blindspiker: Laboratory Blind Spike Sample Analyses

Description

A blind spike program provides samples to a laboratory in order to perform quality control (QC) checks. The samples provided are of a known quantity to the tester. The laboratory is typically uninformed of that the sample provided is a QC sample.

Author(s)

Maintainer: Mark Hogue mark.hogue.chp@gmail.com

See Also

Useful links:

• https://github.com/markhogue/blindspiker

• Report bugs at https://github.com/markhogue/blindspiker/issues

Pipe operator

Description

See magrittr::%>% for details.

Usage

lhs %>% rhs

Value

No return value, called for side effects

Load data associated with the spiked samples and laboratory results from comma separated variable text files.

Description

This function loads spike data and the laboratory results, then processes it so that the 'sample_ID' ties the spike value and the results together, using the 'left_join' function from the 'dplyr' package.

Usage

bs_prep_and_analysis(spike_data, lab_data)

Arguments

Details

False negative results are flagged for laboratory results below the detection level in the analysis of a sample spiked above the detection level. False positives are flagged for laboratory results above the detection when the analyte was not spiked. Error rates are computed with the 'table_false' function.

To use this function, first set up spike value data in a .csv file (in any column order) matching the column headers of the following values:

Required for spike data:

• 'sample_ID' unique identifier, character or numeric

• 'analyte' character data

• 'spike_value' numeric value

• 'spike_units' character data

• 'submission_date' character data that will be converted to date in format YYYY-MM-DD (for example 1999-12-31)

Optional for spike data:

• 'sv_unc' numeric, the uncertainty of the spike value. Default = 0

• 'sv_k' the coverage factor for the spike value uncertainty. Default = 2

• 'provider lab' character name of laboratory providing spiked samples

Required for laboratory results:

• 'sample_ID' must match spike 'sample_ID'

• 'analyte' must match spike 'analyte'

• 'result' numeric value

• 'units' must match 'spike_units'

• 'result_date'

• 'det_lvl' numeric detection level

• 'unc' numeric uncertainty of the laboratory result

#' Optional for lab data:

• 'k' the coverage factor for the result uncertainty. Default = 2

Note that the two data sets (spike values and laboratory results) will be combined by 'sample_ID', and also by 'analyte' if present in both sets. (If the laboratory data includes a non-zero result for an analyte not present in the spike data, that would indicate a false positive.)

Use this function to load the spike data and establish a data frame named 'bs_df'. All plotting and data analysis functions will default to look for this data set.

Value

data frame containing all needed data to be used in subsequent functions.

Examples

example_spike_data <- system.file("extdata", "spikevals.csv", package = "blindspiker")
example_lab_data <- system.file("extdata", "labvals.csv", package = "blindspiker")
example_df <- bs_prep_and_analysis(spike_data = example_spike_data, lab_data = example_lab_data)

quantile-quantile plot

Description

QQ plots by isotope

Usage

plot_qq(select_analyte, dat = bs_df)

Arguments

Value

quantile-quantile plot of laboratory results of spiked samples

Examples

example_spike_data <- system.file('extdata', 'spikevals.csv', package = 'blindspiker')
example_lab_data <- system.file('extdata', 'labvals.csv', package = 'blindspiker')
example_df <- bs_prep_and_analysis(spike_data = example_spike_data, lab_data = example_lab_data)
plot_qq(select_analyte = 'unknownium', dat = example_df)

Plot Run Chart

Description

'plot_run()' produces a Run Chart of the selected analyte. The analyte is selected from the analyte set provided. Results are plotted with error bars (uncertainty with coverage factor of 2) when the result is greater than the detection level. On the ratio version, the uncertainties for the results and spike values are combined as the square root of the sums of the relative uncertainties squared. When original results are plotted, the spike values are shown with a small salmon-colored "+".

Usage

plot_run(
  select_analyte,
  dat = bs_df,
  version = "original",
  log = "n",
  removal_notification = "n"
)

Arguments

Examples

example_spike_data <- system.file('extdata', 'spikevals.csv', package = 'blindspiker')
example_lab_data <- system.file('extdata', 'labvals.csv', package = 'blindspiker')
example_df <- bs_prep_and_analysis(spike_data = example_spike_data, lab_data = example_lab_data)
plot_run(select_analyte = 'unknownium', dat = example_df)

Plot turnaround time

Description

Plot time for laboratory analysis by date (result date from lab - spiked sample submitted date)

Usage

plot_tat(select_analyte, dat = bs_df, target_days = 60)

Arguments

Value

turn-around-time plot

Examples

example_spike_data <- system.file('extdata', 'spikevals.csv', package = 'blindspiker')
example_lab_data <- system.file('extdata', 'labvals.csv', package = 'blindspiker')
example_df <- bs_prep_and_analysis(spike_data = example_spike_data, lab_data = example_lab_data)
plot_tat(select_analyte = 'unknownium', dat = example_df, target_days = 60)

Count combinations of spiked samples

Description

A summary of spiked samples is provided based on combinations of interest.

Usage

spike_combos(analytes, dat = bs_df)

Arguments

Value

table of combinations of analytes in blind spikes

Examples

example_spike_data <- system.file('extdata', 'spikevals.csv', package = 'blindspiker')
example_lab_data <- system.file('extdata', 'labvals.csv', package = 'blindspiker')
example_df <- bs_prep_and_analysis(spike_data = example_spike_data, lab_data = example_lab_data)
spike_combos(analytes = c('unknownium', 'Sr-90'), dat = example_df)

Make table of false positives and false negatives in laboratory results

Description

Make a table with estimated confidence intervals for false negatives and false positives for an analyte in the data set.

Usage

table_false(select_analyte, dat = bs_df)

Arguments

Details

False negatives are the number of laboratory results that missed a spiked value. For the false negative rate, the numerator is the number of laboratory results less than detection level for spiked samples. The denominator is the number or spiked samples.

False positives are the number of laboratory results above detection level when the analyte identified by the laboratory was not in the spiked sample. For the false positive rate, the numerator is the number of false positives. The denominator is the number of false positives plus number of true negatives.

The total error rate is the total number of laboratory results with either false negative or false positive results divided by the total number of laboratory results.

Value

table of false positive and false negative results

Examples

example_spike_data <- system.file('extdata', 'spikevals.csv', package = 'blindspiker')
example_lab_data <- system.file('extdata', 'labvals.csv', package = 'blindspiker')
example_df <- bs_prep_and_analysis(spike_data = example_spike_data, lab_data = example_lab_data)
table_false(select_analyte = "Sr-90", dat = example_df)

Make table summarizing the blind spikes.

Description

Make a table showing how many blind spike samples contained one or more analytes.

Usage

table_spike(dat = bs_df)

Arguments

Value

table of all analytes in blind spike samples

See Also

For combinations of analytes spiked in a single sample, see 'spike_combos'.

Examples

example_spike_data <- system.file('extdata', 'spikevals.csv', package = 'blindspiker')
example_lab_data <- system.file('extdata', 'labvals.csv', package = 'blindspiker')
example_df <- bs_prep_and_analysis(spike_data = example_spike_data, lab_data = example_lab_data)
table_spike(dat = example_df)