Heat wave and Ozone, Mediation analysis on cardiac hospitalization

Creating Database only with summer months

villes_s<-list()

for(i in 1:length(villes)) {
  villes_s[[i]]<-villes[[i]][villes[[i]]$saison=="summer",]
  villes_s[[i]]$temps <- ave(villes_s[[i]]$time,villes_s[[i]]$annee,  FUN = seq_along)
  }

names(villes_s)<-cities

table of N obs per months to check to have only summer month

for (i in 1:length(villes)){
print(table(villes_s[[i]]$mois))
}

## 
##   1   2   3   4   5   6   7   8   9  10  11  12 
##   0   0   0   0   0 540 558 558 540   0   0   0 
## 
##   1   2   3   4   5   6   7   8   9  10  11  12 
##   0   0   0   0   0 540 558 558 540   0   0   0 
## 
##   1   2   3   4   5   6   7   8   9  10  11  12 
##   0   0   0   0   0 540 558 558 540   0   0   0 
## 
##   1   2   3   4   5   6   7   8   9  10  11  12 
##   0   0   0   0   0 540 558 558 540   0   0   0 
## 
##   1   2   3   4   5   6   7   8   9  10  11  12 
##   0   0   0   0   0 540 558 558 540   0   0   0 
## 
##   1   2   3   4   5   6   7   8   9  10  11  12 
##   0   0   0   0   0 540 558 558 540   0   0   0 
## 
##   1   2   3   4   5   6   7   8   9  10  11  12 
##   0   0   0   0   0 540 558 558 540   0   0   0 
## 
##   1   2   3   4   5   6   7   8   9  10  11  12 
##   0   0   0   0   0 540 558 558 540   0   0   0 
## 
##   1   2   3   4   5   6   7   8   9  10  11  12 
##   0   0   0   0   0 540 558 558 540   0   0   0 
## 
##   1   2   3   4   5   6   7   8   9  10  11  12 
##   0   0   0   0   0 540 558 558 540   0   0   0 
## 
##   1   2   3   4   5   6   7   8   9  10  11  12 
##   0   0   0   0   0 540 558 558 540   0   0   0 
## 
##   1   2   3   4   5   6   7   8   9  10  11  12 
##   0   0   0   0   0 540 558 558 540   0   0   0 
## 
##   1   2   3   4   5   6   7   8   9  10  11  12 
##   0   0   0   0   0 540 558 558 540   0   0   0 
## 
##   1   2   3   4   5   6   7   8   9  10  11  12 
##   0   0   0   0   0 540 558 558 540   0   0   0 
## 
##   1   2   3   4   5   6   7   8   9  10  11  12 
##   0   0   0   0   0 540 558 558 540   0   0   0 
## 
##   1   2   3   4   5   6   7   8   9  10  11  12 
##   0   0   0   0   0 540 558 558 540   0   0   0 
## 
##   1   2   3   4   5   6   7   8   9  10  11  12 
##   0   0   0   0   0 540 558 558 540   0   0   0

Creating Heat Wave variable

trshld95<-c()

for (i in 1:length(villes_s)){
  trshld95[i]<-quantile(villes_s[[i]]$tempmax, probs=c(0.975),na.rm=TRUE)
  villes_s[[i]]$heat_wave<-0
  villes_s[[i]]$heat_wave[which(villes_s[[i]]$tempmax > trshld95[i])]<- 1 
}

Number of Heat Wave for each city:

	Heat wave=0	Heat wave=1
bm	2143	53
bordeaux	2143	53
clermont	2142	54
grenoble	2141	55
lehavre	2142	54
lille	2143	53
lyon	2142	54
marseille	2142	54
montpellier	2145	51
nancy	2142	54
nantes	2141	55
nice	2142	54
paris	2144	52
rennes	2142	54
rouen	2143	53
strasbourg	2143	53
toulouse	2141	55

# creation of no2 variable of today and previous day
#function filter {stats}
for (i in 1:length(villes_s)){
villes_s[[i]]<-transform(villes_s[[i]], no2moy = as.vector(filter(no2,sides = 1, filter = rep(1, 2))/2), no2Lag1=Lag(no2,1),no2Lag2=Lag(no2,2))
}

Analysis

BOOTSTRAP METHOD

Code:

## create dataframe without missing values
for (i in 1:length(villes_s)){
  villes_s[[i]]<-na.omit(villes_s[[i]][, c("cardiaque","o3","Jours","Vacances","hol","heat_wave","no2moy")])
  #villes_s[[i]]<-villes_s[[i]][villes_s[[i]]$Dates>="2007-06-01",]
  villes_s[[i]]$time<-1:nrow(villes_s[[i]])
} 
   

    
  # start bootstrap
     
  nreps=1000 #number of bootstrap reps
  results<-matrix(NA,nrow = 17,ncol=10)
  colnames(results)<-c("city","NDE","NDE_2.5","NDE_97.5","NIE","NIE_2.5","NIE_97.5","PMM","PMM_2.5","PMM_97.5")
  
for (i in (1:length(villes_s))){
  boot.nde<- rep(NA,nreps)
  boot.nie <- rep(NA,nreps)
  boot.pmm <- rep(NA,nreps)
  for (rep in 1:nreps)
      
    {
      dat_resample <- sample(1:nrow(villes_s[[i]]),nrow(villes_s[[i]]),replace=T)
      dat_boot <- villes_s[[i]][dat_resample,]
        
      # outcome model - Poisson Regression: Mortality, HW and O3
      m.outcome<-glm(cardiaque~heat_wave+o3+ns(time,df=round(8*length(time)/365.25))+Jours+Vacances+hol,data=dat_boot,family=quasipoisson)
      
      # mediator model - multinomial linear regression
      m.mediator<-lm(o3~heat_wave+no2moy+ns(time,df=round(8*length(time)/365.25))+Jours+Vacances+hol,data=dat_boot) 
      
      # save coefficients
      theta1<-coef(m.outcome)[2]
      
      theta2<-coef(m.outcome)[3]
     
      beta1<-coef(m.mediator)[2]
      
      # estimate CDE and NIE
      
      boot.nde[rep] <- exp(theta1) #NDE
      
      boot.nie[rep] <-exp(theta2*beta1) #NIE 
      
      boot.pmm[rep] <-boot.nde[rep]*(boot.nie[rep]-1)/(boot.nde[rep]*boot.nie[rep]-1) #NIE 
                   
      
    } #end bootstrap

  results[i,1]<-cities[[i]]
  results[i,2]<-median(boot.nde, na.rm=T)
  results[i,3]<-quantile(boot.nde, 0.025, na.rm=T)
  results[i,4]<-quantile(boot.nde, 0.975, na.rm=T)
  results[i,5]<-median(boot.nie, na.rm=T)
  results[i,6]<-quantile(boot.nie, 0.025, na.rm=T)
  results[i,7]<-quantile(boot.nie, 0.975, na.rm=T)
  results[i,8]<-median(boot.pmm, na.rm=T)
  results[i,9]<-quantile(boot.pmm, 0.025, na.rm=T)
  results[i,10]<-quantile(boot.pmm, 0.975, na.rm=T)
  
     }
  
  # output CDE and NIE and confidence intervals

Results:

res<-data.frame(results)

for (i in (2:10)){
res[,i]<-as.numeric(as.character(res[,i]))
}

res[,2:10]<-round(res[,2:10],digits = 4)

kable(res)%>%kable_styling()%>%
  column_spec(1, bold = T, border_right = T)%>%
  column_spec(4, border_right = T)%>%
  column_spec(7, border_right = T)

city	NDE	NDE_2.5	NDE_97.5	NIE	NIE_2.5	NIE_97.5	PMM	PMM_2.5	PMM_97.5
bm	0.9912	0.8741	1.1250	0.9869	0.9303	1.0469	0.0538	-6.2033	5.5744
bordeaux	0.9241	0.8388	1.0095	1.0192	0.9881	1.0517	-0.2497	-2.2694	2.2138
clermont	1.0433	0.9089	1.1749	1.0046	0.9756	1.0346	0.0442	-2.6958	2.1599
grenoble	0.9862	0.8496	1.1334	1.0314	0.9994	1.0608	0.2455	-6.6852	8.3387
lehavre	0.9790	0.8525	1.1211	0.9965	0.9493	1.0434	-0.0080	-3.8968	3.8469
lille	0.9489	0.8496	1.0466	1.0084	0.9781	1.0407	-0.0901	-3.3829	3.5985
lyon	0.9727	0.8896	1.0612	0.9940	0.9738	1.0132	0.0761	-2.5442	2.5705
marseille	0.9810	0.8940	1.0765	1.0010	0.9958	1.0082	-0.0085	-0.8857	0.7113
montpellier	1.1683	1.0184	1.3126	1.0000	0.9964	1.0031	-0.0003	-0.0300	0.0322
nancy	0.8429	0.7383	0.9461	0.9976	0.9662	1.0263	0.0130	-0.1881	0.2582
nantes	1.0690	0.9392	1.1927	1.0040	0.9731	1.0346	0.0496	-1.0756	2.2856
nice	0.8918	0.7622	1.0247	0.9987	0.9850	1.0059	0.0103	-0.1279	0.2523
paris	0.9669	0.9079	1.0246	0.9965	0.9825	1.0089	0.0643	-1.5928	1.6282
rennes	0.8451	0.6707	1.0455	1.0133	0.9653	1.0666	-0.0543	-1.6832	1.7839
rouen	1.0148	0.9100	1.1320	0.9966	0.9543	1.0356	0.0053	-4.2066	5.7058
strasbourg	1.0342	0.9231	1.1387	0.9961	0.9643	1.0293	-0.0787	-4.5787	3.7572
toulouse	1.0323	0.9155	1.1389	0.9977	0.9759	1.0184	-0.0191	-2.1384	2.5741

Forestplot

Natural Direct Effect

Natural Indirect Effect

Heat wave and Ozone, Mediation analysis on cardiac hospitalization - 97.5 perc

Anna Alari

17/07/2020

Creating Database only with summer months

Creating Heat Wave variable

Analysis

BOOTSTRAP METHOD

Forestplot