NHS England has, they tell us, a Learning from Complaints Strategy. They are now publishing statistics on their management of user contacts – a development we certainly welcome. What do we learn from the Quarter 12014/15 results? The complaint cause data are presented in big cake slice diagrams with uninformative general headings – Medical; Dental. There is no learning here unless we can get more detail. Comparing the 2013/14 quarter’s data to this latest one tells us nothing – some are more some are less. Why these changes? Should we be worried?
A Treasure Chest of Data?
Another baffling table on page 6 headed Complaint Frequency by Area Team tries to normalise the same two quarters’ complaint data against – what exactly? The untitled column shows a set of figures which in London are 1 in 13,000 varying up to 1 in 45,000 in Durham Darlington and Tees. Is this GP visits or proportions of population complaining? Not only is the column is unlabelled but there is no reference to the data in the text. There is a column showing whether the frequency is up or down. Good? Bad? Meaning can only come with more detail and more cross-tabulation that relates say “Medical” or a subset of the data within that broad category to an area (geographic/ clinical practice) – isolating particular causes for concern using standard epidemological/ statistical approaches as one way of unlocking the treasure chest of data?
Change Over Time
X-tabs are about a snapshot in time. What we also need – and this report begins to supply it – is to see how things move over time with a tracking survey. Such a survey was done for almost 900 complainants whose cases had been closed in quarter 1. Four tables track satisfaction and one the readiness to recommend of the respondents, showing the year to date and figures for each week. The ‘recommend’ data is around 80%+ which is encouraging, but then perhaps people do see them as the only show in town. Two tables look as though they are reporting on the same thing, but the results vary widely for what appears to be much the same question asked of the same people – one table is headed “how satisfied were you with how long it took to resolve your enquiry or complaint?” ( I have corrected the misprint in the heading) and other ‘how satisfied were you with the length of time it took us to respond to your enquiry or complaint?'” I suppose the difference for those keen on textual analysis is between ‘resolve’ and ‘respond’? In a complaint environment, the two words may means very much the same thing to people contacting to complain. The report proudly claims that 34% of complaints have been resolved with 25 working days – rather less impressive when stated as 66% were not resolved in this time frame.
Healthwatch England Challenge
The data has been challenged as presenting too rosy a picture by Healthwatch England. People are still too scared to complain – HWE estimates that over half of those who encounter a problem bother to complain. Complainant data is incomplete unless the experience of the non-complainant or the reasons for non-complaint are understood and this reluctance overcome. While the recommend and satisfaction levels are at first glance respectable, Healthwatch England say that people are still confused as to which of the 70 organisations they could take their complaint to and that many find the system deeply unresponsive e.g 50% did not even receive an apology. The support system for complainants has to be urgently overhauled and the report makes a number of process recommendations such as a facility for the system to ‘pause’ a complaint if those involved are diverted by some other pressing problem in their lives.
But these failings set out by HWE are not reasons to stop collecting data. The information becomes more valuable and better understood over time as people become more comfortable with the numbers. It is part of the process whereby complaint management becomes an integral and accepted part of the way organisations manage contact with their users – complaints are regarded not as dangerous, destructive and an assault on a system which has enough problems anyway – thank you very much – but positive, useful and acted upon. Decent data well presented and used for learning is vital to that process.