Weekly Trends in ED Admissions
Lessons from an undisclosed NHS Trust
I’ve now boiled the non-COVID adult (18 - 99y) admissions data down into 24 diagnostic groupings so we can eyeball trends over the period Jan - Sep 2021. A total of 208,791 admissions were available for analysis offering a total of 226,506 diagnoses (mean rate = 1.08 diagnoses per admission). No distinction has been made between unvaccinated and vaccinated cases since this leads to difficulties in accounting for age-related phased rollout. If vaccinations are causing certain adverse reactions then over the period Jan - Sep 2021 we should see incidence of these rising within the general admission population. Obviously, there may be other reasons why certain conditions rise and fall over time and this must be borne in mind; in particular we need to note the impact of lockdown policies and consequent delays to both diagnostics and treatments as the NHS effectively went into mothballs. Deterioration of the nation’s general health will be a fact of life for decades to come.
Significant Slopes Only!
I am going to start the ball rolling by considering only those weekly time series that managed to achieve a statistically significant positive trend over the period 2021/w1 to 2021/w38. The stats test here was a generalised linear model (GLM) enabling use of a Poisson error structure to handle count data. Many folk would opt for using the ubiquitous ordinary least squares regression (OLSR) built into packages like Excel and whilst this would do a decent job I was brought up ‘proper’ by hardcore statisticians who would have glared at me for stooping to OLSR! GLM it is then, and here are the 10 series that made the grade…
From this table we can see that the lion’s share of ED work consists of handling physical trauma of various sorts (n=72,248) and that the highly statistically significant (p<0.001) trend indicates a rise in these cases at a rate of 30 per week over the period Jan - Sep. This is to be expected for a nation coming out of lockdown and returning to a surreal ‘normal’ - people are going to bump into things. Here’s the actual series, revealing that folk had pretty much got back to normal round about week 24 of 2021…
Next in line for admission volume is the strange category of ‘Null admission’ (n=26,854). These are cases for which no abnormality was detected at triage or instances where the patient simply walked out before being seen. The rate suggests an increase of 11.6 such cases per week. Herewith the corresponding series that shows no sign of let up; this is not the sort of inefficiency we want right now and we may wonder if these are people who would normally have gone to see a GP once upon a time…
Third in terms of departmental workload is general inflammation at n=20,675. This is a wide category for sure, including more or less anything ending with ‘itis’. We are talking any part of the body flaring up and being a general nuisance. These cases are rising at a rate of 6.1 per week and it is worth asking why. Is this the sign of a nation in need of decent primary care or the sign of something dodgy happening with vaccines? Whatever is causing this trend seems to have reached an asymptotic level; that is, everyone who was going to get an ‘itis’ has pretty much now got it and on medication…
Fourth in terms of departmental workload are infections and infestations at n=13,359. Please note that this series does not include SARS-COV-2 infection (this is another story for another day) but it does include normal SARS and influenza along with any other viral, bacterial or parasitic infection. Examples are bacterial meningitis, gonorrhoea, legionella pneumonia, genital herpes simplex, tetanus and sickle cell disease. These cases are rising at a rate of 4.8 per week and the graph suggests this trend stayed strong right up to week 28. Is this a result of people mingling more, a sign of declining general health post-lockdown or some nasty knock-on effect of vaccination?…
Fifth in terms of departmental workload is CNS (central nervous system) at n=8,208. This category covers a wide range of physical disorders and conditions including dementia, MS (multiple sclerosis) and Parkinsonism. Bell’s palsy is in there as is sciatica, epilepsy and vasovagal syncope. Quite why this cluster of diagnoses are rising at a general rate of 1.3 cases per week is not clear but it simply may be people coming out of the woodwork after suffering in silence and fear. This would explain the hump-shaped distribution as things settled down in a freer society. The Behavioural Insights Team have a lot to answer for…
Sixth in terms of departmental workload are (non-COVID) acute respiratory cases at n=5,020, these rising at a steady overall rate of 1.3 cases per week. The slide for this series is pretty darn interesting for it shows an initial seasonal dip (as we may expect) before launching into a rather un-seasonal climb from week 12 to week 26. What on Earth was driving acute respiratory conditions through summer that was not COVID?
Seventh in terms of departmental workload is self harm at n=4,738 and a slight general rise of 0.6 cases per week. This is an emotive subject that has yet to gain traction in a cowardly press. Nailing a practical definition within the context of ED admission diagnosis is not easy (apart from the obvious of overdose). The following diagnoses were considered indicators of actual or potential self harm - I wanted to capture those at a loss in life as well as those driven to suicide:
From the slide below we see that the the initial surge has thankfully declined but, as yet, I have no historic records with which to determine how ‘normal’ rates of 120 - 140 admissions per week are for this NHS Trust…
Eighth in terms of departmental workload is clotting & embolism at n=4,306, with a slight general rise of 0.6 cases per week. This category included conditions such as acute pulmonary oedema, deep venous thrombosis (DVT), transient cerebral ischaemia, stroke and pulmonary embolism (PE). The slide below reveals another of those humped distributions. I find it most odd that these cases would rise between week 1 and week 26 and my guess is that this series is tracking the course of adverse reactions to vaccinations in some people. Saturation will take place when all those who are going to be affected are affected and receiving treatment/medication…
Ninth in terms of departmental workload is cardiac arrhythmia at n=2,295, with a slight general rise of 0.3 cases per week. The slide for this tells a different type of story with an initial rise in cases that quickly plateaued. An interesting one given reports in the mainstream and alternative media but not the most robust of results with the weakest p-value of p=0.028…
Tenth in terms of departmental workload is sepsis at n=1,946, with a slight general rise of 0.4 cases per week. Sepsis includes neutropenic sepsis (aka febrile neutropenia), puerperal sepsis (a condition that may follow childbirth), general sepsis and septic shock. Sepsis is a life-threatening condition with risk factors of being very young or old with an impoverished immune system. The slide below reveals a worrying positive trend that doesn’t seem to be abating, with an explosion of cases from week 26 (w/b 26th Jun - week 31 (w/b 31 Jul). A most worrisome trend unless some clinician can explain why incidence of sepsis should be on the increase (p<0.001)…














I am a fairly new subscriber and absolutely not a statistician or mathematician. That said I’m a bit anal with detail. The conflicting language used by government is one big annoyance to me, the other their ropey and questionable data. I’m really enjoying reading your articles, every day is a school day as they say. To my untrained eye there does seem to be something going on very soon after the big vaccination push-of 2021. Enough to make me raise an eyebrow. Thank you for your efforts and making this area easier for me to navigate.
Nice!
Personally I think all those positive correlations are linked to the vaccination. That also includes the physical and null admissions. Physical because people are losing consciousness, fainted, falling (that is linked to vaccination), and hurting themselves in a different way. I have somebody who I know as an example.
And the rest of admissions are also largely the consequences of vaccination. I don’t buy completely the factor of transitioning from lockdown to a normal life. My estimate is that up to 30% of people may have different level of adverse effects. That inevitably must be reflected in the ER admission. I suggest try to explore this angle.