Oops. And sorry. I’ve been checking over my figures and found that I made a mistake when adjusting the size of the English waiting list for Trust mergers. It turns out the waiting list has grown more significantly than I originally thought, and is now noticeably higher than in recent years.

I’ll expand on the correction in a moment, but first here is the new estimate of the adjusted waiting list:

As you can see, the size of the English waiting list rose sharply in March and is now seasonally bigger than in recent years. The size of the waiting list is significant because it is an early warning of waiting times pressures building up in the NHS. However it is also worth noting that the increase remains comparable to the amount of variation we have already seen in recent years, so as things stand this is an indicator of rising pressure and not an omen of doom.

For comparison, the unadjusted trend is this:

What is the difference between these two charts?

The unadjusted chart shows the raw figures submitted to the Department of Health (number of incomplete pathways, commissioner basis). The adjusted chart is an attempt to strip out the effects of step changes, for instance when a Trust starts including some waiting list patients who have previously been missing from its data, or when a Trust takes a reporting break to fix a serious problem with its data.

Ideally all the data would be perfect all the time, but in practice it is difficult for NHS Trusts to count all their incomplete pathways, because patients are being managed on so many different IT systems. So it is good that Trusts are continuously improving their data recording (even though it causes awkward shifts in the time trends), and this process will continue because there is still some way to go.

The method I used to detect and strip out these step changes has not changed. What has changed is the further adjustment that needs to be made for Trust abolitions and mergers.

For instance, let’s take the giant merger that formed Barts Health NHS Trust in April 2012.

The three previous Trusts (Barts and The London NHS Trust, Newham University Hospital NHS Trust and Whipps Cross University Hospital NHS Trust) all disappeared from the data series after reporting at the end of March 2012. Only the new Trust exists in the current published waiting list data.

When it is detecting step-changes, my software uses the current data as its to-do list when working back to detect step changes. So it automatically detects the new Trust’s appearance in April 2012 and adjusts for it in previous months, but it doesn’t detect the three prior Trusts that aren’t on its to-do list. So we need to make a further adjustment for the three old Trusts to balance the appearance of the new one. The problem was that I had missed some former Trusts out, and in this example I had taken the old Barts and The London into account, but not Whipp’s Cross or Newham. Hence the need to correct the method now.

So I have trawled back through all the old data, to see which Trusts no longer exist. Those that have disappeared in the last three years (and which are all now included in the calculations) are:

• code RAE: Bradford Teaching Hospitals NHS Foundation Trust (not abolished but on a prolonged ‘reporting break’; last reported in Sept 2012 with 16,332 patients on the waiting list);
• code RBF: Nuffield Orthopaedic Centre NHS Trust (merged into Oxford Radcliffe NHS Trust to form Oxford University Hospitals NHS Trust; last reported in October 2011 with 3,529 patients on the waiting list);
• code RCC: Scarborough and North East Yorkshire Health Care NHS Trust (merged into York Teaching Hospital NHS Foundation Trust; last reported in June 2012 with 8,386 patients on the waiting list);
• code RGC: Whipps Cross University Hospital NHS Trust (merged to form Barts Health NHS Trust; last reported in March 2012 with 13,343 patients on the waiting list);
• code RM4: Trafford Healthcare NHS Trust (merged into Central Manchester University Hospitals NHS Foundation Trust; last reported in March 2012 with 5,943 patients on the waiting list);
• code RN1: Winchester and Eastleigh Healthcare NHS Trust (merged into Basingstoke & North Hampshire NHS Foundation Trust; last reported in December 2011 with 13,245 patients on the waiting list);
• code RNH: Newham University Hospital NHS Trust (merged to form Barts Health NHS Trust; last reported in March 2012 with 7,475 patients on the waiting list);
• code RNJ: Barts and The London NHS Trust (merged to form Barts Health NHS Trust; last reported in March 2012 with 25,994 patients on the waiting list).

The effect of including all these former Trusts has been to change the adjusted chart from the version I reported last week:

to the new estimate of the adjusted position:

Hope it’s all fixed now. Apologies again, and do let me know if you think I am still missing something.