One_page_paper
Iyad Sultan
2022-12-03
get data
Download the full code of this paper and upload it as Rmd to your R https://tinyurl.com/onePagePaper
Download data and upload it to the same directory of this file https://tinyurl.com/RHDS1
Download the myScripts.R file and upload to same directory of this file. MAKE SURE TO DOWNLOAD AS CSV not XSLX https://drive.google.com/file/d/1DnTfoMEC5lHek-QMZD1e-vZyfqtAC6di/view?usp=share_link
Use csv file to get all in text
ds<-read.csv("Hodgkin.csv")
cat("Dimensions of data before cropping")## Dimensions of data before croppingdim(ds)## [1] 999 29#View(ds)
ds<-ds[1:285, 1:27]
cat("Dimensions of data after cropping")## Dimensions of data after croppingdim(ds)## [1] 285 27sapply(ds, class) # calsses of all colomns## Date.of.Diagnosis CR.timing
## "character" "character"
## CR_After_2_cycles Age.at.DX
## "integer" "numeric"
## Chemotherapy Regimen.at.Dx
## "character" "integer"
## Total.number.of.cycles Total.Number.after.R.R
## "integer" "integer"
## BMT Radiation
## "character" "character"
## Radiation.with.primary.treatment Risk.Group
## "integer" "character"
## Stage Mets
## "character" "character"
## B.Symptom Bulky
## "character" "character"
## Nationality Outcome
## "character" "character"
## Relapse.Progression Date.of.Last.FU.or.rellapse
## "character" "character"
## Died Date.of.death.or.last.FU
## "integer" "character"
## Event..date.of.relapse Last.Follow.up
## "character" "character"
## Histology Weight
## "character" "numeric"
## Height
## "numeric"colnames(ds)<-janitor::make_clean_names(colnames(ds)) # make all colnames small letters with no spaces
head(ds) # have a peek on the data| date_of_diagnosis | cr_timing | cr_after_2_cycles | age_at_dx | chemotherapy | regimen_at_dx | total_number_of_cycles | total_number_after_r_r | bmt | radiation | radiation_with_primary_treatment | risk_group | stage | mets | b_symptom | bulky | nationality | outcome | relapse_progression | date_of_last_fu_or_rellapse | died | date_of_death_or_last_fu | event_date_of_relapse | last_follow_up | histology | weight | height |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 1/1/2004 | CR after 2nd ABVD | 2 | 15 | 4 ABVD | 1 | 4 | 0 | No | No | 0 | LR | II | No | No | No | Jordanian | Alive in CR | No | 27/07/2007 | 0 | 27/07/2007 | No | 27/07/2007 | MC | 42 | 167 |
| 1/3/2004 | CR after 2nd ABVD | 2 | 14 | 5 ABVD+ 1 COPP | 2 | 6 | 0 | No | No | 0 | HR | IV | No | No | No | Jordanian | Alive in CR | No | 15/05/2011 | 0 | 15/05/2011 | No | 15/05/2011 | NS | 40 | 161 |
| 1/6/2004 | CR after 2nd ABVD | 2 | 7 | 4 ABVD | 1 | 4 | 0 | No | No | 0 | LR | II | No | No | No | Non | Alive in CR | No | 28/07/2004 | 0 | 28/07/2004 | No | 28/07/2004 | NS | 26 | 119 |
| 1/6/2004 | PR after 2 ABVD, CR after 4 ABVD | 1 | 7 | 4 ABVD + 2 COPP | 2 | 6 | 0 | No | Yes | 1 | HR | IV | Yes | Yes | No | Jordanian | Alive in CR | No | 25/03/2012 | 0 | 25/03/2012 | No | 25/03/2012 | NS | 26 | 119 |
| 1/6/2004 | PR after 2 ABVD, CR after 4 ABVD | 1 | 16 | 4 ABVD + 2 COPP | 2 | 6 | 0 | No | Yes | 1 | HR | IV | Yes | Yes | No | Jordanian | Alive in CR | Yes | 24/04/2011 | 0 | 24/04/2011 | Yes 02/01/2007 | 24/04/2011 | NS | 44 | 173 |
| 1/7/2004 | CR after 2nd ABVD | 2 | 12.5 | 4 ABVD | 1 | 4 | 0 | No | No | 0 | LR | II | No | No | No | Jordanian | Alive in CR | No | 01/07/201 | 0 | 01/07/201 | No | 01/07/201 | NS | 37 | 152 |
look at all data before cleaning
exclude=which(str_detect(colnames(ds), "Date|date|Do|Dte|mrn|name|last")) # exclude columns that contain dates or IDs from summary
st_options( dfSummary.na.col = T, dfSummary.graph.col = F, dfSummary.valid.col = F) # options to show missing data but not graph and valid columns
dfSummary(ds[,-exclude],round.digits=2, style="grid", justify="left", plain.ascii=TRUE, append=FALSE, varnumbers=FALSE, , max.distinct.values=10, trim.strings=FALSE, max.string.width=20, split.cells=40, escape.pipe=T)| Variable | Stats / Values | Freqs (% of Valid) | Missing |
|---|---|---|---|
| cr_timing\ [character] | 1\. CR post 2ABVD\ 2\. CR post 4 BEACOPP\ 3\. CR after 2nd ABVD\ 4\. CR post 2 ABVD\ 5\. CR post 4 BEACOP\ 6\. CR post 2 ABVD·\ 7\. PR after 2 ABVD, CR \ 8\. CR AFTER 4 BEACOP·\ 9\. CR after 2 ABVD\ 10\. CR AFTER 2 ABVD·\ [ 30 others ] | \85 (29.8%)\ \67 (23.5%)\ \35 (12.3%)\ \24 ( 8.4%)\ \12 ( 4.2%)\ \ 9 ( 3.2%)\ \ 9 ( 3.2%)\ \ 4 ( 1.4%)\ \ 3 ( 1.1%)\ \ 3 ( 1.1%)\ \34 (11.9%) | 0\ (0.0%) |
| cr_after_2_cycles\ [integer] | Mean (sd) : 1.85 (0.46)\ min < med < max:\ 0 < 2 < 2\ IQR (CV) : 0 (0.25) | 0 : 12 ( 4.2%)\ 1 : 18 ( 6.3%)\ 2 : 255 (89.5%) | 0\ (0.0%) |
| age_at_dx\ [numeric] | Mean (sd) : 11.69 (4.55)\ min < med < max:\ 2 < 12.62 < 18\ IQR (CV) : 8.5 (0.39) | 83 distinct values | 1\ (0.4%) |
| chemotherapy\ [character] | 1\. BEACOPP\ 2\. 4 ABVD\ 3\. 4 ABVD + 2 COPP\ 4\. 3 ABVD\ 5\. 3 ABVD, 2 COPDac\ 6\. 4 ABVD, 2 COPP\ 7\. 4 ABVD, 2 COPDac\ 8\. BEACOPP·\ 9\. 4 ABVD·\ 10\. 4 ABVD, 2 COPdac\ [ 36 others ] | \77 (27.0%)\ \60 (21.1%)\ \40 (14.0%)\ \18 ( 6.3%)\ \12 ( 4.2%)\ \11 ( 3.9%)\ \ 7 ( 2.5%)\ \ 7 ( 2.5%)\ \ 5 ( 1.8%)\ \ 4 ( 1.4%)\ \44 (15.4%) | 0\ (0.0%) |
| regimen_at_dx\ [integer] | Mean (sd) : 1.99 (0.81)\ min < med < max:\ 1 < 2 < 3\ IQR (CV) : 2 (0.41) | 1 : 95 (33.3%)\ 2 : 99 (34.7%)\ 3 : 91 (31.9%) | 0\ (0.0%) |
| total_number_of_cycles\ [integer] | Mean (sd) : 5.67 (1.6)\ min < med < max:\ 3 < 6 < 11\ IQR (CV) : 2 (0.28) | 3 : 21 ( 7.4%)\ 4 : 70 (24.6%)\ 5 : 10 ( 3.5%)\ 6 : 129 (45.3%)\ 7 : 1 ( 0.4%)\ 8 : 49 (17.2%)\ 9 : 1 ( 0.4%)\ 10 : 3 ( 1.1%)\ 11 : 1 ( 0.4%) | 0\ (0.0%) |
| total_number_after_r_r\ [integer] | Mean (sd) : 0.15 (0.67)\ min < med < max:\ 0 < 0 < 5\ IQR (CV) : 0 (4.55) | 0 : 270 (94.7%)\ 2 : 9 ( 3.2%)\ 3 : 1 ( 0.4%)\ 4 : 4 ( 1.4%)\ 5 : 1 ( 0.4%) | 0\ (0.0%) |
| bmt\ [character] | 1\. No\ 2\. Yes | \274 (96.1%)\ \ 11 ( 3.9%) | 0\ (0.0%) |
| radiation\ [character] | 1\. no\ 2\. No\ 3\. No·\ 4\. TBI at relapse\ 5\. Yes\ 6\. yes·\ 7\. Yes (at relapse)\ 8\. Yes at relaps)\ 9\. Yes at relapse\ 10\. Yes at replase·\ 11\. Yes, BONE | \ 1 ( 0.4%)\ \210 (73.7%)\ \ 2 ( 0.7%)\ \ 1 ( 0.4%)\ \ 55 (19.3%)\ \ 7 ( 2.5%)\ \ 1 ( 0.4%)\ \ 1 ( 0.4%)\ \ 3 ( 1.1%)\ \ 3 ( 1.1%)\ \ 1 ( 0.4%) | 0\ (0.0%) |
| radiation_with_primary_treatment\ [integer] | Mean (sd) : 0.32 (0.61)\ min < med < max:\ 0 < 0 < 3\ IQR (CV) : 1 (1.93) | 0 : 213 (74.7%)\ 1 : 58 (20.4%)\ 2 : 10 ( 3.5%)\ 3 : 4 ( 1.4%) | 0\ (0.0%) |
| risk_group\ [character] | 1\. HR\ 2\. HR·\ 3\. HR·\ 4\. IR\ 5\. IR·\ 6\. LR\ 7\. LR· | \120 (42.1%)\ \ 16 ( 5.6%)\ \ 1 ( 0.4%)\ \ 52 (18.2%)\ \ 5 ( 1.8%)\ \ 88 (30.9%)\ \ 3 ( 1.1%) | 0\ (0.0%) |
| stage\ [character] | 1\. II\ 2\. II·\ 3\. IV\ 4\. IIIS\ 5\. III\ 6\. IV·\ 7\. I\ 8\. III·\ 9\. IV·\ 10\. I·\ [ 3 others ] | \94 (33.0%)\ \46 (16.1%)\ \32 (11.2%)\ \26 ( 9.1%)\ \21 ( 7.4%)\ \19 ( 6.7%)\ \16 ( 5.6%)\ \15 ( 5.3%)\ \ 6 ( 2.1%)\ \ 4 ( 1.4%)\ \ 6 ( 2.1%) | 0\ (0.0%) |
| mets\ [character] | 1\. No\ 2\. yes\ 3\. Yes\ 4\. YES· | \228 (80.0%)\ \ 1 ( 0.4%)\ \ 55 (19.3%)\ \ 1 ( 0.4%) | 0\ (0.0%) |
| b_symptom\ [character] | 1\. No\ 2\. yes\ 3\. Yes | \153 (53.7%)\ \ 2 ( 0.7%)\ \130 (45.6%) | 0\ (0.0%) |
| bulky\ [character] | 1\. No\ 2\. yes\ 3\. Yes | \177 (62.1%)\ \ 1 ( 0.4%)\ \107 (37.5%) | 0\ (0.0%) |
| nationality\ [character] | 1\. Jordanian\ 2\. Kuwati\ 3\. Non\ 4\. Syrian | \251 (88.1%)\ \ 1 ( 0.4%)\ \ 32 (11.2%)\ \ 1 ( 0.4%) | 0\ (0.0%) |
| outcome\ [character] | 1\. Alive in CR\ 2\. Alive in CR\ 3\. Alive Off Therapy in\ 4\. AWD\ 5\. Died\ 6\. Died with AML\ 7\. Died in CR/ Sepsis\ 8\. died with disease\ 9\. Died with disease\ 10\. Died with sepsis\ 11\. Lost to followup | \201 (70.5%)\ \ 62 (21.8%)\ \ 8 ( 2.8%)\ \ 1 ( 0.4%)\ \ 1 ( 0.4%)\ \ 1 ( 0.4%)\ \ 1 ( 0.4%)\ \ 4 ( 1.4%)\ \ 1 ( 0.4%)\ \ 1 ( 0.4%)\ \ 4 ( 1.4%) | 0\ (0.0%) |
| relapse_progression\ [character] | 1\. no\ 2\. No\ 3\. Yes | \193 (67.7%)\ \ 67 (23.5%)\ \ 25 ( 8.8%) | 0\ (0.0%) |
| died\ [integer] | Min : 0\ Mean : 0.03\ Max : 1 | 0 : 277 (97.2%)\ 1 : 8 ( 2.8%) | 0\ (0.0%) |
| histology\ [character] | 1\. (Empty string)\ 2\. HD\ 3\. Interfolicular\ 4\. Interfollicular\ 5\. LP\ 6\. LR\ 7\. Lymphocyte Rich\ 8\. MC\ 9\. NS | \ 2 ( 0.8%)\ \ 1 ( 0.4%)\ \ 1 ( 0.4%)\ \ 2 ( 0.8%)\ \ 7 ( 2.7%)\ \ 4 ( 1.5%)\ \ 1 ( 0.4%)\ \ 73 (28.1%)\ \169 (65.0%) | 25\ (8.8%) |
| weight\ [numeric] | Mean (sd) : 35.29 (9.19)\ min < med < max:\ 12 < 37 < 48\ IQR (CV) : 17 (0.26) | 84 distinct values | 0\ (0.0%) |
| height\ [numeric] | Mean (sd) : 146.88 (27.57)\ min < med < max:\ 77 < 152 < 185\ IQR (CV) : 51 (0.19) | 84 distinct values | 0\ (0.0%) |
Look at columns unique values to understand how messy your data
# This code can be made better!
DS<- ds %>% # create a new dataset
select(-exclude) %>% # exclude defined above
select(where(is.character)) # select only character columns
lapply(DS, unique) # get unique elements in each column of DS## $cr_timing
## [1] "CR after 2nd ABVD"
## [2] "PR after 2 ABVD, CR after 4 ABVD"
## [3] "CR after 2nd ABVD, CR after 2 COPP"
## [4] "CR after 2 ABVD"
## [5] "CR after 2 ABVD, CR post AlloBMT"
## [6] "PR after 2nd ABVD"
## [7] "PR after 2 ABVD, CR after 4 ABVD, 2 ICE"
## [8] "PR after 2 ABVD"
## [9] "CR post 2ABVD"
## [10] "PR post2ABVD"
## [11] "CR post 4 BEACOPP"
## [12] "PR post 2ABVD, CR post 4 ABVD"
## [13] "PR post 2ABVD, PR post 4 ABVD"
## [14] "CR post 2ABVD, CR post 2 ICE and ABMT"
## [15] "CR post 2 ABVD"
## [16] "CR post 2 ABVD, then Post 2 COPDac"
## [17] "PR post 4 BEACOPP"
## [18] "CR post 2ABVD, CR post 2 COPP"
## [19] "PR post 2ABVD"
## [20] "CR AFTER 4 BEACOP "
## [21] "CR AFTER 2 ABVD , 2 post GV, ICE"
## [22] "CR AFTER2 ABVD "
## [23] "CR AFTER 2 ABVD "
## [24] "CR AFTER 4 BEACOP"
## [25] "PR post 2 ABVD"
## [26] "CR post 4 BEACOPP, CR post IGEV"
## [27] "CR post 2 ABVD, CR post ICE and ABMT"
## [28] "CR post 3rd regimen"
## [29] "CRpost 2 ABVD"
## [30] "CR after all chemo"
## [31] "CRpostr 4 BEACOPP"
## [32] "CR post 3 BEACOPP and ABVD"
## [33] "CR after 5 BEACOPP, 2ICE"
## [34] "CR post 2ABVD, then post ICE , then post Keytruda"
## [35] "CR post 4 BEACOP"
## [36] "CR post 2 ABVD "
## [37] "PR post 4 BEACOP"
## [38] "PR post 2 ABVD "
## [39] "PR post 2ABVD, CR post 5 cycles"
## [40] "PR post 2ABVD, PR post 4 ABVD, CR post 2 ICE"
##
## $chemotherapy
## [1] "4 ABVD"
## [2] "5 ABVD+ 1 COPP"
## [3] "4 ABVD + 2 COPP"
## [4] "4 ABVD then 4 COPP"
## [5] "4 ABVD + 2 COPP, 2 ICE , ABMT, AlloBMT"
## [6] "4 ABVD+ 2 COPP"
## [7] "4 ABVD "
## [8] "4ABVD, 2COPP, ICE, ABMT"
## [9] "3 ABVD"
## [10] "4 ABVD+ 2 COPP, 2nd line"
## [11] "4ABVD, 2COPP, ICE, ABMT, GV, DHAP XRT"
## [12] "4ABVD, 2COPP, ICE"
## [13] "4ABVD "
## [14] "BEACOPP"
## [15] "4 ABVD, 2 ICE"
## [16] "4 ABVD + 2 COPdac"
## [17] "5 ABVD + 2 COPdac"
## [18] "4 ABVD, 4 COPDac"
## [19] "4 ABVD, 2 COPP"
## [20] "4ABVD"
## [21] "4 ABVD, then 4 COPP"
## [22] "4 ABVD, 2 COPdac"
## [23] "5 ABVD"
## [24] "4 ABVD, 2 COPDac"
## [25] "4 ABVD, 2 COPDac, 3 ICE, 2 GV, ABMT"
## [26] "4 ABVD, 2 COPDac, 2 ICE, ABMT"
## [27] "4 ABVD, 2COPDac, 2 ICE"
## [28] "3 ABVD, (3RICE at relapse)"
## [29] "3 ABVD "
## [30] "4 BEACOP, 2COPP/ABV, 2 ICE"
## [31] "3 ABVD, (4COPDac at relapse)"
## [32] "3 ABVD, 2 COPDac"
## [33] "BEACOPP (IGEV at relape)"
## [34] " BEACOPP"
## [35] "4 ABVD, 2COPDac"
## [36] "5 BEACOPP, 2ICE, Auto BMT"
## [37] "4 ABVD, 2COPDAC , 2 ICE, 2 GV Bend(auto and allo BMT)"
## [38] "4 ABVD, 2COPP"
## [39] " BEACOPP, IGEV, keytrauda, auto BMT after replase)"
## [40] "Chemo, ABMT, AlloBMT"
## [41] "3 ABVD, 2COPP"
## [42] "BEACOPP "
## [43] "3 ABVD, 2 COPP"
## [44] "3 ABVD, 1 COPDac"
## [45] "BEACOPP/ICE ABMT"
## [46] "3 ABVD, 4 COPP, 2 ICE , XRT"
##
## $bmt
## [1] "No" "Yes"
##
## $radiation
## [1] "No" "Yes" "Yes at replase " "no"
## [5] "Yes, BONE" "Yes at relapse" "Yes at relaps)" "Yes (at relapse)"
## [9] "TBI at relapse" "No " "yes "
##
## $risk_group
## [1] "LR" "HR" "HR " "IR" "HR " "IR " "LR "
##
## $stage
## [1] "II" "IV " "II " "IV " "I" "III " "IV" "I " "III" "IIIS"
## [11] "IVS" "II " ""
##
## $mets
## [1] "No" "Yes" "yes" "YES "
##
## $b_symptom
## [1] "No" "Yes" "yes"
##
## $bulky
## [1] "No" "Yes" "yes"
##
## $nationality
## [1] "Jordanian" "Non" "Kuwati" "Syrian"
##
## $outcome
## [1] "Alive in CR" "Alive in CR"
## [3] "Alive Off Therapy in CR" "died with disease"
## [5] "Died with sepsis" "Lost to followup"
## [7] "Died with AML" "Died with disease"
## [9] "AWD" "Died"
## [11] "Died in CR/ Sepsis"
##
## $relapse_progression
## [1] "No" "Yes" "no"
##
## $histology
## [1] "MC" "NS" NA "LP"
## [5] "Interfolicular" "Interfollicular" "Lymphocyte Rich" "LR"
## [9] "HD" ""rm(DS) # delete DS since we are not going to use itData cleaning
cat("Stages before cleaning\n")## Stages before cleaningds %>% count(stage)| stage | n |
|---|---|
| 1 | |
| I | 16 |
| I | 4 |
| II | 94 |
| II | 46 |
| II | 2 |
| III | 21 |
| III | 15 |
| IIIS | 26 |
| IV | 32 |
| IV | 19 |
| IV | 6 |
| IVS | 3 |
ds<- ds %>%
mutate(across(where(is.character), stringr::str_to_upper)) %>% # convert all data to upper case
mutate(across(where(is.character), stringr::str_squish)) # remove spaces from both ends " V " becomes "V" : important step
cat("Stages after cleaning\n")## Stages after cleaningds %>% count(stage)| stage | n |
|---|---|
| 1 | |
| I | 20 |
| II | 142 |
| III | 36 |
| IIIS | 26 |
| IV | 57 |
| IVS | 3 |
clean the column of timing to CR by creating 2 columns
ds<-ds %>%
mutate(best_response=if_else(str_detect(cr_timing, "CR"), "CR", "PR")) %>% # get CR or PR
mutate(best_response_after=str_extract(cr_timing, "(\\d)+")) # get only first number
# count(ds, best_response_after)clean histology
cat("Histology before cleaning\n")## Histology before cleaningds %>%
count(histology)| histology | n |
|---|---|
| 2 | |
| HD | 1 |
| INTERFOLICULAR | 1 |
| INTERFOLLICULAR | 2 |
| LP | 7 |
| LR | 4 |
| LYMPHOCYTE RICH | 1 |
| MC | 73 |
| NS | 169 |
| 25 |
ds <- ds %>%
mutate(histology=recode(histology,
"INTERFOLICULAR"="Int",
"INTERFOLLICULAR"="Int",
"LYMPHOCYTE RICH"="LR"))%>%
replace(., .=="", NA) # this replaces all empty cells with NA
cat("Histology after cleaning\n")## Histology after cleaningds %>%
count(histology)| histology | n |
|---|---|
| HD | 1 |
| Int | 3 |
| LP | 7 |
| LR | 5 |
| MC | 73 |
| NS | 169 |
| 27 |
# unique(ds$histology)clean the radiation column
cat("Radiation before cleaning\n")## Radiation before cleaningds %>% count(radiation)| radiation | n |
|---|---|
| NO | 213 |
| TBI AT RELAPSE | 1 |
| YES | 62 |
| YES (AT RELAPSE) | 1 |
| YES AT RELAPS) | 1 |
| YES AT RELAPSE | 3 |
| YES AT REPLASE | 3 |
| YES, BONE | 1 |
ds <- ds %>% mutate(radiation=if_else(radiation %in% c("YES", "YES, BONE"), "YES", "NO"))
cat("Radiation after cleaning\n")## Radiation after cleaningds %>% count(radiation)| radiation | n |
|---|---|
| NO | 222 |
| YES | 63 |
clean the dates
If some dates are text you may get an error here. Go and check your dates
ds<- ds %>%
mutate(across(starts_with("date"), function(x) lubridate::as_datetime(sapply(x, fix_date)))) # apply fix date function from myScripts.R which should fix any date
cat("Which dates are NA to fix the original data\n")## Which dates are NA to fix the original datalapply(ds%>% select(starts_with("date")) , function(x) which(is.na(x)|x=="9999-09-09")) # this is the value returned by fix_date for NA dates## $date_of_diagnosis
## 2014/011/25
## 184
##
## $date_of_last_fu_or_rellapse
## 01/07/201
## 6
##
## $date_of_death_or_last_fu
## 01/07/201
## 6create dead and relapse column
ds$dead<- str_detect(ds$outcome, "DIED")
ds$event<- ifelse(ds$relapse_progression=="YES" | ds$dead, T, F)calculate OS and EFS
ds<-
ds %>%
mutate(OS=as.numeric(date_of_death_or_last_fu-date_of_diagnosis)/30) %>%
mutate(EFS=as.numeric(date_of_last_fu_or_rellapse-date_of_diagnosis)/30)calculate median follow up
This is defined as the median duration of follow up from diagnosis to last follow up or death for patients who did not die
cat("Median FU in months=")## Median FU in months=summary(ds %>% filter(dead==0) %>% pull(OS))## Min. 1st Qu. Median Mean 3rd Qu. Max. NA's
## -7.90 25.32 48.42 58.80 85.51 179.57 2table 1
ds %>%
select(histology, stage, bulky, b_symptom, mets, radiation, best_response, best_response_after, dead, event, bmt) %>%
tbl_summary(by=best_response, # compare patients according to response
label = list(histology ~ "Histology", # rename rows of the table
stage ~ "Stage",
bulky ~ "Bulky",
b_symptom ~ "B symptoms",
radiation ~ "Radiation",
best_response ~ "Best response",
best_response_after ~ "Number of cycles",
dead ~ "Patient died",
event ~ "Patient had event",
bmt ~ "Bone Marrow Transplantation")) %>%
add_p() # add p values | Characteristic | CR, N = 2721 | PR, N = 131 | p-value2 |
|---|---|---|---|
| Histology | >0.9 | ||
| HD | 1 (0.4%) | 0 (0%) | |
| Int | 3 (1.2%) | 0 (0%) | |
| LP | 7 (2.9%) | 0 (0%) | |
| LR | 5 (2.0%) | 0 (0%) | |
| MC | 69 (28%) | 4 (31%) | |
| NS | 160 (65%) | 9 (69%) | |
| Unknown | 27 | 0 | |
| Stage | 0.7 | ||
| I | 20 (7.4%) | 0 (0%) | |
| II | 136 (50%) | 6 (46%) | |
| III | 35 (13%) | 1 (7.7%) | |
| IIIS | 25 (9.2%) | 1 (7.7%) | |
| IV | 52 (19%) | 5 (38%) | |
| IVS | 3 (1.1%) | 0 (0%) | |
| Unknown | 1 | 0 | |
| Bulky | 101 (37%) | 7 (54%) | 0.3 |
| B symptoms | 123 (45%) | 9 (69%) | 0.090 |
| mets | 52 (19%) | 5 (38%) | 0.15 |
| Radiation | 51 (19%) | 12 (92%) | <0.001 |
| Number of cycles | 0.4 | ||
| 2 | 182 (67%) | 11 (85%) | |
| 3 | 2 (0.7%) | 0 (0%) | |
| 4 | 86 (32%) | 2 (15%) | |
| 5 | 1 (0.4%) | 0 (0%) | |
| Unknown | 1 | 0 | |
| Patient died | 6 (2.2%) | 3 (23%) | 0.005 |
| Patient had event | 26 (9.6%) | 3 (23%) | 0.13 |
| Bone Marrow Transplantation | 9 (3.3%) | 2 (15%) | 0.084 |
| 1 n (%) | |||
| 2 Fisher's exact test; Pearson's Chi-squared test | |||
Donut chart for stages and histologies
A<- ds %>% mutate(var=stage) %>% count(var) %>% filter(!is.na(var)) %>% donut_chart()
B<- ds %>% mutate(histology=ifelse(histology %in% c("MC","NS", "LP"), histology, "Others")) %>% mutate(var=histology)%>% count(var) %>% filter(!is.na(var)) %>% donut_chart()
ggarrange(A,B, labels="AUTO")
Histogram for age and histology
ds %>%
mutate(histology=ifelse(histology %in% c("MC","NS", "LP"), histology, "Others")) %>%
ggplot(aes(x=age_at_dx, fill= fct_rev (fct_infreq(histology))))+
geom_histogram(bins=17)+
theme_classic2()+
labs(x="Age", y="Frequency", fill="Histology")+
theme(legend.position = c(0.2,0.8))+
scale_fill_brewer(palette="Set1")+
scale_x_continuous(breaks=2:18)
Draw the curves of OS and EFS
os<-survfit(Surv(OS, dead)~1, data=ds)
OS<-ggsurvplot(os, data = ds, legend="none", conf.int = F, palette="red", xlab="Months", break.time.by=12)
efs<- survfit(Surv(EFS, event)~1, data=ds)
EFS<- ggsurvplot(efs, data = ds, legend="none", conf.int = F, palette="blue", xlab="Months", break.time.by=12)
plots<- list(OS=os,EFS=efs)
ggsurvplot(plots, data=ds, combine = TRUE,
palette=c("red", "blue"),
break.time.by=12, xlab="Months",
legend=c(0.2,0.15),
legend.title="",
legend.labs=c("Overall Survival", "Event Free Survival"),
surv.scale="percent", xlim=c(0,60))
# compare outcome for those who had PR and CR
source("myScripts.R") # this will upload all functions in myScripts
ds %>% mutate(var=best_response, os=OS/12) %>% OS("according to best response")
## according to best response
## Characteristic Time 1 Time 5
## ──────────────────────────────────────────────────────
## Overall 98% (97%, 100%) 96% (94%, 99%)
## var
## CR 98% (97%, 100%) 97% (95%, 100%)
## PR 92% (79%, 100%) 55% (27%, 100%)
##
## Column names: label, stat_1, stat_2
## P= 1.950382e-06ds %>% mutate(var=best_response, efs=EFS/12) %>% EFS("according to best response")
## according to best response
## Characteristic Time 1 Time 5
## ──────────────────────────────────────────────────────
## Overall 98% (97%, 100%) 88% (84%, 93%)
## var
## CR 98% (97%, 100%) 89% (85%, 94%)
## PR 92% (79%, 100%) 55% (27%, 100%)
##
## Column names: label, stat_1, stat_2
## P= 0.01276748Cox regression for survival
ds$best_response_after_2cycles=ifelse(ds$best_response_after=="2", "2 cycles", "More than 2")
explanatory=Cs(histology, stage, bulky, b_symptom, mets, radiation, best_response, best_response_after_2cycles)
ds<- ds %>% mutate(across(explanatory, as.factor))
ds<-ds %>% mutate(histology=fct_relevel(histology, "NS"),
histology=fct_lump(histology, 4))
dependent = "Surv(OS, dead)"
ds %>% finalfit(dependent, explanatory) ->t
gt(t)| Dependent: Surv(OS, dead) | all | HR (univariable) | HR (multivariable) | |
|---|---|---|---|---|
| histology | NS | 169 (65.5) | - | - |
| LP | 7 (2.7) | 0.00 (0.00-Inf, p=0.999) | 916398.33 (0.00-Inf, p=1.000) | |
| LR | 5 (1.9) | 0.00 (0.00-Inf, p=0.999) | 0.02 (0.00-Inf, p=1.000) | |
| MC | 73 (28.3) | 0.49 (0.05-4.42, p=0.528) | 6.33 (0.50-80.00, p=0.154) | |
| Other | 4 (1.6) | 0.00 (0.00-Inf, p=0.999) | 250848.07 (0.00-Inf, p=1.000) | |
| stage | I | 20 (7.0) | - | - |
| II | 142 (50.0) | 85992615.44 (0.00-Inf, p=0.999) | 0.77 (0.11-5.27, p=0.786) | |
| III | 36 (12.7) | 1.00 (0.00-Inf, p=1.000) | 0.00 (0.00-Inf, p=0.997) | |
| IIIS | 26 (9.2) | 175449624.16 (0.00-Inf, p=0.999) | 0.00 (0.00-Inf, p=0.997) | |
| IV | 57 (20.1) | 365804748.26 (0.00-Inf, p=0.999) | 0.08 (0.01-0.58, p=0.012) | |
| IVS | 3 (1.1) | 0.99 (0.00-Inf, p=1.000) | 0.00 (0.00-Inf, p=1.000) | |
| bulky | NO | 177 (62.1) | - | - |
| YES | 108 (37.9) | 1.39 (0.37-5.16, p=0.627) | 0.68 (0.10-4.64, p=0.692) | |
| b_symptom | NO | 153 (53.7) | - | - |
| YES | 132 (46.3) | 4.16 (0.86-20.04, p=0.075) | 37871418.21 (0.00-Inf, p=0.994) | |
| mets | NO | 228 (80.0) | - | - |
| YES | 57 (20.0) | 3.19 (0.86-11.87, p=0.084) | 6.72 (0.98-46.27, p=0.053) | |
| radiation | NO | 222 (77.9) | - | - |
| YES | 63 (22.1) | 7.71 (1.93-30.86, p=0.004) | 247965289.99 (0.00-Inf, p=0.995) | |
| best_response | CR | 272 (95.4) | - | - |
| PR | 13 (4.6) | 13.61 (3.35-55.36, p<0.001) | 5.46 (0.86-34.68, p=0.072) | |
| best_response_after_2cycles | 2 cycles | 193 (68.0) | - | - |
| More than 2 | 91 (32.0) | 0.25 (0.03-2.04, p=0.197) | 0.14 (0.02-1.31, p=0.085) |