547054
## iter  30 value 1533.547043
## iter  30 value 1533.547043
## final  value 1533.547043 
## converged
## # weights:  39 (24 variable)
## initial  value 1609.467003 
## iter  10 value 1541.426032
## iter  20 value 1525.117706
## final  value 1524.690261 
## converged
## # weights:  39 (24 variable)
## initial  value 1609.467003 
## iter  10 value 1541.772188
## iter  20 value 1524.870079
## final  value 1523.644605 
## converged
## # weights:  39 (24 variable)
## initial  value 1609.467003 
## iter  10 value 1542.881392
## iter  20 value 1523.383037
## final  value 1522.331586 
## converged
## # weights:  39 (24 variable)
## initial  value 1609.467003 
## iter  10 value 1535.869439
## iter  20 value 1525.323795
## iter  30 value 1524.695573
## iter  30 value 1524.695572
## iter  30 value 1524.695572
## final  value 1524.695572 
## converged
## # weights:  39 (24 variable)
## initial  value 1609.467003 
## iter  10 value 1542.596091
## iter  20 value 1529.836582
## final  value 1529.431710 
## converged
## # weights:  39 (24 variable)
## initial  value 1609.467003 
## iter  10 value 1540.775566
## iter  20 value 1527.274828
## iter  30 value 1526.794328
## iter  30 value 1526.794328
## iter  30 value 1526.794328
## final  value 1526.794328 
## converged
## # weights:  39 (24 variable)
## initial  value 1609.467003 
## iter  10 value 1542.574476
## iter  20 value 1524.157403
## iter  30 value 1523.305699
## iter  30 value 1523.305697
## iter  30 value 1523.305697
## final  value 1523.305697 
## converged
## # weights:  39 (24 variable)
## initial  value 1609.467003 
## iter  10 value 1549.429541
## iter  20 value 1533.813443
## final  value 1532.886805 
## converged
## # weights:  39 (24 variable)
## initial  value 1609.467003 
## iter  10 value 1539.035797
## iter  20 value 1527.582605
## final  value 1526.506266 
## converged
## # weights:  39 (24 variable)
## initial  value 1609.467003 
## iter  10 value 1541.400450
## iter  20 value 1524.743727
## final  value 1524.105597 
## converged
## # weights:  39 (24 variable)
## initial  value 1609.467003 
## iter  10 value 1541.376182
## iter  20 value 1525.552266
## final  value 1524.816913 
## converged
##                  Df    AIC
## - Nafiliacion     2 3092.7
## <none>              3093.3
## - INFOm           2 3094.6
## - Nrel            2 3095.3
## - Nutricionales   2 3096.2
## - Nocup           2 3097.4
## - ResidenciaTipo  2 3097.4
## - Vitaminas       2 3097.6
## - Hierbas         2 3101.0
## - INFOnm          2 3101.6
## - Estrato         2 3106.9
## - MAC_Animal      2 3113.8
## - SQUIM           2 3115.1
## # weights:  39 (24 variable)
## initial  value 1609.467003 
## iter  10 value 1542.881392
## iter  20 value 1523.383037
## final  value 1522.331586 
## converged
## 
## Step:  AIC=3092.66
## GAT_MAC$class.f ~ SQUIM + Nocup + Nrel + ResidenciaTipo + Estrato + 
##     INFOnm + INFOm + MAC_Animal + Hierbas + Nutricionales + Vitaminas
## 
## # weights:  36 (22 variable)
## initial  value 1609.467003 
## iter  10 value 1548.134924
## iter  20 value 1535.886722
## final  value 1535.626568 
## converged
## # weights:  36 (22 variable)
## initial  value 1609.467003 
## iter  10 value 1547.040504
## iter  20 value 1526.488754
## final  value 1525.957189 
## converged
## # weights:  36 (22 variable)
## initial  value 1609.467003 
## iter  10 value 1548.649163
## iter  20 value 1526.421085
## final  value 1525.088754 
## converged
## # weights:  36 (22 variable)
## initial  value 1609.467003 
## iter  10 value 1541.258321
## iter  20 value 1527.363418
## final  value 1526.766362 
## converged
## # weights:  36 (22 variable)
## initial  value 1609.467003 
## iter  10 value 1541.820244
## iter  20 value 1532.476720
## final  value 1531.612031 
## converged
## # weights:  36 (22 variable)
## initial  value 1609.467003 
## iter  10 value 1541.891150
## iter  20 value 1528.894735
## final  value 1528.504841 
## converged
## # weights:  36 (22 variable)
## initial  value 1609.467003 
## iter  10 value 1549.987974
## iter  20 value 1526.222150
## final  value 1524.940633 
## converged
## # weights:  36 (22 variable)
## initial  value 1609.467003 
## iter  10 value 1558.168422
## iter  20 value 1535.730526
## final  value 1535.079202 
## converged
## # weights:  36 (22 variable)
## initial  value 1609.467003 
## iter  10 value 1546.452652
## iter  20 value 1528.972255
## final  value 1528.013219 
## converged
## # weights:  36 (22 variable)
## initial  value 1609.467003 
## iter  10 value 1547.396007
## iter  20 value 1526.947712
## final  value 1525.929650 
## converged
## # weights:  36 (22 variable)
## initial  value 1609.467003 
## iter  10 value 1545.285525
## iter  20 value 1526.734759
## final  value 1526.241908 
## converged
##                  Df    AIC
## <none>              3092.7
## - INFOm           2 3093.9
## - Nrel            2 3094.2
## - Nutricionales   2 3095.9
## - Nocup           2 3095.9
## - Vitaminas       2 3096.5
## - ResidenciaTipo  2 3097.5
## - Hierbas         2 3100.0
## - INFOnm          2 3101.0
## - Estrato         2 3107.2
## - MAC_Animal      2 3114.2
## - SQUIM           2 3115.2

summary(modelo_stepwise)

## Call:
## multinom(formula = GAT_MAC$class.f ~ SQUIM + Nocup + Nrel + ResidenciaTipo + 
##     Estrato + INFOnm + INFOm + MAC_Animal + Hierbas + Nutricionales + 
##     Vitaminas, data = GAT_MAC)
## 
## Coefficients:
##            (Intercept)    SQUIMSi Nocup+Ocup    NrelOtra ResidenciaTipoUrbano
## R.Curativa  -0.5438082 -0.6630189  0.3824717 -0.07322162           -0.4893526
## R.Otra       1.9054560 -1.4749685  0.2775858 -0.35000488           -0.2372652
##            EstratoBajo   INFOnmSi   INFOmSi MAC_AnimalSi  HierbasSi
## R.Curativa   0.4642266  0.7453211 0.5453104    0.8465776  0.1697246
## R.Otra      -0.1259393 -0.4553146 0.1752311    0.4397541 -0.2676301
##            NutricionalesSi VitaminasSi
## R.Curativa       0.2731687   0.3381169
## R.Otra           0.3606175   0.3436759
## 
## Std. Errors:
##            (Intercept)   SQUIMSi Nocup+Ocup  NrelOtra ResidenciaTipoUrbano
## R.Curativa   0.5758800 0.3653961  0.1455072 0.1533677            0.1676465
## R.Otra       0.5327052 0.3318646  0.1467346 0.1597286            0.1781030
##            EstratoBajo  INFOnmSi   INFOmSi MAC_AnimalSi HierbasSi
## R.Curativa   0.1506176 0.3995724 0.2577840    0.1743963 0.1390865
## R.Otra       0.1456694 0.3697290 0.2701689    0.1859489 0.1389832
##            NutricionalesSi VitaminasSi
## R.Curativa       0.1378065   0.1378271
## R.Otra           0.1394197   0.1395403
## 
## Residual Deviance: 3044.663 
## AIC: 3092.663

OR

exp(coef(modelo2))

##            (Intercept)   SQUIMSi Nocup+Ocup ResidenciaTipoUrbano EstratoBajo
## R.Curativa   0.5805332 0.5152933   1.465903            0.6130232   1.5907834
## R.Otra       6.7224722 0.2287859   1.319939            0.7887821   0.8816684
##             INFOmSi MAC_AnimalSi HierbasSi NutricionalesSi VitaminasSi
## R.Curativa 1.725144     2.331653 1.1849785        1.314122    1.402304
## R.Otra     1.191521     1.552325 0.7651908        1.434215    1.410122
##             INFOnmSi  NrelOtra
## R.Curativa 2.1071178 0.9293948
## R.Otra     0.6342484 0.7046847

tab_model(modelo2)

	Dependent variable
Predictors	Odds Ratios	CI	p	Response
(Intercept)	0.58	0.19 – 1.80	0.345	R.Curativa
SQUIM [Si]	0.52	0.25 – 1.06	0.070	R.Curativa
Nocup [+Ocup]	1.47	1.10 – 1.95	0.009	R.Curativa
ResidenciaTipo [Urbano]	0.61	0.44 – 0.85	0.004	R.Curativa
Estrato [Bajo]	1.59	1.18 – 2.14	0.002	R.Curativa
INFOm [Si]	1.73	1.04 – 2.86	0.035	R.Curativa
MAC Animal [Si]	2.33	1.66 – 3.28	<0.001	R.Curativa
Hierbas [Si]	1.18	0.90 – 1.56	0.223	R.Curativa
Nutricionales [Si]	1.31	1.00 – 1.72	0.048	R.Curativa
Vitaminas [Si]	1.40	1.07 – 1.84	0.014	R.Curativa
INFOnm [Si]	2.11	0.96 – 4.61	0.062	R.Curativa
Nrel [Otra]	0.93	0.69 – 1.26	0.633	R.Curativa
(Intercept)	6.72	0.19 – 1.80	0.345	R.Curativa
SQUIM [Si]	0.23	0.25 – 1.06	0.070	R.Curativa
Nocup [+Ocup]	1.32	1.10 – 1.95	0.009	R.Curativa
ResidenciaTipo [Urbano]	0.79	0.44 – 0.85	0.004	R.Curativa
Estrato [Bajo]	0.88	1.18 – 2.14	0.002	R.Curativa
INFOm [Si]	1.19	1.04 – 2.86	0.035	R.Curativa
MAC Animal [Si]	1.55	1.66 – 3.28	<0.001	R.Curativa
Hierbas [Si]	0.77	0.90 – 1.56	0.223	R.Curativa
Nutricionales [Si]	1.43	1.00 – 1.72	0.048	R.Curativa
Vitaminas [Si]	1.41	1.07 – 1.84	0.014	R.Curativa
INFOnm [Si]	0.63	0.96 – 4.61	0.062	R.Curativa
Nrel [Otra]	0.70	0.69 – 1.26	0.633	R.Curativa
Observations	1465
R2 / R2 adjusted	0.050 / 0.049

library(gtsummary)
tbl_regression(modelo2, exponentiate = TRUE, add_estimate_to_reference_rows=TRUE)  %>% add_global_p()

Characteristic	OR1	95% CI1	p-value
R.Curativa
SQUIM			<0.001
No	1.00	—
Si	0.52	0.25, 1.05
Nocup			0.027
-Ocup	1.00	—
+Ocup	1.47	1.10, 1.95
ResidenciaTipo			0.012
Rural	1.00	—
Urbano	0.61	0.44, 0.85
Estrato			<0.001
Alto	1.00	—
Bajo	1.59	1.18, 2.14
INFOm			0.074
No	1.00	—
Si	1.73	1.04, 2.86
MAC_Animal			<0.001
No	1.00	—
Si	2.33	1.66, 3.28
Hierbas			0.003
No	1.00	—
Si	1.18	0.90, 1.56
Nutricionales			0.027
No	1.00	—
Si	1.31	1.00, 1.72
Vitaminas			0.020
No	1.00	—
Si	1.40	1.07, 1.84
INFOnm			0.002
No	1.00	—
Si	2.11	0.96, 4.61
Nrel			0.063
Católica	1.00	—
Otra	0.93	0.69, 1.26
R.Otra
SQUIM			<0.001
No	1.00	—
Si	0.23	0.12, 0.44
Nocup			0.027
-Ocup	1.00	—
+Ocup	1.32	0.99, 1.76
ResidenciaTipo			0.012
Rural	1.00	—
Urbano	0.79	0.56, 1.12
Estrato			<0.001
Alto	1.00	—
Bajo	0.88	0.66, 1.17
INFOm			0.074
No	1.00	—
Si	1.19	0.70, 2.02
MAC_Animal			<0.001
No	1.00	—
Si	1.55	1.08, 2.23
Hierbas			0.003
No	1.00	—
Si	0.77	0.58, 1.00
Nutricionales			0.027
No	1.00	—
Si	1.43	1.09, 1.88
Vitaminas			0.020
No	1.00	—
Si	1.41	1.07, 1.85
INFOnm			0.002
No	1.00	—
Si	0.63	0.31, 1.31
Nrel			0.063
Católica	1.00	—
Otra	0.70	0.52, 0.96
1 OR = Odds Ratio, CI = Confidence Interval