GUMCADv3 marks a considerable upgrade to the previous GUMCADv2.  In this note we explain our approach to adopting it for use for Sexual Health Tariff charging and highlight the main changes it will entail.

GUMCADv3 reference material is published here.

In this Note we have made use of the Data Specification and Technical Guidance (GW-505, dated Jul 19)

Background

Why does this note matter to you?  If you use GUMCAD to report your activity for charging purposes, the introduction of GUMCADv3 will mean there are going to be changes in the manner tests and diagnosis are recorded and the way data is handled.  One of the major changes will be the use of SNOMED CT codes, however, there are more subtle structural changes as well.

Historically, the officially accepted version of the GUMCAD report required a record for each activity code.  This meant that much of the data was being duplicated (incurring data redundancy) as common fields were repeated for each activity and where there is redundant data there is room for data conflict or anomaly which need to be handled or ignored.  Furthermore, and more importantly, in order to merge the GUMCADv2 report with the SRHAD report we 'flattened' the GUMCADv2 report into a single row per episode allowing 12 activity fields in an episode using a Transpose process.

This had the advantage that a single episode could be read in a single row of data across both reports and common demographic data could be used between GUMCAD and SRHAD data reports.

Historically, GUMCAD used code 'suffixes' to annotate variations to a code for example, 'X' to indicate where a diagnosis had been made at a different clinic to that which was making the report; it was not always clear if the clinic making the diagnosis actually delivered any intervention since the diagnosis code was associated with the test code.  

GUMCADv3 adopts a different approach, each row of data represents an activity, be it test, diagnosis or intervention.  Associated with that activity can be a number of attributes for example, in the case of diagnosis, a diagnosis site, if it was treated, if the diagnosis is confirmed and if the patient presented as symptomatic.  So, technically, it will be possible to record a large number of activity records for an individual attendance by simply adding another row of data to the report - there is no longer a 12 field limit to the SHHAPT (now SNOMED CT) fields. 

In line with this, SNOMED CT codes will represent each individual test delivered rather than common groups of tests.  In the previous versions of GUMCAD it would be unlikely for an episode record to rise above our predetermined limit of 12 activities, in GUMCADv3 this like to be a regular occurence and so we need to adopt an approach that allows for many more activity codes in an episode.

Our Approach

Therefore, in our interpretation of the GUMCADv3 report we will be adopting the following concepts

• An Episode (or visit) - A collection of records sharing a common Organisation ID, Clinic ID, Patient ID and Attendance Date values.
• An Activity Collection (or a GUMCADv3 record) - A collection of fields relating to each of the Episode Activities within an episode.  We will attribute each of these collections with a unique identifier (using a UUID) based on the GUMCADv3 record submitted.
• Non-Activity Collection - All the fields in an Episode will be grouped together in such a way that the latest recorded record without a null value will take precedence.  These fields will also be available for use in triggering sexual health tariffs currencies.
• We will store data within an Activity collection in the form of Episode ID : Record UUID : Attribute : Value.  For example:
  
  

  Episode ID	Record UUID	Attribute	Value
  1034333|RP5|RP5RE|115686|2020-11-05	8ee78e3b-cf97-451d-8c21-8b9bca9f4c7e	Age	34

SRHAD and GUMCADv2 files uploaded will be treated in a similar manner

Impact

What does this mean?

If you are a commissioner, there will be no change, until it comes time to redesign your Tariff Configuration using GUMCADv3, this can only happen once all your hosted providers have adopted GUMCADv3.

If you are a provider, data management of GUMCADv3 data files becomes a SQL and csv reporting activity, the data transfers less reliably into Excel files.  

All the data extracts you export will remain in the same format as before (we now recommend you use csv format for GUMCADv3 reporting data), however, the Record identifier will not be exported.  The online service will expose data in the Episode ID : Record UUID : Attribute : Value format, this will make it much easier to filter on a particular Episode, Record, Attribute or Value.

This interpretation means that the de-duplication of data will no longer include activity fields and it will be possible to submit multiple instances of the same activity, possibly even with different supporting fields (ie changes in diagnosis confirmation).  This will not have an impact in how the Tariff charges are calculated since each currency can only be charged once in an episode or visit so duplications will only trigger a single instance of a currency.

This implementation approach means the design is fully extensible and we can easily add new data upload formats when required. 

Implementation

Naturally, not all organisations will implement GUMCADv3 at the same time, (nor might they all be successful in their initial implementation!).  So, we have adopted a phased approach to implementation summarised in the table below:

		Data Files Accepted
Phase	Tariff Configuration	GUMCADv2	GUMCADv3	SRHAD
1	GUMCADv2 and SRHAD	X		X
2	GUMCADv2 mapped to GUMCADv3 and SRHAD	X	X	X
3	GUMCADv3 and SRHAD		X	X

 

Phase 1 as is, Tariff Configurations remain descriebd in GUMCADv2 and SRHAD, data is uploaded in GUMCADv2 and SRHAD. 

Phase 2 Tariff Configurations remain described in GUMCADv2 but the codes are mapped one-to-one to GUMCADv3 codes using Appendix 7 (p117) of the guidance document and SRHAD.  In addition, we will retain any deprecated codes critical to maintaining the Tariff Configuration in place, providers will need to work with us to alert us to any specific codes they think they will require to be retained in GUMCADv3.  Data can be uploaded in any combination of GUMCADv3, GUMCADv2 and SRHAD files.

During Phase 2 pay special attention to charges being raised, in particular how tests are accounted for.  We assume 

• 442487003 | Test: chlamydia test equates to the old T1 code
• 1086351000000100 | Test: gonorrhoea test ​equates to the old T2 code
• 1086341000000103 | Test: syphilis test ​equates to the old T3 code
• 171121004 | Test: HIV test ​equates to the old P1A code

For example, these assumptions can lead to T3 and T4 being charged in the same episode or visit unless this combination is expressly excluded in the Invalid Currency Combination ruleset.

Phase 3, will require commissioners to publish new tariff configurations based on GUMCADv3 and SRHAD, at this point GUMCADv2 data will no longer be accepted.  Adopting Phase 3 pre-supposes all providers in a Tariff Configuration have successfully adopted GUMCADv3.

Constraints:

All providers and commissioners are presumed to start at Phase 1.  It is possible for any provider to adopt Phase 2 independently as soon as they feel confident they can report GUMCADv3 data.  Phase 3 is adopted by a Tariff Configuration and will require all Participating Providers of that Tariff Configuration to have successfully adopted GUMCADv3 reporting, preferably, by testing it for Tariff using the Phase 2 approach.  Phase 3 will require commissioners to re-write their Tariff Configuration Triggers to confirm with GUMCADv3 datasets, we will support them in doing this by in the first instance providing a Tariff Configuration with GUMCADv2 codes replaced one-to-one mapping of GUMCADv3 codes.

Phase 2 and 3 will only be available from the new Grouper interface, in order to adopt this please contact Pathway Analytics

Practical Guidance

When handling GUMCADv3 data files ensure:

• data is in csv format, utf-8
• dates to be formatted yyyy-mm-dd
• at the moment, we cannot handle cells demarcated with quotes ie a cell containing a comma might be demarcated with quotes for example a date cell followed by a decimal converted to text might look like: 2021-11-30, "123,456.45",... the only way we could handle this is if the decimal value was a true decimal ie 2021-11-30, 123456.45,...
• the first row is the heading row, last row of the csv file is the last row of the data. - no empty rows added, columns are contiguous, each with a unique heading.
• you should be cautious not to open the csv file in MS Excel without taking precautions not to convert long integer values to scientific notation (use import csv function).

Review your GUMCADv3 data to ensure it includes the necessary local codes required according to your current tariff configuration.

Since each test is coded individually in GUMCADv3, we may need adjust your tariff configuration to ensure the correct currencies are triggered.  Check your first charge report to ensure there are no instances of an episode receiving a combination of T2, T3, T4, T7 or P1A tests.

Carefully review the number of episodes and number of records reported during the upload process.  The number of 'records' counts the rows in the data file uploaded, the number of 'episodes' collates the number of records with a common PatientID, ClinicId and AttendanceDate (ConsultationDate).  If these look wrong, check the accepted data to see what has been read from the file, using the wrong date format will cause episodes not to be collated correctly, malformed files with data that includes commas or blank rows will cause the number of records to be incorrect.

Upload Guide

Please contact enquiries@pathwayanalyitcs.com to setup your account to use GUMCADv3 and book a time to walk through the first upload you make with us.

You will be presented with an upload page like this:

Browse or drag the file to be uploaded to the box that says Drop file here...

Depending on the codes your data files contains select the appropriate button Upload SRHAD and GUMCADv2 or Upload GUMCADv3.

Depending on which button you selected, you will be presented with a list of fields either from the SRHAD and GUMCADv2 dataset or the GUMCADv3 dataset, next to each field is a dropdown list of the column headings from the file you uploaded.  If the column headings you have used are recognised they will be automatically matched to the correct field.  Otherwise you will need to match them.  Green indicates the field has been matched, amber indicates the field is not matched and red indicates the field is not matched and is mandatory - so it must be matched.

Once you are happy the fields are correctly matched.  Click the Validate button.

You can keep an eye on progress in the Grouper Activity box that will appear.  Each circle indicates a file that is being processed.  The files being processed and related to the page you are viewing will be coloured red, others files belonging to other submissions or other organisations will show as blue circles.

A list of uploaded files is available showing you the state of each file and the clinics, records and episodes associated with it.

Note, you can upload any combination of GUMCADv2, GUMCADv3 and SRHAD files in the same submission, however, only GUMCADv2 and SRHAD data can be held in the same file GUMCADv3 must always be in a separate file.

After your data has been checked you might receive error warnings indicating either there are data validation errors or you need to update the clinic addresses for the clinic codes you have used.  You need to clear these errors before you can make a submission.

The Error Correction page by default lists each type of error indicating the number of records affected by it.  Most error types will have a prompt dropdown list or option selection to help you choose a valid correction which will be applied to each record affected.  Click Save All to apply the change for each error.  Alternatively, select the Show All button to look at each record's errors. 

Clinic addresses can be edited as follows, if the clinic code used is a recognised ODS code, try clicking the Clinic Code button to retrieve the registered address, click locate address and save.  Or type in the address manually and click locate address and then save.

Once you have successfully validate your data you will be able to make the submission (or upload and validate another file).

The Sandbox option is only available to Provider+ accounts and it makes a tariff calculation which you can review before it is released to commissioners.   Once you have made a submit with the red button, you can longer make any changes to that month's data.

Once you have files uploaded for a particular month in the new user interface, you will not be able to upload or submit data for that month in the old interface, the message below will be shown: