SAFETY FACTOR A multiplicative factor incorporated in risk assessments or safety standards to allow for unpredictable types of variation, such as variability from test animals to humans, random variation within an experiment, and person-to-person vari- ability. Safety factors are often in the range of 10 to 1000 or even higher magnitudes.
SAFETY STANDARDS Under the requirements of the U.S. Occupational Safety and Health Act (OSHA, 1970), occupational safety and health standard means a stand- ard that requires conditions, or the adoption of one or more practices, means, meth- ods, operations, or processes, reasonably necessary or appropriate to provide safe or healthful employment and places of employment. Safety standards may be adopted by national consensus or established by federal regulation. These standards have been adopted in many other nations besides the United States, although some European and other countries have their own standards, which may be either more or less stringent than those in the United States.266,299,300
There are several varieties of safety standards:
SAMPLE A selected subset of a population. A sample may be random or nonrandom and may be representative or nonrepresentative. Several types of sample can be distinguished, including the following:
SAMPLE, EPSEM (“equal probability of selection method”) A sample selected in such a manner that all the population units have the same probability of selection. A simple random sample is an EPSEM sample; a stratified sample is not unless the probability of selection is the same for all strata.
SAMPLE SIZE DETERMINATION The mathematical process of deciding, before a study begins, how many subjects should be studied. The factors to be taken into account include the incidence or prevalence of the condition being studied, the estimated or putative relationship among the variables in the study, the power that is desired, and the maximum allowable magnitude of type i error.
SAMPLING The process of selecting a number of subjects from all the subjects in a particular group, or “universe.” Statistical inference based on sample results may be attributed only to the population sampled. Any extrapolation to a larger or different population involves judgments about population differences, along with any available data pertaining to the difference, and is not part of conventional statistical inference, although bayesian statistical methods can incorporate these issues.
SAMPLING BIAS Systematic error due to the methods or procedures used to sample or select the study subjects, specimens, or items (e.g., scientific papers), including errors due to the study of a nonrandom sample of a population.
SAMPLING ERROR That part of the total estimation error caused by random influences on who or what is selected for study.
SAMPLING VARIATION Since the inclusion of individuals in a sample is partly determined by chance, the results of analysis in two or more samples will differ in part by chance. This is known as “sampling variation” or more precisely as “random sampling variation.”
SANITARY CORDON See cordon sanitaire.
SANTAYANA SYNDROME The neglect of what might be learned from the many blun- ders and errors contained in medical history.348,349 A term coined and used by A.R. Fein- stein in remembrance of a sentence from George Santayana (1863–1952), a Spanish and American philosopher and poet: “Those who cannot remember the past are con- demned to repeat it.”350 See also interpretive bias.
SARTWELL’S INCUBATION MODEL Philip Sartwell (1908–1999) found that the incubation periods for many communicable diseases tend to have a log-normal distribution, and that the “incubation” periods for certain cancers following certain well-defined external causes also tend to have a log-normal distribution.351 This model is useful but should not be assumed to hold universally. See also incubation period; latency period.
SCALE A device or system for measuring equal portions. A logarithmic scale measures equal powers of 10. Many kinds of scale are used in medicine and epidemiology. From the French and Middle English scale, a ladder.
SCAN STATISTIC A test for detection of clustering over time. A technique used in surveillance epidemiology to detect an unusual rate of occurrence of a disease by com- paring observed number of cases with the expected number on the basis of experience in a recent defined period.
SCATTER DIAGRAM, PLOT (Syn: scattergram) A graphic method of displaying the distribution of two variables in relation to each other. The values for one variable are measured on the horizontal axis and the values for the other on the vertical axis.
SCENARIO-BASED HEALTH RISK ASSESSMENT A variant of population health risk assessment in which the exposure input is not an actual measured or measurable exposure but a plausible, preferably model-generated scenario of future exposure. It is particularly relevant in using existing epidemiological knowledge to prepare plans for the future health impacts of anticipated environmental changes, such as climate change and stratospheric ozone depletion.
SCENARIO BUILDING A method of predicting the future that relies on a series of assumptions about alternative possibilities rather than on simple extrapolation of exist- ing trends.352 Trend lines for demographic composition, morbidity and mortality rates, etc., can then be modified by allowing for each assumption in turn, or combinations of assumptions. The method is claimed to lead to greater flexibility in long-range health planning than simple forecasting that relies only upon extrapolation of trends.
SCIENCE Systematic observation and experiment to explain and predict natural phe- nomena aimed at establishing, enlarging, or confirming knowledge. Science uses obser- vations and/or experiments to make logical inferences, formulate and test hypotheses, and arrive at generalizable conclusions, expressed as testable laws and principles. Sci- ence advances through conjecture or intuition, hypothesis, refutation of deductions from hypotheses, and verification of hypotheses and theories by induction.6,10,22,67,239,268,353 Occasionally science undergoes a paradigm shift as long-established principles and laws are overturned by new discoveries.302
A characteristic of vigorous science is ongoing tests of hypotheses and theories that are taken as “established” in order to detect failings. This activity stems from a point distinguishing science from other systems of knowledge: the idea that all knowledge of the real world is fallible at some level and must not be allowed to become completely unassailable dogma. See also epistemic communities; hypothetico-deductive method; knowledge construction; sociology of scientific knowledge.
SCIENTIFIC MISCONDUCT A class of ethical violations in the conduct of research, gen- erally taken to include falsification, fabrication, fraud, or plagiarism in the proposal, design, implementation, reporting, or review of research. May also include violation of the rights and dignity of participants in research, misuse of research funds, mistreat- ment of scientific colleagues (e.g., in peer review), and failure to report undesired find- ings.37 See also publication bias; reppression bias; suppression bias.
SCIENTOMETRICS The measurement of scientific output and the impact of scientific findings (e.g., on public health policies).40
SCREENING Screening was defined in 1951 by the U.S. Commission on Chronic Illness as, “the presumptive identification of unrecognized disease or defect by the application of tests, examinations or other procedures which can be applied rapidly. Screening tests sort out apparently well persons who probably have a disease from those who prob- ably do not. A screening test is not intended to be diagnostic. Persons with positive or suspicious findings must be referred to their physicians for diagnosis and necessary treatment.”354
The initiative for screening usually comes from an agency or organization rather than from a patient with a complaint. Screening is usually concerned with chronic illness and aims to detect disease not yet under medical care. Screening may identify risk factors, genetic predisposition, and precursors, or early evidence of disease.
The characteristics of a screening test must include accuracy, estimates of yield, precision, reproducibility, sensitivity and specificity, and validity.14,134,355 See also case finding; detectable preclinical period; prevention.
SCREENING LEVEL The normal limit or cutoff point at which a screening test is regarded as positive.
SEASONAL VARIATION Change in physiological status or in disease occurrence that conforms to a regular seasonal pattern.
SECONDARY ATTACK RATE The number of cases of an infection that occur among contacts within the incubation period following exposure to a primary case in relation to the total number of exposed contacts; the denominator is restricted to susceptible contacts when these can be determined. The secondary attack rate is a measure of con- tagiousness and is useful in evaluating control measures. See also attack rate; basic reproductive rate.
SECONDHAND TOBACCO SMOKE See involuntary smoking; environmental tobacco smoke.
SECOND-LINE DRUGS (SLDS) See drug resistance, multiple.
SECTOR In the language used by UN agencies (WHO, UNICEF, etc.), a sector is a defined component of the body politic, such as the health sector, the education sector, the housing sector.
SECULAR TREND (Syn: temporal trend) Changes over a long period of time, generally years or decades. Examples include the decline of tuberculosis mortality and the rise, followed by a decline, in coronary heart disease mortality in many industrial countries in the past 50 years.
SEE Sociedad Española de Epidemiología (the Spanish Society of Epidemiology).
SEER Surveillance, Epidemiology and End Results (a program of the U.S. National Cancer Institute).
SER Society for Epidemiologic Research.
SELECTION In genetics, the force that brings about changes in the frequency of alleles and genotypes in populations through differential reproduction. In epidemiology, the process and procedure for choosing individuals for study, usually by an orderly means such as random allocation.
Selection bias can virtually never be corrected by statistical analysis. It is a common and commonly overlooked problem, not just in epidemiological studies but also in clinical and basic biological studies. See also Berkson’s bias; consent bias; controls, hospital; inception cohort.
SEMI-INDIVIDUALDESIGN Individual-levelstudies (e.g., cohortstudies, cross-sectional studies, case-control studies) in which outcome and covariates are measured at the individual level while exposure is characterized on the aggregate (or ecological) level. Used either because groups share the same exposure or because individual-level expo- sure measures are not available. Frequently used in environmental epidemiology to describe exposure to air, water, or soil pollutants. Not to be confused with ecological studies.107,358
SENSITIVE PERIOD (Syn: critical period) A time during the development of a tissue, organ, or system when it can be permanently changed by harmful influences (e.g., undernutrition, hypoxia, stress). It often coincides with a period of rapid cell division and, for many tissues and systems, occurs before birth. The brain and liver are the main organs that remain plastic after birth.45 The adverse (or protective) effects on health of exposures during a sensitive period may be apparent many years later. See also develop- mental origins hypothesis; plasticity; programming; vulnerability.
SENSITIVITY ANALYSIS A method to determine the robustness of an assessment by examining the extent to which results are affected by changes in methods, models, values of unmeasured variables, or assumptions.360 The aim is to identify results that are most dependent on questionable or unsupported assumptions. See also influence analysis; outliers.
SENSITIVITY TESTING A study of how the final outcome of an analysis changes as a function of varying one or more of the input parameters in a prescribed manner. See also sensitivity analysis.
SENTINEL HEALTH EVENT A condition that can be used to assess the stability or change in health levels of a population, usually by monitoring mortality statistics. Thus, death due to acute head injury is a sentinel event for a class of severe traffic injury that may be reduced by such preventive measures as use of seat belts and crash helmets.
SENTINEL PHYSICIAN, SENTINEL PRACTICE In family medicine, a physician or practice that undertakes to maintain surveillance for and to report certain specific predetermined events, such as cases of certain communicable diseases or adverse drug reactions.
SENTINEL SURVEILLANCE Surveillance based on selected population samples cho- sen to represent the relevant experience of particular groups. This approach is useful in dealing with sensitive issues such as HIV/AIDS or when cooperation levels can be improved through participation of professional organizations, such as colleges or net- works of family physicians, for the early detection of influenza epidemics. In sentinel surveillance, standard case definitions and protocols must be used to ensure validity of comparisons across time and sites despite lack of statistically valid sampling. Sentinel surveillance may include the use of animal sentinels to detect circulation of arboviruses. See also surveillance.
SEQUENTIAL ANALYSIS A statistical method that allows an experiment to be ended as soon as an answer of the desired precision is obtained. Study and control subjects are randomly allocated in pairs or blocks. The result of the comparison of each pair of subjects, one treated and one control, is examined as soon as it becomes available and is added to all previous results.
SERENDIPITY The accidental (and happy) discovery of important new information. A well-known example is Fleming’s discovery of the bactericidal properties of penicillin mold. In case-control studies aimed at testing a specific hypothesis (e.g., about the relationship between tobacco and cancer), questions on other aspects of lifestyle have serendipitously revealed statistically significant associations (e.g., between alcohol con- sumption and certain cancers).
SERIAL INTERVAL (Syn: generation time) The period of time between analogous phases of an infectious illness, in successive cases of a chain of infection, that is spread from person to person.
SEROEPIDEMIOLOGY Epidemiological study or activity based on serological testing of characteristic changes in the serum level of specific antibodies. Latent, subclinical infec- tions and carrier states can thus be detected in addition to clinically overt cases.
SEROLOGY The branch of science dealing with the measurement and characterization of antibodies and other immunological substances in body fluids, particularly serum.
SET A defined group of events, objects, or data that is distinguishable from other groups.
SET THEORY Branch of mathematics and logic dealing with the characteristics and relationships of sets.
SEX RATIO The ratio of one sex to the other. Usually defined as the ratio of males to females (or of the rates observed in males and females).
SF36 Acronym for the 36-item questionnaire derived from the longer set of questions used in household interview surveys conducted by the U.S. National Center for Health Statistics. The SF36 questions measure eight multi-item variables: physical function, social function, role limitation, mental health, energy, vitality, pain, and general percep- tion of health. The instrument has been widely adopted, although some authors have raised doubts about its validity.
SHARPS A jargon term for any sharp object used in a health care setting that is capable of penetrating the skin (e.g., hypodermic needles, scalpel blades, broken glass vials).
SHEA Society for Healthcare Epidemiology of America (www.shea-online.org).
“SHOE-LEATHER” EPIDEMIOLOGY Gathering information for epidemiological studies by direct inquiry among the people, e.g., walking from door to door and asking questions of every householder (wearing out shoe leather in the process).27,313
John Snow (1813–1858) did this when he was investigating the sources of water sup- ply to households in the cholera epidemic in London in 1854; the method has been successfully used in many subsequent epidemic investigations. It is especially useful in investigations of sexually transmitted diseases. Much of the work of the Epidemic Intelligence Service (EIS) is based on shoe-leather epidemiology. EIS officers have a club tie displaying the sole of a shoe with a hole in it. See also natural experiment.
SHRINKAGE ESTIMATION (Syn: Stein estimation, penalized estimation) In statistics, a family of procedures to improve the overall accuracy of multiple estimates. This improvement is made by moving the estimates toward values judged or estimated to be more probable than most of the possible values for the parameters being estimated. The value chosen is often zero, whence the procedures make the estimates smaller, or “shrunk” toward zero. Most shrinkage methods are equivalent to empirical-bayes methods.
SIBLINGS Children borne by the same mother.
SIBSHIP All the brothers and sisters borne by the same mother.
SICKNESS See disease.
SICKNESS “CAREER” The process of decisions made by and/or for a person as he or she becomes symptomatic, defined as sick, seeks informal and professional care, and becomes a patient. It takes place in specific settings, in interaction with other people who, in accordance with their assessment of the problem and taking into account their own needs and the opportunities for alternative courses of action, apply the social norms of their particular group and set expectations for behavior.124–128 See also disease label; semiology.
SIDE EFFECT An effect, other than the intended one, produced by a preventive, diagnos- tic, or therapeutic procedure or regimen. Not necessarily harmful.
SIDESTREAM SMOKE Smoke from combusted tobacco products, usually cigarettes, that is not filtered through the cigarette or the smoker’s respiratory system but directly enters the air, where its toxic and irritant effects on nonsmokers can lead to adverse health effects. See also environmental tobacco smoke; involuntary smoking.
1. In statistics and signal processing, the ratio of explained variation to unexplained (error) variation.
2. A jargon term for the relationship of pertinent findings to that which is extraneous or irrelevant or intrudes because measurement methods or other procedures are insufficiently sensitive.
SIGNIFICANCE, CLINICAL Importance, relevance, or meaning for the care of individuals, who may be—in clinical research—patients. A difference in effect size considered to be important (e.g., by a professional) in medical decisions regardless of the degree of statistical significance. Statistical significance can never be taken to equal clinical significance. For example, when large numbers of subjects are studied, some differences will be statistically significant even if their magnitude or size is small; hence they will be of little importance for patient care. Conversely, when small numbers of subjects are studied, some differences will not be statistically significant even if their magnitude is large; hence they may be of importance for patient care but not detected as such in the analysis.12,14,31,62,63,75
SIGNIFICANCE, PUBLIC HEALTH Importance, relevance, or meaning from a pub- lic health perspective; for example, an environmental factor may have public health significance or importance because of its impact on the burden of disease in a given population. See also significance, clinical.
SIGN TEST A distribution-free test that can be used in combining results of several studies. The test considers the direction of results of individual studies (i.e., whether the associations demonstrated are positive or negative).
SILENCING, GENE See gene silencing.
SIMPSON’S PARADOX The possibility that a measure of association may reverse direction upon stratification by a third variable. In epidemiology, it has been presented as a form of confounding in which the presence of a confounding variable changes the direc- tion of an association. This interpretation is narrower than Simpson’s, however, for he used an example with pure association, with no causality or confounding implied. Simp- son’s paradox can occur in meta-analysis because the sum of the data or results from a number of different studies may be affected by confounding variables that have been excluded by design features from some studies but not others. Simpson’s paradox is not really a paradox but rather an extreme manifestation of the fact that associations may change according to the level of stratification.361 It is an extreme extreme violation of collapsibility, in which results of the data analysis in every mutually exclusive stratum or subgroup are the opposite of the crude results.69,362 See also confounding bias.
SIMULATION The use of a model system (e.g., a mathematical model or an animal model) to approximate the functioning or action of a real system; often used to study the prop- erties of a real system. See also Monte-Carlo study.
SINGLE-PATIENT TRIAL See n-of-one study.
SITUATION ANALYSIS Study of a situation that may require improvement. This begins with a definition of the problem and an assessment or measurement of its extent, severity, causes, and impacts upon the community; it is followed by appraisal of interactions between the system and its environment and evaluations of performance.
SKEW DISTRIBUTION An older and less recommended term for an asymmetrical fre- quency distribution. If a unimodal distribution has a longer tail extending toward lower values of the variate, it is said to have negative skewness; in the contrary case, it is said to have positive skewness. An example is the log-normal distribution.
SLDs (SECOND-LINE DRUGS) See drug resistance, multiple.
SLOW VIRUS Agent causing degenerative (often, neurological) diseases characterized by a long incubation period and a prolonged, slowly progressive course. Multiple sclerosis is possibly a slow virus disease. See also prion.
SMALL FOR GESTATIONAL AGE (SGA) See birth weight
SMOOTHING General term for methods of minimizing irregularities in a set of data. Examples include rounding, kriging, and moving averages.
SNOWBALL SAMPLING A method of selecting for study the members of “hidden” populations, e.g., illicit drug users. Those initially identified are asked to name acquaintances who are added to the sample; these, in turn, are asked to name further acquaintances, and so on until enough numbers are accumulated to give adequate power to the proposed study. The sample is, of course, not random. Compare capture-recapture method.
SOCIAL CLASS A stratum in society composed of individuals and families of equal standing. A method of socially stratifying populations according to occupation, educa- tion, or income. See also life course; socioeconomic classification; class.
SOCIAL DRIFT Downward social class mobility as a result of impaired health, sometimes due to reduced earning potential, mental disorders, or substance abuse. The “social drift hypothesis” also suggests that persons with poor mental health are more likely to move to neighborhoods with poor-quality environments.
SOCIAL EPIDEMIOLOGY A branch or subspecialty of epidemiology that studies the role of social structure and social factors in the production of health and disease in populations; it uses epidemiological principles, reasoning, and methods to study the social determinants and distribution of health states and events. An interface between epidemiology and the social sciences. It includes studies of personal and population health in the social context, behavior, social networks, and determinants of individual and population health such as ethnicity, socioeconomic status, social class, and housing conditions. It may use a life course approach. In intervention studies, it seeks to modify adverse factors and ameliorate determinants that can enhance good health.137,363–365 See also causes in public health sciences; ethnoepidemiology. As mentioned above, this dictionary includes definitions for just a few branches of epidemiology.
SOCIAL MARKETING The use of marketing theory, skills, and practice to achieve social change; to promote the general health or in other public health activities; to raise awareness and induce changes in behavior by individuals and groups aiming at enhanc- ing their level of health.
SOCIAL MEDICINE The practice of medicine concerned with health and disease as a function of group living. Social medicine is concerned with the health of people in rela- tion to their behavior in social groups; as such, it involves care of the individual patient as a member of a family and of other significant groups in everyday life. It is also con- cerned with the health of these groups and with that of the whole community. The “father of social medicine” was Johann Peter Frank (1745–1821), who described many features of this discipline in System einer vollständigen medicinischen Polizey (A System of Complete Medical Police, 1779). After the appointment of John Ryle (1889–1950) as the first professor of social medicine at the University of Oxford, this became the preferred term to describe academic departments dealing with this range of disciplines in the United Kingdom; in the 1970s, the preferred term became community medicine. The Acheson Report (1988) advocated the term public health medicine, which was for some years adopted by the Faculty of Community Medicine of the UK Royal Col- leges of Physicians and many British academic departments; it later became the Faculty of Public Health (FPH) (www.fphm.org.uk). The FPH is the standard setting body for specialists in public health in the United Kingdom; a joint Faculty of the three Royal Colleges of Physicians of the United Kingdom (London, Edinburgh, and Glasgow). See also community medicine; public health.
SOCIAL NETWORK INDEX A measure of the extent to which individuals or groups are connected to or isolated from others (e.g., family, friends, and work colleagues). Health status has been found to be positively associated with the extent of social networks.366
SOCIETAL RISK Probability of harm to the human population, including probability of adverse health effects to descendants and probability of disruption resulting from loss of services, such as industrial plant, loss of material goods, electricity.
SOCIOECONOMIC CLASSIFICATION Arrangement of persons into groups according to such characteristics as prior education, occupation, and income. Upon analysis, this usually reveals a strong connection with health-related characteristics, such as average length of life and risk of dying from certain specific causes.137,363–367
The oldest such classification that is epidemiologically useful is the registrar-general’s (RG’s) occupational classification, developed in 1911 by Stephenson, Registrar-General of England and Wales. This classified all occupations into five groups—the five “social classes.” Social class III is often further subdivided into nonmanual and manual groups, as follows:
This has proven to be a valuable epidemiological tool; social class is a reliable and consistent predictor of health experience and status.
There have been many attempts to develop more refined classifications; however, most refinements require the collection of more detailed information. For example, Hollingshead’s scale requires details about education and income as well as occupation. In developing countries, where up to 90% of the population may be classified under “agriculturalist” or “pastoralist” (farming or herding), other types of classifications have been developed.
One’s prestige in society and attitudes or values (e.g., setting a high value on getting a good education) are generally an integral part of social class or socioeconomic status. Attitudes toward health are often part of the set of values and may explain part of the observed difference in health between social classes.
SOCIOECONOMIC STATUS (SES) Descriptive term for a person’s position in society, which may be expressed on an ordinal scale using such criteria as income, level of education attained, occupation, value of dwelling place, etc.
SOCIOLOGY OF EPIDEMIOLOGY The application of the scientific principles and meth- ods of sociology to the science, discipline, and profession of epidemiology in order to improve understanding of the wider social causes and consequences of epidemiologists’ professional and scientific organization, patterns of practice, ideas, knowledge, and cul- tures (e.g., institutional arrangements, academic norms, scientific discourses, defense of identity and epistemic authority). It also addresses the patterns of interaction of epide- miologists with other branches of science and professions (e.g., clinical medicine, public health, the other health, life and social sciences), and with social agents, organizations, and systems (e.g., the economic, political, and legal systems). The tradition of sociology in epidemiology is rich; the sociology of epidemiology is virtually uncharted (in the sense of not mapped neither surveyed) and unchartered (i.e., not furnished with a charter or constitution). See also epidemiology, demarcation of; epistemic communities; knowledge construction.
SOCIOLOGY OF SCIENTIFIC KNOWLEDGE (SSK) An approach to the understanding of science that focuses on the social causes of the scientists’ convictions, knowledge, and beliefs. It centers on science as knowledge, by contrast with the “constructivist approach,” which is more interested on the constructive elements of scientific production (i.e., it considers science as practice). Both approaches tend to agree that the content of natural science is accessible by way of empirical sociological analysis and should be subjected to it; science is not to be investigated merely as a social institution; science’s epistemic core is a matter of investigation in its own right and should not be studied only by philosophers of science.268 See also epistemic communities; epistemology; interpretive bias; knowledge construction.
SOJOURN TIME (Syn: detectable preclinical period) The interval between detectability at screening and clinical presentation of a condition—i.e., the interval during which the condition is potentially detectable but not yet diagnosed.368
SOUNDEX CODE A sequence of letters used for recording names phonetically, espe- cially in record linkage.
SOURCE OF INFECTION The person, animal, object, or substance from which an infectious agent passes to a host. Source of infection should be clearly distinguished from source of contamination, such as overflow of a septic tank contaminating a water sup- ply or an infected cook contaminating a salad.52,57 See also reservoir of infection.
SPEARMAN’S RANK CORRELATION See correlation coefficient. SPECIFICATION
SPECIFICITY (OF A TEST) See sensitivity and specificity.
SPECTRUM BIAS A problem that may affect a study of diagnostic accuracy when it fails to account for the variation or heterogeneity of the test performance across population subgroups. Failure to recognize and address such heterogeneity will lead to estimates of test performance that are not generalizable to the relevant clinical populations. Bias may occur when diagnostic test performance varies across patient subgroups and the study does not adequately include all subgroups.62,63,321 For instance, overestimation of the sensitivity and specificity of the test will occur when the study includes a healthy group and a group with overt disease. Originally described as a bias,369 the variation is no longer considered to necessarily create a bias; rather, the spectrum of patients is considered a clinically relevant piece of information to be reported accurately and analyzed appropriately (e.g., by stratification). The term spectrum bias would hence be less appropriate than spectrum effect, which reflects the inherent variation in test per- formance among population subgroups.370 In interpreting results of a diagnostic study, assessing the spectrum of patients included will help determine whether results are generalizable to other populations of patients. See also workup bias.
SPECTRUM OF DISEASE The full range of manifestations of a disease; e.g., from pre- cursor states, to subclinical and mild cases, to florid and fulminating disease. The natural history of a disease from onset to resolution.62,63,87 See also inception cohort; induction period.
SPELL OF SICKNESS An episode of sickness with a well-defined onset and termination. As used in the monitoring or surveillance of disease, the spell is often defined by the duration of absence from work or school. See also disease; sickness “career.”
SPLEEN RATE A term used in malaria epidemiology to define the frequency of enlarged spleens detected on survey of a population in which malaria is prevalent. In association with the Hackett spleen classification, it summarizes the severity of malaria endemicity.
SPORADIC Occurring irregularly, haphazardly, from time to time, and generally infre- quently (e.g., cases of certain infectious diseases).
SPOT MAP Map showing the geographic location of people with a specific attribute (e.g., cases of a disease or elderly persons living alone). The making of a spot map is a common procedure in the investigation of a localized outbreak of disease. Inferences from such a map depend on the assumption that the population at risk of developing the disease is fairly evenly distributed over the area or that at least the heterogeneities are known and can be considered in interpreting the map. A refinement is to indicate multiple cases at a single location by a series of short horizontal bars, as John Snow did to mark the location of cases of cholera in the epidemic in London in 1849; the method has been used by innumerable field epidemiologists ever since.
SPREADSHEET A computer matrix of columns and rows in which numerical entries can be made on screen, stored, systematically manipulated, and modified.
STABLE POPULATION A population that has constant fertility and mortality rates, no migration, and consequently a fixed age distribution and constant growth rate; a popu- lation with stable structure. See also stationary population.
STANDARD Something that serves as a basis for comparison; a technical specification or written report drawn up by experts based on the consolidated results of scientific study, technology, and experience, aimed at optimum benefits and approved by a recognized and representative body.
STANDARD DEVIATION A measure of dispersion or variation. It is the most widely used measure of dispersion of a frequency distribution. It is equal to the positive square root of the variance. The mean tells where the values for a group are centered. The standard deviation is a summary of how widely dispersed the values are around this center.
STANDARD ERROR The standard deviation of an estimate. Used to calculate confi- dence intervals.
STANDARD GAMBLE See Von Neumann-Morgenstern standard gamble.
STANDARDIZATION A set of techniques, based on weighted averaging, used to remove as much as possible the effects of differences in age or other confounding variables in comparing two or more populations. The common method uses weighted averaging of rates specific for age, sex, or some other potential confounding variable(s) according to some specified distribution of these variables. There are two main methods, as follows:
Direct method: The specific rates in a study population are averaged, using as weights the distribution of a specified standard population. The directly standardized rate represents what the crude rate would have been in the study population if that population had the same distribution as the standard population with respect to the variable(s) for which the adjustment or standardization was carried out. inverse probability weighting can be seen as a generalization.
Indirect method: This is used to compare study populations for which the specific rates are either statistically unstable or unknown. The specific rates in the standard population are averaged, using as weights the distribution of the study population. The ratio of the crude rate for the study population to the weighted average so obtained is the standardized mortality (or morbidity) ratio, or SMR. The indirectly standardized rate itself is the product of the SMR and the crude rate for
the standard population, but this product is rarely used in etiological studies. A problem that arises with indirect standardization is that different SMRs are based on different weighting schemes (one for each study population) and so are not fully standardized for comparison to one another. As a result, comparisons of SMRs (or indirectly standardized rates) may remain partially confounded by the adjustment variables.12
STANDARDIZED INCIDENCE RATIO (Syn: standardized morbidity ratio) The ratio of the incident number of cases of a specified condition in the study population to the incident number that would be expected if the study population had the same incidence rate as a standard or other population for which the incidence rate is known; this ratio is usually expressed as a percentage. See also standardization (indirect).
STANDARDIZED MORTALITY RATIO (SMR) The ratio of the number of deaths observed in the study group or population to the number that would be expected if the study population had the same specific rates as the standard population. Often multi- plied by 100. See also standardization (indirect).
STANDARDIZED RATE RATIO (SRR) A rate ratio in which the numerator and denomi- nator rates have been standardized (weighted) to the same (standard) population dis- tribution. See also standardization (direct).
STANDARD METROPOLITAN STATISTICAL AREA Because of the extensive interac- tions between a city and its surrounding areas, a unit encompassing both is needed as a base for statistical description. The concept of a standard metropolitan statistical area (SMSA) was introduced in the United States to furnish such a unit. To qualify as an SMSA, an area has to meet criteria related to size, social and economic integration of the city and surrounding county or counties, minimum population density, and mini- mum proportion of the labor force engaged in nonagricultural work.
STARD Standards for Reporting of Diagnostic Accuracy. A consensus checklist and flow diagram aimed at improving the accuracy and completeness of reporting of studies of diagnostic accuracy.95,371,372 See also consort; moose; strobe; trend.
STATIONARY POPULATION A stable population that has a zero growth rate with constant numbers of births and deaths each year.
STATISTICAL ERROR See error.
STATISTICAL INFERENCE See inference.
STATISTICAL MODEL A mathematical model for distribution of samples or data.
STATISTICAL SIGNIFICANCE See significance, statistical.
STATISTICAL TEST A procedure intended to decide whether a statistical hypothesis (which may be about the distribution of one or more populations or variables or the size of an association or effect) should be rejected or not. Statistical tests may be para- metric or nonparametric.
STEIN ESTIMATION See shrinkage estimation
STEM-AND-LEAF DISPLAY A method of presenting numbers in a form resembling a histogram, with multiples of 10 along the “stem” and the integers forming the “leaves.” See also box-and-whiskers plot.
STOCHASTIC PROCESS A process (usually a temporal sequence) that incorporates some element of randomness.
STOPPING RULES In randomized controlled trials and other forms of systematic experiments, stopping rules are laid down in advance, specifying conditions or criteria under which the trial or experiment shall cease or be terminated. For example, in a randomized controlled trial, the unequivocal demonstration of superiority of one regimen over another is the most obvious reason for terminating the trial; a less frequent situation is the demonstration that a regimen causes harm to participants in the trial. The rule must be based on appropriate statistical tests to ensure that the empirically observed results are not due to chance.
STRATEGY, “HIGH-RISK” A clinically oriented approach to preventive medicine that focuses its efforts on needy individuals with the highest levels of the risk factor (‘the deviant minority with high-risk status’) and utilizes the established framework of medi- cal services. “A targeted rescue operation for vulnerable individuals.”72 The aim is to help each person reduce the high level of exposure to a cause or to some intermediate variable. Main strengths of this strategy include that the intervention may be matched to the needs of the individual; it may avoid interference with those who are not at a special risk; it may be accommodated within the ethical and cultural values, organization, and economics, of the health care system; selectivity may increase the likelihood of a cost-effective use of resources. Main weaknesses of the high-risk strategy are that prevention may become medicalized; success may be palliative and temporary; the contribution to overall (population) control of a disease may be small; the preventive intervention may be behaviorally or culturally inadequate or unsustainable; it has a poor ability to predict which individuals will benefit from the intervention.72
STRATEGY, “POPULATION” A public health-oriented approach to preventive medi- cine and public health predicting that a shift in the population distribution of a risk factor will prevent more burden of disease than targeting people at high risk. It starts with the recognition that the occurrence of common exposures and diseases reflects the functioning of society as a whole.72 The approach is more relevant to decrease exposure to (1) certain environmental agents that individuals have little capacity to detect than to (2) risk factors that individuals may generally decide to avoid. Main strengths of this strategy include that it may be radical (“only the social and political approach confronts the root causes”); the societal effects of a distributional shift may be large; it may be more culturally appropriate and sustainable to seek a general change in behavioral norms and in the social values that facilitate their adoption than to attempt to individu- ally change behaviors that are socially conditioned.72 Main limitations of the population strategy are that it offers only a small benefit to each participating individual, which may be wiped out by a small risk; it requires major changes in the economics and mode of functioning of society, which often makes changes unlikely. Individuals generally prefer to pay as late as possible and to enjoy the benefits as soon as possible. Social benefits— which are often achieved through processes with the opposite timing of costs and benefits—may thus be scarcely attractive to the individual. Nevertheless, shared values and targets do exist at the community level.19 See also common good.
STRATIFICATION The process of or result of separating a sample into several subsam- ples according to specified criteria, such as age groups, socioeconomic status, etc. The effect of confounding variables may be controlled by stratifying the analysis of results. For example, lung cancer is known to be associated with smoking. To examine the pos- sible association between urban atmospheric pollution and lung cancer, controlling for smoking, the population may be divided into strata according to smoking status. The association between air pollution and cancer can then be appraised separately within each stratum. Stratification is used not only to control for confounding effects but also as a way of detecting modifying effects. In this example, stratification makes it possible to examine the effect of smoking on the association between atmospheric pollution and lung cancer. See also adjustment; effect modification; standardization.
STRATIFIED RANDOMIZATION A randomization procedure in which strata are identified and subjects randomly allocated within each.This produces a situation intermediate between paired allocation and simple random allocation. See also random allocation; blocked randomization.
STRENGTH OF AN ASSOCIATION See Hill’s criteria of causation; measure of asso- ciation.
STRESS The result of a process through which environmental demands challenge, strain, and exceed the adaptive capacity of a person or community, resulting in psychological, physiological, or clinical changes that place persons at risk for adverse health events. Distributions of stress depend on structural, interpersonal, cognitive, biological, and physical processes. Responses to stressors may favor survival and adaptation.
STROBE Strengthening the Reporting of Observational Studies in Epidemiology. An evidence-based and structured approach to reporting of analytical observational studies. Recommendations on what should be included in an accurate and complete report of cohort studies, case-control studies, and cross-sectional studies (www.cochrane. dk).95,373 See also consort; moose; stard; trend.
STRUCTURED ABSTRACT An abstract or summary of a scientific article or report that is organized or structured in well-defined sections. A typical sequence of sections includes some or all of the following: “Objectives” or “Aims,” “Design,” “Setting,” “Sub- jects,” “Main outcome measures,” “Results,” and “Conclusions.” The structured abstract is intended to be comprehensive and to provide a logical order for the presentation of a scientific communication. Structured abstracts are required by many journals.
STUDY BASE The persons or person-time in which the outcomes of interest are observed. The population experience actually captured (“harvested”) by a study. In case-control studies, cases and controls should be representative of the same base experience.
Some authors, like Oli Miettinen (b.1936),5 distinguish between primary and secondary bases; in the former, the population experience is defined in time and place; in the latter, the cases are defined before the study base is or can be defined. In a clinical trial, the base is the follow-up experience of the patients actually enrolled in the study.
STUDY DESIGN See research design.
SUBCLINICAL DISEASE See disease, subclinical.
SUFFICIENT CAUSE A set of conditions, factors, or events sufficient to produce a given outcome. A complete causal mechanism that does not require the presence of any other determinant in order for an outcome, such as disease, to occur.88 See also association; causality; causation of disease, factors in; component causes; diseases of complex etiology; Evans’s postulates; Hill’s criteria of causation; necessary cause.
SUMMATIVE RATING A rating scale based on measurements of individually scaled items that are monotonically related to an underlying attribute or attributes; the sum of the item scores is approximately linearly related to the attribute.
SUPERINFECTION Fresh infection in a host already infected with a parasite of the same species; a term mainly used in malaria epidemiology.
SUPPRESSION BIAS, OPPRESSION BIAS Bias that result when actions aimed at obstructing the conduct or publication of research produce a bias in the available evi- dence (e.g., on the relationships between exposures and outcomes). An organization, a group, or an individual whose priority interests are not consistent with those of public health can be obstructive. Suppression bias may lead to publication bias. It undermines public health because it distorts and delays the discovery of scientific knowledge on health risks and compromises credibility in science and administrative processes for assessing and preventing exposure to hazards or risks. See also scientific misconduct.
SUPPRESSOR VARIABLE A variable that is causally related to the outcome of inter- est and, because it is associated with a causal variable, suppresses the study exposure’s association with the outcome.12 One variety of confounding variable. For example, smoking can be a suppressor variable for pesticide-related Parkinson’s disease.
SURVEY INSTRUMENT The interview schedule, questionnaire, medical examination record form, etc., used in a survey.
SURVIVAL ANALYSIS A class of statistical procedures for estimating the survival function and making inferences about the effects on it of treatments, prognostic factors, exposures, and other covariates. The proportional hazards model and the Kaplan- Meier estimate are examples of tools for survival analysis.
SURVIVAL CURVE A curve that starts at 100% of the study population and shows the percentage of the population still surviving at successive times for as long as informa- tion is available. May be applied not only to survival as such but also to the persistence of freedom from a disease or complication or some other endpoint.
SURVIVAL FUNCTION (Syn: survival distribution) A function of time, usually denoted by S(t), that starts with a population 100% well at a particular time and provides the percentage of the population still well at later times. Survival functions may be applied to any discrete event; for example, disease incidence or relapse, death, or recovery after onset of disease (in which case the population is initially 100% diseased, and the “sur- vival” function gives the percentage still diseased). See also Kaplan-Meier estimate.
SURVIVAL PROPORTION The proportion of a closed population at risk for a disease that does not become diseased during a specified interval, i.e., 1 minus the incidence proportion.12
SURVIVAL RATE (Syn: cumulative survival rate, survival proportion) The proportion of survivors in a group (e.g., of patients studied and followed over a specified period). The proportion of persons in a specified group alive at the beginning of the time interval (e.g., a 5-year period) who survive to the end of the interval. It is equal to 1 minus the cumulative death rate. May be studied by current or cohort life table methods.
SURVIVAL RATIO The probability of surviving between one age and another; when computed for age groups, the ratios correspond to those of the person-years-lived function of a life table.
SURVIVAL, RELATIVE Adjustment of survival rate for independent cause(s) of death. Multiple regression models of relative survival take into account the mortality from all other causes in each area, permitting better comparisons of survival within and between populations with different life expectancies.375
SURVIVORSHIP STUDY Use of a cohort life table to provide the probability that an event, such as death, will occur in successive intervals of time after diagnosis and, conversely, the probability of surviving each interval. The multiplication of these probabilities of survival for each time interval for those alive at the beginning of that interval yields a cumulative probability of surviving for the total period of study.
In many living organisms, including humans, a clinically and epidemiologically meaningful increase in susceptibility to disease cannot be assumed only on the basis of mechanistic studies (since, for instance, studies often lack design characteristics required to estimate baseline risk and risk differences, and because significant changes in phenotype are prevented by robustness, redundancy, and compensatory mechanisms).207,347,376–378 Assessment of the biological, clinical, and epidemiological coherence of research findings helps to prevent overestimates of susceptibility to disease. See also life course; nonmaleficence; sensitive period.
SUSCEPTIBLE VARIABLE A variable that is potentially confounding in that it is subsequent, not antecedent, to the variable whose effect is being studied. It may or may not be an intermediate variable as well; if it is, special tools such as marginal structural models must be used to adjust for its confounding effects.
SUSTAINABILITY The ability to continue economic, social, cultural, and environmental aspects of human society and the nonhuman environment. The Brundtland Commis- sion, led by the former Norwegian Prime Minister Gro Harlem Brundtland, defined sustainable development as development that “meets the needs of the present without compromising the ability of future generations to meet their own needs.” Common principles in action programs to achieve sustainable development, sustainability, or sustainable prosperity include dealing transparently and systemically with risk, uncer- tainty, and irreversibility; ensuring appropriate valuation, appreciation, and restoration of nature; integration of environmental, social, human, and economic goals in policies and activities; equal opportunity; community participation; conservation of biodiversity and ecological integrity; intergenerational equity; commitment to best practices; no net loss of human capital and natural capital; good governance. To varying degrees all these principles are relevant to epidemiology.111,136–138
SYMBIOSIS The biological association of two or more species to their mutual benefit.
SYNDROME A complex of signs and symptoms that tend to occur together, often char- acterizing a disease.
Under this definition and definition 2 of antagonism, two factors may act synergistically in some persons and antagonistically in others.
SYSTEMATIC ERROR See bias.
SYSTEMATIC REVIEW See review, systematic.