Last week a friend of mine told me of someone she knew who died sadly in a motorcycle accident. At A&E they told the family they'd put it down to covid, and then they wouldn't need a post mortem. As I'm writing this I'm thinking it sounds unbelieveable, but this is someone I know well and trust talking about a family she knows well
Let us suppose we have a death with a diagnosis of a cardiac condition (whose incidence rate is 0.20) and cancer (whose incidence rate is 0.30). Their PROD raw score would be 0.50. Do this summation for all cases then transform the raw score distribution (that ranges from 0.10 to 1.59) into a scaled response of 0 - 100. The idea is to establish a base (pseudo) risk but whether it brings anything new to the table is a different matter for it correlates strongly with the number of diagnoses. What this fiddling has done is reveal under diagnosis in the elderly.
Incidence rate being the same as prevalence among all deaths during this timeframe? Seems like this method weighs more heavily diagnoses that are more common. If so, I could have alopecia contribute to my risk of death more than cancer.
Yes indeed, though the time frame for base calculation was 2017 - 2021 to ensure less common events were represented. Yes indeed, it favours the popular though only 11 diagnostic groupings for major morbidity were used to avoid the alopecia effect.
Partly convinced by this. But maybe you could go by what is listed as the cause of death (excluding covid deaths) and calculate incidence rates for those, then do a similar "lethality fraction".
Could try to validate either method by comparing to country-wide causes of death by age. Though I suspect that older age group issue is real.
Cause of death would be utterly fabulous but that isn't in the EPR dump. In part 2 I explore the relationship between PROD and diagnoses (they're very similar) and discover there is an age related tail-off even in the latter.
The geriatricians I used to work with and for in Gloucestershire as a humble nursing auxiliary back in the early 1980's believed that many of their patients health problems were due to polypharmacy and side effects. They took them off nearly all drugs and placed them under close observations. I've seen some bounce back to life like Lazarus!
Yes, we persuaded my mum to have a statin 'holiday'. One of her friends remarked that she looked 'five years younger'. She was on loads of other drugs, and became sicker and sicker. It breaks my heart looking back :(
Last week a friend of mine told me of someone she knew who died sadly in a motorcycle accident. At A&E they told the family they'd put it down to covid, and then they wouldn't need a post mortem. As I'm writing this I'm thinking it sounds unbelieveable, but this is someone I know well and trust talking about a family she knows well
That puts a whole new spin on matters. Crikey.
Can you give a simple numerical example in calculating PROD. I don't get it. Or if I do get it, it seems like an inappropriate measure.
Let us suppose we have a death with a diagnosis of a cardiac condition (whose incidence rate is 0.20) and cancer (whose incidence rate is 0.30). Their PROD raw score would be 0.50. Do this summation for all cases then transform the raw score distribution (that ranges from 0.10 to 1.59) into a scaled response of 0 - 100. The idea is to establish a base (pseudo) risk but whether it brings anything new to the table is a different matter for it correlates strongly with the number of diagnoses. What this fiddling has done is reveal under diagnosis in the elderly.
Incidence rate being the same as prevalence among all deaths during this timeframe? Seems like this method weighs more heavily diagnoses that are more common. If so, I could have alopecia contribute to my risk of death more than cancer.
Yes indeed, though the time frame for base calculation was 2017 - 2021 to ensure less common events were represented. Yes indeed, it favours the popular though only 11 diagnostic groupings for major morbidity were used to avoid the alopecia effect.
Partly convinced by this. But maybe you could go by what is listed as the cause of death (excluding covid deaths) and calculate incidence rates for those, then do a similar "lethality fraction".
Could try to validate either method by comparing to country-wide causes of death by age. Though I suspect that older age group issue is real.
Cause of death would be utterly fabulous but that isn't in the EPR dump. In part 2 I explore the relationship between PROD and diagnoses (they're very similar) and discover there is an age related tail-off even in the latter.
Maybe it simply meant their prognosis was so poor that they didn't receive a thorough medical investigation - dementia patients for example?
Yep, that makes sense.
The geriatricians I used to work with and for in Gloucestershire as a humble nursing auxiliary back in the early 1980's believed that many of their patients health problems were due to polypharmacy and side effects. They took them off nearly all drugs and placed them under close observations. I've seen some bounce back to life like Lazarus!
Yes, we persuaded my mum to have a statin 'holiday'. One of her friends remarked that she looked 'five years younger'. She was on loads of other drugs, and became sicker and sicker. It breaks my heart looking back :(