Patient Level Costing - not just for geeks

When you work in health information, there is a terrible tendency to get into some bad habits. It’s a sector which is much too easily satisfied. We get data that’s 95% complete and we say ‘hey, that’s not bad.’ At 99% we’re ecstatic.

That’s why it’s been such an eye-opener to work on Patient Level Costing. This is because when it comes to completeness, finance staff only know one acceptable figure – 100%. If there's a discrepancy, then it needs to be small and we have to be able to account for it, to explain it in its entirety.

There’s something refreshing about this uncompromising demand for the highest standards of rigour. It’s made it a lot of fun to spend a lot of time over the last three years in this field.

As it happens, the fundamental principles of Patient Level Costing are an intriguing challenge in themselves. The problem is that you’re trying to reconcile the irreconcilable.

The clinicians who are involved with a particular case take a patient-centric view of what they do. They see that the patient had an X-ray of a leg and then an operation to pin a badly broken bone. He had an anaesthetic during the operation, an antibiotic to combat possible infections and an analgesic to combat the pain. He received nursing care, medical care and an intervention by a surgical team. Consumables were consumed in the operation, a bed was provided, the patient was served food (which may even have been passable).

Unfortunately, the hospital’s General Ledger simply doesn’t recognise any of these categories. There will be entries for drug costs and anaesthetics, there will be entries for staff pay, there will be entries for food and consumables and laundry and cleaning and maintenance. None of these will give anything like a coherent account of what happened to our patient. It’s rather like the first diagram below: two cogs that don’t mesh and turn in opposite directions.

Patient Level Costing therefore has to bring in a third cog, a Costing Engine, to mesh with the other two and make them work together. Such a costing engine has to include complex logical processes to convert the Ledger view of the hospital world into the patient-centric, clinician view. In other words, it takes the values in the ledger and finds a way to translate them into costs and income values that can be assigned to individual patient records.

Now that sounds like an exciting thing for geeks and no-one else. In reality, though, it’s a lot more significant than it sounds. Because carry it off and what you’ve done is to provide a common vocabulary for communication between Management and Clinicians. Currently Management sees that Trauma and Orthopaedics is overspent (that’s just an example, but isn’t it sad that it tends so often to be Trauma and Orthopaedics?). With Patient Level Costing, suddenly the hospital can identify the specific patients where the overspend occurred.

Suddenly the Clinicians can investigate the source of the problem. Is it a mistake not to give an antibiotic as a prophylactic before certain operations – does it lead to significantly higher costs, on average, later? Is it a mistake to use a particular medication to treat a condition which might be dealt with more quickly and cheaply using another, even though it is itself more expensive? Once the costs have been associated with individual patients, it becomes possible to answer those questions.

Which means that the problems can be addressed by the only people who can actually make a difference – clinicians.

So Patient Level Costing isn’t just for geeks. It’s for all of us who’d like to see care made more efficient – and more effective. And that’s basically all of us.