As a Trainer at G2 Speech, Selma sees the benefits of speech recognition to individuals and organisations on a daily basis. She knows the reluctance that some professionals might have and regularly hears the common myths and preconceptions around the technology. Here, Selma puts to bed those myths and tells the truths about the technology in practice.
Often when entering a new clinic, firm, department or ward for the first time I brace myself for the reluctance that might come whilst reminding myself of the many benefits my training and speech recognition can bring. I’m told stories on how the technology worked in a practice a clinician worked in 15 years ago. Below I will attempt to dispel some of those myths and shed light on the real worth of speech recognition.
Myth 1: Speech recognition doesn’t work very well.
Historically, when speech recognition was first introduced the technology was much slower which led to frustrations and meant a lot of time had to be invested in use. Today – the technology has become far more advanced, processors are much faster, and algorithms are much smarter. Speech recognition has become so widely accepted it’s become commonplace in the workplace. Hospitals and legal firms have implemented our solutions to reduce the time taken to create and deliver documentation and spend more time with patients and clients.
Myth 2: As an experienced typist, I’m able to type quicker than the clinician dictates, so there is no real need for a machine to do it for me.
This way well be the case, some secretaries can type with phenomenal speed. If every clinician had their own private secretary who could work the same hours as them and type up dictated letters live and direct, there wouldn’t really be need to change. Unfortunately when clinicians are dictating hundreds of letters a day that need to be typed and sent out to the patient/GP with a two day turnaround, the use of technology can really play a huge role. Backlogs in Trusts prove that there is a need to speed up and streamline the document creation process.
Myth 3: When technology makes mistakes, instead of correcting the odd word here and there, it is quicker and easier to just overwrite the entire document, which is in fact digital dictation.
This is a sure way to deceive the technology, which works on the premise that what you dictate is what you get. When you ensure the text matches your clinician’s dictation 100%, the SpeechReport recognition learns from this and becomes attuned to their unique dictation style, their words and their voice and ensures the recognition of future documents will improve over time.
Myth 4: The technology is very difficult to understand for someone who just doesn’t DO technology.
SpeechReport is very intuitive to use and after just one hour of training, you will have already personalised the app to suit your working preferences. For example, by sorting your worklist and the priority status of documents, by their length and by the clinician that dictated them. You do not need any specific skills to use the system. For example, even logging on to the system can be set up automatically so that once you turn and log on to your PC, SpeechReport is initiated. Once open, you can see all of the outstanding documents that require your attention, in their relevant queues.
Myth 5: The technology can’t cope with my non-native English mother tongue.
This is quite the contrary. The recognition system doesn’t need any initial training of the acoustic model of a user. This means that once out of the box, dictate and the system will recognise your words, irrespective on your accent or dialect, based on statistical probability. Any corrections that are made from initial use will train the system to improve these statistics.
Myth 6: Technology can’t compete with my secretary – with digital dictation I can dictate at any speed, make mistakes and my secretary instinctively knows what I mean and corrects me.
True, technology can’t compete with the experience and excellence of a long-standing, bona fide personal secretary who can modify documents and produce a professional document that goes far beyond the dictation they have heard. However when a Trust is producing thousands of documents a month with a two-three day turnaround time, it isn’t always possible to have that model secretary who can do them all! Looking at the whole document creation process, not just the transcription of a document, is where the value can be compared (from a patient walking into a clinical appointment all the way to them and their GP receiving a copy of that visit).
Recognition reduces this time and therefore saves money. The word perfect letter can be dictated even while the patient is in clinic, and within days their GP has the letter to make the necessary follow up actions.
Also, when transitioning from digital dictation to speech recognition, you may have to make small adjustments to your dictation manner, speak more slowly, articulate well and dictate punctuation, which will take up more seconds. However, the ability to produce and send out a document, even while the patient is in clinic, more than compensates for the seconds lost in comparison to digital dictation.
Myth 7: It isn’t possible for recognition to be able to recognise all the medical terms we use.
In fact, this is false. In order to kick-start the speech recognition system, documents previously generated by a user are fed into the system so that it learns how texts look and what words are used. In addition a medical topic is integrated so that both general and specific medical words form part of the dictionary. The recognition engine works by ‘recognising’ these words when a user dictates. If new words are detected, they are saved in the system and can be added to the medical dictionary.
If you’d like to hear more about what speech recognition can do for you or for a demonstration, contact us today: firstname.lastname@example.org or call (0208) 555 9041
To read about the benefits that Hull and East Yorkshire Hospitals NHS Trust has noted since introducing speech recognition, click here