Philip Sedgwick

Bio: Philip Sedgwick is an academic researcher from St George's, University of London. The author has contributed to research in topics: Randomized controlled trial & Placebo. The author has an hindex of 34, co-authored 241 publications receiving 4659 citations.

Understanding the Hawthorne effect

Abstract: Researchers investigated the effectiveness of patient controlled analgesia for patients presenting to emergency departments with pain from traumatic injuries. A randomised controlled trial with a parallel groups study design was used. The control treatment was usual care, with analgesia titrated by nurses according to hospital guidelines. The participants were patients aged 18-75 years presenting to the emergency department who required intravenous opioid analgesia for the treatment of moderate to severe pain resulting from traumatic injuries, and who were expected to be admitted to hospital for at least 12 hours. In total, 200 adults were recruited and randomised to the intervention (patient controlled analgesia; n=99) and control (routine care; n=101) treatment groups.1 The primary outcome was total pain experienced over the 12 hour study period, recorded using a visual analogue scale. Secondary outcomes included total amount of morphine used. The mean total pain experienced by the intervention group was lower than that experienced by the control group, although the difference was not significant. Participants in the intervention group used significantly more morphine than was administered in the usual care group. It was concluded that patient controlled analgesia provided a reduction in pain, albeit not significant, compared with routine care for emergency department patients with traumatic injuries. The researchers commented that the results of the trial may have been influenced by the Hawthorne effect. Which of the following statements, if any, are true?

Pearson’s correlation coefficient

Abstract: Researchers investigated the relation between the number of involuntary admissions (detentions) for mental disorders a year under the Mental Health Act 1983 and the number of NHS psychiatric beds each year in England. They used hospital episode statistics from 1996 to 2006 in a retrospective analysis. For each year they obtained the number of available NHS psychiatric beds—defined as those beds for patients with mental disorders or learning disabilities—and the number of involuntary admissions for mental disorders in NHS hospital and private facilities combined.1 It was reported that the number of NHS psychiatric beds fell in each successive year and that overall from 1996 to 2006 the number had decreased by 29%. A significant correlation existed between the number of psychiatric NHS beds each year and the combined number of involuntary admissions for mental disorders to NHS and private facilities under the Mental Health Act 1983 (Pearson correlation coefficient r =−0.94 (P<0.001)). Which of the following statements, if any, are true? Statements a and b are true, while c and d are false. The Pearson correlation coefficient measures the strength of linear association between two variables (statement a is true)—in the …

Cross sectional studies: advantages and disadvantages

Abstract: A cross sectional study design was used to investigate the extent of chronic fatigue and the associated psychosocial exposures in a developing country. The setting was a primary health centre catchment area in Goa, India. Participants were women aged 18-50 years. The primary outcome was reporting of fatigue for at least six months. Data on the primary outcome and psychosocial exposures were obtained by a structured interview. The psychosocial exposures that were investigated included mental health and gender disadvantage factors. The presence of anaemia was determined from a blood sample.1 The sampling frame consisted of 8595 eligible women listed in family health registers. In total, 3000 randomly sampled women were invited to participate and 2494 (83%) consented. Recruitment took place from November 2001 to May 2003. It was reported that 12.1% (95% confidence interval 10.8 to 13.4%) of women experienced chronic fatigue. Those psychosocial exposures most strongly associated with chronic fatigue were poor mental health and sexual violence by the husband. Which of the following statements, if any, are true?

Prevalence and incidence

Abstract: A randomised controlled trial investigated whether screening and treating women for chlamydial infection reduced the subsequent occurrence of pelvic inflammatory disease.1 Between 2004 and 2006, vaginal swabs were provided by 2529 sexually active female students. The samples were randomly allocated to immediate testing (screening group), or storage and deferred screening after a year (control group). After screening, treatment for chlamydial infection was offered to women where necessary. All women were able to undertake independent testing for chlamydia during the follow-up period of one year. At baseline, 5.4% of the women in the screened group tested positive for chlamydia compared to 5.9% of the controls. Over the following twelve months, 1.3% of the screened women developed pelvic inflammatory disease compared with 1.9% of the controls. Which of the following statements, if any, are true?

Spearman’s rank correlation coefficient

Abstract: Researchers examined the association between trends in antidepressant prescribing and suicide rates between 1991 and 2000 in Australia.1 A retrospective analysis of national databases was undertaken. Participants were aged 15 years or more. The primary outcomes were trends in suicide rates and antidepressant prescribing, according to sex and 10 year age groups. The trend in suicide within each age group was measured by the difference between the suicide rates per 100 000 people in two five year periods (1986-90 and 1996-2000). Trends in antidepressant prescribing were assessed by the change in defined daily dose per 1000 days, as indicated by the difference between 1991 and 2000. A positive trend in suicide rates or antidepressant prescribing within an age group represented an increase from 1991 to 2000. The researchers reported that although overall national rates of suicide did not fall significantly, the incidence decreased in older men and women and increased in younger adults. Rates of antibiotic prescribing increased across all age groups in both men and women. The association between trends in suicide rates and antidepressant prescribing were measured by Spearman’s rank correlation coefficient. There was an inverse correlation between trends in antidepressant prescribing and suicide; with the largest declines in suicide in the age groups with the greatest increase in exposure to antidepressants. The association was significant in women ( rs =−0.74; P<0.05) but not in men ( r s=−0.62; P<0.10). It was concluded that an increase in antidepressant prescribing may be a proxy marker for improved overall management of depression. If so, increased prescribing of selective serotonin reuptake inhibitors in general practice may have a quantifiable benefit on the mental health of the population. Which of the following statements, if any, are true?

The advantages of the Matthews correlation coefficient (MCC) over F1 score and accuracy in binary classification evaluation

Abstract: To evaluate binary classifications and their confusion matrices, scientific researchers can employ several statistical rates, accordingly to the goal of the experiment they are investigating. Despite being a crucial issue in machine learning, no widespread consensus has been reached on a unified elective chosen measure yet. Accuracy and F1 score computed on confusion matrices have been (and still are) among the most popular adopted metrics in binary classification tasks. However, these statistical measures can dangerously show overoptimistic inflated results, especially on imbalanced datasets. The Matthews correlation coefficient (MCC), instead, is a more reliable statistical rate which produces a high score only if the prediction obtained good results in all of the four confusion matrix categories (true positives, false negatives, true negatives, and false positives), proportionally both to the size of positive elements and the size of negative elements in the dataset. In this article, we show how MCC produces a more informative and truthful score in evaluating binary classifications than accuracy and F1 score, by first explaining the mathematical properties, and then the asset of MCC in six synthetic use cases and in a real genomics scenario. We believe that the Matthews correlation coefficient should be preferred to accuracy and F1 score in evaluating binary classification tasks by all scientific communities.

Reading Mendelian randomisation studies: a guide, glossary, and checklist for clinicians

Abstract: Mendelian randomisation uses genetic variation as a natural experiment to investigate the causal relations between potentially modifiable risk factors and health outcomes in observational data. As with all epidemiological approaches, findings from Mendelian randomisation studies depend on specific assumptions. We provide explanations of the information typically reported in Mendelian randomisation studies that can be used to assess the plausibility of these assumptions and guidance on how to interpret findings from Mendelian randomisation studies in the context of other sources of evidence

The COMET Handbook: Version 1.0

Abstract: The selection of appropriate outcomes is crucial when designing clinical trials in order to compare the effects of different interventions directly. For the findings to influence policy and practice, the outcomes need to be relevant and important to key stakeholders including patients and the public, health care professionals and others making decisions about health care. It is now widely acknowledged that insufficient attention has been paid to the choice of outcomes measured in clinical trials. Researchers are increasingly addressing this issue through the development and use of a core outcome set, an agreed standardised collection of outcomes which should be measured and reported, as a minimum, in all trials for a specific clinical area. Accumulating work in this area has identified the need for guidance on the development, implementation, evaluation and updating of core outcome sets. This Handbook, developed by the COMET Initiative, brings together current thinking and methodological research regarding those issues. We recommend a four-step process to develop a core outcome set. The aim is to update the contents of the Handbook as further research is identified.

Stress and cardiovascular disease

Abstract: The physiological reaction to psychological stress, involving the hypothalamic-pituitary-adrenocortical and sympatho-adrenomedullary axes, is well characterized, but its link to cardiovascular disease risk is not well understood. Epidemiological data show that chronic stress predicts the occurrence of coronary heart disease (CHD). Employees who experience work-related stress and individuals who are socially isolated or lonely have an increased risk of a first CHD event. In addition, short-term emotional stress can act as a trigger of cardiac events among individuals with advanced atherosclerosis. A stress-specific coronary syndrome, known as transient left ventricular apical ballooning cardiomyopathy or stress (Takotsubo) cardiomyopathy, also exists. Among patients with CHD, acute psychological stress has been shown to induce transient myocardial ischemia and long-term stress can increase the risk of recurrent CHD events and mortality. Applications of the 'stress concept' (the understanding of stress as a risk factor and the use of stress management) in the clinical settings have been relatively limited, although the importance of stress management is highlighted in European guidelines for cardiovascular disease prevention.

Controlled Interventions to Reduce Burnout in Physicians A Systematic Review and Meta-analysis

Abstract: Importance Burnout is prevalent in physicians and can have a negative influence on performance, career continuation, and patient care. Existing evidence does not allow clear recommendations for the management of burnout in physicians. Objective To evaluate the effectiveness of interventions to reduce burnout in physicians and whether different types of interventions (physician-directed or organization-directed interventions), physician characteristics (length of experience), and health care setting characteristics (primary or secondary care) were associated with improved effects. Data Sources MEDLINE, Embase, PsycINFO, CINAHL, and Cochrane Register of Controlled Trials were searched from inception to May 31, 2016. The reference lists of eligible studies and other relevant systematic reviews were hand searched. Study Selection Randomized clinical trials and controlled before-after studies of interventions targeting burnout in physicians. Data Extraction and Synthesis Two independent reviewers extracted data and assessed the risk of bias. The main meta-analysis was followed by a number of prespecified subgroup and sensitivity analyses. All analyses were performed using random-effects models and heterogeneity was quantified. Main Outcomes and Measures The core outcome was burnout scores focused on emotional exhaustion, reported as standardized mean differences and their 95% confidence intervals. Results Twenty independent comparisons from 19 studies were included in the meta-analysis (n = 1550 physicians; mean [SD] age, 40.3 [9.5] years; 49% male). Interventions were associated with small significant reductions in burnout (standardized mean difference [SMD] = −0.29; 95% CI, −0.42 to −0.16; equal to a drop of 3 points on the emotional exhaustion domain of the Maslach Burnout Inventory above change in the controls). Subgroup analyses suggested significantly improved effects for organization-directed interventions (SMD = −0.45; 95% CI, −0.62 to −0.28) compared with physician-directed interventions (SMD = −0.18; 95% CI, −0.32 to −0.03). Interventions delivered in experienced physicians and in primary care were associated with higher effects compared with interventions delivered in inexperienced physicians and in secondary care, but these differences were not significant. The results were not influenced by the risk of bias ratings. Conclusions and Relevance Evidence from this meta-analysis suggests that recent intervention programs for burnout in physicians were associated with small benefits that may be boosted by adoption of organization-directed approaches. This finding provides support for the view that burnout is a problem of the whole health care organization, rather than individuals.