Module 5 - Clear-text linkage project simulation

In this Module, we will conduct a more complex CLEAR TEXT linkage project involving four synthetic, clear-text datasets.

The project will simulate a linkage of Primary (GP) to Secondary health care data. We will imagine that the linkage project will be used to support a specific Research Project – for example, one that investigates Chronic Heart Failure (CHF):

• Secondary health care data will comprise three synthetic datasets of Death registrations, Hospital admissions records (HOSP) and Emergency department presentations (EMER).
• Primary care data is also simulated and is supplied as clear text General Practice (GP) data.

Please note that encoded versions of these same datasets are also available as part of this Training Package, so it is possible to undertake a PPRL version of this same linkage project but using encoded PII data instead of clear text data. For instructions on how to undertake a PPRL version of this simulated Primary-to-Secondary care linkage, please refer to Module 8.

info

This module is part of a series of training modules that can be used with any LinXmart deployment. However, the user will require permissions to create Linkage Projects, load data for linkage extract linkage map results.

All required data files are available from the main self guideded training page.

Getting started

Step 1: Connect to the LinXmart web interface by typing in the LinXmart web address. You will see a screen similar to this:

Step 2: Ensure that your installation of LinXmart has a Data Provider called TRAINDP. To do this, click on the CATALOG tab at the top of the page.

If this Data Provider is not displayed, please follow the setup instructions for this series of training modules before continuing.

With LinXmart now aware of the Data Provider called TRAINDP and its related Event Types, you are ready to create a more complex linkage project that links the clear text versions of these datasets.

Creating a Linkage Project

Step 3: Click on the PROJECTS tab at the top of the web page. This will present you with the dashboard of current projects (shown below). At the bottom of the list, click on Add Project.

Step 4: Create your project (e.g. XX_CHF_CLEAR) as shown below and click Save to finish. Do NOT tick the Privacy Preserving box, as this project is NOT a PPRL project.

Adding data to a Linkage Project

Step 5: Click on your XX_CHF_CLEAR project to bring up the Project Details page.

You are presented with an empty project (as shown below).

Before data can be loaded into this Project, one or more Event Types need to be attached to the Project and a suitable Match configuration needs to be defined.

Step 6: Click on the Attach some Event Types link presented to you below the recorded statistics for the empty project (shown below).

Click on the Event Types field. You will be presented with a list of Event Types that are already known to (and defined in) LinXmart, including those supplied by our Data Provider, TRAINDP.

Check the tickbox next to Data Provider TRAINDP and Event Type that corresponds to the clear text dataset TDP_DEATH (TDP_DEATH).

Also check the tickboxes for all the other clear text datasets – TDP_HOSP, TDP_EMER and TDP_GP.

Click OK to confirm your selection. Then click Attach.

This will cause these types of datasets – and ONLY these types of datasets – to be associated with your Linkage Project. (LinXmart will only permit these types of datasets to be uploaded and linked in your project. This ensures separation and security of individual linkage projects in LinXmart.)

Using the default Match strategy

LinXmart projects are automatically created with two default matching strategies – which can be used as the basis for many linkages.

These strategies are suitable for clear text projects because they operate on 'standard' linkage fields and make use of standard comparators such as the Jaro-Winkler comparator.

Step 7: To view these strategies, navigate to the Project dashboard and return to your project (PROJECTS -> select XX_CHF_CLEAR). Then click on Settings (as below), then click on Match configuration.

You will notice that the default Match configuration comprises two strategies – one which blocks on a combination of Date of Birth and Sex (shown below), the second which blocks on a combination of Soundex of Family Name and First Initial of Given Name. Note that in both strategies, the fields being compared are the default, clear-text linkage fields that LinXmart is already aware of.

The weights being used here are also the defaults which are assigned by LinXmart. The threshold value (21) is also a LinXmart default. For the purposes of this Module, we will use these defaults; however, you can modify any or all settings as your project progresses.

Check Linkage Field-Source Data Item alignment

Before proceeding to upload data, it is a good idea to confirm that the incoming data items align with what is expected by the Linkage Project (as specified in Import formats). These checks ensure that the 'right variables' have been uploaded and mapped to the appropriate default linkage fields in LinXmart.

(Some cleaning and standardisation of incoming data items can also be defined here – but we'll show you that feature at another time.)

Step 8: To check alignment of linkage variables, navigate to the Project dashboard and select your project (PROJECTS -> click on XX_CHF_CLEAR). Then click on Settings (as below), then click on Data Cleaning and Encoding.

Under the section labelled Linkage Fields, you will see the following:

Review the list of Linkage Fields. Questions to ask yourself:

• Are these fields the complete set of all data items supplied by the Data Provider(s)?
• Have I accidently left anything off?
• Are incoming data items correctly aligned with each other and to Linkage Fields?

If you have answered Yes to these questions, and you are satisfied with the default Matching strategies (and threshold values), then you are ready to proceed to uploading and linking your first dataset.

Linkage files of Clear text data

We will start the linkage by uploading the Death records.

(Question: Why might we load Deaths first?)

The relevant input file can be found as TDP_DEATH_clear.csv - located in your Training folder: TRAINDP Clear Text Datasets.

You may wish to view this file in a text editor to see its contents. It is a comma delimited file containing a unique record identifier, a unique person ID and various PII fields. For convenience, the format of this file (called the Import Format) has already been defined within LinXmart (refer Vital_Stats_IF).

Step 9: If you are not already on your Project dashboard, navigate to it. Then click on Settings. Then Data Sources. This should take you to your Event Type Data Sources page (see below).

Click the Load icon on the TDP_DEATH row on this page.

Step 10: Click on the Choose File icon and select the TDP_DEATH_clear.csv file from the relevant Data folder. Click Load. This may take a minute while the dataset is uploaded to the LinXmart server from your local machine.

You will be taken to a Load Data Confirmation page.

Please ignore the Group by JLK option (default is NO) and select Load. This will begin the linkage process; its process can be monitored in real-time through the ACTIVITY tab.

Monitoring your load-linkage job

Step 11: To see how this load-come-linkage job is progressing, navigate to the Job Monitoring screen by clicking on the ACTIVITY tab at the top of the screen.

Step 12: Select the ENVELOPES sub-tab. Information regarding the current job's status will be appear under 'All Envelopes'.

The data in this file was internally linked or de-duplicated, identifying records within the file belonging to the same person. As there were no other datasets in the Project, this is the only linkage that took place. A green tick in the Status column indicates that the job ran to completion (green is good).

Clear Text Linkage continued: Adding Hospital records

Next, we upload the clear text hospital admissions data: TDP_HOSP_clear.csv.

This file is found in the relevant Training Data folder; there are about 25,000 records in this file. The format of this file (called the Import Format) is identical to the format of the TDP_DEATH data and uses the same LinXmart Import Format.

Step 13: To upload the data, first navigate to the Project dashboard (PROJECTS -> select XX_CHF_CLEAR) and then click on Data Sources. Select the TDP_HOSP upload icon.

Step 14: Follow similar steps as above to find the hospital data file in your folder system, then confirm the load.

Once uploaded, you can monitor progress of the Linkage job through the ACTIVITY screen.

Step 15: When the job has completed, go back to the Project Dashboard and review the results. Do this by clicking on the Report icon of the recently complete Linkage Job (shown below).

Clear Text Linkage continued: Adding Emergency records

Step 16: Follow similar steps (as above) to find and upload clear text hospital data (replace TDP_HOSP with TDP_EMER, where applicable, and use TDP_EMER_clear.csv).

Once uploaded, progress can be monitored through the ACTIVITY tab -> ENVELOPE screen.

Step 17: When the Linkage job has completed, go back to the Project Dashboard and review the results from the linkage. Again, do this by clicking on the Report Icon associated with the Job that has just completed (as shown previously).

Clear Text Linkage continued: Adding GP records

Finally, we will add the clear text General Practice data into the project. The data file: TDP_GP_clear.csv is found in the relevant Training Data folder (there are about 50,000 records in this file). The format of this file is identical to the format of the TDP_HOSP data and uses the same LinXmart Import Format: Vital_Stats_IF.

Step 18: To upload the data, first navigate to the Project dashboard (PROJECTS -> select XX_CHF_CLEAR) and then click on Data Sources. Select the TP_GP_clear upload icon.

Step 19: Follow similar steps as above to find and upload the clear text general practice data (replace TDP_HOSP with TDP_GP where applicable; TDP_GP_clear.csv).

Step 20: Once again, monitor progress of the Linkage job through the ACTIVITY -> ENVELOPES screen.

Step 21: When the linkage job has completed, go back to the Project Dashboard and review the results from the linkage.

Should I vary my threshold or change weights?

Decisions about varying the threshold are typically made after reviewing Linkage Reports and by eye-balling matches that sit on or near the cutoff value (the current threshold value is 21). These review points can occur at the end of EVERY linkage job, or they can take place at certain progress points in the project, or both.

Remember that linkage is an iterative process – it is unlikely that you will achieve optimal linkage results on your first run!

Here are some ideas for refining your linkage and improving results:

• Look at the distribution of record-pairs in the linkage reports. Do these 'look' right? Are there some particular confidence levels with a high number of pairs? Might you be tapping into the distribution of True Non-Matches?
• Look at the structure of low scoring groups (view through Quality Review) and reflect on the weight of each pair in that group.
• Look at some groups with low saturation (view through Quality Review). How are the records coming together here? Could setting a higher threshold, for example, eliminate wide diameter or bridge-type group structures?
• Look at the linkage fields, comparators and weights in the match strategies (match config) – consider if an alternative comparator should be used; do any comparators and/or probabilities need to be adjusted?

Welcome to the art and science of data linkage!

What about my Grouping Strategy?

So far, we haven't talked much about the Grouping strategy. Under the default configuration, all datasets are grouped after the matching phase using a Merge method of grouping. However, in this Module and after having linked together your secondary care datasets (Deaths, Hospital and Emergency), you may want to revisit the Grouping configuration setting and consider using a Weighted Best-Link method.

Weighted best-link grouping is particularly useful when linking against a pre-existing repository of linked records for which you have confidence in their linkage quality (i.e. a Spine). Weighted best-link grouping essentially allows you to leverage this pre-existing quality and it will NOT permit an incoming record to join two pre-existing groups in the Spine.

Think about why you would want this scenario, and what the benefits might be to switching to a Weighted best-link grouping method.

Changing the grouping method is easy (shown below); however, once you have made the change you may need to rollback and re-run the previous linkage of the GP data to the secondary care datasets, as the earlier linkage was run using a Merge grouping method.

Step 22: To change the Grouping method, first navigate to the Project dashboard (PROJECTS -> select XX_CHF_CLEAR), click on Settings, then Group Configuration.

Step 23: Edit the configuration, select the Weighted Best-Link method and Save your change.

Step 24: Now go back to the ACTIVITY screen and rollback the previous Linkage job involving GP data (rollbacks are explained in Module 3).

Step 25: When the rollback is complete, re-do the linkage of GP data (i.e. follow the steps in the Adding GP records section above).

Generating Project-level Reports

You are already familiar with reports that are generated as part of a Linkage Job. These Job-level reports describe results from that linkage job ONLY.

However, there may be times when a whole-of-project view (report) is required - this is especially true toward the end of a project or when a linkage project reaches certain milestones.

In this section, we will describe how to generate Project-level reports.

Step 26: Go to the Project dashboard (PROJECTS -> select XX_CHF_CLEAR), and click on Reports (on right, just below Quality Review).

Step 27: On this Reporting screen, select the type of Report that you want e.g. Project Linkage Detail. Add some text in the comments section to describe what report you are generating and why you want it. Then click Generate.

This will fire off a Reporting Job. The progress of this Reporting Job can be monitored through the ACTIVITY tab -> JOB.

(Please note that the Project Linkage Report may take some time to run, as it is generating a lot of project-level statistics that will be included in the final report!)

Step 28: Another useful Project-level Report is the Data Summary Report. This can be generated in a similar way – by selecting Data Summary as the Report Type, then adding descriptive text in the comments and then clicking on Generate (shown below).

When these Project-level Reporting Jobs have completed, they will appear in the bottom portion of the Report screen (shown below). You can inspect the detail of each Report by clicking on the Report icon on the right of each Report (also shown below).

Please note that Project-level Reporting Jobs can be run at any time during a Linkage Project. If you have a Linkage Job running at the time that you request a Project-level Report, then the Reporting Job will be queued until the active Linkage Job is completed.

The progress of any Reporting Job can be monitored through the ACTIVITY tab -> JOB.