(www.improvingmedicalstatistics.com)                                          
     
home   author    contact   web site dedication
  

         

......................m.........................

Specific guide to this web site for:


 1.  Medical School
      Educators 
      in Statistics


 2.  Medical Students

 3.  Science media writers

 4.  High School & College
     Statistic Teachers


   Misadventures:


1. Harvard led MI study

2. JACC study 

   (J. of Amer. Coll.
   Cardio.)


3. NEJM cath study

4. Amer. J. of Cardio.
    review of literature

5.
ALLHAT
    controversy
 

6.
Oat bran study

7.
Pregnancy & Alcohol

8.
Are Geminis really
   
different?
      
9. Columbia 'Miracle' Study  
                                                 

Additional Topics:

Celebrex

Limitations of Meta-Analyses

Large Randomized Clinical Trials

Tale of Two Large
Trials

Advocate meta-analyses

Network meta-analyses






 

 

 

How Medical Science Advances When There Are Problems in the Scientific Literature

Introduction:  One might ask, if there are numerous problems in the scientific literature, why not just throw up our collective hands?  Why not take  medical information from sources not supported by randomized clinical trials or experimental studies?

How does medical science progress and correct studies that are suboptimal or improperly interpreted?

There is a self correcting aspect to medical science.  A trial may be poorly done, inadequately powered (too few patients to be reliable), or have faulty conclusions not derived directly from the trial data. However, if it is a topic of significant clinical interest, there may be multiple subsequent clinical trials, and often with larger numbers of patients. Future trials may be better constructed in order to answer prior controversies. In addition, there may be an improvement in the treatment modality being studied.  If there truly is a benefit with a particular treatment strategy, it frequently will ultimately be shown with the subsequent accumulated trial evidence, despite an inaccurate individual trial. 

 

The Self-Correcting Nature of Science

One of the major reasons why medical treatment and medical knowledge have successfully advanced is that the information is based on the scientific method. The scientific method relies upon reproducible experimental results and studies.  Theories of what constitute the best treatment can be verified or found to be false by subsequent well done and well interpreted clinical trials. Clinical randomized trials, as well as basic science experiments can build upon prior experimental findings. The field of medicine can progress over time far more easily than descriptive fields such as sociology or economics where the experimental verification of theories is much more difficult to conduct.

“10 Years – Later”   

To show the self correcting nature of medical research over time, an examination of what has subsequently occurred or is speculated to occur in the future for each of the statistical problems reviewed in this web site is as follows:
  
Example 1   Angioplasty vs. Medication and Adequate Aspirin dose for unstable angina. "10 years later"

Example 2  Extremely poor statistical analysis in an article in the Journal of the American College of Cardiology.   
"10 years later"

Example 3  Very poor article in the New England Journal of Medicine. "10 years later" 

Example- 4  Incorrect review of literature regarding dosage of aspirin required for patients presenting with a threatened heart attack. "10 years later"

Example- 5 An influential trial (ALLHAT) that made recommendations far beyond what its data warranted in regards to the initial treatment of hypertension.  
"10 years later
"

Example- 6  Underpowered widely publicized study poorly done study suggesting oat bran had no significant cholesterol lowering properties. "10 years later"

Example- 7  Pregnancy and low levels of alcohol ingestion "10 years later"

Example- 8  Inappropriate and unreliable subgroup analyses being performed. "10 years later"  

back  

 

ImprovingMedicalStatistics.com