Q: How does the audit system check for MIs in the last year?
Answer: This functionality has recently been added to the Housekeeping set. Consider 2 Read Codes for MI, say G30y, one dated 1994 and another 2002. How do we determine if the second entry is in fact another MI or simply a review of the old MI episode. The key is in that word 'Episode'. Read codes are assigned an Episode Type - First, New or Other in Miquest terms. We are interested in the First or New types entered in the last year. Unfortunately errors creep in to recording these, hence the need for Housekeeping. MI codes entered using templates or directly do not give control over the episode type attached and so there is a danger of missing these. They need in EMIS to be added as problem titles and the New / First episode attribute added when the consultation is saved. Conversely, when saving consultations it is possible to record an episode as New instead of Review (Other). The sheets CHD3 & 4 are designed to give lists of patients which need these checks performing. As the number of patients having MIs in any year is relatively small this should not be a long task!
To keep the number that will need checking to a minimum the following is recommended:-
a) In EMIS use the Group function to group previous MIs under G3 for IHD as the active problem and use this as problem titles for reviews
b) For new MIs add MI as the problem title to the scanned discharge letter and relevant consultations, saving as a New or First episode as appropriate. After the acute episode end the MI problem, making it past significant, and group under G3 as above.
Q: We have a patient who has anginal like chest pains but has had normal angiography - should he be added to the CHD register?
Answer: Most practices seem to have a few patients who fall into this grey area. There are several ways of looking at this. Firstly you could say we are only going to add to the register those where we know there is an atherosclerotic plaque in the coronary vessels causing symptoms. The problem with this is that not all patients with classical symptoms where there really is little doubt about the diagnosis will have definitive investigations done. Others have experimented with assigning a degree of uncertainty to a diagnosis (as in later versions of Read). This however complicates the analysis and you still have to decide what degree of uncertainty you are going to tolerate in the audit. Another way of looking at it is to say we will add a patient to the register if we are convinced enough of their presentation that it is safest to give them the benefit of the doubt and treat them. We may want to assign uncertainty to the diagnosis but we still have to decide what we will do with a patient! If we are going to treat them we should audit that treatment. This is probably a more useful way of dealing with the above example. If the diagnosis turns out to be wrong later we can replace erroneous codes with [D] Chest Pain codes with free text explaining the error.