Gout management in Hospital
Jacklyn Chin
David Fong
27th August 2020
library(ggplot2) # plotting
library(highcharter) # plotting
library(dplyr) # data pliers
library(tidyr) # 'tidy' up data
library(gmodels) # contingency tables
library(mice) # imputation
library(corrplot) # correlation plot
library(lme4) # linear models
library(ztable) # cross-table display
options(ztable.type="html")
library(magrittr) # pipes
library(janitor) # clean names
library(huxtable)Read the data
81 entries
and ‘clean up’ the names
d <- read.csv("SerumUricAcid.csv",
skip = 3, # a few rows to skip at the beginning
check.names = TRUE, # avoid duplicate names
stringsAsFactors = FALSE
)
d <- d[seq(from = 1, to = 81),] # 81 entriesClean names
‘Clean up the names’ - e.g. spaces changed to underscores ’_’
names(d) <- make_clean_names(names(d))
names(d)[1] “number”
[2] “age”
[3] “gender”
[4] “ethnicity”
[5] “hospital_campus”
[6] “hopc”
[7] “primary_care_team”
[8] “unit_of_admission”
[9] “grouped_units_of_admission”
[10] “rheum_input”
[11] “admission_date”
[12] “discharge_date”
[13] “duration_of_stay”
[14] “number_of_active_joints_involved”
[15] “joint_s_involved”
[16] “mtp”
[17] “ankle”
[18] “knee”
[19] “elbow”
[20] “wrist”
[21] “mcp”
[22] “tophi”
[23] “therapeutic_agents”
[24] “nsaid”
[25] “nsaid_type”
[26] “nsaid_dose”
[27] “duration”
[28] “colchicine”
[29] “total_col”
[30] “duration_col”
[31] “mean_daily_col”
[32] “d1_col”
[33] “d2_col”
[34] “total_daily_dose_mg”
[35] “course”
[36] “prednisolone”
[37] “total_pred”
[38] “duration_pred”
[39] “mean_daily_pred”
[40] “d1_pred”
[41] “d2_pred”
[42] “course_1”
[43] “course_2”
[44] “wean”
[45] “ia_steroid_injections”
[46] “dose”
[47] “comments”
[48] “x_diuretics_used”
[49] “thiazide_diuretics”
[50] “dose_1”
[51] “loop_diuretics”
[52] “k_sparing_diuretics”
[53] “other_diuretics”
[54] “dose_2”
[55] “anti_hypertensives”
[56] “x_anti_htn”
[57] “b_blocker”
[58] “ac_ei”
[59] “arb”
[60] “ccb”
[61] “a_blocker”
[62] “a_b_blocker”
[63] “moxonidine”
[64] “dose_3”
[65] “x_anti_chol_used”
[66] “statins”
[67] “fibrates”
[68] “ezetimibe”
[69] “other”
[70] “cholesterol_lowering_therapy”
[71] “dose_4”
[72] “x_anti_platelet_therapy”
[73] “aspirin”
[74] “clopidogrel”
[75] “total_number”
[76] “dose_5”
[77] “htn”
[78] “t2dm”
[79] “cvd”
[80] “chol”
[81] “ckd”
[82] “ckd_stage”
[83] “concurrent_infections”
[84] “weight”
[85] “height”
[86] “bmi”
[87] “obesity”
[88] “x_total_comorbidities”
[89] “x0”
[90] “x1”
[91] “x2”
[92] “x3”
[93] “x4”
[94] “x5”
[95] “x6”
[96] “total_3_or_more”
[97] “duration_of_therapy”
[98] “clinical_asssesed_improvement”
[99] “patient_reported_improvemnet”
[100] “ult_on_admission_y_n”
[101] “dose_2_2”
[102] “commenced_during”
[103] “dose_1_2”
[104] “after_discharge”
[105] “dose_2_2_2”
[106] “change_dose”
[107] “old_dose”
[108] “new_dose”
[109] “ceased_withheld”
[110] “dose_3_2”
[111] “serum_urate_range”
[112] “serum_urate_level_mmol_l”
[113] “interval_to_next_flare”
[114] “d_c_instructions_present”
[115] “discharge_instructions”
[116] “factors_complicating_gout_treatment” [117] “none”
[118] “infections”
[119] “diuresis”
[120] “aki”
[121] “trauma”
[122] “cognitive_impairment”
[123] “acidosis”
[124] “other_rheum”
[125] “chemo”
[126] “surgery”
[127] “comments_2”
Some variables are ‘categorical’
…rather than ordered.
categorical_cols <- c("gender", "primary_care_team", "ethnicity", "grouped_units_of_admission", "hospital_campus", "hopc",
"rheum_input", "mtp", "ankle", "knee", "elbow", "wrist", "mcp", "tophi",
"nsaid", "colchicine", "prednisolone", "wean", "ia_steroid_injections",
"thiazide_diuretics", "loop_diuretics", "k_sparing_diuretics",
"b_blocker", "ac_ei", "arb", "ccb", "a_blocker", "a_b_blocker", "moxonidine",
"statins", "fibrates", "ezetimibe",
"aspirin", "clopidogrel", "htn", "t2dm", "cvd", "chol", "ckd",
"obesity", "clinical_asssesed_improvement", "patient_reported_improvemnet",
"ult_on_admission_y_n", "commenced_during", "after_discharge", "change_dose",
"ceased_withheld", "d_c_instructions_present",
"none", "infections", "diuresis", "aki", "trauma", "cognitive_impairment",
"acidosis", "other_rheum", "chemo", "surgery")
d[categorical_cols] <- lapply(d[categorical_cols], factor) # apply 'factor' to those columnssummary(d)## number age gender ethnicity hospital_campus hopc
## Min. : 1 Min. :24.00 0:53 0:81 0:29 1:35
## 1st Qu.:21 1st Qu.:60.00 1:28 1:50 2:46
## Median :41 Median :77.00 2: 2
## Mean :41 Mean :70.98
## 3rd Qu.:61 3rd Qu.:83.00
## Max. :81 Max. :95.00
##
## primary_care_team unit_of_admission grouped_units_of_admission rheum_input
## 0:79 Length:81 0: 2 0:64
## 1: 2 Class :character 1:15 1:17
## Mode :character 2:46
## 3:15
## 4: 2
## 5: 1
##
## admission_date discharge_date duration_of_stay
## Length:81 Length:81 Min. : 0.050
## Class :character Class :character 1st Qu.: 1.600
## Mode :character Mode :character Median : 5.680
## Mean : 8.896
## 3rd Qu.:11.090
## Max. :53.780
##
## number_of_active_joints_involved joint_s_involved mtp ankle
## Min. :1.000 Length:81 0 :45 0 :56
## 1st Qu.:1.000 Class :character 1 :33 1 :22
## Median :1.000 Mode :character NA's: 3 NA's: 3
## Mean :1.716
## 3rd Qu.:2.000
## Max. :5.000
##
## knee elbow wrist mcp tophi therapeutic_agents nsaid
## 0 :57 0 :70 0 :62 0 :62 0 :68 Min. :0.000 0:71
## 1 :21 1 : 8 1 :16 1 :16 1 :12 1st Qu.:1.000 1:10
## NA's: 3 NA's: 3 NA's: 3 NA's: 3 NA's: 1 Median :1.000
## Mean :1.395
## 3rd Qu.:2.000
## Max. :3.000
##
## nsaid_type nsaid_dose duration colchicine total_col
## Length:81 Min. : 50 Min. :1.000 0:30 Min. : 0.500
## Class :character 1st Qu.:400 1st Qu.:1.000 1:51 1st Qu.: 1.500
## Mode :character Median :400 Median :2.000 Median : 3.000
## Mean :365 Mean :2.222 Mean : 4.019
## 3rd Qu.:400 3rd Qu.:3.000 3rd Qu.: 5.000
## Max. :400 Max. :4.000 Max. :14.000
## NA's :71 NA's :72 NA's :42
## duration_col mean_daily_col d1_col d2_col
## Min. : 1.000 Min. :0.2500 Min. :0.2500 Min. :0.0000
## 1st Qu.: 2.000 1st Qu.:0.5000 1st Qu.:0.5000 1st Qu.:0.5000
## Median : 5.000 Median :0.6000 Median :1.0000 Median :0.5000
## Mean : 5.615 Mean :0.7702 Mean :0.9559 Mean :0.5343
## 3rd Qu.: 8.000 3rd Qu.:1.0000 3rd Qu.:1.5000 3rd Qu.:0.7500
## Max. :21.000 Max. :1.5000 Max. :2.0000 Max. :1.0000
## NA's :42 NA's :37 NA's :30 NA's :30
## total_daily_dose_mg course prednisolone total_pred
## Length:81 Length:81 0:29 Min. : 25.0
## Class :character Class :character 1:52 1st Qu.: 50.0
## Mode :character Mode :character Median :101.2
## Mean :126.4
## 3rd Qu.:150.0
## Max. :420.0
## NA's :29
## duration_pred mean_daily_pred d1_pred d2_pred
## Min. : 1.000 Min. : 6.548 Min. : 5.00 Min. : 0.00
## 1st Qu.: 3.000 1st Qu.:14.125 1st Qu.:15.00 1st Qu.:15.00
## Median : 5.000 Median :16.201 Median :25.00 Median :20.00
## Mean : 7.885 Mean :19.706 Mean :23.32 Mean :18.61
## 3rd Qu.: 9.250 3rd Qu.:25.000 3rd Qu.:25.00 3rd Qu.:25.00
## Max. :30.000 Max. :50.000 Max. :50.00 Max. :50.00
## NA's :29 NA's :29 NA's :29 NA's :29
## course_1 course_2 wean ia_steroid_injections
## Length:81 Length:81 0 :33 0:79
## Class :character Class :character 1 :19 1: 2
## Mode :character Mode :character NA's:29
##
##
##
##
## dose comments x_diuretics_used thiazide_diuretics
## Min. :40 Length:81 Min. :0.0000 0:64
## 1st Qu.:40 Class :character 1st Qu.:0.0000 1:17
## Median :40 Mode :character Median :1.0000
## Mean :40 Mean :0.8148
## 3rd Qu.:40 3rd Qu.:1.0000
## Max. :40 Max. :3.0000
## NA's :79
## dose_1 loop_diuretics k_sparing_diuretics other_diuretics
## Length:81 0:40 0:73 Min. :0
## Class :character 1:41 1: 8 1st Qu.:0
## Mode :character Median :0
## Mean :0
## 3rd Qu.:0
## Max. :0
##
## dose_2 anti_hypertensives x_anti_htn b_blocker ac_ei arb
## Length:81 Length:81 Min. :0.000 0:39 0:71 0:68
## Class :character Class :character 1st Qu.:1.000 1:42 1:10 1:13
## Mode :character Mode :character Median :1.000
## Mean :1.309
## 3rd Qu.:2.000
## Max. :4.000
##
## ccb a_blocker a_b_blocker moxonidine dose_3 x_anti_chol_used
## 0:51 0 :77 0:79 0:75 Length:81 Min. :0.0000
## 1:30 1 : 3 1: 2 1: 6 Class :character 1st Qu.:0.0000
## NA's: 1 Mode :character Median :0.0000
## Mean :0.5062
## 3rd Qu.:1.0000
## Max. :3.0000
##
## statins fibrates ezetimibe other cholesterol_lowering_therapy
## 0:43 0:80 0:79 Min. :0 Length:81
## 1:38 1: 1 1: 2 1st Qu.:0 Class :character
## Median :0 Mode :character
## Mean :0
## 3rd Qu.:0
## Max. :0
##
## dose_4 x_anti_platelet_therapy aspirin clopidogrel total_number
## Length:81 Min. :0.000 0:60 0:72 Min. :0.000
## Class :character 1st Qu.:2.000 1:21 1: 9 1st Qu.:1.000
## Mode :character Median :3.000 Median :3.000
## Mean :2.901 Mean :2.642
## 3rd Qu.:4.000 3rd Qu.:4.000
## Max. :7.000 Max. :7.000
##
## dose_5 htn t2dm cvd chol ckd ckd_stage
## Length:81 0:23 0:49 0:27 0:45 0:20 Min. :0.000
## Class :character 1:58 1:32 1:54 1:36 1:61 1st Qu.:1.000
## Mode :character Median :2.000
## Mean :2.049
## 3rd Qu.:3.000
## Max. :5.000
##
## concurrent_infections weight height bmi
## Min. :0.0000 Min. : 43.70 Min. : 78.0 Min. :26.25
## 1st Qu.:0.0000 1st Qu.: 72.70 1st Qu.:154.8 1st Qu.:30.39
## Median :0.0000 Median : 81.40 Median :158.0 Median :34.17
## Mean :0.3247 Mean : 82.48 Mean :154.4 Mean :33.78
## 3rd Qu.:1.0000 3rd Qu.: 92.90 3rd Qu.:164.5 3rd Qu.:37.03
## Max. :1.0000 Max. :131.40 Max. :173.0 Max. :42.67
## NA's :4 NA's :22 NA's :67 NA's :67
## obesity x_total_comorbidities x0 x1 x2
## 0 : 3 Min. :0.00 Min. :1 Min. :1 Min. :1
## 1 :15 1st Qu.:2.00 1st Qu.:1 1st Qu.:1 1st Qu.:1
## NA's:63 Median :4.00 Median :1 Median :1 Median :1
## Mean :3.16 Mean :1 Mean :1 Mean :1
## 3rd Qu.:4.00 3rd Qu.:1 3rd Qu.:1 3rd Qu.:1
## Max. :6.00 Max. :1 Max. :1 Max. :1
## NA's :72 NA's :73 NA's :71
## x3 x4 x5 x6 total_3_or_more
## Min. :1 Min. :1 Min. :1 Min. :1 Min. :0.0000
## 1st Qu.:1 1st Qu.:1 1st Qu.:1 1st Qu.:1 1st Qu.:0.0000
## Median :1 Median :1 Median :1 Median :1 Median :1.0000
## Mean :1 Mean :1 Mean :1 Mean :1 Mean :0.6667
## 3rd Qu.:1 3rd Qu.:1 3rd Qu.:1 3rd Qu.:1 3rd Qu.:1.0000
## Max. :1 Max. :1 Max. :1 Max. :1 Max. :1.0000
## NA's :70 NA's :55 NA's :70 NA's :75
## duration_of_therapy clinical_asssesed_improvement patient_reported_improvemnet
## Min. : 0.0 0 : 6 0 : 3
## 1st Qu.: 3.0 1 :55 1 :52
## Median : 6.0 NA's:20 2 : 1
## Mean : 7.5 NA's:25
## 3rd Qu.:10.0
## Max. :30.0
## NA's :3
## ult_on_admission_y_n dose_2_2 commenced_during dose_1_2
## 0:55 Min. : 10.0 0 :43 Min. : 50.0
## 1:26 1st Qu.: 50.0 1 :12 1st Qu.: 50.0
## Median :100.0 NA's:26 Median :100.0
## Mean :128.5 Mean :112.5
## 3rd Qu.:175.0 3rd Qu.:100.0
## Max. :400.0 Max. :300.0
## NA's :55 NA's :69
## after_discharge dose_2_2_2 change_dose old_dose new_dose
## 0 :40 Min. : 40.00 0 :23 Min. : 50.00 Min. :150.0
## 1 :15 1st Qu.: 50.00 1 : 3 1st Qu.: 75.00 1st Qu.:175.0
## NA's:26 Median : 50.00 NA's:55 Median :100.00 Median :200.0
## Mean : 67.27 Mean : 83.33 Mean :183.3
## 3rd Qu.:100.00 3rd Qu.:100.00 3rd Qu.:200.0
## Max. :100.00 Max. :100.00 Max. :200.0
## NA's :70 NA's :78 NA's :78
## ceased_withheld dose_3_2 serum_urate_range serum_urate_level_mmol_l
## 0 :24 Min. :100 Min. :0.000 Min. :0.1300
## 1 : 2 1st Qu.:150 1st Qu.:1.000 1st Qu.:0.4500
## NA's:55 Median :200 Median :2.000 Median :0.5600
## Mean :200 Mean :1.444 Mean :0.5556
## 3rd Qu.:250 3rd Qu.:2.000 3rd Qu.:0.6700
## Max. :300 Max. :2.000 Max. :0.8700
## NA's :79 NA's :20
## interval_to_next_flare d_c_instructions_present discharge_instructions
## Min. :0 0:38 Length:81
## 1st Qu.:0 1:43 Class :character
## Median :0 Mode :character
## Mean :0
## 3rd Qu.:0
## Max. :0
## NA's :68
## factors_complicating_gout_treatment none infections diuresis aki trauma
## Length:81 0:48 0:49 0:73 0:62 0:79
## Class :character 1:33 1:32 1: 8 1:19 1: 2
## Mode :character
##
##
##
##
## cognitive_impairment acidosis other_rheum chemo surgery comments_2
## 0:68 0:80 0:80 0:80 0:79 Length:81
## 1:13 1: 1 1: 1 1: 1 1: 2 Class :character
## Mode :character
##
##
##
## Serum urate vs Already on Urate Lowering Therapy
serum_urate_level_mmol_l vs ult_on_admission_y_n
Boxplot
p <- ggplot(
data = d,
aes(
group = ult_on_admission_y_n,
x = ult_on_admission_y_n,
y = serum_urate_level_mmol_l
)
) + geom_boxplot()
p + stat_summary(fun = mean, geom = "point", shape = 23, size = 4)## Warning: Removed 20 rows containing non-finite values (stat_boxplot).## Warning: Removed 20 rows containing non-finite values (stat_summary).
same plot with highcharter
highchart() %>%
hc_xAxis(type = "category") %>%
hc_add_series_list(
data_to_boxplot(
data = d,
variable = serum_urate_level_mmol_l,
group_var = ult_on_admission_y_n
)
) %>%
hc_exporting(enabled = TRUE) # allow 'save'Wilcox unpaired
(same as Mann-Whitney U)
wilcox.test(
serum_urate_level_mmol_l ~ ult_on_admission_y_n,
data = d,
exact = FALSE, paired = FALSE
)##
## Wilcoxon rank sum test with continuity correction
##
## data: serum_urate_level_mmol_l by ult_on_admission_y_n
## W = 445, p-value = 0.09536
## alternative hypothesis: true location shift is not equal to 0Rheumatology input
Age
d_notRheumPt <- d %>%
filter(grouped_units_of_admission != 0)
p <- ggplot(
data = d_notRheumPt,
aes(
group = rheum_input,
x = rheum_input,
y = age
)
) + geom_boxplot()
p + stat_summary(fun = mean, geom = "point", shape = 23, size = 4)
Unit of admission
ztable(table(d_notRheumPt$rheum_input, d_notRheumPt$grouped_units_of_admission)) %>%
makeHeatmap(palette="Blues")| 0 | 1 | 2 | 3 | 4 | 5 | |
|---|---|---|---|---|---|---|
| 0 | 0 | 14 | 36 | 13 | 1 | 0 |
CrossTable(d_notRheumPt$rheum_input, d_notRheumPt$grouped_units_of_admission,
fisher = TRUE)##
##
## Cell Contents
## |-------------------------|
## | N |
## | Chi-square contribution |
## | N / Row Total |
## | N / Col Total |
## | N / Table Total |
## |-------------------------|
##
##
## Total Observations in Table: 79
##
##
## | d_notRheumPt$grouped_units_of_admission
## d_notRheumPt$rheum_input | 1 | 2 | 3 | 4 | 5 | Row Total |
## -------------------------|-----------|-----------|-----------|-----------|-----------|-----------|
## 0 | 14 | 36 | 13 | 1 | 0 | 64 |
## | 0.281 | 0.043 | 0.059 | 0.237 | 0.810 | |
## | 0.219 | 0.562 | 0.203 | 0.016 | 0.000 | 0.810 |
## | 0.933 | 0.783 | 0.867 | 0.500 | 0.000 | |
## | 0.177 | 0.456 | 0.165 | 0.013 | 0.000 | |
## -------------------------|-----------|-----------|-----------|-----------|-----------|-----------|
## 1 | 1 | 10 | 2 | 1 | 1 | 15 |
## | 1.199 | 0.183 | 0.253 | 1.013 | 3.457 | |
## | 0.067 | 0.667 | 0.133 | 0.067 | 0.067 | 0.190 |
## | 0.067 | 0.217 | 0.133 | 0.500 | 1.000 | |
## | 0.013 | 0.127 | 0.025 | 0.013 | 0.013 | |
## -------------------------|-----------|-----------|-----------|-----------|-----------|-----------|
## Column Total | 15 | 46 | 15 | 2 | 1 | 79 |
## | 0.190 | 0.582 | 0.190 | 0.025 | 0.013 | |
## -------------------------|-----------|-----------|-----------|-----------|-----------|-----------|
##
##
## Fisher's Exact Test for Count Data
## ------------------------------------------------------------
## Alternative hypothesis: two.sided
## p = 0.1492576
##
## p <- ggplot(
data = d_notRheumPt,
aes(
group = grouped_units_of_admission,
x = grouped_units_of_admission,
y = rheum_input
)
) + geom_boxplot()
p + stat_summary(fun = mean, geom = "point", shape = 23, size = 4)
Gender
p <- ggplot(
data = d_notRheumPt,
aes(
group = rheum_input,
x = rheum_input,
y = gender
)
) + geom_boxplot()
p + stat_summary(fun = mean, geom = "point", shape = 23, size = 4)
ztable(table(d_notRheumPt$rheum_input, d_notRheumPt$gender)) %>%
makeHeatmap(palette="Blues")| 0 | 1 | |
|---|---|---|
| 0 | 40 | 24 |
CrossTable(d_notRheumPt$rheum_input, d_notRheumPt$gender,
fisher = TRUE)##
##
## Cell Contents
## |-------------------------|
## | N |
## | Chi-square contribution |
## | N / Row Total |
## | N / Col Total |
## | N / Table Total |
## |-------------------------|
##
##
## Total Observations in Table: 79
##
##
## | d_notRheumPt$gender
## d_notRheumPt$rheum_input | 0 | 1 | Row Total |
## -------------------------|-----------|-----------|-----------|
## 0 | 40 | 24 | 64 |
## | 0.107 | 0.207 | |
## | 0.625 | 0.375 | 0.810 |
## | 0.769 | 0.889 | |
## | 0.506 | 0.304 | |
## -------------------------|-----------|-----------|-----------|
## 1 | 12 | 3 | 15 |
## | 0.458 | 0.882 | |
## | 0.800 | 0.200 | 0.190 |
## | 0.231 | 0.111 | |
## | 0.152 | 0.038 | |
## -------------------------|-----------|-----------|-----------|
## Column Total | 52 | 27 | 79 |
## | 0.658 | 0.342 | |
## -------------------------|-----------|-----------|-----------|
##
##
## Fisher's Exact Test for Count Data
## ------------------------------------------------------------
## Sample estimate odds ratio: 0.420881
##
## Alternative hypothesis: true odds ratio is not equal to 1
## p = 0.2405739
## 95% confidence interval: 0.06926226 1.78356
##
## Alternative hypothesis: true odds ratio is less than 1
## p = 0.1629917
## 95% confidence interval: 0 1.483282
##
## Alternative hypothesis: true odds ratio is greater than 1
## p = 0.9489704
## 95% confidence interval: 0.09274648 Inf
##
##
## Campus
ztable(table(d_notRheumPt$rheum_input, d_notRheumPt$hospital_campus)) %>%
makeHeatmap(palette="Blues")| 0 | 1 | 2 | |
|---|---|---|---|
| 0 | 16 | 46 | 2 |
CrossTable(d_notRheumPt$rheum_input, d_notRheumPt$hospital_campus,
fisher = TRUE)##
##
## Cell Contents
## |-------------------------|
## | N |
## | Chi-square contribution |
## | N / Row Total |
## | N / Col Total |
## | N / Table Total |
## |-------------------------|
##
##
## Total Observations in Table: 79
##
##
## | d_notRheumPt$hospital_campus
## d_notRheumPt$rheum_input | 0 | 1 | 2 | Row Total |
## -------------------------|-----------|-----------|-----------|-----------|
## 0 | 16 | 46 | 2 | 64 |
## | 1.577 | 0.745 | 0.089 | |
## | 0.250 | 0.719 | 0.031 | 0.810 |
## | 0.593 | 0.920 | 1.000 | |
## | 0.203 | 0.582 | 0.025 | |
## -------------------------|-----------|-----------|-----------|-----------|
## 1 | 11 | 4 | 0 | 15 |
## | 6.729 | 3.179 | 0.380 | |
## | 0.733 | 0.267 | 0.000 | 0.190 |
## | 0.407 | 0.080 | 0.000 | |
## | 0.139 | 0.051 | 0.000 | |
## -------------------------|-----------|-----------|-----------|-----------|
## Column Total | 27 | 50 | 2 | 79 |
## | 0.342 | 0.633 | 0.025 | |
## -------------------------|-----------|-----------|-----------|-----------|
##
##
## Fisher's Exact Test for Count Data
## ------------------------------------------------------------
## Alternative hypothesis: two.sided
## p = 0.001793077
##
## Generalized Linear Model
rheum_consulted <- glm(
rheum_input ~ gender + age + hopc +
grouped_units_of_admission + duration_of_stay +
number_of_active_joints_involved +
mtp + knee + ankle + elbow + wrist + mcp + tophi + hospital_campus,
data = d_notRheumPt,
family = "binomial"
)
summary(rheum_consulted)##
## Call:
## glm(formula = rheum_input ~ gender + age + hopc + grouped_units_of_admission +
## duration_of_stay + number_of_active_joints_involved + mtp +
## knee + ankle + elbow + wrist + mcp + tophi + hospital_campus,
## family = "binomial", data = d_notRheumPt)
##
## Deviance Residuals:
## Min 1Q Median 3Q Max
## -1.7833 -0.3433 -0.1003 -0.0021 2.5494
##
## Coefficients:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) -0.28625 4.17324 -0.069 0.94531
## gender1 -0.66057 1.33779 -0.494 0.62147
## age -0.11567 0.06245 -1.852 0.06400 .
## hopc2 -2.45800 1.47677 -1.664 0.09602 .
## grouped_units_of_admission2 8.81032 5.13955 1.714 0.08649 .
## grouped_units_of_admission3 7.59034 4.73530 1.603 0.10895
## grouped_units_of_admission4 5.47775 4.41705 1.240 0.21492
## grouped_units_of_admission5 29.36716 3956.18437 0.007 0.99408
## duration_of_stay 0.06196 0.05161 1.200 0.22995
## number_of_active_joints_involved -0.10115 1.04492 -0.097 0.92289
## mtp1 2.46534 2.09531 1.177 0.23936
## knee1 2.34039 1.74589 1.341 0.18008
## ankle1 -0.31765 1.20794 -0.263 0.79257
## elbow1 1.32866 1.76821 0.751 0.45240
## wrist1 -0.98977 2.78596 -0.355 0.72239
## mcp1 1.40558 2.18495 0.643 0.52003
## tophi1 0.57479 1.98761 0.289 0.77244
## hospital_campus1 -3.84722 1.40344 -2.741 0.00612 **
## hospital_campus2 -17.21744 2738.76798 -0.006 0.99498
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for binomial family taken to be 1)
##
## Null deviance: 75.503 on 75 degrees of freedom
## Residual deviance: 37.096 on 57 degrees of freedom
## (3 observations deleted due to missingness)
## AIC: 75.096
##
## Number of Fisher Scoring iterations: 16predictors <- d_notRheumPt[,
c("gender", "hopc", "grouped_units_of_admission", "duration_of_stay",
"number_of_active_joints_involved", "mtp", "knee", "ankle",
"elbow", "wrist", "mcp", "tophi")
] %>%
mutate_all(function(x) as.numeric(as.character(x))) %>%
rename(c("admission_unit" = "grouped_units_of_admission", "number_of_joints" = "number_of_active_joints_involved"))
rheum_consult_predictor_cor <- cor(predictors, use = "complete.obs") %>% round(3)corrplot(rheum_consult_predictor_cor,
type = "upper", order = "hclust",
tl.col = "black", tl.srt = 45)
number_of_active_joints_involved heavily correlated with other variables (e.g. specific active joints), and little role in the model so far. Remove from model
rheum_consulted2 <- glm(
rheum_input ~ gender + age + hopc +
grouped_units_of_admission + duration_of_stay +
mtp + knee + ankle + elbow + wrist + mcp + tophi + hospital_campus,
data = d_notRheumPt,
family = "binomial"
)
summary(rheum_consulted2)##
## Call:
## glm(formula = rheum_input ~ gender + age + hopc + grouped_units_of_admission +
## duration_of_stay + mtp + knee + ankle + elbow + wrist + mcp +
## tophi + hospital_campus, family = "binomial", data = d_notRheumPt)
##
## Deviance Residuals:
## Min 1Q Median 3Q Max
## -1.78574 -0.33979 -0.10061 -0.00203 2.53697
##
## Coefficients:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) -0.32087 4.17283 -0.077 0.93871
## gender1 -0.65035 1.33132 -0.488 0.62520
## age -0.11545 0.06242 -1.850 0.06436 .
## hopc2 -2.48505 1.45719 -1.705 0.08813 .
## grouped_units_of_admission2 8.80777 5.13768 1.714 0.08646 .
## grouped_units_of_admission3 7.64779 4.70608 1.625 0.10414
## grouped_units_of_admission4 5.46766 4.40872 1.240 0.21490
## grouped_units_of_admission5 29.43561 3956.18433 0.007 0.99406
## duration_of_stay 0.06258 0.05120 1.222 0.22161
## mtp1 2.33793 1.61774 1.445 0.14841
## knee1 2.23359 1.34621 1.659 0.09708 .
## ankle1 -0.38842 0.96136 -0.404 0.68619
## elbow1 1.25526 1.60128 0.784 0.43309
## wrist1 -1.18133 2.00109 -0.590 0.55496
## mcp1 1.28271 1.78640 0.718 0.47273
## tophi1 0.52529 1.94328 0.270 0.78692
## hospital_campus1 -3.82934 1.39009 -2.755 0.00587 **
## hospital_campus2 -17.18216 2726.17009 -0.006 0.99497
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for binomial family taken to be 1)
##
## Null deviance: 75.503 on 75 degrees of freedom
## Residual deviance: 37.105 on 58 degrees of freedom
## (3 observations deleted due to missingness)
## AIC: 73.105
##
## Number of Fisher Scoring iterations: 16MTP, MCP, wrist and knee heavily correlated. Wrist has highest p-value, remove.
rheum_consulted3 <- glm(
rheum_input ~ gender + age + hopc +
grouped_units_of_admission + duration_of_stay +
mtp + knee + ankle + elbow + mcp + tophi + hospital_campus,
data = d_notRheumPt,
family = "binomial"
)
summary(rheum_consulted3)##
## Call:
## glm(formula = rheum_input ~ gender + age + hopc + grouped_units_of_admission +
## duration_of_stay + mtp + knee + ankle + elbow + mcp + tophi +
## hospital_campus, family = "binomial", data = d_notRheumPt)
##
## Deviance Residuals:
## Min 1Q Median 3Q Max
## -1.77035 -0.41455 -0.08472 -0.00324 2.66260
##
## Coefficients:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) 0.27945 3.91236 0.071 0.94306
## gender1 -0.63887 1.32724 -0.481 0.63027
## age -0.11671 0.06286 -1.857 0.06336 .
## hopc2 -2.67012 1.44774 -1.844 0.06513 .
## grouped_units_of_admission2 8.42781 4.96895 1.696 0.08987 .
## grouped_units_of_admission3 7.48614 4.60485 1.626 0.10401
## grouped_units_of_admission4 5.18832 4.27806 1.213 0.22522
## grouped_units_of_admission5 29.48513 3956.18410 0.007 0.99405
## duration_of_stay 0.06305 0.05092 1.238 0.21561
## mtp1 2.38834 1.53796 1.553 0.12044
## knee1 2.00904 1.23129 1.632 0.10275
## ankle1 -0.39640 0.95221 -0.416 0.67719
## elbow1 1.04387 1.45840 0.716 0.47414
## mcp1 0.76786 1.49519 0.514 0.60757
## tophi1 0.80929 1.74321 0.464 0.64247
## hospital_campus1 -4.14402 1.36583 -3.034 0.00241 **
## hospital_campus2 -17.48335 2774.45440 -0.006 0.99497
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for binomial family taken to be 1)
##
## Null deviance: 75.503 on 75 degrees of freedom
## Residual deviance: 37.508 on 59 degrees of freedom
## (3 observations deleted due to missingness)
## AIC: 71.508
##
## Number of Fisher Scoring iterations: 16Remove tophi
rheum_consulted4 <- glm(
rheum_input ~ gender + age + hopc +
grouped_units_of_admission + duration_of_stay +
mtp + knee + ankle + elbow + mcp + hospital_campus,
data = d_notRheumPt,
family = "binomial"
)
summary(rheum_consulted4)##
## Call:
## glm(formula = rheum_input ~ gender + age + hopc + grouped_units_of_admission +
## duration_of_stay + mtp + knee + ankle + elbow + mcp + hospital_campus,
## family = "binomial", data = d_notRheumPt)
##
## Deviance Residuals:
## Min 1Q Median 3Q Max
## -1.69306 -0.33141 -0.09629 -0.00420 2.73713
##
## Coefficients:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) 0.24670 3.66209 0.067 0.94629
## gender1 -0.81242 1.27201 -0.639 0.52303
## age -0.10922 0.05896 -1.852 0.06397 .
## hopc2 -2.66329 1.44707 -1.840 0.06570 .
## grouped_units_of_admission2 7.98936 4.61115 1.733 0.08316 .
## grouped_units_of_admission3 7.08557 4.25130 1.667 0.09558 .
## grouped_units_of_admission4 4.94464 3.99790 1.237 0.21616
## grouped_units_of_admission5 29.60474 3956.18388 0.007 0.99403
## duration_of_stay 0.05698 0.04842 1.177 0.23926
## mtp1 2.47558 1.56524 1.582 0.11374
## knee1 1.90612 1.21347 1.571 0.11623
## ankle1 -0.38933 0.94592 -0.412 0.68064
## elbow1 1.32982 1.32276 1.005 0.31473
## mcp1 0.74476 1.48427 0.502 0.61583
## hospital_campus1 -4.01107 1.28938 -3.111 0.00187 **
## hospital_campus2 -17.35513 2755.82841 -0.006 0.99498
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for binomial family taken to be 1)
##
## Null deviance: 75.503 on 75 degrees of freedom
## Residual deviance: 37.719 on 60 degrees of freedom
## (3 observations deleted due to missingness)
## AIC: 69.719
##
## Number of Fisher Scoring iterations: 16Remove ankle
rheum_consulted5 <- glm(
rheum_input ~ gender + age + hopc +
grouped_units_of_admission + duration_of_stay +
mtp + knee + elbow + mcp + hospital_campus,
data = d_notRheumPt,
family = "binomial"
)
summary(rheum_consulted5)##
## Call:
## glm(formula = rheum_input ~ gender + age + hopc + grouped_units_of_admission +
## duration_of_stay + mtp + knee + elbow + mcp + hospital_campus,
## family = "binomial", data = d_notRheumPt)
##
## Deviance Residuals:
## Min 1Q Median 3Q Max
## -1.66930 -0.32910 -0.09162 -0.00376 2.73932
##
## Coefficients:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) 0.10480 3.50344 0.030 0.97614
## gender1 -0.75566 1.26347 -0.598 0.54979
## age -0.11025 0.05882 -1.874 0.06090 .
## hopc2 -2.68787 1.44305 -1.863 0.06251 .
## grouped_units_of_admission2 8.09503 4.55939 1.775 0.07582 .
## grouped_units_of_admission3 7.22488 4.17178 1.732 0.08330 .
## grouped_units_of_admission4 4.72190 3.85917 1.224 0.22112
## grouped_units_of_admission5 29.88186 3956.18382 0.008 0.99397
## duration_of_stay 0.05570 0.04920 1.132 0.25755
## mtp1 2.46803 1.56690 1.575 0.11523
## knee1 1.94319 1.20620 1.611 0.10718
## elbow1 1.33800 1.29972 1.029 0.30327
## mcp1 0.71053 1.45453 0.488 0.62520
## hospital_campus1 -4.02251 1.29698 -3.101 0.00193 **
## hospital_campus2 -17.22134 2750.95865 -0.006 0.99501
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for binomial family taken to be 1)
##
## Null deviance: 75.503 on 75 degrees of freedom
## Residual deviance: 37.892 on 61 degrees of freedom
## (3 observations deleted due to missingness)
## AIC: 67.892
##
## Number of Fisher Scoring iterations: 16Remove mcp
rheum_consulted6 <- glm(
rheum_input ~ gender + age + hopc +
grouped_units_of_admission + duration_of_stay +
mtp + knee + elbow + hospital_campus,
data = d_notRheumPt,
family = "binomial"
)
summary(rheum_consulted6)##
## Call:
## glm(formula = rheum_input ~ gender + age + hopc + grouped_units_of_admission +
## duration_of_stay + mtp + knee + elbow + hospital_campus,
## family = "binomial", data = d_notRheumPt)
##
## Deviance Residuals:
## Min 1Q Median 3Q Max
## -1.74568 -0.34874 -0.10401 -0.00616 2.59590
##
## Coefficients:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) 0.37799 3.28264 0.115 0.90833
## gender1 -0.67582 1.23210 -0.549 0.58334
## age -0.10085 0.05379 -1.875 0.06082 .
## hopc2 -2.58757 1.41001 -1.835 0.06648 .
## grouped_units_of_admission2 7.22730 3.87316 1.866 0.06204 .
## grouped_units_of_admission3 6.57193 3.66833 1.792 0.07321 .
## grouped_units_of_admission4 4.06899 3.37426 1.206 0.22786
## grouped_units_of_admission5 29.39374 3956.18329 0.007 0.99407
## duration_of_stay 0.05361 0.04785 1.120 0.26261
## mtp1 2.15235 1.37762 1.562 0.11820
## knee1 1.90839 1.19544 1.596 0.11040
## elbow1 1.41440 1.27651 1.108 0.26785
## hospital_campus1 -3.88066 1.22519 -3.167 0.00154 **
## hospital_campus2 -17.43573 2761.94026 -0.006 0.99496
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for binomial family taken to be 1)
##
## Null deviance: 75.503 on 75 degrees of freedom
## Residual deviance: 38.128 on 62 degrees of freedom
## (3 observations deleted due to missingness)
## AIC: 66.128
##
## Number of Fisher Scoring iterations: 16Remove duration of stay
rheum_consulted7 <- glm(
rheum_input ~ gender + age + hopc +
grouped_units_of_admission +
mtp + knee + elbow + hospital_campus,
data = d_notRheumPt,
family = "binomial"
)
summary(rheum_consulted7)##
## Call:
## glm(formula = rheum_input ~ gender + age + hopc + grouped_units_of_admission +
## mtp + knee + elbow + hospital_campus, family = "binomial",
## data = d_notRheumPt)
##
## Deviance Residuals:
## Min 1Q Median 3Q Max
## -1.5969 -0.3925 -0.1047 -0.0082 2.7805
##
## Coefficients:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) -0.23106 3.15990 -0.073 0.94171
## gender1 -1.12352 1.19045 -0.944 0.34528
## age -0.08749 0.04969 -1.761 0.07830 .
## hopc2 -1.87108 1.18388 -1.580 0.11400
## grouped_units_of_admission2 6.98348 3.70982 1.882 0.05978 .
## grouped_units_of_admission3 5.87455 3.37290 1.742 0.08156 .
## grouped_units_of_admission4 4.22140 3.31672 1.273 0.20310
## grouped_units_of_admission5 28.81446 3956.18298 0.007 0.99419
## mtp1 2.43201 1.35822 1.791 0.07336 .
## knee1 1.81182 1.15403 1.570 0.11642
## elbow1 1.24849 1.24201 1.005 0.31479
## hospital_campus1 -3.66839 1.17593 -3.120 0.00181 **
## hospital_campus2 -16.60371 2639.30167 -0.006 0.99498
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for binomial family taken to be 1)
##
## Null deviance: 75.503 on 75 degrees of freedom
## Residual deviance: 39.485 on 63 degrees of freedom
## (3 observations deleted due to missingness)
## AIC: 65.485
##
## Number of Fisher Scoring iterations: 16remove age
rheum_consulted8 <- glm(
rheum_input ~ gender + hopc +
grouped_units_of_admission +
mtp + knee + elbow + hospital_campus,
data = d_notRheumPt,
family = "binomial"
)
summary(rheum_consulted8)##
## Call:
## glm(formula = rheum_input ~ gender + hopc + grouped_units_of_admission +
## mtp + knee + elbow + hospital_campus, family = "binomial",
## data = d_notRheumPt)
##
## Deviance Residuals:
## Min 1Q Median 3Q Max
## -1.51242 -0.39440 -0.15725 -0.05777 2.83855
##
## Coefficients:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) -3.5390 2.3644 -1.497 0.13445
## gender1 -1.2200 1.1067 -1.102 0.27029
## hopc2 -0.8651 0.9921 -0.872 0.38322
## grouped_units_of_admission2 2.7863 2.0730 1.344 0.17892
## grouped_units_of_admission3 2.9564 2.3556 1.255 0.20947
## grouped_units_of_admission4 4.2105 3.2180 1.308 0.19074
## grouped_units_of_admission5 23.9609 3956.1811 0.006 0.99517
## mtp1 1.7887 1.1239 1.592 0.11148
## knee1 1.5128 1.0118 1.495 0.13488
## elbow1 1.8809 1.1539 1.630 0.10309
## hospital_campus1 -2.8558 0.9540 -2.993 0.00276 **
## hospital_campus2 -16.5539 2778.3657 -0.006 0.99525
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for binomial family taken to be 1)
##
## Null deviance: 75.503 on 75 degrees of freedom
## Residual deviance: 43.120 on 64 degrees of freedom
## (3 observations deleted due to missingness)
## AIC: 67.12
##
## Number of Fisher Scoring iterations: 16remove admission group
rheum_consulted9 <- glm(
rheum_input ~ gender + hopc +
mtp + knee + elbow + hospital_campus,
data = d_notRheumPt,
family = "binomial"
)
summary(rheum_consulted9)##
## Call:
## glm(formula = rheum_input ~ gender + hopc + mtp + knee + elbow +
## hospital_campus, family = "binomial", data = d_notRheumPt)
##
## Deviance Residuals:
## Min 1Q Median 3Q Max
## -1.4369 -0.5732 -0.3192 -0.1197 2.8192
##
## Coefficients:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) -0.3887 0.9682 -0.402 0.68803
## gender1 -1.3344 0.9641 -1.384 0.16633
## hopc2 -0.6497 0.7606 -0.854 0.39294
## mtp1 1.0968 0.8454 1.297 0.19454
## knee1 0.9808 0.8192 1.197 0.23119
## elbow1 1.4818 0.9799 1.512 0.13047
## hospital_campus1 -2.5629 0.8347 -3.070 0.00214 **
## hospital_campus2 -15.7524 1659.7199 -0.009 0.99243
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for binomial family taken to be 1)
##
## Null deviance: 75.503 on 75 degrees of freedom
## Residual deviance: 53.501 on 68 degrees of freedom
## (3 observations deleted due to missingness)
## AIC: 69.501
##
## Number of Fisher Scoring iterations: 15just some joints and campus
rheum_consulted10 <- glm(
rheum_input ~ mtp + knee + elbow + hospital_campus,
data = d_notRheumPt,
family = "binomial"
)
summary(rheum_consulted10)##
## Call:
## glm(formula = rheum_input ~ mtp + knee + elbow + hospital_campus,
## family = "binomial", data = d_notRheumPt)
##
## Deviance Residuals:
## Min 1Q Median 3Q Max
## -1.4504 -0.6155 -0.3526 -0.2531 2.6294
##
## Coefficients:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) -1.2335 0.7119 -1.733 0.08318 .
## mtp1 0.6779 0.7786 0.871 0.38395
## knee1 1.1778 0.8083 1.457 0.14508
## elbow1 1.8609 0.9621 1.934 0.05308 .
## hospital_campus1 -2.1913 0.7327 -2.991 0.00278 **
## hospital_campus2 -16.0281 1648.5621 -0.010 0.99224
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for binomial family taken to be 1)
##
## Null deviance: 75.503 on 75 degrees of freedom
## Residual deviance: 55.936 on 70 degrees of freedom
## (3 observations deleted due to missingness)
## AIC: 67.936
##
## Number of Fisher Scoring iterations: 15just grouped units of admission and campus
rheum_consulted11 <- glm(
rheum_input ~ grouped_units_of_admission + hospital_campus,
data = d_notRheumPt,
family = "binomial"
)
summary(rheum_consulted11)##
## Call:
## glm(formula = rheum_input ~ grouped_units_of_admission + hospital_campus,
## family = "binomial", data = d_notRheumPt)
##
## Deviance Residuals:
## Min 1Q Median 3Q Max
## -1.1185 -0.4018 -0.4018 -0.2113 2.5127
##
## Coefficients:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) -1.4548 1.1027 -1.319 0.18708
## grouped_units_of_admission2 1.3147 1.1565 1.137 0.25564
## grouped_units_of_admission3 0.6773 1.3592 0.498 0.61826
## grouped_units_of_admission4 2.6227 1.9909 1.317 0.18773
## grouped_units_of_admission5 21.3567 3956.1805 0.005 0.99569
## hospital_campus1 -2.3359 0.7286 -3.206 0.00135 **
## hospital_campus2 -17.4260 2797.4420 -0.006 0.99503
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for binomial family taken to be 1)
##
## Null deviance: 76.794 on 78 degrees of freedom
## Residual deviance: 56.369 on 72 degrees of freedom
## AIC: 70.369
##
## Number of Fisher Scoring iterations: 16CrossTable(d_notRheumPt$number_of_active_joints_involved, d_notRheumPt$rheum_input, fisher = TRUE)##
##
## Cell Contents
## |-------------------------|
## | N |
## | Chi-square contribution |
## | N / Row Total |
## | N / Col Total |
## | N / Table Total |
## |-------------------------|
##
##
## Total Observations in Table: 79
##
##
## | d_notRheumPt$rheum_input
## d_notRheumPt$number_of_active_joints_involved | 0 | 1 | Row Total |
## ----------------------------------------------|-----------|-----------|-----------|
## 1 | 38 | 7 | 45 |
## | 0.065 | 0.279 | |
## | 0.844 | 0.156 | 0.570 |
## | 0.594 | 0.467 | |
## | 0.481 | 0.089 | |
## ----------------------------------------------|-----------|-----------|-----------|
## 2 | 19 | 5 | 24 |
## | 0.010 | 0.043 | |
## | 0.792 | 0.208 | 0.304 |
## | 0.297 | 0.333 | |
## | 0.241 | 0.063 | |
## ----------------------------------------------|-----------|-----------|-----------|
## 3 | 3 | 1 | 4 |
## | 0.018 | 0.076 | |
## | 0.750 | 0.250 | 0.051 |
## | 0.047 | 0.067 | |
## | 0.038 | 0.013 | |
## ----------------------------------------------|-----------|-----------|-----------|
## 4 | 2 | 0 | 2 |
## | 0.089 | 0.380 | |
## | 1.000 | 0.000 | 0.025 |
## | 0.031 | 0.000 | |
## | 0.025 | 0.000 | |
## ----------------------------------------------|-----------|-----------|-----------|
## 5 | 2 | 2 | 4 |
## | 0.475 | 2.026 | |
## | 0.500 | 0.500 | 0.051 |
## | 0.031 | 0.133 | |
## | 0.025 | 0.025 | |
## ----------------------------------------------|-----------|-----------|-----------|
## Column Total | 64 | 15 | 79 |
## | 0.810 | 0.190 | |
## ----------------------------------------------|-----------|-----------|-----------|
##
##
## Fisher's Exact Test for Count Data
## ------------------------------------------------------------
## Alternative hypothesis: two.sided
## p = 0.3854947
##
## rheum_consulted12 <- glm(
rheum_input ~ number_of_active_joints_involved + hospital_campus,
data = d_notRheumPt,
family = "binomial"
)
summary(rheum_consulted12)##
## Call:
## glm(formula = rheum_input ~ number_of_active_joints_involved +
## hospital_campus, family = "binomial", data = d_notRheumPt)
##
## Deviance Residuals:
## Min 1Q Median 3Q Max
## -1.1944 -0.5314 -0.3565 -0.3565 2.3611
##
## Coefficients:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) -0.9586 0.6286 -1.525 0.12731
## number_of_active_joints_involved 0.3328 0.2787 1.194 0.23241
## hospital_campus1 -2.0980 0.6633 -3.163 0.00156 **
## hospital_campus2 -15.9403 1696.7344 -0.009 0.99250
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for binomial family taken to be 1)
##
## Null deviance: 76.794 on 78 degrees of freedom
## Residual deviance: 62.973 on 75 degrees of freedom
## AIC: 70.973
##
## Number of Fisher Scoring iterations: 15Gout therapy
CrossTable(d$prednisolone, d$rheum_input, fisher = TRUE)##
##
## Cell Contents
## |-------------------------|
## | N |
## | Chi-square contribution |
## | N / Row Total |
## | N / Col Total |
## | N / Table Total |
## |-------------------------|
##
##
## Total Observations in Table: 81
##
##
## | d$rheum_input
## d$prednisolone | 0 | 1 | Row Total |
## ---------------|-----------|-----------|-----------|
## 0 | 23 | 6 | 29 |
## | 0.000 | 0.001 | |
## | 0.793 | 0.207 | 0.358 |
## | 0.359 | 0.353 | |
## | 0.284 | 0.074 | |
## ---------------|-----------|-----------|-----------|
## 1 | 41 | 11 | 52 |
## | 0.000 | 0.001 | |
## | 0.788 | 0.212 | 0.642 |
## | 0.641 | 0.647 | |
## | 0.506 | 0.136 | |
## ---------------|-----------|-----------|-----------|
## Column Total | 64 | 17 | 81 |
## | 0.790 | 0.210 | |
## ---------------|-----------|-----------|-----------|
##
##
## Fisher's Exact Test for Count Data
## ------------------------------------------------------------
## Sample estimate odds ratio: 1.028103
##
## Alternative hypothesis: true odds ratio is not equal to 1
## p = 1
## 95% confidence interval: 0.2996486 3.852664
##
## Alternative hypothesis: true odds ratio is less than 1
## p = 0.624778
## 95% confidence interval: 0 3.158698
##
## Alternative hypothesis: true odds ratio is greater than 1
## p = 0.5986041
## 95% confidence interval: 0.3556597 Inf
##
##
## CrossTable(d$ia_steroid_injections, d$rheum_input, fisher = TRUE)##
##
## Cell Contents
## |-------------------------|
## | N |
## | Chi-square contribution |
## | N / Row Total |
## | N / Col Total |
## | N / Table Total |
## |-------------------------|
##
##
## Total Observations in Table: 81
##
##
## | d$rheum_input
## d$ia_steroid_injections | 0 | 1 | Row Total |
## ------------------------|-----------|-----------|-----------|
## 0 | 64 | 15 | 79 |
## | 0.040 | 0.151 | |
## | 0.810 | 0.190 | 0.975 |
## | 1.000 | 0.882 | |
## | 0.790 | 0.185 | |
## ------------------------|-----------|-----------|-----------|
## 1 | 0 | 2 | 2 |
## | 1.580 | 5.949 | |
## | 0.000 | 1.000 | 0.025 |
## | 0.000 | 0.118 | |
## | 0.000 | 0.025 | |
## ------------------------|-----------|-----------|-----------|
## Column Total | 64 | 17 | 81 |
## | 0.790 | 0.210 | |
## ------------------------|-----------|-----------|-----------|
##
##
## Fisher's Exact Test for Count Data
## ------------------------------------------------------------
## Sample estimate odds ratio: Inf
##
## Alternative hypothesis: true odds ratio is not equal to 1
## p = 0.04197531
## 95% confidence interval: 0.7275513 Inf
##
## Alternative hypothesis: true odds ratio is less than 1
## p = 1
## 95% confidence interval: 0 Inf
##
## Alternative hypothesis: true odds ratio is greater than 1
## p = 0.04197531
## 95% confidence interval: 1.118547 Inf
##
##
## hopc <- glm(
ia_steroid_injections ~ rheum_input,
data = d,
family = "binomial"
)
summary(hopc)##
## Call:
## glm(formula = ia_steroid_injections ~ rheum_input, family = "binomial",
## data = d)
##
## Deviance Residuals:
## Min 1Q Median 3Q Max
## -0.50033 -0.00003 -0.00003 -0.00003 2.06885
##
## Coefficients:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) -21.57 3654.05 -0.006 0.995
## rheum_input1 19.55 3654.05 0.005 0.996
##
## (Dispersion parameter for binomial family taken to be 1)
##
## Null deviance: 18.755 on 80 degrees of freedom
## Residual deviance: 12.315 on 79 degrees of freedom
## AIC: 16.315
##
## Number of Fisher Scoring iterations: 20CrossTable(d$ia_steroid_injections, d$rheum_input, fisher = TRUE)##
##
## Cell Contents
## |-------------------------|
## | N |
## | Chi-square contribution |
## | N / Row Total |
## | N / Col Total |
## | N / Table Total |
## |-------------------------|
##
##
## Total Observations in Table: 81
##
##
## | d$rheum_input
## d$ia_steroid_injections | 0 | 1 | Row Total |
## ------------------------|-----------|-----------|-----------|
## 0 | 64 | 15 | 79 |
## | 0.040 | 0.151 | |
## | 0.810 | 0.190 | 0.975 |
## | 1.000 | 0.882 | |
## | 0.790 | 0.185 | |
## ------------------------|-----------|-----------|-----------|
## 1 | 0 | 2 | 2 |
## | 1.580 | 5.949 | |
## | 0.000 | 1.000 | 0.025 |
## | 0.000 | 0.118 | |
## | 0.000 | 0.025 | |
## ------------------------|-----------|-----------|-----------|
## Column Total | 64 | 17 | 81 |
## | 0.790 | 0.210 | |
## ------------------------|-----------|-----------|-----------|
##
##
## Fisher's Exact Test for Count Data
## ------------------------------------------------------------
## Sample estimate odds ratio: Inf
##
## Alternative hypothesis: true odds ratio is not equal to 1
## p = 0.04197531
## 95% confidence interval: 0.7275513 Inf
##
## Alternative hypothesis: true odds ratio is less than 1
## p = 1
## 95% confidence interval: 0 Inf
##
## Alternative hypothesis: true odds ratio is greater than 1
## p = 0.04197531
## 95% confidence interval: 1.118547 Inf
##
##
## Secondary gout
medications - sub-groups
hopc <- glm(
hopc ~ thiazide_diuretics + loop_diuretics + k_sparing_diuretics + other_diuretics + b_blocker + ac_ei + arb + ccb + a_blocker + a_b_blocker + moxonidine + statins + fibrates + ezetimibe + aspirin + clopidogrel,
data = d,
family = "binomial"
)
summary(hopc)##
## Call:
## glm(formula = hopc ~ thiazide_diuretics + loop_diuretics + k_sparing_diuretics +
## other_diuretics + b_blocker + ac_ei + arb + ccb + a_blocker +
## a_b_blocker + moxonidine + statins + fibrates + ezetimibe +
## aspirin + clopidogrel, family = "binomial", data = d)
##
## Deviance Residuals:
## Min 1Q Median 3Q Max
## -2.0637 -0.8569 0.4377 0.9571 1.7163
##
## Coefficients: (1 not defined because of singularities)
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) -0.81275 0.47640 -1.706 0.088 .
## thiazide_diuretics1 -0.04582 0.76115 -0.060 0.952
## loop_diuretics1 0.86983 0.73998 1.175 0.240
## k_sparing_diuretics1 0.50609 1.02291 0.495 0.621
## other_diuretics NA NA NA NA
## b_blocker1 1.06880 0.74852 1.428 0.153
## ac_ei1 -0.81184 0.86521 -0.938 0.348
## arb1 0.05892 0.75315 0.078 0.938
## ccb1 0.47112 0.60193 0.783 0.434
## a_blocker1 17.08547 2703.61784 0.006 0.995
## a_b_blocker1 -0.30652 2.07750 -0.148 0.883
## moxonidine1 -0.17586 1.17114 -0.150 0.881
## statins1 -0.58266 0.67757 -0.860 0.390
## fibrates1 18.18195 6213.87804 0.003 0.998
## ezetimibe1 -1.30688 4791.75481 0.000 1.000
## aspirin1 0.74546 0.75919 0.982 0.326
## clopidogrel1 0.71422 1.28213 0.557 0.577
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for binomial family taken to be 1)
##
## Null deviance: 109.650 on 79 degrees of freedom
## Residual deviance: 87.632 on 64 degrees of freedom
## (1 observation deleted due to missingness)
## AIC: 119.63
##
## Number of Fisher Scoring iterations: 16medication - major groups
hopc2 <- glm(
hopc ~ (x_diuretics_used > 0) + (x_anti_htn > 0) + (x_anti_chol_used > 0) + (x_anti_platelet_therapy > 0),
data = d,
family = "binomial"
)
summary(hopc2)##
## Call:
## glm(formula = hopc ~ (x_diuretics_used > 0) + (x_anti_htn > 0) +
## (x_anti_chol_used > 0) + (x_anti_platelet_therapy > 0), family = "binomial",
## data = d)
##
## Deviance Residuals:
## Min 1Q Median 3Q Max
## -1.7660 -1.1241 0.6871 0.9319 1.9728
##
## Coefficients:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) -1.7918 0.7638 -2.346 0.019 *
## x_diuretics_used > 0TRUE 0.7359 0.5876 1.252 0.210
## x_anti_htn > 0TRUE 1.3233 1.5018 0.881 0.378
## x_anti_chol_used > 0TRUE -0.7141 0.5649 -1.264 0.206
## x_anti_platelet_therapy > 0TRUE 1.0558 1.7113 0.617 0.537
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for binomial family taken to be 1)
##
## Null deviance: 110.791 on 80 degrees of freedom
## Residual deviance: 94.576 on 76 degrees of freedom
## AIC: 104.58
##
## Number of Fisher Scoring iterations: 4conditions
hopc3 <- glm(
hopc ~ htn + t2dm + cvd + chol + ckd,
data = d,
family = "binomial"
)
# not including obesity
summary(hopc3)##
## Call:
## glm(formula = hopc ~ htn + t2dm + cvd + chol + ckd, family = "binomial",
## data = d)
##
## Deviance Residuals:
## Min 1Q Median 3Q Max
## -1.7925 -0.9102 0.7158 0.9316 1.7590
##
## Coefficients:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) -0.88609 0.59028 -1.501 0.13332
## htn1 0.35097 0.57476 0.611 0.54144
## t2dm1 0.62101 0.57099 1.088 0.27677
## cvd1 1.89829 0.64642 2.937 0.00332 **
## chol1 0.01944 0.56125 0.035 0.97237
## ckd1 -0.77248 0.76730 -1.007 0.31405
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for binomial family taken to be 1)
##
## Null deviance: 110.791 on 80 degrees of freedom
## Residual deviance: 96.053 on 75 degrees of freedom
## AIC: 108.05
##
## Number of Fisher Scoring iterations: 4hopc_ckd <- glm(
hopc ~ ckd,
data = d,
family = "binomial"
)
# not including obesity
summary(hopc_ckd)##
## Call:
## glm(formula = hopc ~ ckd, family = "binomial", data = d)
##
## Deviance Residuals:
## Min 1Q Median 3Q Max
## -1.366 -1.366 1.000 1.000 1.264
##
## Coefficients:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) -0.2007 0.4495 -0.446 0.655
## ckd1 0.6335 0.5203 1.218 0.223
##
## (Dispersion parameter for binomial family taken to be 1)
##
## Null deviance: 110.79 on 80 degrees of freedom
## Residual deviance: 109.30 on 79 degrees of freedom
## AIC: 113.3
##
## Number of Fisher Scoring iterations: 4co-morbidities
n_comorbidities_hopc <- table(d$x_total_comorbidities, d$hopc)
ztable(n_comorbidities_hopc) %>%
makeHeatmap(palette="Blues")| 1 | 2 | |
|---|---|---|
| 0 | 8 | 1 |
CrossTable(d$x_total_comorbidities, d$hopc,
fisher = TRUE)##
##
## Cell Contents
## |-------------------------|
## | N |
## | Chi-square contribution |
## | N / Row Total |
## | N / Col Total |
## | N / Table Total |
## |-------------------------|
##
##
## Total Observations in Table: 81
##
##
## | d$hopc
## d$x_total_comorbidities | 1 | 2 | Row Total |
## ------------------------|-----------|-----------|-----------|
## 0 | 8 | 1 | 9 |
## | 4.346 | 3.307 | |
## | 0.889 | 0.111 | 0.111 |
## | 0.229 | 0.022 | |
## | 0.099 | 0.012 | |
## ------------------------|-----------|-----------|-----------|
## 1 | 4 | 4 | 8 |
## | 0.085 | 0.065 | |
## | 0.500 | 0.500 | 0.099 |
## | 0.114 | 0.087 | |
## | 0.049 | 0.049 | |
## ------------------------|-----------|-----------|-----------|
## 2 | 2 | 8 | 10 |
## | 1.247 | 0.949 | |
## | 0.200 | 0.800 | 0.123 |
## | 0.057 | 0.174 | |
## | 0.025 | 0.099 | |
## ------------------------|-----------|-----------|-----------|
## 3 | 7 | 4 | 11 |
## | 1.062 | 0.808 | |
## | 0.636 | 0.364 | 0.136 |
## | 0.200 | 0.087 | |
## | 0.086 | 0.049 | |
## ------------------------|-----------|-----------|-----------|
## 4 | 10 | 16 | 26 |
## | 0.136 | 0.103 | |
## | 0.385 | 0.615 | 0.321 |
## | 0.286 | 0.348 | |
## | 0.123 | 0.198 | |
## ------------------------|-----------|-----------|-----------|
## 5 | 4 | 7 | 11 |
## | 0.119 | 0.091 | |
## | 0.364 | 0.636 | 0.136 |
## | 0.114 | 0.152 | |
## | 0.049 | 0.086 | |
## ------------------------|-----------|-----------|-----------|
## 6 | 0 | 6 | 6 |
## | 2.593 | 1.973 | |
## | 0.000 | 1.000 | 0.074 |
## | 0.000 | 0.130 | |
## | 0.000 | 0.074 | |
## ------------------------|-----------|-----------|-----------|
## Column Total | 35 | 46 | 81 |
## | 0.432 | 0.568 | |
## ------------------------|-----------|-----------|-----------|
##
##
## Fisher's Exact Test for Count Data
## ------------------------------------------------------------
## Alternative hypothesis: two.sided
## p = 0.007648878
##
## comorbidities_hopc_df <- as.data.frame(n_comorbidities_hopc) %>%
rename(n_comorbidities = Var1, hopc = 2)
comorbidities_hopc_df| n_comorbidities | hopc | Freq |
|---|---|---|
| 0 | 1 | 8 |
| 1 | 1 | 4 |
| 2 | 1 | 2 |
| 3 | 1 | 7 |
| 4 | 1 | 10 |
| 5 | 1 | 4 |
| 6 | 1 | 0 |
| 0 | 2 | 1 |
| 1 | 2 | 4 |
| 2 | 2 | 8 |
| 3 | 2 | 4 |
| 4 | 2 | 16 |
| 5 | 2 | 7 |
| 6 | 2 | 6 |
comorbidities_hopc_df_wide <- pivot_wider(
data = comorbidities_hopc_df,
names_from = hopc,
values_from = Freq
) %>%
rename(c("Primary" = "1", "Secondary" = "2")) %>%
mutate(Proportion_Secondary = Secondary / (Primary + Secondary))
comorbidities_hopc_df_wide| n_comorbidities | Primary | Secondary | Proportion_Secondary |
|---|---|---|---|
| 0 | 8 | 1 | 0.111 |
| 1 | 4 | 4 | 0.5 |
| 2 | 2 | 8 | 0.8 |
| 3 | 7 | 4 | 0.364 |
| 4 | 10 | 16 | 0.615 |
| 5 | 4 | 7 | 0.636 |
| 6 | 0 | 6 | 1 |
highchart() %>%
hc_chart(type = "column") %>%
hc_title(text = "Co-morbidities and Secondary Gout") %>%
hc_xAxis(title = list(text = "Number of co-morbidities")) %>%
hc_yAxis(title = list(text = "n")) %>%
hc_plotOptions(column = list(
dataLabels = list(enabled = FALSE),
stacking = "normal",
enableMouseTracking = FALSE)
) %>%
hc_series(list(name = "Primary",
data = comorbidities_hopc_df_wide$Primary),
list(name = "Secondary",
data = comorbidities_hopc_df_wide$Secondary)
) %>%
hc_exporting(enabled = TRUE) # allow 'save'highchart() %>%
hc_chart(type = "column") %>%
hc_title(text = "Co-morbidities and Secondary Gout") %>%
hc_xAxis(title = list(text = "Number of co-morbidities")) %>%
hc_yAxis_multiples(list(title = list(text = "n")),
list(title = list(text = "Proportion secondary"),
opposite = TRUE, min = 0, max = 1)
) %>%
hc_plotOptions(column = list(
dataLabels = list(enabled = FALSE),
stacking = "normal",
enableMouseTracking = FALSE)
) %>%
hc_series(list(name = "Primary",
data = comorbidities_hopc_df_wide$Primary,
yAxis = 0),
list(name = "Secondary",
data = comorbidities_hopc_df_wide$Secondary,
yAxis = 0)
) %>%
hc_add_series(name = "Secondary proportion",
data = comorbidities_hopc_df_wide$Proportion_Secondary,
type = "line", color = "red",
yAxis = 1
) %>%
hc_exporting(enabled = TRUE) # allow 'save'hopc4 <- glm(
hopc ~ x_total_comorbidities,
data = d,
family = "binomial"
)
summary(hopc4)##
## Call:
## glm(formula = hopc ~ x_total_comorbidities, family = "binomial",
## data = d)
##
## Deviance Residuals:
## Min 1Q Median 3Q Max
## -1.6184 -1.2776 0.6711 0.9302 1.5805
##
## Coefficients:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) -0.9111 0.4964 -1.835 0.0664 .
## x_total_comorbidities 0.3812 0.1419 2.687 0.0072 **
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for binomial family taken to be 1)
##
## Null deviance: 110.79 on 80 degrees of freedom
## Residual deviance: 102.81 on 79 degrees of freedom
## AIC: 106.81
##
## Number of Fisher Scoring iterations: 4Change in ULT therapy vs. rheum_input
ULT_category_rheum <- d %>%
mutate(ULT_change_category =
case_when(
ult_on_admission_y_n == 1 & change_dose == 0 & ceased_withheld == 0 ~ "On, no change",
ult_on_admission_y_n == 1 & change_dose == 1 ~ "On and changed",
ult_on_admission_y_n == 1 & ceased_withheld == 1 ~ "On and ceased",
ult_on_admission_y_n == 0 & commenced_during == 0 & after_discharge == 0 ~ "None, not started",
ult_on_admission_y_n == 0 & commenced_during == 1 ~ "None, started during admission",
ult_on_admission_y_n == 0 & after_discharge == 1 ~ "None, started on discharge",
TRUE ~ "Everything else"
)) %>%
mutate(ULT_change_category = factor(ULT_change_category)) %>%
select(ULT_change_category, rheum_input)
summary(ULT_category_rheum)## ULT_change_category rheum_input
## None, not started :28 0:64
## None, started during admission:12 1:17
## None, started on discharge :15
## On and ceased : 2
## On and changed : 3
## On, no change :21CrossTable(ULT_category_rheum$ULT_change_category, ULT_category_rheum$rheum_input, fisher = TRUE)##
##
## Cell Contents
## |-------------------------|
## | N |
## | Chi-square contribution |
## | N / Row Total |
## | N / Col Total |
## | N / Table Total |
## |-------------------------|
##
##
## Total Observations in Table: 81
##
##
## | ULT_category_rheum$rheum_input
## ULT_category_rheum$ULT_change_category | 0 | 1 | Row Total |
## ---------------------------------------|-----------|-----------|-----------|
## None, not started | 21 | 7 | 28 |
## | 0.057 | 0.215 | |
## | 0.750 | 0.250 | 0.346 |
## | 0.328 | 0.412 | |
## | 0.259 | 0.086 | |
## ---------------------------------------|-----------|-----------|-----------|
## None, started during admission | 10 | 2 | 12 |
## | 0.028 | 0.107 | |
## | 0.833 | 0.167 | 0.148 |
## | 0.156 | 0.118 | |
## | 0.123 | 0.025 | |
## ---------------------------------------|-----------|-----------|-----------|
## None, started on discharge | 12 | 3 | 15 |
## | 0.002 | 0.007 | |
## | 0.800 | 0.200 | 0.185 |
## | 0.188 | 0.176 | |
## | 0.148 | 0.037 | |
## ---------------------------------------|-----------|-----------|-----------|
## On and ceased | 2 | 0 | 2 |
## | 0.111 | 0.420 | |
## | 1.000 | 0.000 | 0.025 |
## | 0.031 | 0.000 | |
## | 0.025 | 0.000 | |
## ---------------------------------------|-----------|-----------|-----------|
## On and changed | 0 | 3 | 3 |
## | 2.370 | 8.924 | |
## | 0.000 | 1.000 | 0.037 |
## | 0.000 | 0.176 | |
## | 0.000 | 0.037 | |
## ---------------------------------------|-----------|-----------|-----------|
## On, no change | 19 | 2 | 21 |
## | 0.349 | 1.315 | |
## | 0.905 | 0.095 | 0.259 |
## | 0.297 | 0.118 | |
## | 0.235 | 0.025 | |
## ---------------------------------------|-----------|-----------|-----------|
## Column Total | 64 | 17 | 81 |
## | 0.790 | 0.210 | |
## ---------------------------------------|-----------|-----------|-----------|
##
##
## Fisher's Exact Test for Count Data
## ------------------------------------------------------------
## Alternative hypothesis: two.sided
## p = 0.0435109
##
## ULT_category_rheum2 <- d %>%
mutate(ULT_change_category =
case_when(
ult_on_admission_y_n == 1 & change_dose == 0 & ceased_withheld == 0 ~ "Indifferent",
ult_on_admission_y_n == 1 & change_dose == 1 ~ "Good",
ult_on_admission_y_n == 1 & ceased_withheld == 1 ~ "Bad",
ult_on_admission_y_n == 0 & commenced_during == 0 & after_discharge == 0 ~ "Bad",
ult_on_admission_y_n == 0 & commenced_during == 1 ~ "Good",
ult_on_admission_y_n == 0 & after_discharge == 1 ~ "Good",
TRUE ~ "Everything else"
)) %>%
mutate(ULT_change_category = factor(ULT_change_category)) %>%
select(ULT_change_category, rheum_input)
summary(ULT_category_rheum2)## ULT_change_category rheum_input
## Bad :30 0:64
## Good :30 1:17
## Indifferent:21ULT_category_rheum2b <- ULT_category_rheum2 %>% filter(ULT_change_category != "Indifferent")
CrossTable(ULT_category_rheum2b$ULT_change_category, ULT_category_rheum2b$rheum_input, fisher = TRUE)##
##
## Cell Contents
## |-------------------------|
## | N |
## | Chi-square contribution |
## | N / Row Total |
## | N / Col Total |
## | N / Table Total |
## |-------------------------|
##
##
## Total Observations in Table: 60
##
##
## | ULT_category_rheum2b$rheum_input
## ULT_category_rheum2b$ULT_change_category | 0 | 1 | Row Total |
## -----------------------------------------|-----------|-----------|-----------|
## Bad | 23 | 7 | 30 |
## | 0.011 | 0.033 | |
## | 0.767 | 0.233 | 0.500 |
## | 0.511 | 0.467 | |
## | 0.383 | 0.117 | |
## -----------------------------------------|-----------|-----------|-----------|
## Good | 22 | 8 | 30 |
## | 0.011 | 0.033 | |
## | 0.733 | 0.267 | 0.500 |
## | 0.489 | 0.533 | |
## | 0.367 | 0.133 | |
## -----------------------------------------|-----------|-----------|-----------|
## Column Total | 45 | 15 | 60 |
## | 0.750 | 0.250 | |
## -----------------------------------------|-----------|-----------|-----------|
##
##
## Fisher's Exact Test for Count Data
## ------------------------------------------------------------
## Sample estimate odds ratio: 1.191262
##
## Alternative hypothesis: true odds ratio is not equal to 1
## p = 1
## 95% confidence interval: 0.3159736 4.590214
##
## Alternative hypothesis: true odds ratio is less than 1
## p = 0.7239983
## 95% confidence interval: 0 3.775563
##
## Alternative hypothesis: true odds ratio is greater than 1
## p = 0.5
## 95% confidence interval: 0.3815898 Inf
##
##
##