Friday, September 29, 2006
Online Biostatistics Course
An online biostatistics course from Tufts University, StatisticalPractice.com
Wednesday, September 20, 2006
The maths may be simple but intuition is more use
By John Kay
Published: August 28 2006 20:03
Our ancestors gambled around camp-fires on the savannahs thousands of years ago. The ancient Greeks had games of chance, embryonic financial markets and some of the finest mathematicians who have ever lived. But they did not discover the rather simple arithmetic of probability and the methods we use today are no more than 300 years old.
Even today, many people struggle to learn probability theory and experts make mistakes in applying it. This can have disastrous results, as in the imprisonment of innocent mothers whose children had been victims of unlikely, but not that unlikely, sequences of accidental death. Stephen Jay Gould is only one of many science writers to have observed that human minds are not well adapted to dealing with issues of probability.
The two-box problem, which I described last week, is one in which a knowledge of probabilities seems to be a hindrance rather than a help. The puzzle offers you the choice of two boxes, one containing more money than the other. Once you have made a decision, you are shown what is in your preferred box. Do you stick with your original choice, or switch?
You have so little information that there seems no rational basis for decision. Yet the dilemma is real. In our personal lives, in business and finance, in employment and in house-hunting, we repeatedly encounter two-box problems – should we stick with what we know or switch to something about which we know much less?
The extraordinary feature of the two-box problem is that there is a strategy that seems better than always switching or always sticking, one that beats random choice even in a situation of almost total ignorance.
Before the game starts, focus on a sum of money. It does not matter what the amount is – say, £100. The “threshold strategy” is to switch if the box you choose contains less than £100 and to stick if it contains more. The threshold strategy gives you a better-than-even chance of getting the larger sum. It does so for any value of the threshold you choose.
If both boxes have less than £100, or more than £100, then the probability that you get the larger sum from your random choice remains one-half. But if one box has less than £100 and the other has more than £100, adopting the threshold strategy makes sure you get the larger sum. Since there is at least a possibility that the amounts in the boxes lie in this range, the threshold strategy must increase your chance of winning.
So how should you set the threshold? There are two criteria. How likely is it that the box with the smaller amount of money has an amount below, but not too much below, the figure you set? There may not be much benefit in setting a very high or very low threshold. So if you have some idea, however vague, about the range of possible contents, you can tweak the threshold strategy to your advantage. Second, choose a threshold in the range of sums of money that would make a real difference to you: if £20,000 would not transform your life but £50,000 would, then go for £50,000. The range from £25,000 to £50,000 is the range in which the benefit from switching might be greatest.
The curious feature of the threshold strategy is that the maths is surprising but the intuition is familiar. In real-life search problems the principle of “be realistic but look for something that will make a difference” represents typical behaviour. In the version of the two-box problem I described last week, where you had the additional information that one box contained twice as much money as the other, there seemed always to be an argument for switching; the potential gain is always twice the seemingly equally likely potential loss. But this conclusion is wrong. Once more, our intuition runs ahead of our mathematical understanding.
Probability theory works well for a limited class of – mostly artificial – problems, such as coin tossing and roulette. But the real world is much more open-ended and there is usually fundamental uncertainty about both the nature of the outcomes and the process that gives rise to them. Perhaps the reason we do not use probability theory much is that it is not all that useful.
http://www.johnkay.com
Source: Financial Times
http://www.ft.com/cms/s/0171c18a-36b5-11db-89d6-0000779e2340.html
Published: August 28 2006 20:03
Our ancestors gambled around camp-fires on the savannahs thousands of years ago. The ancient Greeks had games of chance, embryonic financial markets and some of the finest mathematicians who have ever lived. But they did not discover the rather simple arithmetic of probability and the methods we use today are no more than 300 years old.
Even today, many people struggle to learn probability theory and experts make mistakes in applying it. This can have disastrous results, as in the imprisonment of innocent mothers whose children had been victims of unlikely, but not that unlikely, sequences of accidental death. Stephen Jay Gould is only one of many science writers to have observed that human minds are not well adapted to dealing with issues of probability.
The two-box problem, which I described last week, is one in which a knowledge of probabilities seems to be a hindrance rather than a help. The puzzle offers you the choice of two boxes, one containing more money than the other. Once you have made a decision, you are shown what is in your preferred box. Do you stick with your original choice, or switch?
You have so little information that there seems no rational basis for decision. Yet the dilemma is real. In our personal lives, in business and finance, in employment and in house-hunting, we repeatedly encounter two-box problems – should we stick with what we know or switch to something about which we know much less?
The extraordinary feature of the two-box problem is that there is a strategy that seems better than always switching or always sticking, one that beats random choice even in a situation of almost total ignorance.
Before the game starts, focus on a sum of money. It does not matter what the amount is – say, £100. The “threshold strategy” is to switch if the box you choose contains less than £100 and to stick if it contains more. The threshold strategy gives you a better-than-even chance of getting the larger sum. It does so for any value of the threshold you choose.
If both boxes have less than £100, or more than £100, then the probability that you get the larger sum from your random choice remains one-half. But if one box has less than £100 and the other has more than £100, adopting the threshold strategy makes sure you get the larger sum. Since there is at least a possibility that the amounts in the boxes lie in this range, the threshold strategy must increase your chance of winning.
So how should you set the threshold? There are two criteria. How likely is it that the box with the smaller amount of money has an amount below, but not too much below, the figure you set? There may not be much benefit in setting a very high or very low threshold. So if you have some idea, however vague, about the range of possible contents, you can tweak the threshold strategy to your advantage. Second, choose a threshold in the range of sums of money that would make a real difference to you: if £20,000 would not transform your life but £50,000 would, then go for £50,000. The range from £25,000 to £50,000 is the range in which the benefit from switching might be greatest.
The curious feature of the threshold strategy is that the maths is surprising but the intuition is familiar. In real-life search problems the principle of “be realistic but look for something that will make a difference” represents typical behaviour. In the version of the two-box problem I described last week, where you had the additional information that one box contained twice as much money as the other, there seemed always to be an argument for switching; the potential gain is always twice the seemingly equally likely potential loss. But this conclusion is wrong. Once more, our intuition runs ahead of our mathematical understanding.
Probability theory works well for a limited class of – mostly artificial – problems, such as coin tossing and roulette. But the real world is much more open-ended and there is usually fundamental uncertainty about both the nature of the outcomes and the process that gives rise to them. Perhaps the reason we do not use probability theory much is that it is not all that useful.
http://www.johnkay.com
Source: Financial Times
http://www.ft.com/cms/s/0171c18a-36b5-11db-89d6-0000779e2340.html
Saturday, September 09, 2006
Drugmakers: A Dose Of Reality : To become a blockbuster, a drug will have to prove to insurers that it's cost-effective
JUNE 19, 2006
By David Balekdjian and Michael J. Russo
An increasing number of new drugs since 2003 have fallen drastically short of sales expectations, garnering as little as 4% of the billions they were expected to take in. These dire results have blindsided drugmakers and spooked investors. Why have so many new medicines, widely projected to be blockbusters, fizzled at launch? The short answer: managed care.
It's no secret that the clout of managed-care providers (insurance companies and pharmacy benefit managers who use the power of the purse to hold down health costs) has been growing for almost a decade. But many pharmaceutical and biotech companies failed to recognize just how much that transformation challenged their own business model. Now they're scrambling as managed care is increasingly determining the winners -- and the big losers -- in pharma land.
The strategy used by managed-care players, called outcomes-based access (OBA), is central to understanding the frequent disconnect between industry predictions and which drugs actually end up making money. Make no mistake, pharma remains a growth sector. But drugmakers and analysts must adapt to current industry realities if they hope ever to return to the days of predictable profit growth.
OBA IS A POWERFUL APPROACH that managed-care companies use to decide which drugs to cover. The impact of this new paradigm cannot be overstated. The drugs that managed care favors enjoy fewer access restrictions and lower co-pays, while nonpreferred drugs are difficult, expensive, or sometimes even impossible for a patient to get. When Bruckner Group's annual Payer Trends Study first characterized OBA in 2001, most drugmakers considered OBA a passing fad. Today it's the industry standard.
The basis of OBA is a simple question: Which medicines give the health-care system the most bang for the buck? The insurance industry seeks to find drugs that produce the best results at the lowest cost. The process for determining this is more complicated than just comparing prices. For example, drugs that let patients avoid expensive surgeries or emergency room visits get a value bonus, while those with costly side effects are penalized. So drugmakers need to find every last penny their drug saves the insurance system, since only those found to be most valuable to payers are likely to become hits.
That's a significantly higher hurdle than drug manufacturers historically have faced. Indeed, many pharma managers were reared in an era when cost wasn't their primary worry, and there's continuing public concern that insurers have exceeded their bounds by overriding physicians' patient-care decisions. Yet today health insurers put the burden of proof on pharmaceutical companies to show in dollars and cents how their new drugs are cost-effective to payers. Nifty science and catchy advertising mean little to payers if they don't produce cost savings or take treatment to a higher level. Improvements that provide little more than patient convenience at a higher price, like reducing the dose from twice a day to once a day, are valuable only if a drugmaker can prove that once-a-day treatment means fewer skipped doses and improved patient health.
"If a health plan is spending $100 to $1,000 or more per month for a drug, we need to know that we are getting value from the expenditure," explains Dr. Robert Seidman, chief pharmacy officer at managed-care leader WellPoint Inc.
Drugmakers caught unaware of or unprepared for OBA are suffering. Expected blockbusters like Biogen Idec's Amevive (a psoriasis treatment offering new science at a hefty price without better results) and Pfizer's Caduet (a combo-pill of two existing drugs) have become major disappointments because of OBA.
Sadly, few drugmakers or investors routinely incorporate OBA considerations into their decisions. In Bruckner Group's 2006 OBA Manufacturer Index, which measures the ability to bring to market drugs with high health-care value, only 3 of the top 20 pharma and biotech companies get a grade of C+ or better. That's why drugmakers must keep bolstering their products' ability to either significantly improve care or save the system money, or both. Doing so will lead to managed-care acceptance and larger revenues. Ignoring these increasingly powerful customers, on the other hand, is the quickest route to financial illness.
--
David Balekdjian and Michael J. Russo are partners at Bruckner Group, which provides strategic advice to pharmaceutical and biotech company executives.
Source: BusinessWeek
http://www.businessweek.com/magazine/content/06_25/b3989121.htm
Chinese version: BusinessWeek China
http://www.businessweekchina.com/article_p.php?BusinessweekID=993
By David Balekdjian and Michael J. Russo
An increasing number of new drugs since 2003 have fallen drastically short of sales expectations, garnering as little as 4% of the billions they were expected to take in. These dire results have blindsided drugmakers and spooked investors. Why have so many new medicines, widely projected to be blockbusters, fizzled at launch? The short answer: managed care.
It's no secret that the clout of managed-care providers (insurance companies and pharmacy benefit managers who use the power of the purse to hold down health costs) has been growing for almost a decade. But many pharmaceutical and biotech companies failed to recognize just how much that transformation challenged their own business model. Now they're scrambling as managed care is increasingly determining the winners -- and the big losers -- in pharma land.
The strategy used by managed-care players, called outcomes-based access (OBA), is central to understanding the frequent disconnect between industry predictions and which drugs actually end up making money. Make no mistake, pharma remains a growth sector. But drugmakers and analysts must adapt to current industry realities if they hope ever to return to the days of predictable profit growth.
OBA IS A POWERFUL APPROACH that managed-care companies use to decide which drugs to cover. The impact of this new paradigm cannot be overstated. The drugs that managed care favors enjoy fewer access restrictions and lower co-pays, while nonpreferred drugs are difficult, expensive, or sometimes even impossible for a patient to get. When Bruckner Group's annual Payer Trends Study first characterized OBA in 2001, most drugmakers considered OBA a passing fad. Today it's the industry standard.
The basis of OBA is a simple question: Which medicines give the health-care system the most bang for the buck? The insurance industry seeks to find drugs that produce the best results at the lowest cost. The process for determining this is more complicated than just comparing prices. For example, drugs that let patients avoid expensive surgeries or emergency room visits get a value bonus, while those with costly side effects are penalized. So drugmakers need to find every last penny their drug saves the insurance system, since only those found to be most valuable to payers are likely to become hits.
That's a significantly higher hurdle than drug manufacturers historically have faced. Indeed, many pharma managers were reared in an era when cost wasn't their primary worry, and there's continuing public concern that insurers have exceeded their bounds by overriding physicians' patient-care decisions. Yet today health insurers put the burden of proof on pharmaceutical companies to show in dollars and cents how their new drugs are cost-effective to payers. Nifty science and catchy advertising mean little to payers if they don't produce cost savings or take treatment to a higher level. Improvements that provide little more than patient convenience at a higher price, like reducing the dose from twice a day to once a day, are valuable only if a drugmaker can prove that once-a-day treatment means fewer skipped doses and improved patient health.
"If a health plan is spending $100 to $1,000 or more per month for a drug, we need to know that we are getting value from the expenditure," explains Dr. Robert Seidman, chief pharmacy officer at managed-care leader WellPoint Inc.
Drugmakers caught unaware of or unprepared for OBA are suffering. Expected blockbusters like Biogen Idec's Amevive (a psoriasis treatment offering new science at a hefty price without better results) and Pfizer's Caduet (a combo-pill of two existing drugs) have become major disappointments because of OBA.
Sadly, few drugmakers or investors routinely incorporate OBA considerations into their decisions. In Bruckner Group's 2006 OBA Manufacturer Index, which measures the ability to bring to market drugs with high health-care value, only 3 of the top 20 pharma and biotech companies get a grade of C+ or better. That's why drugmakers must keep bolstering their products' ability to either significantly improve care or save the system money, or both. Doing so will lead to managed-care acceptance and larger revenues. Ignoring these increasingly powerful customers, on the other hand, is the quickest route to financial illness.
--
David Balekdjian and Michael J. Russo are partners at Bruckner Group, which provides strategic advice to pharmaceutical and biotech company executives.
Source: BusinessWeek
http://www.businessweek.com/magazine/content/06_25/b3989121.htm
Chinese version: BusinessWeek China
http://www.businessweekchina.com/article_p.php?BusinessweekID=993
Saturday, August 19, 2006
Clinical Trials: Phase 1, 2 , 3 and 4 ---(4)
PHASE 4
Phase 4 trials are done after a drug has been shown to work and has been granted a license. So they are looking at drugs that are already available for doctors to prescribe, rather than new drugs that are still being developed.
The main reasons pharmaceutical companies run phase 4 trials are to find out
* More about the side effects and safety of the drug
* What the long term risks and benefits are
* How well the drug works when it’s used more widely than in clinical trials
There is more information about the different phases of clinical trials on the website of The Association of the British Pharmaceutical Industry (ABPI).
Source: Cancer Research UK
http://www.cancerhelp.org.uk/trials/understanding/default.asp?page=73
Phase 4 trials are done after a drug has been shown to work and has been granted a license. So they are looking at drugs that are already available for doctors to prescribe, rather than new drugs that are still being developed.
The main reasons pharmaceutical companies run phase 4 trials are to find out
* More about the side effects and safety of the drug
* What the long term risks and benefits are
* How well the drug works when it’s used more widely than in clinical trials
There is more information about the different phases of clinical trials on the website of The Association of the British Pharmaceutical Industry (ABPI).
Source: Cancer Research UK
http://www.cancerhelp.org.uk/trials/understanding/default.asp?page=73
Clinical Trials: Phase 1, 2 , 3 and 4 ---(3)
PHASE 3
These trials compare new treatments with the best currently available treatment (the standard treatment). They may compare
* A completely new treatment with the standard treatment
* Different doses or ways of giving a standard treatment
* A new schedule of treatment with the standard one
Phase 3 trials are usually much larger than phase 1 or 2. This is because differences in success rates may be small. So, you would need many patients in the trial to show the difference.
For example, 6% more people get a remission with a new treatment compared to standard treatment. If a phase 3 trial gave the new treatment to 50 people and the standard treatment to 50, on average, there may be 3 more remissions in the new treatment group. The 2 groups would not look that different. If they gave each treatment to 5,000 people, there could be 300 more remissions in the new treatment group.
Sometimes phase 3 trials involve thousands of patients in many different hospitals and even different countries.
Randomisation
Phase 3 trials are usually randomised. This means the researchers put the people taking part into 2 groups at random. One group gets the new treatment and the other the standard treatment. There is more about randomisation and different types of trials in this section.
Overviews
Trial overviews are studies that combine all the results from phase 3 trials of a new treatment. They are sometimes called meta-analyses. The idea is to get a broader picture of how well a treatment works. The more data (information) you have, the more accurate the results are likely to be.
Source: Cancer Research UK
These trials compare new treatments with the best currently available treatment (the standard treatment). They may compare
* A completely new treatment with the standard treatment
* Different doses or ways of giving a standard treatment
* A new schedule of treatment with the standard one
Phase 3 trials are usually much larger than phase 1 or 2. This is because differences in success rates may be small. So, you would need many patients in the trial to show the difference.
For example, 6% more people get a remission with a new treatment compared to standard treatment. If a phase 3 trial gave the new treatment to 50 people and the standard treatment to 50, on average, there may be 3 more remissions in the new treatment group. The 2 groups would not look that different. If they gave each treatment to 5,000 people, there could be 300 more remissions in the new treatment group.
Sometimes phase 3 trials involve thousands of patients in many different hospitals and even different countries.
Randomisation
Phase 3 trials are usually randomised. This means the researchers put the people taking part into 2 groups at random. One group gets the new treatment and the other the standard treatment. There is more about randomisation and different types of trials in this section.
Overviews
Trial overviews are studies that combine all the results from phase 3 trials of a new treatment. They are sometimes called meta-analyses. The idea is to get a broader picture of how well a treatment works. The more data (information) you have, the more accurate the results are likely to be.
Source: Cancer Research UK
Clinical Trials: Phase 1, 2 , 3 and 4 ---(2)
PHASE 2
About 7 out of every 10 (70%) new treatments tested at phase 1 make it to phase 2 trials. These trials may be done on people who all have same sub-type of the disease, or with several different sub-types of the disease. Phase 2 trials are done to find out
* If the new treatment works well enough to test in phase 3
* Which sub-types of the disease it is effective against
* More about side effects and how to manage them
* More about the most effective dose to use
Although these treatments have been tested at phase 1, you may still have side effects that are not known about. Drugs can affect people in different ways.
Phase 2 trials are often larger than phase 1. There may be up to 50 people taking part. If the results of phase 2 trials show that a new treatment may be as good as existing treatment, or better, it then moves to phase 3.
Source: Cancer Research UK
About 7 out of every 10 (70%) new treatments tested at phase 1 make it to phase 2 trials. These trials may be done on people who all have same sub-type of the disease, or with several different sub-types of the disease. Phase 2 trials are done to find out
* If the new treatment works well enough to test in phase 3
* Which sub-types of the disease it is effective against
* More about side effects and how to manage them
* More about the most effective dose to use
Although these treatments have been tested at phase 1, you may still have side effects that are not known about. Drugs can affect people in different ways.
Phase 2 trials are often larger than phase 1. There may be up to 50 people taking part. If the results of phase 2 trials show that a new treatment may be as good as existing treatment, or better, it then moves to phase 3.
Source: Cancer Research UK
Clinical Trials: Phase 1, 2 , 3 and 4 ---(1)
PHASE 1
These are the earliest trials in the life of a new drug or treatment. They are usually small trials, recruiting anything up to 30 patients (often a lot less).
When laboratory testing shows a new treatment might help treat a certain disease, phase 1 trials are done to find out:
* The safe dose range
* The side effects
* How the body copes with the drug
* If the treatment shrinks cancer
The first patient to take part will be given a very small dose of the drug. If all goes well, the next person will get a slightly higher dose. With each patient taking part, the dose will gradually be increased and the effect that has will be monitored. Any side effects will be recorded.
In a phase 1 trial, you may have lots of blood tests, as the researchers look at how the drug is affecting you. And at how your body copes with, and gets rid of the drug.
People entering phase 1 trials often have severe disease and have usually had all the treatment available to them. This is because they may benefit from the new treatment in the trial, but many won't. The aim of the trial is to look at doses and side effects. This work has to be done first, before we can test the potential new treatment to see if it works. Phase 1 trials are important because they are the first step in finding new treatments for the future.
Source: Cancer Research UK
These are the earliest trials in the life of a new drug or treatment. They are usually small trials, recruiting anything up to 30 patients (often a lot less).
When laboratory testing shows a new treatment might help treat a certain disease, phase 1 trials are done to find out:
* The safe dose range
* The side effects
* How the body copes with the drug
* If the treatment shrinks cancer
The first patient to take part will be given a very small dose of the drug. If all goes well, the next person will get a slightly higher dose. With each patient taking part, the dose will gradually be increased and the effect that has will be monitored. Any side effects will be recorded.
In a phase 1 trial, you may have lots of blood tests, as the researchers look at how the drug is affecting you. And at how your body copes with, and gets rid of the drug.
People entering phase 1 trials often have severe disease and have usually had all the treatment available to them. This is because they may benefit from the new treatment in the trial, but many won't. The aim of the trial is to look at doses and side effects. This work has to be done first, before we can test the potential new treatment to see if it works. Phase 1 trials are important because they are the first step in finding new treatments for the future.
Source: Cancer Research UK
