22 December 2014 07:50


News
Twitter RSS Newsletter Send to a friend
2

Johns Hopkins risk tool used in South

5 March 2012   Rebecca Todd

More than 200 GP practices in South Central England are getting quarterly risk stratification and predictive modelling information on their patient populations.

The South Central PCT Alliance is using the Johns Hopkins University Adjusted Clinical Groups tool to support risk stratification of the population, disease profiling, case management and resource management.

Alliance programme lead Alan Thompson said that of these GPs were mostly using the tool for case management at the moment.

The tool combines information from primary and secondary care. It is used to produce quarterly reports that are accessed via a secure website, with only those working within a practice able to see underlying, patient level data.

Community matrons can also be given access to data from a number of different practices in order to create case management lists.

“There's a belief that GPs know about all the patients that need to be referred to case management / community matron services, but in Buckinghamshire the system is finding patients the GPs didn’t know about,” Thompson said.

Buckinghamshire was one of the pilot sites for ACG, deploying it in October 2010. Thompson said figures showed that of the 100 highest risk patients known in one practice, an additional 17 were identified as being suitable for additional support by the ACG tool.

“There’s real value added from identifying ten patients in a practice that previously a GP or community matron was not really aware of - either they were not quite sick enough to be on the radar or people were accessing healthcare via secondary care and bypassing the GP,” Thompson explained.

Patient consent for the use of their data is implied, with information about the service provided via posters and leaflets in practices. If they decide to opt out, patients are assigned a Read Code so their data is not extracted.

Around 520 practices across nine primary care trusts could have access to the tool as part of the alliance.

Thompson said two primary care trusts had not really participated in the programme until now, because of other priorities or concerns about data security and information governance.

However, he said they had since been reassured and he was expecting a surge of sign-ups from GPs in those areas.

The contract for the system was in place until March 2013, when clinical commissioning groups will formally be established.

However, Thompson predicted that the tool would continue to be commissioned by CCGs in the area as the Department of Health was pushing the use of risk stratification to reduce non-elective admissions.

“Most CCGs will want to continue to use risk stratification software of some kind, we think our ‘at scale’ solution provides lots of benefits. I would be surprised if people didn’t want that to continue,” he added.

“What we will see from the CCGs over the next 18 months is increasing use of other parts of the ACG functionality. They want to have a better understanding of how the whole population is stratified in terms of risk.”

The South Central PCT Alliance was the first customer of the ACG tool which has since been deployed in other regions.

Thompson said the group was linking with those users to share their experiences with other regions and learn from them as well.

He also hoped to get a study started in Oxfordshire where there was some academic interest in looking at the impact of the tool on healthcare service usage.

 


Related Articles:

2 News: Nuffield issues predictive risk guide | 18 November 2011
News: Risk model applied to social care | 24 February 2011
News: DH quits predictive risk modelling | 10 August 2011
Last updated: 4 April 2012 12:38

© 2012 EHealth Media.


Follow

Benefit analysis?

spandango 145 weeks ago

Do we have any evidence regarding ACGs cost saving prowess...any cost/benefit analysis?

The society of actuaries report into risk assessment certainly didn't rate ACGs very highly (http://www.soa.org/files/pdf/risk-assessmentc.pdf), in fact the model comes out below average in terms of actual risk prediction.


Reply
Flag
 
Like

This comment is:

Email address:

Submit

Follow

Post

 

spandango
Expertise:
Policy maker/analyst
Sector:
Other
Approved posts:
3
Likes:
0
My EHI score:
3
Reward badges:

Frequency?

broken 145 weeks ago

I would have thought a more regular refresh of data e.g monthly would allow for more active case mgmt for those individuals identified at risk?


Reply
Flag
 
Like

This comment is:

Email address:

Submit

Follow

Post

 

broken
Expertise:
IT professional
Sector:
Public sector
Approved posts:
3
Likes:
0
My EHI score:
3
Reward badges:
 
 
[1]

About us | Advertise | Terms and conditions | Community and comment rules | Privacy policy | Cookie policy | Contact us
Digital Health Intelligence Limited
EHealth Insider is managed and maintained by Digital Health Intelligence © 2014
Registered Office: Southbank House, Black Prince Road, London SE1 7SJ
Registered No. 9257440 | Vat No. 198 3531 71