Swiss Health

Deep Pockets and Long Lives in Switzerland

A September 20, 2020 broadcast of the PBS Newhour profiled Switzerland’s high-performing health care system. I decided to take a closer look with some exploratory data analysis.
- https://www.pbs.org/newshour/show/how-switzerland-delivered-health-care-for-all-and-kept-its-private-insurance
According to the Newshour:
- “The average Swiss pays more out-of-pocket for things like co-pays than the average American”
- “The Swiss live about five years longer, on average, than we do”
I focused on out-of-pocket (household) spending as a percentage of total health spending, as well as life expectancy.
All data is from the World Bank World Development Indicators
- https://databank.worldbank.org/source/world-development-indicators

Set-up includes data.table, fpp3, and kableExtra packages

# Paths
projpath <- "C:/Users/dancu/Documents/Fall2022_ADEC743002"
rawdata <- file.path(projpath, "RawData")
output <- file.path(projpath, "Output")

Read in files

First read in out-of-pocket spending as a percentage of total health expenditure
I “melted” the year columns into row values, and did some basic clean-up, mainly renaming columns.
Melting the columns positioned the out-of-pocket % as a column of data with yearly row labels.
For comparison, I included Switzerland, the US, the UK and all 27 countries in the EU. I realize many other combination of countries could be considered as well.

# Out-of-pocket spending 
pocketper<-read.csv(file.path(rawdata, "pocketper.csv"))
head(pocketper,3)

##   Country.Name Country.Code
## 1      Austria          AUT
## 2      Belgium          BEL
## 3     Bulgaria          BGR
##                                                   Series.Name       Series.Code
## 1 Out-of-pocket expenditure (% of current health expenditure) SH.XPD.OOPC.CH.ZS
## 2 Out-of-pocket expenditure (% of current health expenditure) SH.XPD.OOPC.CH.ZS
## 3 Out-of-pocket expenditure (% of current health expenditure) SH.XPD.OOPC.CH.ZS
##   X2015..YR2015. X2016..YR2016. X2017..YR2017. X2018..YR2018. X2019..YR2019.
## 1       19.09111       19.24335       20.63871       19.59343       19.07191
## 2       19.34007       18.00682       17.80459       18.23685       18.17289
## 3       42.50994       43.25471       43.36105       40.51738       38.97313

tail(pocketper) #shows empty rows

##                                        Country.Name Country.Code
## 30                                   United Kingdom          GBR
## 31                                                              
## 32                                                              
## 33                                                              
## 34 Data from database: World Development Indicators             
## 35                         Last Updated: 12/22/2022             
##                                                    Series.Name
## 30 Out-of-pocket expenditure (% of current health expenditure)
## 31                                                            
## 32                                                            
## 33                                                            
## 34                                                            
## 35                                                            
##          Series.Code X2015..YR2015. X2016..YR2016. X2017..YR2017.
## 30 SH.XPD.OOPC.CH.ZS       16.08236       16.28811       16.73583
## 31                               NA             NA             NA
## 32                               NA             NA             NA
## 33                               NA             NA             NA
## 34                               NA             NA             NA
## 35                               NA             NA             NA
##    X2018..YR2018. X2019..YR2019.
## 30       16.99979       17.07361
## 31             NA             NA
## 32             NA             NA
## 33             NA             NA
## 34             NA             NA
## 35             NA             NA

pocketper<-data.table(pocketper)
pocketper<-pocketper[1:30] #remove empty rows at the bottom

# Melt year columns into rows
pocketper.m <- data.table::melt(data = pocketper,
                                id.vars = c("Country.Name","Series.Code"),
                                measure.vars = c("X2015..YR2015.", "X2016..YR2016.", "X2017..YR2017.", "X2018..YR2018.","X2019..YR2019."))  

# Clean up table, rename column
pocketper.m2<-pocketper.m[,c(1,3,4)]
names(pocketper.m2)[names(pocketper.m2) == "value"] <- "OutPocketPct"

Read in, melt, and clean up life expectancy at birth

# Life Expectancy
le<-read.csv(file.path(rawdata, "le.csv"))
head(le,3)

##   Country.Name Country.Code                             Series.Name
## 1      Austria          AUT Life expectancy at birth, total (years)
## 2      Belgium          BEL Life expectancy at birth, total (years)
## 3     Bulgaria          BGR Life expectancy at birth, total (years)
##      Series.Code X2015..YR2015. X2016..YR2016. X2017..YR2017. X2018..YR2018.
## 1 SP.DYN.LE00.IN       81.19024       81.64146       81.64390       81.69268
## 2 SP.DYN.LE00.IN       80.99268       81.43902       81.49268       81.59512
## 3 SP.DYN.LE00.IN       74.61463       74.81220       74.81463       74.96341
##   X2019..YR2019.
## 1       81.89512
## 2       81.99512
## 3       75.11220

le<-data.table(le)
le<-le[1:30]

# Melt year columns into rows
le.m <- data.table::melt(data = le,
                         id.vars = c("Country.Name","Series.Code"),
                         measure.vars = c("X2015..YR2015.", "X2016..YR2016.", "X2017..YR2017.", "X2018..YR2018.","X2019..YR2019.")) 

le.m2<-le.m[,c(1,3,4)]

names(le.m2)[names(le.m2) == "value"] <- "LifeExp"

Merge spending and life expectancy data

# Check missing values before merging
sapply(pocketper.m2, function(x) sum(is.na(x)))

## Country.Name     variable OutPocketPct 
##            0            0            0

sapply(le.m2, function(x) sum(is.na(x)))

## Country.Name     variable      LifeExp 
##            0            0            0

health<-merge(pocketper.m2,le.m2)

# rename year values
health[variable == "X2015..YR2015."]$variable<-"2015"
health[variable == "X2016..YR2016."]$variable<-"2016"
health[variable == "X2017..YR2017."]$variable<-"2017"
health[variable == "X2018..YR2018."]$variable<-"2018"
health[variable == "X2019..YR2019."]$variable<-"2019"

names(health)[names(health) == "variable"] <- "Year"

Quick review of new table

# Glance at the new table
health[Country.Name=="Switzerland"]

##    Country.Name Year OutPocketPct  LifeExp
## 1:  Switzerland 2015     25.85815 82.89756
## 2:  Switzerland 2016     26.57211 83.60244
## 3:  Switzerland 2017     26.40589 83.55122
## 4:  Switzerland 2018     27.98195 83.75366
## 5:  Switzerland 2019     25.29318 83.90488

# round decimals
health$LifeExp<-round(health$LifeExp,2)
health$OutPocketPct<-round(health$OutPocketPct,2)

Summary Snapshot

Out of Pocket % of total health spending and Life Expectancy at Birth
- EU, Switzerland, UK, US 2015-2019

as.array(round(summary(health$OutPocketPct),2)) %>% kbl(caption="Out of Pocket %", col.names = NULL) %>% kable_classic(full_width=F,html_font="Times New Roman")

Out of Pocket %
Min.	9.26
1st Qu.	13.01
Median	19.12
Mean	21.57
3rd Qu.	27.64
Max.	44.43

as.array(round(summary(health$LifeExp),2)) %>% kbl(caption="Life Expectancy", col.names = NULL) %>% kable_classic(full_width=F,html_font="Times New Roman")

Life Expectancy
Min.	74.32
1st Qu.	77.86
Median	81.19
Mean	80.05
3rd Qu.	82.07
Max.	83.90

Switzerland and United States Summary Comparison

health[Country.Name=="Switzerland"] %>% kbl(caption="Switzerland Health Spending and LE") %>% kable_classic(full_width=F,html_font="Times New Roman")

Switzerland Health Spending and LE
Country.Name	Year	OutPocketPct	LifeExp
Switzerland	2015	25.86	82.90
Switzerland	2016	26.57	83.60
Switzerland	2017	26.41	83.55
Switzerland	2018	27.98	83.75
Switzerland	2019	25.29	83.90

health[Country.Name=="United States"] %>% kbl(caption="U.S. Health Spending and LE") %>% kable_classic(full_width=F,html_font="Times New Roman")

U.S. Health Spending and LE
Country.Name	Year	OutPocketPct	LifeExp
United States	2015	11.74	78.69
United States	2016	11.63	78.54
United States	2017	11.40	78.54
United States	2018	11.30	78.64
United States	2019	11.31	78.79

Scatterplot: Comparing Nations

I took an average of the 2015-2019 out-of-pocket and life expectancy values to plot them on a scatterplot.

meanout<-aggregate(health$OutPocketPct, by=list(health$Country.Name), FUN=mean)
names(meanout)[names(meanout) == "x"] <- "OutPocketPct"
meanlife<-aggregate(health$LifeExp, by=list(health$Country.Name), FUN=mean)
names(meanlife)[names(meanlife) == "x"] <- "LifeExp"
avgcountry<-merge(meanout,meanlife)
names(avgcountry)[names(avgcountry) == "Group.1"] <- "Country.Name"

ggplot(avgcountry, aes(x = OutPocketPct, y = LifeExp)) +
  labs(title = "EU, Switzerland, UK, US", subtitle="2015-2019 averages",x="Out of Pocket Spending (% of health spending)", y="Life Expectancy (years)")+
  geom_point(size=.75) +
  geom_text(size=2, aes(label=Country.Name)) +
  geom_smooth(method=lm,se=FALSE, fullrange=TRUE, lwd=.75)

## `geom_smooth()` using formula 'y ~ x'

The line of best fit reflects the slightly negative correlation between out-of-pocket spending and life expectancy. While not an outlier, Switzerland bucks this trend.

cor(health$OutPocketPct,health$LifeExp)

## [1] -0.3361479

Conclusions

This combination of high out-of-pocket spending and life expectancy provide a good starting point for more rigorous analysis. Burton-Jeangros, et. al. (2019) praise Switzerland’s consumer-driven and decentralized health marketplace. Perhaps their combination of government support, high volume spending, and market forces amount to an optimal mix for producing strong health outcomes.

Burton-Jeangros, C., Cullati, S., Oris, M., Remund, A., Sieber, S. (2019, August 31). Longer and healthier lives for all? Successes and failures of a universal consumer-driven healthcare system, Switzerland, 1990–2014. International Journal of Public Health, 64, 1173–1181. https://doi.org/10.1007/s00038-019-01290-5