Your article 'Long Term Conditions: Predicting the Future' (2 November 2006) showed the value of measuring the risk of patients experiencing unplanned admissions to hospital, and I thought it would be helpful to highlight other work that is underway

Your article 'Long Term Conditions: Predicting the Future' (2 November 2006) showed the value of measuring the risk of patients experiencing unplanned admissions to hospital, and I thought it would be helpful to highlight other work that is underway

UnitedHealth Europe has developed a risk stratification tool called RISC that looks at the entire PCT population (not simply those at risk of readmission). It thus gives PCTs a tool that they can use to support commissioning as well as to identify patients at high risk of unplanned hospital admission.

We are working with PCTs across the former Trent and Birmingham and the Black Country SHAs in using this tool and are already successfully using GP data to amplify the tool's effectiveness. We believe our tool to be particularly valuable as it uses 'episode treatment groups' (data collected over time that groups episodes of care from different settings e.g. inpatient, outpatient, A&E, and general practice) rather than simple utilisation.

The other powerful aspect of our tool that is not available as part of PARR is a detailed clinical report. One report is a PCT summary of the risk scores, the other is a detailed clinical profile that enables clinicians to understand the drivers of an individual's risk score.

In working with PCTs we have learnt three important lessons about making the tool truly effective:

  • Staff in the information departments of the PCTs need support in ensuring that they can produce high quality data to populate the tool. (We build in customer support including system implementation and training to our contracts.)
  • Case managers using it have to be well trained and mentored and working within a system that is redesigned to reduce unplanned hospital admissions.
  • Risk stratification tools can be used to assist commissioners better understand the needs of their population. It will increasingly make sense to set budgets on the risk of individuals rather than populations.

Finally, the article said that PARR was able to predict 8 out of 10 patients who would have an unplanned admission with a 15% false positive rate, but the BMJ paper on PARR showed that it was actually eight in 100.

Richard Smith, chief executive, UnitedHealth Europe