QALY puts a normalised value scale on mobility and physical pain, however it fails to include the importance of mental health in the analysis, which distorts the end result of the quality of life. Moreover, the analysis does not take into account the quality of life of people that are affected by, let’s say, a family member becoming paraplegic after treatment. Also, the aggregate can vary, depending on how the survey was taken. Was it a phone survey? If yes, not all implications of the ranking might have been clear to the interviewee. As mentioned in the previous post, people, who do not suffer from any of these described outcomes, are likely to overestimate the negative effect on their quality of life.
All of these flaws can lead to a wrong person getting the treatment. So B should get the treatment only because his QALY over the first year is higher than the one from A? That does not sound fait. What if A is really poor and has been on a waiting list for this treatment for almost a year and patient B is a privately insured patient and has not been on the waiting list at all, but still gets the treatment. That’s when QALY can lead to a wrong decision. It does not take into account the patients’ budget constraint or urgency of treatment. Person B can maybe afford to pay for more than half of the treatment himself, but A has to wait because he does not have the money and his need for treatment is probably more urgent through the long wait.
The QALY is also an average and not an individualised result for the patient. It may not reflect any of the patients qualities, because results vary through the way they have been surveyed (as described above). QALY just seems like a very mathematical way of dealing with issues of personal health, but they probably have to be when resources are scarce.
Tuesday, August 17, 2010
Cost Effectiveness Analysis of Health
This summary of cost effectiveness analysis uses the personal QALY for paraplegia done in class. QALYs (Quality-Adjusted Life- Years) uses two distinct variables for measuring – quality and quantity. In the example of a new surgical procedure for spinal cord injuries, we asked ourselves in class – how many years do I get after the operation? And what quality of life will I have? So firstly, we tried to define and combine them. We set four different labels, describing the outcome after the operation.
• (FR) full recovery
• (FFR) full functional recovery
• (PFR) partial functional recovery
• (P) Paraplegia
We decided to include death, to see how paraplegia compares to death on the value scale. The QALY asks you then to rank these labels on a value scale between 0 - ‘least desirable’ and 100 - ‘most desirable’ to get a picture of your personal preferences.
I found this task very difficult and it is also one of my critiques of this analysis, that for a healthy human being it is very difficult to imagine these different health states and a healthy person might overstate them. Someone that is only partially functional and has chronic pain might not scale it as bad as I would, after adjusting to the new circumstances. However, not getting into too many details, I scaled death the lowest at zero and FR at 100, P is very close to death and FFR and PFR are in the upper half of the scale. After discussing our results in class, I realised that there can be a scenario worse than death. Moreover, we discussed our reasons for scaling the labels the way we did and as a result we found that we did not take into account the spacing between the labels. For example, my spacing between FR and FFR was the same as the space between death and P. Does that mean the intensity of both preferences are the same? No, of course not, did I decide. My preference intensity between FR and FFR is lower after I thought about it again, because they are both the better scenarios that I would like to be in after the surgery, but there is a big difference between death and paraplegia. I think I could still have a good life with the support of others and I rather be paraplegic than dead, hence my preference intensity is greater between those two. We rerun the preference ranking and this time I put death at 5 instead of 0 and I also think about the spacing between the labels. In the QALY, analysers take the spacing into account, they are called trade-offs or in class we called them attributes. When deciding the spacing between labels, you have to think about the trade-offs between pain, mobility, depression, social functions and wealth for example. It made the whole analysis a lot more difficult for me.
However, when we thought about the trade-offs and the spacing as well as the different health states, we wondered about the occurrence of depression that is likely to occur, when someone suffers from chronic pain or partial mobility. We decided to make ‘depression’ a label, so with every previous label you get depression on top. We could not settle on the intensity of the depression, because everyone is different and will react differently to the various health outcomes. So we still ranked the new labels (DFR, DFFR, DPFR, DP) as last time, taking into account the spacing (intensity of preferences) and tried to interpret our results. We included depression to refine our analysis, but I found it more confusing and impossible to settle on the intensity of depression. Hence, I have to say that the inclusion of depression is not simple enough for QALY, when it is done on a large scale with many participants.
The introduction of ‘uncertainties’ did bring a new twist to our analysis though. We tried to determine how to one thinks about the outcomes? What value is assigned by the individual? What happens, if the patient is offered a gamble with a probability of p for the good outcome and a probability of (1-p) for the worse outcome? It all depends on how risk averse the person is or if he is a risk lover (which is not very often the case). It is that, if the person is very risk averse than he/she will go for a high probability for a good outcome and a low probability for the worse outcome. In class, we used ‘death’ as the worse outcome to which we compared all other scenarios. We wanted to find a switch point, where the individual is indifferent between both scenarios, so we are able to plot the results on a graph. Again, you have to think about your reasoning for why you are indifferent between both outcomes at these probabilities. It showed that the good outcomes compared to death had a very high switch point compared to paraplegia and death, which had a very low switch point, because some people might value paraplegia very low, hence take a greater gamble as there is not much to lose. To get a conclusive result, we plotted the result on a scale, where probability was on the vertical axis and the value of the outcome was on the horizontal axis. The result, for me, showed an indifference curve of a risk-averse person, where most switch point lye in the upper right of the graph.
Summarising, QALY is a measure of disease burden after treatment, including both the quality and quantity of life lived. It surveys a randomised population and takes the average to decide how to allocate healthcare resources. The average shows medical professionals or the government what treatment has a lower cost to QALY saved ratio, which determines who is more suitable for the treatment. The QALY measure has some flaws, which I will discuss in the second part...
• (FR) full recovery
• (FFR) full functional recovery
• (PFR) partial functional recovery
• (P) Paraplegia
We decided to include death, to see how paraplegia compares to death on the value scale. The QALY asks you then to rank these labels on a value scale between 0 - ‘least desirable’ and 100 - ‘most desirable’ to get a picture of your personal preferences.
I found this task very difficult and it is also one of my critiques of this analysis, that for a healthy human being it is very difficult to imagine these different health states and a healthy person might overstate them. Someone that is only partially functional and has chronic pain might not scale it as bad as I would, after adjusting to the new circumstances. However, not getting into too many details, I scaled death the lowest at zero and FR at 100, P is very close to death and FFR and PFR are in the upper half of the scale. After discussing our results in class, I realised that there can be a scenario worse than death. Moreover, we discussed our reasons for scaling the labels the way we did and as a result we found that we did not take into account the spacing between the labels. For example, my spacing between FR and FFR was the same as the space between death and P. Does that mean the intensity of both preferences are the same? No, of course not, did I decide. My preference intensity between FR and FFR is lower after I thought about it again, because they are both the better scenarios that I would like to be in after the surgery, but there is a big difference between death and paraplegia. I think I could still have a good life with the support of others and I rather be paraplegic than dead, hence my preference intensity is greater between those two. We rerun the preference ranking and this time I put death at 5 instead of 0 and I also think about the spacing between the labels. In the QALY, analysers take the spacing into account, they are called trade-offs or in class we called them attributes. When deciding the spacing between labels, you have to think about the trade-offs between pain, mobility, depression, social functions and wealth for example. It made the whole analysis a lot more difficult for me.
However, when we thought about the trade-offs and the spacing as well as the different health states, we wondered about the occurrence of depression that is likely to occur, when someone suffers from chronic pain or partial mobility. We decided to make ‘depression’ a label, so with every previous label you get depression on top. We could not settle on the intensity of the depression, because everyone is different and will react differently to the various health outcomes. So we still ranked the new labels (DFR, DFFR, DPFR, DP) as last time, taking into account the spacing (intensity of preferences) and tried to interpret our results. We included depression to refine our analysis, but I found it more confusing and impossible to settle on the intensity of depression. Hence, I have to say that the inclusion of depression is not simple enough for QALY, when it is done on a large scale with many participants.
The introduction of ‘uncertainties’ did bring a new twist to our analysis though. We tried to determine how to one thinks about the outcomes? What value is assigned by the individual? What happens, if the patient is offered a gamble with a probability of p for the good outcome and a probability of (1-p) for the worse outcome? It all depends on how risk averse the person is or if he is a risk lover (which is not very often the case). It is that, if the person is very risk averse than he/she will go for a high probability for a good outcome and a low probability for the worse outcome. In class, we used ‘death’ as the worse outcome to which we compared all other scenarios. We wanted to find a switch point, where the individual is indifferent between both scenarios, so we are able to plot the results on a graph. Again, you have to think about your reasoning for why you are indifferent between both outcomes at these probabilities. It showed that the good outcomes compared to death had a very high switch point compared to paraplegia and death, which had a very low switch point, because some people might value paraplegia very low, hence take a greater gamble as there is not much to lose. To get a conclusive result, we plotted the result on a scale, where probability was on the vertical axis and the value of the outcome was on the horizontal axis. The result, for me, showed an indifference curve of a risk-averse person, where most switch point lye in the upper right of the graph.
Summarising, QALY is a measure of disease burden after treatment, including both the quality and quantity of life lived. It surveys a randomised population and takes the average to decide how to allocate healthcare resources. The average shows medical professionals or the government what treatment has a lower cost to QALY saved ratio, which determines who is more suitable for the treatment. The QALY measure has some flaws, which I will discuss in the second part...
Thursday, July 22, 2010
Today's presentation
Hi everyone,
This is a brief summary of my presentation from today:
1950: Streptomycin, Smoking and Sir Austin Bradford Hill
Two unrelated events in 1950(discovery of a cure for tuberculosis and the discovery of smoking as a cause of lung cancer) seperated medicine's past from the coming future.
Before 1950, medical knowledge of causes and treatments of diseases was cumulative wisdom aquired through practice. Austin Bradford Hill changed all that and relied on statistical methods of proof, which became the sole 'scientific truth' after he proved the statistical method on the tuberculosis and lung cancer case. His incentive to run a randomised clinical trial on streptomycin and PAS was his own survival from tuberculosis. The objective is clear, he wanted to do a simple experiment where the efficiacy of a remedy is tested by comparing the outcome in those given it with that in a similar group who are not. Patients with similar age and disease pattern were located to the treatment or control group at random. Bradford Hill's findings were that a combination of streptomycin and PAS led to a 80% survival rate. Since the randomised Controlled Trial (RCT) became a standard way of evaluating new drugs.
He gave a epidemiological proof for the case whether smoking causes lung cancer. This proof involved collecting vital statistics over a certain period and observing the effects via a survey. He found that the dose relationship played a significant role as patients who consumed a higher dose of tobacco had a greater respond to lung cancer.
So, the RCT became the 'dominant' discourse of post-war medicine.
My Questions so far are:
-Is the 'objectivity' of the clinical trial or the RCT really better than the clinical wisdom of the doctor? do we really need these trials to test everything?
- as said in class: how do we randomise these trials? is it ethical?
so feel free to comment or answer, hav a gud weekend everyone...
This is a brief summary of my presentation from today:
1950: Streptomycin, Smoking and Sir Austin Bradford Hill
Two unrelated events in 1950(discovery of a cure for tuberculosis and the discovery of smoking as a cause of lung cancer) seperated medicine's past from the coming future.
Before 1950, medical knowledge of causes and treatments of diseases was cumulative wisdom aquired through practice. Austin Bradford Hill changed all that and relied on statistical methods of proof, which became the sole 'scientific truth' after he proved the statistical method on the tuberculosis and lung cancer case. His incentive to run a randomised clinical trial on streptomycin and PAS was his own survival from tuberculosis. The objective is clear, he wanted to do a simple experiment where the efficiacy of a remedy is tested by comparing the outcome in those given it with that in a similar group who are not. Patients with similar age and disease pattern were located to the treatment or control group at random. Bradford Hill's findings were that a combination of streptomycin and PAS led to a 80% survival rate. Since the randomised Controlled Trial (RCT) became a standard way of evaluating new drugs.
He gave a epidemiological proof for the case whether smoking causes lung cancer. This proof involved collecting vital statistics over a certain period and observing the effects via a survey. He found that the dose relationship played a significant role as patients who consumed a higher dose of tobacco had a greater respond to lung cancer.
So, the RCT became the 'dominant' discourse of post-war medicine.
My Questions so far are:
-Is the 'objectivity' of the clinical trial or the RCT really better than the clinical wisdom of the doctor? do we really need these trials to test everything?
- as said in class: how do we randomise these trials? is it ethical?
so feel free to comment or answer, hav a gud weekend everyone...
Tuesday, July 20, 2010
Subscribe to:
Posts (Atom)