Atlas are not Roadmaps - why big data does not lead to change

The recent edition of the Bulletin of the Royal College of Pathologists contains an article by Bernie Croal that discusses the NHS Atlas of Variation. The Atlas of Variation is an interesting project that takes health data from across the country and displays variation graphically. As discussed in a previous blog, pathology requesting patterns can be a rich source of data for analysis of variation. The NHS Atlas looks at a number of examples at a national level, including TFTs, BNP and Ca 125. The article discusses the large variation seen in these requests across the country, and highlights the role that inappropriate requesting, variation in local clincal pathways, and variation in laboratory practices have to play in this. It is probably hard to justify variation, although sometimes this may reflect differences in local availability of other services (eg. access to echocardiography and BNP). But variation is always interesting and should never be ignored. Variation leads to health inequality, and drives the failure demand that may account for up to 75% of health expenditure. The problem with big data approaches like the Atlas are that they are easy to ignore. There are at least 2 reasons why this occurs. It is hard to know what the root cause of variation is (many reasons for variation). The problems are often at a system level that most people feel disempowered to change - and getting to the root causes requires a level of immersion in the problem that few people with the 'power' to make the change are able to achieve. The second related problem is that it is hard to engage those who are nearest to the root cause with aggregate data - it lacks the personal touch and context that is so important. Not only does the data feel irrelevant to an individual, it is easy to dismiss as someone else's problem. And so we return to the concept of the general pathologist. What we need are people who are skilled at analysing data and expressing it in ways that are meaningful to those that are nearest to the roots of an issue. We need people who are skilled at working with these people to understand this variation, and this probably requires a considerable amount of time reviewing case histories, talking about alternative management options, and ultimately discussing with patients what their experiences are. This granularity and personalisation of data is where pathologists can add real value. Working in this knowledge economy is the way that pathologists will help to turn the atlas into a roadmap.