In Thaliehln/timci: Data management and visualisation for the TIMCI project

\newpage

write("Export withdrawal data and run corresponding quality checks", stderr())

Withdrawal quality checks

db_name <- "withdrawal"

n_raw_withdrawal_records <- 0
raw_withdrawal_is_not_empty <- timci::is_not_empty(raw_withdrawal_data)

n_raw_withdrawal_records <- nrow(raw_withdrawal_data)

There are r n_raw_withdrawal_records records in the raw r db_name database.

raw_withdrawal_data <- raw_withdrawal_data %>%
  dplyr::select(date,
                start,
                end,
                deviceid,
                `page-qr_code_avail`,
                `page-a1_a_4`,
                `page-a1_a_4a`,
                `page-study`,
                `page-withdrawal`,
                `page-withdrawal_reason`,
                `meta-instanceID`)

n_drop_nonvalid_ids <- 0
n_no_doc_child_id_withdrawal <- 0
n_invalid_date_of_withdrawal <- 0

Participant identification

Rule: withdrawals with documented child IDs are kept in the locked Day 0 database, then removed in all the databases where entries are posterior to the date of withdrawal. If withdrawn IDs cannot be found in the locked Day 0 database, it is assumed that the caregiver withdrew during data collection. More information will be collected about this in the withdrawal form.

if (nrow(allday0_data) > 0) {
  if (!is.null(raw_withdrawal_data)) {
    if (nrow(raw_withdrawal_data) > 0) { 
      raw_withdrawal_data$child_id <- ifelse(!is.na(raw_withdrawal_data$'page-a1_a_4'), raw_withdrawal_data$'page-a1_a_4', raw_withdrawal_data$'page-a1_a_4a')
      withdrawal_data2 <- raw_withdrawal_data[!is.na(raw_withdrawal_data$child_id), ]
      withdrawal_data_no_doc <- raw_withdrawal_data[is.na(raw_withdrawal_data$child_id), ]
      n_no_doc_child_id_withdrawal <- nrow(withdrawal_data_no_doc)
      withdrawal_data2$child_id <- gsub("O", "0", withdrawal_data2$child_id)
    }
  }
}

if (nrow(allday0_data) > 0) {
  if (!is.null(raw_withdrawal_data)) {
    if (nrow(raw_withdrawal_data) > 0) { 
      cat(paste0("**", nrow(raw_withdrawal_data), "** withdrawal(s) reported (among which **", nrow(withdrawal_data2),"** withdrawals with documented child IDs)."))
    } else {
      cat("No withdrawal reported.")
    }
  }
} else {
  cat("N/A")
}

day0_data_id_withdrawal <- NULL
non_reconciled_withdrawn_ids <- NULL
`%!in%` <- Negate(`%in%`)

if (nrow(allday0_data) > 0) {
  if (!is.null(raw_withdrawal_data)) {
    if (nrow(raw_withdrawal_data) > 0) {

      # Non reconciled withdrawn IDs
      non_reconciled_withdrawn_ids <- withdrawal_data2[withdrawal_data2$child_id %!in% allday0_data$child_id, ]
      n_drop_nonvalid_ids <- nrow(non_reconciled_withdrawn_ids)

      # Reconciled withdrawn IDs
      day0_data_id_withdrawal <- allday0_data[allday0_data$child_id %in% withdrawal_data2$child_id, ] %>%
          dplyr::select(child_id,
                        fid,
                        date_visit,
                        uuid)
      day0_data_id_withdrawal <- day0_data_id_withdrawal %>%
        merge(withdrawal_data2,
              by = 'child_id',
              all.x = TRUE)

    }
  }
}

if (nrow(allday0_data) > 0) {
  if (!is.null(day0_data_id_withdrawal)) {
    cat(paste0("**", nrow(day0_data_id_withdrawal), "** participant(s) who withdrew found in the locked Day 0 database."))
  }
}

if (nrow(allday0_data) > 0) {
  if (!is.null(non_reconciled_withdrawn_ids)) {
    non_reconciled_withdrawn_ids %>%
      dplyr::select(c("child_id",
                      "date",
                      "page-withdrawal_reason")) %>%
      knitr::kable(col.names = c("Child ID",
                                 "Withdrawal date",
                                 "Documented reason"),
                   caption = "List of withdrawals which could not be reconciled with any child ID in the Day 0 locked database")
  }
}

timci::quality_check_export(non_reconciled_withdrawn_ids,
                            qc_nonreconciled_withdrawals_id,
                            "non_reconciled_withdrawn_ids",
                            qc_dir,
                            "withdrawals were not reconciled")

Invalid date of withdrawal [Context check r qc_withdrawal_before_enrolment]

write(" o Invalid date of withdrawal", stderr())

qc_description <- "The withdrawal date should have happened after enrolment at Day 0."
qc_rule <- action_alert_no_modification
qc_type <- "date_discrepancy"
df <- day0_data_id_withdrawal
col_date1 <- "date"
col_date2 <- "date_visit"
qc_text <- "a date of withdrawal before the enrolment date"
qc_idx <- qc_withdrawal_before_enrolment
qc_export_label <- "withdrawal_before_enrolment"
qc_export_description <- "the date of withdrawal was before the enrolment date"
cat(knitr::knit_child('database_export_sub_quality_check.Rmd',
                      envir = environment(),
                      quiet = TRUE))
n_invalid_date_of_withdrawal <- n_detected

Pseudonymisation

write(" o Pseudonymisation", stderr())

Pseudonymisation is performed using a cryptographic hash function (md5) that takes strings as input and produces a random-like fixed-length output.

day0_data_id_withdrawal <- day0_data_id_withdrawal %>%
  dplyr::select(-"page-a1_a_4",
                -uuid) %>%
  dplyr::rename(uuid = `meta-instanceID`) %>% 
  dplyr::rowwise() %>%
  dplyr::mutate(uuid = ifelse(uuid != "", digest(uuid, algo = crypto_algo), ""),
                child_id = ifelse(child_id != "", digest(child_id, algo = crypto_algo), ""),
                deviceid = ifelse(deviceid != "", digest(deviceid, algo = crypto_algo), "")) %>%
  dplyr::ungroup()

n_cleaned_withdrawal_records <- nrow(day0_data_id_withdrawal)

Data cleaning summary

write(" o Data cleaning summary", stderr())

timci::create_withdrawal_qc_flowchart(n_raw_withdrawal_records,
                                      n_no_doc_child_id_withdrawal,
                                      n_drop_nonvalid_ids,
                                      n_invalid_date_of_withdrawal,
                                      n_cleaned_withdrawal_records)

Data export

write(" o Export withdrawal data", stderr())

timci::dataset_export(raw_withdrawal_data,
                      "99",
                      "withdrawal_data",
                      params$rctls_dir,
                      "Raw withdrawal data")

timci::dataset_export(day0_data_id_withdrawal,
                      "99",
                      "withdrawal_data",
                      locked_db_dir,
                      "Cleaned withdrawal data") 

Thaliehln/timci documentation built on April 8, 2024, 3:38 p.m.