Social media and related digital technology has had a big impact in many industries and sectors, however, it remains largely untapped and under-used in the NHS.
Patient experience is the voice of the patient thus it is important that this voice is heard. Hospitals and Trusts have a Patient Advice and Liaison Service (PALS) team and a dedicated patient experience team. The respective groups are responsible for the patient’s voice being heard, they achieve this by welcoming all feedback and complaints from patients in the form of emails, letters and verbal communication. In addition hospitals have their own patient experience forms and friends and family tests that patients and/or their family are encouraged to complete.
Despite their continued attempts to collect feedback, hospitals are missing out on social media- a medium whereby patients are constantly tweeting and posting about their experience. This can be extended to websites such as patient opinion and NHS choices.
Though most Trusts and hospitals are on Twitter and do respond to individual complaints and feedback, it is largely an unstructured approach. The mass amounts of unstructured data on various public domains present a gold mine of information to use to improve services within your hospital. The next question is therefore how to do this?
The first step is to add structure and organise the data found on external websites. Each patient comment is likely to be unique as it represents their individual experience. Therefore when extracting the information, it is important to formulate themes and groups. A text and sentiment analytics engine is able to take in the free text data and categorise the types of issues being discussed by patients and apply a sentiment rating to depict whether the issues mentioned are in a positive or negative light. If a trust wishes to focus on specific bits of information, whilst omitting the rest, the engine is pre-programmable to pick up data related to pre-set categories. An interactive dashboard can present these findings in the form of word-relationship clouds, bar charts, heat maps and axis charts. The benefits of this include the ability to respond directly back to patients, address a specific recurring issue within a hospital as well as the ability to compare the data with feedback coming in via specific patient experience survey. Moreover, we have come to know that quantitative (tick box) data conflicts with qualitative (free text) data. For example, a patient may say in the friends and family test that they’re likely to recommend their friends and family, however, the free text data shows that same patient discussing various flaws in the service and in their treatment-implying a recommendation may not be in order. Discrepancies such as these are easily identifiable using the dashboard.
DCC are currently in the process of trialing text and sentiment analytic on patient experience surveys with various trusts. If you’re interested in joining the trial or simply would like further information then please do get in contact. Furthermore, feel free to book a workshop (What is a workshop?).