How to record consultations with Vision

Using Vision: A beginners guide for clinicians

Part 1 of an occasional series

If you are new(ish) to Vision, or not so new but still feeling frustrated, then this occasional series is meant to help. It's occasional because it's written in my own time and there are no guarantees I can finish it, at this stage at any rate.

Comment if you find it useful - this may even encourage me to do more!

Why Vision is Different

Vision is different from paper records, and from many other clinical information systems, because it uses a very structured approach to data recording. What this means in practice is that nearly everything you record in Vision hangs off of a Read Term. This scares people at first, because all they want to do is type in text. But, generally speaking, computers are not very good at understanding text so, if you want a high quality (and thus usable) medical record you need to record the stuff you need to record in a way the computer can use it.
The consequence of all this is that unless you understand Read Terms at least a tiny bit you will find Vision a mystery. Most clinicians (but not all) do not understand Read Terms. Some think they do, which can be a good start, but actually they misunderstand them through limited exposure and use. Some understand their structure very well, but not how to use them in clinical practice.
The other big difference with Vision is that it lets you the user, and your practice, configure it and set it up to work as you want it to. This is both a blessing and a curse. It's great to be able to choose, but unless you (or someone in your practice) bothers to learn how and why to use the configuration tools and then implements these changes, it can remain a frustrating mess. OTOH, once you have learned the basics, and set it up, it will work like magic :)
Don't panic, it is actually dead easy to use and to use well, but you will have to learn what you are doing - it won't just magically happen.

Why bother with Read Terms?

Because without them your computer system will just be a jumped up word processor. Unless you tell the computer that the patient has Hypertension using a code, then the computer will not know the patient has Hypertension. We have learned some of this through QOF, I think. But Read terms also contain many, many other codes and descriptions for many, many other pieces of clinical information. Not all of these bits of clinical information are hugely useful on their own. For example, the term 'O/E Blood Pressure Reading' is of limited value if all you do is type in some free text for the systolic and diastolic values. Remember - the computer cannot use text alone; it needs structured data. So Vision associates Read Terms with special forms or 'Data Areas' for capturing all this other data in a useful (i.e. 'structured') way. So, the Read term 'O/E Blood Pressure Reading' is associated with a Blood Pressure form, where you can add lots of other things about a blood pressure if you so wish. Storing Systolic and Diastolic in this way means that they can be searched on, reported on and analysed both for individual patient care, and for other uses such as QOF.
Vision does this for lots of Read terms - associates them with forms that let you capture the other bits of data relevant to that Read term in a structured way. They call them, rather annoyingly, 'Structured Data Areas', which when Vision first came out sounded cool, but in the iPhone age starts to sound a bit geeky. Whatever you call them, this method of associating clinical terms to clinical data is pretty much how things are done (or should be done) in clinical computer systems. It means that a blood pressure in Vision is always understood by the system as a blood pressure. There is only one place to store it (though lots of ways to get there), and that place captures all the bits of data about a blood pressure you may ever wish to record. For those who are interested, all this is the basis of an Information Model.

So if they are so important, how TF do I use them?

And this is what gets folk confused and annoyed, because a) Clinicians like to write free text narrative (except very few of them really do that!); b) Read Terms are 'an abstraction of reality' so, inevitably, there are gaps, errors and ambiguities and c) Finding Read Terms can be tricky, especially if you don't understand how they are organised.
Regarding (a): in fact clinicians use convention and shorthand to record consultation narrative. Clinical notes are full of examples of this (e.g. 'o/e' for On Examination; the use of '0' (zero) prior to text to indicate the absence of something e.g. '0fc' for No finger clubbing), and most clinicians have spent large chunks of their careers learning how to do this. We do not just write long hand text. Vision can be used in a way that supports a lot of this clinical shorthand, making recording a consultation quick and intuitive whilst also gathering lots of useful data. How much of this you use or do is, of course, up to you - you can just enter a string of free text in the comments box, ideally underneath an appropriate Read Term, and this is often a good way to start.
(b) we cannot do much about, but understanding how Read is put together can help you in choosing a suitable term, even if it is not quite what you were trying to say. Qualifying or explaining the Read Term in the free text is entirely sensible, and what the free text is for!
(c) is overcome by this blog posting, which explains the basics of Read, and by the tools in Vision that make it easy to manage the Read Terms dictionary.

What are Read Terms then?

Read Terms are a 'clinical terminology', but you do not need to know this. Better to think of them as a 'Thesaurus' of clinical terms. They are a way of attaching a number to the stuff you do, so computers can make use of it and there are about 120,000 terms. These are organised in broad chapters, which are numbered (coded) beginning with the number '0' (Occupations) then '1' (History / Symptoms) and then continuing up to 9, then through A to Z. The chapters are hierarchical - that is, as you expand them the amount of detail, or how specific the term is, increases. For example:
Chapter 2 contains 'Examination / Signs' terms. It looks like this:

The number (and any dots that follow it) is the Read Code, thus '212A.' is the Read Code. Every Read Code has an associated description or term, in this case it is 'O/E Patient well'. There is other stuff that I am going to ignore for now, such as Preferred Terms and Synonyms, as you can get by without it. Read does, however, also have abbreviations for terms built in, which are typically short forms of the term. These can be up to 10 characters long. For example, 'MI' is an acronym for the Read Code 'G30.. Acute Myocardial Infarction'. This is important because Vision makes use of these as 'keywords' that lets users quickly find the term they want. More later.

The Read Hierarchy

Read is split into broad chapters, as mentioned above. If you examine the picture below, even briefly, you will see that chapters 0,1 and 2 are about history and examinations; chapters 3 to 9 are about stuff that clinicians and others do to, for or about patients; and chapters A onwards is all the disease stuff.
Now, clinicians often get hung up on trying to find a diagnosis for a patient every time they see them. If you are one of these people, please stop! In general practice we manage a lot of chronic disease and a lot of uncertainty. Much of the time there is no diagnosis, or the diagnosis is not new. What you want to record is what the history was, what the findings of your examination were and what you are going to do about it. IOW, terms from chapters 1 and 2, and sometimes 3 to 8.
If you do have a diagnosis, by all means add it and, usefully, abbreviations (or 'keywords') exist for many common ones, such as 'uti' or 'urti'.

Bad things to do with Read

Use 'Chat' and 'Patient Reviewed'

Because people get lost in Read, they latch onto an inappropriate and often useless Read term to give them a comment box to free text in. It is better, when learning, to at least use high level chapter headings about the stuff you have done and capture it there, e.g. '1....History / Symptoms' for free texting the history and '2....Examinations / Signs' for free texting the examination.

Say 'not' in free text

Classically, 'G20.. Myocardial Infarction' as the code, with the free text 'Negative'. Read does not say 'not a thing' very well, but it has some terms for 'negation' and they should be used where possible. Where not possible, record symptoms. For example '1822. Central Chest Pain', free text 'MI Screen negative'.

Free text values and data that should be captured!

For example, adding a blood pressure as free text is not much help:

If there is a 'Structured Data Area' for the thing you are trying to record, use it!