HLTB16 Notes Pt 2 PDF

OCT 2 2024 EPIDEMIOLOGY; THE BASIC SCIENCE OF PUBLIC HEALTH Keywords/Formulas Main points/Thoughts DEFINITIONS OF EPIDEMIOLOGY; diagnostic discipline of public health EPIDEMIOLOGY - Charles-edward Amory Winslow - Charles-Edward...

OCT 2 2024 EPIDEMIOLOGY; THE BASIC SCIENCE OF PUBLIC HEALTH Keywords/Formulas Main points/Thoughts DEFINITIONS OF EPIDEMIOLOGY; diagnostic discipline of public health EPIDEMIOLOGY - Charles-edward Amory Winslow - Charles-Edward Amory - Cause of disease Winslow - Goal is discover causality - John Last - Get and quantify data to argue about it - Identify trends in disease occurrence EPIDEMIOLOGY; Cause of that may influence the need for medical disease and public health services - Goal is discover causality - Evaluate effectiveness of medical and public health interventions USES OF EPIDEMIOLOGY; - Etiology; Look at cause EPIDEMIOLOGY; study of the distribution and of disease determinants of health-related states or events in - Disease spread; specified populations, and the application of this study Characteristics of risk to (prevent) and control of health problems factors - John Last - Disease burden; Disparities USES OF EPIDEMIOLOGY; - Health policy, planning, - Etiology and services; Policies - Disease spread and programs - Disease burden - Health policy, planning, and services 3 ESSENTIAL CHARACTERISTICS OF ETIOLOGY EPIDEMIOLOGY - Independent factor = exposure Person (who?) - Dependent factor = outcome Place (where?) - Studying cause(s) of disease or condition Time (when?) - Is something a risk factor? See the exposure and outcome of the risk factor or agent, or 2 BROAD TYPES OF ascertaining causative factors EPIDEMIOLOGY - Analysing contributing factors 1. Descriptive 2. Analytic DISEASE SPREAD - Need to know characteristics of DESCRIPTIVE agent/causative factor EPIDEMIOLOGY; examining the - Mode of disease transmission distribution of a disease in a - Identify and see spatial and geographic population, and observing the patterns (space and time is important) basic features of its distribution in terms of time, place, and DISEASE BURDEN person - Identifying and analyzing disparities - Deals with outcome - Social - Spatial ANALYTIC EPIDEMIOLOGY; - Geographic testing a specific hypothesis - Reporting on morbidity, disability, injury, about the relationship of a mortality disease to a putative cause, by conducting an epidemiologic HEALTH POLICY, PLANNING, AND SERVICES study that relates the exposure - Aid the planning and development of health of interest to the disease of the services and programs interest - Provide administration and planning data  - Relates outcome to risk - Provide foundation for public health measures factor and policies CHRONIC VS. INFECTIOUS ETIOLOGY; look at cause of disease (beginning of DISEASES something) CHRONIC DISEASE; EPIDEMICS OCCUR WHEN - Difficult to study since its - Host, agent, and env. Factors are not in hard to determine its balance frequency and length of - Due to new agent its time to be present in - Due to change in existing the host - Agent (infectivity, pathogenicity, virulence) - Preventative strategies - Due to change in number of susceptible in the help delay effects to see population - Need to know etiology - Due to environmental changes that affect and factors to know - Transmission of the agent of growth of the disease progression agent - (basing it off of experiences and 3 ESSENTIAL CHARACTERISTICS OF exposures the person EPIDEMIOLOGY has in their life that 1. Person (who?) makes them have the 2. Place (where?) chronic disease) → don’t 3. Time (when?) know when the disease could manifest in a PERSON person’s life despite the - Age, gender, SAB, ethnic group timeframe of exposure - Genetic predisposition (dormancy or immediacy) - Concurrent disease - Diet, physical activity, smoking INFECTIOUS DISEASE; - Risk taking behaviour Transmission of specific - SES, education, occupation pathogenic agent to susceptible host via PLACE 1. Directly from other - Presence of agents/vectors infected humans - Geography 2. Indirectly, through - Climate vectors, airborne - Population density particles or vehicles - Economic development - Lack of this allows for more disease EPIDEMIOLOGICAL spread (lack of restraining PRINCIPLES AND METHODS strategies/resources to fund for - Defining disease prevention) - Understanding - Nutritional practices exposures - Medical practices - Health determinants TIME - Risk factors - Calendar time - Disease frequency - Time since an event - Incidence - Physiologic cycles - Prevalence - Age (time since birth) - Seasonality HEALTH DETERMINANTS; - Temporal trends underlying factors or social and - Ex; influenza going up in winter physical conditions that impact health and disease 2 BROAD TYPES OF EPIDEMIOLOGY - See the relation of 1. Descriptive exposure and outcome 2. Analytic RISK FACTORS; lifestyle or DESCRIPTIVE EPIDEMIOLOGY; examining the environmental factors that distribution of a disease in a population, and observing increase risk of disease the basic features of its distribution in terms of time, - Usually modifiable place, and person - Deals with outcome DISEASE FREQUENCY; - Typical study design: community health survey necessary not only to count (cross-sectional study, descriptive study) number of cases, but relate that number to the size of population ANALYTIC EPIDEMIOLOGY; testing a specific being studied hypothesis about the relationship of a disease to a - Frequency over time and putative cause, by conducting an epidemiologic study space that relates the exposure of interest to the disease of - Useful to count number the interest of people, cases, and the - Relates outcome to risk factor population density - Typical study designs: cohort, case-control - Size of population is - Ex; is cigarette smoking contributing to lung important to know cancer? compared to number of cases (see the relevancy) → John Snow - Epidemiology and john snow INCIDENCE; Rate of new cases - Focus on definitions from class of disease in defined population over defined period of time CHRONIC VS. COMMUNICABLE` (INFECTIOUS) - Change in disease status DISEASE - Healthy → unhealthy - Risk affects incidence ⇒ is the probability that defines the person’s CHRONIC DISEASE; change in health status - Difficult to study since its hard to determine its - How much factors affect frequency and length of its time to be present in people to go from the host healthy→ unhealthy - Epidemiology has had a different role to play in - Only new cases** investigating the causes of the diseases common in older age PREVALENCE; Part of - Cancer, heart disease are different from population who have a specific infectious diseases or acute poisoning characteristic in given time - Thought of part as ageing until period 20th-century found a concern to prevent - Risk does not affect them prevalence - Cancer, heart disease, and other diseases of - Not looking at new ageing do not have single causes cases, but see who HAS - Tend to develop over a period of time, it (including everyone are often chronic and disabling rather overall) than rapidly fatal, and cannot be prevented or cured by any vaccine or EPIDEMIOLOGY PYRAMID ‘magic bullet’ 1. Randomized control - PREVENTION STRATEGIES; learning how to trials > best prevent them or delay them 2. cohort studies > - Healthy eating campaigns 3. case-control studies - Screening campaigns  (give lowest amount of - Prevention → needs understanding of evidence) etiology/cause of disease and factors that influence how it progresses PLACEBO; an inactive - Need data to understand its effects substance similar in appearance - Studying chronic disease to the drug or vaccine being - Epidemiologic studies of these chronic tested diseases are much more complicated - Treatment group; new and difficult than investigations of acute drug to try outbreaks of infectious diseases toxic - Control group; placebo to contamination try that gives same - Many ‘Risk factors’ to cause chronic results from another diseases certified drug/vaccine - Long period over which these diseases - Purpose; new treatment develop also contributes to the difficulty is different from placebo of determining the causative factors - Meaning- does - Epidemiologists must determine which my drug reduce of a person’s many experiences over the incidence of the previous decades are relevant, and illness compared what significant exposures might have to placebo occurred 10 or 20 years ago → potential increase in person’s risk of RANDOMIZATION AND developing the disease today BLINDING - RANDOMISED; controls - Heart disease bias and equalises known-unknown factors that might affect results - DOUBLE-BLIND; neither patient and doctor knows treatment is administered - Prevent the possibility that doctors might interpret the patient’s condition differently if they know how the patient is being treated - Highest form of evidence to get - Randomized trials are expensive but gives more results due to whatever results you get back COHORT STUDIES - All healthy participants - Original cohort study (important to study (at the beginning) to know) - See if they develop a disease (of interest) over - Smoking cessation time - Purpose; want to see a change in health status  over time - Resource intensive ⇒ Takes a lot of time and resources to follow them - Can measure incidence and quantify risk CASE-CONTROL STUDIES - Looking at who has the disease/case, ascertaining exposure status - Case–control - Cases = unhealthy people - Example of display data over time (have disease) - Temporal example - Controls = health - Cause of lung cancer is observed when people the hypothesis that patients are also - Having similar smokers (coincidence? I think not!!) characteristics but healthy-unhealthy difference INFECTIOUS DISEASE; Transmission of specific TYPES OF STUDIES → TYPES pathogenic agent to susceptible host OF BIASES - Synonymous - Average results with - Infectious agents may be transmitted to observatory studies (lack humans either; of knowing timeframe of 1. Directly from other infected humans diseases) 2. Indirectly, through vectors, airborne - Experimental studies are particles or vehicles high levels of evidence - Lots of randomisation = lots of errors = lack of biases EPIDEMIOLOGICAL PRINCIPLES AND METHODS - Defining disease RANDOM ERROR; value of the - Understanding exposures sample measurement diverges - Health determinants due to chance alone, from that of - Risk factors the true population value - Disease frequency - Causes inaccurate - Incidence measures of association - Prevalence - Never be completely - Methods: Types of epidemiological studies eliminated since we can study only a sample of DEFINING DISEASE the population - No way to define absolute health so need to know whether a certain individual not included 3 major sources of random error in the health standard should be considered as 1. Individual biological healthy or not healthy variation - Need to define a disease to see if people fit into 2. Sampling error that definition (from healthy to unhealthy) 3. Measurement error - Some diseases are easier to identify than others INDIVIDUAL VARIATION; - Based on resources to be able to define always occurs and no this measurement is perfectly - Availability/quality of data are accurate considerations - Registry of how much individuals have SAMPLING ERROR; is usually records of this disease and their method caused by the fact that a small of being healed from it sample is not representative of - Epidemiologists use the term ‘disease’ broadly all the population’s variables - ‘Health outcome’ is more accurate but difficult - Best way to reduce description of what is to be studied sampling error is to - For example; Epidemiologists might increase the side of the study the frequency and distribution of study high blood cholesterol, which is not a disease but it is related to the risk of a MEASUREMENT ERROR; can heart attack, or they might study injuries be reduced by stringent due to traffic accidents, which are not protocols, and making individual diseases but are certainly significant to measurements as precise as health possible - Epidemiological studies may point to ways of preventing the negative health outcome SYSTEMATIC ERROR (BIAS); occurs in epidemiology when ‘EXPOSURES’ results differ in systematic 1. Health determinants manner from the true values 2. Risk factors - Study with small systematic error said to HEALTH DETERMINANTS; underlying factors or have high accuracy social and physical conditions that impact health and disease BIAS; ‘any systematic error in - See the relation of exposure and outcome the design, conduct or analysis - Social, economic, cultural, environmental of a study that results in a (outside health sector) mistaken estimate of an expsoure’s effect on the risk of RISK FACTORS; lifestyle or environmental factors that disease’ increase risk of disease - Usually modifiable 3 PRINCIPLE BIASES OF - Examples; improve access to greenspace, SYSTEMATIC ERRORS; care, resources, healthy food options 1. Selection bias 2. measurement/information DISEASE FREQUENCY; necessary not only to count bias number of cases, but relate that number to the size of 3. Confounding bias population being studied - How are diseases distributed? SELECTION BIAS; occurs when - Frequency over time and space there is a systematic difference - Useful to count number of people, between the characteristics of cases, and the population density the people selected for a study - Size of population is important to know and the characteristics of those compared to number of cases (see the who are not relevancy) - When participants select themselves for a study INCIDENCE; Rate of new cases of disease in defined due to being unwell or population over defined period of time because they are - Change in disease status / new cases; healthy particularly worried about → unhealthy an exposure = selection - Risk affects incidence ⇒ is the probability that bias defines the person’s change in health status - How much factors affect people to go MEASUREMENT BIAS; can from healthy→ unhealthy occur when the individual - Only new cases** measurements or classifications of disease or exposure are PREVALENCE; Part of population who have a specific inaccurate – stuff from labs not characteristic in given time period measured correctly of what they - Risk does not affect prevalence are supposed to measure - Not looking at new cases, but see who - Any data coming from lab HAS it (including everyone overall) is wrong due to poor - More broad; incidence and prevalence related calibration of tools to each other - Relationship depends on how long INFORMATION BIAS; Lots of people live with the disease interpretation from how the data - Disease with high incidence could have low should be read from whoever prevalence if people recover from it rapidly, or if looks at it (lack of objective bias) they die from it in a short period of time - Quality and extent of - However for chronic diseases that are not information obtained is lethal (arthritis) the prevalence will be much different for exposed higher than incidence persons than for - Most useful in assessing the societal impact of non-exposed persons → a disease and planning for healthcare services significant bias can be introduced WHO, WHEN, WHERE? - Likely to occur in historical cohort WHO? (person) studies where - Rate of new cases that quantifies disease information is spread gotten from past - Characterises disease victims by such factors records - (age, sex, race, and economic status) - As discussed in - For example, incidences of cancer and heart connection with disease are greater in older people; measles randomized trials, in any and chicken pox occur more often in the young cohort studies, it is - Older women and younger men are more likely essential that the quality to suffer broken bones than old men and young of the information women obtained be comparable in both exposed and WHEN? (time) non-exposed individuals - Disease changing over time; incidence - Frequency; increasing, decreasing, stability? OBSERVER BIAS; if the - Incidence of lung cancer in american investigator, lab. Technician or men, for example, increased steadily participant, knows the exposure from the 1930s–1990 (peaking and status → knowledge can beginning to decreasing) influence measurements and - Incidence of respiratory infections yielded results always higher in the winter - To avoid this bias, - Posing another kind of when question → measurement can be seasonal variations in incidence made in a blind or double-blind fashion WHERE? (place) - Blind study = - Looks at comparisons of disease frequency in investigators do not know different places how participants are - May also look at comparisons between urban classified and rural populations - Hypothesis that fluoride protects against tooth CONFOUNDING BIAS; decay arose from the observation that dental distortion of the association cavities were less common in children who between an exposure and health lived in parts of the country that had high outcome by an extraneous, third concentrations of fluoride in water variable called a confounder - Statistics on causes of death in different - Confounding is an countries can be very suggestive in generating important concept in hypotheses about the causes of disease epidemiology due to causing an over- or HOW TO STUDY EPIDEMIOLOGY? under-estimate of the (EPIDEMIOLOGY PYRAMID) observed association - Perform a study between exposure and - Randomized control trials > health outcome - cohort studies > - Having (confounding - case-control studies (give lowest amount of variables) other evidence) variables influencing - Need to know for test results that yields biassed results HILL’S 9 CRITERIA TO SATISFY CAUSALITY 1. Strength of association (effect size) 2. Consistency of data (reproducibility) 3. Specificity 4. Temporality 5. Dose-response (biological gradient) 6. Biological plausibility 7. Coherence 8. Experimental evidence 9. Analogy INTERVENTION STUDIES - Randomized control studies are important to know Questions - Studies are conducted in very much same way as those of laboratory experiments on animals Index: - They are usually done to test a new treatment for a disease such as a chemotherapy drug for ** everything in neon blue is the cancer, or preventive measure (vaccines) main summarization the prof - In clinical trials, one group is exposed to the said for each headline intervention, while a control group is not exposed ** everything in light blue is kind - Investigators then watch and wait to see of important but just as context? whether the response of the treatment group is For the light blue (still good to different from that of the control group skim though) TERMINOLOGY OF CLINICAL TRIALS ** everything in red will be on the - PLACEBO; an inactive substance similar in exam appearance to the drug or vaccine being tested - Treatment for a disease already known ** everything in yellow highlights to exist? Trials may compare the new might be on exam treatment with the existing treatment - Purpose of the placebo; prevent ** everything in black is basically subjects from knowing whether they are copied from the slides so choose receiving the intervention whichever you want to focus on - Many trials over the years have found for extra reading/understanding that up to a third of patients respond to the placebo as if it were the intervention, reporting that they feel better or that they suffered side effects - Drug being tested must show the high response rate than the placebo if it is to be considered effective - Treatment group; new drug to try - Control group; placebo to try that gives same results from another certified drug/vaccine - Purpose; new treatment is different from placebo - Meaning- does my drug reduce the incidence of illness compared to placebo RANDOMIZATION AND BLINDING - RANDOMISED; controls bias and equalises known-unknown factors that might affect results - DOUBLE-BLIND; neither patient and doctor knows treatment is administered - Prevent the possibility that doctors might interpret the patient’s condition differently if they know how the patient is being treated - Highest form of evidence to get - Randomized trials are expensive but gives more results due to whatever results you get back - Hard to do a randomized control trial due to ethical reasons - But still want to know incidence changes due to exposure - Food and Drug Administration (FDA) requires that the safety and effectiveness of new drug must be demonstrated in a properly conducted clinical trial before it can be approved for marketing COHORT STUDIES - Ideal but expensive (and takes a lot of time) - All healthy participants (at the beginning) - See if they develop a disease (of interest) over time - Purpose; want to see a change in health status over time - Can measure incidence and quantify risk - Resource intensive ⇒ Takes a lot of time and resources to follow them - Observational studies  - Since such experiments are not possible for most hypotheses that epidemiologists want to test, methods have been devised by which investigators can link exposures to results by observation alone, without actively intervening in the lives of the study of subjects - (so basically let the subjects live on in their daily lives with having exposure to the disease, but it will take a lot of time to observe the exposure rate and activity of the disease in the group of people first introduced to the disease) - Observe over period of time to see whether those who were exposed to the factor being studied are more likely to develop the disease than those who were not - Example; Measure if they are smokers or non-smokers → Over time, want to measure new cases of getting disease over time CASE-CONTROL STUDIES - Observational studies - Looking at who has the case, ascertaining exposure status - Matching a case to control - Cases = unhealthy people (have disease) - Controls = health people - Having similar characteristics but healthy-unhealthy difference - Start out by measuring exposure and watching for the development of disease, case-control studies start with people who are already ill and look back to determine their exposure - More efficient in terms of focusing on smaller number of people and can be completed quickly - Investigator asks all participants same questions concerning extent of their exposure to the factors hypothesised to have caused the disease - Pros; see how many people are affected - Cons; can’t see the timeframe the exposure started (get an accurate reading when the cases first got the disease, unspecified temporality measurement) PROBLEMS AND LIMITS OF EPIDEMIOLOGY - Clinical trials are randomised (help determine causality) → gives high evidence - If cannot clinical, then cohort (see rates of incidence) - If no resource/time, then case-control study  (start with a group of cases, match to controls, and record the exposures) - Cannot estimate risk with a case-control study** - Quantity you get from case-control study; odds-ratio - Case-control give similar quantities to risk but not giving risk since not starting with healthy people - First; observe possible association between an exposure and an illness - Second; developing a hypothesis about a cause and effect relationship - Third; testing hypothesis through formal epidemiologic study - Experimental studies are high levels of evidence - Lots of randomisation = lots of errors = lack of biases RANDOM ERROR; value of the sample measurement diverges due to chance alone, from that of the true population value - Causes inaccurate measures of association - Never be completely eliminated since we can study only a sample of the population - 3 major sources of random error 1. Individual biological variation 2. Sampling error 3. Measurement error - Individual variation; always occurs and no measurement is perfectly accurate - Sampling error; is usually caused by the fact that a small sample is not representative of all the population’s variables - Best way to reduce sampling error is to increase the side of the study - Measurement error; can be reduced by stringent protocols, and making individual measurements as precise as possible - Investigators need to understand the measurement methods being used in the study, and the errors that these methods can cause - Ideally, labs should be able to document the accuracy and precision of their measurements by systematic quality control procedures SYSTEMATIC ERROR (BIAS); occurs in epidemiology when results differ in systematic manner from the true values - Study with small systematic error said to have high accuracy - Accuracy not affected by sample size - BIAS; ‘any systematic error in the design,  conduct or analysis of a study that results in a mistaken estimate of an expsoure’s effect on the risk of disease’ - Over 30 types of bias that have been identified 3 PRINCIPLE BIASES OF SYSTEMATIC ERRORS; 1. Selection bias 2. measurement/information bias 3. Confounding bias SELECTION BIAS; occurs when there is a systematic difference between the characteristics of the people selected for a study and the characteristics of those who are not - When participants select themselves for a study due to being unwell or because they are particularly worried about an exposure = selection bias - Example; well known that people who respond to an invitation to participate in a study on the effects of smoking differ in their smoking habits from non-responders, the latter are usually heavier smokers - In studies of children’s health, parental cooperation is required (people can choose to enter their child or not in a study) MEASUREMENT BIAS; can occur when the individual measurements or classifications of disease or exposure are inaccurate – stuff from labs not measured correctly of what they are supposed to measure - Any data coming from lab is wrong due to poor calibration of tools - There are many sources of measurement bias and their effects are of varying importance - For instance, biochemical or physiological measurements are never completely accurate and different laboratories often produce different results on the same specimen - If specimens from the exposed and control groups are analysed randomly for different laboratories, there is less chance for systematic measurement bias than in the situation where all specimens from the exposed group are analysed in one laboratory and all those from the control group are analysed in another - INFORMATION BIAS; Lots of interpretation from how the data should be read from whoever looks at it (lack of objective bias) - Quality and extent of information obtained is different for exposed persons than for non-exposed persons → significant bias can be  introduced - Likely to occur in historical cohort studies where information is gotten from past records - As discussed in connection with randomized trials, in any cohort studies, it is essential that the quality of the information obtained be comparable in both exposed and non-exposed individuals OBSERVER BIAS; if the investigator, lab. Technician or participant, knows the exposure status → knowledge can influence measurements and yielded results - To avoid this bias, measurement can be made in a blind or double-blind fashion - Blind study = investigators do not know how participants are classified - Blinding is a kind of observer bias - Blinding physicians to study - physicians might know beforehand how a drug works, and if they knew the patient is receiving a drug → affects their idea of their administration and treatment plan for the patient ⇒ not giving raw results CONFOUNDING BIAS; distortion of the association between an exposure and health outcome by an extraneous, third variable called a confounder - Confounding is an important concept in epidemiology due to causing an over- or under-estimate of the observed association between exposure and health outcome - Having (confounding variables) other variables influencing results that yields biassed results - Distortion introduced by a confounding factor can be large - Can change the apparent direction of an effect - Problem but accounted during analysis - How to consider confounding variable when checking epidemiology; 1. Exposure 2. Outcome 3. Confounder = affects both exposure and outcome and extorts relationship (extorted results) CONFOUNDING VARIABLE (IMAGE)  - Randomized experiments are typically preferred over observational studies or experimental studies that lack randomization because they allow for more control - Common problem; non-randomized studies may have other variables influencing results - = confounding variables - Elated to both the explanatory variable and response variable - Characteristic that varies between cases and is related to both the explanatory and response variables → lurking variable or third variable WHEN IS A 3RD VARIABLE A CONFOUNDER? - Factor A → disease B? Potential factor X (confounder) if the following are true; 1. Factor X is a known risk factor for disease B 2. Factor X is associated with factor A, but is not a result of factor A EXAMPLE; ICE CREAM CONSUMPTION CAUSE HOME INVASIONS? CONFOUNDING VARIABLE TO ICE CREAM-HOME INVASIONS COFFEE AND CHD RELATIONSHIP CONFOUNDING VARIABLE TO COFFEE-CHD CAUSALITY ESTABLISHING CRITERIA? ⇒ Hill’s criteria background HILL’S CRITERIA BACKGROUND - 1965; Austin Bradford Hill made the 9 factors that constitute the current standard for determining causality - In his article, Hill expanded upon criteria that had previously been set forth in the report Smoking and Health (1964) by the United States Surgeon General - It is important to note that satisfying these criteria may lend support for causality, but failing to meet some criteria does not necessarily provide evidence against causality,  either - Hill’s causal criteria should be viewed as guidelines, not as a “checklist” that must be satisfied for a causal relationship to exist. HILL’S 9 CRITERIA TO SATISFY CAUSALITY 1. Strength of association (effect size) 2. Consistency of data (reproducibility) 3. Specificity 4. Temporality 5. Dose-response (biological gradient) 6. Biological plausibility 7. Coherence 8. Experimental evidence 9. Analogy STRENGTH OF ASSOCIATION - Cause-effect relationship - Don’t assume that a strong association alone is indicative of causality, as the presence of strong confounding may erroneously lead to a strong causal association - Measured via risk ratios, rate ratios, or odds ratios. - Hill believed that causal relationships were more likely to demonstrate strong associations than were non-causal agents. - Smoking and lung cancer is a perfect example where risk ratios, rate ratios, and odds ratios are in the 20 to 40 range when comparing smokers to non-smokers. - However, weak associations as demonstrated by the risk ratio, rate ratio, or odds ratio should not be taken as an indication of non-causality. - This is particularly true when the outcome of interest is common - Example; A common outcome that exhibits a weak association to smoking is cardiovascular disease (CVD). Yet even with a weak association, evidence supports the casual nature between smoking and the development of CVD CONSISTENCY OF DATA - Greater consistency = greater causality - Did the case-control study that might not be generalisable/applicable to other countries → if its repeatable in another population, there is a relationship there in being able to reproduce similar results - Reproducibility of experiment/collected data and results across countries is possible = greater consistency  - Lack of consistency does not mean causal association as some causal agents are causal only in the presence of other cofactors SPECIFICITY - Associations are more likely to be causal when they are specific - Exposures → ONLY ONE disease - Criterion has been proven to be invalid in a number of instances - Example; smoking - Evidence clearly demonstrates that smoking does not lead solely to lung carcinogenesis but to a myriad of other clinical disorders ranging from emphysema to heart disease - On other hand, there are certain situations where a 1-to-1 relationship exists (certain pathogens are necessary to produce a specific disease) TEMPORALITY - Everyone agrees this needs to be fulfilled - Exposure must be there/fulfilled before outcome - Lack of temporality rules out causality - Example; atrial fibrillation and pulmonary embolism - Current wisdom supports that pulmonary embolism causes atrial fibrillation, recent evidence and plausible biological hypothesis suggests that the reverse could be true - Determining the proper course of care may hinge upon discovering if pulmonary emboli can indeed precede and thus perhaps cause the development of atrial fibrillation DOSE-RESPONSE - ‘If a dose response is seen, it is more likely that the association is causal’ - Presence of dose-response relationship between an exposure and outcome provides good evidence for a causal relationship - Absence should not be taken as evidence against such relationship - Some diseases or health outcomes do not display a dose-response relationship with a causal exposure - May demonstrate a threshold association where a given level of exposure is required for disease or health outcome initiation, and any additional exposure does not affect the  outcome BIOLOGICAL PLAUSIBILITY - Basic science lab results to prove causality - Not unusual for epidemiological conclusions to be reached in the absence of evidence from the lab, particularly in situations where the epidemiological results are the first evidence of a relationship between exposure and outcome - However, further support for causal relationship with the addition of a reasonable biological mode of action, even though basic science data may not yet be available COHERENCE - Causal association needs data that should not oppose the current evidence - However, one should be cautious in making definite conclusions regarding causation, since it is possible that conflicting information is incorrect or highly biassed EXPERIMENTAL EVIDENCE - Experimental design supports causality due to experimenter altering cause and effect - See the difference in risk to chosen exposed participants, determining the exposure causes the outcome - Experimental evidence results from many areas; the lab, epidemiological studies, preventive-clinical trials - Ideally, epidemiologists would like experimental evidence obtained from a well-controlled study - randomized trials - Types of studies that can support causality by demonstrating that ‘altering the cause alters the effect’ ANALOGY - Speculative in nature and dependent on the subject opinion of the researcher - Example; infection may cause fever, not all fevers are due to infection - Absence of analogies should not be taken as evidence against causation OTHER CONSIDERATIONS TO CAUSALITY - Critically important to have a thorough understanding of the literature to determine if any other plausible explanations have been considered and tested previously Summary  - Epidemiology - Uses of epidemiology - Etiology - Disease spread - Disease burden - Health policies - 2 types of broad epidemiology - Descriptive - Analytic - Chronic vs. infectious disease and their etiologies - Defining disease - Understanding exposures - Health determinants - Risk factors - Disease frequency - Incidence - Prevalence - Methods: Types of epidemiological studies - Randomized control study - Cohort study - Case-control study - Random error - 3 major sources of random error 1. Individual biological variation 2. Sampling error 3. Measurement error - Bias / systematic error - 3 principle biases 1. Selection bias 2. measurement/information bias 3. Confounding bias - Confounding variable - Hill’s 9 criteria test of causality OCT 9 2024 SURVEILLANCE AND THE ROLE OF DATA IN PUBLIC HEALTH Keywords/Formulas Main points/Thoughts HEALTH STATISTICS; BACKGROUND monitoring and analyzing health - Need to know what's going on in community to of community by collecting quantify what evidence you need to support the health data health policy you want to establish - Identify risk groups, identify health SURVEILLANCE; systematic threats, create policies to prevent health collection, analysis, and timely endangerment dissemination of information on - Health statistics; monitoring and analyzing population health to those who health of community by collecting health data need to know so action can be - Vital part of public health’s assessment taken function - Provides information - Used to identify special risk groups about patterns of health, - Detect new threats disease, changes in - Plan public health programs these patterns, - Evaluate their success prevention methods - Prepare government budgets  based on patterns - Statistics collected by federal, state, and local government are the raw 2 TYPES OF SURVEILLANCE material for research on epidemiology, - Passive env. Health, social and behavioural - Active factors in health, and for medical care system ROLE OF CLINICIANS IN SURVEILLANCE SURVEILLANCE; systematic collection, analysis, and - The primary care timely dissemination of information on population physician must inform the health to those who need to know so action can be public health service by taken reporting occurrences of - Provides information about patterns of health, certain contagious disease, changes in these patterns, prevention diseases that they see methods based on patterns - First line of surveillance - Guide prevention and control efforts as well as that prevent potential contribute to planning health services and epidemics evaluating their impact - Health influenced by many factors, surveillance PASSIVE SURVEILLANCE; role data coming from many sources; of the surveillance agency → - Vital stat. Data, birth and deaths waiting for reports to come to - Environmental data on air and water them quality - Legislation requires - Health service indicators, such as physicians and hospital discharges laboratories to report - Census information on the population, them to the public health such as income, language, and ethnic agencies once they are group suspected/diagnosed - More of routine work on regional and - Hospitals and billing data provincial health authorities as well as - Primary clinician the public health agency of canada - Mortality data involves surveillance - Birth and death certificates 2 TYPES OF SURVEILLANCE - Sentinel surveillance - Passive - Active ACTIVE SURVEILLANCE; Gathering the data in real-time ROLE OF CLINICIANS IN SURVEILLANCE - Resource-intensive, only - The primary care physician must inform the done when finding a public health service by reporting occurrences specific purpose of certain contagious diseases that they see - Surveys - Primary care physicians are typically the ones - Census data who see the initial cases in what may become an epidemic CENSUS OF POPULATION; - They may see more people than usual provides detailed statistical presenting with a particular condition, portrait of canada and its people and patients might mention other by their demographic, social, people with same symptoms and economic characteristics - Funded through federal/provincial system - Clinicians play super important role through SENTINEL SURVEILLANCE; surveillance system Clinicians keeping watching for disease of interest PASSIVE SURVEILLANCE; role of the surveillance - Selected clinicians gather agency → waiting for reports to come to them data and pass them onto - Legislation requires physicians and laboratories  those responsible for to report them to the public health agencies surveillance once they are suspected/diagnosed - Used to report rare - Allows public health to identify possible events outbreaks early and implement timely prevention and control measures DATA ACCURACY - Reports may take the form of routinely - Never perfect collected data - Need a limitation section - Hospital discharge summaries - Might be biassed in some - Mortality data way - Physician billing data - Can contextualize - Assembles information on notifiable evidence to make policy diseases must be reported (judged by when getting information public health significance) from data - Some provinces have legislation that requires notification of possible outbreaks, even if the COMPUTERS IN PUBLIC disease itself is not notifiable HEALTH - For example, the Quebec Public Health - Computers are Law demands that “any physician who overtaking most suspects the presence of a threat to the administration and health of the population must notify the documentation appropriate public health director.” procedures and handling of data in the medical HOSPITAL AND BILLING DATA field to allow for efficiency - As a resource for passive surveillance, hospital and data discharge summaries can provide useful accessibility-quickness information on patterns of disease and on the therapies being used DATA CONFIDENTIALITY - Availability of services greatly influence their - Discrimination and use, comparisons in data between places and stigma over time are of limited value for disease - Access to information surveillance - Physicians billing data can be used, but new PII (PERSONAL IDENTIFIABLE methods of physician remunerations and INFORMATION); any inaccurate or missing diagnosis limit the representation that permits the usefulness of this data source identity of an individual to whom the information applies to be VITAL STATISTICS reasonably inferred by either - Birth (and death) certification is important direct or indirect means - Your receipt creation and refund on - Directly identifies an earth individual - Birth certificates contain information - An agency tends to supplied by mother about the child’s identify specific family, including names, addresses, individuals in conjunction ages, race-ethnicity, and education with other elements levels (indirect identification) - Medical and health information is suplied by hospitals, doctors, or other PHI (PERSONAL HEALTH birth attendants concerning prenatal INFORMATION); oral or written care, birth weight, medical risk factors, information about individual complications of labour and delivery, - Relates to the individual’s obstetrical procedures, and physical or mental health, abnormalities in the newborn including family health - Main use is for public health research, history providing the data that can be used to  - Relates to the provision relate features of mother and her of health care, including pregnancy to the health of the child the identification of persons providing DEATH CERTIFICATES - Information on death certificates is subject to EPIDEMIC CURVES; Plotting number of uncertainties epidemic curves and gathering - Dpeends on how well the informant information on the geographical knew the deceased and the distribution of cases form circumstances of the death important steps in characterizing - Example; information on parents, an outbreak and in judging the educaiton, and occupation may not be likelihood of its transition into full known if the decedent is an elderly epidemic person with no surviving relatives - Often difficult in the accuracy and consistency TYPES OF OUTBREAKS with which causes of death are specified BASED ON SUSPECTED - Innaccurate diagnoses are common EXPOSURES; - Absense of autopsy - Common source - Exact cause of death may not be known outbreak - Exact time of death may not be specific - Point source - If a number of conditions contribute to the fatal - Continuous process, underlying causes and immediate source causes may be confused - Intermittent - For some conditions (AIDS/suicide), - Index case and limited cause of death may be misstated spread deliberately by the local official because - Propagated spread of social stigma COMMON SOURCE ACTIVE SURVEILLANCE; Gathering the data in OUTBREAK; people exposed to real-time one noxious influence - Active collection of data - Point source outbreak - Resource-intensive, only done when finding a - Continuous source specific purpose outbreak - Example; Canadian Paediatric Society routinely sends letters to every POINT SOURCE OUTBREAK; paediatrician asking them to report on When the exposure is very brief, cases of rare conditions, such as acute most people get sick at one flaccid paralysis, to assess the success incubation period following of polio vaccination → the Society then exposure reports the data to the Public Health Agency of Canada CENSUS DATA - Data collected through the vital statistics system and other methods must be converted into rates or proportions if they are to be useful for many public health purposes - Calculation requires information on the number of people in the population being referred to - Number serves as the denominator - Vital statistic is used as a numerator - Example; calculate age-adjusted or - Source may occur for a age-specific rates, necessary to know brief time or it may how people are in each age group persist  CENSUS IN CANADA CONTINUOUS SOURCE; - Information on the population denominators sometimes the exposure to a required for interpreting most sureilance common source can be information comes from the censes prolonged - First national canadian decennial was carried out in 1871 and there has been a census every 10 years since, in years ending in 1 - Since 1956, there has been as additional census in the years ending in 6 - Both censuses cover the entire population and collect basic demographic data - In addition, more detailed information is collected from a random 20% sample of - Cases rise over an population, covering range of demographic, extended period → social and economic topics (~50 questions) but coming from a common/ not including health single source - Longer and flatter → CENSUS OF POPULATION; provides detailed longer duration of statistical portrait of canada and its people by their exposures and variation demographic, social, and economic characteristics between people in incubation periods SENTINEL SURVEILLANCE; Clinicians keeping - Curve ending → people watching for disease of interest are treated or get - Selected clinicians gather data and pass them immunity onto those responsible for surveillance - Relative flatness of the - Used to report rare events curve → infection comes - Passive surveillance from a common source & - Help improve quality of care NO person-to-person - Trusted participants are carefully chosen to spread trust data with - Example; the Canadian Primary Care INTERMITTENT EXPOSURE; Sentinel Surveillance Network links irregular pattern of cases that selected family health teams via an reflects the timing and extent of electronic record surveillance system. repeated exposures - This can be used both to report rare events (such as side-effects from COVID-19 immunizations) and to help improve the quality of care (as with monitoring inappropriate use of antibiotics) - If sample of physicians is carefully designed, estimates can be made of the population incidence of the event of interest without the need to survey the entire population - May not be initially clear whether this is a common HEALTH SURVEYS source - Sample large random participants to get - Or there are several feedback from sources - Smaller samples do better estimates for - BUT successive peaks epidemiology in reasonable, economic do not become larger constraints and merge - Active surveys - Epidemic curves - Canadian community health survey  suggests (CCHS) and national census non-transmissible - Every 2 years gathers data on general health condition and health habits from a large random population sample INDEX CASE AND LIMITED - In the intervening years, collects data on SPREAD; person-to-person specific health topics from smaller samples spread showing single index - Surveys can target specific groups, case infecting other people such as injection drug users or people within incubation period with particular diagnosis, to document changes in patterns of behaviour that may affect disease or transmission NEED LOTS OF DATA; - Need lots to make better decisions and observations on being alert to plan for potential epidemics - Early notification of communicable diseases cases is a classic use of public health - Point source with information to protect the public health secondary transmission - Need for public health intervention to control - Outbreak wanes when other problems may not be obvious without an the infected people no analysis of data longer transmit the - Explains the Institute of Medicine infection to other committee’s insistence on the susceptible people due to importance of assessment as a core successful control function of public health measures - Public health leaders are increasingly stressing - Isolation importance of planning, setting goals, and - Quarantine managing public health programs to met these goals PROPAGATED SPREAD; - Process that requires data at the local infection from an index case, but state, and federal levels secondary cases of disease act - Example; a community may not as sources to infect new people recognize that it has a problem with that become new sources for unintended pregnancy and low-birth another case weight unless it analyzes the data from birth certificates, comparing local data with statewide or national averages - Recognition of the problem might persuade local public health leaders to consider school-based birth control education and services DATA ACCURACY - Not all data is perfect - Peaks are initially - Need a limitation section separated by one - Might be biassed in some way incubation period → - Can contextualize evidence to make policy merges into one large when getting information from data peak - Most other sources of health information, - Epidemic continues until relying as they do on surveys or voluntary the remaining pool of reports, are even more incomplete or subject to susceptible individuals bias - Errors in reporting cause of death certificates, a  shrinks or until control prime example of the error are especially measures take effect worrisome for public health in that mortality data have such strong influence on planning and priority setting for public health programs Questions COMPUTERS IN PUBLIC HEALTH - Computers are overtaking most administration and documentation procedures and handling of data in the medical field to allow for efficiency and data accessibility-quickness - Computers are extensively used in the analysis of public health data and new applications are continually improving the timeliness and accessibility of data - Weekly report of notifiable diseases are transmitted electronically to the CDC allowing prompt response to new outbreaks - Laboratory results are reported electronically, facilitating the rapid identification of bacteria and viral strains that may be causing illness in scattered locations around the country - Databases are kept up-to-date by electronic filings can provide rapid feedback on the effectiveness of new public health interventions as well as help detect emerging problems - New information technology, public health informatics, has vastly improved the accessibility of public health information to public health workers and the general public DATA CONFIDENTIALITY - Discrimination and stigma - Access to information - Conflict for the need of confidentiality and the need for open access to information has been played out over various aspects of the AIDS epidemic - Because HIV-positive individuals feared for good reason - Might be discriminated against if people learned of their infection - Public health practitioners were concerned that patients would refuse to be tested unless confidentiality was ensured - Rules for reporting HIV were handled differently from other communicable diseases; anonymous testing was allowed, and system for reporting cases to many state health departments and  the CDC was modified to maintain anonymity PII (PERSONAL IDENTIFIABLE INFORMATION); any representation that permits the identity of an individual to whom the information applies to be reasonably inferred by either direct or indirect means - Directly identifies an individual (based on whatever is known about them, information belonging to them, in relation to the individual) - An agency tends to identify specific individuals in conjunction with other elements (indirect identification) PHI (PERSONAL HEALTH INFORMATION); oral or written information about individual - Relates to the individual’s physical or mental health, including family health history - Relates to the provision of health care, including the identification of persons providing - Types of care; - Is a plan of service for individuals requiring long-term care - Relates to payment or eligibility for health care - Relates to the donation of body parts or bodily substances or derived from the testing or examination of such parts or substances - Is the individual’s health number - Identifies an individual’s substitute decision-maker - Any other information about an individual that is included in a record containing personal health information is also included in definition - Protected under Personal Health Information Protection Pact EXAMPLES OF SURVEILLANCE DATA - Patterns of disease development in a population; epidemic curves POPULATION OF DISEASE DEVELOPMENT IN A POPULATION; EPIDEMIC CURVES - When our efforts to prevent disease fail and an outbreak develops, cases arise over time in various patterns described by “epidemic curves”. - These plot the numbers of new cases arising over time, a population equivalent to the natural history of a disease for an individual case. - The natural history of a population outbreak is most evident in infectious disease, but also occurs in situations  such as a chemical spill that causes respiratory disease or, on a much longer time-scale, in non-communicable, chronic diseases. - Shape of the resulting epidemic curve(s) can suggest the - nature of the disease - mode of transmissions - temporal pattern, - the curve shows the magnitude of the outbreak (the number of cases), - the likely incubation period for the condition, - Reveal outliers (in time and perhaps in place) - To characterize different types of outbreak, the Centers for Disease Control classify epidemic curves based on the suspected type of exposure - Types of outbreaks based on suspected types of exposure; - Common source outbreak - Point source - Continuous source - Intermittent - Index case and limited spread - Propagated spread - Plotting epidemic curves and gathering information on the geographical distribution of cases form important steps in characterizing an outbreak and in judging the likelihood of its transition into full epidemic - Information on the success of policy reactions to changing disease patterns is gained from health services research COMMON SOURCE OUTBREAK; people exposed to one noxious influence - Point source outbreak - Continuous source outbreak POINT SOURCE OUTBREAK; When the exposure is very brief, most people get sick at one incubation period following exposure  - Point source outbreak example above^^ - Source may occur for a brief time or it may persist - Example; Staphylococcus aureus poisoning from tainted food at a wedding dinner (awkward for the hosts). This produces a single curve that wanes quickly, as long as there is no person-to-person spread CONTINUOUS SOURCE; sometimes the exposure to a common source can be prolonged - Continuous source outbreak example above^^ - Contaminated water supply - Restaurant failing to fix faulty ref. System - Cases rise over an extended period but still originate from a common or single source - Resulting epidemic curve becomes longer and flatter → longer duration of exposures and  variation between people in incubation periods - Curve ends when the source of the contamination is corrected, or when all susceptible people develop immunity - Relative flatness of the curve → infection comes from a common source & NO person-to-person spread - The number of cases would mount over time as one person infects others INTERMITTENT EXPOSURE; irregular pattern of cases that reflects the timing and extent of repeated exposures - May not be initially clear whether this is a common source - Example; Industrial contaminant emitted at intervals - Or there are several sources - Example; series of outbreaks of food poisoning occurring at different summer camps for children - Gaps between the outbreaks could initially suggest person-to-person transmission followed by incubation period - BUT successive peaks do not become larger and merge (as they would if the outbreaks were due to infectious spread; mass infection) - Epidemic curves suggests non-transmissible condition INDEX CASE AND LIMITED SPREAD; person-to-person spread showing single index case infecting other people within incubation period  - Point source with secondary transmission - Outbreak wanes when the infected people no longer transmit the infection to other susceptible people due to successful control measures - Isolation - Quarantine - Medication - Vaccines? PROPAGATED SPREAD; infection from an index case, but secondary cases of disease act as sources to infect new people that become new sources for another case - Taller peaks in each generation of secondary and tertiary cases - Peaks are initially separated by one incubation period, tending to merge into one large peak - Epidemic continues until the remaining pool of susceptible individuals shrinks or until control measures take effect  - Example; measles spreading from person to person Summary - Surveillance definition - Passive - Hospitals and billing data - Primary clinician - Mortality data - Birth and death certificates - Sentinel surveillance - Active - Surveys - Census data - Data accuracy and confidentiality - Computers in public health/medical field - PII and PHI - Epidemic curves - Types of outbreaks based on suspected types of exposure; - Common source outbreak - Point source - Continuous source - Intermittent - Index case and limited spread - Propagated spread TUTORIAL WORK; QUESTIONS ON MIDTERM - How to define risk - Risk; probability that event occurs in a time period (health status change; healthy → exposure → outcome of disease) - Define this as the probability that an individual with certain characteristics will experience a health status change over time - Free of outcome in beginning - Has outcome in the ending - Important to measure incidence - Quantity to help estimate risk - Cohort studies; start with population, measure exposure, group exposed, assess association that have exposure with outcome, and the non-exposed to the outcome - Assesses incidence (sees those that are healthy and the exposed) - Advantage; Temporal association that they didn’t have it at first and now have they have it (causal inference) - Disadvantage; expensive and can be lengthy to see results (need to see results in exposed participants over an extended period of time, and might come out with no results) - Problems with rare diseases in cohort studies - Choosing population? Seeing the legibility and not wasting the worth of the participants based on the length of the study and the age-criteria of the participants (10-20 good length for a cohort studies, so older people being the participants are better) - Risk factors vs. hard outcome - Hard outcome= diseases themselves (cancer, hiv, aids) - Risk factors= increase hard outcome probability. Not a definite guarantee of a hard outcome, but more exposure to the risk factor increases the risk of having it, but not get it definitely. - Not the actual disease but the disease causer itself - Used in intervention studies since much shorter (expensive though) - Case-control studies; control the outcome than the cases - Advantage; more effective and efficient since no waiting for individual to develop disease since they have the disease (questionnaires) - Good for studying rare diseases - Disadvantage; difficult to find appropriating controls based on how many characteristics they are matched with another person  - Exposure and outcome; difficult to assess when they got it since they already have it, and temporal trends - Cannot estimate incidence directly through this type of approach (did not start with healthy and cannot measure how long and their exposure rate to disease) - How to remember what you did to cause the risk to manifest years ago (assessed via exercise) - Self-report bias; they don’t know how it started, but know what they have currently - Observational study design can help estimate this quantity for different exposures - Determine whether there is an association between exposure and outcome