Difference between revisions of "Journal:Timely delivery of laboratory efficiency information, Part I: Developing an interactive turnaround time dashboard at a high-volume laboratory"

From LIMSWiki
Jump to navigationJump to search
(Saving and adding more.)
(Saving and adding more.)
Line 43: Line 43:


Historically, the NHLS CDW TAT reports generated were static and reported only the mean TAT. Turnaround time data have a positive skewness, that is, a long tail to the right, meaning that the mean will be greater than the median. This implies that TAT data reported previously<ref name="HawkinsLab07" /><ref name="CoetzeeUsing18">{{cite journal |title=Using laboratory data to categorise CD4 laboratory turn-around-time performance across a national programme |journal=African Journal of Laboratory Medicine |author=Coetzee, L.-M.; Cassim, N.; Glencross, D.K. |volume=7 |issue=1 |at=665 |year=2018 |doi=10.4102/ajlm.v7i1.665 |pmid=30167387 |pmc=PMC6111574}}</ref>, reporting the mean TAT, masks good performance while concealing poor efficiency. Further, neither the current LIS nor CDW reports enable detailed analysis of the [[information]] or drilling down to laboratory- or test-level data for additional information about TAT efficiencies. Data presented at the first conference of the African Society for Laboratory Medicine in Cape Town, South Africa, in 2012, reported daily laboratory test volumes and mean TAT for authorized results, stratified by individual laboratories<ref name="DruryUsing12">{{cite journal |title=Using Central Data Warehouse (CDW) Reports for Monitoring CD4 Laboratory Workload and Related Turn- Around-Time (TAT) |journal=Proceedings of the First International Conference of the African Society for Laboratory Medicine |author=Drury, S.; Coetzee, L.-M.; Cassim, N. et al. |year=2012 |doi=10.13140/2.1.1277.7922}}</ref>, providing a snapshot of performance. This enabled review of CD4 laboratory efficiency for a national program and provided important insights into laboratory operations.
Historically, the NHLS CDW TAT reports generated were static and reported only the mean TAT. Turnaround time data have a positive skewness, that is, a long tail to the right, meaning that the mean will be greater than the median. This implies that TAT data reported previously<ref name="HawkinsLab07" /><ref name="CoetzeeUsing18">{{cite journal |title=Using laboratory data to categorise CD4 laboratory turn-around-time performance across a national programme |journal=African Journal of Laboratory Medicine |author=Coetzee, L.-M.; Cassim, N.; Glencross, D.K. |volume=7 |issue=1 |at=665 |year=2018 |doi=10.4102/ajlm.v7i1.665 |pmid=30167387 |pmc=PMC6111574}}</ref>, reporting the mean TAT, masks good performance while concealing poor efficiency. Further, neither the current LIS nor CDW reports enable detailed analysis of the [[information]] or drilling down to laboratory- or test-level data for additional information about TAT efficiencies. Data presented at the first conference of the African Society for Laboratory Medicine in Cape Town, South Africa, in 2012, reported daily laboratory test volumes and mean TAT for authorized results, stratified by individual laboratories<ref name="DruryUsing12">{{cite journal |title=Using Central Data Warehouse (CDW) Reports for Monitoring CD4 Laboratory Workload and Related Turn- Around-Time (TAT) |journal=Proceedings of the First International Conference of the African Society for Laboratory Medicine |author=Drury, S.; Coetzee, L.-M.; Cassim, N. et al. |year=2012 |doi=10.13140/2.1.1277.7922}}</ref>, providing a snapshot of performance. This enabled review of CD4 laboratory efficiency for a national program and provided important insights into laboratory operations.
A recent evaluation of TAT for HIV programs reported a methodology to further categorize laboratory TAT performance using three additional measures<ref name="CoetzeeUsing18" />: (1) median TAT, (2) 75th percentile TAT (tail size), and (3) percentage within cut-off TAT. These data were graphically presented using a scatter plot of percentage of samples within the TAT cut-off (x-axis) against the TAT 75th percentile (y-axis), categorized into four quadrants of performance to help identify the level of laboratory performance in a national program. This approach made it easier to identify both good performers and outliers in the same analysis.<ref name="CoetzeeUsing18" /> The report was generated in Excel and included the raw monthly data, a scatter and bar graph, and a summary table of all laboratories per business region, the 75th percentile, the percentage within cut-off values, and TAT component 75th percentile values. This report was primarily distributed to managers of all HIV-related operational programs, that is, CD4, viral load and tuberculosis testing, and early infant diagnosis for review and intervention, and not shared across the network of testing laboratories.
Senior management in Gauteng, South Africa expressed a need for a TAT monitoring tool that would enable them to better manage their laboratories and identify sites with poor TAT performance. Given the static nature of the historic TAT reporting in the organization, an interactive system that offered information to enable review of performance, including outliers (tail size assessment), while confirming that sites were meeting cutoffs, would be a useful tool to enable business and laboratory managers alike to monitor their efficiency via TAT performance in real time. The concepts already developed and in use as Excel reports for the HIV programs were the starting point for developing a reporting dashboard for use across multiple disciplines and tests done throughout the network of 241 testing laboratories of the NHLS in South Africa.
The aim of this study was to develop an easy-to-use information system, in a dashboard format. This would enable weekly reporting of TAT data as a snapshot of performance. To achieve this, a number of changes to current TAT reporting had to be addressed. These changes included (1) moving from program-specific, single-test TAT reporting previously used<ref name="CoetzeeUsing18" /> to a specific set of high-volume tests, (2) adopting TAT measures reported by Coetzee ''et al.''<ref name="CoetzeeUsing18" />, (3) identifying dashboard software to use, and (4) identifying the target users.<ref name="CoetzeeUsing18" /> The specific set of tests (or "basket" of commonly requested tests) should be representative across the primary pathology disciplines.
This article sets out to describe the process followed to develop the TAT dashboard—using available software—that could provide a weekly summary of national, business unit, and laboratory-level TAT performance for a basket of tests. For the purposes of this article, data from a single participating tertiary laboratory were used to illustrate the data distribution, as it represents an example of a testing facility that performed all tests reviewed in the prescribed national basket. This was done to show the respective levels of drilling functionality of the dashboard and to iterate the interactive properties, while demonstrating how the dashboard can be used to assess performance and identify outliers for intervention.
==Methods==
===Ethical considerations===
Ethics clearance for this study was obtained from the University of the Witwatersrand (M1706108). Only anonymized laboratory data, containing no patient identifiers, were used for the study.
===Study design===
A retrospective descriptive study design was used to analyze and report laboratory TAT data for a specific set of tests (Table 1).
[[File:Tab1 Cassim AfricanJLabMed2020 9-2.jpg|600px]]
{{clear}}
{|
| STYLE="vertical-align:top;"|
{| border="0" cellpadding="5" cellspacing="0" width="600px"
|-
  | style="background-color:white; padding-left:10px; padding-right:10px;"| <blockquote>'''Table 1.''' Sample of a turnaround time dashboard table that lists the outcomes for the basket of tests, South Africa, 2018</blockquote>
|-
|}
|}





Revision as of 21:26, 18 January 2021

Full article title Timely delivery of laboratory efficiency information, Part I: Developing an interactive turnaround time dashboard at a high-volume laboratory
Journal African Journal of Laboratory Medicine
Author(s) Cassim, Naseem; Tepper, Manfred E.; Coetzee, Lindi M.; Glencross, Deborah K.
Author affiliation(s) National Health Laboratory Service, University of the Witwatersrand
Primary contact Email: naseem dor cassim at wits dot ac dot za
Year published 2020
Volume and issue 9(2)
Article # a947
DOI 10.4102/ajlm.v9i2.947
ISSN 2225-2010
Distribution license Creative Commons Attribution 4.0 International License
Website https://ajlmonline.org/index.php/ajlm/article/view/947/1482
Download https://ajlmonline.org/index.php/ajlm/article/download/947/1479 (PDF)

Abstract

Background: Mean turnaround time (TAT) reporting for testing laboratories in a national network is typically static and not immediately available for meaningful corrective action and does not allow for test-by-test or site-by-site interrogation of individual laboratory performance.

Objective: The aim of this study was to develop an easy-to-use, visual dashboard to report interactive graphical TAT data to provide a weekly snapshot of TAT efficiency.

Methods: An interactive dashboard was developed by staff from the National Priority Programme and Central Data Warehouse of the National Health Laboratory Service in Johannesburg, South Africa, during 2018. Steps required to develop the dashboard were summarized in a flowchart. To illustrate the dashboard, one week of data from a busy laboratory for a specific set of tests was analyzed using annual performance plan TAT cutoffs. Data were extracted and prepared to deliver an aggregate extract, with statistical measures provided, including test volumes, global percentage of tests that were within TAT cutoffs, and percentile statistics.

Results: Nine steps were used to develop the dashboard iteratively, with continuous feedback for each step. The data warehouse environment conformed and stored laboratory information system (LIS) data in two formats: (1) fact and (2) dimension. Queries were developed to generate an aggregate TAT data extract to create the dashboard. The dashboard successfully delivered weekly TAT reports.

Conclusion: Implementation of a TAT dashboard can successfully enable the delivery of near real-time information and provide a weekly snapshot of efficiency in the form of TAT performance to identify and quantitate bottlenecks in service delivery.

Keywords: turn-around time, laboratory efficiency, interactive dashboard, indicators, performance assessment

Introduction

|Turnaround time (TAT) is an important performance indicator for a laboratory service. It refers to the time from first registration in a laboratory to a result released in the laboratory information system (LIS).[1] Historically, within the National Health Laboratory Service (NHLS) of South Africa, TAT reporting was provided in annual and quarterly static management reports generated by the LIS, TrakCare[2], which also provided ad hoc reporting for use at the laboratory level. These reports are printed to provide a snapshot of TAT reporting and are suited for staff working at the laboratory level. At a national level, the corporate data warehouse (CDW) of the NHLS collated global TAT data from over 266 testing laboratories based on predetermined, annual performance plan cutoffs. National TAT cutoffs are set by expert committees of different pathology disciplines, with final confirmation from senior management before implementation. These cutoffs are set with provisions for all levels of the service laboratory: from low-volume laboratories with limited test repertoires to high-volume testing laboratories with extensive test offerings, including specialized testing, such as viral load testing. However, a large percentage of NHLS laboratories have 24-hour service and have emergency units in the hospitals in which they are housed; such laboratories have locally stricter TAT cut-offs for emergency and other local tests than are reflected in the national cutoffs for all samples.

Historically, the NHLS CDW TAT reports generated were static and reported only the mean TAT. Turnaround time data have a positive skewness, that is, a long tail to the right, meaning that the mean will be greater than the median. This implies that TAT data reported previously[1][3], reporting the mean TAT, masks good performance while concealing poor efficiency. Further, neither the current LIS nor CDW reports enable detailed analysis of the information or drilling down to laboratory- or test-level data for additional information about TAT efficiencies. Data presented at the first conference of the African Society for Laboratory Medicine in Cape Town, South Africa, in 2012, reported daily laboratory test volumes and mean TAT for authorized results, stratified by individual laboratories[4], providing a snapshot of performance. This enabled review of CD4 laboratory efficiency for a national program and provided important insights into laboratory operations.

A recent evaluation of TAT for HIV programs reported a methodology to further categorize laboratory TAT performance using three additional measures[3]: (1) median TAT, (2) 75th percentile TAT (tail size), and (3) percentage within cut-off TAT. These data were graphically presented using a scatter plot of percentage of samples within the TAT cut-off (x-axis) against the TAT 75th percentile (y-axis), categorized into four quadrants of performance to help identify the level of laboratory performance in a national program. This approach made it easier to identify both good performers and outliers in the same analysis.[3] The report was generated in Excel and included the raw monthly data, a scatter and bar graph, and a summary table of all laboratories per business region, the 75th percentile, the percentage within cut-off values, and TAT component 75th percentile values. This report was primarily distributed to managers of all HIV-related operational programs, that is, CD4, viral load and tuberculosis testing, and early infant diagnosis for review and intervention, and not shared across the network of testing laboratories.

Senior management in Gauteng, South Africa expressed a need for a TAT monitoring tool that would enable them to better manage their laboratories and identify sites with poor TAT performance. Given the static nature of the historic TAT reporting in the organization, an interactive system that offered information to enable review of performance, including outliers (tail size assessment), while confirming that sites were meeting cutoffs, would be a useful tool to enable business and laboratory managers alike to monitor their efficiency via TAT performance in real time. The concepts already developed and in use as Excel reports for the HIV programs were the starting point for developing a reporting dashboard for use across multiple disciplines and tests done throughout the network of 241 testing laboratories of the NHLS in South Africa.

The aim of this study was to develop an easy-to-use information system, in a dashboard format. This would enable weekly reporting of TAT data as a snapshot of performance. To achieve this, a number of changes to current TAT reporting had to be addressed. These changes included (1) moving from program-specific, single-test TAT reporting previously used[3] to a specific set of high-volume tests, (2) adopting TAT measures reported by Coetzee et al.[3], (3) identifying dashboard software to use, and (4) identifying the target users.[3] The specific set of tests (or "basket" of commonly requested tests) should be representative across the primary pathology disciplines.

This article sets out to describe the process followed to develop the TAT dashboard—using available software—that could provide a weekly summary of national, business unit, and laboratory-level TAT performance for a basket of tests. For the purposes of this article, data from a single participating tertiary laboratory were used to illustrate the data distribution, as it represents an example of a testing facility that performed all tests reviewed in the prescribed national basket. This was done to show the respective levels of drilling functionality of the dashboard and to iterate the interactive properties, while demonstrating how the dashboard can be used to assess performance and identify outliers for intervention.

Methods

Ethical considerations

Ethics clearance for this study was obtained from the University of the Witwatersrand (M1706108). Only anonymized laboratory data, containing no patient identifiers, were used for the study.

Study design

A retrospective descriptive study design was used to analyze and report laboratory TAT data for a specific set of tests (Table 1).


Tab1 Cassim AfricanJLabMed2020 9-2.jpg

Table 1. Sample of a turnaround time dashboard table that lists the outcomes for the basket of tests, South Africa, 2018


References

  1. 1.0 1.1 Hawkins, R.C. (2007). "Laboratory turnaround time". The Clinical Biochemist 28 (4): 179–94. PMC PMC2282400. PMID 18392122. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2282400. 
  2. "TrakCare Lab Enterprise". InterSystems. https://www.intersystems.com/products/trakcare/trakcare-lab-enterprise. Retrieved 03 December 2018. 
  3. 3.0 3.1 3.2 3.3 3.4 3.5 Coetzee, L.-M.; Cassim, N.; Glencross, D.K. (2018). "Using laboratory data to categorise CD4 laboratory turn-around-time performance across a national programme". African Journal of Laboratory Medicine 7 (1): 665. doi:10.4102/ajlm.v7i1.665. PMC PMC6111574. PMID 30167387. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6111574. 
  4. Drury, S.; Coetzee, L.-M.; Cassim, N. et al. (2012). "Using Central Data Warehouse (CDW) Reports for Monitoring CD4 Laboratory Workload and Related Turn- Around-Time (TAT)". Proceedings of the First International Conference of the African Society for Laboratory Medicine. doi:10.13140/2.1.1277.7922. 

Notes

This presentation is faithful to the original, with only a few minor changes to presentation. Grammar was cleaned up for smoother reading. In some cases important information was missing from the references, and that information was added.