Epidemic and population patterns in the Chinese Empire (243 B.C.E. to 1911 C.E.): quantitative analysis of a unique but neglected epidemic catalogue.

A catalogue of dates and places of major outbreaks of epidemic diseases, that occurred in the Chinese Empire between 243 B.C.E. and 1911 C.E., combined with corresponding demographic data, provides a unique opportunity to explore how the pressure of epidemics grew in an agrarian society over 2000 years. This quantitative analysis reveals that: (1) the frequency of outbreaks increased slowly before the 12th century and rapidly thereafter, until 1872; (2) in the first millennium of our era, the people of China lived for decades free of major epidemics; in the second millennium, major outbreaks occurred every couple of years, but were localized; (3) in the more recent centuries, these outbreaks were as common, but disseminated to more places. This evolution, closely matching the demographic growth, was similar in the north and south of China, and therefore may have been similar in other regions of the world.

The James Lind Library: explaining and illustrating the evolution of fair tests of medical treatments.

The James Lind Library (www.jameslindlibrary.org) has been established to improve public and professional general knowledge about fair tests of treatments in healthcare and their history. Its foundation was laid ten years ago at the Royal College of Physicians of Edinburgh, and its administrative centre is in the College's Sibbald Library, one of the most important collections of historic medical manuscripts, papers and books in the world. The James Lind Library is a website that introduces visitors to the principles of fair tests of treatments, with a series of short, illustrated essays, which are currently available in English, Arabic, Chinese, French, Portuguese, Russian and Spanish. A 100-page book-- Testing Treatments--is now available free through the website, both in English and in Arabic and Spanish translations. To illustrate the evolution of ideas related to fair tests of treatments from 2000 BC to the present, the James Lind Library contains key passages and images from manuscripts, books and journal articles, many of them accompanied by commentaries, biographies, portraits and other relevant documents and images, including audio and video files. New material is being added to the website continuously, as relevant new records are identified and as methods for testing treatments evolve. A multinational, multilingual editorial team oversees the development of the website, which currently receives tens of thousands of visitors every month.

A cohort study of drug users' compliance with zidovudine treatment.

OBJECTIVES: Determine whether patients who have acquired the human immunodeficiency virus through injecting drug use receive less antiviral medication (zidovudine) than comparable patients of other risk groups considering access to, acceptance of, and compliance with treatment. DESIGN: Historical cohort study. SETTING: Human immunodeficiency virus outpatient clinic. PATIENTS: Human immunodeficiency virus-infected subjects eligible for zidovudine treatment between January 1, 1989, and January 1, 1992, comparing injecting drug users (IDUs) with non-IDUs ("others"). MAIN OUTCOME MEASURES: Proposal, acceptance, start of, and compliance with zidovudine treatment. RESULTS: One hundred fifty-one IDUs and 162 other human immunodeficiency virus-positive subjects became eligible for zidovudine treatment between January 1, 1989, and January 1, 1992. Both groups were proposed zidovudine as often, but zidovudine treatment was refused by 14.9% of IDUs compared with 7.1% of others (P = .029). The IDUs needed considerably more time than other subjects to accept zidovudine therapy (median delay between indication and start of zidovudine treatment, 61 days vs 30 days, P = .0001). After accepting, IDUs were as compliant with treatment as others: 81.3% vs 83.2% were good compliers, and rises of mean corpuscular volume of erythrocytes after 3 and 6 months of treatment were similar in both groups. Former drug users and IDUs receiving methadone were started on zidovudine treatment more often and complied better with treatment than active drug users. Absence of housing and presence of psychiatric diagnosis (both more prevalent in IDUs) were associated with less zidovudine treatment and worse compliance. CONCLUSION: Injecting drug users tend to delay the start of zidovudine treatment. However, once they have started, their compliance is no worse than the compliance of patients from other risk groups. These results have important implications for clinical trials, medical care, and public health.

Diagnosis of pulmonary embolism by a decision analysis-based strategy including clinical probability, D-dimer levels, and ultrasonography: a management study.

BACKGROUND: Assessment of the clinical probability of pulmonary emboli sm, plasma D-dimer measurement, and lower-limb venous compression ultrasonography have all been advocated in the workup of suspected pulmonary embolism, to minimize the requirement for pulmonary angiography in patients with nondiagnostic lung scans. However, their contribution has not been assessed prospectively. METHODS: Three hundred eight consecutive patients who came to the emergency department with suspected pulmonary embolism were managed according to a diagnostic protocol that included clinical probability assessment, lung scan, and sequential noninvasive tests: plasma D-dimer measurement by enzyme-linked immunosorbent assay (a concentration <500 microgram/L ruled out pulmonary embolism) and lower-limb B-mode venous compression ultrasonography (a positive finding was diagnostic of venous thromboembolism). Patients without pulmonary embolism according to the diagnostic workup did not receive anticoagulant treatment. The safety of this approach was assessed by a 6-month follow-up. RESULTS: of the 308 patients, 106 (34%) had a diagnostic lung scan (normal in 43 and high probability in 63). For the remaining 202 patients, noninvasive workup was diagnostic in 125 (62%). Pulmonary embolism was ruled out by a low clinical probability and a nondiagnostic scan in 48 patients and a D-dimer level less than 500 microgram/L in 53; pulmonary embolism was established by a high clinical probability and a nondiagnostic scan in seven patients and by a finding of a deep vein thrombosis on ultrasonography in 17. Therefore, only 77 of these 202 patients underwent pulmonary angiography (negative in 55; positive in 22). At 6-month follow-up (completed for 99.4% of the study population), only two of the 199 patients in whom the diagnostic protocol had ruled out pulmonary embolism (1.0% [95% confidence interval, 0.1 to 3.6]) had a thromboembolic event (pulmonary embolism, one; deep vein thrombosis, one). CONCLUSIONS: This decision analysis strategy yielded a definitive noninvasive diagnosis in 62% of patients with a nondiagnostic scan and appears to be safe.

Cost-effectiveness of noninvasive diagnostic aids in suspected pulmonary embolism.

BACKGROUND: Noninvasive instruments such as plasma D-dimer measurement (DD) and lower-limb compression ultrasonography (US) are being increasingly advocated to reduce the number of necessary angiograms in patients having suspected pulmonary embolism (PE) and a nondiagnostic lung scan. We therefore designed a decision analysis model (1) to evaluate the cost-effectiveness of combining these noninvasive diagnostic aids with lung scan and angiography in the diagnosis of PE and (2) to determine the optimal sequence and combination of tests taking into account the clinical probability of PE. METHODS: We performed a cost-effectiveness analysis based on literature data, including data from a management study in our institution. Six diagnostic strategies were compared with the reference, ie, lung scan followed when nondiagnostic (low or intermediate probability) by angiography. In all strategies, PE was ruled out by a normal or near-normal scan, a negative DD (plasma level below 500 micrograms/L), or a negative angiogram. Pulmonary embolism was diagnosed and anticoagulant treatment was undertaken in the presence of a high-probability lung scan, deep vein thrombosis showed by US, or a positive angiogram. In case of a nondiagnostic scan (low or intermediate probability), patients could be either treated or not treated, or undergo other tests, according to the selected strategy. RESULTS: Under baseline conditions (prevalence of PE, 35%), strategies combining DD and US with lung scan, angiography being done only in case of an inconclusive noninvasive workup (DD level > 500 micrograms/L, normal US, and nondiagnostic lung scan), were most cost-effective. This approach yielded a 9% incremental cost reduction and a 37% to 47% decrease in the number of necessary angiograms compared with the reference strategy (scan +/- angiography). For patients with a low clinical probability of PE (< or = 20%), withholding treatment from those with a low-probability lung scan without performing an angiogram proved safe and highly cost-effective (30% cost reduction), provided US showed no deep vein thrombosis. CONCLUSION: The DD test and US are cost-effective in the diagnostic workup of PE, whether performed after or before lung scan, thus allowing centers devoid of lung scanning and/or angiography facilities to screen patients with suspected PE and avoid costly referrals. In patients with a low clinical probability, a low-probability lung scan, and a normal US, treatment may be withheld without resorting to angiography.

[Determinants of hypercholesterolemia in Geneva: should the focus be on genes or environment?]. / Déterminants de l'hypercholestérolémie a Genève: doit-on se concentrer sur res gènes ou l'environnement?

Do genetic (G), or environmental (E), factors primarily determine blood lipid concentrations in the Geneva population? On-going research on the causes of hypercholesterolemia in populations based on data collected by the Bus Santé Survey is summarized. About 2/3 of the HDL-cholesterol variance could not be explained by the most important 5 of 10 environmental factors and 9 of 275 common genetic variants identified in the analyses. The remaining 1/3 of the variance was explained mainly by obesity, smoking, alcohol intake, age, and gender. The common polymorphisms played a much smaller role. GxG, GxE, and ExE interactions were the weakest determinants. Environmental factors appear to be the main determinants of hypercholesterolemia in populations. Measurement of genetic traits for clinical or public health purposes is currently not useful.

Dietary habits of smokers, people who never smoked, and exsmokers.

A large database on hospital patients with illnesses not related to tobacco or alcohol was used to investigate the dietary habits of males and females who never smoked, who were exsmokers, and who currently smoked. Smoking was positively related to meat consumption and negatively related to cereal consumption in males. Both male and female smokers consumed fewer vegetables and fruits but more alcohol and coffee than did people who never smoked. Exsmokers' diets were similar to those of people who never smoked. These results emphasize the importance of ruling out potential confounders or effect modifiers when studying the role of meat, milk, fruits, vegetables, cereal, coffee, or alcohol intake in the etiology of tobacco-related diseases.

Thyroid hormones and duration of ovulatory activity in the etiology of breast cancer.

We conducted the present study with the hypothesis that conflicting reports on the association between mild hypothyroidism and breast cancer may be due to failure to consider the potential interaction between thyroid and ovarian hormones. Seventy-three cases of breast cancer and 75 hospital controls were studied. The overall matched multivariate odds ratio of breast cancer for the lowest tertile of free T4 (< or = 1.10 ng/dl) versus the two other tertiles was 1.7 (95% confidence limits, 0.6-5.0). However, there was a statistically significant linear trend (P = 0.04) in the odds ratios for breast cancer related to subnormal free T4 levels across tertiles of duration of ovulatory activity. These results suggest that women combining low levels of circulating free T4 with long duration of ovulatory activity may be at increased risk for this disease.

Incidence of status epilepticus in French-speaking Switzerland: (EPISTAR).

BACKGROUND: To determine the incidence, risk factors, and case fatality rate of status epilepticus (SE) in the French-speaking part of Switzerland. METHODS: Between October 1, 1997 and September 30, 1998 all cases of SE referred to all the hospitals in the six cantons of the French-speaking part of Switzerland were identified by physicians working in emergency rooms, intensive care units, and electroencephalography departments; neurologists; and pediatricians from all hospitals in the area. Each case was validated and classified according to seizure type and etiology. RESULTS: Over 1 year, 172 cases were identified, of whom 74 had a history of epilepsy (42.4%). The crude and standardized annual incidence rate were 9.9/100,000 (95% CI, 8.4 to 11.4) and 10.3/100,000 (95% CI, 8.7 to 11.9). The incidence rate was higher among children < 1 year of age and adults > 65 years, and among men than women. There were 108 cases of acute symptomatic SE (incidence: 6.2 per 1000), 49 cases of remote symptomatic SE, and 15 cases of unknown etiology. Case fatality rate was 7.6%. CONCLUSIONS: The standardized incidence rate of SE in the French-speaking part of Switzerland was lower than that reported in Rochester, MN (18.3/100,000) and in the white population of Richmond, VA (20/100,000). The discrepancy may stem from the lack of a homogeneous, rigorous, and pragmatic definition of SE and the efficient management of acute repetitive seizures in this area.

The differential effect of education and occupation on body mass and overweight in a sample of working people of the general population.

PURPOSE: To assess whether two indicators of social class, education and occupation, have independent and/or synergistic effects in determining the body mass and overweight. METHODS: Body mass index (BMI), education, and occupation were assessed in a survey of 1767 men and 1268 women from a representative sample of currently working people of the general population of Geneva, Switzerland. Education and occupation were categorized as low, medium, and high. Overweight was defined as BMI > or = 25 kg/m(2). RESULTS: The prevalence of overweight was 52.1% in men and 28.7% in women. Men with overweight were more likely to have low education while women with overweight had lower education and lower occupation. Education and occupation were inversely related to BMI in both genders and, in women, had a synergistic effect (p-value for the interaction = 0.03). CONCLUSIONS: Education and occupation have independent and, in women, synergistic effects on BMI. The two indicators may express different mechanisms through which low social class is related to high body mass.

P. C. A. Louis and the birth of clinical epidemiology.

Pierre Charles Alexandre Louis (1787-1872) has been the direct or indirect mentor of influential U.S. and English scientists in public health, epidemiology, medicine, and biostatistics during the 19th and 20th century. Louis was primarily a clinician, but his name has been more closely associated with the history of epidemiology. There is an imbalance between the fame of Louis's scientific contribution and the scarcity of in depth analyses of his work. The controversy related to the usage of bloodletting for treating inflammatory diseases is one of the most famous episodes of Louis's work. A modern analysis of Louis's data confirms his original conclusions; that is, early bloodletting seems to reduce the duration of a pneumonitis disease in patients who survive from this disease but may also increase the overall short-term mortality. Some basic algebraic mistakes in computations of means tend to attenuate the verdict against bloodletting, but these are trivial relative to the rigor of the overall demonstration and to Louis's methodological contribution to clinical observation. Louis firmly belongs to the history of epidemiology. He deserves a similar place in the history of medicine.

The influence of patient and physician gender on prescription of psychotropic drugs.

There is a wealth of evidence that more psychotropic drugs are prescribed to female than to male patients. We studied the psychotropic drug prescription of 15 male and 9 female physicians to 2493 patients in the Ambulatory Care Clinic of a University hospital. The relative odds (RO) of benzodiazepine prescription to a female compared to a male patient was the same when the physician was a male or a female (adjusted RO: 1.8; 95% CL: 1.5-2.1). But female patients were significantly more likely to be prescribed psychotropic drugs of all types from a female than from a male physician (adjusted RO: 1.3; 95% CL: 1.0-1.7). Information on patient diagnosis or on severity of symptoms would be needed to determine whether the present results reflect gender differences in medical care needs, overprescription of psychotropic drugs to females or underprescription of psychotropic drugs to males.

Interaction fallacy.

We define interaction fallacy as the situation in which heterogeneity of odds ratios suggests an interaction that does actually not exist among the corresponding risk ratios. We provide a hypothetical example of interaction fallacy between the presence of the germ-line BRCA1 mutation, age at first live birth, and breast cancer risk based on data from the literature. More generally, we present a set of hypothetical conditions under which interaction fallacy may occur.

Relation of environmental tobacco smoke to diet and health habits: variations according to the site of exposure.

It has been postulated that the relationship of environmental tobacco smoke (ETS) exposure to cancer or cardiovascular diseases may be confounded by social class or diet because women exposed to ETS by their smoker spouse belong to lower social classes and have an unhealthy diet. In a population survey in Geneva, Switzerland, 914 female never-smokers were interviewed about sociodemographic factors, health habits including a semiquantitative food frequency questionnaire, and exposure to ETS according to the site (home, work, leisure). Compared to women unexposed to ETS, those exposed to ETS at work ate less fibers, cereals, vegetables, lean meat, had a lower intake of iron and beta-carotene, and had a lower total energy intake; women exposed during leisure time ate less cereals, drank less skim milk, and had a lower intake of complex carbohydrates. But the diet of women exposed at home did not differ from the diet of those unexposed to ETS. Thus, "living with a smoker" in Geneva does not necessarily imply adopting his health and dietary habits. We conclude that confounding factors of the association of ETS and disease vary according to site and populations and therefore should not be invoked as a systematic source of bias in all studies.

Smoking prevalence in neighborhood and hospital controls: implications for hospital-based case-control studies.

It is widely believed that the prevalence of smoking among hospital patients is greater than that of the general population because many conditions for which patients are hospitalized are caused by or associated with smoking, and that this increased prevalence may bias results of case-control studies of tobacco-related diseases. For this reason, many authors have suggested excluding from the control series patients hospitalized for tobacco-related illnesses. The present study investigated potential selection bias for hospital compared to neighborhood controls in studying tobacco-related diseases. The 709 cases from six U.S. cities had tobacco-related cancers or myocardial infarction. They were individually matched to one hospital control and to one neighbor. After excluding patients with tobacco-related diseases, hospital controls were less often current smokers and more often former smokers than neighborhood controls. Among male ever smokers, hospital controls tended to smoke more cigarettes per day than neighborhood controls. Compared with the U.S. population, there was an overrepresentation of smokers in neighborhood controls rather than an under-representation of smokers in hospital controls. Relative risk estimates varied according to type of control. Choosing between hospital and neighborhood controls in case-control studies should be dictated by criteria related to the study hypothesis, participation, or cost. In particular, exclusion of hospital controls with diseases known to be tobacco-related seems to be a successful strategy for reducing selection bias.

Body mass index compared to dual-energy x-ray absorptiometry: evidence for a spectrum bias.

The objective of this study was to evaluate the accuracy of body mass index (BMI) in the diagnosis of obesity. The relationship of this weight-for-height index to body composition was determined using dual-energy x-ray absorptiometry (DEXA) in 226 Caucasian subjects. BMI-obesity was defined as a BMI greater than 27.8 kg/m2 in males and greater than 27.3 kg/m2 in females. BMI-obesity was weakly sensitive (males: 12.5%, females 13.6%) compared with diagnosis by DEXA, defined as percent body fat mass (%BF) greater than 20% in males and greater than 25% in females. Conversely, the specificity of BMI-obesity was high (males and females: 100%). When analyzing subgroups of subjects according to weight, sensitivity was higher among heavier subjects than among lighter ones. In both sexes and in all subgroups, the specificity was 100%. The clinical implication of this spectrum bias is that, in men or women weighing less than 80 kg, measures of obesity other than BMI, such as bioelectrical impedance, should be preferred.

Epidemiology of herpes simplex virus in children by detection of specific antibodies in saliva.

OBJECTIVES: To facilitate the study of the prevalence of herpes simplex virus (HSV) infection and its determinants in children, we developed a noninvasive saliva test. METHODS: A capture enzyme-linked immunosorbent assay (ELISA) for the detection of IgG to HSV in saliva was developed, validated against a commercial serum ELISA in 110 children and 187 adults and used in a cross-sectional population-based study including 2,048 children ages 1 to 17 years, recruited in day-care centers and schools of Geneva, Switzerland. Demographic and socioeconomic determinants of HSV prevalence were studied. RESULTS: The sensitivity and specificity of the saliva assay were 94.1 and 95.5%, respectively, compared with the commercial serum ELISA. Participation in the cross-sectional study was 86.6%. The overall prevalence of anti-HSV IgG was 23.91%. It increased with age up to 7 years, reaching a plateau at 35% without evidence for day-care or school transmission. The main determinants of prevalence were region of national origin and parents' professional category. CONCLUSIONS: This new saliva-based assay proved its feasibility in the first population-based study of HSV prevalence in children that uses saliva, confirmed its validity by identifying determinants of prevalence consistent with previous reports and yielded new information, such as the lack of influence of day-care attendance, in the population studied.

Misclassification bias related to definition of menopausal status in case-control studies of breast cancer.

It is often assumed, but has not been consistently observed, that some characteristics of reproductive history are specifically related to breast cancer of pre- or postmenopausal onset. To determine whether inconsistent reports may be due to differences in definition of menopause, we computed the relative odds (RO) of breast cancer for nulliparity, age at first live birth, family history of breast cancer and prior history of benign breast disease, separately in pre- and postmenopausal women, using seven different definitions of menopause. Results show that (i) relative odds of breast cancer and their confidence intervals may vary according to definitions of menopause; (ii) age-based definitions of menopause are associated with moderate differential misclassification bias between cases and controls; (iii) nulliparity, late age at first birth and family history of breast cancer seem to be specific risk factors for pre- but not postmenopausal breast cancer when cutoff for menopausal status is 10 years or more after last menses; and (iv) when information on menstrual history is not available, 50 years of age may be the best proxy for all menses-based definitions of menopause. We conclude that inconsistent findings on the effect of menopausal status in the association of breast cancer with some reproductive factors are partly due to statistical imprecision and differential misclassification bias associated with different age-based or menses-based definitions of menopause. Researchers should either test whether their conclusions hold using several definitions of menopause or give a biological rationale for the choice of a given definition of menopause.

Effect of occupational and recreational activity on the risk of colorectal cancer among males: a case-control study.

Epidemiological studies have consistently demonstrated that either occupational or recreational physical activity is protective against colon cancer. However, it is unclear whether recreational activity is similarly protective among those who engage in high or low occupational activity. We therefore compared 440 male cases of colorectal cancer with 1164 male hospital patients. Occupational activity was defined according to job title, while recreational activity was assessed by questionnaire for three different periods of life. Occupational activity was protective with respect to colorectal cancer irrespective of whether one engaged in recreational activity at any different period of life. In contrast, recreational activity, performed at 20-44 years of age appeared to decrease colon cancer risk by 10-25% irrespective of the intensity of job activity. The present results suggest that, although we observed a larger effect with occupational activity than with recreational activity, middle-aged men may reduce their risk of colorectal cancer if they exercise when they are not working. These findings need to be confirmed in other populations.

Diet and socioeconomic position: does the use of different indicators matter?

OBJECTIVES: To describe the association of diet and socioeconomic position and to assess whether two different indicators, education and occupation, independently contribute in determining diet. METHODS: A community-based random sample of men and women residents of Geneva canton, aged 35 to 74, participated in a survey of cardiovascular risk factors conducted annually since 1993. Lifetime occupational and educational history and a semi-quantitative food frequency questionnaire were obtained from 2929 men and 2767 women. RESULTS: Subjects from lower education and/or occupation consumed less fish and vegetables but more fried foods, pasta and potatoes, table sugar and beer. Iron, calcium, vitamin A and vitamin D intake were lower in the lower educational and occupational groups. Both indicators significantly contributed to determining a less healthy dietary pattern for those from low social class. The effects of education and occupation on dietary habits were usually additive and synergistic for some food groups. CONCLUSION: Assessing both education and occupation, improves the description of social class inequalities in dietary habits, as they act, most of the time, as independent factors.