Digital Dictation or Voice Recognition?

dictation_or_voice_recognitionLast week, a GP shared his experience on the blogosphere in a post titled “How I improved the accuracy of our digital dictation software”. Great title that keeps you wanting more, except for the fact that:

a) You can’t improve the accuracy of a digital dictation. It just works.
b) The only thing you can improve is the accuracy of your voice recognition system, which is what this brave GP did. And I say “brave” because from what I could read, it wasn’t a small job.

 The reason why it took significant efforts (and creativity) from our GP to “train” his system is simple: general medicine is very close to normal language, therefore it is harder for the system to decide whether the 40 year-old man came into the surgery “with a swollen knee cap” or “with a mole on his wee back”. Some things are better left unknown, colon and period included.

So you get the point: voice recognition technology is tightly linked with something I personally despised throughout high school: the probability theory. In other words, if you are a Radiologist firing jargon all day long, then yes, VR is most definitely for you as the system, based on its built-in Radiology dictionary, won’t be tempted to turn “Dolichoectasia of the basilar artery” into “Dolly got ecstasy off the Brazilian eatery”.

Have you got the voice?

GPs often ask us to put them onboard voice recognition right away, without going through the digital dictation phase first. And when we ask them why, we tend to hear this:

  • “Because my secretary doesn’t understand a word of what I am saying.”
    (But if she can’t, what more do you expect from a computer?)
  • “Because I don’t have time and I need a computer to do the job for me.”
    (But if you don’t have time, how on earth will you find some to train the system to your voice?)

Which brings us to yet another major difference between digital dictation and voice recognition:

  • DD is not voice dependent and requires no voice training.
  • SR is completely voice dependent and requires voice training. So it will work better for some than others, depending on accents, tone of voice, pronunciation, etc.

The price is right

But it’s not the right question to ask.

In his post, our brave GP talks about pricing as in “cost” when he should think investment and return on the latter – which would yield brand new questions such as: “since I started correcting my own reports, did it steal hours away from my patients? If yes, how many? On the other hand, is my secretary able to dedicate one day a week to more revenue-generating activities?

The Money spent /Money saved ratio remains our only true guide, provided we know where to look. And once again, DD and VR don’t quite open the same windows:

  • VR can be measured in secretarial time and accelerated turnaround times. Period.
  • DD impacts the entire workflow: ability to reorganise secretarial workload in real-time, automation of report distribution, tracking and audit trail, reporting on referral trends. Reduced turnaround times is a clear benefit, yes, but one amongst many more (patient details imported automatically from the practice management system into the report, ability to even out workload amongst secretaries to compensate for peaks or sick leaves, ability to track phantom referrals in a few clicks every month, etc.)

The X Factor

Workflow Management: it’s the central framework that supports everything else, and the only true answer to the VR or DD question. If you don’t have it, then all you get is a stack of isolated reports flying at the first wind blowing – no tracking, no auditing; a rather dangerous move in any healthcare environment, not to mention in the new NHS.

Traditionally, DD and workflow management are one and the same. If you purchase a digital dictation system, you should get all the management and reporting tools that go with it. And right there, expecting your referral turnaround to be reduced from days to a couple of hours is more than realistic. Ask other practices using DD and they will tell you.

Then you can add the voice recognition ingredient to turn hours into minutes in some specific cases (i.e.: VR can be an interesting alternative to the secretarial route for some specialty doctors or in case of an emergency). It’s sweet as sugar but that’s typically Chapter 2 in your ROI scenario.

Voice recognition? Yes, most definitely; perfect icing on the digital dictation and workflow management cake, where the biggest (and tastiest) efficiency gains are to be expected.

Cheers,
The Equalizer

Related case studies:
The Westmoreland GP Centre
The Humbleyard Practice

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

CRESCENDO SYSTEMS LIMITED - Unit 1, Fiveways Business Centre, Feltham, TW13 7AQ - UNITED KINGDOM
Tel: +44(0) 1932 789433 - info@crescendosystems.co.uk

%d bloggers like this: