Health Care Metrics / Commensurability

[overtone]

For the record the US citizens pay roughly twice as much for their healthcare and don't have significantly better outcomes; by some measures- like age at death- they do worse.

The best statistic I have found for quick and dirty comparison is not lifespan, which varies by many factors, but length of survival after diagnosis with one of the few dozen major diseases that can kill you.

 

It works for me because it incorporates, summarizes the net effects of, and corrects for, a whole lot of factors difficult to measure individually, including preventive care (early diagnosis), ability to afford therapy (such as drugs, better food, etc), quality of diagnosis and treatment (too early hospital release, home care etc), tendency to impose worthless or more expensive treatments and therapies for whatever reasons, poverty leading to increased disease vulnerability, and so forth. In other words, it corrects for the common tendency to confuse input with output, expense and sophisticated capability with delivered benefit.

 

And it works because it splits out differences between different cultures and medical care establishments in handling different diseases - stroke in Japan, diabetes in Polynesia, cardiovascular disease in the US, cirrhosis in Russia, cholera prevalence here and there - that can warp comparison stats.

 

So somewhere on the internet, easy to find a couple of years ago, is a color coded chart featuring the 34 First World Country health care systems, compared by the survival times after diagnosis of each of many diseases (the big killers), these times coded as above, at (within one standard deviation), or below, the median of the other 33. A three value code: green, yellow, red. And countries vary by disease - some countries do better with certain diseases, worse with others, etc. Some countries (easily guessed) do better with almost all diseases. {edit in: found a related and similar chart, not quite as pointed but that I believe derives from the same data set, here: http://jama.jamanetwork.com/article.aspx?articleid=1710486 figure 4}

 

The US is arguably, by eyeballing the prevalence of reds and greens and yellows, the worst of the lot, overall. Not just bad - the worst of the 34. One can make an argument for Turkey, but that's its only rival for the bottom performance. Almost no above median ranks, for any disease (iirc maybe one, two?) . Many below median ranks. It's not just that the US doesn't get good value for its extraordinary expenditures, but that it doesn't get First World standard performance from its system. The US system is lousy, way below average, even before the burden of its enormous cost is figured in.

 

And this is something one can blame, pretty much, within reason, on the Republican Party. Of course various rightwing conservative Democrats played a part (especially before 1960) and many circumstances come into play (especially: race) but the primary obstacle to establishing one of the known superior setups for medical care delivery in the US for the past forty years or so has been the Republican Party.

Edited February 25, 2016 by overtone

[CharonY]

Actually survival length after diagnosis is also quite problematic. The issue here is that in different systems there different guidelines for disease screens. For instance, some countries emphasize very early detection of cancers. However, by diagnosing cancer earlier, the the survival time is pushed upward, although the overall trajectory may not change at all. For instance, assume that there is a form of cancer in which everyone, regardless of therapy dies after 10 years. If you diagnose it 5 years into the development your length of survival after diagnosis is 5. However, assume a country implements a more sensitive method which diagnoses at the onset, your survival length may increase to 10 years, without actually affecting the overall outcome (also called lead time bias).

 

There is a bit of discussion what metric would work better and AFAIK there is no firm consensus beside that many metrics, including access should be included. For example, if a significant proportion of the population has no access to early diagnosis, their deaths would affect life expectancy but not survival length. There is also age-adjusted 5-year survival of cancer. By this metric USA is not doing too bad, however, that, too is confounded by the above mentioned issue. I.e. people with access to health care in the US tend to be overdiagnosed. Another measure is the so-called age-standardized avoidable mortality (i.e. deaths that could have been averted by correct prevention/intervention strategies). But this has similar issues with other aggregate measures.

 

What is clear, however, is that cost and inequality in access is something that skews health trajectories quite a bit in the USA compared to other OECD countries. I have not seen the US system performing that bad if cost is taking out of the equation, though.

[iNow]

The challenge, of course, is that cost is the pivotal component of the argument here, and what makes the US so obscenely out of step because... evidence be damned, the free market!! and ick! government, booo!... or something.

[CharonY]

Of course. Most analyses would agree that cost (and therefore, access) is what creates the problems in the US health system. If you have the money, you can get some of the best medical care.

[StringJunky]

The challenge, of course, is that cost is the pivotal component of the argument here, and what makes the US so obscenely out of step because... evidence be damned, the free market!! and ick! government, booo!... or something.

Yes, it's outrageous the difference in drugs on both sides of the pond. Hep C treatment Harvoni costs, on average in bulk, $1125/pill in the US against $641/pill in the UK; they are made in the US. They cost a couple of dollars each to make. Both sides are getting taken to the cleaners but the US population are getting screwed for more. For a new treatment like this, with many patients needing it, the UK NHS budgets, I think, about $35 000 for a year's treatment without too many qualms; $277/pill in this scenario at 90 days treatment length. For the UK the price is over twice the preferred price and four times if the NHS had to pay US prices.

Edited February 25, 2016 by StringJunky

[overtone]

Actually survival length after diagnosis is also quite problematic. The issue here is that in different systems there different guidelines for disease screens. For instance, some countries emphasize very early detection of cancers. However, by diagnosing cancer earlier, the the survival time is pushed upward, although the overall trajectory may not change at all.

But all I am doing is comparing health care systems. Even all else being equal in prognosis, treatment, or amelioration of suffering, (which is true of no common disease) availability of earlier diagnosis is better, a characteristic of a better system - T or F?

 

 

For example, if a significant proportion of the population has no access to early diagnosis, their deaths would affect life expectancy but not survival length.

I'm assuming that a situation in which an entire country's people suffered no net penalty for a major fraction of them lacking access to early diagnosis of major diseases is rare.

 

What is clear, however, is that cost and inequality in access is something that skews health trajectories quite a bit in the USA compared to other OECD countries. I have not seen the US system performing that bad if cost is taking out of the equation, though.

Take cost out of the equation and you could probably get first class boutique medical care in rural Zaire. The dark side of the moon.

 

One strength of the statistic is that it does not depend on the reason for the poor performance.

Edited February 26, 2016 by overtone

[CharonY]

But all I am doing is comparing health care systems. Even all else being equal in prognosis, treatment, or amelioration of suffering, (which is true of no common disease) availability of earlier diagnosis is better, a characteristic of a better system - T or F?

 

 

Not necessarily. There is actually quite a bit of a discussion about whether early diagnosis actually helps in better health outcomes. And maybe somewhat counter-intuitively the answer is: "not necessarily". This is true for certain types of cancers with very slow progression, for example. In some cases early aggressive treatment may even result in slightly worse outcomes. Another example is breast cancer, for which in the US a yearly screening is recommended, whereas in the UK only all three years screens are conducted. In addition they start earlier in the US (40) than the UK (50) This is partially due to some British studies that have shown that additional screening may not increase treatment outcome, but may actually lead to overtreatment. This is often due to false positive detections, which obviously increase with additional screens. The reason why frequent screens were popularized in the US were actually an observed increase in survival from diagnosis with earlier screens. However, cohort studies showed that the survival characteristic of UK and US cancer patients were pretty much the same (the slopes of survival rate curves were basically equal, if you see what I mean). That actually indicates that the early diagnosis on the US has not resulted in better health outcome, but rather the difference between those two curve were mostly (if not entirely) due to lead time bias. I.e. the added survival time is just due to the early diagnosis, but the disease progression remains the same. In these cases there is a window at which detection can lead to improvement but pushing it earlier does not anymore. Of course for other diseases earlier diagnosis can massively improve quality of life and/or survival. Therefore this a question that cannot be answered with true or false without being wrong in some aspect.

 

Just to reiterate, different approaches to diagnosis (even, or maybe especially with same subsequent treatment) can skew the survival time, but may not affect overall health outcome . As such, there is no single metric that is suitable to estimate the overall quality of a health system. And I agree, cost and access are two major components of a functional health system. Best treatment options mean nothing if you cannot have it.

 

Edit: I should add that at least in the area of cancer newer studies have shown that after accounting for lead time bias there have been increased survival, mostly due to improved treatment options, rather than diagnosis. Also in the two given examples (prostrate and breast cancer) the US actually outperforms Western Europe in terms of survival, but at a higher cost. But I believe if lung cancer is added the US loses out (2010ish data, IIRC). As an additional note defining adequate measures for health care quality is the subject of quite a bit of research, especially in the UK and Canada.

Edited February 26, 2016 by CharonY

[John Cuthber]

Meanwhile, the Americans' spend twice as much on healthcare with outcomes that, at best, are not twice as good.

And the Right wing in the UK want to introduce the same system here, which isn't a competent thing to do.

[overtone]

Not necessarily. There is actually quite a bit of a discussion about whether early diagnosis actually helps in better health outcomes. And maybe somewhat counter-intuitively the answer is: "not necessarily". This is true for certain types of cancers with very slow progression, for example.

You're missing the point. 1) The availability of earlier diagnosis of major diseases helps - even dramatically - sometimes. 2) In a health care system in which early diagnosis of major diseases is often not available, this benefit is often not available. 3) There is no major disease in which earlier diagnosis itself harms the patient by shortening their lifespan after diagnosis

 

4) So health care systems can be compared by survival times after diagnosis for major diseases without worrying about whether early diagnosis actually helps, significantly, in every single disease. You won't go far wrong, get an inverted comparison, etc., overall. If your comparison verdict ends up riding on just one or two cancers, it probably isn't significant anyway. We were looking for a rough but indicative statistic for comparing entire systems.

 

Your only concern would be overtreatment, excessive diverted resources - and this would only apply if the more expensive country were the one featuring the meaninglessly longer survival times, which is not the case - really, seriously, not the case - with the US.

Edited February 26, 2016 by overtone

[StringJunky]

 

Not necessarily. There is actually quite a bit of a discussion about whether early diagnosis actually helps in better health outcomes. <snip>

Interesting post, CharonY, never thought of it like that. I think you would agree, that, even though the end date of a person's life doesn't change, the earlier they know the more time they have to make the most of the the time they have left.

[CharonY]

Interesting post, CharonY, never thought of it like that. I think you would agree, that, even though the end date of a person's life doesn't change, the earlier they know the more time they have to make the most of the the time they have left.

 

That is actually at the center of the debate. The fundamental issue is that our diagnostic abilities are beginning to outstrip our ability to make proper prognosis. Quite a few diagnostic tools are so sensitive that one could identify aberrations at at stage where we have no clue whether it will actually become symptomatic. This bit is crucial, as the bias does not apply anymore when the disease becomes symptomatic. I.e. we may find indications of a diseases before it affects the organism and we cannot predict whether it ever will within the lifetime of the patient.

 

Ultimately it depends on how the medical professional and the patient deal with it. For example, some may recommend biopsies which may turn up negatively. Depending on the type of cancer the procedure may be more or less invasive. Likewise the screen itself may can be harmful to some degree. While it is unlikely to be fatal, it will affect quality of life without actually improving the health situation. Or, if positive, one might recommend chemo. This has massive impact on quality of life and may also result in hidden damage e.g. in liver of kidneys, that may prove to be problematic later in life. Yet it is quite possible that without treatment the patient may have no issues at all or at least no worse outcomes. Especially for rare diseases (i.e. where the likelihood of true detection is low), the morbidity of exams and follow-up may outstrip potential benefits. And in cases where there are no benefits the net outcome may be negative, if the quality of life and health of the patient is considered and not only the survival. This depends highly on the type of disease and the patient characteristics of course. But it is often not easy to assess at which point the downsides outstrip the benefits.

 

Quite a few have now accepted that a "wait and see" approach may be appropriate that closely monitors progression and only intervene when indicated. The main issue here is that being diagnosed with cancer can be huge mental burden. Note how you mentioned "the time they have left"? It is quite possible that especially in early diagnosis the cancer may never progress to a lethal state, but depending on how it is conveyed and the mindset of the patient it may still be huge, albeit unnecessary blow. The patient may indeed rather die with cancer than from cancer. This is a real issue (Fowler et al 2006, J Gen Intern Med).

 

And maybe more On-topic, the whole procedure costs money. As we have established and all agree, an ideal health care system should maximize health benefit at an affordable cost. However, unnecessary follow-ups can be costly and, as pointed out, unnecessary. There is a reason why NHS has reduced the screening intervals, whereas in the US there is a push to increase screening (or keep it on a high level).

 

I think it boils down to the question of how potentially harmful these procedures are relative to their benefit and whether we can educate the medical professional and the patients in a way that they can use the information in a positive way.

 

 

You're missing the point. 1) The availability of earlier diagnosis of major diseases helps - even dramatically - sometimes. 2) In a health care system in which early diagnosis of major diseases is often not available, this benefit is often not available. 3) There is no major disease in which earlier diagnosis itself harms the patient by shortening their lifespan after diagnosis

 

4) So health care systems can be compared by survival times after diagnosis for major diseases without worrying about whether early diagnosis actually helps, significantly, in every single disease. You won't go far wrong, get an inverted comparison, etc., overall. If your comparison verdict ends up riding on just one or two cancers, it probably isn't significant anyway. We were looking for a rough but indicative statistic for comparing entire systems.

 

Your only concern would be overtreatment, excessive diverted resources - and this would only apply if the more expensive country were the one featuring the meaninglessly longer survival times, which is not the case - really, seriously, not the case - with the US.

 

1-3) are partially true, but 4 does not follow from them. You have agreed that a good measure of a health care system needs at least two elements. Health outcome and cost (the Zimbabwe example you provided).

 

Now your argument seems to be that the survival time after diagnosis is able to capture these elements. But there are several issues with it.The basic one is that typically that the length of survival depends on many factors and not only the treatment and most studies are not able to follow patients up until their deaths. I would be surprised if there are actually country comparisons of this metric, but I would kindly request them, if you have them available. Actually providing the actual diseases would also help to figure out whether lead time bias has an overall effect. It certainly has for cancer, but I could try to figure it out for the other diseases.

 

That being said, what is more commonly used in studies is an x-year (typically 5 or 10, sometimes longer) survival after diagnosis. My guess is that the data you have seen has actually been adjusted for a number of factors as it is well-recognized that the raw numbers are useless for comparison. For example, the USA actually outperforms Europe in most cancer forms (see Gatta Cancer 2000) using SEER and EUROCARE data. Which runs counter to your (and also my) arguments. A few subsequent studies have pointed out that this is at least partially due to the fact that in the US more aggressive screens are performed. Based on your argument (4) the US should therefore have a better system than most European ones because a) there are more screens and b) the survival rate is higher. But as I mentioned, a) creates the bias that increases b).

 

In other studies such as CONCORD-1 and CONCORD-2 the USA exhibits highest survival rates (typically within the top 5). However, even by using a slightly different metric such as by estimating the ratio of survival found in cancer patients relative to the expected background mortality rate, the authors were able to find systematic inequalities in the US system. The authors still acknowledged that leas time bias may still inflate the outcome. Taking all-cause survival rates overestimates US performance, for example. Note that these studies do not even try to assess health care performance, just to provide a comparative view on cancer.

 

Either way, the main point is that there is still an ongoing discussion on which metrics to use to analyze a health care system and while certain measures perform well to highlight certain aspects, the fall short in others. And again, if we really only took the simple survival rate, the US system would be one of the best in the world. But, as John said:

 

 

 

Meanwhile, the Americans' spend twice as much on healthcare with outcomes that, at best, are not twice as good.

Edited February 26, 2016 by CharonY

[overtone]

That {futile diagnosis} is actually at the center of the debate.

Not mine.

 

 

This bit is crucial, as the bias does not apply anymore when the disease becomes symptomatic. I.e. we may find indications of a diseases before it affects the organism and we cannot predict whether it ever will within the lifetime of the patient.

That is a response and treatment decision, that can be made regardless of diagnosis. There is no intrinsic value in ignorance. One does not improve one's health care, in abstract, via ignorance. There's no intrinsic gain.

 

 

1-3) are partially true, but 4 does not follow from them.

1-3 are simply and completely true, and yes, 4 does follow.

 

I did not agree that cost was a necessary factor in any stage of any comparison of health care systems. I excluded it. As the chosen statistic makes possible, quite elegantly, one can roughly but informatively compare health care systems independently of cost, and I wish to do so as one part or stage in my approach. I wish to do that because I'm usually talking to fellow Americans and one common argument in the US is that we may pay a lot more overall but we get the best care available overall. The fact that we don't, statistically, is telling - in my opinion, and in my experience dealing with Americans. The efficiency argument is muddled, with Americans, unless that initial situation has been made clear.

 

 

 

Now your argument seems to be that the survival time after diagnosis is able to capture these elements. But there are several issues with it.The basic one is that typically that the length of survival depends on many factors and not only the treatment and most studies are not able to follow patients up until their deaths.

Whatever they are, they are summarized in the overall statistic of survival after diagnosis. The goal is comparison of the systems overall, the big picture, not particular treatments.

 

 

For example, the USA actually outperforms Europe in most cancer forms (see Gatta Cancer 2000) using SEER and EUROCARE data.

Quote from the abstract:

 

 

The authors considered 12 cancers: -

- The authors analyzed 738,076 European and 282,398 U.S. patients, whose disease was diagnosed in 1985-1989, obtained from 41 EUROCARE cancer registries in 17 countries and 9 U.S. SEER registries. Relative survival was estimated to correct for competing causes of mortality.

Exactly. Don't do that, was my point. Don't use five year survival rates, and don't use rates "estimated to correct for competing causes of mortality", and don't compare just a few cancers or whatever. Take the big, summary, overall numbers by disease for the whole slew of major diseases and throw them up on a chart like the one I linked, and look at it.

 

 

And again, if we really only took the simple survival rate, the US system would be one of the best in the world.

If the statistic I used was accurately compiled, and my memory sound, (and the chart I could find, and linked, supports it) that is false. That was the point. Edited February 26, 2016 by overtone

[CharonY]

Uhm the article you linked showed a completely different metrics. death rate, years of life lost, years lived with diasbilities, life expectancy at birth and healthy life expectancy. None of which are the metric you provided. In fact, the numbers are derivatives based on expected life expectancy measures. So no, it does not support your point. In fact, the author actually highlights that if one takes survival as a measure the USA outperforms many OECD countries

 

Five-year survival for breast cancer and colorectal cancer are higher in the United States than in many OECD countries.75 Although the number of conditions for which the United States has below-average age-standardized rates is small, it does point out that poor health outcomes in the United States are not preordained. There are potential roles for public health programs, access to high-quality medical care, and policy and legislation in addressing both diseases and risk factors.

 

Thus the USA underperforms if we look at disease and injury burden rather than actual survival. It should be noted that injury, including road injury were among the leading factors of years lost (as well as hear disease, lung cancer pulmonary disease etc. The disadvantage of this metric is that some of the burden would be better addressed by e.g. improving infrastructure and environmental safety rather than the health care system itself. I.e. it is difficult to be sure that just improving health care would change the overall numbers significantly (but that is a different discussion).

 

 

Now the other issue is that you assume that there is a survival metric that goes measured beyond a fixed number of years. But you seem to be under the impression that by lumping the values together somehow the performance would suddenly drop. While it is not impossible (as it may be a function of frequency for example), I would be wondering why a value that that does not allow us to see where a system underperforms is more helpful? In fact, the discussion would be easier if you could point to the study/chart you are thinking about so we do not need guess around so much. You may be right, after all, but I would like to see the data.

[overtone]

Uhm the article you linked showed a completely different metrics. death rate, years of life lost, years lived with diasbilities, life expectancy at birth and healthy life expectancy. None of which are the metric you provided.

Figure 4 (as specified) shows the general pattern in a related statistic, displayed in the same general code. I have been unable to find the original chart - hate relying on memory, but no help for it.

 

 

In fact, the author actually highlights that if one takes survival as a measure the USA outperforms many OECD countries

The "five year survival rates" are always "estimated to correct for competing causes of mortality" and always apply to a few pre-selected cancers. That's one way Americans fool themselves about their health care system. The better statistic helps avoid this.

 

 

Now the other issue is that you assume that there is a survival metric that goes measured beyond a fixed number of years. But you seem to be under the impression that by lumping the values together somehow the performance would suddenly drop. While it is not impossible (as it may be a function of frequency for example), I would be wondering why a value that that does not allow us to see where a system underperforms is more helpful?

Because it makes no advance assumptions about what is and is not to be considered, thereby avoiding a common route to self-deception. Looking at the forest - that will guide your selection of trees to inspect.

 

How about an illustration (real life, my circle): 1) Guy tears his knee up in an accident. He's 63 (private insurance age in the US). The aftercare of the surgery involves release from the original hospital before he can handle getting to an ordinary household bathroom, and therefore three transfers to different facilities for reasons of insurance coverage details - in the confusion, he is sent home without having restored security in the knee, and without professional help. On his way into the front door of his house he falls, and tears up the other knee. Bedridden while his knees heal, his cardiovascular health (already marginal) deteriorates.

2) Gal suffers an intestinal blockage (later indications of gluten intolerance behind the event). She's 68, under Medicare, but it doesn't cover everything - she also has private extended coverage. The surgery involves complications, and recovery after multiple surgeries involves IV hydration and feeding. For insurance reasons she is transferred to two different facilities, in the course of which some key paperwork describing the setup of the IV she was provided (insurance specifications) is lost, in eventual consequence of which she suffers dehydration twice. After the second time a different IV setup is provided, and her two daughters instructed in its use, as she is sent home without a nurse. The daughters turn out to be competent and responsible, so that her next dehydration episode does not come until after the hydration IV has been recalled, which was done on a time table without a home nurse visit or followup. She was not drinking enough water. Her symptoms were unclear in origin, and this dehydration is not diagnosed in an initial doctor visit with a new doc unfamiliar with her case. Diagnosed later, it requires hospitalization, and another round of home IV monitoring by daughters. A year later, another dehydration episode, cause unclear, another misperception by another emergency doctor unfamiliar with her case, this time insistence via daughters gets her into the hospital more quickly. Her heart and kidneys have now been damaged somewhat by multiple dehydrations.

 

The observation is this: both these people are likely to die of heart disease, or possibly car accidents made more likely, with normal survival stats attached to the medical care of this heart disease, car accident trauma, etc A statistic that omits their survival time after their earlier diagnosis and surgeries will miss the health care system issues in their deaths.

Edited February 26, 2016 by overtone

[CharonY]

 

Please give up on the "five year survival rates" which are always "estimated to correct for competing causes of mortality" and always apply to a few pre-selected cancers. That's one way Americans fool themselves about their health care system. The better statistic helps avoid this.

 

No, actually this what large consortia (not US-based) have been using to assess cancer survival. Note that you suggested that survival after diagnosis would be the best measure to do so, with which I disagreed. The only difference is that you mentioned it should be survival length after diagnosis, which to my knowledge is not monitored anywhere. However,, the figure 4 in your link uses a completely different metric. It does not indicate survival but years of life lost (YLL). If that is the one you remember, that is fine. It just means that your initial post was wrong in stating that you were discussing survival after diagnosis and would explain the presented issues here (so there is no need to vigorously defend early diagnostics as it actually opposes your point).

 

YLL itself is not a raw number (such as counting the time between diagnosis and death, for which AFAIK there are no large databases), but the derivative of the number of deaths (due to a given condition) and the standard life expectancy at age of death. A better and more commonly used measure is DALY (disability adjusted life years) which takes YLL and years lived with disability into account. DALY therefore present the overall health burden due to disability, ill health or death. This is part of what I am alluding to, as this measure does not only take death into account, but overall disease burden. Obviously, a good health care system should not only keep people alive, but also healthy. Similarly there is DALE (disability-adjusted life expectancy) which is often used for international comparisons. It indicates the number of years expected to be in full health. This probably comes closest to your idea of a lump value. The disadvantage is that it does not tell you anything about the type of disease or issues that a population may face.

 

This btw. is a common trend. The US perform well in metrics that are based on survival of diseases, but fare less well when it comes to measures that are associated or derived from life expectancy. More detailed studies typically indicate that low-income, Afro-American and Native Americans are most severely affected.

 

Edit: also lousy access to support surrounding pregnancies was also found to be a major factor in the US and the crippling of planned parenthood is definitely not going to make it better.

Edited February 26, 2016 by CharonY

[overtone]

The only difference is that you mentioned it should be survival length after diagnosis, which to my knowledge is not monitored anywhere. However,, the figure 4 in your link uses a completely different metric. It does not indicate survival but years of life lost (YLL).

Yeah, as I mentioned four or five times I was unable to find the original chart, and posted instead a similarly compiled chart using a related statistic exhibiting more or less the same pattern.

 

If you simply wish to question my memory, as I mentioned is a problem here, fine. That's valid. If you don't think the basic pattern - survival time after diagnosis, compiled by disease - is at all visible or discernable or supported by what I did post, and you doubt it exists, Ok. If you don't like the argument, however, altering it as you persist in doing is bootless.

 

 

Similarly there is DALE (disability-adjusted life expectancy) which is often used for international comparisons. It indicates the number of years expected to be in full health. This probably comes closest to your idea of a lump value.

It's not close at all, of course.

 

 

The US perform well in metrics that are based on survival of diseases, but fare less well when it comes to measures that are associated or derived from life expectancy. More detailed studies typically indicate that low-income, Afro-American and Native Americans are most severely affected.

Which indicates problems with the "survival of disease" metrics. Unless Americans are getting sick earlier and more often than others, or only those who haven't been sick are dying young, they are not in fact surviving disease longer in at least the case of some major, common diseases.

Edited February 26, 2016 by overtone

[CharonY]

No I think you are still slightly missing the point. The reason why I am astonished is in almost all studies I am aware of that utilize some form of survival after diagnosis overestimate the performance of the US system. On the other hand, measures that are based on life expectancy (as the one you cited) show a different picture.

 

 

Which indicates problems with the "survival of disease" metrics. Unless Americans are getting sick earlier and more often than others, or only those who haven't been sick are dying young, they are not in fact surviving disease longer in at least the case of some major, common diseases.

 

I am not sure what you mean with "survival of disease metrics". Do you mean the the survival time after diagnosis? That one is problematic and overestimates the US health care system for reasons outlined a couple of times. But I am not sure to which data set you may be referring to (again, the one cited does not measure survival, see below).

 

 

One fundamental issue is also that for many diseases survival time is not tracked.

The reason why YLL is used (in the study you cited) is because it basically categorizes the cause of death (e.g. coronoary disease, lung cancer or car accident) looks at the time of death and estimate how much longer the patient would have lived if he/she hadn't died. For this, data is easier to find (although comparisons in different systems is still difficult).

 

All in all, I think we are in agreement (many posts ago actually) about the actual pattern, but I think the discussion also shows that the type of metric matters and different measures are better in identifying underlying issues than others. There is simply no one-size-fits all solution that accurately ranks performance of health care, but by using different data sets it is possible to assess general trends. And these show that the biggest issue of the US is unequal access to health care (or no access at all).

 

While we are at the topic, YLL and other life expectancy measures tend to weigh children and infant death higher. And actually that is one of the areas where the US system seems to be failing massively as I mentioned before. Various health measures have shown that preterm births and associated health issues as well as infant survival in the US compares very unfavorably compared to other OECD nations. In this context the pro-life movement and associated policies are probably a major factor.

[overtone]

I am not sure what you mean with "survival of disease metrics".

The metrics referred to as "survival of disease" in the post I responded to - five year survival rates after cancer treatment, etc. They mislead, and this is visible in their incongruity with the larger mortality statistics.

 

 

Do you mean the the survival time after diagnosis? That one is problematic and overestimates the US health care system for reasons outlined a couple of times

That one I haven't been able to find and link, but in memory it rates the US very low - median or under among the 34, for almost every cause of death charted. Essentially, the worst or among the worst of the 34. I would hate to think it overestimates the US.

 

The chart pattern I recall is similar to that linked, for a related statistic, but had the US even lower - below Poland, etc.

Edited February 27, 2016 by overtone

[CharonY]

Based on the ones I have seen I am pretty sure that it mus be a one that is based on life-expectancy measures or one of its derivatives. One of the reasons is that in the USA infant mortality is one of the highest in the developed countries this is one of man reports. Of note is that some claim that this is entirely due to different measures of the definition of live birth, but even when corrected the USA is doing pretty bad in OECD standards (just somewhere below Poland).

 

The US has also some of the highest pre-term births. What the USA is doing surprisingly well is giving high-end care to ensure survival of pre-terms, but once that the infants are post ~37 weeks, their survival rate is way lower than in other OECD countries. Other studies have shown that minorities, especially African American are disproportionately affected by this issue. It is no surprise that states with better access to health care (such as Massachusetts and Vermont) have the lowest infant mortality rates, whereas Mississippi and Alabama have more than double of that.