Improving data and analytics to enhance practice and patient experience

Ideal Health share their experiences and explore some of the challenges and potential solutions for delivering data, analytics and information that creates an accurate picture of a patient’s care and enables improvements in practice, patient experience and safety.

The NHS challenges with data and analytics

Many people will have experience in ICU of data collection onto a giant paper chart, with a nurse analysing this in her head. A clinician could literally glance at the chart which was updated hourly and would know instantly how the patient was doing or what was needed.

The ultimate intention for many Trusts is to replace this physical chart with an electronic version but this is often met with strong resistance as the EPR solution is not as intuitive.

New digital systems are improving all the time, but there are still fundamental issues to be resolved including over complicated visualisations, system integration limitations and unintuitive system workflows. These all lead to poor compliance and data quality errors which can then lead to ineffective patient pathway management. This can have a knock-on effect on patient care.

Technology may have changed the way data is captured but it also needs to support the information needs of the practitioner in a way that is practical and intuitive.

More attention to intuitive EPR design would provide less reliance on training and leave more time for patient care. Although great progress has been made there is still some work to do in this area.

Using data and analytics to drive improvements in patient experience

Trust information departments have often focused on production and validation of statutory data sets to support commissioning. There are challenges faced in trying to move beyond this and identify what else is needed.

Data collection and analytics have been largely retrospective, reactive and resource heavy with large teams of people required to do queries, corrections and reporting. Clinical data sets exist, if not in isolation, then without much stewardship. They were often implemented without the ability to embed clinical terminology, and eluded efforts to evaluate quality and completeness.

Current trends to share these clinical data sets can run aground based on lack of semantic interoperability, requiring further mapping and maintenance based on the new data context. Analytics and the information generated needs to become much more predictive, proactive and automated, getting real-time information alongside clinical decision support.

When clinical systems are implemented, the objective is to create an accurate picture of what has happened with a patient’s care. We need confidence to be able to say that everything in the record DID happen and everything that’s not in the record DIDN’T happen. For clinicians, certainty and provenance of data are crucial.

Analytics has a key role to play. At the individual level we can ascertain that something is going wrong in a clinical system, but without analytics, it will be difficult to establish a pattern across a group of patients to gain insight.

Examples of where things have gone wrong –

• When free text clinical notes weren’t disabled, some users started using the function but didn’t tell anyone else.
• The implementation of a new system didn’t include switching off the old system so both were used at the same time.
• Different ways of recording a parameter were used in different systems ie no pain = 0 in some places and 1 in others

Each of these examples will cause major problems.

Approaches and Solutions

When looking at EPR maturity and adoption, as well as analytics maturity assessments, Ideal sees the same issues time and again.

HIMSS is a tool which tells us a lot about data and analytics issues and can also help with solutions.

• Through the HIMSS EMRAM (O-EMRAM) and AMAM assessments we can identify at Trust level whether there is less than complete integration of multiple front-end solutions.
• Integration is a key solution to data issues. Complete and reliable replication of patient and visit entities throughout the data journey is essential. It is important that data is shared and visible at all touchpoints along a care pathway/patient journey, not just siloed within applications or organisations.
• If uncertainty of meaning is introduced by sharing data, there’s a problem that has the potential to grow, the further away from source the data gets. Need to avoid probabilistic matching of activity to visits and implementation shortcuts – such as PDF or textual results – leading to data being unusable beyond its initial purpose.
• We can identify Data Architecture problems, such as issues with design, and unintended system usage. A ‘Right Function/Right System’ approach can ensure that analytics isn’t trying to do things that should properly live in the originating system.
• The HIMSS CCMM assessment can also be used to look at data sharing and continuity of care across Integrated Care Systems and evaluate not just the technical areas requiring improvement, but also other areas such as governance, patient experience and the clinical perspective. Improved data sharing and analytics at ICS level can support population management, prevention and personalisation.
• HIMSS is a good diagnostic tool for analysing the gaps in systems and data flows. By identifying the issues and gaps, we can develop recommendations and solutions for real improvements that will help both front line staff and patients receive the data and information that they need.

Engaging people and creating trust in the data

The two biggest assets of a healthcare system are its people and its data. Valuing our data and enabling trust in it is crucial.

Recognising the importance of data improves investment in it and extraction of value from it. It is important to manage the quality of data and do it right first time, to start to build trust in it. Every £ spent in the design phase, saves £100 in the management of it.

Good engagement with all impacted stakeholders is also critical to ensure involvement and a willingness to make improvements work. Stakeholders should be involved in the specification and design decisions, so they have realistic expectations, and the solution fits their needs. They should be part of the Design Authority and design assurance process.

Strong change management and a comprehensive communications and engagement strategy needs to be in place to create understanding, awareness and support for what the system is going to deliver. By setting out the case for change early on this can alter any negative perception that the system is an imposition and adds to workload. The overall objectives need to be explained and accepted and should answer the ‘what’s in it for me?’ question.

The benefits to the user and to the patient need to be clear. Strong engagement leads to a successful change in behaviour, with increased adoption, compliance in system usage and correct data input, as well as improved reporting, information and decision making.

The key steps involved in improving data and mitigating the causes

• Listen to the client, as in most cases they will know what the issues are. Our approach is based on working with the clients and not telling them what to do. There are incredible talented and dedicated staff across the NHS who need extra support and don’t have time or capacity to focus on resolving data issues.
• Once a basic understanding from the client is established, the next step is to look at the data and do logic-based tests to identify inconsistencies and errors.
• Then through validation, trends and patterns of anomalies become clearer which helps to establish the root causes. This allows for mitigating action to be taken, such as training, better standard operating procedures, correcting poor system usage and improving workflows.
• Ongoing system monitoring and management enables early issue identification and mitigation and helps avoid crisis management
• It is about cultural change and ensuring clinicians see EPRs as a help not a hindrance
• Optimisation work gets amazing results as its about getting the most out of existing EPRs to find efficiencies that can lead to shorter clinic times and more accurate patient records which leads to better overall patient care.

Why is Master Data Management so important and what can improve it?

Gartner describes MDM as “a technology-enabled discipline in which business and IT work together to ensure the uniformity, accuracy, stewardship, semantic consistency, and accountability of the enterprise’s official, shared master data assets.”

Master Data Management principles require data ownership and stewardship of end-to-end data flows.

• Ensuring ownership – Assigned responsibility for data quality and completeness, focussed on departments.
• Data flow assurance.
• Semantic interoperability – taking data out of context is dangerous: incomplete SNOWMED implementation, and poor coding and classification can exacerbate this.
• Identifying and resolving issues
• Optimisation to avoid future issues
• Move from fixing the data to fixing the systems or processes. Poor or incomplete system integration creates orphaned procedures and data silos. Good integration, interoperability, and implementation are critical to improve data and practice.

Benefits Summary

There are many challenges associated with data and analytics and we have considered the potential solutions of better integration and data architecture, improved master data management, root cause analysis, ongoing optimisation and the important people element of managing change, engaging staff, encouraging involvement and gaining trust.

There are many benefits if these solutions are put in place, including –

1. Optimised systems and efficient workflows
2. Effective capacity management and patient flow
3. Fewer data validators as data is correct first time.
4. Breakdown of siloed working
5. Consistent, visible data
6. Better data and analytics that inform decision-making about care
7. Greater trust in data and recognition of its value
8. Advances in population management, prevention and personalisation
9. Better patient care and experience