Analysis Oral Health Latvia 2016

Packages and dataset

require(tidyverse)
require(stargazer)
require(ggthemes)

df <- df[, !duplicated(colnames(df))]

df <- df %>% 
        select(-c(C17V:Trauma_37))

dim(df)

Exploratory Data Analysis

addmargins(table(df$RegionName, df$`1_gender`))

         
             F    M  Sum
  Kurzeme  145  144  289
  Latgale  157  137  294
  Pieriga  198  215  413
  Riga     292  353  645
  Vidzeme   86  116  202
  Zemgale  153  142  295
  Sum     1031 1107 2138

FAS

Prevalence caries

Children with at least one tooth with D1

nrow(subset(df, D1T > 0))

[1] 2023

Children with at least one tooth with D3

nrow(subset(df,D3T > 0))

[1] 937

Children with at least one tooth with D5

nrow(subset(df,D5T > 0))

[1] 470

Children with at least one tooth with F

nrow(subset(df, FT > 0))

[1] 1411

Children with at least one tooth with M

nrow(subset(df, MT > 0))

[1] 36

Children with at least one tooth with D1MFT

nrow(subset(df, D1MFT > 0))

[1] 2105

Children with at least one tooth with D3MFT

nrow(subset(df, D3MFT > 0))

[1] 1705

DMFS

All

df %>% 
  summarise_each(funs(mean, median, sd) , D1S:Sealants)

df %>% 
  summarise_each(funs(quantile(., probs = 0.25)) , D1S:Sealants)

df %>% 
  summarise_each(funs(quantile(., probs = 0.75)) , D1S:Sealants)

Recode D3MFT in 0, 1

df <- df %>% 
        mutate( bin.D3T = ifelse(D3MFT == 0, 0, 1))
df$bin.D3T <- as.factor(df$bin.D3T)

Comparison by gender, DMFT and DMFS

df %>% 
        group_by(`1_gender`) %>% 
        summarise_each(funs(mean) , D1T:Sealants)

t.test(df$D1T~df$`1_gender`)

    Welch Two Sample t-test

data:  df$D1T by df$`1_gender`
t = -1, df = 2000, p-value = 0.2
alternative hypothesis: true difference in means is not equal to 0
95 percent confidence interval:
 -0.59  0.14
sample estimates:
mean in group F mean in group M 
           5.74            5.96 

t.test(df$D3T~df$`1_gender`)

    Welch Two Sample t-test

data:  df$D3T by df$`1_gender`
t = 1, df = 2000, p-value = 0.3
alternative hypothesis: true difference in means is not equal to 0
95 percent confidence interval:
 -0.0537  0.1744
sample estimates:
mean in group F mean in group M 
          0.919           0.858 

t.test(df$D5T~df$`1_gender`)

    Welch Two Sample t-test

data:  df$D5T by df$`1_gender`
t = -0.5, df = 2000, p-value = 0.6
alternative hypothesis: true difference in means is not equal to 0
95 percent confidence interval:
 -0.112  0.066
sample estimates:
mean in group F mean in group M 
          0.400           0.423 

t.test(df$FT~df$`1_gender`)

    Welch Two Sample t-test

data:  df$FT by df$`1_gender`
t = 2, df = 2000, p-value = 0.08
alternative hypothesis: true difference in means is not equal to 0
95 percent confidence interval:
 -0.0227  0.3572
sample estimates:
mean in group F mean in group M 
           2.12            1.95 

t.test(df$MT~df$`1_gender`)

    Welch Two Sample t-test

data:  df$MT by df$`1_gender`
t = 0.8, df = 2000, p-value = 0.4
alternative hypothesis: true difference in means is not equal to 0
95 percent confidence interval:
 -0.00801  0.02024
sample estimates:
mean in group F mean in group M 
         0.0233          0.0172 

t.test(df$D1MFT~df$`1_gender`)

    Welch Two Sample t-test

data:  df$D1MFT by df$`1_gender`
t = -0.06, df = 2000, p-value = 1
alternative hypothesis: true difference in means is not equal to 0
95 percent confidence interval:
 -0.469  0.440
sample estimates:
mean in group F mean in group M 
           9.20            9.21 

t.test(df$D3MFT~df$`1_gender`)

    Welch Two Sample t-test

data:  df$D3MFT by df$`1_gender`
t = 2, df = 2000, p-value = 0.1
alternative hypothesis: true difference in means is not equal to 0
95 percent confidence interval:
 -0.0439  0.4650
sample estimates:
mean in group F mean in group M 
           3.46            3.25 

t.test(df$D5MFT~df$`1_gender`)

    Welch Two Sample t-test

data:  df$D5MFT by df$`1_gender`
t = 1, df = 2000, p-value = 0.2
alternative hypothesis: true difference in means is not equal to 0
95 percent confidence interval:
 -0.061  0.361
sample estimates:
mean in group F mean in group M 
           2.54            2.39 

t.test(df$D1S~df$`1_gender`)

    Welch Two Sample t-test

data:  df$D1S by df$`1_gender`
t = -2, df = 2000, p-value = 0.1
alternative hypothesis: true difference in means is not equal to 0
95 percent confidence interval:
 -1.610  0.161
sample estimates:
mean in group F mean in group M 
           12.3            13.0 

t.test(df$D3S~df$`1_gender`)

    Welch Two Sample t-test

data:  df$D3S by df$`1_gender`
t = 0.4, df = 2000, p-value = 0.7
alternative hypothesis: true difference in means is not equal to 0
95 percent confidence interval:
 -0.113  0.162
sample estimates:
mean in group F mean in group M 
           1.03            1.00 

t.test(df$D5S~df$`1_gender`)

    Welch Two Sample t-test

data:  df$D5S by df$`1_gender`
t = -1, df = 2000, p-value = 0.3
alternative hypothesis: true difference in means is not equal to 0
95 percent confidence interval:
 -0.2595  0.0761
sample estimates:
mean in group F mean in group M 
          0.572           0.664 

t.test(df$FS~df$`1_gender`)

    Welch Two Sample t-test

data:  df$FS by df$`1_gender`
t = 0.5, df = 2000, p-value = 0.6
alternative hypothesis: true difference in means is not equal to 0
95 percent confidence interval:
 -0.249  0.436
sample estimates:
mean in group F mean in group M 
           3.28            3.19 

t.test(df$MS~df$`1_gender`)

    Welch Two Sample t-test

data:  df$MS by df$`1_gender`
t = 0.6, df = 2000, p-value = 0.5
alternative hypothesis: true difference in means is not equal to 0
95 percent confidence interval:
 -0.0480  0.0901
sample estimates:
mean in group F mean in group M 
         0.1096          0.0885 

t.test(df$D1MFS~df$`1_gender`)

    Welch Two Sample t-test

data:  df$D1MFS by df$`1_gender`
t = -1, df = 2000, p-value = 0.2
alternative hypothesis: true difference in means is not equal to 0
95 percent confidence interval:
 -1.789  0.435
sample estimates:
mean in group F mean in group M 
           17.2            17.9 

t.test(df$D3MFS~df$`1_gender`)

    Welch Two Sample t-test

data:  df$D3MFS by df$`1_gender`
t = 0.2, df = 2000, p-value = 0.8
alternative hypothesis: true difference in means is not equal to 0
95 percent confidence interval:
 -0.428  0.522
sample estimates:
mean in group F mean in group M 
           4.99            4.95 

t.test(df$D5MFS~df$`1_gender`)

    Welch Two Sample t-test

data:  df$D5MFS by df$`1_gender`
t = 0.1, df = 2000, p-value = 0.9
alternative hypothesis: true difference in means is not equal to 0
95 percent confidence interval:
 -0.395  0.441
sample estimates:
mean in group F mean in group M 
           3.96            3.94 

t.test(df$Sealants~df$`1_gender`)

    Welch Two Sample t-test

data:  df$Sealants by df$`1_gender`
t = 0.4, df = 2000, p-value = 0.7
alternative hypothesis: true difference in means is not equal to 0
95 percent confidence interval:
 -0.0473  0.0695
sample estimates:
mean in group F mean in group M 
          0.148           0.137 

df.long <- df %>%
        gather("byTooth", "valueByTooth", D1T:D5MFT) %>% 
        gather("bySurface", "valueBySurface", D1S:Sealants)
df.long$byTooth <- ordered(df.long$byTooth, levels = c(
        "D1T", "D3T", "D5T", "FT",
        "MT", "D1MFT", "D3MFT", "D5MFT"))
df.long$bySurface <- ordered(df.long$bySurface, levels = c(
         "D1S", "D3S", "D5S", "FS", "MS",
         "D1MFS", "D3MFS", "D5MFS", "Sealants"))

df.long %>% 
        ggplot(aes(factor(byTooth), valueByTooth)) +
        geom_boxplot(aes( fill = `1_gender` ) ) + 
        theme_minimal() +
        labs(title = " ", x = " ", y = "Zobi", color = "Dzimums\n")+
        ggsave("./plots/dmftByGender.png", width=8, height=6, dpi=250)

df.long %>% 
        ggplot(aes(factor(bySurface), valueBySurface)) +
        geom_boxplot(aes(fill = factor(`1_gender`))) + 
        theme_minimal() +
        labs(title = " ", x = " ", y = "Zobi", color = "Dzimums\n") +
        ggsave("./plots/dmfsByGender.png", width=8, height=6, dpi=250)

by region

by FAS

Sugar

Diet