(BQ) Part 2 book “Modern epidemiology” has contents: Surveillance, using secondary data, field methods in epidemiology, ecologic studies, social epidemiology, infectious disease epidemiology, genetic and molecular epidemiology, nutritional epidemiology, environmental epidemiology,… and other contents.
P1: TNL GRBT241-22 Printer: RRD GRBT241-v4.cls January 18, 2008 12:5 Section Special Topics 457 IV P1: TNL GRBT241-22 Printer: RRD GRBT241-v4.cls January 18, 2008 12:5 458 P1: TNL GRBT241-22 Printer: RRD GRBT241-v4.cls January 18, 2008 12:5 CHAPTER 22 Surveillance James W Buehler History of Surveillance 460 Objectives of Surveillance 462 Approaches to Surveillance Descriptive Epidemiology of Health Problems 462 Links to Services 464 Links to Research 464 Evaluation of Interventions 465 Planning and Projections 466 Education and Policy 467 Summary 467 Elements of a Surveillance System Case Definition 467 Population under Surveillance 469 Cycle of Surveillance 470 Confidentiality 470 Incentives to Participation 470 Surveillance Ethics 471 Summary 471 467 472 Active versus Passive Surveillance 472 Notifiable Disease Reporting 472 Laboratory-Based Surveillance 472 Volunteer Providers 473 Registries 473 Surveys 473 Information Systems 474 Sentinel Events 475 Record Linkages 475 Combinations of Surveillance Methods 475 Summary 476 Analysis, Interpretation, and Presentation of Surveillance Data 476 Analysis and Interpretation Presentation 478 Attributes of Surveillance Conclusion 479 476 479 P eople who manage programs to prevent or control specific diseases need reliable information about the status of those diseases or their antecedents in the populations they serve The process that is used to collect, manage, analyze, interpret, and report this information is called surveillance Surveillance systems are networks of people and activities that maintain this process and may function at local to international levels Because surveillance systems are typically operated by public health agencies, the term “public health surveillance” is often used (Thacker and Berkelman, 1988) Locally, surveillance may provide the basis for identifying people who need treatment, prophylaxis, or education More broadly, surveillance can inform the management of public health programs and the direction of public health policy (Sussman et al., 2002) When new public health problems emerge, the rapid implementation of surveillance is critical to an effective early response Likewise, as public health agencies expand their domain to include a broader spectrum of health problems, establishing surveillance is often a first step to inform priority setting for new programs Over time, surveillance is used to identify changes in the nature or extent of health problems and the effectiveness of public health interventions As a result, surveillance systems may grow from simple ad hoc arrangements into more elaborate structures 459 P1: TNL GRBT241-22 Printer: RRD GRBT241-v4.cls January 18, 2008 12:5 460 Section IV ● Special Topics The modern concept of surveillance was shaped by programs to combat infectious diseases, which depended heavily on legally mandated reporting of “notifiable” diseases (Langmuir, 1963) Health problems now monitored by surveillance reflect the diversity of epidemiologic inquiry and public health responsibilities, including acute and chronic diseases, reproductive health, injuries, disabilities, environmental and occupational health hazards, and health risk behaviors (Thacker and Berkelman, 1988) An equally diverse array of methods is used to obtain information for surveillance, ranging from traditional case reporting to adapting data collected primarily for other purposes, such as computerized medical care records Surveillance systems are generally called on to provide descriptive information regarding when and where health problems are occurring and who is affected—the basic epidemiologic parameters of time, place, and person The primary objective of surveillance is most commonly to monitor the occurrence of disease over time within specific populations When surveillance systems seek to identify all, or a representative sample of, occurrences of a health event in a defined population, data from surveillance can be used to calculate incidence rates and prevalence Surveillance can characterize persons or groups who are affected by health problems and identify groups at highest risk Surveillance is often used to describe health problems themselves, including their manifestations and severity, the nature of etiologic agents (e.g., antibiotic resistance of microorganisms), or the use and effect of treatments Populations under surveillance are defined by the information needs of prevention or control programs For example, as part of a hospital’s program to monitor and prevent hospital-acquired infections, the target population would be patients receiving care at that hospital At the other extreme, the population under surveillance may be defined as the global population, as is the case for a global network of laboratories that collaborate with the World Health Organization in tracking the emergence and spread of influenza strains (Kitler et al., 2002) For public health agencies, the population under surveillance usually represents residents within their political jurisdiction, which may be a city, region, or nation All forms of epidemiologic investigation require a balance between information needs and the limits of feasibility in data collection For surveillance, this balance is often the primary methodologic challenge As an ongoing process, surveillance depends on long-term cooperation among persons at different levels in the health delivery system and coordinating agencies Asking too much of these participants or failing to demonstrate the usefulness of their participation threatens the operation of any surveillance system and wastes resources Another dimension of this balance lies in the interpretation of surveillance data, regardless of whether surveillance depends on primary data collection or adaptation of data collected for other purposes Compared with data from targeted research studies, the advantage of surveillance data is often their timeliness and their breadth in time, geographic coverage, or number of people represented To be effective, surveillance must be as streamlined as possible As a result, surveillance data may be less detailed or precise compared with those from research studies Thus, analyses and interpretation of surveillance data must exploit their unique strengths while avoiding overstatement HISTORY OF SURVEILLANCE The modern concept of surveillance has been shaped by an evolution in the way health information has been gathered and used to guide public health practice (Table 22–1) (Thacker and Berkelman, 1992; Eylenbosch and Noah, 1988) Beginning in the late 1600s and 1700s, death reports were first used as a measure of the health of populations, a use that continues today In the 1800s, Shattuck used morbidity and mortality reports to relate health status to living conditions, following on the earlier work of Chadwick, who had demonstrated the link between poverty and disease Farr combined data analysis and interpretation with dissemination to policy makers and the public, moving beyond the role of an archivist to that of a public health advocate In the late 1800s and early 1900s, health authorities in multiple countries began to require that physicians report specific communicable diseases to enable local prevention and control activities, such as quarantine of exposed persons or isolation of affected persons Eventually, local reporting systems coalesced into national systems for tracking certain endemic and epidemic infectious diseases, and the term surveillance evolved to describe a population-wide approach to monitoring health and disease P1: TNL GRBT241-22 Printer: RRD GRBT241-v4.cls January 18, 2008 Chapter 22 ● ˆT A B L E ˆ 12:5 Surveillance 461 ˆ ˆ 22–1 Key Events in the History of Public Health Surveillance Date Late 1600s 1700s 1840–1850 1839–1879 Late 1800s 1925 1935 1943 Late 1940s 1955 1963 1960s 1980s 1990s and 2000s 2001 Events von Leibnitz calls for analysis of mortality reports in health planning Graunt publishes Natural and Political Observations Made upon the Bills of Mortality, which defines disease-specific death counts and rates Vital statistics are used in describing health increases in Europe Chadwick demonstrates relationship between poverty, environmental conditions, and disease Shattuck, in report from Massachussets Sanitary Commission, relates death rates, infant and maternal mortality, and communicable diseases to living conditions Farr collects, analyzes, and disseminates to authorities and the public data from vital statistics for England and Wales Physicians are increasingly required to report selected communicable diseases (e.g., smallpox, tuberculosis, cholera, plague, yellow fever) to local health authorities in European countries and the United States All states in the United States begin participating in national morbidity reporting First national health survey is conducted in the United States Cancer registry is established in Denmark Implementation of specific case definition demonstrates that malaria is no longer endemic in the southern United States Active surveillance for cases of poliomyelitis demonstrates that vaccine-associated cases are limited to recipients of vaccine from one manufacturer, allowing continuation of national immunization program Langmuir formulates modern concept of surveillance in public health, emphasizing role in describing health of populations Networks of “sentinel” general practitioners are established in the United Kingdom and The Netherlands Surveillance is used to target smallpox vaccination campaigns, leading to global eradication WHO broadens its concept of surveillance to include a full range of public health problems (beyond communicable diseases) The introduction of microcomputers allows more effective decentralization of data analysis and electronic linkage of participants in surveillance networks The Internet is used increasingly to transmit and report data Public concerns about privacy and confidentiality increase in parallel with the growth in information technology Cases of anthrax associated with exposure to intentionally contaminated mail in the United States lead to growth in “syndromic surveillance” aimed at early detection of epidemics Adapted from Thacker SB, Berkelman RL History of public health surveillance In: Halperin W, Baker EL, Monson RR Public Health Surveillance New York: Van Nostrand Reinhold, 1992:1–15; and Eylenbosch WJ, Noah ND Historical aspects In: Eylenbosch WJ, Noah ND, eds Surveillance in Health and Disease Oxford: Oxford University Press, 1988:1–8 Important refinements in the methods of notifiable disease reporting occurred in response to specific information needs In the late 1940s, concern that cases of malaria were being overreported in the southern United States led to a requirement that case reports be documented This change in surveillance procedures revealed that malaria was no longer endemic, permitting a shift in public health resources and demonstrating the utility of specific case definitions In the 1960s, the usefulness of outreach to physicians and laboratories by public health officials to identify cases of disease and solicit reports (active surveillance) was demonstrated by poliomyelitis surveillance during the P1: TNL GRBT241-22 Printer: RRD GRBT241-v4.cls January 18, 2008 12:5 462 Section IV ● Special Topics implementation of a national poliomyelitis immunization program in the United States As a result of these efforts, cases of vaccine-associated poliomyelitis were shown to be limited to recipients of vaccine from one manufacturer, enabling a targeted vaccine recall, calming of public fears, and continuation of the program The usefulness of active surveillance was further demonstrated during the smallpox-eradication campaign, when surveillance led to a redirection of vaccination efforts away from mass vaccinations to highly targeted vaccination programs Throughout the 1900s, alternatives to disease reporting were developed to monitor diseases and a growing spectrum of public health problems, leading to an expansion in methods used to conduct surveillance, including health surveys, disease registries, networks of “sentinel” physicians, and use of health databases In 1988, the Institute of Medicine in the United States defined three essential functions of public health: assessment of the health of communities, policy development based on a “community diagnosis,” and assurance that necessary services are provided, each of which depends on or can be informed by surveillance (Institute of Medicine, 1988) In the 1980s, the advent of microcomputers revolutionized surveillance practice, enabling decentralized data management and analysis, automated data transmission via telephone lines, and electronic linkage of participants in surveillance networks, as pioneered in France (Valleron et al., 1986) This automation of surveillance was accelerated in the 1990s and early 2000s by advances in the science of informatics and growth in the use of the Internet (Yasnoff et al., 2000) In the early 2000s, the increasing threat of bioterrorism provided an impetus for the growth of systems that emphasized the earliest possible detection of epidemics, enabling a timely and maximally effective public health response These systems involve automation of nearly the entire process of surveillance, including harvesting health indicators from electronic records, data management, statistical analysis to detect aberrant trends, and Internet-based display of results Despite this emphasis on informatics, the interpretation of results and the decision to act on surveillance still requires human judgment (Buehler et al., 2003) While the balance between privacy rights and governments’ access to personal information for disease monitoring has been debated for over a century, the increasing automation of health information, both for medical care and public health uses, has led to heightened public concerns about potential misuse (Bayer and Fairchild, 2000; Hodges et al., 1999) This concern is exemplified in the United States by the implementation in 2003 of the privacy rules of the Health Insurance Portability and Accountability Act of 1996, which aim to protect privacy by strictly regulating the use of electronic health data yet allowing for legitimate access for public health surveillance (Centers for Disease Control and Prevention, 2003a) In the United Kingdom, the Data Protection Act of 1998, prompted by similar concerns, has called into question the authority of public health agencies to act on information obtained from surveillance (Lyons et al., 1999) As the power of information technologies grow, such controversies regarding the balance between public health objectives and individual privacy are likely to increase in parallel with the capacity to automate public health surveillance OBJECTIVES OF SURVEILLANCE DESCRIPTIVE EPIDEMIOLOGY OF HEALTH PROBLEMS Monitoring trends, most often trends in the rate of disease occurrence, is the cornerstone objective of most surveillance systems The detection of an increase in adverse health events can alert health agencies to the need for further investigation When outbreaks or disease clusters are suspected, surveillance can provide a historical perspective in assessing the importance of perceived or documented changes in incidence Alternatively, trends identified through surveillance can provide an indication of the success of interventions, even though more detailed studies may be required to evaluate programs formally For example, the effectiveness of the national program to immunize children against measles in the United States has been gauged by trends in measles incidence Following the widespread use of measles vaccine, measles cases declined dramatically during the 1960s In 1989–1990, however, a then-relatively large increase in measles cases identified vulnerabilities in prevention programs, and subsequent declines demonstrated the success of redoubled vaccination efforts (Centers for Disease Control and Prevention, 1996) (Fig 22–1) P1: TNL Printer: RRD January 18, 2008 Chapter 22 500 ● 12:5 Surveillance 463 Vaccine licensed MEASLES — by year, United States, 1981–1996 Reported Cases (Thousands) GRBT241-v4.cls 450 400 Reported Cases (Thousands) GRBT241-22 350 300 250 30 25 20 15 10 1981 1986 1991 1996 Year 200 150 100 50 1961 1966 1971 1976 1981 1986 1991 1996 Year FIGURE 22–1 ● Measles, by year of report, 1961–1996, United States (Reproduced from Centers for Disease Control and Prevention Summary of notifiable diseases, United States, 1996 Morb Mortal Wkly Rep 1996;45:43.) Information on the common characteristics of people with health problems permits identification of groups at highest risk of disease, while information on specific exposures or behaviors provides insight into etiologies or modes of spread In this regard, surveillance can guide prevention activities before the etiology of a disease is defined This role was demonstrated in the early 1980s, when surveillance of the acquired immunodeficiency syndrome (AIDS) provided information on the sexual, drug using, and medical histories of people with this newly recognized syndrome Surveillance data combined with initial epidemiologic investigations defined the modes of human immunodeficiency virus (HIV) transmission before HIV was discovered, permitting early prevention recommendations (Jaffe et al., 1983) Equally important, the observation that nearly all persons with AIDS had an identified sexual, drug-related, or transfusion exposure was effective in calming public fears about the ways in which the disease was not transmitted, i.e., that the presumed infectious agent was not transmissible via casual contact or mosquito bites Detection of outbreaks is an often-cited use of surveillance In practice, astute clinicians commonly detect outbreaks before public health agencies receive and analyze information on case reports This pattern has been often been the case for clusters of new diseases, including toxic shock syndrome, legionnaires disease, and AIDS Contacts between health departments and clinicians engendered by surveillance, however, can increase the likelihood that clinicians will inform health departments when they suspect that outbreaks are occurring Some outbreaks may not be recognized if individual clinicians are unlikely to encounter a sufficient number of affected persons to perceive an increase in incidence In such instances, surveillance systems that operate on a broad geographic basis may detect outbreaks Such detection occurred in 1983 in Minnesota, where laboratory-based surveillance of salmonella infections detected an increase in isolates of a particular serotype, Salmonella newport Subsequent investigation of these cases documented a specific pattern of antibiotic resistance in these isolates and a link to meat from cattle that had been fed subtherapeutic doses of antibiotics to promote growth (Holmberg, 1984) The results of this investigation, which was triggered by findings from routine surveillance in one state, contributed to a national reassessment of policies in the United States regarding the use of antibiotics in animals raised for human consumption P1: TNL GRBT241-22 Printer: RRD GRBT241-v4.cls January 18, 2008 464 12:5 Section IV ● Special Topics The development of so-called syndromic surveillance systems to detect bioterrorism-related epidemics as quickly as possible has emphasized automated tracking of disease indicators that may herald the onset of an epidemic These systems monitor nonspecific syndromes (e.g., respiratory illness, gastrointestinal illness, febrile rash illness) and other measures (e.g., purchase of medications, school or work absenteeism, ambulance dispatches) that may increase before clinicians recognize an unusual pattern of illness or before illnesses are diagnosed and reported Whether these approaches offer a substantial advantage over traditional approaches to epidemic detection has been controversial (Reingold, 2003) Data may also be collected on the characteristics of the disease itself, such as the duration, severity, method of diagnosis, treatment, and outcome This information provides a measure of the effect of the disease and identification of groups in whom the illness may be more severe For example, surveillance of tetanus cases in the United States in 1989–1990 documented that deaths were limited to persons >40 years of age and that the risk of death among persons with tetanus increased with increasing age This observation emphasized the importance of updating the immunization status of adults as part of basic health services, particularly among the elderly (Prevots et al., 1992) Among patients with end-stage kidney disease receiving care in a national network of dialysis centers in the United States, surveillance of a simple indicator that predicts the risk of morbidity and reflects the sufficiency of dialysis (reduction in blood urea levels following dialysis) identified centers with subpar performance levels For those centers with relatively poor performance, targeted quality improvement efforts led to subsequent improvement (McClellan et al., 2003) By describing where most cases of a disease occur or where disease rates are highest, surveillance provides another means for targeting public health interventions Depicting surveillance data using maps has long been a standard approach to illustrate geographic clustering, highlight regional differences in prevalence or incidence, and generate or support hypotheses regarding etiology A classic example is the use of maps by John Snow to support his observations that cholera cases in London in 1854 were associated with consumption of drinking water from a particular well, the Broad Street pump (Brody et al., 2000) In the United States, men of African descent have higher rates of prostate cancer compared with other men, and death rates for prostate cancer are highest in the Southeast (Fig 22–2) This observation, coupled with observations that farmers are at increased risk for prostate cancer and that farming is a common occupation in affected states, prompted calls for further investigation of agricultural exposures that may be linked to prostate cancer (Dosemeci et al., 1994) LINKS TO SERVICES At the community level, surveillance is often an integral part of the delivery of preventive and therapeutic services by health departments This role is particularly true for infectious diseases for which interventions are based on known modes of disease transmission, therapeutic or prophylactic interventions are available, and receipt of a case report triggers a specific public health response For example, notification of a case of tuberculosis should trigger a public health effort to assure that the patient completes the full course of therapy, not only to cure the disease but also to minimize the risk of further transmission and prevent recurrence or emergence of a drug-resistant strain of Mycobacterium tuberculosis In countries with sufficient public health resources, such a report also prompts efforts to identify potential contacts in the home, workplace, or school who would benefit from screening for latent tuberculosis infection and prophylactic therapy Likewise for certain sexually transmitted infections, case reports trigger investigations to identify, test, counsel, and treat sex partners Thus, at the local level, surveillance not only provides aggregate data for health planners, it also serves to initiate individual preventive or therapeutic actions LINKS TO RESEARCH Although surveillance data can be valuable in characterizing the basic epidemiology of health problems, they seldom provide sufficient detail for probing more in-depth epidemiologic hypotheses Among persons reported with a disease, surveillance may permit comparisons among different groups defined by age, gender, date of report, etc Surveillance data alone, however, not often P1: TNL GRBT241-22 Printer: RRD GRBT241-v4.cls January 18, 2008 Chapter 22 ● 12:5 Surveillance 465 NYC Age-specific rate per 100,000 population >268.3–369.7 >205.4–268.3 >33.1–205.4 >1.5– 33.1 0.1– 1.5 FIGURE 22–2 ● Prostate cancer death rates, by place of residence, black males, age 70 years, 1988–1992 United States (Reproduced from Pickle LW, Mungiole M, Jones GK, White AA Atlas of United States Mortality Hyattsville, MD: National Center for Health Statistics; 1996 DHHS Publication No (PHS) 97-1015, p 67.) provide a comparison group of people without the health problem in question Nonetheless, surveillance can provide an important bridge to researchers by providing clues for further investigation and by identifying people who may participate in research studies This sequence of events occurred shortly after the detection of an epidemic of toxic shock syndrome in 1979 Rapidly initiated surveillance illustrated that the outbreak was occurring predominantly among women and that disease onset was typically during menstruation (Davis et al., 1980) This finding led to case-control studies that examined exposures associated with menstruation These studies initially found an association with tampon use and subsequently with use of a particular tampon brand This information led to the recall of that tampon brand and recommendations concerning tampon manufacture (Centers for Disease Control, 1990d) EVALUATION OF INTERVENTIONS Evaluation of the effect of public health interventions is complex Health planners need information about the effectiveness of interventions, yet full-scale evaluation may not be feasible By charting trends in the numbers or rates of events or the characteristics of affected persons, surveillance may provide a comparatively inexpensive and sufficient assessment of the effect of intervention efforts In some instances, the temporal association of changes in disease trends and interventions are so dramatic that surveillance alone can provide simple and convincing documentation of the effect of an intervention Such was the case in the outbreak of toxic shock syndrome, when cases fell sharply following removal from the market of the tampon brand associated with the disease (Fig 22–3) In other instances, the role of surveillance in assessing the effect of interventions is less direct For example, the linkage of information from birth and death certificates is an important tool in the surveillance of infant mortality and permits monitoring of birth-weight-specific infant death rates This surveillance has demonstrated that in the United States, declines in infant mortality during the latter part of the 20th century were due primarily to a reduction in deaths among small, prematurely born infants Indirectly, this decline is a testament to the effect of advances in specialized obstetric and newborn care services for preterm newborns In contrast, relatively little progress has been made in reducing the proportion of infants who are born prematurely (Buehler et al., 2000) P1: TNL GRBT241-22 Printer: RRD GRBT241-v4.cls January 18, 2008 12:5 466 Section IV ● Special Topics FIGURE 22–3 ● Reported cases of toxic shock syndrome, by quarter: United States, January 1, 1979, to March 31, 1990 (Reproduced from Centers for Disease Control Reduced incidence of menstrual toxic-shock syndrome—United States, 1980–1990 Morb Mortal Wkly Rep 1990;39:421–424.) Following recognition of widespread HIV transmission during the late 1980s and early 1990s in Thailand, the Thai government instituted a multifaceted national HIV prevention program Surveillance data demonstrated that one element of this program—aggressive promotion of condom use for commercial sex encounters—was associated with an increase in condom use and parallel declines in HIV and other sexually transmitted infections among military conscripts, one of several sentinel populations among whom HIV trends had been monitored Although this observation provides compelling support for the effectiveness of the condom promotion strategy, it is impossible to definitively parse attribution among various program elements and other influences on HIV risk behaviors (Celentano et al., 1998) PLANNING AND PROJECTIONS Planners need to anticipate future demands for health services Observed trends in disease incidence, combined with other information about the population at risk or the natural history of a disease, can be used to anticipate the effect of a disease or the need for care During earlier years of the global HIV epidemic, widespread transmission was not manifest because of the long interval between the asymptomatic phase of HIV infection and the occurrence of severe disease In Thailand, HIV prevention programs noted earlier were prompted by findings from a comprehensive system of HIV serologic surveys during a period when the full effect of HIV infection on morbidity and mortality was yet to be seen These surveys, established to monitor HIV prevalence trends, revealed a dramatic increase in HIV infections among illicit drug users in 1988, followed by subsequent increases among female sex workers, young men entering military service (most of whom were presumably infected through sexual contact with prostitutes), women infected through sexual contact with their boyfriends or husbands, and newborn infants infected through perinatal mother-to-infant transmission (Weninger et al., 1991) The implications of these data, both for the number of future AIDS cases and the potential for extension of HIV transmission, prompted the prevention program Techniques for predicting disease trends using surveillance data can range from the application of complex epidemiologic models to relatively simple strategies, such as applying current disease rates to future population estimates The World Health Organization used this latter strategy to predict global trends in diabetes through 2025, applying the most recently available age- and country-specific diabetes prevalence estimates obtained from surveillance and other sources to P1: TNL/OTB GRBT241-INDEX P2: IML/OTB QC: IML/OTB GRBT241-v4.cls T1: IML February 5, 2008 Printer: RRD 14:38 744 Impact fractions, 67 See also Attributable fractions; Imprinting, 639 Incentives, 498, 498t Incidence, 47–48 model-based estimates, 439–440 Incidence density See Incidence rate Incidence measures, 33–36, 40–45 prevalence v., 46 relations among, 41–45 Incidence odds, 41 case-control studies, 113 risk-based measures and, 58 models, 394–395 Incidence proportion, 33, 40–41 See also Risk case-control studies, 113–122 causes and, 11, 11t component causes and, 12, 12t case-cohort study of, 123–124 models See Risk models Incidence rate, 33, 34–35, 40, 41, 42 absolute, 40 case-control studies, 113 incidence times v., in special populations, 37f, 39 models, 395–398 person time and, 34, 101 of population, 34 proper interpretation of, 35–36 of recurrent events, 36 Incidence rate difference See Rate difference Incidence rate ratio See Rate ratio Incidence studies See Cohort studies Incidence time, 33–34 models, 396–398 ratios, 53, 397 Incident case-control studies, 94 See Case-control studies Income, 535 Incomplete matching, 177 See also Marginal matching; Partial matching Incremental plots (Slope plots), 311–312, 312f Incubation periods, 551 agents with, infection occurrence, 557, 557f persistent infection v., progression-related states, 551 Independence See also Association; D-Separation conditional, 190 marginal, 184,188 statistical, 184–185 Independent censoring, 34–35, 45, 289–290 Independent competing risks, 44–45, 289–290 Independent error, 138 Independent outcome, 239 Independent variables See also Covariate; Regressor in regression, 383 statistically, 184–185 Index condition, counterfactual measures and, 55 Index level, regressors and, 407 Indicator variables, 407–408 Indirect adjustment See Standardized morbidity ratio Indirect effects, 200 direct effect v., 545, 545f sufficient cause model and, 13 Indirectly adjusted rate, 665 Individual matching, 170–171, 178 Individual risks, 40 Index Individual studies, statistical reanalysis of, meta-analysis, 659–668 Individual-level analysis, 512 Induction period, 15–17, 603 analysis of, 300–302 Induction, 18–19 Inductivism, 18–19 Refutationism v., 20 Inequality, 540–541 Infection, 549 See also Preinfectiousness; Subclinical infection cure, factors influencing, 559–560 death from, factors influencing, 559–560 disease v., infectious agents and, 551 factors influencing, 557–559 measures of frequency, 553–555 occurrence of, factors influencing, 557–559 occurrence of exposure, 556–557 persistent, factors influencing, 559–560 subclinical, 551 transmission, factors influencing, 561–562 transmission-related states, 551, 552 Infectious agents exposure occurrence, factors influencing, 555–556 persistent infection, 559 Infectious disease emerging, outbreak investigations, 553 epidemiology, 549–563 process, states of, 551 progression axis, 549, 550f public health, 617 surveillance and, 460 transmission models, 562–563 Infectivity factors, infection occurrence, 556 Inferential statistics, 221 See also Confidence limits; P-values data descriptors v., 216 Infertile worker effect, 622 Infertility, 624–628 Inflammatory bowel disease, genetic susceptibility, 577–578 Influence analysis, 221 meta-analysis and, 675 sensitivity, 221–222 Information accuracy, comparability of, 121 needs, epidemiologic investigation, 460 systems, surveillance, 474 weighting, 334–337 Information bias, 137–146 See also Misclassification Information-weighted averaging, 334–337 See also Inverse-variance weighting Informative cluster size, recognized loss, 630 Informed consent, 90 Initiator, in carcinogensis, 16 In-person interviews, questionnaires and, 503 Instrumental variables, 91, 202–204 Instrument-outcome association, 203 Intensity of disease, 34 See also hazard, incidence rate Intensity scoring, 449 Intentional selection, bias from, 196–198 Intent-to-treat analysis (ITT), 203, 647 P1: TNL/OTB GRBT241-INDEX P2: IML/OTB QC: IML/OTB GRBT241-v4.cls T1: IML February 5, 2008 Printer: RRD 14:38 Index Interaction, 10, 13, 71–83 See also Biologic Interaction; Public health interaction; Statistical interaction; Synergism analysis of, 298–300 Interaction contrast (IC), 75–79, 298–299 Interaction response types, 79–80, 80t Interaction terms, 406 See also Product terms Interactive voice response (IVR), 500 Intercept See Zero level, special handling in trend evaluation Intercommunity comparisons, social class, 611–612 Intermediate variables, 131–134, 186–188, 545–546 adjusting for, problems of, 131–134, 200–202, 260, 545–546 in meta-analysis, 656 Internal reference group, ratios derived, meta-analysis, 665–666 Internal validity, 128–129 International ecologic studies, limitations, 581 Internet See also Software sites, follow-up techniques, 508 surveillance data, 478 surveillance systems, 471 Interval estimation, 157–158, 163–167 See also Bayesian intervals; Confidence intervals; Likelihood intervals Intervening variables See Intermediate variables Intervention measure, 648–649 Intervention studies, recruitment to, 494 Interview response rates, epidemiologic research, 497–498 Interviews See also Recall bias longitudinal studies, 503–504 response rates and, 509 techniques and training, 504 Inverse-probability weighting (IPW), 266, 444–445, 449 See also Doubly robust estimation; Marginal models; Model-based standardization exposure models, standardization using, 444–445 and propensity scores, 445 stabilized, 446 Inverse-variance weighting, 334–337, 670–671 See also Information-weighted averaging IP See Incidence proportion IPW See Inverse-probability weighting Item non-response See Missing values, handling ITT association See Intent-to-treat association IVR See Interactive voice response Join points, of spline, 411 Joint action, 82 See also Interaction, concepts of Joint analysis, of multiple exposure levels, 323 Joint confidence regions, 235–236, 324, 324f Joint effects, model searching, 419 Joint hypothesis, 237, 307, 322–327 Joint null hypothesis, Pearson χ statistic, 307 Joint P-value, 235 Joint trend statistic, 326 Kaplan-Meier formula, 42–43, 290–292 censoring and, 45, 290 Kernels, 314, 318–320, 440, 441 Knots, of spline, 411 Kuhn, Thomas, philosophy of, 21 745 Laboratory-based surveillance, 472–473 Lagged analysis, 301 Large-sample methods, 216, 220, 225–230, 240–253, 263, 267–282, 307, 314–316, 422 and model fitting, 422–423 person-time data, 240–245 pure count data, 245–253 sparse data and, 263 trend statistics, 314–316 Latency of effects, 16, 603 preinfectiousness v., transmission-related states, 551 reactivation disease, 559 Latency periods, exposure assessment, 603 Latent effects model, 546 Latent period, 16, 603 See also Induction period Levels of analysis, ecologic studies, 512–513 Levels of inference, ecologic studies, 513 Levels of measurement, ecologic studies, 512 Life-course model, 546 Life expectancy, 34 “Life trajectory” effect, 547 Life-table analysis, 37, 290 See also Log-rank test; Product-limit estimator Likelihood function, 164, 165f, 227–228 conditional, 257, 272–273, 422, 433 partial, 422 statistical theory and, 227–229 Likelihood inference, pure, 164–165, 227–229 Likelihood intervals, pure, 164–165, 229 Likelihood ratio, 228–231, 425–426 in Bayesian analysis, 230–231 confidence limits, 229–230 statistics, 229–230 tests, 229–230, 425–426 Linear hypothesis, 314 Linear models, 392, 395, 396, 398 See also Generalized linear models in ecologic analysis, 517, 520 in hierarchical regression, 436 for odds, 395 for rates, 396 for risks, 392 for trends, 398, 402–403 Linear predictor, 416 Linear risk models, 392, 395, 398, 402–403 Linear spline, 411 trend, 411 Link function generalized linear models, 416 secondary data, 486 Linkage analysis, 565 List-based sampling, 496 Locally linear regression, 441 LOESS procedures, 442 Log additive model See Log-linear model, Multiplicative model Logarithmic curve, 398 Logarithmic plot scales, 310–313, 313f Logic checks, 215, 509 Logistic distributions, 370 P1: TNL/OTB GRBT241-INDEX P2: IML/OTB QC: IML/OTB GRBT241-v4.cls T1: IML February 5, 2008 Printer: RRD 14:38 746 Logistic models, 394–395, 399, 414–415, 442 adjacent-category, 414–415 conditional, 433 continuation ratio, 415 cumulative-odds, 415 group-specific, 442 exact, 422 hierarchical, 435–439 marginal, 446 for matched data, 434 ordinal, 414–415 polytomous, 413–414 proportional-odds, 415 stratified, 433 Logistic transform, 246, 391 Logit-logistic distribution, 371 Logit-normal distribution, 371 Logit, 231, 246, 391 Logit models, 395, 414–415 See also Logistic models Log-likelihood function, 165 Log-linear hypothesis, 314 Log-linear models, 395, 416, 420 See also Multiplicative models for counts, 423, 444 ecologic studies and, 517 genetic factors, 638 for incidence times, 396 for odds, 394, 416 for rates, 396, 416 for risks, 393, 416 Log-log risk model, 417 Log-rank test, 295 Log-risk scale, 74 Longitudinal causal modeling, 46, 453–455 Longitudinal cohort studies, 569 See also Cohort studies Longitudinal data, 46 modeling, 451–455 Longitudinal monitoring, 489 Longitudinal sampling, 116 See also Density sampling Longitudinal studies See also Cohort studies follow-up techniques, 507–508 interviews, 503–504 Longitudinal trends, 606, 608 Log-likelihoods See Deviance statistics; Likelihood functions; Likelihood ratios Log-rank test, 295 Lorenz curve, 540, 541f, 542 Loss to follow-up, 100–101, 108–109, 289 See also Censoring Low birth-weight paradox, 633 Lower P-value, 220 Lower-tailed P-value, 152, 220 LOWESS procedures, 442 Lung cancer, 18, 103, 263, 604, 607f, 609 Lyme disease, culture technique and, 553 “M” diagram, 188f Malformations, prevalence and, 47 Mantel trend statistic, 314–316 Mantel-Haenszel methods, 271 case-cohort, 281 homogeneity assumption, 271 Index model fitting, 433 odds ratio, 276, 276t rate difference, 273 rate ratio, 273 risk difference, 274–275 risk ratio 274–275, 281, 285 sparse-data, 272, 275, 281 statistic for person-time data, 277 statistic for pure-count data, 278 statistic for polytomous exposures, 307 study size, 679 survival analysis, 294–295 two-stage data, 282 Mapping, disease rates, 608–609 Marginal averages, 387 Marginal independence, 184 Marginal matching, 182 Marginal models, 442, 446, 447, 450, 451, 544–545 See also Inverse probability weighting; Marginal structural models; Model based standardization Marginal structural models (MSM), 208, 446, 454–455 Marginally unbiased, 191 Marginal outcome modeling, 446 Markov-chain Monte Carlo (MCMC), 342 Matched data See also Matched pair; Matching analysis of, 283–288 modeling, 434–435 Matched designs, 171–182 See also Matching Matched pair analysis, 284–288 case-control data, 287–288 cohort data, 284–287 odds ratio, 287–288 risk ratio, 285 test, 286 Matched randomization, 175 Matching, 171–179 See also Matched data; Matched pairs; Overmatching in case-control studies, 175–179 cost of, 178 selection bias and, 174–176, 205, 205f in cohort studies, 174–175 effect of, 171–174, 205 information accuracy, indicators of, 181 marginal, 182 partial, 177, 182 purpose of, 171–174 stratification variables, 150 Matrix adjustment methods for misclassification, 360–361 Maximum likelihood estimate (MLE), 164, 221, 228, 271–272 See also Likelihood function; Likelihood ratio homogeneity assumption, 271 of homogenous measure, 271 model fitting and, 421, 425 overdispersion and, 422 priors, 342 score statistic and, 226 Maximum likelihood test statistic, 230 See Wald statistic Maximal models, 421 MCMC See Markov-chain Monte Carlo program McNemar test statistic, 286, 288 P1: TNL/OTB GRBT241-INDEX P2: IML/OTB QC: IML/OTB GRBT241-v4.cls T1: IML February 5, 2008 Printer: RRD 14:38 Index MCSA See Monte-Carlo sensitivity analysis of biases MDR method See Multifactor dimensionality-reduction method Measles vaccine prevention programs and, 462, 463f register, secondary data, 485–486 Measurement error, 137–138, 346, 352–353, 361 See also Misclassification in dietary and nutritional assessment, 591–592, 595, 596–597 Measures of association, 51–70, 57 See also Association Measures of effect v., 59–60, 59t, 185, 385–388 standardized measures of, 67–69 Measures of body composition, anthropometry, 594–595 Measures of occurrence, 33–35, 40, 46 Measures of effect, 51–56, 59–70; See also Causal effects; Effects; Effect-measure modification defining exposure in, 55 generalized, 66–67 measures of association v., 59–60, 59t, 185, 385–388 null state and, 54–56 regression, 385–386 relations among, 60–62 standardized, 67–69, 386–388 theoretical nature of, 54–55 Measures of frequency, infection and disease, 553–555 Median-unbiased estimates, 221, 224, 253, 255, 257 Mediating variables See Intermediate variables Medical testing, 643, 644t Medical-record abstracts, 499 Mendelian transmission, 565 Menopause, 622–623 Menstrual cycle, 623 Meta-analysis, 652–682 See also Publication bias goals of, 654–655 nature of, 652–653 protocol for, 655 quality scores, 679, 681 quantification of effects, 657–658 role and limitations, 681 statistical methods, 668–677 study identification, 656–657 summary statistics for, 670t vote counting, 680–681 Meta-regression methods, 673–677 Methods, classes of, 220 Mid-P-values, 232–233, 255–257 v continuity corrections, 232–233 Migrant studies, nutritional exposures, 582 Migration across groups, ecologic studies, 527 Mill, John Stuart, philosophy of, 19 Minimally sufficient conditioning sets, 192 Minimal models, 420–421 Miscarriage studies, 629 Misclassification, 138–144 analysis of, 352–361, 372 bias related to, quantification of, 352–361, 488 of confounders, 144–145 dependent, 138, 143, 145, 360 of multiple variables, 360 in meta-analysis, 662–663 multiple-bias analyses and, 363 Misspecification See Model specification; Specification bias; Specification error 747 Missing at random (MAR), 219 Missing completely at random (MCAR), 219 Missing data, 215, 219 bias, 199–200, 200f, 346 methods, 219 Missing-data indicators, bias from, 199–200, 219 Mixed effects model, 130, 529 See also Hierarchical regression in meta regression, 677 MLE See Maximum-likelihood estimate Mobility See Social mobility Model accuracy, 389–390, 419, 449–450 Model averaging, 427 Model-based estimates, 424, 439–440 Model-based standardization, 442–446; See also Inverse probability weighting; Marginal modeling; Regression standardization Model checking, 423–429, 425t Model combination, doubly robust estimation, 450–451 Model diagnostics, model checking and, 423, 427–428 Model fitting, 389–390, 421–423 background example, 390 Model parameter, 389 Model selection, 419–421 hierarchical regression and, 437–438 confounder scoring and, 449–450 Model sensitivity analysis, 346, 429 Model specification, 382, 389–390, 420–421, 449–450 See also Model selection Modifiers See Effect measure modification Modular questionnaire, multipurpose studies and, 503 Molecular epidemiology, 564–579 Monitoring of disease, 488–490, 490f Monotone trend, 308 Monte-Carlo sensitivity analysis (MCSA), 364–378 Bayesian analysis v., 378–380 combined, 376–378 intervals, 365 probabilistic bias analysis and, 364 simulation histograms, 368f Mortality rates follow-up, 109–110 monitoring disease, 488–490, 489f patterns of, 33, 33t time-series analyses, 611 Mosquito, environmental reservoir, 555 Moving To Opportunity study (MTO study), 547 Moving averages, 317–321 See also Smoothers and Smoothing categorical estimates as, 321 MRFIT See Multiple Risk Factor Intervention Trial MSM See Marginal structural model MTO study See Moving To Opportunity study Multidimensional outcome, 56 Multifactor dimensionality-reduction (MDR) method, gene-disease association, 572 Multifactorial causation and etiology, 5, 14 Multigeneration registers, secondary data, 484 Multilevel analyses, 513 and designs, ecologic studies, 528–530 Multilevel modeling, 337, 435–439, 529, 530, 542–544 See also Hierarchical regression Multi-linear regression, 674t Multiple bias analyses, 363–364, 376–378 P1: TNL/OTB GRBT241-INDEX P2: IML/OTB QC: IML/OTB GRBT241-v4.cls T1: IML February 5, 2008 Printer: RRD 14:38 748 Multiple comparisons, 234–237, 322–327 hierarchical regression and, 237 single-comparisons v., 326–327 in trend analyses, 307, 326 Multiple control groups, case-control studies, 122, 322 Multiple correlation, 424–425 Multiple diseases, analysis of See Multiple outcomes analysis Multiple group designs, ecologic studies, 514 Multiple group ecologic analysis, confounders, 518–519 Multiple group ecologic studies, within-group misclassification, 526 Multiple imputation, 219 Multiple inference procedures See Multiple comparisons Multiple logistic model, 401 See also Logistic models extensions of, 413–416 single-logistic model v., 401 Multiple outcomes analysis, 321–323, 322t, 325–326 See also Longitudinal data; Recurrent events simultaneous analysis of, 325, 325t Multiple probabilistic bias analysis, Monte-Carlo analyses and, 376, 377t, 378 Multiple regression, 384 models, 400–408 See also specific models, e.g., Logistic models trend models, 408–413 Multiple Risk Factor Intervention Trial (MRFIT), 92 Multiple testing See Multiple comparisons Multiplicative models, 404–405 See also Exponential models; Log-linear models Multiplicative-intercept models conditional fitting, 433 cumulative studies, 431 density studies, 430–431 prevalence studies, 431 Multipurpose studies, modular questionnaire, 503 Multistage model, 82 Multivariable modeling See Multiple regression; Multivariate regression Multivariate outcome, with competing risks, 56 Multivariate regression, 388 Mycobacterium avium, infectivity and, 558 Mycobacterium leprae, culture technique and, 553 Mycobacterium tuberculosis infection occurrence, 556 surveillance, 464 Nails, biochemical measurements, 594 Naive direct-effect analysis, 201 Narrative historical approach, 548 National Death Index (NDI), tracing, epidemiologic studies, 507 National genetic testing committees, terminology by, 566, 566t National Health and Nutrition Examination Survey See NHANES survey Natural direct effect, 200–201 Natural log function, exponential function v., 416 Naturalism, consensus and, 21–22 NDI See National Death Index Nearest-neighbor windows, 321 Necessary cause, 7, Negative adjustments, treatment of, in bias analysis, 373 Neighborhood controls, 117, 121 Index Neisseria meningitidis, colonization of, 556 Nelson-Aalen estimator, 292 See also Exponential formula; Kaplan-Meier formula Neonatal mortality, 633–635, 634f Nested case-control studies, 114, 122–123 and missing data models, 432 Nested confidence intervals, 157f Nested indicator coding, 407 Next-step investigations, surveillance data, 478 Neyman-Pearson hypothesis testing, 150, 153–155 See also Hypothesis test; P-values; Statistical significance statistical estimation and, 157–158 NHANES survey (National Health and Nutrition Examination Survey), 569, 611 surveillance, 474 Nodes See Vertices Nominal-scale variables, 214 Nonadherence See adherence Noncausal associations, 26 Noncollapsibility, 58, 62 See also Collapsibility confounding v., 62 Noncompliance See compliance g-estimation and, 454 Nondependent error, 138 See Independent error Nondifferential misclassification, 139, 372–373 of disease, 142–143 of exposure, 139–142 incidence-rate difference and, incidence-rate ratio and, 139t pervasiveness of misinterpretation, 143 requirements for, 355–356 with three exposure categories, 141t with two exposure categories, 140t Nonexperimental studies, 93–99 types of, 87–88 Noninformative priors, 167, 335–337, 343, 348, 379–380 Nonnested models, 426 Non-normal priors, 340, 370–372 Nonparametric bootstrapping, model checking and, 429 Nonparametric regression, 421, 440–442 See also Smoothers and Smoothing Nonrandomized clinical studies, 641, 650 Nonsignificance of a statistical test, proper interpretation of, 151–152 In meta-analysis, 657, 680 Notifiable disease reporting, 472 surveillance and, 460 Novum Organum, scientific method and, 18 Null hypergeometric distribution, 256, 257 Null hypothesis, 60, 88, 151, 156, 314 bias and, 144 P-value, 153, 315, 454 test for, 158 true, cross-level bias, 523, 525t Null state, effect measure and, 54–56 Null studies, publication bias and, meta-analysis and, 678–679 Nutrient content, measurement of, epidemiologic studies, 586–588 Nutrient, definition complexity, 580 P1: TNL/OTB GRBT241-INDEX P2: IML/OTB QC: IML/OTB GRBT241-v4.cls T1: IML February 5, 2008 Printer: RRD 14:38 Index Nutritional epidemiology, 580–597 methodological issues, 595–597 Nutritional exposures, 581–586 Observational studies, 19, 87–88, 332–333 confounding and, 129 Occam’s razor, 389 Occupation, as social variable, 537 Occupational cohort study, 495 Occupational epidemiology, 599 Occurrence measures, 33–50 Occurrence time, 33 See also Incidence time Odds models, 394–395 Odds ratio, 61, 62, 74 case-control studies, 127 pseudo-frequencies and, 113–114 Oil disease, 603 “Omic” tools, 564–566 One step study, 646 One-sided P-value, 156 One-sided tests, 156 One-tailed P-value, 152 Open populations, 38, 101 See also Dynamic population closed populations v., 38 steady state, 38–39 Open-ended categories, confounder stratification, 218, 304–305, 410 “Optimal” control distribution, case-control studies, 176, 177 Ordered variables, categorization of, 303–305 Ordering of classification, multiple-bias analyses and, 363 Ordinal logistic models, 414–415 Ordinal category scores, 312–313, 410 Outbreak bias, 553 Outbreak investigations, emerging infectious diseases, 553 Outcome events See also Causal effects; Effects; Outcome variables sufficient-component cause model and, timing of, exposure category and, 107–108 Outcome measures See specific measures, e.g., Incidence times; Rates Outcome models exposure models v., 449–450 standardization using, 443 Outcome scores, 447–448 See also Confounder scores Outcome transformations, 400 Outcome variables, 382–383 See also Causal effects; Effects; Outcome events; Regressand in meta-analysis, 655–656 in regression v causation, 383 transformed, 400 Outliers, 410 Overadjustment, 181, 260 Overconclusiveness, meta-analysis and, 677 Overdispersion, 422 Overmatching, 179–181, 260 bias and, 180–181 cost efficiency, 181 statistical efficiency and, 179–180 Ovulation, 623 Pair matching, 283–289 Pairwise blocking, 174 Parallel linear trends, 404 749 Partial likelihood, 422 Partial-Bayes analysis See Semi-Bayes analysis Partial matching, 177, 182 Partially ecologic analysis, 512, 512f Passive surveillance, 476 active surveillance v., 472 Pathway model, 546 Patterns of exposure, 66 attributable fractions, 67 distinguishing, 67 Pearson global test statistics, 427 Pearson χ statistic, 307 Penalized splines, 440–441 Penalized estimation, 271, 331, 341, 421, 423, 436, 437–439 See also Hierarchical regression; Shrinkage estimation maximum likelihood v., 271 Per protocol analysis, randomized controlled trials, 648 Percentile boundaries, problems with, 217–218, 264, 303–305 Perfect compatibility, in causal graphs, 191 Perinatal mortality, 635–636 Period effect, 606, 608 Persistent infection (Chronic infection), 552 incubation period v., progression-related states, 551 infectious agents and, 559 Persistent organic pollutants, 616–617 Personal identifiers, secondary data access, ethics of, 491 Person-count cohort data, 240 Person-time, 294 See also Follow-up analysis, follow-up data, 287 classifying, 102–103 data large-sample methods, 240–245 small-sample methods, 253–255 unstratified data with, 243, 243t distribution, 49–50, 68 exposure, 106 follow-up data, disease misclassification, 358–359 immortal, 106–107 rate, 34 at risk, 34, 39 units, classification of, 218–219 weights, Mantel-Haenszel estimation, 275 Physical examinations, epidemiologic studies, 504–505 Placebo, 91 equipoise and, 89 response, 91 Plausibility, causal inference and, 28–29 PMR See Proportional mortality ratio Point estimate, 156 Point prevalence, 46 Point sources, clustering analysis, 613 Point-source outbreaks, epidemic outbreaks, 552 Poisson distribution (model), 241–243 clusters, 612 Poisson regression, 421–422 See also Exponential models for rates case-parent studies, genotype distribution and, 575 extravariation, 422 genetic factors, 638 Policy, and bias analysis, 347, 380 Policy, surveillance data, 467 P1: TNL/OTB GRBT241-INDEX P2: IML/OTB QC: IML/OTB GRBT241-v4.cls T1: IML February 5, 2008 Printer: RRD 14:38 750 Polynomial regression, 410 Polytomous exposures, and outcomes, analysis of, 303–327 Polytomous logistic models, 413–414 Pooled estimate, 271 Popper, Karl, philosophy of, 20 Population See also Case selection; Complete populations; Confounding; General-population; Impact fractions; Incidence rate; Population-based case-control studies; Source population; Superpopulation; Target population closed, 36–37 cohorts v., 38 concepts of, 383 individual rates and, 34–35 open, 38, 101 regression, 383 at risk, 34 of recurrent events, 36 steady state, 38–39 source, 114–115, 128–129, 383 target, 60, 90, 146–147, 383, 471 under surveillance, 460 types of, 36–39 Population-attributable fractions, 67, 295–297 Population-attributable risk percent, 67 Population-average models, 544–545 Population-averaged regression, 387 Population-based case-control studies, 114, 123 See also Nested case-control studies Population-based prevalence study, 569 Population-based surveillance systems, 469–470 Population effects, 52 Population incidence rate, 34, 35 Population time at risk, 34 Post hoc ergo propter hoc fallacy, 19 Posterior probability, 23, 166–167, 330–331, 337, 345, 364, 379–380 Postexposure events, person-time allocation, 107 Potential biases See Systematic errors Potential outcomes, 54, 385–386 See also Counterfactual outcomes binary exposure variables and, 75, 76t causal models and, 60 causation, 18 model, 59–60, 59t in regression, 385–388 in standardization, 387–388 sufficient-cause models v., 81–82, 81f Poverty, 536 Power models, 410 Power of statistical test, 153–154 Pr See Probability Precision data stratification and, 150 definition of, 1498–149 statistical significance and, 162, 162f weighting, 271, 334–337, 670–671 See also Information-weighted averaging Prediction, 421 Prediction accuracy, 389–390, 436 Predictive values diagnostic tests, 643–644,644t Index exposure misclassification, 138, 353–354 negative, 138, 353, 643–644,644t positive, 138, 353, 643–644, 644t screening tests, 643–644, 644t sensitivity and specificity v., 357–358, 643–644, 644t surveillance systems, 479 Pregnancy See also Subclinical pregnancy loss complications, 632 hormone, 621 loss, 628–631 reproductive epidemiology, 620, 621 Preinfectiousness, latency v., 551 See also Infection Pretesting, nonresponse and, 499 Prevalence, 46–48 case-control studies, 97, 127 duration of disease, 47–48 etiologic research and, 46–47 model-based estimates, 439–440 multiplicative-intercept models, 431 odds, 48 pool, 46, 47 proportion, 46, 47 rate, 46 ratios, 69 studies, 88, 97 Preventable fractions, 54, 67 See also Attributable fractions Prevention programs, measles vaccine and, 462, 463f Preventive effects, 8, 9, 17, 54, 59, 65 See also Causal effects; Potential outcomes Primary base study, 114–115 Primary data, secondary data v., 481 Prior data, 337–342 See also Data priors; Priors Prior information, in model searching, 419–420 See also Bayesian analysis; Priors Prior limits, 330 See also Priors Prior model, in hierarchical regression, 436 Prior probability See Priors Prior parameters, Bayesian analysis and, 330 Prior standard deviation, hierarchical regression, 436–437 Prior data, 337–342 as diagnostic device, 338–339, 341–342 frequentist interpretation, 337–338 methods, extensions, 340 and reverse Bayes analysis, 338–339 Priors, 23, 166–167, 330–332, 337–343, 366–373, 378–380 See also Prior data; Probability; Probability distributions cautions, 342–343 correlated, 371–372 noninformative, 167, 335–337, 343, 348, 379–380 reference, 343 realistic, 335–336, 337, 365, 369–370, 371, 373 Privacy rights, surveillance, 462 Probabilistic bias analysis, 364–380 Probabilistic sensitivity analysis (PSA), 364–365 See also Monte-Carlo sensitivity analysis Bayesian analysis v., 378–380 Probability, 9–10 See also Likelihood; Probability distributions and densities; Risk conditional, 184–185 data, 331 frequency, 10, 331–332 P1: TNL/OTB GRBT241-INDEX P2: IML/OTB QC: IML/OTB GRBT241-v4.cls T1: IML February 5, 2008 Printer: RRD 14:38 Index in graphs, 187–190 marginal, 184 mass functions, 378 models for counts, 240–247 population, 184–185, 387 posterior, 23, 166–167, 330–331, 337, 345, 364, 379–380 prior, 23, 166–167, 330–332, 337–343, 366–373, 378–380 propensity, 10 standardized, 442–445 subjective, 10, 23–24, 165–167, 330–331 unconditional, 184 Probability distributions and densities, 184–185, 222–225, 370–372, 378–380 binomial, 223–225, 245–247, 254 graphical models for, 186–195 exact statistics and, 222–225 log F, 370 logistic, 367f, 370 logit logistic, 371 log normal, 371 logit normal, 371 normal, 367f parameter, 223 Poisson, 241–245 population, 184–185, 387 posterior, 23, 166–167, 330–331, 337, 345, 364, 379–380 prior, 23, 166–167, 330–332, 337–343, 366–373, 378–380 trapezoidal, 367f, 371 uniform, 365, 369 Probability of causation (of a case), 64–65, 297–298 Probit model, 395 Product terms, 402–407 See also Statistical interaction biologic interactions v., 407 interpreting, 405–406 trends and, 404–405 Product-limit estimator, 290–292 Nelson-Aalen estimator v., 292 Product-limit formula, 42–43, 42f, 42t, 43, 290–292 censoring and, 45, 290 Prognostic scores, 447 See also Confounder scores; Outcome scores Progression axis, infectious disease and, 549, 550f Progression-related studies, 553–560 Promoter, in carcinogenesis, 16 Propagated outbreaks, epidemic outbreaks, 552 Propensity scores, 445, 448–450 See also Confounder scores; Doubly robust estimation; Exposure scores inverse-weighting by, 445, 448 Proportional hazards model See Cox model Proportional mortality ratio (PMR), 97–98 Proportional mortality studies, 97–99, 127 Proportional odds model, 415 Prospective ascertainment, 96 Prospective studies, retrospective studies v., 95–97 Prostate cancer, surveillance, 464, 465f PSA See Probabilistic sensitivity analysis Pseudo-denominators, in case-cohort studies, 252–253 Pseudo-frequencies, 113–114, 125 Pseudo-likelihood, 422 751 Pseudo-rates, 113–114 Pseudo-risks, 123–124 Puberty, 622–623 Public health bias analysis and, 347, 380 effects, 14 functions of, 462 interaction, 83 interventions, evaluation, 465–466, 466f laws, notifiable disease reporting, 472 services, planning and projections, 466–467 surveillance, history of, 460–462, 461t Public resources, secondary data access, ethics of, 491 Publication bias In meta-analysis, 656–657, 678–679 from study exclusion, 679 Pure count data See Count Data Pure direct effect, 200–201 Pure likelihood equation, 230 Pure likelihood inference, 228 Pure likelihood limits, 229 P-values, 151–153, 156–163, 364 See also Confidence intervals and limits; Evidence of absence of an effect; Hypothesis tests; Joint P-value; Lower-P-value; Mantel trend P-value; Mid-P-values; Nonsignificance; One-sided P-value; Significance tests; Two-sided P-value Bayesian theory and, 166 confidence intervals and, 158–159 continuity corrected, 232 data and, 216 deviance, 229–230 exact, 220 function, 158–159, 160, 160f, 161f, 239 hypothetical data, 158t interpretation, 151–152, 159–163 likelihood ratio, 229–230 linear hypothesis, 314–316 lower-tailed, 152 Mantel trend, 314–316 mid, 232–233 misinterpretations, 151–152 multiple testing, 234–237, 322–327 null hypothesis and, 153, 315, 454 one-sided, 156 from Pearson χ statistic, 307 practice guidelines, 162–163 random error and, 156 score, 225–226 test statistics, 220–221 trend, 314–316 two-sided, 156, 221, 234 two-tailed, 152 types, 152–153 upper-tailed, 152 Wald, 220, 226–227 Quadratic model, 398 Quadratic spline, 411–412 Qualitative statistical interaction, 79–80 biologic interaction v., 299 Qualitative tally (vote counting), in meta-analysis, 680 Quality scoring, 681 P1: TNL/OTB GRBT241-INDEX P2: IML/OTB QC: IML/OTB GRBT241-v4.cls T1: IML February 5, 2008 Printer: RRD 14:38 752 Quantification of effects, 51–52 See also Effect measures in meta-analysis, 657–658 Quantiles, 217–218, 304 Quantitative criterion, variables and, 261 Quasi-experiments, 88–89, 547 Quasi-likelihood, 422 Questionnaires See also Data coding questionnaires; Data entry questionnaires; Food-frequency questionnaires; Modular questionnaire; Self-administered questionnaires; Semiquantitative food-frequency questionnaires administration method, 500–501 data and, 214 modular, multipurpose studies, 503 respondent burden, 503–504 R See Effective reproductive number (R) R2 , 424–425 Race/ethnicity, as social variable, 534–535 Racism, 537–538 Random allocation See Randomization Random coefficient models, 331, 423, 426, 427, 439 See also Hierarchical regression Random digit dialing (RDD), 117–118, 496 Random effects models, 452, 542–544, 628 See also Hierarchical regression in meta-analysis, 675–677 Random error, 24, 128, 148–167, 332–333; See also Confidence Limits; Probability distributions; P-values; Random sampling adjustment for, in bias analysis, 346, 366–369, 376, 379 conventional statistical analysis, 364 distributions, 222–225, 241–247, 421 P-values, 151–153, 156 P-value functions, 158–159 statistical precision and, 148–149 study size and, 149–150 systematic errors v., 346 Random sampling, 10, 213, 220, 222, 224, 346 See also Sampling error; Selection bias of cases, case-control studies, 115 of controls, case-control studies, 114, 122–123 for validation data, 361 Random variation, components, 148–149 See also Random error Randomization (Random allocation), 88–89 clinical studies, 641 d-separated unconditionally v., 187 experiments and, 168–169 of exposure assignment, 346 intervention studies, recruitment to, 494 Randomized experiments See Experimental studies Randomized recruitment, 604 Randomized trials, 18, 203f See also Experimental studies dietary hypothesis, 585–586 subject selection, 647–648 therapeutic interventions and, 646–649 Rare disease assumption in case-control studies, 114, 125 Rate fractions, 54, 63 See also Attributable fractions etiologic fractions v., 63–64, 297–298 probability of causation v., 64–65, 297–298 Index Rate, 34–36, 39–40 See also specific types, e.g., Case fatality rate; Incidence rate; Prevalence; Mortality rates standardized, 49–50 Rate difference, 52–53, 277 crude data, 243, 244, 245 Mantel-Haenszel, 273–274 standardized, 67–69, 266–267 Rate models, 395–398 in meta-analysis, 663 Rate ratio (RR), 61, 62, 116, 241 crude data, 243, 244, 245, 251, 254–255 ecologic studies and, 517, 518f Mantel-Haenszel, 273–274 relation to other ratio measures, 250 standardized, 67–69, 267–269 unconditional v conditional, 272 Ratio measures, 53 See also Rate ratio; Relative risk; Risk ratio relations among, 250 Rationally coherent probability assessments, 166 See also Bayesianism RDD See Random-digit dialing Reactivation disease, latency and, 559 Reanalysis, for meta-analysis and, 659–668, 669t Recall, in interviews, 96 Recall bias, 111, 112, 138, 356–357 birth defects, 637 Recentering variables, 392 Recognized loss, 628–630 Record abstraction, 499 Record linkage, 486 in surveillance, 475, 476 Record-level adjustment, in bias analysis, 373 Recurrent events, recurrent outcomes, 36, 452–453 See also Longitudinal data Red blood cells, biochemical measurements, 594 Reference level, regressors and, 407 Reference model, in checking model fit, 425 Reference priors, 343 See also Noninformative priors Referent, for causes, Refutationism, 20–21, 24, 30 Registries, use of in epidemiologic research, 481–491 surveillance, 473 of vaccinations, 488 Regressand, 383–384 See also Outcome variables in multivariate regression, 388 Regression, 382 Bayesian, 388 binary, 383–384 causal, 385–386 frequentist, 382–383, 388–389 hierarchical, 237, 263, 435–439, 529, 630 meta-analytic, 673–677 multiple, 384 multivariate, 388 ridge, 331, 423, 436, 437, 439 See also Shrinkage estimation standardization, 386–388, 442–445 Regression analysis, 381–382, 418–456 Regression measures of effect, 385–388 Regression functions, 382–389 P1: TNL/OTB GRBT241-INDEX P2: IML/OTB QC: IML/OTB GRBT241-v4.cls T1: IML February 5, 2008 Printer: RRD 14:38 Index Regression models, 381–417 See also specific entries, e.g., Generalized linear models; Linear models; Logistic models; Log-linear models biologic interactions and, 406–407 contingency table data and, 216 in meta-analysis, 658–659, 673–677 smoothed curves and, 321 transformed, 400 Regression splines, 410–412 See also Nonparametric regression; Power models; Smoothers and Smoothing Regressor, 383–384 transformed, 398–400 Regressor-specific predictions, fitted model v., 423 Relative birth weight, 635 Relative difference, 54, 65 Relative effect measures, 52 Relative excess measures, 53–54 Relative risk scale, meta-analysis and, 670 Relative risk (RR), 53, 67 See also Rate ratio; Risk ratio adjustment of, in Bayesian-analysis, 336 factorization of, meta-analysis, 660–661 relations among ratio effect measures, 60–61 single two-way table, 334–335, 334t Relative tests of regression models, 425 Repeated-measures, 451 Representativeness generalizability and, 146–147 in case-control studies, 120–121 surveillance systems, 479 Reproductive endpoints, 621–622 Reproductive epidemiology, 620–641 Reproductive systems, 620 Reproductively unhealthy worker effect, 622 Resampling, 368–369, 376 model checking and, 428–429 Rescaling variables, 392–393 Research biomarkers, ELSI and, 567–568 conduct, 21 links to, surveillance, 464–465 objective, 32–33 questions, meta-analysis, 655 resources, secondary data access, 491 secondary data access, ethics of, 490–491 sponsors, suppression bias and, 678 staff, occupational cohort study, 495 Research synthesis See Bias analysis; Meta-analysis Reservoirs, infectious agents and, 555 Residual confounding, 69, 198–199, 199f, 259 bias quantification and, 198–199 Residual degrees of freedom global tests of fit, 427 meta-regression methods, 673 Residual distributions, model fitting and, 421–422 Residual effects, 435 Residual sum of squares, 427 meta-regression, 673 Register-based search, secondary data, 488 Respondent burden of questionnaires, 503–504 Response rates epidemiologic research, 497–499 interviews and, 509 753 Response types, causal and preventive, 59–60 and additivity, 78–79 cohorts and, equivalent interaction contrast and, 80t sufficient cause models and, 82 Restriction in study design, 169 Retention time of toxic exposures, 602–603 Retrospective ascertainment, 96, 112 See also Recall bias Retrospective cohort selection, 109 Retrospective studies, prospective studies v., 95–97, 112 Reference event, in incidence times, 33 Reverse causation, 28, 30, 622 Reverse-Bayes analysis, 338–339 Reverse-continuation-ratio model, 415 Ridge regression, 331, 423, 436, 437, 439 See also Hierarchical regression; Shrinkage estimation Risk, 9–10 analysis, 347 assessment, 347 difference, 52–53 environmental hazards, 614 estimation, 290–293 fraction, 54, 65 history, 41, 42f models, 395 ratio, 61, 73, 116 score, 447 standardized, 49–50 Risk difference, 52–53, 260–261 crude data, 247, 249, 250 Mantel-Haenszel, 274–276 standardized, 68, 266–267 Risk-difference additivity, biologic interactions and, 78–79, 298–299, 406 Risk factors coronary heart disease, 14 lung cancer, 18 non-causal, 28 Risk fraction, 54 See also Excess fraction Risk models, 393–395, 417 Risk period, 33, 34 Risk ratio, 61, 73, 116 case-cohort data, 252–253, 280–281 crude data, 246, 247, 250 Mantel-Haenszel, 274–276 standardized, 68, 267–269 Risk scores, 447 See also Confounder scores, Outcome scores Risk-set sampling, 124, 125 RR See Rate ratio; Relative risk; Risk ratio Run-in phase, subject selection, randomized controlled trials, 647 Running mean, 317 Running-weighted regression curve, 321 See also Nonparametric regression; Smoothers and Smoothing Russell, Bertrand, philosophy of, 19 Salk vaccine field trial, Salmonella, infection occurrence, factors influencing, 556 Salmonella newport, surveillance, 463 Sample-size, considerations, 239 See also Study size model fitting and, 422–423 P1: TNL/OTB GRBT241-INDEX P2: IML/OTB QC: IML/OTB GRBT241-v4.cls T1: IML February 5, 2008 Printer: RRD 14:38 754 Sampling error, 149, 346 See also Random error; Random sampling; Selection bias fractions, case-control studies, 115, 123–125, 430–432 rates, case-control studies, 113, 430 variation, 149 SARS See Severe acute respiratory syndrome Saturated regression model, 427 Scale dependence of effect-measure modification, 72–74 of regression coefficients, 392–393 Scaling, of graphs, 310–311, 312–313, of regressors, 392–393 Scanned forms, 500 Scatterplot smoothers, 321 See also Nonparametric regression; Smoothers and Smoothing Scatterplots, 216, 321 of test statistics, 681 Schwarz information criterion See Bayesian information criterion Scientific inference, philosophy of, 18–25 Scientific proof, impossibility of, 24–25 Score limits, 226 Score statistics, 225–226 See also Mantel-Haenszel statistics for case-cohort data, 252 for case-control data, 251 for person-time data, 242–245 for pure count data, 245–250 Score test, 225–226 Scoring methods for confounder control See Confounder scoring Scoring of categories, 312–313, 410 Screening tests, studies, 642–646 SD See Standard deviation SE See Standard error Seasonal variation, 605, 608, 609, 611 See also Cyclic patterns Second stage model, hierarchical regression and, 438 Second stage standard deviations, hierarchical regression, 436 Secondary attack rate, 560–561 Secondary data, 481–491 access, ethics of, 490–491 analysis, complete populations, 482–483 use of, examples, 484 validity and, 483 Second-stage model, 435–437 See also Hierarchical regression Secular trends, migrant studies and, nutritional exposures, 582 Segregation analysis, genetic, 565 Segregation, racial/social, 539 Selection bias, 2, 134–137, 186, 193, 193f, 194, 362–363 See also Berksonian bias; Collider bias; Publication bias; Sampling error; Source population; Target population analysis of, 362–363, 375–376, 662–663 case-control studies, matching, 205, 205f complete populations, 482–483 confounding v., 136–137 graphical representation and analysis, 192–198 and publication bias, 654–655 Index quantitatively v qualitatively, 136 substudies and, 488 Selection time, study design and, 122 Self-administered questionnaires, 500 Self-reported disease, surveillance, 473 Self-selection bias, 134 Semen quality, 623–624 Semi-Bayes analysis, 337, 345, 378–380, 437, 439 See also Bayesian analysis; Hierarchical regression; Shrinkage estimation biases and, 378–380 Semilogarithmic plotting, 310, 311f Semiquantitative food-frequency questionnaires, 591 Sensitivity in bias analysis, 354–360 of diagnostic or screening test, 643–644 of exposure measurement method, 138 influence analysis and, 221–222 in probabilistic bias analysis, 372–375 relation of predictive values to, 357–358 surveillance systems, 479 Sensitivity analysis, 208, 355, 488 See also Bias analysis base priors and, 380 Bayesian analysis, 342, 378–380 external adjustment process and, 351, 351t meta-analysis, 661–662, 675 multiple bias analyses and, 363 unmeasured confounders, 349 Sentinel events, surveillance, 475 Sequelae prevention, clinical trials and, 90 Serial interval, state of infectiousness, 550f, 560 Serotyping, 553 Seroconversion, HIV, 557 Seventh Day Adventists studies, 582 Severe acute respiratory syndrome (SARS), infection occurrence, factors influencing, 556 Sex/gender, as social variable, 535 Sharp null hypothesis, 60 Shep’s relative difference, 65 Short-term recall, food intake, 589 Shrinkage estimation, 331, 337, 423, 436, 437, 439 See also Bayesian analysis; Hierarchical regression confounder control and, 263 Siblings records, secondary data, 484 Side effects of treatments See Adverse events Sign test, 680 Significance tests, 150, 151–153 confidence intervals v., 157–158 model searching and, 419 Simplifying search, model searching, 419 Simulation intervals, 365 Simultaneous analysis, 323–327 See also Multiple comparisons of exposure levels, 323–324 of multiple outcomes, 325–327 single-comparison v., 326–327 of trends, 326 Simultaneous misclassification, 145–146 Sine curve, cyclic patterns, 608 Single study group large-sample methods, 245–247, 247t person-time data, large-sample methods, 240–243 small-sample statistics, person time data and, 253–254 P1: TNL/OTB GRBT241-INDEX P2: IML/OTB QC: IML/OTB GRBT241-v4.cls T1: IML February 5, 2008 Printer: RRD 14:38 Index Single-regression models, 401 multiple-regression models v., 401–402 SIRs See Standardized incidence ratios Small-sample methods, 220, 239, 253–257, 276 See also Exact statistics; Sparse data person-time data, 253–255 count data, 255–257 Smoking 1, 2, 8–10, 13–14, 24, 26–30, 52, 55–56, 64–66, 117, 118, 119, 121, 122, 251, 263, 264, 266–267, 274, 280, 424, 483, 488, 608, 616 in pregnancy, 602, 624, 632, 634, 635 Smoothed curves, 319–321, 410–412 Smoothers and Smoothing, 313–14, 317–321, 410–412, 421, 438–442 See also Moving averages; Nonparametric regression; Power models; Spline models fractional polynomial, 410 graphical, 313–314 with hierarchical regression, 438–439 kernels, 314, 318–320, 440, 441 locally linear, 441 power model, 410 regression splines, 410–412 for surveillance data, 478 variable span, 321, 442 windows, 317–321 Smoothing parameter, 441–442 Smoothing splines, 440–441 See also Nonparametric regression; Spline models SMR See Standardized morbidity ratio SNFT See Structural nested failure-time models SNM models See Structural nested mean models Snow’s natural experiment, 94, 95 See also Cholera, Snow’s study of Social capital, 541–542 Social class, 537, 548 intercommunity comparisons, 611–612 Social epidemiology, 532–548 analytic approaches, 542–548 covariate assessment, 534–542 exposure, 534–542 Social factors, 534–542 aggregate level, 538–542 individual level, 534–542 Social mobility, 539, 547 Software for Bayesian analysis, 338 for hierarchical modeling, 436 problems in distinguishing models, 402 Source population, 114–115, 128–129, 383 See also Sampling error; Selection bias; Superpopulation; Target population confounding and, 129–133 subject identification, 496 target population v., 129, 383 Sparse data See also Small-sample methods bias from, 263, 386, 422, 433 methods for, 272–278, 280–281, 283–288, 294, 296, 314, 316, 381, 386, 425, 433–435, 446–451 Spatial-data analysis, time-trend designs v., 609–610 Special-exposure cohorts, 109–110 Specification accuracy, 449–450 Specification biases, 346 Specification error, 346 755 Specificity, as causal criterion, 27–28 Specificity in bias analysis, 354–360 of diagnostic or screening test, 643–644 of exposure measurement method, 138 in probabilistic bias analysis, 372–375 relation to predictive values, 357–358 Spline models, 410–412, 440–441 See also Nonparametric regression; Power models; Smoothers and Smoothing Spontaneous abortion, 621, 628 Square-root model, 398 Stability assumption in causal diagrams, 191 Standard, choice of, 266 Standard deviation (SD) See also Variance flaws in conventional estimates, 239, 346–347 flaws in rescaling by, 393, 681 Standard distribution, 49–50, 68–69, 265–266, 386–388 See also Standardization Standard error (SE) See also specific estimates, e.g., Risk ratio of externally adjusted estimates, meta-analysis, 661–662 impact of model searching on, 419 Standardization, 49–50, 67–69, 265–269, 386–388, 442–445 See also Inverse-probability weighting; Standard distribution; Standardized measures; Standardized morbidity ratio in case-control studies, 269, 445 by exposure models, 444–445 by full-data models, 444 by outcome models, 443 of regression, 386–388 Standardized coefficients, 681 See also Standard deviation, flaws in rescaling by Standardized incidence ratio (SIR), 68 See Standardized morbidity ratio Standardized measures, 33, 386–388 See also Standardization of association, 67–69 of effect, 67–69, 386–388 of outcome, 33, 387 probability, 442–445 rate, 49, 67, 265, 442 rate difference, 68, 266 rate ratio, 68, 267, 269 regression, 386–388, 442–445 risk, 50, 265 risk difference, 68, 266 risk ratio, 68, 267 regression coefficients v., 443–444 Standardized morbidity ratio (SMR), 68–69, 241–242, 267, 269, 664–665 Standardized mortality ratio, 68 See Standardized morbidity ratio State of infectiousness, 550f, 560 assessing risk of, 561–563 progression-related infectious process and, 550f, 560 Stationary population, 38–39, 47 Statistical approximations, accuracy of, 220, 225, 227–229, 243, 245, 247, 249, 250, 324, 334, 340, 342 Statistical biases, variable selection, 263, 419 Statistical efficiency, overmatching and, 176t, 179–180 Statistical estimation, 156–167 P1: TNL/OTB GRBT241-INDEX P2: IML/OTB QC: IML/OTB GRBT241-v4.cls T1: IML February 5, 2008 Printer: RRD 14:38 756 Statistical independence, 184–185 absence of open paths and, rules linking, 187–190 and d-separation rules, 189–190 Statistical inference, philosophies of See Bayesianism; Frequentist methods; Pure likelihood inference Statistical interaction, 72–74, 402–404 See also Effect-measure modification; Heterogeneity; Product terms and biologic interactions, 82–83, 298–300, 407 effect-measure modification v., 72 qualitative, 79–80 Statistical precision, 149 Statistical significance, 151 See also Hypothesis tests; P-values; Significance tests; inference and, 160f, 161 Statistical tests See Hypothesis tests; P-values; Significance tests Statistics, epidemiology and, 167 Steady state population, 38–39 Stein estimation, 331, 423, 439 See also Bayesian analysis; Hierarchical regression; Shrinkage estimation Stepwise model searching, 419, 420 Stereotype model, 415 Stillbirth mortality, 635–636 Stratification of data, and precision, 150 Stratified analysis, 258–302 problems with, 179 Stratified count data, polytomous exposure, 305, 306t Stratified logistic model, 433 Stratified models, 432–433 Stratified null hypothesis, P-values, 277–278 Stratified person-count data See Stratified count data, polytomous exposure Stratified person-time data, polytomous exposure, 305, 305t Stratified pure count data, P-values, 278 Stratum-specific estimates, 258, 264–265 measures, overall measures v., 62 models, 433 Strength associations, causal inference, 26–27 Strength of effects, 6f, 10–13, 14t Streptococcus pneumoniae, laboratory-based surveillance, 472–473 Stress, during pregnancy, 485 Structural equations, 60, 208, 453 Structural nested models, 453–455 See also G-estimation Structural nested failure-time (SNFT) models, 453–454 Structural nested mean (SNM) models, 453 Study See also Epidemiologic studies base, in case-control studies, 112 efficiency, 150 apportionment ratios and, 169–171 endpoints, 648–649 exposure, levels of, 440 hypothesis, person-time and, 102 populations, variations, 493–494 primary base, secondary base v., 114–115 size binomial likelihood, 247 formulas, 149–150 Index meta-analysis and, 679 random error and, 149–150 small, publication bias and, 679 variables, in meta-analysis, 655–656 Subclinical infection, clinical disease v., 551 Subclinical pregnancy loss, 630–631, 631f Subcutaneous fat, biochemical measurements, 594 Subinterval-specific incidence rates, survival proportion and, 45 Subject selection case-control studies, 120–121, 496–497 community trials, 494 and recruitment, epidemiologic studies, 493–499 run-in phase, randomized controlled trials, 647 Subject tracing, 109, 157 Subjective Bayesianism, 165–166, 328–333 See also Bayesianism; Subjective probability frequentism v., 166–167, 329–333 Subjective probability, 10, 23–24, 165–167, 330–331 Subjects, classification of, 218–219 Subject-selection bias, meta-analysis, 656–657 Subject-specific models, 544 See also Group-specific models Substudies, selection bias and, 488 Sufficient cause, 13 disease and, 6, 6f mechanisms, disease proportion and, 13–15 Sufficient component cause model, 6–18 biologic interactions under, 80–81 epidemiology and, 8–9, 8f indirect effects and, 13 potential-outcome v., 81–82, 81f scope of, 17–18 Sufficient conditioning sets, 192 Superpopulation, 151, 152, 383 for regression, 383 “Superspreaders,” infection occurrence, 556 Suppression bias, 678 Surgeon General’s Smoking and Health, 29 Surrogate confounders, 130 Surveillance, 459–480 approaches to, 472–476 history of, 460–462, 461t objectives, 462–467 Surveillance data analysis and interpretation, 476–478 assessing completeness, 478 education and policy, 467 presentation of, 478–479 Surveillance methods, combinations of, 475–476 Surveillance reports, surveillance data, 478 Surveillance systems, 459 attributes of, 479 cycles of, 470 elements of, 467–472 participation incentives, 470–471 SARS, 480 Surveys, 473 See also Cross-sectional studies; Prevalence studies Survival analysis, 41, 290–294, 291t See also Accelerated failure-time model; Cox model average incidence time v., 45 subinterval-specific incidence rates and, 45 Survival proportion, 40 P1: TNL/OTB GRBT241-INDEX P2: IML/OTB QC: IML/OTB GRBT241-v4.cls T1: IML February 5, 2008 Printer: RRD 14:38 Index Survival time See Incidence time Survivor bias, 198 Susceptibility measures, 64–65 Susceptibility factors, infection occurrence, 556 Syndromic surveillance systems, bioterrorism-related epidemics and, 464 Synergism, 76–82 See also Biologic interaction Synthetic meta-analysis, problem with, 654 Systematic errors (biases), 128–146, 345–380 random errors v., 346 Tabular analysis See stratified analysis Tabular checks, regression model checking and, 423–424 Tabular data, simultaneous statistics for, 323 Tabular presentation, surveillance data, 478–479 Target population, 60, 90, 146–147, 383 See also Source population; Sampling error; Selection bias confounding and, 60 generalizability of trial results, 90 matching, 172 source population v., 129, 383 surveillance systems, 471 TDT See Transmission disequilibrium test Temporal ambiguity, ecologic studies, 527 Temporality, causal inference and, 28 Test hypothesis, 152–154, 156, 165, 220 See also Null hypothesis Test of fit, 425–427 Test statistics, 152, 219–237 Test validity, screening tests, 642, 643, 643t See also Misclassification Test-based method for extraction of estimates, meta-analysis, 660 Tetanus, surveillance, 464 Therapy, studies of, 646–650 Time at risk, 102 time of exposure v., 102 Time clustering, 606–607, 607f Time unit, incidence rates and, 36 Time windows, induction analysis, 300–302 Time-dependent exposures, 300–302, 397–398, 452 See also Longitudinal data Time-dependent covariates, 397–398, 452 Time-related disease patterns, 606–608 Time-series analyses, 517, 610–611 Time-stratified referent periods, case-crossover studies, 605, 605f Time-to-pregnancy studies, 625–628 Time-trend designs ecologic studies, 515–517 spatial-data analysis v., 609–610 Time-varying covariates, 397–398, 452 See also Longitudinal data Tissue analysis, biochemical measurements, 594 Tobacco See Smoking Total energy intake, dietary intake and, 596–597 Townsend and Carstairs deprivation indices, 538 Toxic shock syndrome, surveillance, 465, 466f Tracing, epidemiologic studies, 109, 507 Transformed regressors, 398–400 Transformed-outcome models, 400 generalized linear models v., 416 Transmission axis, 549, 550f Transmission disequilibrium test (TDT), 638 757 Transmission, factors influencing, 561 Transmission-related states, infection and, 551, 552 Transmission-related studies, epidemiologic issues in, 560–563 Trapezoidal distribution, 371 Treatment assignment, clinical trials and, 90 Trend See also Dose response analysis, 314–317 dose-response and, 308–314, 308t graphing, 308–14 modeling, product terms and, 404–405 P-value, 314–316 in surveillance data, 477–478, 477f Trend models, 398–399, 408–413 category codes in, 409 hierarchical, 438 trend variation models, 412–413 Trend statistics, 314–316 Triple-blind studies, 91 Tumor heterogeneity, 577 Tumor promoter, 16 Twin studies, 15, 568–569 Two-binomial model large-sample methods, 247–250 Wald confidence limits, 250 Two-by-two tables, 247, 247t, 345 See also Contingency tables, data analysis and Two-phase studies See Two-stage studies Two-Poisson model, 245 Two-sided confidence interval, 158, 221 Two-sided P-value, 156, 221 Two-stage studies, 127, 496, 604 analysis of, 281–282, 432 Two-tailed P-value, 152 Type I error (alpha error), 153, 154 DSMB and, 91 gene-disease association, 571–572 Type II error (beta error), 153, 154, 155f Typhoid fever, evolution of, 553, 554f Uncertainty, 329, 331, 346–347, 364–365, 370, 376, 378, 379, 380 Uncertainty analysis, 346–347 See also Bayesian analysis; Bias analysis; Uncertainty Unconditional analysis, conditional analysis v., 272–273 Unconditional d-separation, 187–188 See also d-Separation Unconditionally unbiased, in causal diagrams, 191 Unconditional probability, 184 See also Probability Uncontrolled confounders, analysis of, 348–352, 365–371 Unconfounded dependence, in causal diagrams, 193 Unconfounded direct effect, diagram, 201f Unemployment, 537 Unexposed group, bias and, in trend evaluation, 317 Unexposed time in exposed subjects, cohort studies, 104 Unhealthy worker effect, 622 See also Healthy worker effect Uniformity of effect measures, 61, 259 See also Homogeneity Uniform prior distributions, 365, 369 Univariate analysis See Descriptive statistics Univariate exposure transforms, 398–399 Unmatched case-control studies, modeling in, 429–432 P1: TNL/OTB GRBT241-INDEX P2: IML/OTB QC: IML/OTB GRBT241-v4.cls T1: IML February 5, 2008 Printer: RRD 14:38 758 Upper P-value, 220 Upper-tailed P-value, 152, 220 U.S Surgeon General’s Smoking and Health, 29 UV-B radiation, cancer and, 617 Vaccines and autism, secondary data, 485–486 Vacuous models, 390–391 Validation data, 361 dietary assessment methods, 591–592 secondary data, 487 Validity See also Bias; Generalizability; Precision; specific validity topics, e.g., Confounding; Selection bias; Sparse-data bias of biomarkers, 566 external, 128 in epidemiologic studies, 128–147 of estimation, 128–129 ethical considerations v., 89 internal, 128 secondary data and, 483 Variables See also specific types, e.g., Regressors in causal graphs, 186–187 in regression and causation, 383 Variable selection, 261–263, 419–421 Variable-span smoothers, 321, 442 See also Nonparametric regression; Smoothers and Smoothing Variance See also specific estimates, e.g., Standardized estimates bootstrap estimation, 429, 443 and confounder scores, 447 estimates, 149, 225, 226, 231, 243, 246, 248, 253, 254, 256, 422, 440, 443 model, 436 residual, 422, 425 Variance component models See Hierarchical regression Vectors, 384–385 Vertical scaling of graphs, 310–311 Vertices, in causal diagrams, 187 Virulence factors, 558 Volunteer providers, surveillance networks, 473 Vote counting, in meta-analysis, 680–681 Index Waiting time, 39 See also Incidence time Wald limits, 227 for binomial data, 246 for case-cohort data, 252–253 for case-control data, 250–252 for person-time data, 240–245 for pure count data, 245–250 Wald statistic, 220, 226–227, 279 Wald test, 227 Weighted averaging for Bayesian analysis, 334–337 by information, 334–337 for pooling across strata, 271 for smoothing, 318–319, 441 for meta-analysis, 670–671 by inverse probability of selection See inverse-probability weighting by inverse variance, 271, 334–337, 670–671 by precision, 271, 334–337, 670–671 by quality scores, 681 in standardization, 49–50, 67–69, 265–269, 386–388 Weighted regression, 441 See also Inverse-probability weighting in meta-analysis, 663, 668 Weights in meta-analysis, 670–671, 681 in standardization, 49–50, 67–69, 265–267, 386–388 Weight-specific mortality, 633–635, 634f WHO See World Health Organization Windows, See also Smoothers and Smoothing in smoothing, 317–321 Within-community time-series analysis, air pollution, 616 Within-group bias, ecologic studies, 520 Within-group misclassification, ecologic studies, 526–527, 526f Withdrawals See Loss to follow-up; Censoring Woolf method for pooling, 271 See also Information-weighted averaging World Health Organization (WHO) surveillance, 466, 474 Zero level, in trend evaluation, 316–317 Z-ratio, 220 See Wald statistic Z-score See Wald statistic; Z-ratio ...P1: TNL GRBT241 -22 Printer: RRD GRBT241-v4.cls January 18, 20 08 12: 5 458 P1: TNL GRBT241 -22 Printer: RRD GRBT241-v4.cls January 18, 20 08 12: 5 CHAPTER 22 Surveillance James W Buehler... often P1: TNL GRBT241 -22 Printer: RRD GRBT241-v4.cls January 18, 20 08 Chapter 22 ● 12: 5 Surveillance 465 NYC Age-specific rate per 100,000 population >26 8.3–369.7 >20 5.4 26 8.3 >33.1 20 5.4 >1.5– 33.1... 1988–99, and 1999 20 00 seasons In Surveillance Summaries, October 25 , 20 02 Morb Mortal Wkly Rep 20 02: 51(No SS-7):6.) P1: TNL GRBT241 -22 Printer: RRD GRBT241-v4.cls January 18, 20 08 12: 5 478 Section