library(shiny)
library(forcats)
library(tidyverse)
library(here)
library(hablar)
library(janitor)
library(gt)Appendix F — Functional CRF Data Quality Checks
This document goes over quality checks for each the Functional Testing Case Report Form (CRF). The headings in the sidebar help the user navigate to their desired content.
F.1 Read in Data and write functions
F.1.1 Load Libraries
F.1.2 Function
Write a function to remove columns where all rows have NA, this is will remove duplicate columns for the the Thoracotomy/TKA cohort
not_all_na <- function(x) any(!is.na(x))F.2 CRF Quality checks
F.2.1 functional-testing Assessment Form
F.2.2 Read in data
We will call this func
# read_csv function results in parsing errors, we will use read.csv instead
func <- read.csv(here(
"data",
"functional-testing",
"functional-testing-2024-11-06.csv"
)) %>%
retype()Remove test records
test_records <- c(
"10000",
"15000",
"20000",
"25000",
"40000",
"50000",
"60000",
"70000",
"80000",
"90000",
"100000",
"110000",
"120000"
)
func <- func %>%
filter(!record_id %in% test_records)Create a column for cohort type called “cohort”
func <- func %>%
mutate(
cohort = case_when(
record_id >= 10000 & record_id < 15000 | record_id >= 25000 ~ "TKA",
TRUE ~ "Thoracic"
)
)F.2.3 Data Dictionary
Read in data dictionary and remove duplicate field names
func_dict <- read_csv(here(
"data",
"functional-testing",
"functional-testing-Data-Dictionary-2024-11-06.csv"
)) %>%
distinct(field_name, .keep_all = TRUE)F.2.4 New field name(s):
Add the field name “cohort” to the data dictionary
# Create field names
cohort_new_row <- data.frame(
field_name = "cohort",
field_type = "Character",
field_note = "Type of surgical cohort",
select_choices_or_calculations = "TKA,Thoracic"
)
# Add the new row after the last row
func_dict <- func_dict %>%
slice(1:nrow(.)) %>%
add_row(.after = nrow(.), !!!cohort_new_row)F.2.5 TKA Functional Assessment
tka_func <- func %>%
select(
record_id,
guid,
redcap_data_access_group,
redcap_event_name,
redcap_repeat_instrument,
redcap_repeat_instance,
starts_with("walk"),
starts_with("tsts"),
functional_testing_complete,
cohort
)tka_func <- tka_func %>%
filter(cohort == "TKA") %>%
filter(redcap_repeat_instrument == "functional_testing") %>%
filter(functional_testing_complete == 2 & !is.na(redcap_repeat_instance)) %>%
group_by(record_id, redcap_event_name) %>%
top_n(1, redcap_repeat_instance) %>%
ungroup() %>%
select(where(not_all_na))Keep records of subjects who completed the test.
frdata3 <- tka_func %>%
filter(walk10completeyn == 1 & functional_testing_complete == 2)F.2.5.1 10m Walk Test:
F.2.5.1.1 Flag 1:
Check for discrepancy or missing values in the first and the second initial pain ratings.
ferror1 <- frdata3 %>%
mutate(init_pain_diff = walk10initialpainscl - walk10initialpainscl1) %>%
filter(init_pain_diff != 0 | is.na(init_pain_diff)) %>%
add_column(
error_type = "Walk test:Discrepancy or missing values in the first and the second initial pain ratings"
) %>%
add_column(errors = "error") %>%
select(
record_id,
redcap_data_access_group,
redcap_repeat_instrument,
redcap_repeat_instance,
error_type,
errors
)F.2.5.1.2 Flag 2:
Check for discrepancy or missing values in the first and the second final pain ratings.
ferror2 <- frdata3 %>%
mutate(final_pain_diff = walk10finalpainscl - walk10finalpainscl1) %>%
filter(final_pain_diff != 0 | is.na(final_pain_diff)) %>%
add_column(
error_type = "Walk test:Discrepancy or missing values in the first and the second initial pain ratings"
) %>%
add_column(errors = "error") %>%
select(
record_id,
redcap_data_access_group,
redcap_repeat_instrument,
redcap_repeat_instance,
error_type,
errors
)F.2.5.1.3 Flag 3:
Check for missing values or discrepancy between the first and the second walk time.
ferror3 <- frdata3 %>%
retype() %>%
mutate(walk_time_diff = walk10time - walk10time1) %>%
filter(walk_time_diff != 0 | is.na(walk_time_diff)) %>%
add_column(
error_type = "Walk test:Missing values or discrepancy between the first and the second walk time"
) %>%
add_column(errors = "error") %>%
select(
record_id,
redcap_data_access_group,
redcap_repeat_instrument,
redcap_repeat_instance,
error_type,
errors
)F.2.5.1.4 Flag 4:
Check if the reason for test not completed was not specified.
ferror4 <- tka_func %>%
filter(walk10completeyn == 0) %>%
filter(is.na(walk10incompletereason)) %>%
add_column(
error_type = "Walk test:If the reason for test not completed was marked off"
) %>%
add_column(errors = "error") %>%
select(
record_id,
redcap_data_access_group,
redcap_repeat_instrument,
redcap_repeat_instance,
error_type,
errors
)F.2.5.1.5 Flag 5:
Check for records with missing values for any assistance if the walk test was completed.
ferror5 <- frdata3 %>%
filter(walk10completeyn == 1) %>%
filter(is.na(walk10assistyn)) %>%
add_column(
error_type = "Walk test:missing values for any assistance if the walk test was completed"
) %>%
add_column(errors = "error") %>%
select(
record_id,
redcap_data_access_group,
redcap_repeat_instrument,
redcap_repeat_instance,
error_type,
errors
)F.2.5.1.6 Flag 6:
For subjects who used any assistance during the walk test, check for records with type of assistance left “unchecked”.
ferror6 <- frdata3 %>%
filter(walk10assistyn == 1) %>%
filter(
walk10assist_cane___1 == 0 &
walk10assist_crutch___1 == 0 &
walk10assist_perssuppt___1 == 0 &
walk10assist_other___1 == 0 &
walk10assist_walkder___1 == 0
) %>%
add_column(error_type = "Walk test:type of assistance unchecked") %>%
add_column(errors = "error") %>%
select(
record_id,
redcap_data_access_group,
redcap_repeat_instrument,
redcap_repeat_instance,
error_type,
errors
)F.2.5.2 5 times sit to stand test (5tsts)
F.2.5.2.1 Flag 7:
Check for missing bp values if 5tsts was completed.
frdata.sit <- tka_func
ferror.bp <- frdata.sit %>%
filter(tstscompleteyn == 1 & is.na(tstsbpscreen)) %>%
add_column(error_type = "5tsts:missing bp values if 5tsts was completed") %>%
add_column(errors = "error") %>%
select(
record_id,
redcap_data_access_group,
redcap_repeat_instrument,
redcap_repeat_instance,
error_type,
errors
)keep records for subjects who completed the test.
frdata.sit1 <- frdata.sit %>%
filter(tstscompleteyn == 1 & functional_testing_complete == 2)F.2.5.2.2 Flag 8:
Check for discrepancy or missing values in the first and the second initial pain rating.
ferror1.sit <- frdata.sit1 %>%
mutate(init_pain_diff.sit = tstsprepainscl - tstsprepainscl1) %>%
filter(init_pain_diff.sit != 0 | is.na(init_pain_diff.sit)) %>%
add_column(
error_type = "5tsts:discrepancy or missing values for the first and the second initial pain rating "
) %>%
add_column(errors = "error") %>%
select(
record_id,
redcap_data_access_group,
redcap_repeat_instrument,
redcap_repeat_instance,
error_type,
errors
)F.2.5.2.3 Flag 9:
Check for discrepancy or missing values in the first and the second final pain rating.
ferror2.sit <- frdata.sit1 %>%
mutate(final_pain_diff.sit = tstspostpainscl - tstspostpainscl1) %>%
filter(final_pain_diff.sit != 0 | is.na(final_pain_diff.sit)) %>%
add_column(
error_type = "5tsts:discrepancy or missing values for the first and the second final pain rating "
) %>%
add_column(errors = "error") %>%
select(
record_id,
redcap_data_access_group,
redcap_repeat_instrument,
redcap_repeat_instance,
error_type,
errors
)F.2.5.2.4 Flag 10:
Check for discrepancy or missing values the first and the second activity time.
ferror3.sit <- frdata.sit1 %>%
retype() %>%
mutate(sit_time_diff = tststime - tststime1) %>%
filter(sit_time_diff != 0 | is.na(sit_time_diff)) %>%
add_column(
error_type = "5tsts:discrepancy or missing values the first and the second activity time"
) %>%
add_column(errors = "error") %>%
select(
record_id,
redcap_data_access_group,
redcap_repeat_instrument,
redcap_repeat_instance,
error_type,
errors
)F.2.5.2.5 Flag 11:
Check if the reason for test not completed was not specified.
ferror4.sit <- frdata.sit %>%
filter(tstscompleteyn == 0) %>%
filter(is.na(tstsnonreasonyn)) %>%
add_column(
error_type = "5tsts:if the reason for test not completed was not specified"
) %>%
add_column(errors = "error") %>%
select(
record_id,
redcap_data_access_group,
redcap_repeat_instrument,
redcap_repeat_instance,
error_type,
errors
)F.2.5.2.6 Flag 12:
Check if the test was not completed but was Initiated, and the number of reps completed (n/5) were not specified)
ferror5.sit <- frdata.sit %>%
filter(tstscompleteyn == 0) %>%
filter(tstsnonreasonyn == 1) %>%
filter(is.na(tstsnumbrepsyn)) %>%
add_column(
error_type = "5tsts:If the test was not completed but was Initiated, and the number of reps completed (n/5) were not specified"
) %>%
add_column(errors = "none") %>%
select(
record_id,
redcap_data_access_group,
redcap_repeat_instrument,
redcap_repeat_instance,
error_type,
errors
)F.2.5.2.7 Flag 13:
Check for records with missing values for any assistance if the test was completed.
ferror6.sit <- frdata.sit1 %>%
filter(tstscompleteyn == 1) %>%
filter(is.na(tstsassistyn)) %>%
add_column(
error_type = "5tsts:missing values for any assistance if the test was completed"
) %>%
add_column(errors = "error") %>%
select(
record_id,
redcap_data_access_group,
redcap_repeat_instrument,
redcap_repeat_instance,
error_type,
errors
)F.2.5.2.8 Flag 14:
For subjects who used any assistance during the walk test, check for records with type of assistance left “unchecked”.
ferror7.sit <- frdata.sit1 %>%
filter(tstsassistyn == 1) %>%
filter(tstsassist_1___1 == 0 & tstsassist_2___1 == 0) %>%
add_column(error_type = "5tsts:type of assistance unchecked") %>%
add_column(errors = "error") %>%
select(
record_id,
redcap_data_access_group,
redcap_repeat_instrument,
redcap_repeat_instance,
error_type,
errors
)F.2.5.2.9 Biomarkers for the TKA cohort
F.2.5.2.9.1 Movement Evoked Pain (MEP):
MEP refers to pain that occurs or is worsened by movement (Berardi et al., 2022). MEP can not be computed if any of the pain ratings are missing.
Berardi, G., Frey-Law, L., Sluka, K. A., Bayman, E. O., Coffey, C. S., Ecklund, D., Vance, C. G. T., Dailey, D. L., Burns, J., Buvanendran, A., McCarthy, R. J., Jacobs, J., Zhou, X. J., Wixson, R., Balach, T., Brummett, C. M., Clauw, D., Colquhoun, D., Harte, S. E., … Wandner, L. D. (2022). Multi-site observational study to assess biomarkers for susceptibility or resilience to chronic pain: The acute to chronic pain signatures (A2CPS) study protocol. Frontiers in Medicine, 9. https://doi.org/10.3389/fmed.2022.849214
F.2.5.2.9.2 10m Walk Test MEP:
MEP 10m walk test= Final pain rating - Initial pain rating
tka_func <- tka_func %>%
mutate(mep_walk = walk10finalpainscl - walk10initialpainscl)F.2.5.2.9.3 5TSTS Test MEP:
MEP 5TSTS test = Final pain rating - Initial pain rating
tka_func <- tka_func %>%
mutate(mep_5tsts = tstspostpainscl - tstsprepainscl)F.2.5.2.10 New field name(s):
Add field names for the computed biomarkers to the Functional Testing data dictionary
# Create field names
walk_mep_new_row <- data.frame(
field_name = "mep_walk",
field_type = "numeric",
field_note = "Final pain rating - initial pain rating (10m walk test) "
)
tsts_mep_new_row <- data.frame(
field_name = "mep_5tsts",
field_type = "numeric",
field_note = "Final pain rating - initial pain rating (5TSTS test)"
)
# Add the new row after the last row
func_dict <- func_dict %>%
slice(1:nrow(.)) %>%
add_row(.after = nrow(.), !!!walk_mep_new_row) %>%
add_row(.after = nrow(.), !!!tsts_mep_new_row)F.2.5.2.11 Create functional assessment error report for the TKA cohort.
# Specify the common prefix
func_error <- "ferror"
# Find data frames in the global environment with the specified prefix
func_list <- mget(ls(pattern = paste0("^", func_error)))
# Combine the data frames using bind_rows
tka_func_report <- bind_rows(func_list) %>%
pivot_wider(names_from = "error_type", values_from = "errors") %>%
mutate_all(~ replace_na(., ""))tka_func_report %>%
gt() %>%
tab_header(
title = md("**TKA Functional Assessment Error Report**")
) %>%
tab_options(
table.font.size = px(12),
column_labels.font.size = px(12)
) %>%
tab_style(
style = list(cell_fill(color = "#F4F4F4")),
locations = cells_body(columns = record_id)
)| TKA Functional Assessment Error Report | ||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| record_id | redcap_data_access_group | redcap_repeat_instrument | redcap_repeat_instance | 5tsts:missing bp values if 5tsts was completed | Walk test:Discrepancy or missing values in the first and the second initial pain ratings | 5tsts:discrepancy or missing values for the first and the second initial pain rating | 5tsts:discrepancy or missing values for the first and the second final pain rating | Walk test:Missing values or discrepancy between the first and the second walk time | 5tsts:discrepancy or missing values the first and the second activity time | Walk test:missing values for any assistance if the walk test was completed | 5tsts:missing values for any assistance if the test was completed | 5tsts:type of assistance unchecked |
| 10135 | uchicago | functional_testing | 1 | error | ||||||||
| 10308 | uchicago | functional_testing | 1 | error | error | |||||||
| 10341 | uchicago | functional_testing | 1 | error | ||||||||
| 25014 | university_of_mich | functional_testing | 1 | error | ||||||||
| 25048 | university_of_mich | functional_testing | 1 | error | ||||||||
| 25051 | university_of_mich | functional_testing | 1 | error | ||||||||
| 25053 | university_of_mich | functional_testing | 1 | error | error | |||||||
| 25081 | university_of_mich | functional_testing | 1 | error | ||||||||
| 25090 | university_of_mich | functional_testing | 2 | error | error | |||||||
| 25095 | university_of_mich | functional_testing | 1 | error | ||||||||
| 25158 | university_of_mich | functional_testing | 1 | error | ||||||||
| 25166 | university_of_mich | functional_testing | 1 | error | ||||||||
| 25171 | university_of_mich | functional_testing | 2 | error | ||||||||
| 25224 | university_of_mich | functional_testing | 1 | error | ||||||||
| 10055 | uchicago | functional_testing | 1 | error | ||||||||
| 10363 | uchicago | functional_testing | 1 | error | error | |||||||
| 10696 | uchicago | functional_testing | 1 | error | error | |||||||
| 10040 | uchicago | functional_testing | 1 | error | error | error | ||||||
| 10147 | uchicago | functional_testing | 1 | error | ||||||||
| 10688 | northshore | functional_testing | 1 | error | error | error | ||||||
| 10598 | uchicago | functional_testing | 1 | error | ||||||||
| 25141 | university_of_mich | functional_testing | 1 | error | ||||||||
| 10321 | uchicago | functional_testing | 1 | error | ||||||||
| 10436 | uchicago | functional_testing | 1 | error | ||||||||
| 25108 | university_of_mich | functional_testing | 2 | error | ||||||||
| 25204 | university_of_mich | functional_testing | 1 | error | ||||||||
| 25013 | university_of_mich | functional_testing | 1 | error | ||||||||
F.2.5.2.12 Save:
Save “tka_func” and data dictionary as .csv files in the folder named “reformatted_functional”
write_csv(
tka_func,
file = here::here(
"data",
"functional-testing",
"Reformatted",
"reformatted_tka_func.csv"
)
)
write_csv(
func_dict,
file = here::here(
"data",
"functional-testing",
"Reformatted",
"updated_func_dict.csv"
)
)F.2.6 Thoracotomy cohort Functional Assessment:
thor_func <- func %>%
select(
record_id,
guid,
redcap_data_access_group,
redcap_event_name,
redcap_repeat_instrument,
redcap_repeat_instance,
starts_with("ftdbc"),
functional_testing_mcc2_v01_complete,
cohort
)thor_func <- thor_func %>%
filter(cohort == "Thoracic") %>%
filter(redcap_repeat_instrument == "functional_testing_mcc2_v01") %>%
filter(
functional_testing_mcc2_v01_complete == 2 & !is.na(redcap_repeat_instance)
) %>%
group_by(record_id, redcap_event_name) %>%
top_n(1, redcap_repeat_instance) %>%
ungroup() %>%
select(where(not_all_na))Keep records of subjects who completed the test.
m2frfunc3 <- thor_func %>%
filter(ftdbctestcmpltyn == 1)F.2.6.0.1 Flag 1:
Check for discrepancy or missing values in the first and the second initial pain ratings.
m2frerror1 <- m2frfunc3 %>%
mutate(init_pain_diff = ftdbcdeepbrthinitscl - ftdbcdeepbrthinitscl2) %>%
filter(init_pain_diff != 0 | is.na(init_pain_diff)) %>%
add_column(
error_type = "Deep breathing & coughing:Discrepancy or missing values in the first and the second initial pain ratings"
) %>%
add_column(errors = "error") %>%
select(
record_id,
redcap_data_access_group,
redcap_repeat_instrument,
redcap_repeat_instance,
error_type,
errors
)F.2.6.0.2 Flag 2:
Check for discrepancy or missing values in the first and the second final pain ratings.
m2frerror2 <- m2frfunc3 %>%
mutate(final_pain_diff = ftdbcdeepbrthfinalscl - ftdbcdeepbrthfinalscl2) %>%
filter(final_pain_diff != 0 | is.na(final_pain_diff)) %>%
add_column(
error_type = "Deep breathing & coughing:Discrepancy or missing values in the first and the second initial pain ratings"
) %>%
add_column(errors = "error") %>%
select(
record_id,
redcap_data_access_group,
redcap_repeat_instrument,
redcap_repeat_instance,
error_type,
errors
)F.2.6.0.3 Flag 3:
Check for missing values or discrepancy between the first and the second cough pain.
m2frerror3 <- m2frfunc3 %>%
mutate(cough_diff = ftdbccoughfinalscl - ftdbccoughfinalscl2) %>%
filter(cough_diff != 0 | is.na(cough_diff)) %>%
add_column(
error_type = "Deep breathing & coughing: Missing values or discrepancy between the first and the second cough pain"
) %>%
add_column(errors = "error") %>%
select(
record_id,
redcap_data_access_group,
redcap_repeat_instrument,
redcap_repeat_instance,
error_type,
errors
)F.2.6.0.4 Flag 4:
Check if the reason for test not completed was not specified.
m2frerror4 <- thor_func %>%
filter(ftdbctestcmpltyn == 0) %>%
filter(is.na(ftdbctestcmpltno)) %>%
add_column(
error_type = "If the reason for test not completed was not specified"
) %>%
add_column(errors = "error") %>%
select(
record_id,
redcap_data_access_group,
redcap_repeat_instrument,
redcap_repeat_instance,
error_type,
errors
)F.2.6.0.5 Flag 5:
Check if the “functional_testing_mcc2_v01_complete” is missing but the test completion status is available.
m2frerror5 <- thor_func %>%
filter(!is.na(ftdbctestcmpltyn)) %>%
filter(is.na(functional_testing_mcc2_v01_complete)) %>%
add_column(error_type = "missing functional_testing_mcc2_v01_complete") %>%
add_column(errors = "error") %>%
select(
record_id,
redcap_data_access_group,
redcap_repeat_instrument,
redcap_repeat_instance,
error_type,
errors
)F.2.6.0.6 Create functional assessment error report for the Thoracotomy cohort.
# Specify the common prefix
m2func_error <- "m2frerror"
# Find data frames in the global environment with the specified prefix
m2func_list <- mget(ls(pattern = paste0("^", m2func_error)))
# Combine the data frames using bind_rows
thor_func_report <- bind_rows(m2func_list) %>%
pivot_wider(names_from = "error_type", values_from = "errors") %>%
mutate_all(~ replace_na(., ""))thor_func_report %>%
gt() %>%
tab_header(
title = md(
"**Thoracotomy Cohort Functional Assessment Error Report: No Errors**"
)
) %>%
tab_options(
table.font.size = px(12),
column_labels.font.size = px(12)
) %>%
tab_style(
style = list(cell_fill(color = "#F4F4F4")),
locations = cells_body(columns = record_id)
)| Thoracotomy Cohort Functional Assessment Error Report: No Errors | |||
|---|---|---|---|
| record_id | redcap_data_access_group | redcap_repeat_instrument | redcap_repeat_instance |
F.2.6.0.7 Biomarkers for the Thoraic cohort:
Post baseline data is needed to compute MEP for the thoracic cohort, hence could not be computed.
F.2.6.0.8 Save:
Save “thor_func” and data dictionary as .csv files in the folder named “reformatted_functional”
write_csv(
thor_func,
file = here::here(
"data",
"functional-testing",
"Reformatted",
"reformatted_thor_func.csv"
)
)
write_csv(
func_dict,
file = here::here(
"data",
"functional-testing",
"Reformatted",
"updated_func_dict.csv"
)
)