Revisiting Patient Experience
Collecting a patient’s experience of care whilst in hospital is extremely important. The feedback a patient gives can serve as an important learning curve for a health institution. All hospitals collect patient experience, however, some find it difficult to manage patient experience data, whilst others are unable to act with the data they’ve attained simply because their existing system do not allow them to visualise that data in a way make appropriate changes can be made.
DCC have been involved in numerous patient experience projects and we have a great deal of understanding in how best to manage patient experience data in a way that permits a hospital to make improvements.
Colchester Hospital Cancer Treatment Experience
Using a current example DCC will attempt to explain how data received from patient experience can be used to its maximum ability.
Healthwatch Essex is currently running a patient experience project at Colchester General Hospital and Essex County Hospital. They wish to gather data on the treatment former cancer patients received whilst in their care, for up to the last four years. Such an example is an apt one to use because they have an online and paper version of their form.
As a paper copy of the questionnaire is not available, we’ll look at the online version with the assumption it follows the same format and asks the same questions. The questionnaire is eight questions long and contains a mix of closed and open ended questions.
The online form is easily explainable and works seamlessly, business rules and validations would be put in place to ensure that the data entered is accurate and all compulsory questions are made apparent before an individual is allowed to submit the form. Upon clicking submit, we would route the data directly to your system/database where it would be stored.
The greater benefits line in automating the paper version of the form. In the traditional methods either a PALS team or a patient experience team would receive the form and manually enter the data into a database. The disadvantages of such a method are that it is a time consuming process and prone to human error in re-entering the data. Data automation would eliminate the manual entry aspect. The forms would be scanned and extracted using methods of optical mark recognition for closed-ended questions and intelligent word recognition for the free text, open-ended questions. If you’re further interested in learning about extraction techniques, click here
Furthermore, whilst the software is verifying the data, it is at this stage business rules and validations would come into effect. Any word or character the software is unsure of; our team of highly skilled verifiers would intervene and correct discrepancies. Anyway, enough of the technical side of things, once the data has been extracted and verified it is directly exported to your preferred storage location.
Hospitals use many different types of systems and databases to store their data. We have found some systems failing, for example, in extracting specific data they want, unable to view that data graphically and in an interactive way or there are cost additional implication to the aforementioned elements.
DCC’s interactive dashboard is able to pull in data from the paper and online form and display it on a dashboard. The dashboard is fully flexible and is able to highlight and extract specific data they want and ignore the rest. Visualisation of the data is in the form of graphs and charts. Moreover, the data can be organised using tabs. In the example of Colchester and Essex Hospital there is the possibility of organising the tabs on the dashboard by year for the four years they’re measuring and compare the results year by year to establish how experiences have progressed or regressed. In addition, you could pull in the free text data- this further adds clarity as a user can then proceed to understand the reason behind any progress/regress. The last question on the form asks the patient to give their consent on an informal follow up interview. It is possible to export the data from the informal interviews onto the dashboard for further in-depth comparisons to help healthcare professionals understand the reasoning behind their patient’s answers. Lastly, results between the two different hospitals can be separated for further comparison purposes. Depending on the types of data being dealt with and the specific information that needs extraction, contrasting and comparison capabilities are almost endless.
Text and Sentiment Analytics
A text and sentiment analytics engine is a notch up in the scale. Such an engine is able to rate data from free text answers, such as the open-ended questions and the formal interviews, on a sophisticated 11 point sentiment scale. Not only does this tell you what the patients are saying but also explains what they’re feeling. The unique aspect of the engine is that it understands the relationships between words in a sentence and can automatically develop themes and categorise them. Using a simple example, if a patient were to say “The hospital was untidy, but the nurse friendly”, the text engine would firstly pick up the themes (cleanliness and staff) and apply a sentiment to each of these themes. All themes are then exported to the interactive dashboard and visualised in four distinct ways:
– Bar chart: The bar chart would show each category/theme compared to the amount of times the theme was mentioned. The purpose of this is so that a user understands the most talked about topics.
– Word relationship cloud: The word relationship cloud would again show the amount of times a theme was mentioned (the size of the word would depict this). In addition, however, each word would be colour coded to represent the sentiment felt towards those themes as well as making relationships between words apparent.
– Double axis chart. The double axis chart is essentially the word relationship cloud in a chart format. It depicts the amount of times a theme is mentioned, like the bar chart, but compares it with the overall sentiment rating.
– Heat map: The heat map is another unique way of displaying the data. Each theme is represented in a block. The size of the blocks depicts the amount of times the theme is mentioned and its colour the sentiment.
The dashboard can illustrate an overall picture or individual sketch; it’s up to you and is fully flexible. The example given is taken from the Clarabridge Dashboard engine.
We thoroughly hope it was useful going through the methods of processing patient experience data. The Colchester form was used for no particular reason except that is a current project and the practical use of an example which includes closed and opened-ended questions as well as online and paper forms would help enhance understand.
Therefore, if your health institution is planning a specific patient experience project or would like assistance in their on-going projects, book your free educational knowledge share workshop. The workshop serves the purposes of further detailing you on best practise methods in patient experience projects as well as the technologies available to you to enhance your process.