flawed research

[Guest guest]

Thought the group would find this interesting. This was forwarded to

me by a friend and patient:

 

The great majority of published research is so deeply flawed that it

should be considered essentially worthless. So says John Ioannidis,

PhD, an epidemiologist at the University of Ioannina School of

Medicine in Greece, in the August, 2005, issue of the journal PLoS

(Public Library of Science) Medicine.

 

" For most study designs and settings, it is more likely for a

research claim to be false than true, " he states in the study's

summary. " Moreover, for many current scientific fields, claimed

research findings may often be simply accurate measures of the

prevailing bias. "

 

Ioannidis singles out several types of study that are particularly

likely to lead to a worthless result, namely:

 

Studies with a small sample size ('underpowered' studies);

Studies with a small effect size, i.e., studies in which the drug

or technique under investigation leads to improvement in only a small

percentage of patients – and it should be remembered that FDA

approval was granted to many of the newer, much vaunted 'targeted'

anti cancer drugs such as Avastin, Erbitux and Iressa on the basis of

response rates that were at best in the 10-20 percent range;

Studies whose outcomes are poorly or subjectively defined, such as

studies which use so-called 'surrogate endpoints' such as tumor

shrinkage to measure outcome, instead of using objective measures

such as death – and the vast majority of clinical studies in the

cancer field employ surrogate endpoints;

Studies in which financial conflict of interest is a factor, such

as is very commonly the case in biomedical research;

Studies in which the researchers are prejudiced by being unduly

wedded to a particular outcome - and this too is far commoner than

one might think;

Studies of a topic that is currently 'hot' – the hotter the field,

the more teams are working against one another and competing to be

the first to publish, and this is likely to lead to selective

reporting only of positive results.

Very often, Ioannidis points out, several of these factors are

working in concert with one another. For example, researchers in a

'hot' field are more likely to be prejudiced in favor of one or other

hypothesis, as well as to be competing fiercely with other teams to

be the first to publish. As he sees it, though, the bottom line is

unambiguous: most research findings are false for most research

designs and in most fields.

 

It is clear that there is a lamentable lack of objectivity and merit

in most of the published research on which clinical medicine depends.

I have made it my life's work to study the medical literature

critically and to question the basis for cancer treatments that have

become universally adopted without ever having been shown to prolong

life. I have written and published extensively on the subject of

cancer and its treatment, including compiling a comprehensive series

of individual reports on more than 200 different cancer diagnoses –

The Moss Reports – each one of which examines both the standard

treatment options that are likely to be offered for a particular

cancer diagnosis, and the possible alternative and complementary

approaches to that disease.

 

 

[Guest guest]

Zev

How true, somthing we need to also keep in mind

alon

 

<zrosenbe wrote:

Thought the group would find this interesting. This was forwarded to

me by a friend and patient:

 

The great majority of published research is so deeply flawed that it

should be considered essentially worthless. So says John Ioannidis,

PhD, an epidemiologist at the University of Ioannina School of

Medicine in Greece, in the August, 2005, issue of the journal PLoS

(Public Library of Science) Medicine.

 

" For most study designs and settings, it is more likely for a

research claim to be false than true, " he states in the study's

summary. " Moreover, for many current scientific fields, claimed

research findings may often be simply accurate measures of the

prevailing bias. "

 

Ioannidis singles out several types of study that are particularly

likely to lead to a worthless result, namely:

 

Studies with a small sample size ('underpowered' studies);

Studies with a small effect size, i.e., studies in which the drug

or technique under investigation leads to improvement in only a small

percentage of patients – and it should be remembered that FDA

approval was granted to many of the newer, much vaunted 'targeted'

anti cancer drugs such as Avastin, Erbitux and Iressa on the basis of

response rates that were at best in the 10-20 percent range;

Studies whose outcomes are poorly or subjectively defined, such as

studies which use so-called 'surrogate endpoints' such as tumor

shrinkage to measure outcome, instead of using objective measures

such as death – and the vast majority of clinical studies in the

cancer field employ surrogate endpoints;

Studies in which financial conflict of interest is a factor, such

as is very commonly the case in biomedical research;

Studies in which the researchers are prejudiced by being unduly

wedded to a particular outcome - and this too is far commoner than

one might think;

Studies of a topic that is currently 'hot' – the hotter the field,

the more teams are working against one another and competing to be

the first to publish, and this is likely to lead to selective

reporting only of positive results.

Very often, Ioannidis points out, several of these factors are

working in concert with one another. For example, researchers in a

'hot' field are more likely to be prejudiced in favor of one or other

hypothesis, as well as to be competing fiercely with other teams to

be the first to publish. As he sees it, though, the bottom line is

unambiguous: most research findings are false for most research

designs and in most fields.

 

It is clear that there is a lamentable lack of objectivity and merit

in most of the published research on which clinical medicine depends.

I have made it my life's work to study the medical literature

critically and to question the basis for cancer treatments that have

become universally adopted without ever having been shown to prolong

life. I have written and published extensively on the subject of

cancer and its treatment, including compiling a comprehensive series

of individual reports on more than 200 different cancer diagnoses –

The Moss Reports – each one of which examines both the standard

treatment options that are likely to be offered for a particular

cancer diagnosis, and the possible alternative and complementary

approaches to that disease.