Read codes, disease registers and the quality framework
It is simplistic to think the disease registers so vital to the quality framework can be based on Read codes alone, says Dr Jeremy Gray
GP negotiators have claimed that disease registers can be set up in
10 minutes using Read codes. GPC negotiator
Dr Laurence Buckman, for example, has said an acceptable level of accuracy can be achieved using Read codes to set up initial registers.
I have to say I think these views are over-simplistic. Research by the Government-funded Battersea Research Group has shown that basing IHD registers purely on Read codes misses one patient in three with the condition.
A substantial number of further checks are needed to ensure the register is accurate, and our study, which covered 46 practices, showed an average three-partner practice might need to invest up to a further 27 hours of work. This includes Read code checks and note sifting.
So much for setting up a register in 10 minutes.
For IHD, the latest version of the Read codes is G3. In theory an electronic disease register can be built doing a search to include all patients with a G3 code in their computer records. These patients will then create the IHD population from which the new contract points will be calculated.
But as I have indicated, we have shown that up to 30 per cent of this disease register may be wrong. Our research has shown 15 per cent of patients who have IHD may not be coded as such, while a further 15 per cent of coded patients have no evidence of the disease and should not be on the register.
What are the implications of coding patients with G3 who do not have ischaemic heart disease?
First, it is this group that will cause a shortfall in quality framework points if their management does not meet the quality criteria.
If they do not have IHD then they should not be on the register as they may not meet all the quality framework indicators.
Second, patients without IHD are likely to be overtreated with drugs such as statins as the real IHD population will be aiming for lower total cholesterol values than patients being treated with statins as primary prevention. There will be a lower threshold for adding additional medications such as hypertensive drugs in this group with the added risks of side-effects and complications
Finally, these patients may receive less preventive primary care in terms of dietary and exercise advice as they are perceived as already on medication.
How does a wrong coding occur?
Patients are miscoded for a variety of reasons, including the following:
· Administrative errors, especially when coding new patient registrations
· Using an angina term when the patient presents with chest pain that is subsequently not shown to be IHD
· Difficulty in removing all incidences of the G3 code from the patient record.
The other side of the coin exists too. Patients with IHD may not be coded with the G3 code and may not appear on the disease register. These patients should be receiving the full power of secondary prevention both in terms of follow-up arrangements in primary care, lowered threshold for referral to secondary care and for drug treatment but have been shown to miss out on this care if they are not on the register.
How should practices find patients with IHD who are not coded?
We have previously shown nearly all patients can be found using a five-drug search, assuming the practice does its prescribing on the computer as most do, and that the system can search on groups of individual drugs.
We calculated that it takes on average 41 minutes of work to find one new case and one case is found for every five sets of notes examined at a cost of £10.20 per case found. A drug search using a lesser number of drugs will be less effective but more time-efficient.
In addition, patients may be coded as having had an angiogram or CABG but not as having IHD. This is more likely to occur in paperless practices where mail is scanned and coded for the investigations but not the underlying disease.
Updating the disease register
Another flaw in the new contract's thinking that a G3 search will give you your disease register is the failure to recognise a disease register is a living entity.
As patients leave the list or die and new IHD patients join the list, or existing patients develop IHD, the register population will change.
A significant amount of work needs to be done to keep the register 'current' and this work needs to be incorporated into the everyday routines of the practice otherwise tedious paper searches will need to be repeated.
The new contract has allowed GPs to build on the work already achieved under the national service framework. There needs to be recognition that the quality of the disease register will benefit from having objective evidence recorded alongside the G3 code
of real disease. Patient care will improve if
this is done.