2018 EADPH

Packages

library(tidyverse)
library(viridis)
# install.packages("ggsignif")
library(ggsignif)

dataset

df <- read_csv("https://docs.google.com/spreadsheets/d/e/2PACX-1vQi_9CH1iGd36xWS5ECw4czoj780kJaTWOY5CgwW3fXCYkU7I9cKYTUpNY2I-k5QTF4tC18tlSLgpjP/pub?gid=663675586&single=true&output=csv")

Missing column names filled in: 'X30' [30], 'X31' [31]Duplicated column names deduplicated: 'gadi kops beigsanas' => 'gadi kops beigsanas_1' [22]Parsed with column specification:
cols(
  .default = col_integer(),
  `Ud es:` = col_character(),
  `gadi kops beigsanas` = col_number(),
  `gadi kops beigsanas_1` = col_number(),
  X30 = col_character(),
  X31 = col_character(),
  Min = col_character(),
  Q1 = col_character(),
  Average = col_number(),
  SD = col_number(),
  `P<0.001` = col_character()
)
See spec(...) for full column specifications.

Data cleaning

df <- janitor::clean_names(df)
df <- janitor::remove_empty(df, c("rows", "cols"))

Graphs

Recommendations

df %>% 
  filter(!is.na(recommendations)) %>% 
  ggplot(aes(x = ud_es, y = recommendations, fill = ud_es)) +
  geom_boxplot() + 
  labs(title = "Recommendations about oral hygiene given by dental hygienist and dentists", 
       y = "Value", 
       x = "Group", 
       caption="*** p value < 0.01") +
  theme_classic() + 
  scale_fill_manual(values=c("#ffcc03", "#00377c")) + 
  geom_jitter(alpha = 0.1) + 
  geom_signif(comparisons = list(c("Dental hygienist", "Dentist")),
               map_signif_level = TRUE, textsize=6) + 
  coord_cartesian(ylim = c(0, 21)) + 
  theme(legend.position="none") 

NA

Attitude

df %>% 
  filter(!is.na(recommendations)) %>% 
  ggplot(aes(x = ud_es, y = attitude, fill = ud_es)) +
  geom_boxplot() + 
  labs(title = "Attitude about oral hygiene given by dental hygienist and dentists", 
       y = "Value", 
       x = "Group", 
       caption="NS. p > 0.05") +
  theme_classic() + 
  scale_fill_manual(values=c("#ffcc03", "#00377c")) + 
  geom_jitter(alpha = 0.1) + 
  geom_signif(comparisons = list(c("Dental hygienist", "Dentist")),
               map_signif_level = TRUE, textsize = 4) + 
  coord_cartesian(ylim = c(0, 28)) + 
  theme(legend.position="none")

Dati

dati <- read_csv("https://docs.google.com/spreadsheets/d/e/2PACX-1vQi_9CH1iGd36xWS5ECw4czoj780kJaTWOY5CgwW3fXCYkU7I9cKYTUpNY2I-k5QTF4tC18tlSLgpjP/pub?gid=411400485&single=true&output=csv")

Parsed with column specification:
cols(
  .default = col_character(),
  N = col_integer()
)
See spec(...) for full column specifications.

Weighted data

weighted <- read_csv("https://docs.google.com/spreadsheets/d/e/2PACX-1vQi_9CH1iGd36xWS5ECw4czoj780kJaTWOY5CgwW3fXCYkU7I9cKYTUpNY2I-k5QTF4tC18tlSLgpjP/pub?gid=663675586&single=true&output=csv")

Missing column names filled in: 'X30' [30], 'X31' [31]Duplicated column names deduplicated: 'gadi kops beigsanas' => 'gadi kops beigsanas_1' [22]Parsed with column specification:
cols(
  .default = col_integer(),
  `Ud es:` = col_character(),
  `gadi kops beigsanas` = col_number(),
  `gadi kops beigsanas_1` = col_number(),
  X30 = col_character(),
  X31 = col_character(),
  Min = col_character(),
  Q1 = col_character(),
  Average = col_number(),
  SD = col_number(),
  `P<0.001` = col_character()
)
See spec(...) for full column specifications.

Changes through time

Proportion tests

38% of dental hygienists and 68% of dentists do not recommend any toothbrushing technique.

prop.test(dy, dyt)

    2-sample test for equality of proportions with continuity
    correction

data:  dy out of dyt
X-squared = 16.881, df = 1, p-value = 3.98e-05
alternative hypothesis: two.sided
95 percent confidence interval:
 -0.441945 -0.158055
sample estimates:
prop 1 prop 2 
  0.38   0.68 

3 dentists (1.11%) and no dental hygienists would recommend toothpaste without fluoride.

prop.test(dy, dyt)

Chi-squared approximation may be incorrect

    2-sample test for equality of proportions with continuity
    correction

data:  dy out of dyt
X-squared = 0.010962, df = 1, p-value = 0.9166
alternative hypothesis: two.sided
95 percent confidence interval:
 -0.04163458  0.01943458
sample estimates:
prop 1 prop 2 
0.0000 0.0111 

55% of dental hygienists and 30% of dentists would recommend toothpaste with fluoride in a concentration of 1000 ppm or more, the rest of specialists (45% of dental hygienists and 70% of dentists) would not give any recommendations about required fluoride concentration in a toothpaste.

prop.test(dy, dyt)

    2-sample test for equality of proportions with continuity
    correction

data:  dy out of dyt
X-squared = 11.785, df = 1, p-value = 0.000597
alternative hypothesis: two.sided
95 percent confidence interval:
 0.1074305 0.3925695
sample estimates:
prop 1 prop 2 
  0.55   0.30 

Majority of dental team specialists would recommend interdental cleaning (95% of dental hygienists and 88% of dentists).

prop.test(dy, dyt)

    2-sample test for equality of proportions with continuity
    correction

data:  dy out of dyt
X-squared = 2.3144, df = 1, p-value = 0.1282
alternative hypothesis: two.sided
95 percent confidence interval:
 -0.01668946  0.15668946
sample estimates:
prop 1 prop 2 
  0.95   0.88