The purpose of the friends and family test is to ask a patient whether they would recommend the service to their friends and family if they were to require similar care. The results published of the friends and family test enable service providers to design and change services to suit the needs of the patients.
The friends and family test consists of 10 questions and is standardised over all service providers who utilise it. The test was first used by NHS funded acute services for inpatients, since then has been extended to women who use maternity services (October 2013) and now will be extended further in three different stages:
– GP services from 01/12/2014
– Mental health and community services from 01/01/2015
– Dental practices, patient transport services, acute hospital outpatient and day cases: 01/04/2015.
Past data from friends and family test is recorded on excel spreadsheets for public viewing and can be found on the NHS statistical domain. It is important to use this data in a way that can drive positive changes in the NHS for the betterment for patients and their treatment.
Text and sentiment analytic
The first 9 questions asked on the test are single response questions and it is the last question which is a free text comment box requesting patients to give reasons behind their single response answer. It is vital to ask this question because it provides context, reasoning and background behind patient responses.
Single response questions are extremely rigid, therefore it is not possible to cover all tenets of a patients care and treatment and there is no opportunity for patients to voice their opinion in a respected matter, hence this begs the needs to give them the opportunity to write their own comments. What may transpire is that a patient may seemingly give positive responses in the single response and a slightly more negative response in the comment box, which appears to be contradictory. A patient may be “very likely” to recommend the service but may still have qualms about food or a member of staff.
It is vital for service providers to appropriately process and analyse the free text data because it is these comments that will be the biggest driving factor behind improvement plans, changes in the design and delivery of services and so forth.
Text and sentiment analytics is an engine that is fit for this purpose. A text and analytics engine is the application of sentiment and emotion to free text responses, giving health providers an indication as to how strongly patients felt behind a certain element of care they had received. The engine uses an 11 point sentiment scale and carefully analyses the words in the comments in order to determine the sentiment and the strength of that sentiment. For example, the comments “the food was bad” and “the food was disgusting” are exactly the same sentiment, however, the word “disgusting” is stronger in representing that sentiment hence it would be rated higher in the sentiment scale than the term “bad” and it is to this detail the engine can ascertain how patients feel about aspects of their care and how strong that feeling is.
Furthermore, the engine is able to formulate themes and patterns from the text, which works in two ways. Firstly, if there are specific types of information a health provider would like, then this can pre-coded into the engine so it picks up those specific themes. Alternatively, the engine can be left to do its own natural and organic theming.
The end result is a dashboard that depicts the data and identified themes in four unique ways:
– Bar chart measuring the different themes identified against the amount of time each theme was mentioned allowing an healthcare organisation to understand the most heavily talked about topics.
– Word relationship cloud. Rather than forming the word cloud from a count of words, the cloud displays different themes as well as relationships between words for added accuracy. The colour of the word relationships depicts sentiment, while the size of the word represents the frequency of its mention.
– Double-axis chart, similar to the bar chart, however compares each theme against the total volume of sentiment received for that theme i.e. a combined value of sentiment per theme.
– Heat map showing different themes in different shaped and sized blocks, as with the word cloud, colour depicts sentiment and size depicts volume of theme.
The dashboard is organic and can be added to and combined with different data sets as required and preferred.
The numerous benefits for applying a text and sentiment engine include:
– Better justification behind decisions and changes implemented due to the ability to look at at data on a granular level.
– Better justification can be provided when applying for various grants such as the Nursing Technology Fund, once again due to the granular abilities.
– Identification of small issues, when corrected, has a major impact on a patients care and treatment that cannot be seen via manual processing.
– No manual processing/data entry involved resulting in savings in time and money as well as fully accurate data.
– Dashboard is user friendly.
– Better treatments, therefore happier patients.
Contact us to learn more or alternatively book your free onsite assessment where DCC are able to prepare a dashboard using your own data to demonstrate the powers of automation and the text and sentiment analytics engine.