Discovery in Action Case Study CoCreation of Discovery

Discovery in Action: Case Study: -Co-Creation of Discovery Dashboards with Discovery Users Examples Showcased: - • Clinical Profiles: Endoscopies • BADs Dashboard Developments • Hospital Spell Analysis

Introduction: - Dashboard Co Creation • Discovery is an information system that provides approved users with access to a range of comparative healthcare information to support performance and quality improvement in Health Boards across Scotland. It is an ongoing collaboration between NHS Boards, the Scottish Government, and NHS National Services Scotland. • The Discovery Team aim to build dashboards that provide useful, informative visualisations that allow users to get the most out of the information that is held nationally. • A co-creation approach to developing these dashboards is followed where development is in collaboration with the end users to ensure that the final dashboards are as useful as possible. • This case study will showcase a few examples of this co-creation approach

Discovery Dashboards Co-Creation Examples Showcased: Dashboard Development: Clinical Profiles: Endoscopy New Visualisation: Indicator: BADs Dashboard Development: Hospital Spell Analysis Dashboard

Example Dashboard Development CLINICAL PROFILES: - ENDOSCOPY Situation • The JAG (Joint Advisory Group on Gl endoscopy) provides accreditation of high quality gastrointestinal endoscopy services. • Part of this accreditation process requires multiple audits on an ongoing basis of patients who have undergone these procedure, including: • Audit of readmissions within 8 days • Audit of deaths within 30 days of an endoscopic procedure • High Level Analysis + individual patient audits to identify if outcomes were Definite; Probable; Possible; Unlikely Background • There was no national reporting process for this which was leading to multiple, time consuming , local audits of varying quality. • Desire to develop a robust national reporting solution to allow endoscopy centres to move towards accreditation.

Example Dashboard Development CLINICAL PROFILES: - ENDOSCOPY Actions • The Discovery Wrap-around Service were approached by a lead clinician from an NHS Health Board to develop proof of concept analysis based on nationally available data. This was then validated by comparing to patient level data in local health board endoscopy systems. • Any differences were investigated and explained to the satisfaction of clinicians involved. • Development of a Endoscopy specific suite of dashboards was proposed to the Discovery Oversight Board and formally approved. Results • An Endoscopy specific suite of dashboards were developed in conjunction with clinicians in responded to their needs. • Benchmarking dashboards developed at Level 1, including multiple visualisations and comparisons across agreed outcome measures. • Patient listing dashboards available for users with approved Level 3 access. Allowing patient specific audits.

Results: - Level 1 Benchmarking Dashboards 1. 2. Benchmarking dashboards developed at Level 1, including multiple visualisations and comparisons across agreed outcome measures. Links to Patient Listings in level 3 dashboards.

Results: - Level 3 Patient Listing Dashboards A Level 3 Dashboard helps identify specific patient information, which is necessary for detailed audits that are required for accreditation. This information is easily downloadable. Outcomes • Robust dashboard solution developed that has clinical buy in, saves clinical time, highlights variation, improves clinical processes and the likelihood of JAG accreditation.

Example Dashboard Development BADS: - EXTRA VISUALISATIONS Situation • The British Association of Day Case Surgeons have identified a target % for a number of elective procedures that they would expect to be completed as a day case. • The Discovery team included a BADs indicator at level 1 that showed the HB and hospitals overall BADs rate compared to other Health Boards/hospitals. A level 2 dashboard provided functionality for the user to drill down to specific BADs groups and BADs procedures and compare their performance against the BADs target and an amalgamated BADs peer rate. Background • Users provided positive feedback on these dashboards when they were released. However they suggested a number of developments that would make the visualisations even more useful. Actions • The Discovery team took these suggestions on board and developed in conjunction with the users new visualisations for this indicator. • A BADs specialty split that is now available within the Level 1 overview dashboards. This allows BADs activity and rates to be compared across all HB and Hospitals in Scotland for each specialty. • In the Level 2 dashboard users can now easily compare BADs rates for hospitals within a health board and against their hospital peer rate within one view.

Results/Outcomes: - New Visualisations Released at Level 1 and Level 2 1. A BADs specialty split that is now available within the Level 1 overview dashboards. This allows BADs activity and rates to be compared across all HB and Hospitals in Scotland for each specialty. 2. In the Level 2 dashboard users can now easily compare BADs rates for hospitals within a health board and against their hospital peer rate within one view.

Example Dashboard Development HOSPITAL SPELL ANALYSIS DASHBOARD Situation • The Discovery Wrap-Around Service was approached to provide analytical support for a Health Board’s Respiratory Managed Clinical Network looking for potentially productive opportunities. During this work it became apparent that the existing length of stay information in Discovery could only take you so far and that bespoke analysis was needed to respond to some of the questions that the MCN were interested in. • Based on the work with the MCN the Discovery Team agreed to develop a proof of concept dashboard that would provide more functionality that would respond better to a hospital clinical/service improvements team requirements. Background • Originally average LOS information in Discovery was available at a episodic or specialty spell level. However a patient could have multiple specialty spells and episodes within a hospital spell even though it still the same patient journey and relate to the same condition/procedure. • Zero day stays were excluded from the original LOS indicators and small proportions of patients with long LOS could skew the average shown. Activity trend information was also not easily available.

Example Dashboard Development HOSPITAL SPELL ANALYSIS DASHBOARD Actions • Based on the original Managed Clinical Network feedback and bespoke reports a proof of concept dashboard was developed to better meet a hospital clinical/service improvements teams requirements. • This development was part of a Sprint process where dashboard versions were presented to a team of users to provide feedback and advice on. The Discovery team took these suggestions on board and developed the Hospital Spell Analysis Dashboard. Benefits that have been highlighted for this dashboards are: • Greater Transparency: Removes the vagaries of different patient journeys. Greater range of filters to allow users to drill down to more detailed levels as required. • Hospital Spell Focus: This provides users with access to information around processes they control. • Diagnostic, Procedure, Specialty and Consultant Analysis : - Users can group consultant teams or specific diagnoses or procedures together and compare mean/median hospital lengths of stay for these specific conditions/procedures across hospitals and Scotland.

Results: - Hospital Spell Analysis Dashboard Released 1. Improved Trend Analysis: User can identify trends for Hospital Admissions or Bed Days and easily drill down to analyse specific conditions/procedures. LOS denominator is now hospital admissions as opposed to specialty spells or episodes, which is clearer and more understandable. 1 b. Hospital Length of Stay Profiles: Users can identify differences among length of stay profiles across hospitals in their HB area and against Scotland. Median LOS is available. Users are able to include/exclude zero day lengths of stay and patients with long lengths of stay, which may skew the mean. 2. User can easily identify the high volume diagnosis and procedures for all hospital admissions, available for all discharging specialties and admission types. Analysis available by Spells or Bed days.

Case Study: - Summary • The Discovery Tool is continually improving and evolving with new content being developed constantly. • The Discovery Team want to make the content as useful as possible to our users and believe the best way to do this is via a co-creation approach with the people who are going to use the information available. • If there are specific new developments that you feel would be useful or any improvements we can make to current content please get in touch.

For Discovery support and more information on accessing and using NSS Discovery, email us at: nss. discovery@nhs. net or visit our website: www. nssdiscovery. scot. nhs. uk