Epidemiological surveys
Epidemiology seeks to quantify the frequency of a health event in a population,
And to determine its biological, medical, environmental, socio-economic causes , its prognosis , its treatments , its prevention , its cost
Very broad field which uses statistics and provides statistics (WHO, UNICEF, etc.)
- DIFFERENT CLASSIFICATIONS OF EPIDEMIOLOGICAL SURVEYS
- Descriptive surveys
Objective : To provide health statistics concerning populations
They study the frequency and distribution of health indicators or risk factors and their variations according to:
- of time,
- geographic areas
- population groups.
They allow us to raise, from these observations , hypotheses on the risk factors of diseases.
- Etiological investigations
They aim to highlight a relationship between one or more exposure factors and one or more diseases and to provide arguments in favor of a causal interpretation of this relationship (strength of the association, presence of a dose-effect relationship).
We need to discuss: biases and confounding factors
- Experimental investigations: Therapeutic trial
They aim to:
- To demonstrate the superiority of one treatment over others
- To demonstrate equivalence They allow causality to be affirmed They are prospective
There are other rankings
So-called “observational” epidemiological studies
A descriptive study allows a state of health to be described An analytical study (Etiological) allows an explanation A prognostic study allows an evolution to be predicted
Experimental studies
Therapeutic or clinical trials
- DESCRIPTIVE SURVEYS
Allows the calculation of health indicators of a population, namely mortality, lethality, morbidity (prevalence and incidence), etc.
- Prevalence surveys
These are cross-sectional surveys that estimate the number of cases present in a population at a given time.
(Example: “on a given day” prevalence survey of nosocomial infections in healthcare establishments).
The prevalence rate is:
P= M/N (between 0 and 1)
M = number of cases (patients)
N = total number of subjects (sick + non-sick) They are used to measure the burden of care on the health system
- Incidence surveys
These are longitudinal surveys (following a group of subjects over time) that estimate the number of new cases of disease in a population, during a given period. Ex/ cancer registry
The incidence rate is the ratio of the number of new cases of disease recorded to the number of people likely to be affected in the population, during a given period.
TI = M/PT
M = number of new cases (patients) during the study period PT = number of person-times (person-years, most often)
For example, we are talking about an annual incidence rate of 10 cases per 100,000 people per year. They are used to detect the start of epidemics.
- ETIOLOGICAL (OR ANALYTICAL) SURVEYS
Etiological studies analyze the relationships between exposure to a risk factor and a health condition . Etiological studies are always comparative (they compare two different groups either on the presence of the disease or on the presence of the risk factor).
A risk factor is a characteristic associated with a higher likelihood of disease. The causal role can only be established through experimentation.
- Exposed/Non-Exposed Surveys
Exposed/non-exposed surveys consist of comparing the proportion of sick people (or deaths) observed between a group of subjects exposed to a risk factor and a group of subjects not exposed to this risk factor, initially free from the disease.
The exposed/non-exposed survey can be:
- prospective (the groups are formed at the beginning of the study and the subjects are followed up over the years following their inclusion)
- retrospective (the inclusion of subjects is done from a date in the past sufficiently distant so that the disease has had time to develop and the follow-up period is shortened (date of hiring of employees of a factory for example).
The exposed/non-exposed survey allows the relative risk of the disease (RR) to be estimated.
The advantages:
Selection on exposure, regardless of disease
Disadvantages: the survey can take a long time, the costs are high and there is a risk of loss of follow-up.
| Exposed/Not Exposed | Cancer + | Cancer – | Total |
Ethylism + | 73 | 927 | 1,000 |
Ethylism – | 37 | 963 | 1,000 |
Total | 110 | 1,890 | 2,000 |
- Calculation of relative risk
Example: Lung cancer and alcoholism
Cancer in alcoholics:
R1 = 73/1000
Cancer in non-alcoholics:
R0 = 37/1000
Relative Risk (RR)
= R1/R0 =1.97
An exposed subject has 1.97 (approximately 2) times more “chance” of developing cancer than an unexposed subject. Interpretation of relative risk (RR)
RR > 1: Exposure to the factor increases the frequency of the disease: The factor is a risk factor
RR = 1: Exposure to the factor does not influence the frequency of the disease
RR < 1: Exposure to the factor reduces the frequency of the disease: The factor is a protective factor
- Excess risk and attributable risk
- Excess risk = R1-R0
- Attributable risk = (R1-R0)/R1= (RR-1)/RR
- Case-witness investigations
Case-control studies consist of comparing the frequency of previous exposure to one (or more) risk factor(s) in a group of “cases” suffering from the disease studied, and in a group of “controls” free from it.
Risk factor measurement is performed retrospectively. Case-control studies are preferably aimed at rare diseases.
The relative risk of disease cannot be estimated in a case-control study. An Odds ratio (OR) is calculated.
Advantages: quick response (short study period)
Disadvantages: Memory effect, possibility of error (bias) in the selection of subjects
| lung cancer (Case) | non-sick (Witnesses) | total | |
| smokers | 133(a) | 102600(b) | 102733 |
| non-smokers | 3(c) | 42800(d) | 42803 |
| total | 136 | 145400 | 145536 |
- Calculation of the odds ratio or odds ratio OR = ad / bc
OR = (133*42800)/(3*102600)= 18.4
Interpretation of OR
Same as a RR
We can test whether the association is due to chance:
Statistical test of the existence of the link on the table Or calculation of a confidence interval of the OR
- Prognostic surveys ( survival study type)
Allows a cohort of subjects to be followed and the number and time to occurrence of an event to be assessed (survival curve).
Example: the evolution of patients infected with HIV with regard to the appearance of Kaposi’s sarcoma.
- Patients are included in the study when they are discovered to be infected with HIV.
- We follow these patients regularly over time, for example every year on the anniversary of the discovery, and for each person we note the presence or absence of Kaposi’s
- We thus obtain for each patient a follow-up period (which is not the same for each patient) and a status (presence or absence of Kaposi)
- Some patients are sometimes lost to follow-up
- Experimental investigations: Therapeutic trial
Context
The development of a new drug follows different phases
- Basic development without human testing
Synthesis, determination of animal toxicity, efficacy on models (cell cultures, animal models, etc.), effect on reproduction (teratogenic effect, effect on fertility, etc.)
- Development in humans
- Phase 1 : In healthy volunteers, search for adverse effects, choice of galenic form (bioavailability)
- Phase 2 : In the patient Dose/action curve,
- Phase 3 : In the patient’s real treatment situation
- Phase 3 ends with the marketing authorization (MA) application. The drug is available for sale and can then be prescribed.
- Phase 4 : Pharmacovigilance. This involves monitoring the effectiveness and safety of the treatment with periodic reassessment of the “benefit/risk”
Placebo effect
Beneficial effect of the treatment provided outside of any real physiological effect “The simple fact of having seen the doctor makes me feel better”
In the therapeutic trial, we hope that the drug being tested works better than “the crumb of bread” or a simple meeting with the doctor.
Placebo effect = parasitic element that we are trying to get rid of
Compliance with regulations and ethical rules
Regulations depend on the country
Certain attitudes are prohibited in certain countries: Voluntary Termination of Pregnancy, use of certain genetic techniques, etc.
Ethical rules are universal (Interest of ethics committees)
- Respect for human and children’s rights
- Taking into account patient consent
Causality
The existence of a significant link is not necessarily a causal link (existence of a confounding factor)
Only comparative, randomized, prospective, double-blind studies are highly likely to conclude in terms of causality.
The Criteria of Causality
Austin Bradford Hill 1965: 9 criteria 1- Strength of association Level of correlation Importance of difference
- Consistency
External: Reproducibility, consistency of results with other work, teams, populations Internal: protocol, consideration of biases
- Time The effect precedes the cause
- Dose-effect relationship
- Experimental evidence
- Specificity of the association
- Biological coherence Depending on the natural history of the disease , pathophysiology
- Biological plausibility (in view of available knowledge)
- Analogy (With other events linked to another pathology)
Reproducibility of results
In general, we find in the literature several investigations which deal with the same problems and which do not always find the same results.
We then use meta-analysis
Meta-analysis
- A meta-analysis is a statistical approach that combines the results of a series of independent studies on a given problem.
- Meta-analysis allows for a more precise analysis of data by increasing the number of cases studied and drawing an overall conclusion.
- This approach is widely used in medicine for the overall interpretation of sometimes contradictory clinical studies. It also allows for the detection of methodological biases in the studies analyzed.
- However, it can itself be subject to publication bias, as researchers may be less likely to publish a study concluding that there is no result.
Epidemiological surveys
Wisdom teeth can cause pain if they erupt crooked.
Ceramic crowns offer a natural appearance and great strength.
Bleeding gums when brushing may indicate gingivitis.
Short orthodontic treatments quickly correct minor misalignments.
Composite dental fillings are discreet and long-lasting.
Interdental brushes are essential for cleaning narrow spaces.
A vitamin-rich diet strengthens teeth and gums.