Care preference documentation within 48hrs of admission, age
>74.
Patient as unit of analysis. Care Provider is random effect, so we can explain the difference in variance explained by care providers compared to that explained by other effects in the model.
The following Illness Severity Scores (github) are available:
These will be used as well as:
library("ggplot2") ## Graphing
library("GGally") ## Extension to ggplot
library("reshape2") ## for melt()
library("lme4") ## for Hierachal Models
library("sjPlot") ## Lovely Presentation of Model Outputattending_check to ensure that all patients’ notes documented during any hospital admission have at least a single attending physician as part of the care team, attending_check will go through each hospital admission and keep only those admissions with attendings who have logged a patient note that was captured by MIMIC-III
dat <- read.csv("~/nqf_caregivers/data/20180607_EOL_data_ICU.csv", header = T, stringsAsFactors = F)
## Load CAREGIVERS Table for join on CGID
cg <- read.csv("~/nqf_caregivers/data/mimic/CAREGIVERS.csv",
header = T, stringsAsFactors = F)
## Change column name of "NOTEEVENTS.DESCRIPTION" to explicitly mention that it describes the note
colnames(dat)[which(colnames(dat) == "DESCRIPTION")] <- "NOTE_DESCRIPTION"
## Change column name of "CAREGIVERS. DESCRIPTION" to explicitly mention that it describes the careprovider
colnames(cg)[which(colnames(cg) == "DESCRIPTION")] <- "CG_DESCRIPTION"TEXT as the column is unnecessary but very large.## Remove ROW_ID from CG
cg$ROW_ID <- NULL
## Remove TEXT
dat$TEXT <- NULL
## Clean ethnicity
dat[(grepl("WHITE|PORTUGUESE", dat$ETHNICITY)),]$ETHNICITY <- "WHITE"
dat[(grepl("ASIAN", dat$ETHNICITY)),]$ETHNICITY <- "ASIAN"
dat[(grepl("BLACK", dat$ETHNICITY)),]$ETHNICITY <- "BLACK"
dat[(grepl("HISPANIC", dat$ETHNICITY)),]$ETHNICITY <- "HISPANIC"
dat[(grepl("MIDDLE|NATIVE|MULTI|DECLINED|UNABLE|OTHER|NOT",dat$ETHNICITY)),]$ETHNICITY <- "UNKNOWN"
## Clean Marital Status
dat$MARITAL_STATUS[dat$MARITAL_STATUS == ""] <- "UNKNOWN (DEFAULT)"
dat$MARITAL_STATUS[dat$MARITAL_STATUS == "UNKNOWN (DEFAULT)"] <- "UNKNOWN"
## Clean Religion
## Christianity-based
dat$RELIGION[dat$RELIGION == "7TH DAY ADVENTIST"] <- "CHRISTIAN"
dat$RELIGION[dat$RELIGION == "CATHOLIC"] <- "CHRISTIAN"
dat$RELIGION[dat$RELIGION == "CHRISTIAN SCIENTIST"] <- "CHRISTIAN"
dat$RELIGION[dat$RELIGION == "EPISCOPALIAN"] <- "CHRISTIAN"
dat$RELIGION[dat$RELIGION == "GREEK ORTHODOX"] <- "CHRISTIAN"
dat$RELIGION[dat$RELIGION == "CHRISTIAN SCIENTIST"] <- "CHRISTIAN"
dat$RELIGION[dat$RELIGION == "PROTESTANT QUAKER"] <- "CHRISTIAN"
dat$RELIGION[dat$RELIGION == "ROMANIAN EAST. ORTH"] <- "CHRISTIAN"
## Others
dat$RELIGION[dat$RELIGION == "UNITARIAN-UNIVERSALIST"] <- "OTHER/UNSPECIFIED"
dat$RELIGION[dat$RELIGION == "NOT SPECIFIED"] <- "OTHER/UNSPECIFIED"
dat$RELIGION[dat$RELIGION == "OTHER"] <- "OTHER/UNSPECIFIED"
dat$RELIGION[dat$RELIGION == "UNOBTAINABLE"] <- "OTHER/UNSPECIFIED"
## Merge
dat <- merge(dat, cg, by = "CGID")
## Factor
dat <- within(dat, {
ETHNICITY <- factor(ETHNICITY)
MARITAL_STATUS <- factor(MARITAL_STATUS)
RELIGION <- factor(RELIGION)
FIRST_CAREUNIT <- factor(FIRST_CAREUNIT)
CGID <- factor(CGID)
CG_DESCRIPTION <- factor(CG_DESCRIPTION)
SUBJECT_ID <- factor(SUBJECT_ID)
})length(unique(dat$CGID))## [1] 497tmp <- attending_check(dat)
length(unique(tmp$CGID))## [1] 493## Initial model to inform prior probabilities
m_initial <- glmer(CIM.machine ~
AGE +
ETHNICITY +
MARITAL_STATUS +
RELIGION + ## Failing to converge...
(1 | CGID),
data = tmp,
family = binomial,
control = glmerControl(optimizer = "bobyqa"), ## Good optimizer to avoid non-convergence
nAGQ = 10) ## Default value 1, higher values increase estimate accuracy## Warning in optwrap(optimizer, devfun, start, rho$lower, control =
## control, : convergence code 1 from bobyqa: bobyqa -- maximum number of
## function evaluations exceeded## Warning in checkConv(attr(opt, "derivs"), opt$par, ctrl = control
## $checkConv, : Model failed to converge with max|grad| = 0.393964 (tol =
## 0.001, component 1)## Warning in checkConv(attr(opt, "derivs"), opt$par, ctrl = control$checkConv, : Model is nearly unidentifiable: very large eigenvalue
## - Rescale variables?;Model is nearly unidentifiable: large eigenvalue ratio
## - Rescale variables?## View
summary(m_initial)## Generalized linear mixed model fit by maximum likelihood (Adaptive
## Gauss-Hermite Quadrature, nAGQ = 10) [glmerMod]
## Family: binomial ( logit )
## Formula: CIM.machine ~ AGE + ETHNICITY + MARITAL_STATUS + RELIGION + (1 |
## CGID)
## Data: tmp
## Control: glmerControl(optimizer = "bobyqa")
##
## AIC BIC logLik deviance df.resid
## 12658.0 12780.4 -6312.0 12624.0 9887
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -3.5865 -0.8254 -0.4054 0.8606 3.0488
##
## Random effects:
## Groups Name Variance Std.Dev.
## CGID (Intercept) 1.024 1.012
## Number of obs: 9904, groups: CGID, 493
##
## Fixed effects:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) -3.36459 0.67356 -4.995 5.88e-07 ***
## AGE 0.04775 0.00476 10.032 < 2e-16 ***
## ETHNICITYBLACK 0.55746 0.15251 3.655 0.000257 ***
## ETHNICITYHISPANIC 1.10065 0.24710 4.454 8.42e-06 ***
## ETHNICITYUNKNOWN 0.36274 0.16559 2.191 0.028479 *
## ETHNICITYWHITE 0.46241 0.13301 3.476 0.000508 ***
## MARITAL_STATUSMARRIED -0.40540 0.11662 -3.476 0.000508 ***
## MARITAL_STATUSSEPARATED 1.01880 0.29001 3.513 0.000443 ***
## MARITAL_STATUSSINGLE -0.12321 0.12469 -0.988 0.323105
## MARITAL_STATUSUNKNOWN -0.36164 0.16936 -2.135 0.032739 *
## MARITAL_STATUSWIDOWED -0.07357 0.11817 -0.623 0.533567
## RELIGIONCHRISTIAN -1.11058 0.53592 -2.072 0.038240 *
## RELIGIONHINDU 0.25637 0.77267 0.332 0.740044
## RELIGIONJEWISH -0.96491 0.53815 -1.793 0.072969 .
## RELIGIONMUSLIM -3.38761 1.21704 -2.783 0.005378 **
## RELIGIONOTHER/UNSPECIFIED -0.95891 0.53538 -1.791 0.073279 .
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1##
## Correlation matrix not shown by default, as p = 16 > 12.
## Use print(x, correlation=TRUE) or
## vcov(x) if you need it## convergence code: 1
## Model failed to converge with max|grad| = 0.393964 (tol = 0.001, component 1)
## Model is nearly unidentifiable: very large eigenvalue
## - Rescale variables?
## Model is nearly unidentifiable: large eigenvalue ratio
## - Rescale variables?sjt.glmer(m_initial)| CIM.machine | ||||
| Odds Ratio | CI | p | ||
| Fixed Parts | ||||
| (Intercept) | 0.03 | 0.01 – 0.13 | <.001 | |
| AGE | 1.05 | 1.04 – 1.06 | <.001 | |
| ETHNICITY (BLACK) | 1.75 | 1.30 – 2.35 | <.001 | |
| ETHNICITY (HISPANIC) | 3.01 | 1.85 – 4.88 | <.001 | |
| ETHNICITY (UNKNOWN) | 1.44 | 1.04 – 1.99 | .028 | |
| ETHNICITY (WHITE) | 1.59 | 1.22 – 2.06 | <.001 | |
| MARITAL_STATUS (MARRIED) | 0.67 | 0.53 – 0.84 | <.001 | |
| MARITAL_STATUS (SEPARATED) | 2.77 | 1.57 – 4.89 | <.001 | |
| MARITAL_STATUS (SINGLE) | 0.88 | 0.69 – 1.13 | .323 | |
| MARITAL_STATUS (UNKNOWN) | 0.70 | 0.50 – 0.97 | .033 | |
| MARITAL_STATUS (WIDOWED) | 0.93 | 0.74 – 1.17 | .534 | |
| RELIGION (CHRISTIAN) | 0.33 | 0.12 – 0.94 | .038 | |
| RELIGION (HINDU) | 1.29 | 0.28 – 5.88 | .740 | |
| RELIGION (JEWISH) | 0.38 | 0.13 – 1.09 | .073 | |
| RELIGION (MUSLIM) | 0.03 | 0.00 – 0.37 | .005 | |
| RELIGION (OTHER/UNSPECIFIED) | 0.38 | 0.13 – 1.09 | .073 | |
| Random Parts | ||||
| τ00, CGID | 1.024 | |||
| NCGID | 493 | |||
| ICCCGID | 0.237 | |||
| Observations | 9904 | |||
| Deviance | 11680.560 | |||
## Initial model to inform prior probabilities
m_icu <- glmer(CIM.machine ~
AGE +
ETHNICITY +
MARITAL_STATUS +
RELIGION + ## Failing to converge...
FIRST_CAREUNIT +
(1 | CGID),
data = tmp,
family = binomial,
control = glmerControl(optimizer = "bobyqa"), ## Good optimizer to avoid non-convergence
nAGQ = 10) ## Default value 1, higher values increase estimate accuracy## Warning in optwrap(optimizer, devfun, start, rho$lower, control =
## control, : convergence code 1 from bobyqa: bobyqa -- maximum number of
## function evaluations exceeded## Warning in checkConv(attr(opt, "derivs"), opt$par, ctrl = control
## $checkConv, : Model failed to converge with max|grad| = 0.46684 (tol =
## 0.001, component 1)## Warning in checkConv(attr(opt, "derivs"), opt$par, ctrl = control$checkConv, : Model is nearly unidentifiable: very large eigenvalue
## - Rescale variables?;Model is nearly unidentifiable: large eigenvalue ratio
## - Rescale variables?## View
summary(m_icu)## Generalized linear mixed model fit by maximum likelihood (Adaptive
## Gauss-Hermite Quadrature, nAGQ = 10) [glmerMod]
## Family: binomial ( logit )
## Formula:
## CIM.machine ~ AGE + ETHNICITY + MARITAL_STATUS + RELIGION + FIRST_CAREUNIT +
## (1 | CGID)
## Data: tmp
## Control: glmerControl(optimizer = "bobyqa")
##
## AIC BIC logLik deviance df.resid
## 12629.4 12780.7 -6293.7 12587.4 9883
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -3.6281 -0.8239 -0.3848 0.8595 3.5402
##
## Random effects:
## Groups Name Variance Std.Dev.
## CGID (Intercept) 1 1
## Number of obs: 9904, groups: CGID, 493
##
## Fixed effects:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) -3.730517 0.679931 -5.487 4.10e-08 ***
## AGE 0.048223 0.004781 10.085 < 2e-16 ***
## ETHNICITYBLACK 0.521314 0.152622 3.416 0.000636 ***
## ETHNICITYHISPANIC 1.089231 0.248201 4.388 1.14e-05 ***
## ETHNICITYUNKNOWN 0.371401 0.165602 2.243 0.024914 *
## ETHNICITYWHITE 0.445360 0.132932 3.350 0.000807 ***
## MARITAL_STATUSMARRIED -0.387393 0.116756 -3.318 0.000907 ***
## MARITAL_STATUSSEPARATED 1.032241 0.288688 3.576 0.000349 ***
## MARITAL_STATUSSINGLE -0.100163 0.124847 -0.802 0.422387
## MARITAL_STATUSUNKNOWN -0.356539 0.169592 -2.102 0.035524 *
## MARITAL_STATUSWIDOWED -0.048843 0.118450 -0.412 0.680082
## RELIGIONCHRISTIAN -0.992153 0.535635 -1.852 0.063984 .
## RELIGIONHINDU 0.260893 0.768140 0.340 0.734125
## RELIGIONJEWISH -0.843354 0.538022 -1.568 0.116996
## RELIGIONMUSLIM -3.326771 1.222678 -2.721 0.006511 **
## RELIGIONOTHER/UNSPECIFIED -0.847496 0.534991 -1.584 0.113164
## FIRST_CAREUNITCSRU -0.488832 0.173488 -2.818 0.004837 **
## FIRST_CAREUNITMICU 0.273727 0.077723 3.522 0.000429 ***
## FIRST_CAREUNITSICU 0.264767 0.115104 2.300 0.021434 *
## FIRST_CAREUNITTSICU 0.420698 0.141915 2.964 0.003032 **
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1##
## Correlation matrix not shown by default, as p = 20 > 12.
## Use print(x, correlation=TRUE) or
## vcov(x) if you need it## convergence code: 1
## Model failed to converge with max|grad| = 0.46684 (tol = 0.001, component 1)
## Model is nearly unidentifiable: very large eigenvalue
## - Rescale variables?
## Model is nearly unidentifiable: large eigenvalue ratio
## - Rescale variables?sjt.glmer(m_icu)| CIM.machine | ||||
| Odds Ratio | CI | p | ||
| Fixed Parts | ||||
| (Intercept) | 0.02 | 0.01 – 0.09 | <.001 | |
| AGE | 1.05 | 1.04 – 1.06 | <.001 | |
| ETHNICITY (BLACK) | 1.68 | 1.25 – 2.27 | <.001 | |
| ETHNICITY (HISPANIC) | 2.97 | 1.83 – 4.83 | <.001 | |
| ETHNICITY (UNKNOWN) | 1.45 | 1.05 – 2.01 | .025 | |
| ETHNICITY (WHITE) | 1.56 | 1.20 – 2.03 | <.001 | |
| MARITAL_STATUS (MARRIED) | 0.68 | 0.54 – 0.85 | <.001 | |
| MARITAL_STATUS (SEPARATED) | 2.81 | 1.59 – 4.94 | <.001 | |
| MARITAL_STATUS (SINGLE) | 0.90 | 0.71 – 1.16 | .422 | |
| MARITAL_STATUS (UNKNOWN) | 0.70 | 0.50 – 0.98 | .036 | |
| MARITAL_STATUS (WIDOWED) | 0.95 | 0.76 – 1.20 | .680 | |
| RELIGION (CHRISTIAN) | 0.37 | 0.13 – 1.06 | .064 | |
| RELIGION (HINDU) | 1.30 | 0.29 – 5.85 | .734 | |
| RELIGION (JEWISH) | 0.43 | 0.15 – 1.24 | .117 | |
| RELIGION (MUSLIM) | 0.04 | 0.00 – 0.39 | .007 | |
| RELIGION (OTHER/UNSPECIFIED) | 0.43 | 0.15 – 1.22 | .113 | |
| FIRST_CAREUNIT (CSRU) | 0.61 | 0.44 – 0.86 | .005 | |
| FIRST_CAREUNIT (MICU) | 1.31 | 1.13 – 1.53 | <.001 | |
| FIRST_CAREUNIT (SICU) | 1.30 | 1.04 – 1.63 | .021 | |
| FIRST_CAREUNIT (TSICU) | 1.52 | 1.15 – 2.01 | .003 | |
| Random Parts | ||||
| τ00, CGID | 1.000 | |||
| NCGID | 493 | |||
| ICCCGID | 0.233 | |||
| Observations | 9904 | |||
| Deviance | 11653.818 | |||
## Initial model to inform prior probabilities
f_icu <- glmer(CIM.machine ~
AGE +
ETHNICITY +
MARITAL_STATUS +
RELIGION + ## Failing to converge...
FIRST_CAREUNIT +
ADMISSION_LOCATION +
ADMISSION_TYPE +
(1 | CGID),
data = tmp,
family = binomial,
control = glmerControl(optimizer = "bobyqa"), ## Good optimizer to avoid non-convergence
nAGQ = 10) ## Default value 1, higher values increase estimate accuracy## Warning in optwrap(optimizer, devfun, start, rho$lower, control =
## control, : convergence code 1 from bobyqa: bobyqa -- maximum number of
## function evaluations exceeded## Warning in checkConv(attr(opt, "derivs"), opt$par, ctrl = control
## $checkConv, : Model failed to converge with max|grad| = 1.56219 (tol =
## 0.001, component 1)## Warning in checkConv(attr(opt, "derivs"), opt$par, ctrl = control$checkConv, : Model is nearly unidentifiable: very large eigenvalue
## - Rescale variables?;Model is nearly unidentifiable: large eigenvalue ratio
## - Rescale variables?## View
summary(f_icu)## Generalized linear mixed model fit by maximum likelihood (Adaptive
## Gauss-Hermite Quadrature, nAGQ = 10) [glmerMod]
## Family: binomial ( logit )
## Formula:
## CIM.machine ~ AGE + ETHNICITY + MARITAL_STATUS + RELIGION + FIRST_CAREUNIT +
## ADMISSION_LOCATION + ADMISSION_TYPE + (1 | CGID)
## Data: tmp
## Control: glmerControl(optimizer = "bobyqa")
##
## AIC BIC logLik deviance df.resid
## 12408.8 12610.5 -6176.4 12352.8 9876
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -3.6553 -0.8222 -0.2387 0.8388 4.4339
##
## Random effects:
## Groups Name Variance Std.Dev.
## CGID (Intercept) 1.003 1.002
## Number of obs: 9904, groups: CGID, 493
##
## Fixed effects:
## Estimate Std. Error z value
## (Intercept) -6.071121 0.803722 -7.554
## AGE 0.044664 0.004871 9.169
## ETHNICITYBLACK 0.530336 0.153676 3.451
## ETHNICITYHISPANIC 1.074456 0.247904 4.334
## ETHNICITYUNKNOWN 0.442682 0.168108 2.633
## ETHNICITYWHITE 0.507861 0.133482 3.805
## MARITAL_STATUSMARRIED -0.387152 0.118663 -3.263
## MARITAL_STATUSSEPARATED 1.406281 0.317763 4.426
## MARITAL_STATUSSINGLE -0.118201 0.127064 -0.930
## MARITAL_STATUSUNKNOWN -0.444483 0.170714 -2.604
## MARITAL_STATUSWIDOWED -0.039255 0.120435 -0.326
## RELIGIONCHRISTIAN -1.223663 0.564632 -2.167
## RELIGIONHINDU 0.027039 0.795014 0.034
## RELIGIONJEWISH -1.071856 0.567615 -1.888
## RELIGIONMUSLIM -3.621371 1.237211 -2.927
## RELIGIONOTHER/UNSPECIFIED -1.077147 0.564309 -1.909
## FIRST_CAREUNITCSRU -0.271377 0.178386 -1.521
## FIRST_CAREUNITMICU 0.312865 0.078928 3.964
## FIRST_CAREUNITSICU 0.304097 0.116536 2.609
## FIRST_CAREUNITTSICU 0.469925 0.144031 3.263
## ADMISSION_LOCATIONEMERGENCY ROOM ADMIT 0.313420 0.176291 1.778
## ADMISSION_LOCATIONPHYS REFERRAL/NORMAL DELI 0.578317 0.329927 1.753
## ADMISSION_LOCATIONTRANSFER FROM HOSP/EXTRAM 0.226384 0.187421 1.208
## ADMISSION_LOCATIONTRANSFER FROM OTHER HEALT -12.388770 64.000726 -0.194
## ADMISSION_LOCATIONTRANSFER FROM SKILLED NUR -4.013686 0.844157 -4.755
## ADMISSION_TYPEEMERGENCY 2.553068 0.358987 7.112
## ADMISSION_TYPEURGENT 2.276738 0.464781 4.899
## Pr(>|z|)
## (Intercept) 4.23e-14 ***
## AGE < 2e-16 ***
## ETHNICITYBLACK 0.000559 ***
## ETHNICITYHISPANIC 1.46e-05 ***
## ETHNICITYUNKNOWN 0.008455 **
## ETHNICITYWHITE 0.000142 ***
## MARITAL_STATUSMARRIED 0.001104 **
## MARITAL_STATUSSEPARATED 9.62e-06 ***
## MARITAL_STATUSSINGLE 0.352242
## MARITAL_STATUSUNKNOWN 0.009223 **
## MARITAL_STATUSWIDOWED 0.744469
## RELIGIONCHRISTIAN 0.030221 *
## RELIGIONHINDU 0.972869
## RELIGIONJEWISH 0.058979 .
## RELIGIONMUSLIM 0.003422 **
## RELIGIONOTHER/UNSPECIFIED 0.056289 .
## FIRST_CAREUNITCSRU 0.128185
## FIRST_CAREUNITMICU 7.37e-05 ***
## FIRST_CAREUNITSICU 0.009068 **
## FIRST_CAREUNITTSICU 0.001104 **
## ADMISSION_LOCATIONEMERGENCY ROOM ADMIT 0.075428 .
## ADMISSION_LOCATIONPHYS REFERRAL/NORMAL DELI 0.079625 .
## ADMISSION_LOCATIONTRANSFER FROM HOSP/EXTRAM 0.227088
## ADMISSION_LOCATIONTRANSFER FROM OTHER HEALT 0.846511
## ADMISSION_LOCATIONTRANSFER FROM SKILLED NUR 1.99e-06 ***
## ADMISSION_TYPEEMERGENCY 1.14e-12 ***
## ADMISSION_TYPEURGENT 9.66e-07 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1##
## Correlation matrix not shown by default, as p = 27 > 12.
## Use print(x, correlation=TRUE) or
## vcov(x) if you need it## convergence code: 1
## Model failed to converge with max|grad| = 1.56219 (tol = 0.001, component 1)
## Model is nearly unidentifiable: very large eigenvalue
## - Rescale variables?
## Model is nearly unidentifiable: large eigenvalue ratio
## - Rescale variables?sjt.glmer(f_icu)| CIM.machine | ||||
| Odds Ratio | CI | p | ||
| Fixed Parts | ||||
| (Intercept) | 0.00 | 0.00 – 0.01 | <.001 | |
| AGE | 1.05 | 1.04 – 1.06 | <.001 | |
| ETHNICITY (BLACK) | 1.70 | 1.26 – 2.30 | <.001 | |
| ETHNICITY (HISPANIC) | 2.93 | 1.80 – 4.76 | <.001 | |
| ETHNICITY (UNKNOWN) | 1.56 | 1.12 – 2.16 | .008 | |
| ETHNICITY (WHITE) | 1.66 | 1.28 – 2.16 | <.001 | |
| MARITAL_STATUS (MARRIED) | 0.68 | 0.54 – 0.86 | .001 | |
| MARITAL_STATUS (SEPARATED) | 4.08 | 2.19 – 7.61 | <.001 | |
| MARITAL_STATUS (SINGLE) | 0.89 | 0.69 – 1.14 | .352 | |
| MARITAL_STATUS (UNKNOWN) | 0.64 | 0.46 – 0.90 | .009 | |
| MARITAL_STATUS (WIDOWED) | 0.96 | 0.76 – 1.22 | .744 | |
| RELIGION (CHRISTIAN) | 0.29 | 0.10 – 0.89 | .030 | |
| RELIGION (HINDU) | 1.03 | 0.22 – 4.88 | .973 | |
| RELIGION (JEWISH) | 0.34 | 0.11 – 1.04 | .059 | |
| RELIGION (MUSLIM) | 0.03 | 0.00 – 0.30 | .003 | |
| RELIGION (OTHER/UNSPECIFIED) | 0.34 | 0.11 – 1.03 | .056 | |
| FIRST_CAREUNIT (CSRU) | 0.76 | 0.54 – 1.08 | .128 | |
| FIRST_CAREUNIT (MICU) | 1.37 | 1.17 – 1.60 | <.001 | |
| FIRST_CAREUNIT (SICU) | 1.36 | 1.08 – 1.70 | .009 | |
| FIRST_CAREUNIT (TSICU) | 1.60 | 1.21 – 2.12 | .001 | |
| ADMISSION_LOCATION (EMERGENCY ROOM ADMIT) | 1.37 | 0.97 – 1.93 | .075 | |
| ADMISSION_LOCATION (PHYS REFERRAL/NORMAL DELI) | 1.78 | 0.93 – 3.40 | .080 | |
| ADMISSION_LOCATION (TRANSFER FROM HOSP/EXTRAM) | 1.25 | 0.87 – 1.81 | .227 | |
| ADMISSION_LOCATION (TRANSFER FROM OTHER HEALT) | 0.00 | 0.00 – 12506702398686011825011417270416144874365282942976.00 | .847 | |
| ADMISSION_LOCATION (TRANSFER FROM SKILLED NUR) | 0.02 | 0.00 – 0.09 | <.001 | |
| ADMISSION_TYPE (EMERGENCY) | 12.85 | 6.36 – 25.96 | <.001 | |
| ADMISSION_TYPE (URGENT) | 9.74 | 3.92 – 24.23 | <.001 | |
| Random Parts | ||||
| τ00, CGID | 1.003 | |||
| NCGID | 493 | |||
| ICCCGID | 0.234 | |||
| Observations | 9904 | |||
| Deviance | 11429.559 | |||