top of page
  • Writer's pictureSimon Beaumont

Disaggregating measures - Personalising content

Working in healthcare I regularly come across dashboards that aggregate a metric to a single number, quoting % compliance over a given period of time and RAG rating the number to symbolise full, partial or non compliance. Whilst this approach can have its merits (although in my humble opinion, not many) when needing to present data strategically and provide an high level oversight of performance I think it often leads to closed conversations, with leaders gaining false assurance from am aggregated number, forgetting that within that number that may be patient level variation that could warrant further investigation and action. 

For example, imagine you are a Senior Manager, and you are presented with the chart below that analyses performance for a specific measure, with performance analysed at an aggregated team level. The high level measure analyses the % of patients on a team's caseload who have been risk assessed. The bar represents this measure with a simple RAG rating applied against a target of 95% for full compliance and 90% for partial compliance. The dot, displayed on the secondary y axis, represents the % of the caseload who have a risk assessment AND have been reviewed in the last 12 months.

From this chart you may assume that performance is strong in teams 1,2,4 and 5 and there is small improvement required in team 6. You're eye may naturally be drawn to team 3, showing non compliance and a lower % of patients reviewed in the last 12 months. The trouble with taking this approach to managing performance is that there is important information not visible to you in the visualisation as the data has been aggregated at Team level.

An alternative approach is to visualise individual patient level compliance, but still at team level, so the consumer of the data can then fully understand the inherent risk across all of the patients being cared for by a team. This can be done through using a Tableau Box and Whisker plot, with each dot plot representing an individual team and the axis representing the length of time in months since a patient was last risk assessed and the colour of the dot representing the patient's individual compliance status (with red representing no risk assessment has ever been completed).

Now when viewing the patient level compliance visualisation you may come to a different conclusion as to which team requires the highest level of management support. Team 5 is showing the highest level of patient variation with one patient having been treated for the nearly 4 years since their risk assessment was last reviewed (by the way this example is all based upon fictional data to prove the point!).


It is through personalising data that, often, the maximum benefit of visualisations and analysis is realised. Many performance measures within an organisation relate to people, whether they be consumers, patients or staff, and every consumer of a dashboard is a person so surely it makes sense to connect people with data about people, immediately achieving an emotional connection with the data and, as such, leading to prompter and more effective action.

19 views0 comments

Recent Posts

See All


bottom of page