Tackling the Elective Surgery Backlog – Does Data Science Hold the Answers?

Tackling the Elective Surgery Backlog – Does Data Science Hold the Answers?

Richard Vize

On 15th July 2021 Together for Health in partnership with Oracle brought together IT, Data and Operational leaders from across the NHS to discuss the potential for data analytics and data science in supporting NHS leaders and clinical decision-makers to tackle the elective surgery backlog. In this article we summarise the key themes and findings of the discussion. 

The size of the challenge

Four million fewer people completed elective treatment pathways in England in 2020 compared with 2019 (down from 16 million to 12 million), while six million fewer people were referred for elective care, presenting a serious additional risk for post-pandemic planning. There are now 4.7 million people waiting for treatment, including 387,000 who have been waiting more than a year. It is feared that there could be as many as 3000 additional deaths over the coming years as result of delays in treating cancer during the pandemic.

Many aspects of the pandemic have exacerbated health inequalities, including the impact on elective care. The number of completed treatment pathways in the most deprived areas of England fell by 31% compared with 26% in the least deprived areas. 

There were also differences between specialties, with trauma and orthopaedics – one of the higher volume specialties – experiencing a drop of about 75% in May 2020, while neurology and thoracic medicine had much smaller percentage falls.

The delays in providing treatment are leaving many people in chronic pain, and are creating additional care needs, including greater costs for social care.

All this demonstrates the range of factors that could be considered when analysing data around elective care and prioritising patients for treatment – individual need, health inequalities and disparities between different disciplines all have their merits, and all have their own ethical implications. 

Understanding the problem

In exploiting the potential of data science it is vital to understand who and what is represented in the data, and where there are biases, uncertainty and gaps. For example, there are big differences between patient data and population data, which may drive different conclusions on issues such as the wider determinants of health.

Few trusts have a detailed understanding of the elective backlog they face, underlining the need for far greater collaboration between the business intelligence and operational teams. There are around 14,000 data analysts in the NHS, but a lot of their time is consumed in low value, routine reporting which could be automated, enabling them to focus on bigger questions.

Linking data

Much of the power in data analytics comes from linking datasets, such as patient information with housing and education to understand what is driving health inequalities. This might show, for example that a woman with breast cancer from a deprived community is more likely to have a mastectomy instead of a less invasive procedure compared with a woman from a more affluent area, or a man from a deprived area is more likely to end up with a colostomy bag than have their bowel repaired.

The need for board-level leadership

If data is going to play a central role in driving operations, leadership has to come from the board. Data analysts need to build relationships and know who to lobby internally to build a coalition of support. Don’t be shy of building the profile of the business intelligence team.

Analytics needs to be a separate function from IT. The institution needs to be brave rather than risk-averse in its attitude to information governance. 

Transparency is key

With data driving important operational decisions, openness and transparency about what is being done and why is essential for maintaining confidence and trust with both the public and NHS staff. What algorithms do and the impact they have need to be explained clearly.

Transparency also encourages organisations to keep it simple, rather than introducing layers of unnecessary complexity which are impenetrable to the public, and be clear on success criteria.

Transparency ensures algorithms are open to scrutiny and challenge, helping root out biases.


  • Andi Orlowski, Director, The Health Economics Unit, Senior Advisor, Population Health Management, NHS England
  • Michael Connaughton, Head of Analytics & Big Data EMEA, Oracle
  • Ellen Coughlan, Programme Manager – Analytical Capability, The Health Foundation
  • Dr Marc Farr, Chief Analytical Officer, East Kent Hospitals University NHS Foundation Trust
  • Tabby Boydell, Elective Care Recovery and Transformation Specialist, The Royal Orthopaedic Hospital NHS Foundation Trust

Follow Together for Health: