Return to the Home
Page    Information Resources

MAKING SENSE OF MEDICAL RESEARCH

An Alternative and Complementary Medicine Resource Guide

CONTENTS:

teek.jpg (3467 bytes)
Introduction
 
     teeush.jpg (2789 bytes) Background
 
     teeqpu.jpg (4050 bytes) Conflict of Interest
 
     teepsh.jpg (3181 bytes) Contradictory Research
 
     teeush.jpg (2789 bytes) Built-In Bias
 
     teek.jpg (3467 bytes) Special Remarks on Alternative Medicine Research
 
     teepsh.jpg (3181 bytes) Questions to Ask
 
     teeush.jpg (2789 bytes) Perspective
teepsh.jpg (3181 bytes) Research Design and Terminology
 
     teeush.jpg (2789 bytes) Observational Studies
     
 teeqpu.jpg (4050 bytes) Experimental / Interventional / Clinical Studies
      teek.jpg (3409 bytes) Bias: Selection bias and procedural bias
teeqpu.jpg (4050 bytes) Books
teepsh.jpg (3181 bytes) Journals
teeush.jpg (2789 bytes)
Government Resources
teek.jpg (3467 bytes) Professional Organizations
teeqpu.jpg (4050 bytes) Web resources

green_bar.gif (1576 bytes)

INTRODUCTION

How do we look at medical research? What is it for? While the answers to these questions may at first seem obvious, it turns out that there are a large number of ways that research is used and interpreted. Research is in a sense another language – a language which many of us associate with “the search for the truth,” but which can also be used, like all forms of communication, to obscure and to misdirect. The purpose of this resource guide is to introduce ways to understand the language and aims of research, both positive and negative, and in the process to clarify the terminology and methods used by researchers, as well as some of the issues involved in experimental design. It will also look at some of the contexts in which research and its reporting now operate.

Background

Over the last few decades, market forces have increasingly entered into the scientific process. What is the scientific process? It is the search for truth, what is real. Science comes from the Latin word for “to know.” The market involves the narrower aim of selling what is real – or, for that matter, selling anything as though it were real.

Recently stories about the famous Bell Laboratories told of the rise and fall of a scientific group devoted to basic research-- the scientists were salaried and obtained no direct benefit from their discoveries other than the joy and prestige of the hunt for knowledge. The lab was funded by the enormous profits from Bell Telephone’s monopoly in phone communications. A culture conducive to the belief in science for knowledge rather than short-term money-making was “in the air,” and many remarkable discoveries were made by the lab. It has now closed due to tightened competitiveness in business (and the break-up of the “Bells”), and a changing technological culture that increasingly values money and expedience as the measure of science’s value.

In universities, basic science formerly received considerable support, and scientists pursued “the search for knowledge for its own sake.” Funding came more from university and government programs than from companies. Now, many scientists have learned, along with their employers, that money—and lots of it—is to be made in the private market. The result is that science is increasingly motivated by the pursuit of money, both for the profit of the university and for the scientists who make discoveries and go on to start their own companies. Scientists indeed are under competitive pressures within departments to produce real returns, and these are measured, one way or another, in terms of the “bottom line.” This has always been true to some extent, but the influence of money has increased dramatically, (just as it has in publishing, the arts, and education), as costs of operation increase dramatically while government support wanes.

Over the past decade, universities have lost their majority status in study development. Increasingly studies are performed directly under the aegis of companies testing their own products. More ominously, these companies are handing over the design and publication of studies to “contract research organizations” (known as CROs), which in turn are increasingly owned by marketing and advertising firms. Research can be promulgated and designed to try to increase the odds that it will produce results useful for advertising, both to the public and to medical and alternative practitioners.

Why is this important? Because in reading research reports, or news of these reports in the popular media, it helps to know how they have come about. The pressure of the market has produced a lot of bad studies in all variations from outright falsification to subtle “re-interpretation” of results. Normal statistical variation in outcomes can be used to “cherry-pick” favorable results for publication while discarding the rest. Focus can be placed on results likely to be positive while ensuring that negatives are not even examined. Several editors of academic journals now acknowledge the problem of research sponsored by pharmaceutical companies.
See Angell and Smith.

There remain many good studies as well. There remain many sincere scientists, good institutions, and good companies with positive results to report. For these, research is used to confirm and communicate what they have truly found. How are we to tell? An understanding of some of the factors involved is helpful.

Conflict of Interest

A growing amount of attention is being paid to the issue of conflict of interest in studies. Institutional Review Boards (IRB) are a necessary screening step for any research involving human subjects. These IRBs increasingly request applicant investigators to submit answers to questions about conflict of interest, usually framed in terms of financial interests, such as stock holdings involving the product being tested, or direct employment (and bonuses) by the company testing the product. The dictionary definition of conflict of interest is broader, including anyone who might personally benefit in any way from their actions. In other words, almost everyone involved in research has some degree of conflict of interest. The question is, to what degree, and more to the point, is it recognized? Most applicant investigators claim that they have absolutely no conflict of interest which means some of the influences that might affect the reading, recording, and interpretation of results go unrecognized.

An example of the blinding effect of conflict of interest has been found for decades in American medicine in the practice of giving “perks” to doctors. More than pens marked with drug names, these have gone so far as cash payments and junkets to resorts. When surveys were done of doctors as to how they felt this affected their treatment choices, most answered “not at all.” (When asked how it might affect other doctors, many allowed that there probably was some degree of inducement.) Studies of actual practice choice strongly showed that doctors chose the “supportive product” more often. (Why else would the companies have spent the money this way?) In recent years the practice has been severely curtailed. However, doctors are still paid “honoraria” as consultants by drug companies to give presentations on various diseases, during which it is understood that the company’s products will be mentioned favorably.

An example is the case of Neurontin, an anti-seizure medicine which came to be used for all sorts of pain syndromes after intensive marketing efforts employing honoraria for doctor presentations at conferences, and numerous articles in medical magazines. The uses of the medicine proliferated as doctors seeking solutions for difficult pain problems kept hearing and reading the name. Yet the drug, even when useful in these indications, may cause dramatic side-effects, and has never gone through appropriate studies for these various “off label” applications.

Another case involves the continued sale of some COX-2 inhibitors (such as Celebrex) for pain, even after increased rates of heart disease were found in users. The FDA was considering pulling all these medicines off the market. Ten (of about 30) of the review committeee members turned out to have been consultants to the companies that make these drugs, and nine of them voted in favor of allowing marketing to continue. Even when these drugs were introduced years ago, the manufacturers hyped up questionable benefits (such as the supposed prevention of gastrointestinal (GI) bleeding, and “improved” pain control compared to standby non-steroidal medications such as ibuprofen and naproxen) and minimized the harms found even in their own studies – harms which included the increase in heart disease. Neither GI bleeding nor pain control have been shown to be significantly improved. This “marketing by misinformation” was achieved through the use of researchers and physicians who were placed into the position of personally benefiting by giving positive support for this class of medicines.

It doesn’t stop there. Schools of treatment methods can encourage flawed studies in support of their mission. Editors of medical publications can have strong beliefs, or friendships with people involved in certain product areas, which can cause deficient reports to be published in what would appear to be reputable journals. Reporters at newspapers can have particular axes to grind, and in the process misread studies in journals and misreport them on the front page. These are all effects of conflict of interest.

Money, prestige, fear for one’s job, the desire to be right, or to advance one’s career – all can influence the outcome of research, especially if not seen and factored in. Finding the unvarnished truth requires that the researcher see through the varnish when performing and interpreting the research. It also requires a strong ethical basis for practicing research.

Anyone looking at published research is helped by indications of conflicts that may have affected the outcomes. Increasingly, this information is being included to a very limited extent in footnotes about the authors.

Contradictory Research

Even when research is well conducted, there may be confusing and contradictory findings. Research sponsored by private companies will often disagree with results found by independent institutions. The former often have more of a vested interest in a particular outcome.

Analyses of research into the use of prostate specific antigen testing and mammography for cancer screening have generated many differing conclusions as to their effectiveness. Some of this has to do with the way the research was done, and which research is chosen for review, and some has to do with the biases of the reviewers. Urologists are much more strongly inclined to conclude that prostate specific antigen (PSA) testing is beneficial at early ages, in part due to their conflict of interest – they profit from the procedures that follow positive tests. This position has been convincingly challenged in the book Should I be Tested for Cancer? The American Cancer Society has determined by its reviews that mammography screening for breast cancer should begin at the age of forty, while the professional societies of internists and family physicians find that they should not routinely begin before the age of fifty. Of further interest is the fact that among different countries (having differing cultures), the same body of studies can be reviewed in various ways to arrive at contradictory screening recommendations, including recommending against screening.

Cultural issues can fog the interpretation of information. For example, there is much controversy surrounding research efforts on the cardiovascular-protective nature of red wine, which began after lower rates of heart disease were noted in the French population. Currently, there is no conclusive evidence determining any exact compound responsible for the health benefits of red wine. Some research points to grape derivatives in red wine or in grape skins; other research points to alcohol itself as the beneficial component. But this all may be beside the point, since review of French death statistics indicates that it may have been customary in France to declare death as due to other causes than the heart, for cultural reasons. Thus, red wine may be very “effective” in France, but perhaps not so elsewhere.

Many questions arise among both the public and professionals in response to research findings and exaggerated claims that are often confusing and contradictory. What conclusions and generalizations can be made from the research literature? Should practitioners change treatment protocols on the basis of new research findings? Might the public have access to all the data to assess it for themselves?

Built-in Bias

Sometimes studies are designed with inherent bias. An example of such built-in bias is previous research on hormone replacement therapy (HRT). These studies concluded that HRT had cardiovascular benefits for women in menopause. It now turns out that women selected for these studies (“selection bias”) had to be extremely healthy and fit. Later, more carefully designed studies showed that HRT had no cardiovascular benefits.

“Lead time bias” has created many early results seemingly favorable to early PSA screening. Since prostate cancer is so slow to develop in most men, many will outlive it before even noticing symptoms. Early screening can apparently result in early “cure” of a disease which might never have had any effect on many or even most people. The lead time for diagnosis caused by the test has created a large group of “diseased” men who might not in fact have a disease -- most of whom instead may have a normal manifestation of ageing - see The Last Well Person.

For further discussion, please see the Bias section below, which outlines many of the numerous forms of bias in research.

Don't believe everything you read - Commercial suppliers frequently state that research shows that their product is effective, and safe. Schools of therapeutic practice do the same as to their techniques. It is important to question what sort of research their claims are based on. Is it based on simple case reports without crossover, or randomized, placebo controlled clinical trials, or somewhere in between? (These terms are explained in clear, simple language in the "Research Design and Terminology" section.) Was the form of the research appropriate to the product or method of treatment?

If suppliers make many claims for their product, be wary. “If it seems too good to be true, then it probably is.” There has never been a cure-all or panacea. If practitioners refuse to fully explain their protocol so that it can be repeated and researched by others, be very wary. If they don’t give specific references to claimed studies so that they can be looked into, or if they refuse to explain what is in their product, be wary. If they refuse to give details of a device, how it works, or how it was studied, again, be wary.

Special Remarks on Alternative Medicine Research

Trials performed on alternative and complementary means of treatment have only fairly recently become widely encouraged. The support from a relatively new government agency, The National Center for Complementary and Alternative Medicine at the National Institutes of Health (NIH), is nevertheless miniscule when compared to the NIH as a whole. Backing of alternative medicine research is largely the domain of private individuals, companies, and schools of practice (homeopathy, Traditional Chinese Medicine, etc.). As a consequence, most studies are not performed by those with long experience in research and with good financial resources at their disposal. Many studies are of inadequate design, especially with regard to number of subjects (too few) and length of time studied (often only a few weeks), but aim for the appearance of remarkable results. The conclusions, often written for quotation in advertising, overreach dramatically, or leave out obvious and important considerations. In other words, all of the problems that affect standard research affect alternative research too. In time, it is hoped, there will be better ethical standards and independent oversight along with increased independent funding for quality research in this deserving area. Maybe this will also encourage alternative medicine researchers to develop more appropriate scientific methods for their field of inquiry.

This is not meant to be discouraging, although it may seem so. The ethical standards of alternative researchers are no worse than those of anyone else. Their experience in research, however, is generally less – with many encouraging exceptions. The freshness of this research could lead to a healing of ethics throughout medical research if taken on in the right way.

Questions to Ask

What was the purpose of the research?
What was the rationale?
Was the research properly designed for the purpose?
Who sponsored and paid for it? Why?
Who performed the research and what did they get out of it?
What was their authority and experience?
What were the biases?
Was the analysis appropriate? Were the conclusions justifiable and valid?
Was the work reviewed independently? Was it repeated anywhere else?
See also the The Resource Guide: How to Assess the Credibility of Medical Information on the Web

Perspective

Much of this introduction has had to do with warnings because it is so very easy for scientists as well as the public to misunderstand the meaning of research. Research serves a very important purpose. It allows the measurement of conditions, safety and effect in a way that can be communicated without relying on pre-existing belief, especially when people are looking for knowledge and understanding well beyond their own immediate experience, for selecting new paths, whether of research or of treatment. It is important, though, to move from unquestioning belief in all research to appropriate skepticism. A winnowing process can be applied to reports of research so that the nourishing kernels remain.

[Back to Contents]

green_bar.gif (1576 bytes)

RESEARCH DESIGN AND TERMINOLOGY

Study is a broad term inclusive of everything that can be examined objectively.  Generally, there are two major classifications of studies, observational and experimental.

Observational Studies
A study where one or more groups of patients or subjects are observed and characteristics about the patients are recorded.  In this type of study, no intervention (e.g., drugs, procedures, or dietary restrictions) is introduced.   Therefore, there are no control subjects.

Case Study: A simple descriptive report of interesting characteristics of one unique subject.   This type of study takes place over a relatively short period of time, is conducted by a doctor or scientist, and is considered to be anecdotal evidence.
Case Series: A type of case study that reports on at least five cases.  This type of study is useful for indicating that a trend may be present and deserves more formal study.
Retrospective: These studies begin with the presence or absence of a disease or outcome and then look backward in time to try to detect possible causes or risk factors.  To improve quality, retrospective studies may use the case-control method.
Case-Control: The experimental and control groups are matched as closely as possible with regard to age, sex, geographic location, or any other factor that may have influenced the variable being examined.
Prospective: When subjects are followed forward in time.  This type of study can be both observational or experimental.
Cohort: Refers to a group of subjects who have some defining characteristic in common and who remain part of this group over an extended period of time.  The common characteristic is suspected of being a precursor or risk factor for a disease or health effect.   Since the direction of inquiry in cohort studies are forward in time, they are prospective studies.
Cross-sectional: These studies analyze data collected on a group of subjects at a given point in time rather than over an extended period of time, and are designed to determine what is happening presently, not what happened in the past.  One type of of cross-sectional study is a survey.
Survey: A study that reports the results of interviewed subjects who have not received any intervention.  Surveys try to determine if there are any commonalties - or correlations/associations - between various diseases, behaviors, etc.
Positive Correlation: Demonstrates a directly-proportional relationship.  In other words, as factor x increases, factor y increases.
Negative Correlation: Demonstrates an inversely-proportional relationship.  In other words, as factor x increases, factor y decreases.
Epidemiology: The study of disease patterns and the factors that influence them.
Prevalence: An epidemiological term describing how many people at a given point in time have a given disease or condition.
Incidence: An epidemiological term describing how many new cases of a given disease or condition occur over a given period of time.

Experimental / Interventional / Clinical Studies
These are studies where an experimental intervention is introduced to a group of subjects to determine the efficacy of certain procedures or treatments. This type of study is also called a trial, and can be either controlled or uncontrolled. Due to the nature of this study design, trials are always prospective.

Preclinical Trials: Subjects are non-human, such as animals, bacteria, cells, etc.
Clinical Trials: Subjects are human
Controlled Clinical Trials: At least two groups are compared to determine the efficacy of an intervention.  The experimental / treated group receives the intervention while the control / placebo group does not.
Randomized Clinical Trial: Subjects are randomly assigned to either the experimental group or the control group.  Apart from the intervention, these two groups are treated equally. Thus, any differences between the two groups can be attributed to the intervention and not other factors. This method provides the strongest evidence for determining cause and provides the best evidence that the result was due to the intervention.
Nonrandomized Clinical Trial: When subjects are not assigned to either the experimental or control group at random.  This method renders the study open to many sources of bias.  Results from theses studies are highly questionable and are generally considered to be much weaker than randomized clinical trials since they do not screen for bias.
Self-Controlled Clinical Trial: The same group of subjects are used for both the experimental and control groups.  This method is vulnerable to bias since subjects may change their behavior, which may lead to improvement in their condition.  In this case, the improvement is due to the attention the subject receives from the study, and not the intervention itself.  To improve this research design, a cross-over trial can be used.
Cross-Over Trial: Subjects are assigned to either the experimental or control group.  After a period of time, both groups receive no intervention at all, and then the two groups switch into the opposite groups.  By this method, some of the bias that may normally be present can be avoided.
Externally-Controlled Clinical Trial: In this study, controls external to the study are used.  For example, the results of another investigator's research or patients that the investigator has previously treated in another manner may be used as the controls in the study.
Placebo-Controlled Clinical Trial: A study that administers an inactive pill or procedure to the control group during the trial period.  Placebos are given to maintain the same settings and environment for both control and experimental groups.  However, a "placebo effect" may occur, meaning that if a patient believes s/he is receiving an active treatment, s/he may experience effects (positive or negative), even if there is nothing in the placebo that could be the cause.  Because the average placebo effect is approximately 33% (ranging from 3% to 65%), research results must be significantly more effective than the placebo effect to demonstrate that a procedure or drug works.
Food and Drug Administration (FDA) Trials: The FDA is a consumer protection agency of the United States Federal Government which mandates that specific trials on human subjects be performed before receipt of FDA approval.  There are four types of drug studies:
Phase I Trial: Usually conducted on a small number of healthy volunteers without a control.  Tests for best dosage, potential side effects and safety.
Phase II Trial: Tests a drug with known dose and side effects on people with the condition or disease to be treated.  A control group may or may not be used.  This phase supplies preliminary data on whether the treatment works, how the drug helps the condition, and supplements the safety data of Phase I.
Phase III Trial: Usually randomized and controlled.  Assesses efficacy, safety, and dosage compared with standard treatments or a placebo.
Phase IV Trial: Not routinely required, this phase is done after drugs are approved by the FDA and sold to the public.  These studies are either randomized or surveys that attempt to evaluate long-term benefits and risks.
Investigational New Drug (IND) application and New Drug Application (NDA): Phase I, II, and III trials are usually performed as part of this application to the FDA.  After the Phase III trial is completed, the manufacturer can submit an NDA application which requests permission to market the drug.
Uncontrolled Clinical Trial: These studies are more likely to be used to compare procedures rather than the efficacy of drugs. The problem with this type of study is that it must be assumed the procedure described is the best since there is no control to compare it to.

Bias
Bias is a systematic error involved in the selection of subjects, research design, or in the measurement of a characteristic. Whether intentional or unintentional, bias may influence experimental findings. There are many types of bias, which can be categorized into two major types: selection and procedural:

    Selection Bias

Prevalence Bias: Occurs when a disease or condition is characterized by early fatalities, as some subjects die before they are diagnosed.  Prevalence bias may also result if there is a time gap  between exposure / onset of the condition and selection of subjects, and the "worst" cases have died.
Admission Rate Bias (Berkson's Fallacy): Occurs when patient admission rates vary between the experimental and control groups, especially in studies that use hospitalized patients for both groups.  For example, this type of bias is present when hospitalized patients with the risk factor (experimental group) are admitted at a higher rate than the control group.
Nonresponse Bias:This type of bias occurs most often with surveys.  If there is no follow-up of people who did not respond to the survey, it is difficult to determine if the responses received are typical of the general population.
Membership Bias: Occurs when there are pre-existing groups, because often, one or more of the characteristics that cause the people to belong to the groups are related to the outcome of interest.  For example, in studying the clinical effects of smoking on cancer, it may be that smoking is not the actual cause of cancer, but rather a trait that is more common among smokers.
Procedural Selection Bias: Frequently occurs in studies that are not randomized or in studies using historical or external controls.  Treatment assignments are made on the basis of certain patient characteristics, which results in dissimilarity between treatment groups.

    Procedural Bias occurs when study groups are not treated similarly.  This may happen when one group receives more attention or care, which may change the attitude of one group.  This causes an imbalance in the study results.

Recall Bias: Occurs when subjects are asked to recall certain events.  Subjects in one group may be more likely to remember the event than those in the other group.
Detection Bias: Occurs with the advent of new technology and equipment for detection of disease.  If new technology allows for earlier detection, survival for these patients will appear to be longer because the condition was diagnosed earlier.
Compliance Bias: Occurs when patients find it easier to comply with one treatment as compared to another.   This may cause data to be skewed towards the favorable treatment, even it is not as effective.

In order to have an effective and statistically significant study, bias must be avoided.  There are several procedures that can be incorporated into the research design to avoid bias:

Case Control: See above.
Randomization: See above.
Cross-Over Techniques: See above.
Blinding/Masking: When the researchers and /or subjects are prevented from knowing what group each subject is in.
Single-Blind: When the researchers, but not the subjects, know which group the subjects are in.   Non-surgical studies are usually single-blinded (if not double-blinded) since the subjects' knowledge of receiving treatment or placebo is likely to affect results.
Double-Blind: When neither researchers nor subjects know which group the subjects are in.  This is made possible by coding all the data, and this code is broken at the end of the experiment.   This procedure is especially important since researchers can give unconscious cues that may influence subjects' responses independently of the treatment.
Statistical Significance: Research findings are significant when it can be shown statistically that they are unlikely to have occurred by chance.
Sample Size: The number of subjects in a study.  Generally, a sample size should be large enough to produce a p value of <0.05.
P value: There is always a probability that study findings may have occurred by chance, rather than due to the intervention.  The p (probability) value determines this probability.   To understand the significance of a p value, move the decimal point two places to the right.  For example, a p value of <0.01 (where "<" stands for "less than") means that the probability that the results occurred by chance is less than 1%.  Generally, p values <0.05 are considered to be statistically significant.
Confounding Variables / Confounding Factors / Confounders: These are characteristics that may influence experimental results.  A large sample size helps to control for confounding factors.  For example, small study groups may not have an even distribution of a characteristic that is known to affect or cause the result under investigation.  With larger study groups, confounding factors are likely to be evenly distributed between the two groups.
Null Hypothesis: A statement claiming that there is no difference between the assumed research results and the population average.  In other words, a null hypothesis is a pre-specified statement saying that the treatment under investigation will have no effect on the subjects.
Type I (Alpha) Error: When results suggest that a treatment works when in fact it does not work - a false positive.  This occurs when one rejects the null hypothesis when it is true.
Type II (Beta) Error: When results suggest a treatment does not work when in fact it actually does - a false negative.  This error occurs when the null hypothesis is accepted but is actually false.

With the understanding of all these terms, the gold standard for a medical research design is widely recognized as a prospective, randomized, double-blinded, placebo-controlled trial with a sample size large enough to produce a p value of 0.5 or less.

Although it is unlikely for many studies to actually follow such a research design, it is useful to know the gold standard and for trials of both conventional and alternative medicine to be compared to it.

[Back to Contents]

green_bar.gif (1576 bytes)

NOTE: The following resource listings are not intended to be comprehensive, nor to be used as a guide for treatment.   They are provided for information only.  The resources are selected and categorized to help you with your own research.

BOOKS

TEXTBOOKS

amazonlogo.gif (1557 bytes) George Lewith, Wayne B. Jonas, and Harald Walach
Clinical Research in Complementary Therapies
Churchill Livingston, 2002.
Authoritative contributions on the theory and practice of research in complementary medicine, with specific applications to many areas of alternative therapy.

Introduction to Research Elizabeth Depoy and Laura Gitlin
Introduction to Research: Understanding and Applying Multiple Strategies
Mosby-Year Book, 1998
A textbook for students or beginning researchers, describing the thoughts and actions involved in the conduct of naturalistic, experimental, social, or behavioral research.

  Understanding Health Care Outcomes Research  Robert L. Kane
Understanding Health Care Outcomes Research
Jones and Bartlett Publishers, 2004
A textbook for the investigator and administrator, provides an introduction to the utility of outcomes data in the process of making clinical decisions.

piantadosi.gif (10194 bytes) Steven Piantadosi
Clinical Trials: A Methodologic Perspective
John Wiley and Sons, 1997
Explains the basic statistical methods used in clinical design, bias and random error, sample size, and the study cohort.

  Clinical Trials  Stuart J. Pocock
Clinical Trials: A Practical Approach
Wiley-Interscience, 1997
A comprehensive text on the principles and practice of clinical trials.   This book describes the design, analysis, and interpretation of clinical trials in a non-technical manner and provides a general perspective on their historical development, current status, and future strategy.

 

GENERAL AUDIENCE

The Truth About the Drug Companies

Marcia Angell
The Truth About the Drug Companies: How They Deceive Us and What to Do About It
Random House, 2005
Powerful testimonial from the previous Editor of the New England Journal of Medicine of the growing corruption of the pharmaceutical industry and how they moved from manufacturing useful drugs to becoming vast marketing corporations

berman.gif (11097 bytes) Adriane Fugh-Berman, M.D.
Alternative Medicine: What Works
Lippincott, Williams and Wilkins, 1997
A comprehensive, easy to read review of the scientific evidence on alternative medicine.  Contains an excellent chapter explaining scientific terms and research design.

 Evaluating Research Articles  Ellen Girden
Evaluating Research Articles from Start to Finish
Sage, 2001
Gives examples of studies of varying quality, poses questions as an exercise, and discusses answers in a separate section.

The Last Well Person Nortin M. Hadler
The Last Well Person
LMcGill-Queen's University Press, 2004
MD discusses how constant monitoring and unnecessary interventions by a self-serving medical industry turn healthy people into patients

Statistics - A Spectator Sport  Richard Jaeger
Statistics: A Spectator Sport
Sage Publications, 1990
Excellent resource for deciphering and understanding statistical data to help you make an informed decision. The primary focus is on teaching the user of statistical information what statistics are, what they mean, and how to use and interpret them appropriately.

The Best Alternative Medicine Kenneth Pelletier, M.D.
The Best Alternative Medicine
Fireside, 2002
An outline of effective alternatives including available scientific evidence.

Studying a Study and Testing a Test: How to Read the Medical Evidence Richard Riegelmann
Studying a Study and Testing a Test: How to Read the Medical Evidence
Lippincott, Williams and Wilkins, 2004
How to evaluate medical literature, for students and practitioners.

  Should I be Tested for Cancer?   H. Gilbert Welch
Should I Be Tested for Cancer? Maybe Not and Here’s Why
University of California Press, 2004
Cancer expert speaks out against common cancer screening tests and challenges the view that the best defense against cancer is to catch it early.

 
ESSAY

Public Library of Science Richard Smith, MD
Medical Journals Are an Extension of the Marketing Arm of Pharmaceutical Companies
PloS Medicine, May 2005
http://medicine.plosjournals.org/perlserv/?request=get-document&doi=10.1371/journal.pmed.0020138

Former Editor of the British Medical Journal speaks out against the manipulation of respected journals by the pharmaceutical industry.

[Back to Contents]

green_bar.gif (1576 bytes)

JOURNALS

American Journal of Epidemiology (AJE)
http://aje.oxfordjournals.org/

The AJE is produced by the Society for Epidemiologic Research and is devoted to the publication of empirical research findings and methodologic developments in the field of epidemiology.  It is aimed at epidemiologists, health workers, and clinicians.   Searches on this site link to abstracts and synopsis of commentaries.

Annals of Epidemiology
http://www.journals.elsevierhealth.com/periodicals/aep

A peer-reviewed, international journal devoted to epidemiological research and methodological development.  The journal emphasizes the application of epidemiologic methods to issues that affect the distribution and determinants of human illness in diverse contexts.

Contemporary Clinical Trials: Design, Methods, and Analysis
http://www.elsevier.com/locate/conclintrial
The official journal of the Society for Clinical Trials. Search by tables of contents and abstracts or by keyword.

Journal of Clinical Epidemiology
http://www.sciencedirect.com/science/journal/08954356

This monthly publication provides timely, authoritative studies developed from the interplay of clinical medicine, epidemiology, biostatistics and pharmacoepidemiology.   Articles are oriented toward methodology, clinical research, or both.

[Back to Contents]

green_bar.gif (1576 bytes)

GOVERNMENT RESOURCES

National Center for Complementary and Alternative Medicine
http://nccam.nih.gov/health/clearinghouse/
NCCAM Clearinghouse
P.O. Box 7923
Gaithersburg, MD 20898
Toll-Free: (888) 644-6226
Email: info@nccam.nih.gov
The NIH center for the coordination of alternative medicine research.

Food and Drug Administration (FDA)
http://www.fda.gov/
HFI-40
Rockville, MD 20857
Toll-Free: (888) 463-6332
Email: webmail@oc.fda.gov
The FDA is a consumer protection agency of the United States Federal Government, which mandates that specific trials on food, medicines, medical devices, and other products be performed before receipt of FDA approval.

National Center for Research Resources (NCRR)
http://www.ncrr.nih.gov/
National Institutes of Health
Bethesda, MD 20892-5662
Phone: (301) 435-0888
Email: ospio@ncrr.nih.gov
The NCRR creates, develops, and provides a comprehensive range of human, animal, technological, and other resources to enable biomedical research advances. NCRR seeks scientific knowledge that will lead to better health and reduced illness and disability for our nation's citizens.

National Institutes of Health (NIH)
http://www.nih.gov/health/
Bethesda, MD 20892
Links to credible health information and research supported by the N.I.H. Click on "MEDLINEplus" for information on hundreds of diseases and conditions and wellness issues.

[Back to Contents]

green_bar.gif (1576 bytes)

PROFESSIONAL ORGANIZATIONS

American College of Epidemiology (ACE)
http://www.acepidemiology.org/
4101 Lake Boone Trail, Suite 201
Raleigh, NC 27607
Phone: (919) 787-5181
Fax: (919) 787-4976
Email: info@acepidemiology.org
Professional organization dedicated to continued education and advocacy for epidemiologists in their efforts to promote public health.

Society for Clinical Trials (SCT)
http://www.sctweb.org/
600 Wyndhurst Ave.
Baltimore, MD 21210
Phone: (410) 433-4722
Fax: (410) 435-8631
Email: sctbalt@aol.com
An international professional organization dedicated to the development and dissemination of knowledge about the design and conduct of clinical trials and related health care research methodologies.

Society for Healthcare Epidemiology of America (SHEA)
http://www.shea-online.org/

19 Mantua Rd.
Mt. Royal, NJ 08061
Phone: (856) 423-0087
Fax: (856) 423-3420
Email: sheahq@talley.com
Organized to foster the development and application of the science of health care epidemiology which is broadly defined as any activity designed to study and/or improve patient care outcomes in any type of healthcare institution or setting.

[Back to Contents]

green_bar.gif (1576 bytes)

WEB RESOURCES

NOTE: Promotional and commercial sites are not included in this listing unless they provide significant impartial information resources.

CenterWatch
http://www.centerwatch.com/MAIN.HTM
A searchable listing of over 2,100 ongoing clinical trials categorized by therapeutic area and geographic region;  includes information on clinical research at the National Institutes of Health.

ClinicalTrials.gov
http://clinicaltrials.gov/ct/gui/c/r
Developed by the U.S. National Institutes of Health, this site provides patients, family members, and members of the public with current information about clinical research studies.  Click on "clinical trials" to learn more about this topic, or search the clinical trials database.

Cochrane Collaboration Reviews
http://www.cochrane.org/reviews/index.htm

You can view the abstracts (summaries) by searching on topic or title; view a full alphabetical list or browse by review group and topic.

Electronic Orange Book
http://www.fda.gov/cder/ob/default.htm
Maintained by the U.S. Food and Drug Administration, this is a database of approved drug products with therapeutic equivalence evaluations.

NCCAM Clinical Trials Site
http://nccam.nih.gov/clinicaltrials/factsheet/
This is an excellent starting point for finding out more about alternative medical research, provided by the NIH National Center for Complementary and Alternative Medicine.

Resources in Epidemiology and Biostatistics
http://www.usm.maine.edu/~ams/epi/linkepi.htm
Links to journals and other resources in epidemiology and biostatistics.

Rosenthal Center Directory of Databases
http://www.rosenthal.hs.columbia.edu/Databases.html

A compilation of established sources in the U.S., Europe and Asia, designed to facilitate research by both professionals and the public.

Search the Studies
http://clinicalstudies.info.nih.gov/
Links to information on clinical research: what it is, who can participate, and referrals.

Statistics Glossary
http://www.stats.gla.ac.uk/steps/glossary/
Definitions to statistical and research methods terminology, with links to related words

[Back to Contents]

green_bar.gif (1576 bytes)
Last updated: June 30, 2005
green_bar.gif (1576 bytes)



© Copyright © 1998-2010 Alternative Medicine Foundation Inc., Potomac, Maryland.
All rights reserved.