Talk:Chelsea and Westminster Hospital NHS Foundation Trust

A&Egraph

The a&e graph is extremely misleading as it looks at pure patient numbers. It is captioned as "four hour target" but there is no four hour target for numbers of patients, it is a proportion and the target is 95%, and despite the scary looking graph the target was met in 2018/19Q2. There is no reason given for why the lines are colored according to government. It could equally be before and after the social care bill, or better yet no colour at all.--82.132.233.152 (talk) 14:56, 5 November 2018 (UTC)

No statistician would accept that actual numbers are misleading. These numbers were recorded nationally as part of the monitoring of the four hour target. I am afraid I have no control over the colour. That was done by the person who made the graphs. But it seems to be generally agreed that performance against NHS targets declined markedly - though not in this case - in 2010 and more so from 2015, and that this was a consequence of political decisions. Rathfelder (talk) 18:15, 5 November 2018 (UTC)

Well I'm a statistician and showing raw numbers is ridiculous and gives no information, what if the population doubled? The NHS's own target is 95% less than 4 hours, there is no mention of raw patient numbers. It would be like looking at raw number of deaths from cancer and saying that they'd massively increased when it's just because the population has increased.--Jess97xx (talk) 18:18, 5 November 2018 (UTC)

Also the graphs are tagged as your own work, is that not so?--Jess97xx (talk) 18:20, 5 November 2018 (UTC)

• The graphs were produced by an IT specialist from UCLH who wanted me to post them on his behalf. They are based on NHS England published figures. They use the raw numbers, not %s. As a statistician are you suggesting that the population served by Chelsea and Westminster has doubled? Rathfelder (talk) 18:28, 5 November 2018 (UTC)

Well at C&W there were 22119 A&E attendances in 2005/6 Q2[1] and 87158 attendances in 2017/18 Q2,[2] so the number of attendances more than quadrupled actually. This is not evident at all from the graph. If you want to show various trusts' performance against the NHS standard then you have to look at the percentage waiting over four hours. That is the standard. Otherwise not only is this poor quality, it is WP:OR and should not be on Wikipedia; even though the numbers are publish it presents an analysis and interpretation not in the original source.--Jess97xx (talk) 18:39, 5 November 2018 (UTC)

P.S. just to be clear I'm not disputing that A&E performance has decreased over the past 10 or so years and that is shown in good quality sources e.g. [3] I'm just against bad graphs and misleading stats--Jess97xx (talk) 18:41, 5 November 2018 (UTC)

The graph does not present any analysis at all. It's not original research because it was all in the Health Service Journal. The only originality is collecting the figures and turning them into a graph. The analysis should be in the article. Why not put these details into the article? They don't match what has happened elsewhere. Manchester thought it was in trouble because attendance has been going up at 4% a year, but this is clearly different. And I'm sure the local population hasn't quadrupled. This sounds like organisational changes. Rathfelder (talk) 22:16, 5 November 2018 (UTC)

The graph makes it appear that waiting times have increased dramatically in the past couple of years whereas in fact the trust has managed to keep to its target despite a massive increase in demand, probably because of closures of nearby A&Es such as Hammersmith and Central Middlesex. The colouring is obviously to suggest an association between government and waiting times; they are all captioned with "Waiting longer in A&E", they look more like something from an election leaflet. But worst of all, they are meaningless: Like I said, it's true that A&Es have gotten worse at meeting the 95% target, but the graphs do not show that, and the original figures do not mention the government so it is an added analysis. It is all interesting of course but for Wikipedia we should be quoting secondary sources, not presenting our own analysis. Just out of interest, the figures are available for all trusts, how did you/the IT specialist decide which trusts to make graphs for? And if there is an article in the HSJ about how A&E targets are (not) being met then we can mention that and reference it.--Jess97xx (talk) 22:58, 5 November 2018 (UTC)

We can only make graphs for trusts which have not been merged or reorganised during the period, as figures not available for them. I dont think it's analysis to show a change of government by a change of colour. NB I think you should stop removing material which is referenced. If you think it needs comment or explanation please add what is necessary.Rathfelder (talk) 23:17, 5 November 2018 (UTC)

OK no probs. I removed the nursing times bit because it was almost verbatim from the source, and the bit about CQC reports from 2015 because West Mid wasn't part of chelwest then, but I edited it instead now you put it back. I'm sure there are myriad reasons why A&E targets are not being met including closure of A&Es putting pressure on the remaining ones, changes in staffing levels/recruitment problems, population growth and ageing population, changes in funding, difficulty in getting GP appointments, 'defensive medicine' increasing GP referrals to A&E and many more; these are things that NHSE and CCGs would/should be analysing, and if a report or article exists then we can certainly reference it. The graphs are polished and nicely presented but fatally flawed, and paint a picture that is at best oversimplified and at worst grossly misleading. This applies to graphs for the other trusts too but is particularly egregious in the case of chelwest because, as I said, it met its most recent target, while the graph makes it look like it's A&E is a complete fiasco.--Jess97xx (talk) 23:44, 5 November 2018 (UTC)

It's interesting that the C&W graph looks rather different than most - see https://commons.wikimedia.org/wiki/Category:Health_statistics_in_the_United_Kingdom. But I don't think we can say its misleading. It's basically the official assessment of performance taken from an officially source. It's no more misleading than the target itself. But it may require more explanation. Similarly if attendance has quadrupled that needs comment. Does it include walk-in centres and the like?Rathfelder (talk) 08:09, 6 November 2018 (UTC)

I've already explained why it is misleading. It is not the official assessment of performance, as I explained the target is 95% seen within 4 hours, not "the blue bit's higher up".[4] As for the walk-in centres, there are different types of A&E and the data are split for type 1 and other, at least in some of the later datasets. Writing an in-depth analysis of A&E performance is probably outside the scope of these articles, especially as they're pretty short to begin with. I didn't realize there were so many of those graphs; I'm sorry to say they are all flawed and basically meaningless because they are only looking at the numerator, not the denominator.--Jess97xx (talk) 21:19, 9 November 2018 (UTC)

Why not write about it being misleading at NHS targets? These are the figures NHS England publish after all. Rathfelder (talk) 22:48, 9 November 2018 (UTC)

I think you may have misunderstood, the graph is misleading because it doesn't reflect the targets--Jess97xx (talk) 23:28, 9 November 2018 (UTC)

I understand I think. This data is presented by NHS England in respect of the targets. If its misleading its the publication of the data, not putting it into a graph.Rathfelder (talk) 00:02, 10 November 2018 (UTC)

The data they publish isn't misleading because, for one thing it's just data, it is what it is, and secondly they don't only publish the number of attendances involving waiting over 4 hours, they publish the total number of attendances and further break this down by department type. They also publish the calculated percentage of attendances less than 4 hours, which is the target. The data isn't misleading, putting one part of it into a graph and titling all the graphs "Waiting longer in A&E" then colouring the line by government is misleading. The data actually goes back to 2002, do you know why your friend/colleague chose to start from 2005?--Jess97xx (talk) 08:48, 10 November 2018 (UTC)

Against Deletion

I was encouraged to create this article because of the new article link on List of NHS Trusts. I also used Imperial College Healthcare NHS Trust as a template. The trusts are both in south west London, so it seems reasonable to create an article for Chelsea and Westminster since the Imperial College one has been accepted. The organisation has over 13000 members, and I assume is the employer for several thousand staff. Also see Category:NHS hospital trusts for other similar pages Alexd (talk) 04:09, 14 September 2009 (UTC)