dqLib

dqLib is an R package for data quality assessment and reporting. dqLib provides methods for calculating data quality metrics and generating reports on detected data quality issues, especially in CORD-MI.

Acknowledgement: This work was done within the “Collaboration on Rare Diseases” of the Medical Informatics Initiative (CORD-MI) funded by the German Federal Ministry of Education and Research (BMBF), under grant number: 01ZZ1911R, FKZ-01ZZ1911R

Data Quality Metrics and Reports

• dqLib provides functions for creating specific reporting scripts that enable user to select desired data quality dimensions and indicators. The data quality reports provide adequate information to find the data quality violations and the causes of these violations.

• The following data quality dimensions, indicatos and parameters are implemented:

  Dimension	Data Quality Indicator Name
  completeness	item completeness rate, value completeness rate, orphaCoding completeness rate
  plausibility	orphaCoding plausibility rate, range plausibility rate
  uniqueness	RD case unambiguity rate, RD case dissimilarity rate
  concordance	concordance of RD cases, concordance of tracer cases

  No.	Data Quality Parameter Name	Description
  P1	missing data items	number of missing data items per year
  P2	mandatory data items	number of mandatory items per year
  P3	missing data values	number of missing data values per year
  P4	available data values	number of available data values per year
  P5	missing orphacodes	number of missing Orphacodes per year
  P6	tracer diagnoses	number of tracer RD diagnoses per year
  P7	implausible links	number of implausible code links per year
  P8	checked for outliers	number of checked data values for outliers per year
  P9	outliers	number of detected outliers per year
  P10	ambigous RD cases	number of ambigous RD cases per year
  P11	RD cases	number of RD cases per year
  P13	duplicated RD cases	number of duplicated RD cases per year
  P14	tracer cases	number of tracer RD cases per year
  P15	inpatient cases	number of inpatient cases per year
  P16	RD cases rel. frequency	relative frequency of inpatient RD cases per year
  P17	tracer cases rel. frequency	relative frequency of inpatient tracer RD cases per year
  P18	available cases	number of available cases per year
  P19	available patients	number of available patients per year
  P20	orphacodes	number of available orphacodes per year
  P21	orpha-coded cases	number of available orpha-coded cases per year
  P22	unambigous RD cases	number of unambigous RD cases per year

• The following references are required to assess the quality of RD documentation:

  • Current Version of Alpha-ID-SE Terminology [1]

  • A reference for tracer diagnoses such as the list provided in [2].

    [1] BfArM - Alpha-ID-SE [Internet]. [cited 2022 May 23]. Available from: https://www.bfarm.de/EN/Code-systems/Terminologies/Alpha-ID-SE/_node.html

    [2] List of Tracer Diagnoses Extracted from Alpha-ID-SE Terminology [Internet]. 2022 [cited 2022May 24]. Available from: https://doi.org/21.11101/0000-0007-F6DF-9

---

Installation

You can install dqLib from local folder with:

devtools::install_local("./dqLib")

You can also install it directly from github with:

devtools::install_github("https://github.com/medizininformatik-initiative/dqLib")

Example

Here are examples for data quality analysis and reporting using this package

• cordDQCheck.R for generating data quality reports in CORD-MI.
• Here you can see the resulting files

Note

The default data quality dimensions are completeness, plausibility, uniqueness and concordance. Howerver, this framework allows the user to select desired quality dimensions and indicators as well as to generate user defined DQ reports.

To cite dqLib, please use the following BibTeX entry:

@software{Tahar_dqLib,
author = {Tahar, Kais},
title = {{dqLib}},
url = {https://github.com/KaisTahar/dqLib}
year = {2021}
}

See also: CORD-MI