Jeremy Howick

Philosophers of science have insisted that evidence of underlying mechanisms is required to support claims about the effects of medical interventions. Yet evidence about mechanisms does not feature on dominant evidence-based medicine “hierarchies.” After arguing that only inferences from mechanisms (“mechanistic reasoning”)—not mechanisms themselves—count as evidence, I argue for a middle ground. Mechanistic reasoning is not required to establish causation when we have high-quality controlled studies; moreover, mechanistic reasoning is more problematic than has been assumed. Yet where the problems can be overcome, mechanistic reasoning can and should be used as evidence.

Since its introduction just over two decades ago, evidence-based medicine (EBM) has come to dominate medical practice, teaching, and policy. There are a growing number of textbooks, journals, and websites dedicated to EBM research, teaching, and evidence dissemination. EBM was most recently defined as a method that integrates best research evidence with clinical expertise and patient values and circumstances in the treatment of patients. There have been debates throughout the early 21st century about what counts as good research evidence between EBM proponents and philosophical critics and even within the EBM community itself. Similar controversy arises about the relative worth of patient values and clinical expertise (and how these can be integrated). EBM has also evolved in ways that have come under scrutiny. Specifically, policymakers have used EBM research methodology to increase the relative importance of clinical guidelines that some clinicians have argued are tyrannical. Philosophers have addressed all of these controversies, and with very few exceptions have been critical of EBM. In addition most philosophical attention has been on the epistemic role of Randomization and evidence hierarchies, with relatively little attention being paid to the role of Diagnosis, expertise, patient values, and Systematic Reviews within EBM.

Unlike its friendly cousin the placebo effect, the nocebo effect (the effect of expecting a negative outcome) has been almost ignored. Epistemic and ethical confusions related to its existence have gone all but unnoticed. Contrary to what is often asserted, adverse events following from taking placebo interventions are not necessarily nocebo effects; they could have arisen due to natural history. Meanwhile, ethical informed consent (in clinical trials and clinical practice) has centred almost exclusively on the need to inform patients about intervention risks with patients to preserve their autonomy. Researchers have failed to consider the harm caused by the way in which the information is conveyed. In this paper, I argue that the magnitude of nocebo effects must be measured using control groups consisting of untreated patients. And, because the nocebo effect can produce harm, the principle of non-maleficence must be taken into account alongside autonomy when obtaining (ethical) informed consent and communicating intervention risks with patients.

AI, in the form of artificial carers, provides a possible solution to the problem of a growing elderly population Yet, concerns remain that artificial carers ( such as care-or chat-bots) could not emphathize with patients to the extent that humans can. Utilising the concept of empathy perception,we propose a Turing-type test that could check whether artificial carers could do many of the menial tasks human carers currently undertake, and in the process, free up more time for doctors to offer empathy.

Objectives Surveys in various countries suggest 17% to 80% of doctors prescribe ‘placebos’ in routine practice, but prevalence of placebo use in UK primary care is unknown. Methods We administered a web-based questionnaire to a representative sample of UK general practitioners. Following surveys conducted in other countries we divided placebos into ‘pure’ and ‘impure’. ‘Impure’ placebos are interventions with clear efficacy for certain conditions but are prescribed for ailments where their efficacy is unknown, such as antibiotics for suspected viral infections. ‘Pure’ placebos are interventions such as sugar pills or saline injections without direct pharmacologically active ingredients for the condition being treated. We initiated the survey in April 2012. Two reminders were sent and electronic data collection closed after 4 weeks. Results We surveyed 1715 general practitioners and 783 (46%) completed our questionnaire. Our respondents were similar to those of all registered UK doctors suggesting our results are generalizable. 12% (95% CI 10 to 15) of respondents used pure placebos while 97% (95% CI 96 to 98) used impure placebos at least once in their career. 1% of respondents used pure placebos, and 77% (95% CI 74 to 79) used impure placebos at least once per week. Most (66% for pure, 84% for impure) respondents stated placebos were ethical in some circumstances. Conclusion and implications Placebo use is common in primary care but questions remain about their benefits, harms, costs, and whether they can be delivered ethically. Further research is required to investigate ethically acceptable and cost-effective placebo interventions.

Introduction Current treatments for pain have limited benefits and worrying side effects. Some studies suggest that pain is reduced when clinicians deliver positive messages. However, the effects of positive messages are heterogeneous and have not been subject to meta-analysis. We aimed to estimate the efficacy of positive messages for pain reduction. Methods We included randomized trials of the effects of positive messages in a subset of the studies included in a recent systematic review of context factors for treating pain. Several electronic databases were searched. Reference lists of relevant studies were also searched. Two authors independently undertook study selection, data extraction, risk of bias assessment, and analyses. Our primary outcome measures were differences in patient- or observer-reported pain between groups who were given positive messages and those who were not. Results Of the 16 randomized trials (1703 patients) that met the inclusion criteria, 12 trials had sufficient data for meta-analysis. The pooled standardized effect size was −0.31 (95% confidence interval [CI] −0.61 to −0.01, p = 0.04, I2 = 82%). The effect size remained positive but not statistically significant after we excluded studies considered to have a high risk of bias (standard effect size −0.17, 95% CI −0.54 to 0.19, P = 0.36, I2 = 84%). Conclusion Care of patients with chronic or acute pain may be enhanced when clinicians deliver positive messages about possible clinical outcomes. However, we have identified several limitations of the present study that suggest caution when interpreting the results. We recommend further high-quality studies to confirm (or falsify) our result. FUNDING Alexander Mebius research has been funded through the ERC grant "Philosophy of Pharmacology: Safety, Statistical Standards, and Evidence Amalgamation" (GA: 639276)

Rationale and objectives: Medical humanities are becoming increasingly recognized as positively impacting medical education and medical practice. However, the extent of medical humanities teaching in medical schools is largely unknown. We reviewed medical school curricula in Canada, the UK and the US. We also explored the relationship between medical school ranking and the inclusion of medical humanities in the curricula. Methods: We searched the curriculum websites of all accredited medical schools in Canada, the UK and the US to check which medical humanities topics were taught, and whether they were mandatory or optional. We then noted rankings both by Times Higher Education and U.S. News and World Report and calculated the average rank. We formally explored whether there was an association between average medical school ranking and medical humanities offerings using Spearman's correlation and inverse variance weighting meta-analysis. Results: We identified 18 accredited medical school programmes in Canada, 41 in the UK, and 154 in the US. Of these, nine (56%) in Canada, 34 (73%) in the UK and 124 (80%) in the US offered at least one medical humanity that was not ethics. The most common medical humanities were medical humanities (unspecified), history, and literature (Canada); sociology and social medicine, medical humanities (unspecified), and art (UK); and medical humanities (unspecified), literature and history (US). Higher ranked medical schools appeared less likely to offer medical humanities. Conclusions: The extent and content of medical humanities offerings at accredited medical schools in Canada, the UK and the US varies, and there appears to be an inverse relationship between medical school quality and medical humanities offerings. Our analysis was limited by the data provided on the Universities' websites. Given the potential for medical humanities to improve medical education and medical practice, opportunities to reduce this variation should be exploited.

Debates about the ethics and effects of placebos and whether ‘placebos’ in clinical trials of complex treatments such as acupuncture are adequate rage. Yet there is currently no widely accepted definition of the ‘placebo’. A definition of the placebo is likely to inform these controversies. Grünbaum’s characterization of placebos and placebo effects has been touted by some authors as the best attempt thus far, but has not won widespread acceptance largely because Grünbaum failed to specify what he means by a therapeutic theory and because he does not stipulate a special role for expectation effects. Grünbaum claims that placebos are treatments whose ‘characteristic features’ do not have therapeutic effects on the target disorder. I show that with four modifications, Grünbaum’s definition provides a defensible account of placebos for the purpose of constructing placebo controls within clinical trials. The modifications I introduce are: adding a special role for expectations, insisting that placebo controls control for all and only the effects of the incidental treatment features, relativizing the definition of placebos to patients, and introducing harmful interventions and nocebos to the definitional scheme. I also provide guidance for classifying treatment features as characteristic or incidental

The philosophy of evidence-based medicine -- What is EBM? -- What is good evidence for a clinical decision? -- Ruling out plausible rival hypotheses and confounding factors : a method -- Resolving the paradox of effectiveness : when do observational studies offer the same degree of evidential support as randomized trials? -- Questioning double blinding as a universal methodological virtue of clinical trials : resolving the Philip's paradox -- Placebo controls : problematic and misleading baseline measures of effectiveness -- Questioning the methodological superiority of "placebo" over "active" controlled trials -- Examining the paradox that traditional roles for mechanistic reasoning and expert -- Judgment have been up-ended by EBM -- A qualified defence of the EBM stance on mechanistic reasoning -- Knowledge that versus knowledge how : situating the EBM position on expert clinical judgment -- Moving EBM forward.

Proponents of evidence-based medicine and some philosophers of science seem to agree that knowledge of mechanisms can help solve the problem of applying results of controlled studies to target populations (‘the problem of extrapolation’). We describe the problem of extrapolation, characterize mechanisms, and outline how mechanistic knowledge might be used to solve the problem. Our main thesis is that there are four often overlooked problems with using mechanistic knowledge to solve the problem of extrapolation. First, our understanding of mechanisms is often (and arguably, likely to remain) incomplete. Secondly, knowledge of mechanisms is not always applicable outside the tightly controlled laboratory conditions in which it is gained. Thirdly, mechanisms can behave paradoxically. Fourthly, as Daniel Steel points out, using mechanistic knowledge faces the problem of the ‘extrapolator’s circle’. At the same time, when the problems with mechanistic knowledge have been addressed, such knowledge can and should be used to mitigate (nothing can entirely solve) the problem of extrapolation

The feature of being ‘double blind’, where neither patients nor physicians are aware of who receives the experimental treatment, is universally trumpeted as being a virtue of clinical trials. The rationale for this view is unobjectionable: double blinding rules out the potential confounding influences of patient and physician beliefs. Nonetheless, viewing successfully double blind trials as necessarily superior leads to the paradox that very effective experimental treatments will not be supportable by best (double-blind) evidence. It seems strange that an account of evidence should make a priori judgments that certain claims can never be supported by ‘best evidence’. So far as treatments with large effects go, the claim that they are effective is highly testable and intuitively they should receive greater support from the evidence. In this paper I argue that the two potential confounders ruled out by double blinding are often not actual confounders outside placebo controlled trials of treatments with mild effects and that have subjective outcome measures.

This article critically analyses the concept of evidence in evidence‐based policy, arguing that there is a key problem: there is no existing practicable theory of evidence, one which is philosophically‐grounded and yet applicable for evidence‐based policy. The article critically considers both philosophical accounts of evidence and practical treatments of evidence in evidence‐based policy. It argues that both fail in different ways to provide a theory of evidence that is adequate for evidence‐based policy. The article contributes to the debate about how evidence can and should be used to reduce contingency in science and in policy based on science

Increasing philosophical attention is being directed to the rapidly growing discipline of evidence-based medicine. Philosophical discussions of EBM, however, remain narrowly focused on randomization, mechanisms, and the sociology of EBM. Other aspects of EBM have been all but ignored, including the nature of clinical reasoning and the question of whether it can be standardized; the application of EBM principles to the logic, value, and ethics of diagnosis and prognosis; evidence synthesis ; and the nature and ethics of placebo controls. Philosophical analysis in each of these areas has the potential to enhance the discussion of EBM methodology and practice.

A resilient issue in research ethics is whether and when a placebo-controlled trial is justified if it deprives research subjects of a recognized treatment. The clinicians' moral duty to provide the best available care seems to require the use of ‘active’ controlled trials that use an established treatment as a control whenever such a therapy is available. In another regard, ACTs are supposedly methodologically inferior to PCTs. Hence, the moral duty of the clinical researcher to use the best methods will favor PCTs. In this target article, I analyze the three reasons for believing that ACTs are inferior to PCTs namely: 1) ACTs lack ‘assay sensitivity’; 2) ACTs do not measure absolute effect size; and 3) ACTs require more participants; and I contend that none are acceptable. Consequently the tension between clinical and research ethics dissolves: the moral duty of the clinician to avoid PCTs is unopposed by methodological considerations.

The revised Declaration of Helsinki allows placebo-controlled trials to be used even when there is an established therapy, provided there are adequate ‘methodological’ reasons for doing so. This seems to violate the principle of beneficence: where there is an established therapy, physicians treating patients with a placebo are withholding a known effective therapy. Because of this problem, we hypothesised that clinical researchers may be unwilling to risk violating the principle of beneficence and employ placebo-controlled trials in cases where there is an established therapy. In this paper, we began to investigate this hypothesis. After summarising the arguments for and against using placebo controls in clinical practice, we exploredthe extent to which placebo-controlled trials are used in cases where there is an established therapy. To do this, we conducted as systematic search for all placebo-controlled trials published in 2015 in the five highest impact general medical journals. We identified 70 placebo-controlled trials. Of these, 66 were for indications where there was no established effective therapy. Only four used a placebo control in spite of there being an available effective therapy. The infrequent use of placebo controlled trials where established therapy exists highlights a seeming discrepancy between what the Declaration of Helsinki allows and what clinical investigators believe to be ethically acceptable. The evidence presented in this paper suggests that the Declaration of Helsinki be reconsidered, and perhaps revised, in light of actual practice.

A resilient issue in research ethics is whether and when a placebo-controlled trial is justified if it deprives research subjects of a recognized treatment. The clinicians' moral duty to provide the best available care seems to require the use of ‘active’ controlled trials that use an established treatment as a control whenever such a therapy is available. In another regard, ACTs are supposedly methodologically inferior to PCTs. Hence, the moral duty of the clinical researcher to use the best methods will favor PCTs. In this target article, I analyze the three reasons for believing that ACTs are inferior to PCTs namely: 1) ACTs lack ‘assay sensitivity’; 2) ACTs do not measure absolute effect size; and 3) ACTs require more participants; and I contend that none are acceptable. Consequently the tension between clinical and research ethics dissolves: the moral duty of the clinician to avoid PCTs is unopposed by methodological considerations.

Unlike its friendly cousin the placebo effect, the nocebo effect has been almost ignored. Epistemic and ethical confusions related to its existence have gone all but unnoticed. Contrary to what is often asserted, adverse events following from taking placebo interventions are not necessarily nocebo effects; they could have arisen due to natural history. Meanwhile, ethical informed consent has centred almost exclusively on the need to inform patients about intervention risks with patients to preserve their autonomy. Researchers have failed to consider the harm caused by the way in which the information is conveyed. In this paper, I argue that the magnitude of nocebo effects must be measured using control groups consisting of untreated patients. And, because the nocebo effect can produce harm, the principle of non-maleficence must be taken into account alongside autonomy when obtaining informed consent and communicating intervention risks with patients. There are no data in this work.

In August 2011, a group of medical doctors, ethicists, academic and medical physicists were asked to debate and reach consensus on the potential need for randomised control trials to test charged particle radiation therapy for treating tumours. The outcome of the meeting was a paper recently published in the Journal of Medical Ethics entitled “Position statement on ethics, equipoise and research on charged particle therapy” by Sheehan et al. However 6 of the 30 meeting participants withdrew from authorship of the ‘position statement’ because their views were not adequately represented. The ‘position statement’ did not state our reasons for withdrawing from the statement, which is a considerable omission. We had two principal objections: the case for the benefits to patients and society of randomized trials to test CPRT was not adequately represented, and the complexities and potential harms of CPRT were not clearly stated. In this response we explain and justify our objections. Patients, doctors and policymakers seeking to make independent judgments about whether equipoise exists for the relative benefits of CPRT should therefore read this document alongside the ‘position’ statement