Kathrine Bjerre Løppenthin provided the following data files for the project
list.files(path = "../data")## [1] "Anamnese_baseline_followup.xls"
## [2] "Baseline PSG data NY.xlsx"
## [3] "Baseline data.xls"
## [4] "Follow-up PSG data.xlsx"
## [5] "Follow-up data.xls"
## [6] "Intervention og kontrol.xlsx"
## [7] "Kopi af Indtast s\303\270vndagbog.xls"
## [8] "Variabeloversigt JR-Sleep.doc"
## [9] "data_full.txt"
## [10] "watt-max test_baseline_follow-up.xls"Reply to email on July 1st, 2018: data files on Polysomnography are still missing and will be provided after summer holiday. Data received from KBL on USB stick August 2nd, 2018 at KB.
data_rand_list <- read_excel(path = "../data/Intervention og kontrol.xlsx"
, sheet = 1)
data_rand_list <- dplyr::rename(data_rand_list, "ID_NUM" = "ID-nummer")Check levels
unique(data_rand_list$Randomisering)## [1] "kontrol" "Intervention" "Interventkion" "Kontrol"Relabel factor levels
data_rand_list <- mutate(data_rand_list, Randomisering = fct_collapse(Randomisering, control = c("kontrol", "Kontrol"), intervention = c("Intervention", "Interventkion")))Check new encoding of groups
unique(data_rand_list$Randomisering)## [1] control intervention
## Levels: intervention controlExcel file containing 9 sheets.
First (=anamnese) and last (=Søvndagbog) sheets are empty?
Use instead files Anamnese_baseline_followup.xls and Kopi af Indtast søvndagbog.xls (-see code below).
data_base_psqi <- read_excel(path = "../data/Baseline data.xls", skip = 1
, sheet = 2)
data_base_ess <- read_excel(path = "../data/Baseline data.xls", skip = 1
, sheet = 3)
data_base_bristol <- read_excel(path = "../data/Baseline data.xls", skip = 1
, sheet = 4)
data_base_fys_akt <- read_excel(path = "../data/Baseline data.xls", skip = 1
, sheet = 5)
data_base_dep <- read_excel(path = "../data/Baseline data.xls", skip = 1
, sheet = 6)
data_base_fys_funk <- read_excel(path = "../data/Baseline data.xls", skip = 1
, sheet = 7)
data_base_eq_5d <- read_excel(path = "../data/Baseline data.xls", skip = 1
, sheet = 8)Note that . and blanks have been used to indicate missing values …
data_follow_psqi <- read_excel(path = "../data/Follow-up data.xls", skip = 1, na = "."
, sheet = 1)
data_follow_ess <- read_excel(path = "../data/Follow-up data.xls", skip = 1, na = "."
, sheet = 2)
data_follow_bristol <- read_excel(path = "../data/Follow-up data.xls", skip = 1, na = "."
, sheet = 3)
data_follow_fys_akt <- read_excel(path = "../data/Follow-up data.xls", skip = 1, na = "."
, sheet = 4)
data_follow_dep <- read_excel(path = "../data/Follow-up data.xls", skip = 1, na = "."
, sheet = 5)
data_follow_fys_funk <- read_excel(path = "../data/Follow-up data.xls", skip = 1, na = "."
, sheet = 6)
data_follow_eq_5d <- read_excel(path = "../data/Follow-up data.xls", skip = 1, na = "."
, sheet = 7)Confirmed by reply to email on July 1st, 2018: Remove columns 21+22 for data_follow_psqi.
data_follow_psqi <- data_follow_psqi[, -c(21, 22)]Be careful to read in the following variables (encoded hour:min:sec)
R reads in these variables in Date format. Below we extract only the time part.
However we also make numerical version of relevant variables measuring the duration of an interval.
NB: Make KBL confirm that variable wake_time has the right unit. What does WASO really measure? Could it be that WASO in data corresponds to variable called Awake in Variabeloversigt JR-Sleep. Variable Søvnmønster/beskrivelse not to be used?
data_base_psg <- read_excel(path = "../data/Baseline PSG data NY.xlsx", sheet = 1, na = ".", skip = 1)
data_base_psg <- mutate(data_base_psg,
TST_num = hour(TST) + minute(TST)/60 + second(TST)/3600,
TST = format(TST, "%H:%M:%S"),
Bedtime = format(Bedtime,"%H:%M:%S"),
SO_num = hour(SO) + minute(SO)/60 + second(SO)/3600,
SO = format(SO, "%H:%M:%S"),
sleep_latency_num = hour(sleep_latency) + minute(sleep_latency)/60 + second(sleep_latency)/3600,
sleep_latency = format(sleep_latency, "%H:%M:%S"),
REM_latency_num = hour(REM_latency) + minute(REM_latency)/60 + second(REM_latency)/3600,
REM_latency = format(REM_latency, "%H:%M:%S"),
Wake_time_num = hour(Wake_time) + minute(Wake_time)/60 + second(Wake_time)/3600,
Wake_time = format(Wake_time, "%H:%M:%S")
)
data_base_psg <- dplyr::rename(data_base_psg
, "WASO" ="waso"
, "wake_time" = "Wake_time"
, "wake_time_num" = "Wake_time_num"
, "sleep_effi" = "Sleep_effi") %>% select(-contains("nster/beskr"))data_follow_psg <- read_excel(path = "../data/Follow-up PSG data.xlsx", sheet = 1, na = ".")
data_follow_psg <- mutate(data_follow_psg,
TST_num = hour(TST) + minute(TST)/60 + second(TST)/3600,
TST = format(TST, "%H:%M:%S"),
Bedtime = format(Bedtime,"%H:%M:%S"),
SO_num = hour(SO) + minute(SO)/60 + second(SO)/3600,
SO = format(SO, "%H:%M:%S"),
sleep_latency_num = hour(sleep_latency) + minute(sleep_latency)/60 + second(sleep_latency)/3600,
sleep_latency = format(sleep_latency, "%H:%M:%S"),
REM_latency_num = hour(REM_latency) + minute(REM_latency)/60 + second(REM_latency)/3600,
REM_latency = format(REM_latency, "%H:%M:%S"),
wake_time_num = hour(wake_time) + minute(wake_time)/60 + second(wake_time)/3600,
wake_time = format(wake_time, "%H:%M:%S")
)
data_follow_psg <- select(data_follow_psg, -contains("nster/beskri"))data_base_anamnese <- read_excel(path = "../data/Anamnese_baseline_followup.xls", sheet = 1, na = ".")
data_follow_anamnese <- read_excel(path = "../data/Anamnese_baseline_followup.xls", sheet = 2, na = ".")Renaming variables according to key provided by Kathrine (see file Variabeloversigt).
data_base_anamnese <- plyr::rename(data_base_anamnese
, replace = c("age" = "Age", "tend_joi" = "Tend_joi"
, "job" = "Job", "smoking" = "Smoking") )
data_follow_anamnese <- plyr::rename(data_follow_anamnese
, replace = c("tend_joi" = "Tend_joi"
, "Koffein" = "koffein") )
names(data_base_anamnese)## [1] "Baseline" "ID-NR" "Sex" "Age"
## [5] "Swol_joi" "Tend_joi" "DAS" "HGB"
## [9] "CRP" "Disease_dura" "Comobid" "Educa"
## [13] "Job" "Live_alone" "Smoking" "alcohol"
## [17] "koffein"names(data_follow_anamnese)## [1] "Follow-up" "ID-nr" "Swol_joi" "Tend_joi" "DAS" "HGB"
## [7] "CRP" "koffein"Where to find data?
Excel sheet labelled Søvndagbog fra files Baseline data.xls and Follow-up data.xls does not contain any data.
Not obvious that variables of file Kopi af Indtast søvndagbog.xls correspond to variables listed in Variabeloversigt JR-Sleep.doc for Polysomnografi.
Need to read in and tidy up data from Søvndagbøger
Consult scoringsmanual til PSQI.pdf
apply(data_base_psqi[, -1], 2, unique)## $SP2_Spg_1_BA
## [1] "23.30" "23.00" "24.00" "23.45" "22.30" "22.00" "21.30" "23.15"
## [9] "01.00" "02.00" "01.30" "22.15" "21.00"
##
## $SP2_Spg_2_BA
## [1] "30" "12.5" "45" "75" NA "22.5" "6.5" "5" "20" "25"
## [11] "60" "10" "15" "35" "12" "40"
##
## $SP2_Spg_3_BA
## [1] "07.00" "08.30" "07.45" "07.30" "06.15" "09.00" "06.30" "08.00"
## [9] "06.55" "06.00" "07.15" "05.30" "09.30" "06.10" "5.30" "6.30"
## [17] "7.00"
##
## $SP2_Spg_4_BA
## [1] "7" "6" "5.5" NA "6.5" "5" "7.5" "4.5"
## [9] "2" "3.5" "07.00"
##
## $SP2_Spg_5A_BA
## [1] " 1" " 2" NA " 0" " 3"
##
## $SP2_Spg_5B_BA
## [1] "3" "2" "1"
##
## $SP2_Spg_5C_BA
## [1] "3" "2" "1" "0"
##
## $SP2_Spg_5D_BA
## [1] "0" "1" "3" "2"
##
## $SP2_Spg_5E_BA
## [1] "2" "0" "1" "3"
##
## $SP2_Spg_5F_BA
## [1] "0" "1" "2"
##
## $SP2_Spg_5G_BA
## [1] "3" "1" "2" "0"
##
## $SP2_Spg_5H_BA
## [1] "0" "1" "3"
##
## $SP2_Spg_5I_BA
## [1] "1" "2" "0" "3"
##
## $SP2_Spg_5J_svar_BA
## [1] " 1" NA " 2" " 0" " 3"
##
## $SP2_Spg_5J_BA
## [1] "0" "1" "2" "3"
##
## $SP2_Spg_6_BA
## [1] "1" "0" "2" "3"
##
## $SP2_Spg_7_BA
## [1] "0" "1" "2" "3"
##
## $SP2_Spg_8_BA
## [1] "0" "1" "2" "3"
##
## $SP2_Spg_9_BA
## [1] "1" "3" "2" "0"apply(data_follow_psqi[, -1], 2, unique)## $SP2_Spg_1_FU
## [1] "23.00" "22.30" "22.45" "00.15" "23.30" "22.00" "21.30"
## [8] NA "00.00" "01.00" "0.875" "0.9375" "23"
##
## $SP2_Spg_2_FU
## [1] "5" "10" "30" "17.5" "7.5" "2" "15" "35" "40" NA
## [11] "105" "20" "50" "12.5" "45"
##
## $SP2_Spg_3_FU
## [1] "07.00" "08.00" "06.30" "07.15" "06.15" "06.20" "7" "07.30"
## [9] "06.45" "6" NA "08.15" "05.50" "05.00" "06.00" "60.00"
## [17] "6.15" "6.30" "8.30"
##
## $SP2_Spg_4_FU
## [1] "7" "8" "6" "9" "6.5" "5" "5.5" NA "3" "7.5" "3.5"
##
## $SP2_Spg_5A_FU
## [1] " 1" " 2" " 0" " 3" NA
##
## $SP2_Spg_5B_FU
## [1] " 1" " 2" " 3" NA " 0"
##
## $SP2_Spg_5C_FU
## [1] " 1" " 2" " 3" NA " 0"
##
## $SP2_Spg_5D_FU
## [1] " 2" " 0" " 1" " 3" NA
##
## $SP2_Spg_5E_FU
## [1] " 1" " 0" " 2" " 3" NA
##
## $SP2_Spg_5F_FU
## [1] " 2" " 0" " 1" NA
##
## $SP2_Spg_5G_FU
## [1] " 1" " 3" " 2" " 0" NA
##
## $SP2_Spg_5H_FU
## [1] " 1" " 0" " 2" NA
##
## $SP2_Spg_5I_FU
## [1] " 1" " 2" " 3" " 0" NA
##
## $SP2_Spg_5J_svar_FU
## [1] " 1" NA " 3" " 2" " 0"
##
## $SP2_Spg_5J_FU
## [1] " 1" " 0" " 2" NA " 3"
##
## $SP2_Spg_6_FU
## [1] " 0" " 1" " 3" NA
##
## $SP2_Spg_7_FU
## [1] " 0" " 1" " 2" NA " 3"
##
## $SP2_Spg_8_FU
## [1] " 1" " 0" " 2" NA " 3"
##
## $SP2_Spg_9_FU
## [1] " 3" " 2" " 0" " 1" NAChange 24:00 to 00:00 and correct some obvious errors
data_base_psqi$SP2_Spg_1_BA <- with(data_base_psqi, replace(SP2_Spg_1_BA, SP2_Spg_1_BA == "24.00", "00.00"))
data_base_psqi$SP2_Spg_1_BA## [1] "23.30" "23.00" "00.00" "23.45" "00.00" "23.00" "00.00" "22.30"
## [9] "22.30" "23.00" "22.30" "22.00" "21.30" "23.00" "23.30" "23.30"
## [17] "23.00" "23.30" "23.15" "01.00" "23.30" "22.30" "23.30" "22.30"
## [25] "00.00" "22.30" "02.00" "22.00" "01.30" "22.00" "22.15" "23.00"
## [33] "01.00" "00.00" "21.00" "22.30" "22.30" "00.00"Grundet inkonsistent brug af . eller : har der indsneget sig nogle fejl i indlæsning af data som rettes her (se Excel-ark)
data_follow_psqi$SP2_Spg_1_FU <- with(data_follow_psqi, replace(SP2_Spg_1_FU, SP2_Spg_1_FU == "23", "23.00"))
data_follow_psqi$SP2_Spg_1_FU <- with(data_follow_psqi, replace(SP2_Spg_1_FU, SP2_Spg_1_FU == "0.875", "21.00"))
data_follow_psqi$SP2_Spg_1_FU <- with(data_follow_psqi, replace(SP2_Spg_1_FU, SP2_Spg_1_FU == "0.9375", "22.30"))
data_follow_psqi$SP2_Spg_1_FU## [1] "23.00" "23.00" "22.30" "22.45" "00.15" "22.30" "23.30" "22.30"
## [9] "22.00" "23.30" "22.30" "21.30" "22.00" "22.30" "23.00" "23.30"
## [17] "22.30" "23.30" "23.30" NA "23.30" "22.30" "00.00" "22.30"
## [25] "23.30" "23.30" "01.00" "23.30" "00.00" "22.00" "22.00" "23.00"
## [33] NA NA "21.00" "22.30" "23.00" "23.00"data_base_psqi$SP2_Spg_3_BA <- with(data_base_psqi, replace(SP2_Spg_3_BA, SP2_Spg_3_BA =="5.30", "05.30"))
data_base_psqi$SP2_Spg_3_BA <- with(data_base_psqi, replace(SP2_Spg_3_BA, SP2_Spg_3_BA =="6.30", "06.30"))
data_base_psqi$SP2_Spg_3_BA <- with(data_base_psqi, replace(SP2_Spg_3_BA, SP2_Spg_3_BA =="7.00", "07.00"))
data_base_psqi$SP2_Spg_3_BA## [1] "07.00" "08.30" "07.00" "07.45" "07.30" "06.15" "06.15" "07.00"
## [9] "06.15" "07.30" "07.00" "09.00" "06.30" "06.30" "08.00" "06.55"
## [17] "06.00" "07.30" "07.15" "08.00" "08.00" "06.30" "07.30" "07.30"
## [25] "07.00" "07.15" "05.30" "06.15" "09.30" "06.00" "06.10" "07.00"
## [33] "07.00" "08.00" "05.30" "06.30" "07.00" "06.30"Reply by email on July 1st: confirmed that 60.00 be replaced by 06.00.
data_follow_psqi$SP2_Spg_3_FU <- with(data_follow_psqi, replace(SP2_Spg_3_FU, SP2_Spg_3_FU =="7", "07.00"))
data_follow_psqi$SP2_Spg_3_FU <- with(data_follow_psqi, replace(SP2_Spg_3_FU, SP2_Spg_3_FU =="6.15", "06.15"))
data_follow_psqi$SP2_Spg_3_FU <- with(data_follow_psqi, replace(SP2_Spg_3_FU, SP2_Spg_3_FU =="6.30", "06.30"))
data_follow_psqi$SP2_Spg_3_FU <- with(data_follow_psqi, replace(SP2_Spg_3_FU, SP2_Spg_3_FU =="8.30", "08.30"))
data_follow_psqi$SP2_Spg_3_FU <- with(data_follow_psqi, replace(SP2_Spg_3_FU, SP2_Spg_3_FU =="60.00", "06.00"))
data_follow_psqi$SP2_Spg_3_FU <- with(data_follow_psqi, replace(SP2_Spg_3_FU, SP2_Spg_3_FU =="6", "06.00"))
data_follow_psqi$SP2_Spg_3_FU## [1] "07.00" "08.00" "06.30" "07.00" "07.15" "06.15" "06.30" "07.00"
## [9] "06.20" "07.00" "07.00" "07.00" "06.20" "07.00" "07.30" "06.45"
## [17] "06.00" "07.30" "07.15" NA "08.15" "06.30" "07.00" "07.30"
## [25] "05.50" "08.00" "05.00" "06.30" "08.00" "06.00" "06.00" "06.00"
## [33] NA NA "05.50" "06.15" "06.30" "08.30"data_base_psqi$SP2_Spg_4_BA <- with(data_base_psqi, replace(SP2_Spg_4_BA, SP2_Spg_4_BA =="07.00", "7"))data_base_psqi <- mutate(data_base_psqi, SP2_Spg_1_time_BA = parse_time(gsub("\\.", ":", SP2_Spg_1_BA), format = "%H:%M"), SP2_Spg_3_time_BA = parse_time(gsub("\\.", ":", SP2_Spg_3_BA), format = "%H:%M"), psqi4_in_bed = SP2_Spg_3_time_BA - SP2_Spg_1_time_BA )
data_base_psqi$psqi4_in_bed## Time differences in secs
## [1] -59400 -52200 25200 -57600 27000 -60300 22500 -55800 -58500 -55800
## [11] -55800 -46800 -54000 -59400 -55800 -59700 -61200 -57600 -57600 25200
## [21] -55800 -57600 -57600 -54000 25200 -54900 12600 -56700 28800 -57600
## [31] -57900 -57600 21600 28800 -55800 -57600 -55800 23400data_follow_psqi <- mutate(data_follow_psqi, SP2_Spg_1_time_FU = parse_time(gsub("\\.", ":", SP2_Spg_1_FU), format = "%H:%M"), SP2_Spg_3_time_FU = parse_time(gsub("\\.", ":", SP2_Spg_3_FU), format = "%H:%M"), psqi4_in_bed = SP2_Spg_3_time_FU - SP2_Spg_1_time_FU )
data_follow_psqi$psqi4_in_bed## Time differences in secs
## [1] -57600 -54000 -57600 -56700 25200 -58500 -61200 -55800 -56400 -59400
## [11] -55800 -52200 -56400 -55800 -55800 -60300 -59400 -57600 -58500 NA
## [21] -54900 -57600 25200 -54000 -63600 -55800 14400 -61200 28800 -57600
## [31] -57600 -61200 NA NA -54600 -58500 -59400 -52200data_base_psqi <- mutate(data_base_psqi, psqi1 = SP2_Spg_6_BA)Check carefully conversion to numerical score ….
Confirmed by reply to email on July 1st, 2018 that 6 hours of sleep is converted to the value 2.
data_base_psqi$SP2_Spg_2_BA_num <- cut(as.numeric(data_base_psqi$SP2_Spg_2_BA), breaks = c(0, 15, 30, 60, Inf)
, labels = c(0, 1, 2, 3))
data_base_psqi <- mutate(data_base_psqi, psqi2_tmp = parse_integer(SP2_Spg_2_BA_num) + SP2_Spg_5A_BA)
data_base_psqi <- mutate(data_base_psqi, psqi2 = parse_integer(cut(psqi2_tmp, breaks = c(0, 2, 4, 6, Inf), labels = c(0, 1, 2, 3))))
with(data_base_psqi, table(psqi2, psqi2_tmp))## psqi2_tmp
## psqi2 0 1 2 3 4 5 6
## 0 0 2 10 0 0 0 0
## 1 0 0 0 1 10 0 0
## 2 0 0 0 0 0 7 1Check carefully conversion to numerical score ….
data_base_psqi <- mutate(data_base_psqi, psqi3 = parse_integer(cut(parse_double(data_base_psqi$SP2_Spg_4_BA), breaks = c(-Inf, 5, 6, 7, Inf)
, labels = c(3, 2, 1, 0))))
table(data_base_psqi$psqi3, data_base_psqi$SP2_Spg_4_BA )##
## 2 3.5 4.5 5 5.5 6 6.5 7 7.5
## 0 0 0 0 0 0 0 0 0 4
## 1 0 0 0 0 0 0 3 4 0
## 2 0 0 0 0 3 15 0 0 0
## 3 1 1 1 5 0 0 0 0 0Be very careful with conversion to time differences (bedtime after midnat causes some problems) …
data_base_psqi <- mutate(data_base_psqi, psqi4_in_bed = 3600*24*(psqi4_in_bed < 0) + psqi4_in_bed
, psqi4_sleep_eff = 100 * 3600 * parse_double(SP2_Spg_4_BA) / parse_double(psqi4_in_bed)
, psqi4 = parse_integer(cut(psqi4_sleep_eff, breaks = c(0, 65, 75, 85, Inf), right = FALSE, labels = c(3, 2, 1, 0))))
summary(data_base_psqi$psqi4_sleep_eff)## Min. 1st Qu. Median Mean 3rd Qu. Max. NA's
## 22.22 69.47 80.90 76.36 85.71 100.00 1Check carefully conversion …
choose_vars <- paste("SP2_Spg_5", LETTERS[2:10], "_BA", sep = "")
choose_vars## [1] "SP2_Spg_5B_BA" "SP2_Spg_5C_BA" "SP2_Spg_5D_BA" "SP2_Spg_5E_BA"
## [5] "SP2_Spg_5F_BA" "SP2_Spg_5G_BA" "SP2_Spg_5H_BA" "SP2_Spg_5I_BA"
## [9] "SP2_Spg_5J_BA"data_base_psqi$psqi5_tmp <- apply(select(data_base_psqi, choose_vars), 1, sum)
data_base_psqi <- mutate(data_base_psqi, psqi5 = parse_integer(cut(psqi5_tmp, breaks = c(-Inf, 0, 9, 18, 27, Inf)
, labels = c(-1, 0, 1, 2, 3))))
table(data_base_psqi$psqi5_tmp, data_base_psqi$psqi5)##
## 0 1 2
## 2 1 0 0
## 3 1 0 0
## 4 1 0 0
## 5 2 0 0
## 6 2 0 0
## 7 1 0 0
## 8 2 0 0
## 9 2 0 0
## 10 0 2 0
## 11 0 7 0
## 12 0 3 0
## 13 0 3 0
## 15 0 1 0
## 16 0 4 0
## 17 0 1 0
## 18 0 2 0
## 19 0 0 2
## 20 0 0 1data_base_psqi <- mutate(data_base_psqi, psqi6 = data_base_psqi$SP2_Spg_7_BA)data_base_psqi <- mutate(data_base_psqi, psqi7_tmp = SP2_Spg_8_BA + SP2_Spg_9_BA
, psqi7 = parse_integer(cut(psqi7_tmp, breaks = c(-Inf, 0, 2, 4, 6, Inf), labels = 0:4)))
with(data_base_psqi, table(psqi7, psqi7_tmp))## psqi7_tmp
## psqi7 0 1 2 3 4 5 6
## 0 2 0 0 0 0 0 0
## 1 0 7 7 0 0 0 0
## 2 0 0 0 5 6 0 0
## 3 0 0 0 0 0 10 1data_base_psqi$psqi_total <- apply(select(data_base_psqi, num_range(prefix = "psqi", range = 1:7)), 1, sum)data_follow_psqi <- mutate(data_follow_psqi, psqi1 = SP2_Spg_6_FU)Check carefully conversion to numerical score ….
data_follow_psqi$SP2_Spg_2_FU_num <- parse_integer(cut(as.numeric(data_follow_psqi$SP2_Spg_2_FU), breaks = c( 0, 15, 30, 60, Inf)
, labels = c(0, 1, 2, 3)))
data_follow_psqi <- mutate(data_follow_psqi, psqi2_tmp = SP2_Spg_2_FU_num + SP2_Spg_5A_FU)
data_follow_psqi <- mutate(data_follow_psqi, psqi2 = parse_integer(cut(psqi2_tmp, breaks = c(-Inf, 0, 2, 4, 6)
, labels = 0:3)))
data_follow_psqi$psqi2_tmp## [1] 1 1 3 2 3 2 1 1 1 3 5 4 1 1 5 2 3 0 0 NA 4 2 3
## [24] NA 1 2 5 0 2 1 5 0 NA NA 1 3 3 0data_follow_psqi$psqi2## [1] 1 1 2 1 2 1 1 1 1 2 3 2 1 1 3 1 2 0 0 NA 2 1 2
## [24] NA 1 1 3 0 1 1 3 0 NA NA 1 2 2 0with(data_follow_psqi, table(psqi2, psqi2_tmp))## psqi2_tmp
## psqi2 0 1 2 3 4 5
## 0 5 0 0 0 0 0
## 1 0 10 6 0 0 0
## 2 0 0 0 7 2 0
## 3 0 0 0 0 0 4Check carefully conversion to numerical score ….
data_follow_psqi <- mutate(data_follow_psqi, psqi3 = parse_integer(cut(parse_double(data_follow_psqi$SP2_Spg_4_FU), breaks = c(-Inf, 5, 6, 7, Inf)
, labels = c(3, 2, 1, 0))))
table(data_follow_psqi$psqi3, data_follow_psqi$SP2_Spg_4_FU )##
## 3 3.5 5 5.5 6 6.5 7 7.5 8 9
## 0 0 0 0 0 0 0 0 1 3 1
## 1 0 0 0 0 0 3 7 0 0 0
## 2 0 0 0 2 13 0 0 0 0 0
## 3 1 1 3 0 0 0 0 0 0 0NB: One subject 08-240713 sleeps more hours than he or she stays in bed resulting in sleep efficiency above 100. According to email from KBL on July 1st, 2018 no of hours sleept is change to 8.5 hours to obtain sleep efficiency of 100 %.
data_follow_psqi$SP2_Spg_4_FU <- replace(data_follow_psqi$SP2_Spg_4_FU, data_follow_psqi$ID_NUM == "08-240713", "8.5")Be very careful with conversion to time differences (bedtime after midnat causes some problems) …
data_follow_psqi <- mutate(data_follow_psqi, psqi4_in_bed = 3600*24*(psqi4_in_bed < 0) + psqi4_in_bed
, psqi4_sleep_eff = 100 * 3600 * parse_double(SP2_Spg_4_FU) / parse_double(psqi4_in_bed)
, psqi4 = parse_integer(cut(psqi4_sleep_eff, breaks = c(0, 65, 75, 85, Inf), right = FALSE, labels = c(3, 2, 1, 0))))
summary(data_follow_psqi$psqi4_sleep_eff)## Min. 1st Qu. Median Mean 3rd Qu. Max. NA's
## 39.62 75.00 80.00 79.48 85.71 100.00 3Check carefully conversion …
choose_vars <- paste("SP2_Spg_5", LETTERS[2:10], "_FU", sep = "")
choose_vars## [1] "SP2_Spg_5B_FU" "SP2_Spg_5C_FU" "SP2_Spg_5D_FU" "SP2_Spg_5E_FU"
## [5] "SP2_Spg_5F_FU" "SP2_Spg_5G_FU" "SP2_Spg_5H_FU" "SP2_Spg_5I_FU"
## [9] "SP2_Spg_5J_FU"data_follow_psqi$psqi5_tmp <- apply(select(data_follow_psqi, choose_vars), 1, sum)
data_follow_psqi <- mutate(data_follow_psqi, psqi5 = parse_integer(cut(psqi5_tmp, breaks = c(-Inf, 0, 9, 18, 27, Inf)
, labels = c(-1, 0, 1, 2, 3))))
table(data_follow_psqi$psqi5_tmp, data_follow_psqi$psqi5)##
## 0 1 2
## 3 2 0 0
## 4 2 0 0
## 5 2 0 0
## 6 1 0 0
## 7 4 0 0
## 9 2 0 0
## 10 0 3 0
## 11 0 3 0
## 12 0 2 0
## 13 0 5 0
## 15 0 4 0
## 17 0 2 0
## 18 0 1 0
## 19 0 0 1According to email fra KBL on July 1st, 2018 then question SP2_Spg_5J_svar just contains additional information and should not be used for computing the score.
data_follow_psqi <- mutate(data_follow_psqi, psqi6 = data_follow_psqi$SP2_Spg_7_FU)data_follow_psqi <- mutate(data_follow_psqi, psqi7_tmp = SP2_Spg_8_FU + SP2_Spg_9_FU
, psqi7 = parse_integer(cut(psqi7_tmp, breaks = c(-Inf, 0, 2, 4, 6, Inf), labels = 0:4)))
with(data_follow_psqi, table(psqi7, psqi7_tmp))## psqi7_tmp
## psqi7 0 1 2 3 4 5
## 0 5 0 0 0 0 0
## 1 0 5 8 0 0 0
## 2 0 0 0 2 6 0
## 3 0 0 0 0 0 9data_follow_psqi$psqi_total <- apply(select(data_follow_psqi, num_range(prefix = "psqi", range = 1:7)), 1, sum)select(data_follow_psqi, num_range("psqi", 1:7) )## # A tibble: 38 x 7
## psqi1 psqi2 psqi3 psqi4 psqi5 psqi6 psqi7
## <dbl> <int> <int> <int> <int> <dbl> <int>
## 1 0 1 1 0 1 0 2
## 2 1 1 0 0 0 0 2
## 3 3 2 2 1 1 0 1
## 4 0 1 2 2 1 0 1
## 5 0 2 2 0 1 1 3
## 6 1 1 1 0 0 0 2
## 7 1 1 2 0 1 1 3
## 8 1 1 0 0 1 0 2
## 9 0 1 1 1 0 0 1
## 10 0 2 2 1 2 0 1
## # ... with 28 more rowsHow to handle missing answers?
All items with take data values 0-3
apply(data_base_ess[, -1], 2, table)## $SP3_Spg_1_BA
##
## 0 1 2 3
## 9 10 12 7
##
## $SP3_Spg_2_BA
##
## 0 1 2 3
## 4 9 15 10
##
## $SP3_Spg_3_BA
##
## 0 1 2 3
## 16 16 5 1
##
## $SP3_Spg_4_BA
##
## 0 1 2 3
## 12 12 8 6
##
## $SP3_Spg_5_BA
##
## 0 1 2 3
## 3 3 13 19
##
## $SP3_Spg_6_BA
##
## 0 1 2
## 31 5 2
##
## $SP3_Spg_7_BA
##
## 0 1 2 3
## 14 16 7 1
##
## $SP3_Spg_8_BA
##
## 0 1
## 29 9apply(data_follow_ess[, -1], 2, table)## $SP3_Spg_1_FU
##
## 0 1 2 3
## 6 13 9 7
##
## $SP3_Spg_2_FU
##
## 0 1 2 3
## 6 8 13 8
##
## $SP3_Spg_3_FU
##
## 0 1 2
## 17 15 2
##
## $SP3_Spg_4_FU
##
## 0 1 2 3
## 13 8 9 5
##
## $SP3_Spg_5_FU
##
## 0 1 2 3
## 3 3 14 14
##
## $SP3_Spg_6_FU
##
## 0 1 2
## 27 7 1
##
## $SP3_Spg_7_FU
##
## 0 1 2 3
## 19 7 4 5
##
## $SP3_Spg_8_FU
##
## 0 1 2
## 27 7 1Count number of missing values
apply(data_base_ess[, 1 + 1:8], 2, function(z){sum(is.na(z))})## SP3_Spg_1_BA SP3_Spg_2_BA SP3_Spg_3_BA SP3_Spg_4_BA SP3_Spg_5_BA
## 0 0 0 0 0
## SP3_Spg_6_BA SP3_Spg_7_BA SP3_Spg_8_BA
## 0 0 0apply(data_follow_ess[, 1 + 1:8], 2, function(z){sum(is.na(z))})## SP3_Spg_1_FU SP3_Spg_2_FU SP3_Spg_3_FU SP3_Spg_4_FU SP3_Spg_5_FU
## 3 3 4 3 4
## SP3_Spg_6_FU SP3_Spg_7_FU SP3_Spg_8_FU
## 3 3 3data_base_ess$ess_total <- apply(data_base_ess[, 1 + 1:8], 1, sum)
data_follow_ess$ess_total <- apply(data_follow_ess[, 1 + 1:8], 1, sum)
summary(data_base_ess$ess_total)## Min. 1st Qu. Median Mean 3rd Qu. Max.
## 1.000 6.000 8.000 8.842 11.750 19.000summary(data_follow_ess$ess_total)## Min. 1st Qu. Median Mean 3rd Qu. Max. NA's
## 1.000 5.000 7.000 8.212 11.000 18.000 5Range of SP4_Spg_1_BA should be 0-10
table(data_base_bristol[, 1 + 1])##
## 0 1 2 3 4 5 6 7 8 9
## 1 2 2 8 1 2 4 6 8 3Range of SP4_Spg_2_BA should be 0-7
table(data_base_bristol[, 1 + 2])##
## 0 1 2 3 4 5 6 7
## 1 4 4 3 3 3 1 19Range of SP4_Spg_3_BA should be 0-2
table(data_base_bristol[, 1 + 3])##
## 0 1 2
## 15 19 4Range of SP4_Spg_5_BA- SP4_Spg_20_BA should be 0-3
raw_tab <- apply(data_base_bristol[, 1 + 4:20], 2, table)
raw_tab## $SP4_Spg_4_BA
##
## 0 1 2 3
## 5 17 13 3
##
## $SP4_Spg_5_BA
##
## 0 1 2
## 26 10 2
##
## $SP4_Spg_6_BA
##
## 0 1 2
## 31 6 1
##
## $SP4_Spg_7_BA
##
## 0 1 2 3
## 10 16 11 1
##
## $SP4_Spg_8_BA
##
## 0 1 2 3
## 10 17 10 1
##
## $SP4_Spg_9_BA
##
## 0 1 2 3
## 11 19 6 2
##
## $SP4_Spg_10_BA
##
## 0 1 2
## 10 20 8
##
## $SP4_Spg_11_BA
##
## 0 1 2
## 24 10 4
##
## $SP4_Spg_12_BA
##
## 0 1 2 3
## 10 13 12 3
##
## $SP4_Spg_13_BA
##
## 0 1 2 3
## 10 20 7 1
##
## $SP4_Spg_14_BA
##
## 0 1 2
## 9 24 5
##
## $SP4_Spg_15_BA
##
## 0 1 2 3
## 10 19 7 2
##
## $SP4_Spg_16_BA
##
## 0 1 2
## 16 21 1
##
## $SP4_Spg_17_BA
##
## 0 1 2
## 13 20 5
##
## $SP4_Spg_18_BA
##
## 0 1 2
## 28 5 5
##
## $SP4_Spg_19_BA
##
## 0 1 2 3
## 14 13 8 3
##
## $SP4_Spg_20_BA
##
## 0 1 2 3
## 18 15 4 1Range of SP4_Spg_1_BA should be 0-10
table(data_follow_bristol[, 1 + 1])##
## 0 1 2 3 4 5 6 7 8 9
## 1 1 4 3 4 3 6 7 5 1Range of SP4_Spg_2_BA should be 0-7
table(data_follow_bristol[, 1 + 2])##
## 0 1 2 3 4 5 7
## 1 5 4 5 4 3 13Range of SP4_Spg_3_BA should be 0-2
table(data_follow_bristol[, 1 + 3])##
## 0 1 2
## 10 21 4Range of SP4_Spg_5_BA- SP4_Spg_20_BA should be 0-3
raw_tab <- apply(data_follow_bristol[, 1 + 4:20], 2, table)
raw_tab## $SP4_Spg_4_FU
##
## 0 1 2 3
## 4 18 10 2
##
## $SP4_Spg_5_FU
##
## 0 1
## 24 11
##
## $SP4_Spg_6_FU
##
## 0 1
## 29 6
##
## $SP4_Spg_7_FU
##
## 0 1 2
## 10 18 7
##
## $SP4_Spg_8_FU
##
## 0 1 2 3
## 15 17 2 1
##
## $SP4_Spg_9_FU
##
## 0 1 2
## 16 17 2
##
## $SP4_Spg_10_FU
##
## 0 1 2
## 20 9 6
##
## $SP4_Spg_11_FU
##
## 0 1
## 24 11
##
## $SP4_Spg_12_FU
##
## 0 1 2
## 7 16 12
##
## $SP4_Spg_13_FU
##
## 0 1 2 3
## 9 19 5 2
##
## $SP4_Spg_14_FU
##
## 0 1 2 3
## 16 15 3 1
##
## $SP4_Spg_15_FU
##
## 0 1 2 3
## 12 16 5 2
##
## $SP4_Spg_16_FU
##
## 0 1
## 18 17
##
## $SP4_Spg_17_FU
##
## 0 1 2
## 16 15 4
##
## $SP4_Spg_18_FU
##
## 0 1
## 28 7
##
## $SP4_Spg_19_FU
##
## 0 1 2
## 15 16 4
##
## $SP4_Spg_20_FU
##
## 0 1 2 3
## 16 15 3 1Be careful how to handle NAs …
Tabulate number of NAs
get_psqi_phys <- function(x){
m_score <- c(10, 7, 2, 3)
na_pos <- is.na(x)
res <- sum(x, na.rm = T)/sum(m_score[!na_pos]) * sum(m_score)
if((sum(na_pos) > 1)||(any(is.na(x[1:2])))){res <- NA}
return(res)
}
data_base_bristol$BRAF_phys <- apply(data_base_bristol[2:5], 1, get_psqi_phys)
get_psqi_other <- function(x){
m_score <- rep(3, length(x))
na_pos <- is.na(x)
res <- sum(x, na.rm = T)/sum(m_score[!na_pos]) * sum(m_score)
if((sum(na_pos) > 1)){res <- NA}
return(res)
}
data_base_bristol$BRAF_living <- apply(data_base_bristol[, 1 + 5:11], 1, get_psqi_other)
data_base_bristol$BRAF_cog <- apply(data_base_bristol[, 1 + 12:16], 1, get_psqi_other)
data_base_bristol$BRAF_emo <- apply(data_base_bristol[, 1 + 17:20], 1, get_psqi_other)
data_base_bristol$BRAF_total <- apply(dplyr::select(data_base_bristol, dplyr::contains("BRAF")), 1, sum)
data_follow_bristol$BRAF_phys <- apply(data_follow_bristol[2:5], 1, get_psqi_phys)
get_psqi_other <- function(x){
m_score <- rep(3, length(x))
na_pos <- is.na(x)
res <- sum(x, na.rm = T)/sum(m_score[!na_pos]) * sum(m_score)
if((sum(na_pos) > 1)){res <- NA}
return(res)
}
data_follow_bristol$BRAF_living <- apply(data_follow_bristol[, 1 + 5:11], 1, get_psqi_other)
data_follow_bristol$BRAF_cog <- apply(data_follow_bristol[, 1 + 12:16], 1, get_psqi_other)
data_follow_bristol$BRAF_emo <- apply(data_follow_bristol[, 1 + 17:20], 1, get_psqi_other)
data_follow_bristol$BRAF_total <- apply(dplyr::select(data_follow_bristol, dplyr::contains("BRAF")), 1, sum)summary(data_base_bristol$BRAF_phys)## Min. 1st Qu. Median Mean 3rd Qu. Max. NA's
## 0.00 9.00 14.00 12.54 17.00 20.00 1summary(data_base_bristol$BRAF_living)## Min. 1st Qu. Median Mean 3rd Qu. Max.
## 0.000 2.250 4.500 5.105 7.000 13.000summary(data_base_bristol$BRAF_cog)## Min. 1st Qu. Median Mean 3rd Qu. Max.
## 0.000 2.000 5.000 4.711 6.750 10.000summary(data_base_bristol$BRAF_emo)## Min. 1st Qu. Median Mean 3rd Qu. Max.
## 0.000 1.000 2.500 2.868 5.000 8.000summary(data_base_bristol$BRAF_total)## Min. 1st Qu. Median Mean 3rd Qu. Max. NA's
## 2.00 17.00 25.00 25.35 37.00 49.00 1summary(data_follow_bristol$BRAF_phys)## Min. 1st Qu. Median Mean 3rd Qu. Max. NA's
## 1.00 7.50 13.00 11.67 16.00 21.00 3summary(data_follow_bristol$BRAF_living)## Min. 1st Qu. Median Mean 3rd Qu. Max. NA's
## 0.0 1.0 4.0 3.6 5.0 10.0 3summary(data_follow_bristol$BRAF_cog)## Min. 1st Qu. Median Mean 3rd Qu. Max. NA's
## 0.000 1.500 4.000 4.229 6.000 11.000 3summary(data_follow_bristol$BRAF_emo)## Min. 1st Qu. Median Mean 3rd Qu. Max. NA's
## 0.0 0.0 2.0 2.2 4.0 6.0 3summary(data_follow_bristol$BRAF_total)## Min. 1st Qu. Median Mean 3rd Qu. Max. NA's
## 3.00 10.00 23.00 21.70 29.50 41.53 3Remove all comments from baseline questionnaire (col 2 -)
data_base_fys_akt <- data_base_fys_akt[, 1:2]Check raw data values
table(data_base_fys_akt$SP5_Spg_1_BA)##
## 1 2 3
## 6 21 11table(data_follow_fys_akt$SP5_Spg_1_FU)##
## 1 2 3 4
## 3 16 13 3Range should be 0-3
apply(data_base_dep[, -1], 2, function(z){table(z)})## $SP6_Spg_0_BA
## z
## 0 1 2
## 24 12 1
##
## $SP6_Spg_1_BA
## z
## 0 1 2
## 24 11 3
##
## $SP6_Spg_2_BA
## z
## 0 1
## 23 15
##
## $SP6_Spg_3_BA
## z
## 0 1 2 3
## 5 6 10 17
##
## $SP6_Spg_5_BA
## z
## 0 1 2 3
## 15 16 5 2
##
## $SP6_Spg_6_BA
## z
## 0 1
## 27 11
##
## $SP6_Spg_7_BA
## z
## 0 1 2
## 17 19 2
##
## $SP6_Spg_8_BA
## z
## 0 1 2 3
## 2 10 12 14
##
## $SP6_Spg_9_BA
## z
## 0 1 2
## 36 1 1
##
## $SP6_Spg_00_BA
## z
## 0 1 2
## 33 4 1
##
## $SP6_Spg_00_BA__1
## z
## 0 1 2 3
## 6 17 9 6
##
## $SP6_Spg_01_BA
## z
## 0 1 2 3
## 6 10 11 10
##
## $SP6_Spg_02_BA
## z
## 0 1 2
## 24 11 3
##
## $SP6_Spg_03_BA
## z
## 0 1 2 3
## 27 8 1 1
##
## $SP6_Spg_05_BA
## z
## 0 1 3
## 35 2 1
##
## $SP6_Spg_06_BA
## z
## 0 1 2 3
## 5 6 11 16
##
## $SP6_Spg_07_BA
## z
## 0 1
## 32 6
##
## $SP6_Spg_08_BA
## z
## 0 1
## 20 17
##
## $SP6_Spg_09_BA
## z
## 0 1 2
## 33 4 1
##
## $SP6_Spg_10_BA
## z
## 0 1 2
## 16 14 8Range should be 0-3
apply(data_follow_dep[, -1], 2, table)## $SP6_Spg_0_FU
##
## 0 1 2
## 23 9 2
##
## $SP6_Spg_1_FU
##
## 0 1 2
## 22 7 4
##
## $SP6_Spg_2_FU
##
## 0 1 2
## 23 7 4
##
## $SP6_Spg_3_FU
##
## 0 1 2 3
## 10 5 6 14
##
## $SP6_Spg_5_FU
##
## 0 1 2 3
## 17 14 3 1
##
## $SP6_Spg_6_FU
##
## 0 1 2 3
## 18 14 2 1
##
## $SP6_Spg_7_FU
##
## 0 1 2 3
## 17 12 3 2
##
## $SP6_Spg_8_FU
##
## 0 1 2 3
## 5 7 12 11
##
## $SP6_Spg_9_FU
##
## 0 1 2
## 33 1 1
##
## $SP6_Spg_00_FU
##
## 0 1 2 3
## 30 3 1 1
##
## $SP6_Spg_00_FU__1
##
## 0 1 2 3
## 11 14 5 4
##
## $SP6_Spg_01_FU
##
## 0 1 2 3
## 4 8 14 9
##
## $SP6_Spg_02_FU
##
## 0 1 2 3
## 26 4 4 1
##
## $SP6_Spg_03_FU
##
## 0 1 2 3
## 23 10 1 1
##
## $SP6_Spg_05_FU
##
## 0 1 2 3
## 27 5 1 1
##
## $SP6_Spg_06_FU
##
## 0 1 2 3
## 3 5 12 15
##
## $SP6_Spg_07_FU
##
## 0 1 2
## 27 7 1
##
## $SP6_Spg_08_FU
##
## 0 1 2
## 21 12 2
##
## $SP6_Spg_09_FU
##
## 0 1
## 30 5
##
## $SP6_Spg_10_FU
##
## 0 1 2 3
## 16 13 4 1data_base_dep$CES_D_total <- apply(data_base_dep[, -1], 1, sum)
data_follow_dep$CES_D_total <- apply(data_follow_dep[, -1], 1, sum)KBL has not yet confirmed if conversion of positively / negatively worded items to numbers 0,1,2,3 is correct. Check up on this and change if necessary.
Ignore scoring for aids/devices so far i.e. consider only items 1-20
names(data_base_fys_funk)## [1] "ID_NUM" "SP7_Spg_1_BA"
## [3] "SP7_Spg_2_BA" "SP7_Spg_3_BA"
## [5] "SP7_Spg_4_BA" "SP7_Spg_5_BA"
## [7] "SP7_Spg_6_BA" "SP7_Spg_7_BA"
## [9] "SP7_Spg_8_BA" "SP7_Spg_9_BA"
## [11] "SP7_Spg_10_BA" "SP7_Spg_11_BA"
## [13] "SP7_Spg_12_BA" "SP7_Spg_13_BA"
## [15] "SP7_Spg_14_BA" "SP7_Spg_15_BA"
## [17] "SP7_Spg_16_BA" "SP7_Spg_17_BA"
## [19] "SP7_Spg_18_BA" "SP7_Spg_19_BA"
## [21] "SP7_Spg_20_BA" "Sp7_stok_BA"
## [23] "Sp7_dressing_BA" "Sp7_krykstok_BA"
## [25] "Sp7_specielstol_BA" "Sp7_gangstativ_BA"
## [27] "Sp7_toiletsaede_BA" "Sp7_korestol_BA"
## [29] "Sp7_badestol_BA" "Sp7_kokken_BA"
## [31] "Sp7_badevarelse_BA" "Sp7_spiseredskaber_BA"
## [33] "Sp7_langskaftede_BA" "Sp7_andre_BA"
## [35] "Sp7_ikke_hjalpemidler_BA" "Sp7_spg_21_BA"
## [37] "Sp7_spg_22_BA" "Sp7_spg_23_BA"Check raw data values (allowed range: 0 - 3)
apply(data_base_fys_funk[, 1 + 1:20], 2, table)## $SP7_Spg_1_BA
##
## 0 1
## 28 10
##
## $SP7_Spg_2_BA
##
## 0 1
## 29 8
##
## $SP7_Spg_3_BA
##
## 0 1
## 29 9
##
## $SP7_Spg_4_BA
##
## 0 1
## 27 11
##
## $SP7_Spg_5_BA
##
## 0 1 2
## 24 10 4
##
## $SP7_Spg_6_BA
##
## 0 1 2
## 33 4 1
##
## $SP7_Spg_7_BA
##
## 0 1 2 3 12
## 13 14 9 1 1
##
## $SP7_Spg_8_BA
##
## 0 1
## 32 6
##
## $SP7_Spg_9_BA
##
## 0 1
## 29 9
##
## $SP7_Spg_10_BA
##
## 0 1
## 26 12
##
## $SP7_Spg_11_BA
##
## 0 1 2 3
## 24 2 2 5
##
## $SP7_Spg_12_BA
##
## 0 1
## 31 7
##
## $SP7_Spg_13_BA
##
## 0 1 2 3
## 12 18 5 3
##
## $SP7_Spg_14_BA
##
## 0 1
## 24 14
##
## $SP7_Spg_15_BA
##
## 0 1
## 29 9
##
## $SP7_Spg_16_BA
##
## 0 1 2 3
## 15 18 4 1
##
## $SP7_Spg_17_BA
##
## 0 1 2
## 27 10 1
##
## $SP7_Spg_18_BA
##
## 0 1 2
## 22 12 3
##
## $SP7_Spg_19_BA
##
## 0 1 2
## 26 11 1
##
## $SP7_Spg_20_BA
##
## 0 1 2 3
## 15 12 10 1Correct / change invalid entry 12 in item 7? According to email fra KBL on July 1st, 2018 value 12 should be replaced by 1.
data_base_fys_funk$SP7_Spg_7_BA <- replace(data_base_fys_funk$SP7_Spg_7_BA, data_base_fys_funk$SP7_Spg_7_BA == 12, 1)Confirm method for computing alternative HAQ-DI score
Reply by KBL on July 1st, 2018: AT may decide. According to scoring manual alternative HAQ-DI is the average of each category score given that at least six categories has a score. However, it is not obvious what is meant be the score for a category since each category consists of several sub-categories. Here we use the procedure indicated in the manual under Standard HAQ-DI where category score is the highest sub-category score from each category.
data_base_fys_funk$haq1 <- apply(data_base_fys_funk[, 1 + 1:2], 1, max, na.rm = T)
data_base_fys_funk$haq2 <- apply(data_base_fys_funk[, 1 + 3:4], 1, max, na.rm = T)
data_base_fys_funk$haq3 <- apply(data_base_fys_funk[, 1 + 5:7], 1, max, na.rm = T)
data_base_fys_funk$haq4 <- apply(data_base_fys_funk[, 1 + 8:9], 1, max, na.rm = T)
data_base_fys_funk$haq5 <- apply(data_base_fys_funk[, 1 + 10:12], 1, max, na.rm = T)
data_base_fys_funk$haq6 <- apply(data_base_fys_funk[, 1 + 13:14], 1, max, na.rm = T)
data_base_fys_funk$haq7 <- apply(data_base_fys_funk[, 1 + 15:17], 1, max, na.rm = T)
data_base_fys_funk$haq8 <- apply(data_base_fys_funk[, 1 + 18:20], 1, max, na.rm = T)
data_base_fys_funk$haq_NA <- apply(select(data_base_fys_funk, num_range(prefix = "haq", range = 1:8)), 1, function(z){sum(!is.na(z))})
data_base_fys_funk$haq_total <- apply(select(data_base_fys_funk, num_range(prefix = "haq", range = 1:8)), 1, mean)
data_base_fys_funk$haq_total <- replace(data_base_fys_funk$haq_total, data_base_fys_funk$haq_NA < 6, NA)Check scores for each of eight categories
apply(select(data_base_fys_funk, num_range(prefix = "haq", range = 1:8)), 2, table)## $haq1
##
## 0 1
## 27 11
##
## $haq2
##
## 0 1
## 25 13
##
## $haq3
##
## 0 1 2 3
## 13 14 10 1
##
## $haq4
##
## 0 1
## 29 9
##
## $haq5
##
## 0 1 2 3
## 23 8 2 5
##
## $haq6
##
## 0 1 2 3
## 11 19 5 3
##
## $haq7
##
## 0 1 2 3
## 13 20 4 1
##
## $haq8
##
## 0 1 2 3
## 13 14 10 1Check raw data values (allowed range: 0 - 3)
apply(data_follow_fys_funk[, 1 + 1:20], 2, table)## $SP7_Spg_1_FU
##
## 0 1
## 23 12
##
## $SP7_Spg_2_FU
##
## 0 1
## 28 7
##
## $SP7_Spg_3_FU
##
## 0 1
## 28 7
##
## $SP7_Spg_4_FU
##
## 0 1
## 27 8
##
## $SP7_Spg_5_FU
##
## 0 1 2
## 22 10 3
##
## $SP7_Spg_6_FU
##
## 0 1 2
## 28 6 1
##
## $SP7_Spg_7_FU
##
## 0 1 2 3
## 13 18 3 1
##
## $SP7_Spg_8_FU
##
## 0 1
## 31 4
##
## $SP7_Spg_9_FU
##
## 0 1
## 29 6
##
## $SP7_Spg_10_FU
##
## 0 1
## 24 11
##
## $SP7_Spg_11_FU
##
## 0 1 2 3
## 22 4 1 5
##
## $SP7_Spg_12_FU
##
## 0 1
## 29 6
##
## $SP7_Spg_13_FU
##
## 0 1 2 3
## 12 19 2 2
##
## $SP7_Spg_14_FU
##
## 0 1
## 24 11
##
## $SP7_Spg_15_FU
##
## 0 1 2
## 28 6 1
##
## $SP7_Spg_16_FU
##
## 0 1 2
## 16 16 3
##
## $SP7_Spg_17_FU
##
## 0 1 2
## 27 7 1
##
## $SP7_Spg_18_FU
##
## 0 1 2
## 21 9 5
##
## $SP7_Spg_19_FU
##
## 0 1 2
## 26 8 1
##
## $SP7_Spg_20_FU
##
## 0 1 2 3
## 14 13 7 1Confirm method for computing alternative HAQ-DI score
See comment about for baseline data.
Technical note:
With na.rm = T then max function returns -Inf for an empty vector. A user modified version returning NA for an empty vector is used here.
my_max <- function(x, ...){
if(all(is.na(x))){
return(NA)
}
else{
return(max(x, ...))
}
}
data_follow_fys_funk$haq1 <- apply(data_follow_fys_funk[, 1 + 1:2], 1, my_max, na.rm = T)
data_follow_fys_funk$haq2 <- apply(data_follow_fys_funk[, 1 + 3:4], 1, my_max, na.rm = T)
data_follow_fys_funk$haq3 <- apply(data_follow_fys_funk[, 1 + 5:7], 1, my_max, na.rm = T)
data_follow_fys_funk$haq4 <- apply(data_follow_fys_funk[, 1 + 8:9], 1, my_max, na.rm = T)
data_follow_fys_funk$haq5 <- apply(data_follow_fys_funk[, 1 + 10:12], 1, my_max, na.rm = T)
data_follow_fys_funk$haq6 <- apply(data_follow_fys_funk[, 1 + 13:14], 1, my_max, na.rm = T)
data_follow_fys_funk$haq7 <- apply(data_follow_fys_funk[, 1 + 15:17], 1, my_max, na.rm = T)
data_follow_fys_funk$haq8 <- apply(data_follow_fys_funk[, 1 + 18:20], 1, my_max, na.rm = T)
data_follow_fys_funk$haq_NA <- apply(select(data_follow_fys_funk, num_range(prefix = "haq", range = 1:8)), 1, function(z){sum(!is.na(z))})
data_follow_fys_funk$haq_total <- apply(select(data_follow_fys_funk, num_range(prefix = "haq", range = 1:8)), 1, mean)
data_follow_fys_funk$haq_total <- replace(data_follow_fys_funk$haq_total, data_follow_fys_funk$haq_NA < 6, NA)Check scores for each of eight categories
apply(select(data_follow_fys_funk, num_range(prefix = "haq", range = 1:8)), 2, table)## $haq1
##
## 0 1
## 23 12
##
## $haq2
##
## 0 1
## 26 9
##
## $haq3
##
## 0 1 2 3
## 11 18 5 1
##
## $haq4
##
## 0 1
## 29 6
##
## $haq5
##
## 0 1 2 3
## 19 10 1 5
##
## $haq6
##
## 0 1 2 3
## 11 20 2 2
##
## $haq7
##
## 0 1 2
## 16 15 4
##
## $haq8
##
## 0 1 2 3
## 14 13 7 1Check raw data values (items 1-5)
apply(data_base_eq_5d[, -c(1, 7)], 2, table)## $SP8_Spg_1_BA
##
## 1 2 3
## 24 12 2
##
## $SP8_Spg_2_BA
##
## 1 2
## 27 11
##
## $SP8_Spg_3_BA
##
## 1 2 3
## 16 16 6
##
## $SP8_Spg_4_BA
##
## 1 2 3 4
## 6 20 11 1
##
## $SP8_Spg_5_BA
##
## 1 2
## 29 9apply(data_follow_eq_5d[, -c(1, 7)], 2, table)## $SP8_Spg_1_FU
##
## 1 2 3
## 27 4 4
##
## $SP8_Spg_2_FU
##
## 1 2 3
## 26 6 3
##
## $SP8_Spg_3_FU
##
## 1 2 3 4
## 18 10 5 2
##
## $SP8_Spg_4_FU
##
## 1 2 3 4
## 6 17 11 1
##
## $SP8_Spg_5_FU
##
## 1 2 3
## 23 11 1Check raw data values (items 6)
summary(data_base_eq_5d$SP8_Spg_6_BA)## Min. 1st Qu. Median Mean 3rd Qu. Max.
## 30.00 56.25 75.00 71.42 83.75 100.00summary(data_follow_eq_5d$SP8_Spg_6_FU)## Min. 1st Qu. Median Mean 3rd Qu. Max. NA's
## 20.00 60.00 75.00 69.71 86.50 99.00 3Overvej hvilken version af join der skal benyttes for at sikre mod slåfejl i id nummer
The following data sets are combined (ID_NUM used as key … remember to rename ID variable if necessary).
Check that the correct number of variables are selected from each data set:
data_base <- select(data_base_anamnese, - Baseline) %>% mutate(time = "baseline") %>% rename("ID_NUM" = "ID-NR")
tmp_base <- select(data_base_psqi, c(1, num_range(prefix = "psqi", range = 1:7), matches("psqi_total"))) %>% mutate(time = "baseline")
data_base <- left_join(data_base, tmp_base)## Joining, by = c("ID_NUM", "time")tmp_base <- select(data_base_ess, ID_NUM, ess_total) %>% mutate(time = "baseline")
data_base <- left_join(data_base, tmp_base)## Joining, by = c("ID_NUM", "time")tmp_base <- select(data_base_bristol, ID_NUM, contains("BRAF_")) %>% mutate(time = "baseline")
data_base <- left_join(data_base, tmp_base)## Joining, by = c("ID_NUM", "time")tmp_base <- data_base_fys_akt %>% rename(fys_akt = "SP5_Spg_1_BA")
data_base <- left_join(data_base, tmp_base)## Joining, by = "ID_NUM"tmp_base <- select(data_base_dep, ID_NUM, CES_D_total) %>% mutate(time = "baseline")
data_base <- left_join(data_base, tmp_base)## Joining, by = c("ID_NUM", "time")tmp_base <- select(data_base_fys_funk, c(1, num_range(prefix = "haq", range = 1:7), matches("haq_total")))
data_base <- left_join(data_base, tmp_base)## Joining, by = "ID_NUM"data_base <- left_join(data_base, data_base_eq_5d)## Joining, by = "ID_NUM"data_base <- left_join(data_base, data_base_psg)## Joining, by = "ID_NUM"Display variables of cleaned data set
dim(data_base)## [1] 38 73names(data_base)## [1] "ID_NUM" "Sex" "Age"
## [4] "Swol_joi" "Tend_joi" "DAS"
## [7] "HGB" "CRP" "Disease_dura"
## [10] "Comobid" "Educa" "Job"
## [13] "Live_alone" "Smoking" "alcohol"
## [16] "koffein" "time" "psqi1"
## [19] "psqi2" "psqi3" "psqi4"
## [22] "psqi5" "psqi6" "psqi7"
## [25] "psqi_total" "ess_total" "BRAF_phys"
## [28] "BRAF_living" "BRAF_cog" "BRAF_emo"
## [31] "BRAF_total" "fys_akt" "CES_D_total"
## [34] "haq1" "haq2" "haq3"
## [37] "haq4" "haq5" "haq6"
## [40] "haq7" "haq_total" "SP8_Spg_1_BA"
## [43] "SP8_Spg_2_BA" "SP8_Spg_3_BA" "SP8_Spg_4_BA"
## [46] "SP8_Spg_5_BA" "SP8_Spg_6_BA" "TST"
## [49] "Bedtime" "sleep_effi" "SO"
## [52] "sleep_latency" "REM_latency" "WASO"
## [55] "Arous" "wake_time" "N1"
## [58] "N2" "N3" "REM"
## [61] "AHI" "LM" "PLM"
## [64] "OSA" "Mix_apnea" "Cen_apnea"
## [67] "Hyp_nea" "Desat" "TST_num"
## [70] "SO_num" "sleep_latency_num" "REM_latency_num"
## [73] "wake_time_num"Note that sociodemographic variables are not collected twice implying that anamnese data contains fewer variables at follow up. Make sure to fill in these variables before merging data from baseline and follow up (this is done below).
data_follow <- select(data_follow_anamnese, -contains("Follow-up")) %>% mutate(time = "follow_up") %>% rename("ID_NUM" = "ID-nr")
data_follow <- left_join(data_follow, select(data_base, ID_NUM, Sex, Age, Disease_dura, Comobid, Educa, Job, Live_alone, Smoking, alcohol))## Joining, by = "ID_NUM"tmp_follow <- select(data_follow_psqi, c(1, num_range(prefix = "psqi", range = 1:7), matches("psqi_total"))) %>% mutate(time = "follow_up")
data_follow <- left_join(data_follow, tmp_follow)## Joining, by = c("ID_NUM", "time")tmp_follow <- select(data_follow_ess, ID_NUM, ess_total) %>% mutate(time = "follow_up")
data_follow <- left_join(data_follow, tmp_follow)## Joining, by = c("ID_NUM", "time")tmp_follow <- select(data_follow_bristol, ID_NUM, contains("BRAF_")) %>% mutate(time = "follow_up")
data_follow <- left_join(data_follow, tmp_follow)## Joining, by = c("ID_NUM", "time")tmp_follow <- data_follow_fys_akt %>% rename(fys_akt = "SP5_Spg_1_FU")
data_follow <- left_join(data_follow, tmp_follow)## Joining, by = "ID_NUM"tmp_follow <- select(data_follow_dep, ID_NUM, CES_D_total) %>% mutate(time = "follow_up")
data_follow <- left_join(data_follow, tmp_follow)## Joining, by = c("ID_NUM", "time")tmp_follow <- select(data_follow_fys_funk, c(1, num_range(prefix = "haq", range = 1:7), matches("haq_total")))
data_follow <- left_join(data_follow, tmp_follow)## Joining, by = "ID_NUM"data_follow <- left_join(data_follow, data_follow_eq_5d)## Joining, by = "ID_NUM"data_follow <- left_join(data_follow, data_follow_psg)## Joining, by = "ID_NUM"Be careful to rename questions from EQ-5D from BA and FU to common name. Not done yet.
data_full <- bind_rows(data_base, data_follow)
dim(data_full)## [1] 76 79names(data_full)## [1] "ID_NUM" "Sex" "Age"
## [4] "Swol_joi" "Tend_joi" "DAS"
## [7] "HGB" "CRP" "Disease_dura"
## [10] "Comobid" "Educa" "Job"
## [13] "Live_alone" "Smoking" "alcohol"
## [16] "koffein" "time" "psqi1"
## [19] "psqi2" "psqi3" "psqi4"
## [22] "psqi5" "psqi6" "psqi7"
## [25] "psqi_total" "ess_total" "BRAF_phys"
## [28] "BRAF_living" "BRAF_cog" "BRAF_emo"
## [31] "BRAF_total" "fys_akt" "CES_D_total"
## [34] "haq1" "haq2" "haq3"
## [37] "haq4" "haq5" "haq6"
## [40] "haq7" "haq_total" "SP8_Spg_1_BA"
## [43] "SP8_Spg_2_BA" "SP8_Spg_3_BA" "SP8_Spg_4_BA"
## [46] "SP8_Spg_5_BA" "SP8_Spg_6_BA" "TST"
## [49] "Bedtime" "sleep_effi" "SO"
## [52] "sleep_latency" "REM_latency" "WASO"
## [55] "Arous" "wake_time" "N1"
## [58] "N2" "N3" "REM"
## [61] "AHI" "LM" "PLM"
## [64] "OSA" "Mix_apnea" "Cen_apnea"
## [67] "Hyp_nea" "Desat" "TST_num"
## [70] "SO_num" "sleep_latency_num" "REM_latency_num"
## [73] "wake_time_num" "SP8_Spg_1_FU" "SP8_Spg_2_FU"
## [76] "SP8_Spg_3_FU" "SP8_Spg_4_FU" "SP8_Spg_5_FU"
## [79] "SP8_Spg_6_FU"data_full <- left_join(data_full, data_rand_list)## Joining, by = "ID_NUM"write.table(data_full, file = "../data/data_full.txt", quote = F, row.names = F, sep = "\t")