NHS Comparators reviewed

Originally published in GP magazine .

Written for GP magazine around 2010. This is the version I submitted - the published article may have been edited.

I have a strange relationship with statistics. There is a part of me that is fascinated with playing with data and messing around in Excel. Pivot tables and macros can get me excited. However, the slightly less sad bit of me has no time to do this and just wants important data presented to me in a clear, logical fashion. I’ve struggled getting useful information out of Dr Foster and EMIS QUTE, and while there always looks to be a lot of interesting stuff on the NHS Information Centre website, finding anything of practical use is always time consuming.

Being asked to review the new NHS Comparators website therefore appealed to both sides of me, in that the data miner was excited to have a look at another source of information, but the pragmatist was intrigued by the promise that the information was presented in an easy to use, logical way.

NHS Comparators is advertised as a free comparative analytical service that enables commissioners and providers to improve the quality of care delivered, by benchmarking and comparing activity and costs on a local, regional and national level.

To access NHS Comparators you first need to register your details with the NHS Information Centre’s contact centre by calling 0845 300 6016 or emailing enquiries@ic.nhs.uk and asking for an NHS Comparators log-in. This was easy to do, and when I forgot my password they sorted it for me in a couple of hours - though why doesn’t it just email me the password like almost every other website? I had to register with an nhs.net address, so it should be secure.

NHS Comparators is available from both a real-world (www) connection as well as from an N3 connection, so you can log on from home when bored. There are currently about 200 indicators available, though they are developing the site all the time. Currently data sources on the site include secondary care activity data, GP list size data (population data), Quality and Outcomes Framework data and prescribing data.

When you log in you choose commissioner view or provider view, which to my mind is a little confusing. Effectively the commissioner view allows you to choose a practice and looks at both its QOF achievement but also its patients’ outcomes in secondary care, e.g. COPD admissions. The provider view allows you to choose a hospital and look at how they are doing in an area, e.g. COPD admissions - obviously these patients will come from more than one surgery.

It has a very pretty interface that includes a lot of graphical views and works quickly. I think it’s possibly too easy to get lost in the depths. There are some videos, bizarrely without sound, to watch that explain how to use the system.

The data is broken up into years and it defaults to last year, though you can choose others. A lot of the comparators have cost data as well as activity. The main menu lists all the indicators in the view you have chosen and you can expand all of these or select one. A side panel shows more information on the one you are looking at and gives tips on how to interpret each indicator.

On the whole it seems very good; however, this wouldn’t be a review without some criticism. The bit of me that just wants easy to view information isn’t satisfied. I would like a way of filtering only the indicators that are good or bad. I would like to pick a practice or provider and simply see what they are doing well or poorly in.

I would like the ability to pool practices together for certain things, as we are increasingly working in PBC groups, and comparing to my PBC group may be more important than to the PCT. I would however like to see common themes in a PBC group show up, so that if COPD was an issue across the patch it showed up as such. At the moment you will need to look at each practice in turn and then amalgamate the info.

The comparators are currently expressed in terms of rates per 1,000 practice population and it would be nice to have the ability to look at this on weighted populations as well. As a PCT we have markedly different demographics in areas of the patch and it can be difficult to compare practices.

I have to say the case study they use to illustrate its usefulness really doesn’t explain how it actually helped, and I think some better examples would be useful. If the site continues developing I can see me bookmarking it and visiting it regularly.