UNOS Analysis
Stephanie Tuminello
2024-05-09
Reading in data and patient selection
Data Cleaning
#Main Outcome
table(UNOS2$ACUTE_REJECTION_EPISODE)##
## 1 2 3
## 2170 341 30054UNOS2 <- UNOS2 %>% mutate(reject2 = ifelse(ACUTE_REJECTION_EPISODE == 1 | ACUTE_REJECTION_EPISODE == 2, 1,
ifelse(ACUTE_REJECTION_EPISODE == 3, 0, NA)))
table(UNOS2$reject2) #missing for 25, dropping##
## 0 1
## 30054 2511UNOS2 <- UNOS2 %>% filter(!is.na(reject2)) #32565
#Main Predictor
table(UNOS2$Donor_Recipient_ABO_Match_Level)##
## 1 2
## 30172 2393UNOS2 <- UNOS2 %>% mutate(ABO2 = ifelse(Donor_Recipient_ABO_Match_Level == 1, "Identical",
ifelse(Donor_Recipient_ABO_Match_Level == 2, "Compatible", NA)))
table(UNOS2$ABO2)##
## Compatible Identical
## 2393 30172#Confounders
#Gender
table(UNOS2$TCR_RECIPIENT_GENDER) #does sex of donor always match recipiant?##
## F M
## 12888 19677#R- Age
#table(UNOS2$Recipient_Age)
summary(UNOS2$Recipient_Age)## Min. 1st Qu. Median Mean 3rd Qu. Max.
## 18.00 53.00 61.00 57.46 66.00 84.00#D- Age
#table(UNOS2$DONOR_AGE__YRS_) #still some children donators
summary(UNOS2$DONOR_AGE__YRS_) ## Min. 1st Qu. Median Mean 3rd Qu. Max.
## 6.0 23.0 33.0 35.1 46.0 77.0ggplot(UNOS2, aes(x = Recipient_Age, y = DONOR_AGE__YRS_)) +
geom_point() +
labs(x = "Recipient Age", y = "Donor Age", title = "Scatter Plot of Recipient Age vs Donor Age")
#RD Abs Diff Age
UNOS2$absolute_difference_age <- abs(UNOS2$Recipient_Age - UNOS2$DONOR_AGE__YRS_)
summary(UNOS2$absolute_difference_age)## Min. 1st Qu. Median Mean 3rd Qu. Max.
## 0.00 12.00 24.00 24.56 37.00 66.00hist(UNOS2$absolute_difference_age)
#R- BMI
#table(UNOS2$Calculated_Recipient_BMI)
summary(UNOS2$Calculated_Recipient_BMI)#missing for 31## Min. 1st Qu. Median Mean 3rd Qu. Max. NA's
## 15.00 22.20 25.80 25.62 29.00 44.80 31#D- BMI
#table(UNOS2$Calculated_Donor_BMI)
summary(UNOS2$Calculated_Donor_BMI)#missing for 27## Min. 1st Qu. Median Mean 3rd Qu. Max. NA's
## 10.59 22.49 25.41 26.32 29.14 68.73 27ggplot(UNOS2, aes(x = Calculated_Recipient_BMI, y = Calculated_Donor_BMI)) +
geom_point() +
labs(x = "Recipient BMI", y = "Donor BMI", title = "Scatter Plot of Recipient BMI vs Donor BMI")## Warning: Removed 58 rows containing missing values or values outside the scale range
## (`geom_point()`).
#RD Abs Diff Age
UNOS2$absolute_difference_BMI <- abs(UNOS2$Calculated_Recipient_BMI - UNOS2$Calculated_Donor_BMI)
summary(UNOS2$absolute_difference_BMI)## Min. 1st Qu. Median Mean 3rd Qu. Max. NA's
## 0.00002 2.15405 4.49630 5.47910 7.72303 42.81304 58hist(UNOS2$absolute_difference_BMI)
#scatterplots look random so I can include both variables for donor and recipiant
#Ischemic Time (hours)
#table(UNOS2$Ischemic_Time__hours_)
summary(UNOS2$Ischemic_Time__hours_)#missing for 482## Min. 1st Qu. Median Mean 3rd Qu. Max. NA's
## 0.000 4.200 5.200 5.505 6.400 25.000 482#t- Medicial Condition (1=Intensive Care, 2=Hospitalized but not ICU, 3=Not Hospitalized)
table(UNOS2$TRR_Medical_Condition) #missing for 20##
## 1 2 3
## 3971 3130 25463UNOS2$TRR_Medical_Condition <- as.character(UNOS2$TRR_Medical_Condition)
#Race variables missing from the dataset, only ethnicity (should be noted in the discussion)
#Type of disease
UNOS2 <- UNOS2 %>%
mutate(disease2 = ifelse(Thoracic_Diagnosis_from_TRR_TCR %in% c(1604, 1613, 420, 1605, 1521, 219, 1616, 1617, 1609, 438, 447, 403, 402, 413, 448), "restrictive",
ifelse(Thoracic_Diagnosis_from_TRR_TCR %in% c(1607, 1606, 1608, 1611), "obstructive",
ifelse(Thoracic_Diagnosis_from_TRR_TCR %in% c(1602), "cystic fibrosis ", "Other"))))
table(UNOS2$disease2)##
## cystic fibrosis obstructive Other restrictive
## 2750 8729 2723 18363##Univariate analysis
table1 <- compareGroups(reject2 ~ ABO2 + TCR_RECIPIENT_GENDER + Recipient_Age + DONOR_AGE__YRS_ + absolute_difference_age + Calculated_Recipient_BMI + Calculated_Donor_BMI + absolute_difference_BMI + Ischemic_Time__hours_ + TRR_Medical_Condition + disease2, data = UNOS2)
createTable(table1)##
## --------Summary descriptives table by 'reject2'---------
##
## _____________________________________________________________
## 0 1 p.overall
## N=30054 N=2511
## ¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯
## ABO2: 0.810
## Compatible 2212 (7.36%) 181 (7.21%)
## Identical 27842 (92.6%) 2330 (92.8%)
## TCR_RECIPIENT_GENDER: <0.001
## F 11768 (39.2%) 1120 (44.6%)
## M 18286 (60.8%) 1391 (55.4%)
## Recipient_Age 57.6 (12.2) 55.5 (13.2) <0.001
## DONOR_AGE__YRS_ 35.1 (13.9) 35.0 (14.4) 0.756
## absolute_difference_age 24.6 (14.7) 23.6 (14.8) <0.001
## Calculated_Recipient_BMI 25.6 (4.49) 25.9 (4.69) 0.004
## Calculated_Donor_BMI 26.4 (5.58) 25.9 (5.31) <0.001
## absolute_difference_BMI 5.47 (4.46) 5.60 (4.31) 0.132
## Ischemic_Time__hours_ 5.49 (2.12) 5.72 (2.20) <0.001
## TRR_Medical_Condition: <0.001
## 1 3567 (11.9%) 404 (16.1%)
## 2 2861 (9.52%) 269 (10.7%)
## 3 23625 (78.6%) 1838 (73.2%)
## disease2: <0.001
## cystic fibrosis 2497 (8.31%) 253 (10.1%)
## obstructive 8126 (27.0%) 603 (24.0%)
## Other 2462 (8.19%) 261 (10.4%)
## restrictive 16969 (56.5%) 1394 (55.5%)
## ¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯##Logistic regression
##Unadjusted
model_undj <- glm(reject2 ~ ABO2, data = UNOS2, family = binomial)
summary(model_undj )##
## Call:
## glm(formula = reject2 ~ ABO2, family = binomial, data = UNOS2)
##
## Coefficients:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) -2.50316 0.07731 -32.38 <2e-16 ***
## ABO2Identical 0.02248 0.08026 0.28 0.779
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for binomial family taken to be 1)
##
## Null deviance: 17692 on 32564 degrees of freedom
## Residual deviance: 17692 on 32563 degrees of freedom
## AIC: 17696
##
## Number of Fisher Scoring iterations: 5results <- data.frame(
Odds_Ratio = exp(model_undj$coefficients),
Lower_CI = exp(confint(model_undj )[, 1]),
Upper_CI = exp(confint(model_undj )[, 2])
)## Waiting for profiling to be done...
## Waiting for profiling to be done...print(results)## Odds_Ratio Lower_CI Upper_CI
## (Intercept) 0.0818264 0.07008647 0.09491312
## ABO2Identical 1.0227325 0.87640855 1.20067215##Adjusted
model_adj <- glm(reject2 ~ ABO2 + TCR_RECIPIENT_GENDER + Recipient_Age + DONOR_AGE__YRS_ + Calculated_Recipient_BMI + Calculated_Donor_BMI + Ischemic_Time__hours_ + TRR_Medical_Condition + disease2, data = UNOS2, family = binomial)
summary(model_adj )##
## Call:
## glm(formula = reject2 ~ ABO2 + TCR_RECIPIENT_GENDER + Recipient_Age +
## DONOR_AGE__YRS_ + Calculated_Recipient_BMI + Calculated_Donor_BMI +
## Ischemic_Time__hours_ + TRR_Medical_Condition + disease2,
## family = binomial, data = UNOS2)
##
## Coefficients:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) -1.913429 0.196355 -9.745 < 2e-16 ***
## ABO2Identical -0.012705 0.081132 -0.157 0.876
## TCR_RECIPIENT_GENDERM -0.213838 0.043712 -4.892 9.98e-07 ***
## Recipient_Age -0.012769 0.002165 -5.898 3.67e-09 ***
## DONOR_AGE__YRS_ 0.001016 0.001546 0.657 0.511
## Calculated_Recipient_BMI 0.025590 0.005127 4.991 6.00e-07 ***
## Calculated_Donor_BMI -0.017564 0.004020 -4.370 1.25e-05 ***
## Ischemic_Time__hours_ 0.042134 0.009265 4.548 5.42e-06 ***
## TRR_Medical_Condition2 -0.120886 0.083844 -1.442 0.149
## TRR_Medical_Condition3 -0.258513 0.060718 -4.258 2.07e-05 ***
## disease2obstructive 0.001114 0.101448 0.011 0.991
## disease2Other 0.146952 0.104968 1.400 0.162
## disease2restrictive 0.031642 0.098812 0.320 0.749
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for binomial family taken to be 1)
##
## Null deviance: 17442 on 32025 degrees of freedom
## Residual deviance: 17258 on 32013 degrees of freedom
## (539 observations deleted due to missingness)
## AIC: 17284
##
## Number of Fisher Scoring iterations: 5results2 <- data.frame(
Odds_Ratio = exp(model_adj$coefficients),
Lower_CI = exp(confint(model_adj )[, 1]),
Upper_CI = exp(confint(model_adj )[, 2])
)## Waiting for profiling to be done...
## Waiting for profiling to be done...print(results2)## Odds_Ratio Lower_CI Upper_CI
## (Intercept) 0.1475734 0.1003276 0.2166339
## ABO2Identical 0.9873751 0.8446654 1.1611406
## TCR_RECIPIENT_GENDERM 0.8074789 0.7412478 0.8798066
## Recipient_Age 0.9873124 0.9831472 0.9915260
## DONOR_AGE__YRS_ 1.0010169 0.9979814 1.0040480
## Calculated_Recipient_BMI 1.0259200 1.0156593 1.0362788
## Calculated_Donor_BMI 0.9825898 0.9748242 0.9903058
## Ischemic_Time__hours_ 1.0430345 1.0240529 1.0619327
## TRR_Medical_Condition2 0.8861354 0.7512793 1.0437510
## TRR_Medical_Condition3 0.7721988 0.6862612 0.8707250
## disease2obstructive 1.0011150 0.8211600 1.2222705
## disease2Other 1.1582978 0.9427752 1.4228765
## disease2restrictive 1.0321475 0.8510912 1.2537941