Concepts of Association and Causation in infectious dis epid.pptx

Concepts of Association and Causation in infectious

1. Concepts of Association and Causation in Infectious Disease Epidemiology PRESENTER: Dr Priya Gautam MODERATOR: Dr Manish Kumar Goel FACILITATOR : Dr Nandini Pittala

2. Plan of Presentation Introduction From Association to Causation: Historical Perspective Concept of Association Types of Associations Concepts of Causation Factors for Causation Models of Causation Guidelines for Causality Association vs Causation

3. The Doctor Who Drank Infectious Broth, Gave Himself an Ulcer, and Solved a Medical Mystery - Ulcers long thought to be caused by stress, spicy food. - Marshall & Warren discovered H. pylori in ulcer patients. - Marshall drank H. pylori culture → developed gastritis → cured with antibiotics. - Won Nobel Prize in 2005. This story shows how careful investigation can move us from associations (stress, spicy food) to true causation (H. pylori). “Doctor drinks bacteria to prove ulcers are infectious!”

4. One of the objective of epidemiology is that it seeks to identify causes of disease. An observed association may have several explanations. Observed association ≠ Always Causation Most studies are observational, not experimental. INTRODUCTION

5. Before inferring causation, we must rule out Chance, Bias, and Confounding Only after careful evaluation can a cause–effect relationship be established.

6. Confounding A confounder is factor that is associated with both the exposure and the outcome . It is a third variable that distorts the true relationship between exposure and outcome. Criteria for a Confounder Associated with the exposure. Independently associated with the outcome. Be a risk factor for that disease. Not an intermediate in the causal pathway.

7. Hume (18th c.) Observation only: Cause ≠ directly seen John Snow (1854) Mapping & reasoning: Cholera & water Koch (1884–1890) Postulates: Lab proof of causality Pettenkofer (19th c.) Multifactorial causation: Social & nutrition factors Rivers (1937) Immunology: Antibody timing as evidence Bradford Hill (1965) Nine Criteria for Causality Evans (1967) Five Realities: Host, environment, time MacMahon & Pugh (1970) Web of Causation: Multiple interacting factors Historical Perspective

8. Historical Evolution 1. Hume (18 th century) noted that we are never able in a single instance to discover any connection which binds the effect to the cause. We only observe that one event happens after another 2. John Snow (1854) (Father of Modern Epidemiology): He gave causal inference using mapping (deaths by cholera) and epidemiologic reasoning (contaminated water) rather than laboratory evidence. 3. Koch Postulates (1884–1890) : Lab-based criteria for establishing microbe-disease causality. 4. Pettenkofer (19th century): Emphasized multifactorial causation, challenged simple germ theory by emphasizing factors like poverty, crowding, and nutrition in diseases like tuberculosis. 5. Rivers (1937) : Immunological aspect of causation : The timing and presence of antibodies in a patient’s blood are as important as the virus itself in showing its role in disease 6. Bradford Hill (1965) : Sir Austin Bradford Hill introduced nine epidemiologic criteria (strength, consistency, specificity, temporality, etc.) to infer causality when Koch’s postulates fall short. 7. Evans (1967) further advanced this by introducing the “Five Realities” of acute respiratory diseases, recognizing host susceptibility, environment, and temporal variability in disease causation. 8. MacMahon & Pugh (1970) introduced the "Web of Causation" model—highlighting that disease etiology often involves a complex interplay of multiple factors, not a single agent.

9. In 1840, Henle proposed postulates for causation that were expanded by Koch in 1880s.The postulates for causation were as follow: The agent must be present in every case. The agent must be isolated in pure culture. The agent must reproduce the disease in a healthy host. The agent must be reisolated from the experimental host e.g.: Anthrax was first disease demonstrated to meet these criteria. These postulates are useful in infectious diseases. Issues arise when disease are noninfectious, in such diseases there was no organism that could be isolated, cultured , and grown in susceptible host. So, evidence for causation in noninfectious diseases is hard to find or explain.

10. Rivers criteria 1937 : Modified Kosh Postulates for virus one key point was that a specific immune response should appear after infection. Huebner’s criteria 1957 : Expanded on Rivers, again stressing the role of serological response that is antibody production as evidence that virus is causing disease. Thus in both framework demonstrating an immune response was a major proof of causation. Why absence of immune response invalidates these criteria? Normally HBV infection Hbs antigen appears followed by antibodies but in immunocompromised individual antibody response may be absent. Rivers Hubner serological requirement would fail here even though HBV is the true cause. Infants born to HIV infected mothers may not produce their own antibodies. They may only carry maternal antibodies. Therefore no clear immune response which makes the criteria unreliable even though HIV causes AIDS.

11. The Five Realities of Acute Respiratory Disease- by Evans The same clinical syndrome may be produced by a variety of agents. The same etiologic agent may produce a variety of clinical syndromes. The predominating agent in a given clinical syndrome may vary according to the age group involved, the year, the geographic location, and the type of population. Diagnosis of the etiological agent is frequently impossible on the basis of the clinical findings alone. The cause of a large percentage of common infectious disease syndrome is still unknown.

12. Concept of Association Definition Association may be defined as the concurrence of two variables more often than would be expected by chance. Association does not necessarily imply a causal relationship. To define degree of association between two characteristics, correlation is used.

13. Types of Association Association Spurious Association Indirectly causal Association Directly causal Association Not statistically significant Statistically significant Association Non causal Causal

14. Type of Association Is it real? Is it causal? Mechanism Example Spurious No (false) No Due to bias, chance, confounding Perinatal mortality and hospital birth Indirect Secondary (Non-causal) Yes (true) No Confounder explains link High altitude & Goitre (iodine deficiency is real cause) Indirect causal Yes (true) Yes (but indirect) Through an intermediate factor A →B→D Malnutrition →Low Immunity → TB D irect one to one causal association Yes (true) Yes Single cause → single effect, no intermediate factor Measles virus → Measles Direct Multifactorial causal association Yes (true) Yes Multiple factors act together directly CHD (smoking, HTN, hyperlipidaemia, diabetes, obesity)

15. Concept of Causation A cause of a disease is an event, condition, characteristic or a combination of these factors which plays an important role in producing the disease.* Beaglehole R, Bonita R, Kjellström T. Basic Epidemiology . 2nd ed. Geneva: World Health Organization; 2006 *

16. Factors which play role in causation 1. Predisposing factors- Characteristics which may create a state of susceptibility to a disease agent. Example- age, sex, low immunity etc. 2. Enabling factor- Factors which may favour the development of disease by allowing or promoting exposure. Example- low income, poor nutrition, bad housing etc. Four types of factor play a part in causation of disease. All may be necessary but they are rarely sufficient to cause a particular disease.

17. 3. Precipitating factor- The immediate triggering event or exposure that directly cause the onset of disease. Example- Exposure to virus leading to influenza. 4. Reinforcing factor- Factor that aggravate or contribute to re- occurrence of disease. They maintain or worsen the problem. Example- Repeated exposure to a risk factor i.e. smoking again after quitting.

18. General Models of Causation The most widely used models are: 1. Chain model of Infectious Disease 2. Tetrahedron model 3. The epidemiological triad 4. The wheel 5. The sufficient cause and component causes models (Rothman Component Causal Pie model)

19. Chain model of Infectious Disease

20. Components: (Usually 6 links in the chain) Infectious agent – the microorganism causing disease (e.g., Mycobacterium tuberculosis ). Reservoir – where the agent normally lives and multiplies (e.g., humans with active TB). Portal of exit – how the agent leaves the reservoir (e.g., coughing/sneezing). Mode of transmission – how it spreads to a new host (e.g., airborne droplets). Portal of entry – how the agent enters a new host (e.g., inhalation into lungs). Susceptible host – someone who can get the disease (e.g., malnourished or immunocompromised person). Breaking any link can stop transmission (via vaccination, isolation, masks)

21. Tetrahedron Model Gives a more dynamic view of infectious disease causation. The tetrahedron has 4 parts: Agent – pathogen (same as above) Host – susceptible individual Environment – physical, biological, social factors that favor disease spread Time – duration of exposure, incubation period, seasonality Emphasizes that disease occurs only when agent, host, environment, and time interact.

22. The epidemiological triad model *Global Infectious Disease Epidemiology online network

23. The Wheel of Disease Causation Given by “ Mausner and Kramer” in 1985. Biological environment Host (Human ) Genetic core Biological environment Physical environment Social environment Wheel of causation

24. Web of Causation Given by Manmohan and Pugh (1970). The “web of causation” is a complex interrelationship of multiple factors that contribute to the occurrence of disease. It is not the primary model for infectious diseases. Diseases usually result from multiple interrelated factors (biological, environmental, social, lifestyle). Sometimes, removal or elimination of just only one link or chain may be sufficient to control the disease. Highlights importance of prevention at multiple levels (individual, community, environment).

25. Rothman’s Component Causes and Causal Pies model Causes of disease consist of collection of factors. Factors represent pieces of pie. The whole pie is sufficient cause for disease. The disease may have more than one sufficient cause with each cause being composed of several factors. These factors are called component causes .

26. This disease model differentiates between “necessary” and “sufficient” cause. Necessary cause : Causal factor whose presence is required for the occurrence of the effect. Without that factor the disease never develops. Sufficient cause : The minimum set of conditions factors that is needed to produce a given outcome. In the presence of that factor the disease always develop.

27. Sufficient and necessary cause- A factor is both required and alone capable of producing the disease. Example: Rabies virus → Rabies Presence of rabies virus is necessary (rabies does not occur without it). The virus alone (if inoculated into a susceptible host) is sufficient to cause rabies. Neither sufficient nor necessary- A factor cannot cause disease on its own, nor is it required in every case. Example: Malnutrition → Tuberculosis Malnutrition alone does not cause TB (not sufficient).TB can occur in well-nourished individuals also (not necessary).

28. Necessary but not sufficient- A factor cannot cause disease on its own, and is required in every case. Example: Mycobacterium tuberculosis → Pulmonary tuberculosis The bacillus is necessary (TB cannot occur without it). But infection alone is not sufficient → requires cofactors such as poor immunity, malnutrition, overcrowding Sufficient but not necessary- Example: Hepatitis B virus → Hepatitis Hepatitis B virus is sufficient to cause Hepatitis. But hepatitis is not always caused by HBV (it can also be caused by HAV, HCV).

29. Kenneth Rothman proposed this model (Causal Pie) to explain causation of disease as a combination of different factors. Example- Pulmonary Tuberculosis (TB): Sufficient Cause (the full pie) Set of all factors that together lead to active TB disease. Component Causes (slices of the pie) might include: Mycobacterium tuberculosis infection (necessary cause – appears in every pie). Poor nutrition Overcrowding/poor ventilation HIV infection or other immunosuppression Socioeconomic deprivation (poverty, poor access to health care).

30. Sufficient cause / Minimal causal mechanism “By sufficient cause we mean a complete causal mechanism, a minimal set of conditions and events that are sufficient for the outcome to occur. A sufficient cause may be defined as a set of minimal conditions and events that inevitably produce disease. “Minimal” implies that none of the conditions or events is superfluous.

31. Observed association-Could it be due to selection or information(recall) bias? Could it be due to confounding? Could it be a result of chance? Could it be causal association? Apply guidelines and make judgement No No No Assessing the relationship between a possible cause and an outcome

32. Considerations taken into account when evaluating whether a precisely measured association arose from causal effect List were given by US Surgeon General (1964) which was further strengthened by Bradford Hill (1965) Strength Consistency Specificity Temporality Biological Gradient Plausibility Coherence Experimental Evidence Analogy

33. Strength of Association Strength of association means how strongly an exposure is linked to a disease. According to Hill’s considerations, the stronger the association (i.e., the higher the RR or OR), the more likely it is to be causal, although weaker associations can still be causal, and strong associations could be due to confounding.

34. Helicobacter pylori is found in at least 90% of patients with duodenal ulcer. Examples- HIV & AIDS HIV Exposure ─────► Much Higher Risk ─────► AIDS (Yes) (Relative Risk >10) (Disease develops) HIV Exposure ─────► Very Low Risk ─────► No AIDS (No) (No disease) Example of Strength of association

35. Consistency Consistency refers to the repeated observation of an association in different populations under different circumstances. An association is more likely to be causal when it is consistently observed across different persons, settings, times, and methods. It is demonstrated by several studies giving the same results. If relation is causal, we would expect the finding to be consistent with other data. Lack of consistency does not exclude a causal association.

36. Example- In high-coverage areas (over 95%), measles outbreaks still occurred—especially due to cold-chain failures or waning immunity. Finding : Areas without a second measles vaccine dose experienced outbreaks consistently. Interpretation : Since the same result is seen in different places, it clearly shows that the second dose protects against measles outbreaks.

37. Specificity Implies “one to one” relationship between the cause and effect. Most difficult to establish because: A single cause or factor can give rise to more than one disease. Most disease are due to multiple factors with no possibility of demonstrating one to one relationship. Specificity shows causal relation but lack of specificity does not negate it. Example- Treponema pallidum → Syphilis

38. Strong but non causal association – explained by confounding. For example relation between Down Syndrome and birth rank which is confounded by the relation between Down syndrome and maternal age. Once the confounding factor is identified the association is diminished by adjustment for the factor.

39. Temporality refers to the necessity for a cause to precede an effect in time. Causal attribute/ exposure to the factor must precede the observed effect or diseases. Example - Exposure- Drinking water from contaminated well Outcome- cholera onset Time sequence- exposure (contaminated water) preceded the illness Temporal Association

40. EXAMPLE EXPOSURE OUTCOME TIME SEQUENCE Cholera Drinking water from contaminated source Onset of Cholera Exposure (contaminated water) preceded the illness 2. Food Poisoning Eating food contaminated by staph aureus e.g. dairy, meat Onset of food poisoning symptoms such as vomiting, diarrhoea Symptoms start in 1-6 hours 3. Measles Contact with measles case (droplet infection) Fever, cough, rash Exposure → incubation ~10–14 days → prodromal fever → rash 4. Tuberculosis Inhalation of droplet nuclei from infectious pulmonary TB case Latent or active TB disease Exposure first → latent infection (weeks to years) → later progression to active TB

41. Biological Gradient Biological gradient refers to the presence of a unidirectional dose- response curve. As the dose or level of exposure increases, the risk of disease also increases. Examples – TB household contact → high exposure (larger bacillary load) → high risk of active TB. Cholera → small inoculum may cause mild diarrhoea, large inoculum causes cholera. For non infectious disease- The higher the number of cigarettes smoked per day, the higher the risk of lung cancer and COPD.

42. Dose response relationship A causal relationship is strengthened when increasing levels of exposure are associated with increasing risk (or severity) of disease. Tuberculosis: Longer & closer contact → Higher risk of TB. Cholera: Higher dose of V. cholerae → More severe diarrhoea. Hepatitis B: Higher maternal viral load → Higher chance of neonatal infection

43. Biological Plausibility An observed association is more likely to be causal if it is biologically or medically reasonable, based on existing knowledge of pathophysiology and microbiology.* Examples- Malnutrition → ↓ immunity → ↑ risk of Tuberculosis (M. tuberculosis). HPV infection → Oncoproteins (E6, E7) → Cervical cancer. Vibrio cholerae → Cholera toxin → Profuse watery diarrhoea. Even if we don’t yet know the biological mechanism, the cause–effect link may still exist — our current knowledge may just be limited.