Risk, Chance, and Causation

Risk, Chance, and Causation: Investigating the Origins and Treatment of Disease

MICHAEL B. BRACKEN
Copyright Date: 2013
Published by: Yale University Press
Pages: 288
https://www.jstor.org/stable/j.ctt32bxq3
  • Cite this Item
  • Book Info
    Risk, Chance, and Causation
    Book Description:

    The press and other media constantly report news stories about dangerous chemicals in the environment, miracle cures, the safety of therapeutic treatments, and potential cancer-causing agents. But what exactly is actually meant by "increased risk"-should we worry if we are told that we are at twice the risk of developing an illness? And how do we interpret "reduced risk" to properly assess the benefits of noisily touted dietary supplements?

    Demonstrating the difficulty of separating the hype from the hypothesis, noted epidemiologist Michael Bracken clearly communicates how clinical epidemiology works. Using everyday terms, Bracken describes how professional scientists approach questions of disease causation and therapeutic efficacy to provide readers with the tools to help them understand whether warnings of environmental risk are truly warranted, or if claims of therapeutic benefit are justified.

    eISBN: 978-0-300-18955-1
    Subjects: Health Sciences, History

Table of Contents

  1. Front Matter
    (pp. i-vi)
  2. Table of Contents
    (pp. vii-viii)
  3. ACKNOWLEDGMENTS
    (pp. ix-x)
  4. List of Abbreviations
    (pp. xi-xiv)
  5. ONE Risk, Chance, and Causation: Investigating the Origins and Treatment of Disease
    (pp. 1-14)

    Bertrand Russell was, of course, referring to God’s existence when he decried a lack of evidence, but he might just as well have been referring to the absence of reliable evidence pertaining to risk in many aspects of everyday life and the difficulty that scientists have in producing reliable estimates of risk or, indeed, assessing whether there is any risk at all. How do scientists attempt to separate chance associations, the links between environmental agents and disease, from real ones? And how is an association, once demonstrated, determined to be causal? The public is bombarded on a daily basis by...

  6. TWO Chance and Randomness
    (pp. 15-34)

    Each February America celebrates its national winter pastime with an excess of testosterone-fueled activity known as the Super Bowl, when the winners of the two football conferences, National and American, compete for top dog. Interestingly, the National conference has won the Super Bowl coin toss 13 consecutive times (1998 to 2010), a result likely to happen only once in 8,192 times. Unless you believe that the NFC was able to doctor the coin, or you are convinced your dropped toast always lands with the butter side down, this has to be considered a chance result. How can that happen? If...

  7. THREE Risk
    (pp. 35-50)

    The first two days of our cruise saw little wind and an uneventful passage east from our home port in Casco Bay, first to Christmas Cove in South Bristol a favored sailors’ watering hole with a bar full of ancient burgees from yacht clubs worldwide, and the next day across Muscongus Bay to Allen Island, home of the Wyeth painters, and into Penobscot Bay. Only as we sailed north up Muscle Ridge Channel in the early afternoon did we see a fog bank rolling in from the south, not an unusual occurrence and not of concern as Dix Island was...

  8. FOUR Randomization and Clinical Trials
    (pp. 51-69)

    It is appropriate to begin investigating the methods used to demonstrate causality by considering the type of study design that most readily lends itself to doing so. If we wish to determine whether or not a drug therapy or some other type of medical maneuver causes an illness to improve or, alternatively, causes harm, we use a randomized controlled trial (RCT). This study design is often thought of as the gold standard for medical and social research. Here I discuss it as the touchstone against which other research methodologies will be compared. In subsequent chapters I will describe the biases...

  9. FIVE More Trials and Some Tribulations
    (pp. 70-85)

    Sometimes a traditional clinical trial, the kind discussed in the previous chapter, where patients are individually randomized to one treatment or another, will not work. If patients discuss their experimental treatment with other patients who are in the control group, the latter may then demand to be treated with the “new” treatment. Similarly doctors, on learning about a trial being run in their hospital, may prematurely adopt the experimental therapy making it impossible for their patients to join the trial. In fact, the majority of trials do not show benefit from the experimental therapy, and rushing to adopt the new...

  10. SIX Harm
    (pp. 86-112)

    It is considered unethical to randomize people to test treatments with the intent to see if exposure to a drug or chemical will cause them harm. Institutional Review Boards (IRBs), the present-day arbiters of ethical research practice, would have shown Goldberger the door if he had gone to them requesting permission to randomize people (let alone prisoners) to a diet that might induce pellagra. Fortunately for him, the prisoners who earned their release from prison, and us, Goldberger did not have to obtain consent from a modern IRB. If the principal goal of a study is to see whether a...

  11. SEVEN Screening, Diagnosis, and Prognosis
    (pp. 113-126)

    Not too long ago, the American Cancer Society strongly recommended women practice breast self-examination and, if over age 40, have an annual mammography. For reasons that may seem obvious, early detection of a breast lump that may be cancerous allows more effective treatment and better chance of survival. Case closed? Screening for other conditions, prostate cancer, high blood pressure, high cholesterol, high glucose, and anemia, is now an integral part of modern medicine and health care. But is it that simple: is screening really effective? Are there risks of screening and any possible unintended consequences? Diagnostic and prognostic tests are...

  12. EIGHT A Statistical Sojourn
    (pp. 127-143)

    Causation is not demonstrated or proven by statistical analysis alone; the process for deciding causality is in part empirical, but it is also largely based on the strength of the methodology of the studies from which the data was derived. We have discussed how RCTs lead directly to causation conclusions. There is also an experiential element to recognizing causation; past experience informs us as to what types of preliminary evidence may, or are unlikely to, lead to a causation conclusion. If statistics gives you heartburn by all means skip this chapter, but I encourage you to give it a try;...

  13. NINE Disease Clusters
    (pp. 144-153)

    Some of my earliest memories are of sitting in the Morrison bomb shelter in our living room during air raids.¹ The shelter was designed to protect us so that we could be safely dug out should our house be hit by a bomb. I was born in Bradford, in the English county of Yorkshire, during World War II. The German Luftwaffe flew over us on its way to bomb more important targets in the west, particularly Manchester’s industrial center and the Liverpool docks. When the Royal Air Force intercepted the Luftwaffe the Germans would think it more prudent to drop...

  14. TEN Genetics and the Genome
    (pp. 154-170)

    Philip Larkin, considered one of Britain’s greatest poets of the twentieth century, was right but not for the reasons he thought. It is increasingly recognized that many of the diseases that afflict us are of genetic origin, inherited from our “mum and dad.” Alzheimer’s disease, autism, cancer, hypertension, asthma, all have a strong genetic component to their cause. For some diseases, it has been known for many years that the condition was entirely owed to a genetic origin. Trisomy 21, also known as Down syndrome, is caused by an extra complete chromosome (number 21) that entirely explains the syndrome, although...

  15. ELEVEN The Study of Mankind Is Man: Reflections on Animal Research
    (pp. 171-184)

    On Saturday, May 25, 1940, in Oxford at 11:00 a.m. Norman Heatley administered a carefully measured dose of Streptomyces pyogenes to eight white mice in what would become one of the most famous animal experiments of all time. England was at war, and there were days when invasion seemed imminent and bombing a certainty. Pristine college lawns had been converted to growing vegetables, and the university was blacked out so that no lights shone through the laboratory windows on South Parks Road. At 12:00 noon two mice were given 5ml of penicillin Heatley had extracted from mould juice, two mice...

  16. TWELVE Celebrity Trumps Science
    (pp. 185-199)

    While I sat writing on a beautiful lakeside in Patagonia, the tranquility was shattered by the arrival of a crew filming a TV reality show, the name of which will remain anonymous to protect the guilty. The idea behind the show was unclear, but what we saw were several kayaks being paddled toward the beach where four large polystyrene figures had been erected. Mapuche idols, we were told, but they looked like a cross between Easter Island and the Northwest totems. The local Mapuche people, now generally well assimilated into modern Argentine life, would have been bewildered. After a leisurely...

  17. THIRTEEN Replication and Pooling
    (pp. 200-223)

    To Karl Popper, one of the most influential philosophers of science of the twentieth century, verification and falsifiability were central to the scientific enterprise. It was, he thought, only through replication that the truthfulness of a scientific observation could be assured or rejected. There are two types of replication. One seeks to replicate everything in the original experiment: the study participants, what was done to them, and which study outcome is assessed. All are repeated in the second experiment to see if the same result ensues. The second type of replication varies the experimental condition, perhaps doing the experiment in...

  18. FOURTEEN Bias in Publication and Reporting
    (pp. 224-238)

    Traditionally, scientific knowledge has been obtained by developing hypotheses that can be tested in a study or experiment designed to produce data that will either support or refute it. As Huxley satirically put it in his presidential address to the British Association for the Advancement of Science, science is littered with beautiful but failed hypotheses. In the modern medical sciences, we are less reliant on the results of a single study, preferring to see replication in several of them and their confirmation within the framework of a systematic review and meta-analysis. Nonetheless, the scientific paradigm of hypothesis testing prevails: the...

  19. FIFTEEN Causes
    (pp. 239-259)

    In January 2011, Jared Loughner shot and killed 6 people and critically wounded 13 others, including Representative Gabrielle Giffords, and the editorial pages were full of angst and commentary as to what had caused the tragedy. Supporters of the gun lobby were quick to fault the mental health system, schools were criticized for not reporting the killer’s early signs of sickness, political rhetoric was blamed and, of course, his parents. God was praised for causing Ms. Giffords’s remarkable recovery but not for preventing the shooting in the first place. An obvious cause was rarely mentioned: a mentally ill person with...

  20. SIXTEEN Ultimate Causation
    (pp. 260-276)

    Goldberger, Snow, Doll, and many other public health heroes are celebrated for demonstrating causal relationships between dietary deficiencies, polluted drinking water, cigarette smoking, and a host of other factors shown to cause serious disease. Their discoveries were done at a time when it was not known what was ultimately causing the diseases they studied; there were no blindingly obvious labels about disease-causing products, such as a warning on a mustard jar that it contained mustard. Nonetheless, once shown, or in some cases strongly suspected, that an exposure was causal, it was possible to implement relatively straightforward public health actions: adding...

  21. NOTES
    (pp. 277-302)
  22. BIBLIOGRAPHY AND FURTHER READING
    (pp. 303-304)
  23. INDEX
    (pp. 305-330)