Dirty Little Secret of Health Informatics

I have just posted a new article describing another flagrant deficiency of the medical industrial complex. The title of the article is:

THE DIRTY LITTLE SECRET OF THE HEALTH INFORMATICS INDUSTRY: Without a reliable patient identifier system, electronic health records (EHRs) are worthless; And most systems are unreliable.

Seriously, patient identifiers are at the heart of the push to create electronic health records for everyone. The purpose of the article is to describe the kinds of problems that arise when patient identifiers are unreliable. Most institutions simply do not have in place a sensible system to ensure that every patient has a unique identifier, and that all of the reports for a patient are permanently "held together" by the identifier. The article also describes the features of a patient identifier system, emphasizing the essential role of identifiers in healthcare and biomedical research.

The article is available on my web site at:

http://www.julesberman.info/book/id_deid.htm

© 2011 Jules J. Berman

Tips for writing an evil grant

• Machiavelli's Laboratory is a free ebook that I published on April 13, 2010. It is a satiric discourse on scientific ethics, from the perspective of an unethical scientist. Please don't take any of the advice and opinions in the book (or the excerpts featured in this blog) seriously.

The typical grant reviewer gives your grant a perfunctory once-over, looking for answers to the following simplistic questions:

1. Is the applicant a bona fide scientist? This often boils down to whether the reviewer knows the applicant or knows any of the people who the applicant has worked with, past or present. If the applicant is a complete unknown to the reviewer, it is unlikely that his application will be taken seriously by the reviewer.

2. Does the applicant have a history of completing the grant awards that he has received? If the members of the study section learn that you botched a prior grant, it's an uphill battle to get another chance of funding. One of the best ways of unloading a failing grant is to relocate to a new institution. Many grants are tied to a set of resources that are unique to the host institution. When the principal investigator leaves, the grant stays in the institution. When this happens, the principal investigator is relieved from his obligation to fulfill the grant. The home institution continues to draw down its indirect costs for the grant (i.e., the part of the grant funding that pays for the administrative and physical overhead of supporting the grant), but the incentive to follow the grant's research plan has vanished. In these cases almost everyone wins. The principal investigator has moved on to another job; he can list his severed grant as proof of his fundability, and he can apply for another grant without being held accountable for the failure of his research project. The original host institution continues to collect indirect costs for the duration of the grant. Some lucky researcher in the host institution will be assigned as the new principal investigator for the grant, and this is equivalent to winning a grant, without bothering to compete in the grant application pool. In the case of NIH grants, the only loser is the American taxpayer, who is saddled with paying for a non-productive grant, for the duration of the funding period.

3. Does the applicant have a friend on the study section? Applicants with little or no name recognition in the scientific community can redeem their credibility if they have a champion among their study section. A convivial member of a study section can exert great influence by discussing significant prior contributions from the applicant, and by prophesying the applicant's future productivity. Study section members are asked not to promote the efforts of applicants from their own institutions, but a shrewd study section member often looks for friends who live beyond his own back yard. In winter, a study section member who lives in Wisconsin, might dream of working in San Diego. An evil scientist never misses an opportunity to solidify friendships with influential academics from a desirable location. Exercising his powers of persuasion, he can push the score of a mediocre grant into the "funding" range. Once accomplished, how does the evil scientist take credit for his act? There is no effective way to maintain the confidentiality of study section discussions. It is commonplace for study section members, during their refreshment breaks, to get on their phones and spread the news. The evil scientist will place a call to the applicant, or to a close friend of the successful applicant, relating the thrilling tale of how he saved the grant, just as it was headed for oblivion.

4. Is the grant well written? Most grant applicants do not understand that a grant application is a work of literature, not a technical report. It's not sufficient to provide a correct but perfunctory response to the headings included in the template provided by the funding agency. Like all works of literature, your grant application must provide compelling reasons to read the first sentence, then the next sentence, and so on, until the grant is finished. The biggest mistake that most grant applicants make is to wait until the last possible moment to write the grant. Although it is possible to write a grant application at the last possible moment, it is impossible to re-write a grant application that was written at the last possible moment. Persuasive grants are written, then re-written, and re-written again, until the applicant is satisfied that the goals of the grant will become the goals of the reviewer. A thoughtfully crafted grant proposal is such a rare commodity that it will be likely receive a fundable score, on the basis of literary merit alone.

5. Does the grant address an important scientific problem, that has not been previously solved? This seems obvious, but many grant applications cannot meet this simple condition. A few bad grant applications address problems that have already been solved. More commonly, a grant application will address a problem that has not been solved previously, but which has no scientific importance (i.e., not much reason to solve the problem in the first place). Applicants must unburden themselves of the popular but misguided notion that all unsolved problems have potential merit. The burden of uncovering the merit in a research problem falls on the applicant. He must somehow persuade the study section that a problem that nobody has bothered to solve, throughout human history, is more worthy of funding than the projects offered by his competitors.

6. Does the grant application make any scientific sense? The "sense" of a research project is determined by the scientific paradigm that spawned the central hypothesis. For example, if you believe that there are four essences to the universe (i.e., earth, air, fire, and water), then you might find a project, that measures the power of water to overcome the power of fire, to be rather sensible. If you believe that aliens from other planets walk among us, you might want to fund a registry database that tracks the whereabouts of illegal extraterrestrials. Here's another example. In 2003, NCI set a goal of eliminating death from cancer by the year 2015 (1). In the ensuing years, NCI has not pushed back the expected delivery date. If you believe that cancer will be conquered by the year 2015, then it would make no sense to apply for a cancer research grant that extends into the year 2016 (after cancer has already been cured). For that matter, it would make no sense to fund NCI research grants awarded in the year 2011 with five years of funding; four will suffice. It's best to pander to the biases that prevail in the committee. If your reviewers believe in global warming, and that the oceans will rise 18 inches in the next decade, then you must believe the same. If they don't believe in global warming, then neither should you.

7. Is the grant budget realistic? Don't be cheap. If your grant budget is too small for the work proposed, the review committee will conclude that you don't know what you're doing (which you don't). If it's too large, they'll think the grant is simply a revenue-generating scheme (which it is). What is the correct amount of money to request for your grant? Nobody knows. If you think about it, you'll realize that it is impossible to predict the costs of research; nobody knows what techniques will work before they've actually worked. The best you can do is to plagiarize your budget with the budget of a successfully funded grant from your own institution.

Evil scientists should take note that honesty is not a criteria for grant success. It makes no sense to write a grant proposal that faithfully describes your intended research. The review committee will be composed by your competitors, who will give your grant a low score, and steal your ideas. Under these circumstances, honesty is the worst policy. Feel free to write ridiculous, misleading proposals; even if your grant if rejected, at least you'll have the satisfaction of leading your competitors astray.

Study section members are not endowed with the ability to peer into the future. They cannot determine which grants will yield major advances, and which grants will contribute nothing to science or society. The only thing that a study section member can do is to judge whether the grant meets the general criteria for a well-constructed grant (vida supra). Very few grant applications meet these criteria. Hence, very few grant applications deserve funding. Nonetheless, the available funds will be distributed to thousands of applicants. A Committee of Science and Public Policy report concluded that if you were to switch the review group for a set of grant applications, you would change the group of funded investigators by 25-30% (2). Agencies would save a great deal of money if they pooled the adequately written grants, and awarded funds by lottery. Deep down, everyone knows this.

REFERENCES

[1] Eschenbach AC. NCI sets goal of eliminating suffering and death due to cancer by 2015. Journal of the National Medical Association 95:637-639, 2003.


[2] Garfield E. Essays of an Information Scientist. Current Contents 5:3, Feb. 2, 1987.

- © 2011 Jules Berman

Free ebook update for Machiavelli's Laboratory

I've just prepared an updated ebook version of Machiavelli's Laboratory. It's available for free download at:

http://www.julesberman.info/integ/machfree.htm

Just click the link under the book title, near the top of the web page.

Anyone with a ebook reader should be able to read the MobiPocket file from any mobile device.

If you have a PC, and you've been reading the HTML or the PDF version of Machiavelli's Laboratory, I would highly recommend that you try reading the MobiPocket ebook version on your PC, using freely available ebook reader software.

I personally use MobiPocket desktop reader. It's free, easy to download and install, and provides a much better reading experience than HTML or PDF. Aside from the improved esthetics, The MobiPocket reader provides navigation and search features not available on the PDF version. The MobiPocket site is:

http://www.mobipocket.com/en/downloadsoft/productdetailsreader.asp

My understanding is that there are other "readers" for the PC (usually called desktop readers) widely available on the web.

- Jules J. Berman

Conflict declarations

• Machiavelli's Laboratory is a free ebook that I published on April 13, 2010. It is a satiric discourse on scientific ethics, from the perspective of an unethical scientist. Please don't take any of the advice and opinions in the book (or the excerpts featured in this blog) seriously.This blog entry continues yesterday's blog on Acknowledgments in scientific manuscripts

Editors have a bad record when it comes to ensuring that their published articles are free from financial conflicts (1), (2). Authors seldom volunteer their conflicts of interest, and editors seldom demand disclosures (3). Ensuring integrity is a time-consuming and thankless task. No editor has ever earned a dollar for his journal by uncovering a conflict of interest in a study. Basically, editors are in the business of selling journals, not advancing science.

Until recently, editors did not ask authors to disclose conflicts that might bias a manuscript's conclusions . Nobody really cared. Today, most journals have a "Conflicts" section, but statements are prepared in such a perfunctory fashion that their value is negligible.

Here is an example of about the best "Conflicts" section that you can hope to find in a manuscript.

The author(s) declare that they have no competing interests (4).

The sentence offers a clear statement from the authors, indicating that no conflicts exist. Of course, the editor offers no corroborating statement indicating that he has made a reasonable effort to validate the author's declaration.

Among the many scientific journals, there is no standard "Conflicts" statement. Some statements are barely worth reporting. Here is an example of a statement that could bear some improvement.

The authors declare no competing financial interests (5).

Here, the authors declare that they have no competing financial interests. Whether they have competing interests of a non-financial nature is anyone's guess.

Here's another.

The author reported no potential conflicts of interest (6).

Does this mean that they authors had actual conflicts; not potential conflicts? Or perhaps the statement conveys a completely different message. It may mean that conflicts exist, but they're all so very uninteresting.

Here's another

The authors of this manuscript have no conflict of interest to declare (7).

If I say, "I have no money to give you," it may mean that I have no money whatsoever, or it may mean that none of my money is held for the purpose of giving to you. In this "Conflicts" statement, the authors may have conflicts of interest, but none that they care to declare.

For brevity, the following cannot be beaten.

Competing interests: none declared (8).

Does this mean that the authors declared that they had no conflicts, or does it mean that the authors did not declare anything at all? Perhaps they forgot to respond to the editor's "Competing interests" query.

Why can't the editors of scientific journals do a little editing of their own and compose a useful "Conflicts" statement?

REFERENCES

[1] Smith R. Journals fail to adhere to guidelines on conflicts of interest. BMJ 323:651 (inclusive), 2001.

[2] Gross CP, Gupta AR, Krumholz HM. Disclosure of financial competing interests in randomised controlled trials: cross sectional review. BMJ 326:526-527, 2003.

[3] Hussain A, Smith R. Declaring financial competing interests: survey of five general medical journals. BMJ 323:263-264, 2001.

[4] Gressner OA, Weiskirchen R, Gressner AM. Evolving concepts of liver fibrogenesis provide new diagnostic and therapeutic options. Comparative Hepatology 6:7, 2007.

[5] Landgren O, Kyle RA, Pfeiffer RM. Monoclonal gammopathy of undetermined significance (MGUS) consistently precedes multiple myeloma: a prospective study. Blood 113:5412-5417, 2009.

[6] Slany RK. The molecular biology of mixed lineage leukemia. Haematologica 94:984-993, 2009.

[7] Odell A, Askham J, Whibley C, Hollstein M. How to become immortal: let me count the ways. Aging 2:160-165, 2010.

[8] Bauer AJ, Stratakis CA. The lentiginoses: cutaneous markers of systemic disease and a window to new aspects of tumorigenesis. J Med Genet 42:801-810, 2005.

- © 2011 Jules Berman

Acknowledgments

• Machiavelli's Laboratory is a free ebook that I published on April 13, 2010. It is a satiric discourse on scientific ethics, from the perspective of an unethical scientist. Please don't take any of the advice and opinions in the book (or the excerpts featured in this blog) seriously.

The Acknowledgment section of published papers provides public recognition for those people who contributed to the manuscript. It's a sort of consolation prize for people who weren't included as co-authors.

For those who have never paused to read the Acknowledgment section at the end of a research manuscript, here is a typical example:

"For advice and discussion, we thank L.F. Cai, C. F. Kim, T. Motoi, R. Hobbs, J. Clohessy, T. Yung, A. Carracedo, K. Ito, Pandolfi lab members, B. Clarkson, members of the MSKCC Lung Cancer Oncogenome Group, and members of the DF/HCC Lung Cancer Research Program. We thank M. Asher, T. Matos, and A. Egia for histology services and immunohistochemistry. We thank the MSKCC, University of Iowa, and Dana Farber Cancer Institute flow cytometry core facilities for technical assistance. This work was funded by National Institutes of Health/NCI grants (CA-64593) to P.P.P., and Steps for Breath Fund by the Society of MSKCC and the Thomas G. Labrecque Foundation to M.N." (1)

The authors acknowledged individuals, by name. They listed the laboratories and funding agencies that provided materials, funds, and other tangible support for the project. For their NIH grant, they listed the grant number, CA-64593. NIH expects grantees to reference grant numbers within their manuscripts. This permits administrators to track the research activities of their grantees.

All told, the typical Acknowledgment serves everyone but the authors. Certainly, an evil scientist should be able to turn the acknowledgment section into another opportunity for shameless self-promotion.

Let's look at a few Acknowledgment sections that work for the author.

ACKNOWLEDGMENTS. I thank students, postdocs, and colleagues for critical discussion and encouragement. I am grateful to the National Center for Biotechnology Information for providing the earliest dates associated with each NR sequence (2).

Here, the author has published an Acknowledgment wherein no person or persons are actually acknowledged. The Acknowledgment gives the impression that many people are vested in the success of the project, without actually giving credit to any individuals. By extending the effort to a large, anonymous group of professionals, the Acknowledgment adds a certain gravitas to the manuscript.

Here's an even better Acknowledgment.

ACKNOWLEDGMENTS. We thank the patients, physicians, and investigators who have contributed to our understanding of these disorders, and apologize to our colleagues whose work was not highlighted due to space constraints (3).

This Acknowledgment also fails to actually acknowledge any specific person. Instead, it provides an emotionally charged list of anonymous participants that includes every physician and investigator who has ever worked in the field, and every patient who has ever been afflicted by the investigated disease. Well done, but further improvement is possible.

ACKNOWLEDGMENTS. This work is dedicated to the many colleagues and patients who have contributed over the past several decades to advancing our understanding and treatment of amyloidosis, and to the staffs of the Amyloidosis Foundation and the International Myeloma Foundation whose work is ongoing and so often unheralded (4).

From the first sentence, the authors deftly replace their Acknowledgment with a Dedication. It's brilliant, really. With one bold stroke, the Acknowledgment is transformed from a message of indebtedness to a message of giving. As in the previous examples, no specific persons are actually mentioned in the acknowledgment. The manuscript becomes a gift given selflessly to all of the unnamed little people out there.

The Acknowledgment section has gradually evolved to a point where it's primary purpose is to serve the author and to add credibility to the manuscript. Can we dare hope to raise the level of the Acknowledgment to a higher realm? Yes, with God's help, we can.

Consider this remarkable introduction, wherein the author acknowledges God's role in the manuscript.

ACKNOWLEDGMENTS. I thank God for His grace and inspiration in this work..... (5).

The Acknowledgment continues by thanking a number of coworkers, by name. It ends by thanking the author's wife and his children "for the years of personal sacrifice that are their contribution to this project (5)". The final sentence of the Acknowledgment is another reference to God:

"May God bless them richly (5)."

How can any reviewer reject a manuscript that was inspired by God? To do so would be an act of heresy. Even the most hardened secular humanist will be moved by the years of personal sacrifice endured by the author's family.

So there we have it. The Acknowledgment section is much more than a simple afterthought, grudgingly included in the publisher's boilerplate. Always remember that a well-turned Acknowledgment can save your next manuscript from rejection.

Jump to tomorrow's blog on the Conflicts of Interest statements that appear in journal articles.

REFERENCES

[1] Berger AH, Niki M, Morotti A, et al. Identification of DOK genes as lung tumor suppressors. Nat Genet 42:216-223, 2010.

[2] Levitt M. Nature of the protein universe. Proc Natl Acad Sci USA 106:11079-11084, 2009.

[3] Levine RL, Gilliland DG. Myeloproliferative disorders. Blood 112:2190-2198, 2008.

[4] Comenzo RL. How I treat amyloidosis. Blood 114:3147-3157, 2009.]amyloid_tx.pdf

[5] Bidgood WD. Documenting the Information Content of Images. Proc AMIA Annu Fall Symp 424-428, 1997.


ACKNOWLEDGMENTS

I, Jules Berman, dedicate this blog entry to our universe, whose existence was crucial to the successful creation of this gift to humanity.

- © 2011 Jules Berman



key words: Acknowledgement, Acknowledgements

Another false first

• Machiavelli's Laboratory is a free ebook that I published on April 13, 2010. It is a satiric discourse on scientific ethics, from the perspective of an unethical scientist. Please don't take any of the advice and opinions in the book (or the excerpts featured in this blog) seriously.

Stigler's law of eponymy, "No scientific discovery is named after its original discoverer."

-SM Stigler

One of the most rigid rules of scientific discovery is that the person credited with being "the first" ..... is not.

Christopher Columbus was not the first European to lead an expedition to North America. Alexander Graham Bell was not the first person to invent the telephone. Guglielmo Marconi was not the inventor of radio. Joseph Fourier was not the first mathematician to invent what is now known as the Fourier transform. Einstein did not invent the field of relativity. In all these cases, the credited "firsts" were preceded by others, who made important contributions to the "final" discoveries.

In the field of cancer research, James Ewing (1866-1943) is credited with being the first to use the term "precancer" to describe the early and treatable phase of cancer development.[1] In a 1914 paper, Ewing wrote:[2]

"If inoperable advanced cancer is incurable, and localized cancer eradicable, the disease is preventable by dealing with its preliminary stages. Precancerous lesions are not cancers. Practically they differ enormously from the established disease. They can usually be removed by trivial or safe operations, and they are sometimes amenable to less violent treatment."

Thanks to Google's ngram viewer, we can actually test this assertion by searching for the years in which the term "precancerous" appears in the English literature.

Here is the result:


Click on image for larger view

"Precancerous" was used in 1849 (before Ewing was born) and 1876 (when Ewing was 10 years old) and throughout the latter 19th century (preceding Ewing's 1914 paper).

This only goes to show that ideas percolate through scientific communities before they bubble out as a finished idea. As Carl Sagan said (but perhaps he was not the first to say so), "If you want to make an apple pie from scratch, you must first create the universe."

REFERENCES

1. Cardiff RD. Borowsky AD.Precancer: sequentially acquired or predetermined? Toxicol Pathol 2010 38:171-179, 2010.

2. Ewing J. Precancerous diseases and precancerous lesions, especially in the breast. Medical Record 86, 951-958, 1914.

- © 2011 Jules Berman

Failure overtakes success

• Machiavelli's Laboratory is a free ebook that I published on April 13, 2010. It is a satiric discourse on scientific ethics, from the perspective of an unethical scientist. Please don't take any of the advice and opinions in the book (or the excerpts featured in this blog) seriously.

At the tail end of yesterday's blog, I inserted an output screen from Google's ngram viewer. Google collects an index of the occurrences of words and phrases (up to 5 words in length) found in its' collection of scanned literature. The lengths of phrases are ngrams (e.g., a 1-gram is a single word, a 2-gram is a two-word phrase, and so on).

At the ngram viewer page, anyone can enter a comma-separated list of words and phrases, and Google will produce a graph plotting occurrence frequency over time. Most of their scanned books come from the period following 1800, and Google's default starting data is 1800. You can look at earlier years, or you can confine your search to shorter periods of time.

You can use the ngram layer as a gauge of cultural honesty.

Let's look at the ngram that compares the frequency of "success" and "failure".


Click on image to see larger view

In the nineteenth century, "success" (the blue line) was about 6 times more frequent as "failure". By the end of the twentieth century, "failure" had overtaken "success" and is currently just shy of a tie.

What does this mean? It means that in the artifactual world of literature, our ancestors were somewhat reluctant to write about failure; not so today.

Are modern writers more honest than our predecessors? Have we evolved to the point where we can openly admit our failures?

Of course not. Modern writers have simply learned that they can openly criticize their competitors without incurring the scorn of their peers.

Let's look at the occurrences of the phrases "his failure" and "my failure"


Click on image to see larger view

The occurrence of "my failure" remains at a consistent and low level through the centuries. The occurrence of "his failure" enjoyed a meteoric rise through the mid-twentieth century. Basically, modern writers made the important discovery that they could blame someone else when things went wrong.

- © 2011 Jules Berman