Mednotes

L2 Ethics History of medicine, explanation and the biopsychosocial model

Properties2
tagsethics, lecture
aliasesL2 Ethics History of medicine, explanation and the biopsychosocial model

19 min read

Overview

The lecture uses the ME/CFS controversy to introduce the biopsychosocial model and contrast it with the biomedical model, then asks what makes an explanation “scientific.” Historical case studies (the London cholera outbreaks and Semmelweis’s work on puerperal fever) are used to build Hempel’s deductive-nomological model (DN) of explanation. Dennett’s three explanatory stances, including the intentional stance, are then used to argue that psychological and social-level explanations can be just as rigorous and predictively successful as physical-level ones, answering objections to biopsychosocial explanations of conditions like ME/CFS. The lecture closes on pseudoscience: how to distinguish it from legitimate (if sloppy) science, and how it operates in medicine via Gish galloping and disease mongering, illustrated by fluoridation and female sexual dysfunction case studies.

Biopsychosocial model vs the biomedical model

  • Framing quote (Monbiot, Guardian, 2024): ME/CFS patients (about 250,000 in the UK) have historically been told they can improve by “changing their attitudes”; there is no diagnostic test or validated treatment; the piece calls this psychologisation “the greatest medical scandal of the 21st century.”
  • Engel (1977): biomedical dogma holds that all disease must be conceptualised as derangement of physical mechanism, permitting only two alternatives — reductionist (all behavioural phenomena of disease reduced to physiochemical principles) or exclusionist (whatever cannot be so explained is excluded from the category of disease).
  • Engel’s alternative: understanding the determinants of disease and arriving at rational treatment requires also taking into account the patient, the social context of their life, and the health-care system society has devised to manage illness — i.e. a biopsychosocial model.
  • The core problem identified with the biomedical model: it insists explanations of disease must occur at the level of physical mechanism only.
  • Related distinctions raised: ordinary vs scientific explanation; explanation (why) vs description (what); and whether scientific explanation really requires reduction to the physical level.

Historical case studies and the deductive-nomological model of explanation

London cholera (miasma theory vs Snow). Serious outbreak in 1832; the 1849 outbreak killed 14,317. Miasma theory held cholera was caused by “bad air” (miasmata) from decayed organic matter, passed between people, eradicated by cleansing/scouring; soil at lower elevations was thought to hold more miasmata. An 1849 London chart plotting deaths per 10,000 against elevation above the Thames showed mortality falling steeply from ~100 near river level to under 20 by ~120 feet, then levelling off — but both “observed mortality” and “mortality calculated on miasma theory” tracked closely, so this data alone did not distinguish the two theories. At the Broad Street pump, the only anomaly was the brewery, whose workers drank beer rather than water and were unaffected; Snow repeated the pattern with other populations (e.g. prisons). This was before germ theory (Pasteur, 1861); Snow hypothesised ingestion of a substance caused cholera, without yet explaining it at the physiochemical level.

Semmelweis and puerperal fever. Physician at Vienna General Hospital, 1844–1848. First Maternity Division: 260 of 3157 delivering mothers died of puerperal fever (8.7%). Second Maternity Division: 2.3% mortality. [flag: the slide gives the years for these statistics as “1884” and “1998”, evidently typos for 1846/1848 given the historical context; transcribed as shown, not corrected] Semmelweis tested and refuted a series of hypotheses:

  • “Atmospheric, cosmic-telluric changes” — inconsistent, since it could not explain the difference between the two divisions of the same hospital, and women who gave birth in the street before reaching hospital had a lower death rate than those delivering in the First Division.
  • Overcrowding and diet/general care — rejected because the Second Division was equally crowded.
  • An 1846 Commission blamed “rough handling” by medical students; Semmelweis countered that childbirth itself causes more injury than rough handling, that midwives performed the same examinations without the same mortality, and that halving the number of students briefly halved infections before they rose again.
  • Delivery position (side vs back) — tested directly; infections continued regardless.
  • Visiting priests (5 doors and a ringing bell to reach the First Division vs 1 door for the Second) — refuted by rerouting priests via a different, silent door with no effect on mortality.

In 1847 a colleague was punctured by a scalpel used in an autopsy and died in agony with symptoms matching puerperal fever. Semmelweis reasoned that cadaveric matter was the cause, since he, his colleagues and students went directly from the dissection rooms to the wards. He tested this by handwashing with chlorinated lime; 1848 mortality fell to 1.27%. This explanation additionally accounted for why midwives (who did not perform autopsies) had lower rates, why street births fared better, and why only infants of already-infected mothers became infected — illustrating that a good scientific explanation typically has “explanatory power,” explaining other related phenomena as well.

Semmelweis both refuted hypotheses that conflicted with observed facts (as Snow had) and used predictive testing: derive what should happen if a hypothesis (H) is true, then check. If the predicted observation (I) does not occur, H is false — modus tollens (“If H is true, then so is I. But I is not true. Therefore H is not true.”).

The Deductive-Nomological (DN) model (Hempel, 1966). Scientific explanations have:

  • Explanatory relevance: someone with no prior knowledge of puerperal fever or the Broad Street outbreak could, once given the explanation, accept that the outcome would happen.
  • Testability: e.g. removing the pump handle, handwashing, preventing dissection contact.
  • “Deductive”: the explanandum (E) logically follows from the laws (L) and facts (C).
  • “Nomological”: lawlike — the regularity in nature is what licenses the deduction.
  • Structure: Laws (L1, L2, …) + Facts (C1, C2, …) = Explanans → entails → E (Explanandum).

Worked examples given: (1) Westworld — Law: rational people threatened by violent psychopaths avoid them. Facts: Deloris sees the man in black, has partial memories of him being violent to her, and is rational. Conclusion: Deloris cries and hides from him because he might harm her. (2) Cholera — Law: living closer to the Thames increases the likelihood of contracting cholera; taking water from the Broad Street pump infects with cholera. Facts: the Broad Street pump was supplied by a different company than other pumps and drew water from the Thames where sewage was dumped; there was no other factor common to all who contracted cholera in that area. Conclusion: cholera is caused by something in the water, not miasma. The point of these examples is that psychological/social laws can support the same deductive-nomological structure as physical laws — explanation does not have to reduce to the physical level to be scientific.

The “Semmelweis effect.” His claims were met with tolerance at best and his contract was not renewed in 1849. Contributing factors: he was Hungarian working in Austria; his explanation implicated himself and his colleagues in causing deaths; his explanation was epidemiological rather than physiological (this was pre-germ theory, so it lacked a physical mechanism); and handwashing took time while chlorinated lime irritated hands.

The ME/CFS controversy: an objection to biopsychosocial explanation

  • There are currently no investigative tools or physical signs to confirm or refute CFS, so clinicians must decide how long to look for alternative explanations for fatigue before diagnosing CFS; depression is the most common comorbid condition identified, so a mental-state examination is described as often the most productive single early investigation for unexplained fatigue (Harvey & Wessely, 2009).
  • Harvey and Wessely’s proposed causal pathway: predisposing factors (female, overactive, prior psychiatric disorder, personality factors, early childhood illness, increased doctor use) → a triggering event (viral infection, stress, or other) → fatigue → sustained by maintaining factors (prolonged bed rest, boom-and-bust activity pattern, biological changes) → CFS.
  • PACE trial (White et al., Lancet 2011): found cognitive behaviour therapy (CBT) and graded exercise therapy (GET) could be effective treatments for CFS, but patient organisations reported these treatments could be harmful and preferred pacing and specialist medical care instead.
  • Ongoing controversy: pressure grew on the Lancet to review the “flawed” PACE trial (BMJ, 2018); a UK research watchdog concluded in 2019 the trial had been properly conducted; a 2021 BMJ piece situated the CFS debate alongside long covid.
  • Professor Simon Wessely received the John Maddox Prize for speaking out on ME research despite a “hate campaign and death threats,” but critics argued his work perpetuates the idea that ME is a psychiatric condition and trivialises what they see as a largely physical illness; Wessely stated he has never claimed CFS is “all in the mind” and has published on possible infectious triggers.

The stakes: labelling a condition as psychosocial is experienced by some patients and advocates as dismissive or harmful, even though — per the DN model and the explanatory-stances framework below — a psychosocial-level explanation need not be less scientific or less legitimate than a purely biomedical one.

Levels of explanation: the intentional stance and three explanatory stances

  • The same object or event can be explained at different levels of description — e.g. a key: an alloy, a metallic object, a machined tool, an opener of doors, a key, a hotel-room key, a means of opening a door, a facilitator of illicit liaisons.
  • Dennett’s three explanatory stances (The Intentional Stance):
    • Physical stance: prediction from the level of physical mechanism, e.g. an AI system predicted at the level of physics (0/1 switches); a disease predicted at the cellular level, e.g. hypertension.
    • Design stance: prediction from what something was designed to do, e.g. an AI system predicted on the basis of its design; a disease predicted by design, e.g. laryngospasm.
    • Intentional stance: prediction from beliefs, desires and rationality, e.g. an AI system predicted on the basis of its beliefs and desires (though the lecture notes this is complicated by generative AI); a disease predicted on the basis of beliefs and desires, e.g. asbestosis.

The design-stance slide illustrates the idea with a generic engineering/spacecraft-subsystem schematic; its specific component labels are not explained in the lecture and are not interpreted further here — the diagram is only an example of "prediction from design," not itself lecture content.

  • “Martians and Westworld” thought experiment: Martians who understand the world only via the physical stance, observing humans, would still be struck by how efficiently we predict the behaviour of Westworld hosts and by how efficiently we explain why people become diseased, using beliefs and desires — i.e. psychological and social explanations can be efficient, thorough, and predictively successful, supporting their legitimacy as explanations even though they are not physical-level.

Pseudoscience: distinguishing features and historical background

  • 1847 AMA Code of Medical Ethics: told patients not to weary physicians with irrelevant detail and to feel a lasting sense of gratitude/obligation toward their physician — illustrating an older, paternalistic professional culture operating alongside quackery — and warned that other professionals (e.g. clergy) sometimes lent credibility to “medical empirics” by endorsing nostrums.
  • OED definition: pseudoscience is “a pretended or spurious science; a collection of related beliefs about the world mistakenly regarded as being based on scientific method or as having the status that scientific truths now have.” This yields a two-part test: (1) not scientific; (2) its proponents try to create the impression that it is.
  • Three biochemist “problem cases” used to refine the test:
    • Case 1: a single result later judged by colleagues to be an artefact of experimental error — ordinary scientific error, not pseudoscience.
    • Case 2: repeated sloppy experiments consistently interpreted as supporting an unorthodox, rejected claim about a protein’s role in muscle contraction.
    • Case 3: various sloppy experiments across different areas (including Case 1’s), generally poor-quality work, but with no particular unorthodox theory being promoted.
  • Refined definition (from Case 2): pseudoscience is (1) not scientific, and (2) part of a non-scientific doctrine whose proponents try to create the impression it is scientific — distinguishing genuine but sloppy science (Cases 1 and 3) from pseudoscience proper (Case 2, which persists in promoting an unaccepted doctrine while presenting it as scientific).

Pseudoscience in practice: fluoridation, Gish galloping and disease mongering

  • Fluoridation is used as a worked example of anti-scientific rhetoric: alarmist imagery framing fluoride as an EPA-classified “hazardous waste” and comics warning of toxicity, contrasted with mainstream public-health consensus. Litigated in New Zealand courts: New Health New Zealand Inc v South Taranaki District Council went through the High Court (2014) and Supreme Court (2017). Advocacy groups such as the NZ Health Trust (founded 2002) promote “natural health choices” and lobby against health-related law changes, framing this as protecting individual choice.
  • Gish galloping (named for Duane Gish, biochemist and former VP of the Institute for Creation Research): overwhelming an opponent with a flood of weak or fallacious claims so that rebutting the whole argument requires disproportionate effort. Example: a 700-word, unreviewed anti-fluoridation commentary in the Journal of Primary Health Care made 15–20 claims; the published rebuttal (Broadbent, Wills, McMillan, Drummond & Whyman) took about 12,000 words, three rounds of peer review, five authors and two legal opinions.
  • Disease mongering: “trying to convince essentially well people that they are sick, or slightly sick people that they are very ill” (Payer, 1992) — ordinary problems reframed as medical problems. Examples: baldness was linked with panic and emotional difficulty in publicity prior to the launch of the drug Finasteride; the launch of Roche’s antidepressant Aurorix coincided with news reporting that a million Australians had underdiagnosed “social phobia.”
  • Disease mongering and female sexual dysfunction, a worked timeline:
    • 1992: NIH consensus statement recommended male impotence be termed “erectile dysfunction.”
    • 1998: FDA approved Viagra for erectile dysfunction.
    • 1997–2004: Pfizer trials of Viagra for “female sexual arousal disorder” were inconclusive on benefit.
    • 2015: FDA approved flibanserin — originally developed as an antidepressant, modulating serotonin and dopamine — for hypoactive sexual desire disorder (HSDD) in premenopausal women, despite concern about a suboptimal risk-benefit trade-off; trial endpoints were measured at 4 weeks.
    • RCT result: on flibanserin, “satisfying sexual events” rose from 2.8 to 4.5 per month, versus 2.7 to 3.7 per month on placebo.
    • The lecture’s closing question on this example: “What were they treating?” — raising whether HSDD was a constructed category fitted to a marketable drug rather than a discovered clinical entity.

Self-test

  1. Define the biopsychosocial model and explain how Engel’s 1977 critique of the biomedical model led to it.
  2. Explain why the mortality-by-elevation data from the 1849 London cholera outbreak did not, by itself, distinguish between miasma theory and Snow’s water-borne explanation.
  3. Describe the sequence of hypotheses Semmelweis tested and refuted before concluding that cadaveric matter caused puerperal fever.
  4. Explain why Semmelweis’s cadaveric-matter hypothesis had greater “explanatory power” than the rival hypotheses he had rejected.
  5. Describe the modus tollens logical structure Semmelweis used to test hypotheses, using the delivery-position hypothesis as an example.
  6. List the four features of a scientific explanation in Hempel’s deductive-nomological (DN) model.
  7. Using the DN model’s structure of laws and facts, reconstruct the Broad Street pump case as a scientific explanation of the cholera outbreak.
  8. Distinguish Dennett’s physical, design and intentional explanatory stances, giving one example of each from the lecture.
  9. Explain what the “Martians” thought experiment is meant to show about explanations at the psychological/social level.
  10. Give three factors that, according to the lecture, contributed to the poor reception of Semmelweis’s explanation (the “Semmelweis effect”).
  11. Describe the causal pathway for CFS proposed by Harvey and Wessely (2009), including its stages.
  12. Explain why the PACE trial (White et al., 2011) became controversial, referencing both its findings and the criticisms of it.
  13. State the two-part test for pseudoscience derived from the OED definition, and explain how the “Case 2” biochemist vignette differs from Cases 1 and 3 under this test.
  14. Define Gish galloping and describe the numerical asymmetry between the original anti-fluoridation article and its published rebuttal.
  15. Define disease mongering (Payer, 1992) and describe two examples given in the lecture.
  16. Trace the regulatory history from “erectile dysfunction” (1998) through the female sexual arousal disorder trials to the approval of flibanserin for HSDD (2015), including the RCT outcome for its key endpoint.
  17. Explain how the DN model and Dennett’s three stances together support the claim that framing ME/CFS via the biopsychosocial model need not be less scientific than a purely biomedical explanation.

Answers

Graph View

Backlinks