Q: What are the technical requirements for the data uploading solution?
The solution will use a lightweight data scheduling application (which Beautiful Information will provide) that is installed locally and run as a Windows service. Typically, this will be on hosted in a Windows Virtual Server environment on existing hardware, with access to the internet HTTPS. The application requires a small SQL Server database (SQL Server 2008 onwards, either standalone or a database on a pre-existing solution) and access to SQL tables that contain the required data. The storage and processing requirements are very low and we anticipate that all providers should be able to implement the solution within their existing technical infrastructure with no additional costs incurred.
Historical data is not held in the database, only the current data to be queued and the configuration details, so the data should be less than 20MB. A standard system backup will be sufficient (although keeping a record of the stored procedures with your data extraction queries would be advisable). Logfiles will be regularly cleaned by the application (every 30 days) and the database should not grow as it is not retaining historical data.
Q: How will the required data be extracted?
Providers will need to have the required data available in an SQL database, available for the extraction queries, and the scheduler software will then require access to this database run the necessary queries. There are a number of ways to achieve this, with various levels of complexity. Some providers might be able to query their system directly, while others may use a live mirror or set up a regular extract. Some providers might choose to create a separate SQL view of the required data, and point the scheduler application to that to run the necessary queries. We have worked with various providers using a range of systems, installing and using the data and will provide the necessary assistance to extract the data required.
Q: How will the application deal with any network security?
There is one-way communication to the NHS I Azure Service Bus server, which is not N3 hosted (although it might be in future). NHS I are using Shared Access Signature authentication on this server, and the configuration file downloaded by each Trust will provide a unique endpoint token.
Different organisations have different network security arrangements in place to handle outbound connections (such as authenticated proxies), which may require some adjustments to the configuration and/or local network rules. A configuration tool allows you to add proxy details, which you should be able to obtain from your IT department. Firewall rules may be required to allow outbound HTTPS traffic over Port 443 to the NHSI server at sb://nhsiingestuat.servicebus.windows.net/
Q: How much work is required to implement this the solution?
This will differ from provider to provider, but typically the solution can be implemented within a week. Installing the software and setting up the database is quick, but some troubleshooting might be required to ensure that the service can be automated and an outbound connection can be made to the NHS I server. Providers will then need to create the extract queries in SQL to pull the required metrics together ready for uploading to NHSI. In some cases, this may already be being done but assistance can be provided as necessary.
Q: Do I have to use the BI Scheduler?
No, you can implement a different solution instead; Beautiful Information will also provide support as necessary for this. However, nearly all trusts have found that it is quicker and more efficient to implement the BI scheduler rather than develop their own solution.
Q: Do I need Information Governance approval for sending this data in an automated way?
No, this is aggregate data that you already send manually each day. There is no patient level data and the data is sent securely solely between your trust and NHS I.
Q: Does this data leave the European Economic Area (EEA)?
No, the data stays within the EEA.
Q: Can I check and validate my trust’s data once it has been extracted and submitted to NHSI?
No, this is not currently possible.
Q: What happens if the automated extract fails? Is there a way that we can manually enter the data and then have the scheduler send it?
There is no manual process for this data. In this situation, you should manually populate the database table at your end (for instance via a .csv file) and send this via the scheduler.
Q: How will I know if there’s a problem with the data or if it fails to send?
You will receive a notification from NHSI.
Q: Who will be able to view this data once I send it?
The aim of NHS I is to publish the data at aggregate level through their site. The only change is in how the data is collected.
Q: What technical support is available for this project?
Support for the application during deployment period will provided by Beautiful Information and NHS I. Questions relating to set-up and configuration of the local query agent are being addressed by Beautiful information, NHS I will be supporting the validation service. We would expect that local IT will provide support to ensure the service is operating and Trusts are responsible for the data which is submitted. Updates to the service may be required from time-to-time, but we do not envisage this being a regular occurrence. Beautiful Information (BI) can be contacted at david.mullett@beautifulinformation.org
Q: What if some of these measures are simple to send and others are much more difficult? Do we need to send all data items?
We are keen to understand as part of this project which metrics are most easy to collect and which may be more difficult. We ask that you provide what you can automate and any that are more difficult or only recorded occasionally, such as serious operational issues, are entered manually via the NHS I portal.
Q: Is there a delay in automated data from the scheduler entering the NHSI server or should it be instant?
Lag time should be no more than two minutes.
Q: Where can I find supporting documentation?
Login to the implementation guide: https://nhsi.beautifulinformation.org/
Q: How do I access the implementation guide?
Contact your BI Account Manager (if known), otherwise contact david.mullett@beautifulinformation.org
Q: Who do I contact if I have a metric query?
Metric queries should be forwarded to the NHS I Support Team by email at: nhsi.sitrepsupportteam@nhs.net
Q: Who do I contact if I have a metric definition query?
Metric definition queries should be emailed through to: england.dailysitrep@nhs.net
Q: Where can I find metric definitions and ids?
Login to the implementation guide: https://nhsi.beautifulinformation.org/
Q: How should we deal with empty [null] Metric Value in our returns?
The preferred option is to not send the metric at all, but sending a value of NULL would also be fine.
Q: How do I move my ambulance submissions from the UAT to NHS Improvement’s live server?
Follow the these instructions:
Firstly, make sure that you turn off the service, if you a running one (BI Task Service), that runs the scheduler application and then close the application (BISV_ScheduleServiceBus.exe) in the task manager.
Then update the appconfig3 SQL table to change URL1 to:
sb://nhsiingest.servicebus.windows.net/
Then open the BISV_ScheduleServiceBus.xml file in Notepad and replace the token key with your unique live token which has already been emailed to you.
And then save.
You can then restart the service (or run the scheduled task) to launch the scheduler, and the next time you run the stored procedure it should send to the new destination using the new token.
Please keep an eye on the logs once you have made the changes to check all is looking OK, but contact us if you have any further questions.
If you have further questions, please email david.mullett@beautifulinformation.org