dataquieR example report

Adrian Richter, Stephan Struckmann, Carsten Schmidt

Preface
INTEGRITY
COMPLETENESS
CONSISTENCY
ACCURACY

Preface

This is a brief example report using dataquieR’s functions. For a longer and better elaborated example, please also consider our online example with data from SHIP.

INTEGRITY

Study data

load(system.file("extdata", "study_data.RData", package = "dataquieR"))
sd1 <- study_data

The imported study data consist of:

N = 3000 observations and
P = 53 study variables

Metadata

load(system.file("extdata", "meta_data.RData", package = "dataquieR"))
md1 <- meta_data

The imported meta data provide information for:

P = 53 study variables and
Q = 20 attributes

Applicability

The call of this R-function requires two inputs only:

appmatrix <- pro_applicability_matrix(study_data = sd1, 
                                      meta_data = md1, 
                                      label_col = LABEL)

Heatmap-like plot:

appmatrix$ApplicabilityPlot

COMPLETENESS

Unit missingness

my_unit_missings2 <- com_unit_missingness(study_data  = sd1, 
                                          meta_data   = md1, 
                                          id_vars     = c("CENTER_0", "PSEUDO_ID"), 
                                          strata_vars = "CENTER_0", 
                                          label_col   = "LABEL")

my_unit_missings2$SummaryData

Segment missingness

MissSegs <- com_segment_missingness(study_data = sd1, 
                                    meta_data = md1, 
                                    label_col = "LABEL", 
                                    threshold_value = 5, 
                                    direction = "high",
                                    exclude_roles = c("secondary", "process"))

MissSegs$SummaryPlot

Adding variables for stratification

For some analyses adding new and transformed variable to the study data is necessary.

# use the month function of the lubridate package to extract month of exam date
require(lubridate)
# apply changes to copy of data
sd2 <- sd1
# indicate first/second half year
sd2$month <- month(sd2$v00013)

Static metadata of the variable must be added to the respective metadata.

MD_TMP <- prep_add_to_meta(VAR_NAMES    = "month",
                           DATA_TYPE    = "integer",
                           LABEL        = "EXAM_MONTH",
                           VALUE_LABELS = "1 = January | 2 = February | 3 = March | 
                                          4 = April | 5 = May | 6 = June | 7 = July |
                                          8 = August | 9 = September | 10 = October |
                                          11 = November | 12 = December",
                           meta_data    = md1)

Subsequent call of the R-function may include the new variable.

MissSegs <- com_segment_missingness(study_data = sd2, 
                                    meta_data = MD_TMP, 
                                    group_vars = "EXAM_MONTH", 
                                    label_col = "LABEL", 
                                    threshold_value = 1, 
                                    direction = "high",
                                    exclude_roles = c("secondary", "process"))

MissSegs$SummaryPlot

Item missingness

The following implementation considers also labeled missing codes. The use of such a table is optional but recommended. Missing code labels used in the simulated study data are loaded as follows:

code_labels <- read.csv2(system.file("extdata", 
                                     "Missing-Codes-2020.csv", 
                                     package = "dataquieR"), 
                         stringsAsFactors = FALSE, na.strings = c())

item_miss <- com_item_missingness(study_data      = sd1, 
                                  meta_data       = meta_data, 
                                  label_col       = 'LABEL', 
                                  show_causes     = TRUE, 
                                  cause_label_df  = code_labels,
                                  include_sysmiss = TRUE, 
                                  threshold_value = 80
                                )

The function call above sets the analyses of causes for missing values to TRUE, includes system missings with an own code, and sets the threshold to 80%.

item_miss$SummaryTable

Summary plot of item missingness

item_miss$SummaryPlot

CONSISTENCY

Limit deviations

MyValueLimits <- con_limit_deviations(resp_vars  = NULL,
                                      label_col  = "LABEL",
                                      study_data = sd1,
                                      meta_data  = md1,
                                      limits     = "HARD_LIMITS")

Summary table

MyValueLimits$SummaryTable

Summary plot

# select variables with deviations
whichdeviate <- as.character(MyValueLimits$SummaryTable$Variables)[MyValueLimits$SummaryTable$GRADING == 1]

ggpubr::ggarrange(plotlist = MyValueLimits$SummaryPlotList[whichdeviate], ncol = 2)

Inadmissible levels

IAVCatAll <- con_inadmissible_categorical(study_data = sd1, 
                                          meta_data  = md1, 
                                          label_col  = "LABEL")

Contradictions

checks <- read.csv(system.file("extdata", 
                               "contradiction_checks.csv",
                               package = "dataquieR"), 
                   header = TRUE, sep = "#")

AnyContradictions <- con_contradictions(study_data      = sd1,
                                        meta_data       = md1,
                                        label_col       = "LABEL",
                                        check_table     = checks,
                                        threshold_value = 1)

AnyContradictions$SummaryTable

AnyContradictions$SummaryPlot

ACCURACY

ruol <- dataquieR:::acc_robust_univariate_outlier(study_data = sd1, meta_data = md1, label_col = LABEL)

ruol$SummaryPlotList

## $AGE_0

## 
## $AGE_1

## 
## $SBP_0

## 
## $DBP_0

## 
## $GLOBAL_HEALTH_VAS_0

## 
## $ARM_CIRC_0

## 
## $CRP_0

## 
## $BSG_0

## 
## $DEV_NO_0

## 
## $N_CHILD_0

## 
## $N_INJURIES_0

## 
## $N_BIRTH_0

## 
## $N_ATC_CODES_0

## 
## $ITEM_1_0

## 
## $ITEM_2_0

## 
## $ITEM_3_0

## 
## $ITEM_4_0

## 
## $ITEM_5_0

## 
## $ITEM_6_0

## 
## $ITEM_7_0

## 
## $ITEM_8_0

myloess <- dataquieR::acc_loess(resp_vars = "SBP_0",
                                group_vars = "USR_BP_0",
                                time_vars = "EXAM_DT_0",
                                label_col = "LABEL",
                                study_data = sd1,
                                meta_data = md1)

myloess$SummaryPlotList

## $SBP_0