Revised analyses

Added additional scripts used in producing the revised version of the paper. Some files from the submitted paper were also missing, no rectified.

analysis/describing_trends_part5.R

@@ -0,0 +1,64 @@
+############################################
+### Trends in avoidable hospitalisations ###
+##### Part 5: Descriptives for report ######
+############################################
+
+# Purpose: To create a series of descriptive summary statistic tables for inclusion in the report.
+
+# Libraries
+library(dplyr)
+library(data.table)
+
+## 1. Load and tidy data ##
+
+# Load data
+trends_imd <- fread("../output/measures/standardised_imd_trends.csv") # Load
+trends_imd <- trends_imd[trends_imd$date < "2022-04-01"] # Drop last period
+trends_region <- fread("../output/measures/standardised_region_trends.csv") # Repeat
+trends_region <- trends_region[trends_region$date < "2022-04-01"] 
+trends_eth <- fread("../output/measures/standardised_ethnicity_trends.csv") # And again
+trends_eth <- trends_eth[trends_eth$date < "2022-04-01"] 
+trends_urbrur <- fread("../output/measures/standardised_urbrur_trends.csv") # Last time
+trends_urbrur <- trends_urbrur[trends_urbrur$date < "2022-04-01"] 
+
+# Convert to date format
+trends_imd$date <- as.Date(trends_imd$date) 
+trends_region$date <- as.Date(trends_region$date)
+
+# Shift dates to mid-point of month since 1st of month refers to whole month (akes plots nicer to look at)
+trends_imd$date <- trends_imd$date + 14
+
+## 2. Create tables ##
+
+# Outcome summary statistics #
+
+# Sum values to totals
+all_trends <- trends_imd[, list(admitted = sum(admitted), admitted_acs_all = sum(admitted_acs_all), admitted_acs_acute = sum(admitted_acs_acute), admitted_acs_chronic = sum(admitted_acs_chronic), admitted_acs_vaccine = sum(admitted_acs_vaccine), admitted_eucs = sum(admitted_eucs), pop = sum(pop)), by = c("date")]
+
+# Total admissions over study period
+sum(all_trends$admitted)
+sum(all_trends$admitted_acs_all)
+sum(all_trends$admitted_acs_acute)
+sum(all_trends$admitted_acs_chronic)
+sum(all_trends$admitted_acs_vaccine)
+sum(all_trends$admitted_eucs)
+sum(all_trends$pop)
+
+# Mean monthly rates
+mean(all_trends$admitted)
+mean(all_trends$admitted_acs_all)
+mean(all_trends$admitted_acs_acute)
+mean(all_trends$admitted_acs_chronic)
+mean(all_trends$admitted_acs_vaccine)
+mean(all_trends$admitted_eucs)
+mean(all_trends$pop)
+
+
+# Missing data # 
+(sum(trends_imd$pop[trends_imd$imd_quintile == 0 | is.na(trends_imd$imd_quintile)]) / sum(trends_imd$pop)) * 100 # IMD
+(sum(trends_eth$pop[trends_eth$ethnicity == 0]) / sum(trends_eth$pop)) * 100 # Ethnicity
+# (sum(trends_eth$pop[trends_eth$ethnicity == 5]) / sum(trends_eth$pop)) * 100
+(sum(trends_region$pop[is.na(trends_region$region) ) / sum(trends_region$pop)) * 100 # Region
+(sum(trends_urbrur$pop[trends_urbrur$urban_rural == 0 | trends_urbrur$urban_rural == -1]) / sum(trends_urbrur$pop)) * 100 # Urban/rural
+
+