Selecting Studies for Meta-Analysis: Publication Bias

Meta-analysis is a statistical method of combining the results of individual studies. Meta-analysis may allow a more precise estimate of treatment effects, and may explain differences between the results of individual studies. Although meta-analysis can be a powerful and useful technique, they must be designed carefully to ensure that the results are not misleading.
The identification and selection of individual studies used in a meta-analysis is critical. Publication bias has a direct impact on this selection.
Publication bias refers to the greater likelihood of papers showing positive results to be published than negative results. If the sample of studies selected for meta-analysis is biased, the conclusions drawn from the analysis may be invalid. Turner(2) analyzed the publication status of antidepressant drugs. He found that 97% of the positive studies were published versus only 12% of negative studies. The inclusion of negative studies in the analysis reduced the positive effects of the drugs.
Statistical tests have been developed to identify and assess the impact of publication bias in meta-analysis.
Attempting to locate unpublished studies is time consuming, difficult, and can use methodology hard to assess.(1)
In evaluating the results of meta-analyses (as well as for individual studies) it is important to be aware of who is sponsoring the studies and whether the authors have any vested interests.
Is it possible to overcome the problem of publication bias? The National Institutes of Health keeps a registry of all studies it supports, and the FDA keeps a registry and database in which drug companies must resister all trials they intend to use in applying for marketing approval or for changes in labeling.
Prospective meta-analysis looks forward to meta-analysis in the planning stage of individual trials. Researchers of these trials agree, prior to knowing the results of their studies, to combine their findings when the trials are complete.(3) The researchers agree on trial design as well as outcome measures which facilitates the analysis of results.

1. Rothstein, Hannah, Alexander Sutton & Michael Borenstein, eds. Publication Bias in Meta-Analysis: Prevention, Assessment and Adjustments. 2005. John Wiley & Sons
2. Turner, Erick, et al. Selective Publication of Antidepressant Trials and Its Influence on Apparent Efficacy. N Engl J Med. 2008; 358: 252-60. Trials (BioMed Central) 2011;12:104
3. Turok, David et al. The methodology for developing a prospective meta-analysis in the family planning community.
4. Walker, Esteban, et al. Meta-analysis: Its strengths and limitations. Clev Clin J Med. June 2008; 75(6): 431-439.[tt_news]=360745&cHash=c4f41096946e87f82f539928262b817f

Big Data Advances Personalized Medicine

Personalized medicine is much in the news today. What is it and why is it important?

Until recently, patient treatment has been characterized by a one size fits all approach. Patients with a given condition are given a set medication; many respond as expected, while others either do not respond or react unfavorably.

An important factor in variability in response to treatment is the underlying genetic makeup of individuals. Drug response is affected by the extent (variability) of drug delivery to the sites of drug action as well as by the effectiveness with which the drug interacts with specialized receptors or enzymes. The activity of these processes are influenced by genetics.

Roden, Dan & Alfred George. The Genetic Basis of Variability in Drug Responses. Nature Reviews Drug Discovery, Vol. 1, Jan. 2002, pp. 37-44.

Dramatic reductions in costs of genomic analyses have resulted in real advances in personalized medicine applications. A profile of each individual can be made by associating their unique genetic makeup with treatments that are most likely to respond favorably to that makeup. These personalized treatments (known as targeted therapies) result in more effective treatments at lower cost and time.

A Personalized Medicine World Conference is held each year providing the latest developments in the field. Past speaker videos can be viewed at:

Although the healthcare community recognizes that Big Data could aid in improving patient care, they often do not have the means to use it. This is where information technologies come into the picture. GNS Healthcare is an example.

their computational engine uses “ supercomputers to analyze relationships among multiple types of patient data, including patient population information, electronic medical records, images, clinical outcomes, and other data, learning as it goes.”  Personalized medicine, therefore, is going beyond application of genome analyses to include “real life” data.

Personalized medicine improves diagnostic capability and predictions of outcomes. It can aid investigators to select patients for clinical trials who are most likely to respond to the treatments. For trials in progress, biomarkers can be identified using genotype, gene expression, and patient outcome data. Big Data analytics can identify hidden drug interactions as well as patient characteristics and care processes that affect safety and efficacy.

Personalized medicine has come of age through an explosion of electronically available medical data and advances in computing technology to analyze it.


Big Data Ready to Transform Healthcare

The problems of present-day healthcare are well known—gross inefficiencies, poor outcomes, and inadequate personalized care. A vast amount of information is out there, known as Big Data, that can aid greatly in solving these problems, if only the data could be efficiently accessed, analyzed, summarized, and applied to the healthcare system.

These advances are indeed rapidly taking place. The University of Pittsburgh Medical Center health system is partnering with several information technology companies to create a database that integrates financial, administrative, clinical and genomic information. Although the process is complex, it can lead to simplified solutions.

Another example is the use of IBM’s supercomputer, Watson. The supercomputer was loaded with vast amounts of patient records from the Memorial Sloan-Kettering Cancer Center, as well as from many medical journal publications. Platforms were developed that physicians can access to determine cancer treatment options.

Medicare is serving as an impetus to accelerate the process by requiring hospitals to improve care in three critical areas or to incur penalties: reduce readmission rate of patients, adopt electronic health record systems, and reduce hospital-acquired infections.

Analyses of Big Data promise applications that were not really available before, such as the prediction of disease onset so that interventions can be made before the disease manifests.

Not everyone in the healthcare field is happy about the application of big data analytics in an attempt to reduce costs. Some physicians fear that a change from the traditional fee-for-service system to performance metrics will limit their options in treating patients. The controversy relates to the types of tests or procedures that are ordered or performed and under what situations.

Big Data will be a great assist to advance “personalized medicine.” This will be a topic of another blog.


Bernard, Allen. Healthcare Industry Sees Big Data As More Than a Bandage. CIO, Aug. 5, 2013.

Cerrano, Paul. Why Physicians Don’t Like Big Data. Information Week, August 20, 2013.

Hagland, Mark. Thinking Really Big About Data. Healthcare Informatics 


Big Data Applications

market analysis

“Big Data” is much in the news today. What is it, and what is its significance to the business and medical establishments? Big Data refers to the vast amount of information generated daily, but more specifically to the challenges involved in processing, analyzing, and using the data.

Big Data is being generated from advances in technology including the adoption of large numbers of sensors and smart devices. Big data can be very beneficial in solving business problems, often providing solutions previously inaccessible. To do so can require visionary thinking and talent.

A large amount of Big Data being generated is semi-structured or unstructured data. Quite a bit of this data is lost due to an inability in knowing how to use it.

IBM has developed a platform that addresses this problem. Based on the open-source Hadoop software, this system simplifies, manages, coordinates, and analyzes big Data.

Evaluating Big Data has led to the evolution of a new profession, the Data Scientist, with high levels of both quantitative skills and technical ability.

Virtually all industries can find valuable uses for Big Data, including retail, financial services, manufacturing, government, advertising, media, and energy. Its application for medicine and healthcare will be the topic of my next blog.


Davenport, Thomas & Jill Dyche. Big Data in Big Companies. SAS Institute, May 2013.

Dietrich, David. Big Data Analytics. EMC Education Services. April 4, 2013

Zikopoulos,Paul, et al. Understanding Big Data: Analytics for Enterprise Class Hadoop and Streaming Data. McGraw-Hill Companies, 2012.

Challenges to Open Access

As I mentioned in my previous blog, open access is a means of publication that allows free access to scientific journals for libraries, researchers, businesses, and the general public.

The publishing industry has contended that open access will lower the quality of publications by lack of adequate peer review.

Indeed, a few unfortunate recent incidences have occurred that seem to support this contention. Since financing open access often requires authors to pay for publication, some “predatory” online journals have appeared whose sole purpose is to prey on unsuspecting authors eager to publish. The Scholarly Kitchen

A scandal hit an open access journal published by Bentham Science Publishers. They accepted for publication a phony paper that was computer generated.

These incidences could serve as a wake-up call to the open access movement which is now well established. One of the measures of the impact and value of publications is the number of times an article is cited by other researchers. By this measure, open access articles are doing very well.

In 2012, a bill was introduced in Congress that would have repealed the current open access policy by the National Institutes of Health. This bill caused such a groundswell of protests by research organizations, universities, and even many publishers that it had to be withdrawn. (Online Searcher. Vol. 37, No. 2, March/April, 2013.)


Open Access

David  Olle3

Conducting literature searches is an essential first step to medical writing. Obtaining articles can be an expensive proposition for clients if they are available only through subscription or through online purchase of individual articles. Open access is becoming an increasingly important alternative.

Open access is the practice of providing free availability and unrestricted access via the Internet to peer-reviewed scholarly research. The open access movement began modestly about a decade ago with a small group of scientists and librarians and has now grown to be global in status. Currently about 20% of peer-reviewed articles are available via open access.

Publishers of scientific journals have resisted the movement as it challenges their business model. They often provide free access to articles six months or more after publication, but even then there can be restrictions on their use.

The National Library of Congress’s Pub Med is a popular site for searches. PubMed Central is a free full-text archive of journal articles. Springer is a publisher that has embraced open access, and has acquired another open access site, Biomed Central. The Public Library of Science is another important site. The Directory of Open Access Journals allows searches across a large variety of freely available journals.

Dr. Peter Suber presents the case in Open Access Overview that open access serves the interests of many groups including authors, teachers, libraries, universities, the general public, and even publishers. To this list I would add information professionals and their clients.