We know that not everyone understands health-related data in the way we do, so we asked two of our colleagues, Ashwani Sharma, Healthcare Data Analyst, and Dhirav Patel, Graduate Healthcare Data Analyst, to explain some of its more important aspects to help improve people’s understanding.

What is data?
How does data work?
Contextualising data
Why is data important?
What next?

In this blog post, we will delve into the essence of data and its pivotal role in transforming healthcare.

What is data?

Data refers to raw facts and figures, such as patient counts, hospitalisation durations, and treatment complications, collected from various health service sources. Data serves as the foundation for generating meaningful information.

Using this raw data, information is produced by its collection, manipulation, and analysis. For example, in a healthcare setting clinical data can include:

  • a count of the number of patients treated by a particular hospital or consultant over a year,
  • the number of nights patients spent in hospital for a particular procedure, or
  • the number of times patients suffered a serious complication of their treatment. 

At PHIN we strive to convert these data into meaningful information that can be used for informed decision-making.

By analysing the millions of health information records we collect from healthcare providers and national data sources we use data to help patients make better choices about their treatment options. Our information can also help hospitals and consultants to identify and understand their strengths and weakness.

How does data work?

Digital transformation and ‘big data’ have propelled organisations into a new era of data-driven decision-making and improvements in the delivery of services. Healthcare is increasingly at the forefront of this data science revolution, putting in place digital health systems to measure and integrate vast and complex arrays of data to improve the safe, effective and efficient provision of patient care.

For example, innovative technologies like RFID sensors are being utilised to track the movement and location of various items in a health-related setting, from new-born babies to artificial hip joints. Remote monitoring of blood sugar levels in patients with diabetes is now possible, allowing for proactive and personalised care. And the shift in data management from paper to electronic health records (EHRs).

Innovations like these collect and generate vast amounts of patient data and are empowering data scientists and organisations, such as PHIN, to produce data analytics information for patients, healthcare professionals and other stakeholders. By collecting data electronically (in digital format), it can be analysed more rapidly, in larger quantities and employing more sophisticated tools and methods, including artificial intelligence, predictive modelling and machine learning. This enhances our ability to sift through the raw data to get to the ‘meaningful’ information insights.

Contextualising data

Context plays a critical role in interpreting and applying data effectively. Mere presentation of data without context can be misleading. This is why during the Covid 19 pandemic when the government provided daily briefings, they had experts on hand to explain the information being presented. To overcome this challenge, we emphasise the importance of understanding the circumstances in which data is curated. By knowing the "who, what, where, when, why, and how", we can ensure accurate interpretation and avoid misrepresentation.

You may have heard the phrase: “There are three kinds of lies: lies, damned lies, and statistics”. This emphasises the fact that providing data without context can be misleading – either inadvertently or intentionally. However, when data is given the proper context it becomes a more powerful tool to influence the choices we have and the decisions we make.

As we’ve said, data provides the basis for the information patients and others need to play active roles in the decisions surrounding their healthcare. But, as we’ve also now pointed out, to fully understand its meaning and how best to use it it’s important to know why, how and from where the data collection happens.

Why is data important?

Data enables us to understand the cause-and-effect relationships within healthcare. For instance, we collect and publish information on ‘Adverse Events’ – events that should not happen to patients because of their time in hospital. These events include unplanned transfers, emergency readmissions, returns to theatres, deaths, and serious injuries. 

Unplanned transfers: Where a patient experiences unforeseen and escalating health needs during their admission which their hospital cannot support they may be transferred to another hospital.

Emergency readmissions: Where a patient, who has been previously treated in hospital, has to return to the same hospital as an emergency within 31 days of being discharged for a problem clinically related to their original treatment.

Returns to theatres: Where a patient develops complications following an operation during their admission and must go back to theatre unexpectedly.

Deaths: Where a patient dies, either during their stay in hospital or shortly after discharge, and the death is clinically related to their treatment.

Serious injuries: Where a patient suffers a physical injury whilst in hospital which has resulted in significant harm and/or reduction in their quality of life.

For patients, a simple count of a hospital’s adverse events cannot necessarily tell you how safe that hospital is because it can be influenced by many factors over which it may have little or no control – for example how old, frail or unwell its patients are before starting treatment. Also, one hospital may have more adverse events than another simply because it treats more patients or the type of treatment they provide is complex like Cancer treatment.

To turn this data into useful information, instead of a simple count we could calculate how often these events occur in relation to the number of patients treated and present the data as a rate – such as 5 events for every 1,000 patients. We could also calculate the rate for different age groups to reflect the possibility that some events are more common in older people – such as death. This way information can be tailored to better reflect the experiences of different types of patients and enable the information for different hospitals can be compared.

For healthcare providers these data can also help indicate if they have more adverse events than expected based on the types of patients they treat. Without the ability to compare performance hospitals are unable to best assess whether they are in or out of line with their peers or to have insights into their relative strengths and weaknesses.

What next?

In our next blog post we explain how to use PHIN datasets

Was this article useful?