Many people in business and medicine regard statistics as at best a nuisance, or as a challenging (frequently incomprehensible) and therefore best ignored subject (whenever possible), or at worst as an hinderance to science. Even Einstein liked to say that “God doesn’t gamble.” .

This chapter examines the relationship between the number of reproducing fish and recruitment, number of reproducing females and egg production, and indices of year-class or cohort strength. Stock-recruit data consists of estimates of the number of reproducing fish and the number of subsequent recruits. The problem for biologists is to try to understand the relationship between stock and recruitment at least well enough to know how much the stock can be reduced before recruitment starts to drop. Starting values for stock-recruitment functions may be obtained from parameter estimates from linearized versions of the functions or from a visual fit of the function to data. The chapter shows that a visual of the fitted function with a 95% confidence band is constructed largely. Several authors have demonstrated or discussed how to add additional explanatory variables to a stock-recruitment function to explain additional …

Editor—I was delighted by the new clause 29 in the revised Declaration of Helsinki, which forbids clinical trials comparing drugs against no treatment when an effective treatment exists. I was dismayed that the World Medical Association was retreating from this position.1 From reading Hirsch and Guess's piece in the article about the latest revision of the Declaration of Helsinki, I am concerned that opposition from the pharmaceutical industry may lead to further revision.2 I spent two years on a multicentre research ethics committee, which saw many proposals for trials of active versus no treatment. In some of these cases an effective treatment already existed. I think it wrong to ask people who have come seeking treatment to do without it for the sake of research. People come to doctors for help, not to act as experimental subjects either for scientific curiosity or for drug regulation. For example, a trial was proposed of a drug versus placebo in symptomatic benign prostatic hyperplasia. Men in the trial were to take the placebo for more than a year. We were told that this drug was “equivalent” to another established treatment. What would happen to these men if they were not in the trial? Surely they would be treated. The patient information sheet stated that the drug was available in many other countries in x mg twice daily form, that a new 2x mg formulation was being studied and might show the same effectiveness as the original formulation, and that its effects compared with the effects of dummy capsules had to be tested. This struck me as a complete non sequitur. What is necessary is to compare 2x mg with x mg twice daily. There was no need to test 2x mg against no treatment at all. Another proposal was for a trial of a drug for Paget's disease of the bone. The applicants had already shown that the drug was better than placebo with smaller doses in a larger trial. This was a six month trial with a bone biopsy. Should we really be asking patients to do this? I can understand that clause 29 as currently stated might be interpreted as preventing trials in environments where the best current clinical methods are not available because of their cost. It could be amended to permit such trials without allowing patients to be lured into foregoing proved effective treatment where it exists and is accessible to them.3

CS teargas is one of the most used tools for crowd-control worldwide. Exposure to CS teargas is known to have consequences on protesters’ health (i.e. eye, skin irritation, respiratory problems), but recent concerns have been raised over its potential gender-specific effects. Indeed, field and clinical observations report cases of menstrual cycle issues among female protesters following high exposure to teargas. The hypothesis of a link between teargas exposure and menstrual cycle issues is plausible from a physiological standpoint, but has not yet been empirically investigated. Using data from a cross-sectional study on Yellow Vests protesters’ health in France, we examined the relationship between exposure to teargas and menstrual cycle issues among female protesters (n = 145). Analyses suggested a positive link between exposure and menstrual cycle perturbations. These results constitute first and preliminary evidence that CS teargas may be linked with menstrual cycle among women, which need corroboration given the importance of this issue. We call for further research on the potential effects of CS teargas on women’s reproductive system.

Background: Studies have linked asthma death to either increased or decreased use of medical services.Methods: A population based case-control study of asthma deaths in 1994-8 was performed in 22 English, six Scottish, and five Welsh health authorities/boards. All 681 subjects who died were under the age of 65 years with asthma in Part I on the death certificates. After exclusions, 532 hospital controls were matched to 532 cases for age, district, and date of asthma admission/death. Data were extracted blind from primary care records.Results: The median age of the subjects who died was 53 years; 60% of cases and 64% of controls were female. There was little difference in outpatient attendance (55% and 55%), hospital admission for asthma (51% and 54%), and median inpatient days (20 days and 15 days) in the previous 5 years. After mutual adjustment and adjustment for sex, using conditional logistic regression, three variables were independently associated with asthma death: fewer general practice contacts (odds ratio 0.82 (95% confidence interval (CI) 0.74 to 0.91) per 5 contacts) in the previous year, more home visits (1.14 (95% CI 1.08 to 1.21) per visit) in the previous year, and fewer peak expiratory flow recordings (0.83 (95% CI 0.74 to 0.92) per occasion) in the previous 3 months. These associations were similar after adjustment for markers of severity, psychosocial factors, systemic steroids, short acting bronchodilators and antibiotics, although the association with peak flow was weakened and just lost significance.Conclusion: Asthma death is associated with less use of primary care services. Both practice and patient factors may be involved and a better understanding of these may offer possibilities for reducing asthma death.

EDITOR—We agree with Alderson that authors should recognise that non-significant results are compatible with a range of possible findings.1 Papers in the same issue of the BMJ do not adhere to this good advice. Koivunen et al concluded that adenoidectomy is not effective and cannot be recommended, yet the 95% confidence interval for further …

We often want to compare two estimates of the same quantity derived from separate analyses. Thus we might want to compare the treatment effect in subgroups in a randomised trial, such as two age groups. The term for such a comparison is a test of interaction. In earlier Statistics Notes we discussed interaction in terms of heterogeneity of treatment effect.1–3 Here we revisit interaction and consider the concept more generally. The comparison of two estimated quantities, such as means or proportions, each with its standard error, is a general method that can be applied widely. The two estimates should be independent, not obtained from the same individuals—examples are the results from subgroups in a randomised trial or from two independent studies. The samples should be large. If the estimates are E 1 and E 2 with standard errors SE( E 1) and SE( E 2), then the difference d = E 1- E 2 has standard error SE( d )=√[SE( E …

The study of measurement error, observer variation and agreement between different methods of measurement are frequent topics in the imaging literature. We describe the problems of some applications of correlation and regression methods to these studies, using recent examples from this literature. We use a simulated example to show how these problems and misinterpretations arise. We describe the 95% limits of agreement approach and a similar, appropriate, regression technique. We discuss the difference vs. mean plot, and the pitfalls of plotting difference against one variable only. We stress that these are questions of estimation, not significance tests, and show how confidence intervals can be found for these estimates. Copyright © 2003 ISUOG. Published by John Wiley & Sons, Ltd.

Background: Uncontrolled studies suggest that psychosocial factors and health behaviour may be important in asthma death.Methods: A community based case-control study of 533 cases, comprising 78% of all asthma deaths under age 65 years and 533 hospital controls individually matched for age, district and asthma admission date corresponding to date of death was undertaken in seven regions of Britain (1994-98). Data were extracted blind from anonymised copies of primary care records for the previous 5 years and non-blind for the earlier period.Results: 60% of cases and 63% of controls were female. The median age in both groups was 53. Cases had an earlier age of asthma onset, more chronic obstructive lung disease, and were more obese. 48% of cases and 42% of controls had a health behaviour problem; repeated non-attendance/poor inhaler technique was related to increased risk of death. Overall, 85% and 86%, respectively, had a psychosocial problem. Four psychosocial factors were associated with increased risk of death (psychosis, alcohol/drug abuse, financial/employment problems, learning difficulties) and two with reduced risk (anxiety/prescription of antidepressant drugs and sexual problems). While alcohol/drug abuse lost significance after adjustment for psychosis, other associations appeared independent of each other and of indicators of severity and co-morbidity. None of the remaining 13 factors including family problems, domestic abuse, bereavement, and social isolation were significantly related to risk of asthma death.Conclusion: There was an apparently high burden of psychosocial problems in both cases and controls. The associations between health behaviour, psychosocial factors, and asthma death are varied and complex with a limited number of factors showing positive relationships.

# 1. WHAT DO THE PUBLIC THINK ABOUT THE USE OF THEIR HEALTH INFORMATION? PATIENT ELECTRONIC RECORD: INFORMATION AND CONSENT—THE PERIC PROJECT {#article-title-2} 3921 adults randomly selected from across Great Britain were interviewed. Subjects were asked to assess a selection of 10 out of 200

We show that although there is a significant correlation between intraocular pressure and intracranial pressure in neurosurgical patients, changes in intraocular pressure are a poor predictor of changes in intracranial pressure.

In clinical trials, the statistical concepts of significance and power are used in the determination of sample size for trials. The trialist must provide an estimate of standard deviation and a hypothetical population difference to be detected. This must be modified to deal with the designs encountered in guideline research. These are cluster randomized trials, because the patients of a single doctor or practice form a cluster. The trialist must be able to provide information about the effects of clustering, in the form of an intraclass correlation coefficient.

In recent years odds ratios have become widely used in medical reports—almost certainly some will appear in today's BMJ . There are three reasons for this. Firstly, they provide an estimate (with confidence interval) for the relationship between two binary (“yes or no”) variables. Secondly, they enable us to examine the effects of other variables on that relationship, using logistic regression. Thirdly, they have a special and very convenient interpretation in case-control studies (dealt with in a future note). The odds are a way of representing probability, especially familiar for betting. For example, the odds that a single throw of a die will produce a six are 1 to 5, or 1/5. The odds is the ratio of the probability that the event of interest occurs to the probability that it does not. This is often estimated by the ratio of the number of times that the event of interest occurs to …

Agreement between two methods of clinical measurement can be quantified using the differences between observations made using the two methods on the same subjects. The 95% limits of agreement, estimated by mean difference +/- 1.96 standard deviation of the differences, provide an interval within which 95% of differences between measurements by the two methods are expected to lie. We describe how graphical methods can be used to investigate the assumptions of the method and we also give confidence intervals. We extend the basic approach to data where there is a relationship between difference and magnitude, both with a simple logarithmic transformation approach and a new, more general, regression approach. We discuss the importance of the repeatability of each method separately and compare an estimate of this to the limits of agreement. We extend the limits of agreement approach to data with repeated measurements, proposing new estimates for equal numbers of replicates by each method on each subject, for unequal numbers of replicates, and for replicated data collected in pairs, where the underlying Value of the quantity being measured is changing. Finally, we describe a nonparametric approach to comparing methods.

Since 1991 the BMJ has had a policy of not publishing trials that have not been properly randomised, except in rare cases where this can be justified.1 Why? The simplest approach to evaluating a new treatment is to compare a single group of patients given the new treatment with a group previously treated with an alternative treatment. Usually such studies compare two consecutive series of patients in the same hospital(s). This approach is seriously flawed. Problems will arise from the mixture of retrospective and prospective studies, and we can never satisfactorily eliminate possible biases due to other factors (apart from treatment) that may have changed over time. Sacks et al compared trials of the same treatments in which randomised or historical controls were used and found a consistent tendency for historically controlled trials to yield more optimistic results than randomised trials.2 The use of historical controls can be justified only in tightly controlled situations of relatively rare conditions, such …