The graphic above tells in one way the story of Wales’ coronavirus outbreak.
The purple line uses a seven-day average of coronavirus cases to show how the outbreak rose to a peak on April 9, 17 days after lockdown was announced, when 390 people were diagnosed with Covid-19 on one day.
It then shows how the number of cases being diagnosed dropped more slowly, hit a plateau and then fell again to where we are today – seeing the same levels of cases we first saw before lockdown was announced on March 23.
The yellow line below it shows the number of deaths officially linked to coronavirus by doctors on death certificates following a similar path.
Yet there is also much that this simple line cannot tell us – how the picture has differed across Wales and how many deaths and how many cases have simply never appeared on the official statistics.
That week when Wales saw the greatest loss of life officially linked to this rapacious virus was the week that ended April 17 – meaning those deaths were happening just days after the highest numbers of cases were being diagnosed.
Given what we know about how the length of time, often several weeks, the course of the virus takes, what does this tell us about the state of our awareness of and response to the virus at the time?
It possibly paints a picture of a health and care system lacking testing capacity, grappling to understand what it was up against, desperately trying to do what it could to save lives but lacking the tools to get ahead of the virus and get a picture of what was happening in the community.
In that fateful week – week 16 of the 52-week ONS calendar for recording mortality – 254 of those deaths were in hospitals and 132 in care homes.
Many of those elderly people in care homes would never have been diagnosed. Testing was not offered widely in care homes until May 16. It is also possible that some of those in hospital, at that stage, had not been diagnosed until recently before their death.
Yet the story of coronavirus in Wales is not just a national story. It is also the story of communities and of regions – areas hit at different times and in different ways by the virus.
This short clip is made up maps showing the rate of cases in each council area of Wales on each day since the first patient was officially diagnosed in Swansea. It is a more confusing picture – because it tells a lot of different stories at the same time.
This shows more clearly that slow ebb away of coronavirus cases in Wales’ overall cases masks very different stories in north and south Wales – and markedly different courses in the urban and rural areas of Wales.
Our reporter Laura Clements has already written about the one area of Wales that had very few cases of coronavirus, Ceredigion, and asked what roles geography and the simple but effective contact tracing system set up Ceredigion council played in this.
The focus has been intensely on north Wales in recent weeks. A rise in cases in late May has been followed by the largest numbers of recent deaths being seen in the Betsi Cadwaladr health board.
Increased testing has been suggested as a reason, although this would not explain why there had also been more deaths in recent weeks than in south Wales. An older population and close clusters of care homes also have been proposed – but that would not necessarily explain the timing.
This chart, which compares three of the worst outbreaks in south Wales – Newport, Cardiff and Rhondda Cynon Taf – with Denbighshire, suggests the question we should be asking is what happened in early May to cause case levels to start rising in Denbighshire (and neighbouring areas of north Wales).
Until then, the trajectory of cases had followed a very similar pattern, albeit at a slightly lower level and a few days behind, to the south Wales areas.
Yet this chart suggests that Denbighshire saw a dramatic escalation of cases to hit a much later peak on May 21 – six weeks after the southern regions.
The Welsh Government’s answer in recent weeks has been that testing has played a huge part in the fact that there have appeared to be so many cases in north Wales.
The level of testing in late May and early June certainly far outstrips the number of tests that were being done early on in the outbreak.
People who fear they have the symptoms in north Wales today can far more easily get a test than was the case in March and early April when hospital intensive care units in south Wales say they were a week away from being overrun.
On May 21, the day Denbighshire’s outbreak appears to have peaked, there were 182 tests done in the north Wales council area (which has a population of 95k). In comparison on the day of its peak on April 7, there were 39 tests done in Newport (population 153k).
The number of deaths where Covid-19 was mentioned on the death certificate, as recorded by the Office for National Statistics, in the respective areas is ultimately a far more unbending guide to when the virus was spreading most rapidly.
This chart compares the deaths (in purple) seen in four of the hardest-hit south Wales local authorities (Cardiff, Newport, RCT and Merthyr) with (in yellow) four of the worst-hit north Wales council areas (Wrexham, Denbighshire, Conwy and Gwynedd).
By putting four local authorities together it removes some of the jumps that can be seen from week to week in smaller areas with fewer people. These figures are adjusted for population size to be comparable and show the number of deaths week by week.
This cleary shows that the virus was spreading fastest in south Wales in March and April. It even appears that that both areas saw the greatest spread of the virus at the same time, March to early April as the highest level of deaths in both areas were in the middle of that month.
However the question still remains as to why the number of deaths and cases in north Wales did not fall as they did in the south. The recent outbreaks at the 2 Sisters poultry processing plant on Llangefni and Rowan Foods in Wrexham, as well as the fact deaths are now higher in the north than the south, suggests there is more of the virus circulating in the community in north than south Wales, and the testing is just much more effective at highlighting it.
If there is one other question that the figures put out by the ONS and public health bodies throw up, it’s to ask how bad have been the problems in our care homes.
This scatter chart compares an area’s population with the number of care home deaths since early March. If an area is above an imaginary diagonal line from corner to corner, it’s had unusually high levels of deaths in care homes compared to other local authorities this year.
Merthyr Tydfil, Blaenau Gwent, Monmouthshire, Powys – and also Carmarthenshire, Newport and Rhondda Cynon Taf all stand out.
Yet the heartbreaking scale of the loss of life due to Covid-19 becomes clearer when these numbers are compared to an average year.
To look at just a few of of these, this year Merthyr Tydfil has had 79 deaths in care homes since late February. In a normal year, there would be just 15 deaths in care homes in Merthyr in that time, according to the ONS five-yearly average.
Carmarthenshire has has 223 deaths in care homes since early March where it would normally have just 120. Powys has had 198 deaths in care homes in the last 14 weeks where it would normally have just 86.
And many of these excess care home deaths are almost certainly not recorded on the either the ONS or the Public Health Wales figures as being linked to Covid-19.
According to the ONS, Merthyr has only seen 28 Covid-19 deaths in care homes this year, Powys has only seen 46 and Carmarthenshire 39.
We’re told that Wales has had a true death toll linked to Covid-19 of 2,317, according to the latest figures published by the Office for National Statistics.
Yet is clear that the real toll is higher. There have been 2,592 people die in care homes in Wales since the end of February this year. In a normal year, that would be 1,602.
That is an excess death toll of 910 – and only 656 of them have been linked to Covid-19.
It suggests that looking at care homes alone, the deaths of at least another 254 more people may well be linked to this devastating pandemic, whether through the virus itself or being unable to receive lifesaving care for other conditions.