---
title: "Vorhofkatheter-Statistik"
author: "Daniel Kraus"
date: '2018-12-29'
output:
  ioslides_presentation: default
  slidy_presentation: default
  beamer_presentation: default
---

```{r setup, include=FALSE}
knitr::opts_chunk$set(echo = FALSE, warning = FALSE)
library(tidyverse)

raw_data = read_csv('vhk.csv') %>%
  mutate(Year = lubridate::year(Date))

cath_by_year = raw_data %>% count(Year)
first_year = min(raw_data$Year)
last_year = max(raw_data$Year)
max_y_break = ((max(cath_by_year$n) %/% 10) + 1) * 10

```

## Katheterimplantationen pro Jahr
```{r cath_by_year }
cath_by_year %>%
  ggplot(aes(x = Year, y = n)) +
  geom_col() +
  scale_y_continuous(breaks = seq(from = 0, to = max_y_break, by = 10)) +
  scale_x_continuous(breaks = seq(from = first_year, to = last_year, by = 1)) +
  labs(x = NULL, y = "Anzahl Katheter")
```

## Alter der Patienten
```{r patient_age}
raw_data %>%
  ggplot(aes(group = Year, x = Year, y = Age)) +
  geom_boxplot() +
  coord_cartesian(ylim = c(20, 100)) +
  scale_x_continuous(breaks = seq(from = first_year, to = last_year, by = 1)) +
  scale_y_continuous(breaks = seq(from = 20, to = 100, by = 10))  +
  labs(x = NULL, y = "Jahre")
```

## Geschlecht der Patienten
```{r patient_sex}
raw_data %>% group_by(Year) %>% summarise(PercentFemale = sum(Sex == "weiblich") / n()) %>%
  ggplot(aes(x = Year, y = PercentFemale)) +
  geom_col() +
  scale_x_continuous(breaks = seq(from = first_year, to = last_year, by = 1)) +
  scale_y_continuous(labels = scales::percent_format(accuracy = 1)) +
  labs(x = NULL, y = "Anteil Frauen")
```

## Katheterlokalisation
Ist da ein Trend hin zu immer mehr Kathetern von links?!
```{r insertion_site}
raw_data %>% mutate(Side = factor(Side, levels = c("rechts", "links"))) %>%
  ggplot(aes(x = Year)) +
  facet_grid(InsertionSite ~ Side) +
  geom_bar() +
  labs(x = NULL, y = "Anzahl Katheter")
```

## Anteil der Arztrollen
Um 2014 herum haben einige die Facharztprüfung abgelegt, ist das der Grund für die Auffälligkeit 2015/2016?
```{r percent_residents}
raw_data %>% group_by(Year) %>% 
  summarize(Assistenzarzt = sum(SurgeonRole == "Assistenzarzt") / n(),
            Facharzt = sum(SurgeonRole == "Facharzt") / n(),
            Oberarzt = sum(SurgeonRole == "Oberarzt") / n()) %>%
  gather(key = Role, value = Percent, Assistenzarzt, Facharzt, Oberarzt) %>%
  ggplot(aes(x = Year, y = Percent)) +
  scale_x_continuous(breaks = seq(from = first_year, to = last_year, by = 1)) +
  scale_y_continuous(labels = scales::percent_format(accuracy = 1)) +
  facet_grid(Role ~ .) +
  geom_col() +
  labs(x = NULL, y = "Anteil in den gelegten Kathetern")
```

## Hitparade der Durchleuchtungsdauern
```{r greatest_fluoroscopy}
raw_data %>%
  group_by(Surgeon) %>%
  summarize(FluoroscopyIndex = median(InsertionFluoroscopyDuration, na.rm = TRUE)) %>%
  arrange(desc(FluoroscopyIndex)) %>%
  top_n(-10, FluoroscopyIndex) %>%
  mutate(Surgeon = factor(Surgeon, levels = Surgeon)) %>%
  ggplot(aes(x = Surgeon, y = FluoroscopyIndex)) +
  geom_col() +
  coord_flip() +
  labs(x = NULL, y = "Median der Durchleuchtungsdauer [s]")
```

## Hitparade der Implanteure
```{r greatest_surgeons}
raw_data %>% count(Surgeon) %>% arrange(n) %>% top_n(10, n) %>% mutate(Surgeon = factor(Surgeon, levels = Surgeon)) %>%
  ggplot(aes(x = Surgeon, y = n)) +
  geom_col() +
  coord_flip() +
  labs(x = NULL, y = "Gesamtzahl Katheter")
```

## Hitparade der Assistenten
```{r greatest_assistants}
raw_data %>% count(Assistant) %>% arrange(n) %>% top_n(10, n) %>% mutate(Assistant = factor(Assistant, levels = Assistant)) %>%
  ggplot(aes(x = Assistant, y = n)) +
  geom_col() +
  coord_flip() +
  labs(x = NULL, y = "Gesamtzahl Katheter")
```