ASSIST-CKD

Software development has reached end of life (EOL)

Map of the 23 renal units and pathology laboratories using Assist-CKD

The ASSIST-CKD quality improvement project, was managed by Kidney Research UK and supported by The Health Foundation, and focused on scaling-up an intervention that reports kidney function (eGFR) graphs over time to primary care. The aim was to identify patients at highest risk of CKD progression - a simple concept which not only offers value for money but also a real chance to improve renal patient journeys.

The project involved over twenty renal units and pathology laboratories and their surrounding GP practices, covering an estimated population of 11-12 million people in England, Scotland, Wales and Northern Ireland.This website was provided for participating sites when the project was active and is structured so that it is possible to access software, supporting documents and other materials that were needed by sites carrying out the project.

This site remains active following the successful completion of the project so that materials and software from the project can still be accessed by any NHS sites who wish to continue or any sites who were not involved in the project but are interested in the methodology and tools.

Find Out More

Publications

Quality improvement at scale: evaluation of the drivers and barriers to adoption and sustainability of an intervention to reduce late referral in chronic kidney disease

Read More

Features & Benefits (v4) - Rev 23.02.2017

Over the most recent version from HEFT (v2.2), ASSIST-CKD version 4 carries a number of benefits and a couple of useful new features. These benefits can be broken down into four categories:

User Interface

Modern: The work to date provides an updated, web-like User Interface which brings ASSIST-CKD into the modern age of best design principles and ergonomic design.

Simpler: As a benefit of following modern design principles, in v4 it should be clearer at first glance what each screen is for, where the important areas are and what each input/button does. The inclusion of tooltips and popups allows in-app documentation of critical areas.

Clearer: Graphs are now clearer to view, with the ability to hover over data points to more accurately see the test results. The ability to zoom in and zoom out on a specific location allows the user to gain a more accurate understanding of the patients test history.

Usability

Deployment: v4 is built to deploy in more locations, in more environments, with less fuss. The setup of ASSIST-CKD is now self-explanatory and contained within the application itself - no more configuration files or Control Files, or post-install configuration required.

Database Options: v4 implements connections for three different types of database. Where v2.2 only supported Microsoft SQL Server (MSSQL), v4 will support SQLite and MySQL. This allows greater flexibility for the trusts in choosing the database the application will run on. it can be local, remote, and of a proprietary or open format depending on the individual site’s budget.

Error Reporting: In v2.2, it was common to find errors which went unexplained. V4 aims to report on errors accurately so that problems can be solved faster.

Architecture

Installation: v4 works with the same data set as v2.2. This means that v4 can be installed at a site already using v2.2 with no loss of data or time-consuming upgrade procedure.

Upgrades: Upgrades to both the database and the application are now automated making rolling out new versions/features far easier than in previous versions.

Performance

Large Data Sets: A common issue logged against v2.2 is the application's performance when dealing with large data sets and while outputting graphs. For v4, this problem has been taken into consideration at the ground-level of development. The algorithms have be re-written to best utilise the performance of the database engines.

User Efficiency: Alongside application performance, the user’s ability to work efficiently impacts the overall performance of the project. The aforementioned benefits in design and usability will also allow the user to perform better while using v4. With a more modern and less confusing interface, simpler deployment with less restrictions; and clarity when it comes to errors; the user can get to analysing and reporting on the data quicker and easier than before, and fix problems if they arise with confidence.

Accuracy: v4 improves upon v2.2’s finer details such as what happens when you specify an age range for reporting results. Where v2.2 would only use the year of birth to determine a patient’s age, v4 will compare the actual dates. Another such improvement is in the “check previous” functionality, which turned out to not work quite as expected. These improvements should serve to provide more accurate results while analysing; reducing the workload where possible without skipping over patients who need attention.

Software Releases

The ASSIST-CKD software, initially developed for internal use at HEFT, has been incrementally improved and released under a number of versions. The major version of the application (v2, v3, and v4) indicates the phase of development.

Current Release: 4.4.3

Date: 30th/31st January 2018

Type: Full Distribution

Major improvement to graph construction, display and behaviour; including large alterations to x-axis consistency between patients and fixes for graph zoom and pan functionality. Aimed at improving at-a-glance analysis as a quality of life update for all involved Trusts.

Release Types

Releases fall into three categories:

Type Meaning
Limited Distribution Working versions which were released to one or more NHS Trusts for real-world use, usually designated test sites or individuals experiencing specific problems with prior versions.
Test Build Development versions which were released to internal NHS testers for verification and feedback. Some Test Build releases were used in real-world reporting for short periods of time by certain Trusts, but for these releases it was not recommended.
Full Distribution Working versions which were released to all Trusts involved with the project. These releases can be considered stable.

Version 2

ASSIST-CKD Version 2 was developed internally at HEFT as a proof of concept and had a limited distribution to select Trusts. This phase set the baseline for future development and defined the project's software requirements going forward.

Version Date Type Notes
2.0.0.0 Before 2016 Limited Distribution Baseline release cementing the application core features and improving usability from prior versions.
2.2.0.0 Before 2016 Limited Distribution Feature improvements and bug fixes over the prior version. This release served as the baseline for the next phase of development.
2.2.0.0 (Cornwall) Before 2016 Limited Distribution A unique build created for Cornwall Trusts with some new usability features and bug fixes.

Version 3

The first phase of reconstruction in 2016-2017, designed as a fully backwards-compatible update of the software from a technological perspective, serving as a platform for larger-scale distribution. Designed to match version 2 feature-for-feature, building towards a national release.

Version Date Type Notes
3.0.0 September 2016 Test Build The initial test build for phase two of development. Largely matched feature-for-feature with v2.2.0.0, capable of upgrading v2 databases, serving as a refresh and update of the project as a whole. Supported Microsoft SQL Server, MySQL and SQLite database types for various use cases.
3.1.0 October 4th 2016 Test Build Improved v2 upgrade compatibility, ensuring smooth transition between the two versions.
3.3.1 October 7th 2016 Test Build Comprised of a multitude of bug fixes and usability improvements over the prior version.
3.5.0 October 25th 2016 Limited Distribution Added "Check Previous" functionality for patient filtering, bringing the new development closer to v2.2.0.0. Released to select Trusts experiencing issues with version 2.
3.5.2 October 26th 2016 Limited Distribution An incremental improvement over v3.5.0, fixing issues with the display of "marked" patients. Released as a patch to the trusts who received v3.5.0.
3.6.1 November 9th 2016 Test Build Significant changes to the "Check Previous" functionality, bringing the feature more in line with expectations.
3.7.0 December 15th 2016 Limited Distribution Changed "Check Previous" to "Exclude Previous" to improve the logic going forward.
Improved database query performance.
Altered graph construction behavior.
Added direct-to-email reporting functionality including a built-in address book .
Added Custom Location Type Mapping functionality.
Added some fields to the report templates.
Made a number of changes to data upload validation to allow more flexibility.
3.8.0 January 10th 2017 Limited Distribution Minor fixes for all features added in the prior version.

Version 4

Continued development of version 3, starting with a significant change which made version 3 and version 4 incompatible with one another. Version 4 was improved throughout 2017 and continued into 2018, and provided a number of large-scale distribution versions for national project operation.

Version Date Type Notes
4.0.0 January 2017 Full Distribution The first national release of the rebuilt software, incorporating more v2 upgrade compatibility changes.
4.2.0 February 2017 Full Distribution A number of data validation changes to increase flexibility in uploads. Including the conversion rules for eGFR values of ">90" and ">60".
4.3.0 February 2017 Limited Distribution Graph construction changes and bug fixes according to feedback from various Trusts. Distributed only to Trusts reporting problems.
4.3.5 March 2017 Full Distribution Incorporating the v4.3.0 fixes as well as some more incremental improvements for a new national release.
4.3.6 May 2017 Limited Distribution Fixed an issue with NHS Number upload validation. Distributed only to Trusts who reported the error.
4.3.7 May 2017 Limited Distribution Fixed an issue with SQL Server rejecting upload batches of 1000 rows, causing fatal application errors. Distributed only to Trusts who reported the error, as the issue was dependent on their SQL Server setup.
4.3.8 May 2017 Full Distribution Altered the logic for filtering patients by whether or not they had marked tests, preventing cases where marked patients and unmarked patients (in their lifetimes) would be returned for the same query. Distributed to all Trusts due to the nature of the fix and to spread the prior fixes to all users.
4.3.9 May 19th 2017 Full Distribution Further compatibility fixes for version 2's definition of a "marked" patient, application UI changes, graph colour changes. Altered graph zoom and pan behavior, and added some usability features such as a "Patient X of Y" count on the graph screen and timed warnings for long data loads.
4.4.0 September 22nd 2017 Full Distribution Fix for newlines within upload files causing application freezes.
Fix for error messages not printing properly for upload errors.
Database upgrades to add performance-boosting indexes on the eGFR table.
Performance-boosting query tweaks for the Analyse and graph screens.
Made changes to the "Location Upload" (legacy version 2) feature to bring it closer in line with version 4's "Location Code Mapping", reducing confusion between the two features.
4.4.1 October 17th 2017 Full Distribution Removed "Location Upload" entirely, re-worked the Location Code Mapping feature and associated database queries, and added a bulk Location Code Mapping uploader in the old feature's place.
Improved Analyse to Graph screen transition by pre-loading patient data on the Analyse screen.
Added an application setting "Disable Graph Zoom" to allow disabling of the graph zoom entirely. The feature wasn't working as expected and in some cases causing confusion.
Fixed an issue with graph position on load, bringing consistency to the interface when moving between patients.
Added ODS codes for all Trusts in Wales, and allowed switching of ODS code on the Statistics screen.
4.4.2 December 5th 2017 Full Distribution Fix for a critical issue preventing the use of SQLite databases.
Fix for an issue preventing ODS Code selection on the Statistics screen.
Changed "Disable Graph Zoom" setting to "Enable Graph Zoom" with a default value of "off", thus disabling graph zoom functionality for all users by default. Graph zoom was not working as expected and confusing some users.
Changed graph axis construction logic to try to improve consistency.
Hid "Other Tests Upload" and "Diabetic/Renal Upload" on the Upload screen so as to discourage use of these legacy version 2 features.
Improved upload validation in general, allowing for better error messages and less interruptions/crashes for invalid data.
Updated the application UI to better reflect project colours and branding, and improve usability.
4.4.3 30th/31st January 2018 Full Distribution Major improvement to graph construction, display and behaviour; including large alterations to x-axis consistency between patients and fixes for graph zoom and pan functionality. Aimed at improving at-a-glance analysis as a quality of life update for all involved Trusts.

Software & Lab Pack Download Request

Please click the button below to email us a request for a software and lab pack

Request Pack