Moledule 1 - lecture notes PDF

Preview

Summary

lecture notes...

Description

POPLHLTH MODULE 1 Chapter 1 Epidemiology - The study of how much disease occurs in groups/populations and of the factors that determine difference in disease occurrence between groups/populations -

Used ‘disease’ rather than disease, to encompass any health related event(injury, heart attack, a death) or health related state(diabetes. Disability, quality of life)

-

Epidemiologists tend to study negative events or states, like death and disease, because they are easier to measure than positive states of health such as degrees of wellbeing.

-

Epidemiologists count dis-ease occurrences in populations of people. An ‘occurrence’ (of dis-ease)’ describes the transition from a ‘non-dis-eased state’ to a ‘dis-eased state’.

-

If the transition easily observable, like a road traffic injury or crushing central chest pain caused by a heart attack, then epidemiologists usually count the occurrences as the number of events over a period of time(incidence) if the transition is not easily observable, like transitioning from a non-diabetic to a diabetic state, then epidemiologists count the occurrences as the number of people with the dis-ease ‘state’ at a point in time(prevalence)

-

-

A population is any group of people who share a specified common factor.

-

Measuring dis-ease occurrence in a population can inform health service planners about types of health services required for populations, including health promotion, dis-ease prevention, disease diagnosis and treatment. - one group : health service required - more than one group : identify the causes or predictors

Numerators and denominators Denominator - the number of people in a study population Numerator - the number of people from the study population in whom dis-ease occurs (outcomes)

-

Epidemiology studies begins by defining the denominator Determine of the occurrence is different between groups

The gate frame Exposure Group Occurrence (EGO) (EGO = a/EG) Comparison Group Occurrence’ or CGO (CGO = b/CG). One could measure the occurrence of ‘no dis-ease’ in EG (= c/EG) and CG (=d/CG) and this is done in some studies, particularly diagnostic test accuracy studies (discussed later). Converting numerical outcomes in to categorical - in order to to the calculation

-

-

Numerical measures are converted into categorical measures Example above, numerical measures of salt consumption could be divided into two or more categories (e.g. high and low intake) The outcomes are categorical

-

When numerical outcomes are not converted into categorical measures, often calculate the mean or median level of the outcome (e.g. average blood pressure) in EG and CG.

-

More than one EG is possible but always only one CG

Incidence and prevalence

Incidence - It is often possible, and useful, to observe when a person transitions from a non-dis-ease to a dis-ease state (e.g. when a heart attack occurs) and epidemiologists usually count the number of these types of events that occur over a period of time. - Alway has unit - Vertical arrow - measure the occurrence of di-ease that have an easily observable onset - outcome to be categorical variable Prevalence - In other situations, it is only possible, and also more useful, just to determine if, not

-

-

when, the transition has occurred, yes or no, such as classifying someone as overweight or having diabetes. ount) how many people are overweight, or have diabetes at a specified point in time. Always does not has unit Horizontal arrow The dis-ease occurrence cannot easily to be observed and counted *** measure the amount of water in the pool, but you don’t know how much has been falling into the pool. You do not know when they developed the dis-ease, but you know at a particular point in time a measurable proportion of the study population has the dis-ease. A static measure that measure at a point in time Size of the pool is depends on the rate the dizzle falls into the pool and how much water is lost from the pool People can leave the prevalence pool by death or cured Less useful than incidence because it is dependent on the death and cure rate Outcome can be categorical(proportion) or numerical(mean or median)

-

Often measure the prevalence of disease at 2 point of time and calculate the change in prevalence The difference in prevalence between 2 points is in fact a measure of the incidence

-

Can be measured at the beginning or other period during the study

Comparing EGO and CGO

-

Wrong reporting, using the comparison difference in the wrong place Collect the right apparent link is important What is actually causing it to happen what is the effect

-

comparisons of dis-ease occurrence : ‘estimates of association’ between an exposure and an outcome.

-

Ratio of Occurrences (EGO ÷ CGO) relative risk (RR) no unit Difference in Occurrences (EGO – CGO) or vice versa ((CGO – EGO) r isk difference (RD) also refer as an absolute risk same units as the EGO and CGO

RISK RATIO - commonly known as RELATIVE RISK (RR = EGO ÷ CGO) - explain a relative risk by stating that the risk in one (specified) group is, say 2 times higher than in another (specified) group. - A relative risk can be any number greater than zero.

-

-

If there is no difference in the risk or occurrences of a dis-ease between the two groups being compared (i.e. EGO = CGO), then the relative risk = 1.0 (i.e. when the RR = 1, there is no difference in the effect of E and C on the study outcome; this is often known as the no-effect’ value). A relative risk that is less than 1.0 can also be expressed as a Relative Risk Reduction (RRR) – because it is reduced below 1.0 (i.e. the no-effect value). if the relative risk is greater than 1.0, it can be expressed as a Relative Risk Increase (RRI). s relative to non-smokers, suggesting that smoking increases the risk of heart

RISK DIFFERENCE - also known as ABSOLUTE RISK DIFFERENCE (RD = EGO - CGO) - When describing the relative risk, always state which risk (i.e. in which group) is being subtracted from which other risk. - A risk difference can be any number between minus infinity and plus infinity. - If there is no difference between the groups compared (EGO = CGO), the risk difference = 0 units (i.e. a RD = 0 demonstrates no difference in effect of E and C on the study outcome; this is also known as the ‘no-effect value;’ remember that the equivalent no-effect value for a RR = 1.0). Errors - Errors can occur due to problems with the study recruitment, design and implementation, analyses, or due to chance.

random errors : errors caused by chance 9  5% Cl is used to describe the amount of random error in the study results - non-random errors(biases or systematic errors) : errors caused by problems with how the study is designed or conducted RAMBOMAN - Acronym to identify where non-random errors can occur in epidemiological studies. - The letters refer to parts of the GATE frame (i.e. aspects of the study design) where errors can occur. -

Recruitment - Who was recruited/selected into the study? - Is the recruited/selected populations/groups representative to the study finding?

-

external validity error, because when it is present the study findings may not be applicable to a wider population

-

This type of error is particularly important when the main objective of the study is to measure the characteristics of a real population (the ‘Eligibles’), but the Participants (P) who are recruited are not representative of the Eligibles.

For example, consider a study in which the objective is to measure the prevalence of participation in sport at school among all New Zealand school children (the Eligibles) . For this study to be valid, the investigators must make sure that a representative sample of all New Zealand school children are recruited. The best way to do this is to obtain a list of all school children and choose a random sample of children from the list. If however the investigators only recruited participants from schools that require all children to participate in sport, then the prevalence of sport participation in the study participants will overestimate the true prevalence among all school children in New Zealand, because not all schools expect all children to participate in sport.

-

Of note, even if the investigators recruit participants correctly (i.e. a representative sample from a list of all NZ school children), it is still possible to recruit a non-representative sample just by chance alone, particularly if the sample is small. This is known as a random sampling error and is discussed later in this chapter under ‘Random Error.’

-

selection bias occurs when many/all of the participants who are allocated to the Exposure Group are recruited from a very different source from the participants allocated to the Comparison Group - confounding error caused by the allocation process

-

if the non-responders are different from the responders, this can cause a recruitment error ( also known as a non- response bias). There is no specific level of response (i.e. the response rate) that is considered unacceptable, but a response rate of less than about 70% of those invited could cause a significant recruitment error in prevalence studies like the physical activity in school children study described above.

Allocation Measurement error - Were the study participants correctly and successfully allocated into EG and CG? - EG and CG are measure incorrectly - When this occur, some participants will be incorrectly allocated to the wrong group

Confounding - Were the EG and CG similar at the beginning of the study? -

Common cause of bias in non-randomised studies

-

In the ideal epidemiological study, the participants allocated to the Exposure Group and to the Comparison Group would be so similar that, in the absence of the specific ‘exposure’ being studied (e.g. smoking or a new drug), the two groups would have the same occurrence of outcomes (i.e. EGO would equal CGO).

-

So, in this ideal study, if you do find a difference between EGO and CGO, it is likely to be caused by the study ‘exposure’ because everything else about the two groups (EG and CG) would be the same. However, if the exposure and comparison groups differ in other ways - not just the study ‘exposure’ - and if these other differences also have an effect on the study outcome, then i t is not possible to know whether the study exposure or the other factors caused EGO and CGO to differ. These other factors are known as confounders.

Random process allocation - randomised controlled trials (RCTs) experimental - the investigator control the allocation process -

In RCTs, the study investigators, in effect, flip a coin for each participant.

-

The purpose of this randomisation process is to give all participants an equal chance of being allocated to EG or CG so that all groups are similar at the beginning of the study. (reduce confounding)

Measurement allocation - Observational study -

Investigators does not determine how the participants are allocated ‘natural experiment’

-

Participants are allocated into EG and CG by exposing to a factor and take some measurements, the allocation is according to these measurements

-

In observational studies it is important to measure the exposures as accurately as possible, which can often be difficult.

-

allocation measurement errors, participants allocated into wrong groups due to wrong measurements(honesty, embarrassment) well-designed and validated questionnaires or using biological tests

-

In observational studies the exposure and comparison groups are frequently quite different from each other in many respects. We usually try to adjust for these confounders in the analyses. Therefore, it is also important to collect sufficient information about any other differences between EG and CG so it is possible to adjustment for them. - further adjustment to reduce confounders

-

confounding can be caused by errors in how participants are allocated to EG and CG, confounding will be present in almost every observational study. This is due to the fact that, unless participants are randomly allocated to EG and CG, there will almost always be a number of differences between EG and CG that could influence the study outcome

-

The best way to reduce the likelihood of confounding is to conduct an RCT in which participants are randomly allocated to EG and CG. Randomisation is a very effective allocation method for producing two groups (EG and CG) with similar characteristics. If the study is big enough, random allocation will result in similar numbers in EG and CG.

-

RCTs, particularly small ones, randomly allocating participants may not produce groups with similar characteristics, just by chance alone. Therefore it is always important to check for differences between EG and CG at the beginning of a study – this is called a ‘baseline comparison’ and should be done whether the study has allocated participants by randomisation or by measurement.

-

Confounding can still occur in a large RCT if the random allocation process is not done properly. ( surgeon and envelope example) concealment of allocation is usually done by phone, fax or using the internet, which further conceals the allocation decision from the participant and investigator.

-

Studies comparing concealed with unconcealed allocation have been shown to produce quite different results, with unconcealed allocation tending to exaggerate the benefits of the ‘desired’ treatment.

-

While a large RCT with concealed random allocation is the best way to minimise confounding, randomisation is only possible when the exposure intervention investigated is considered to be at least as safe as the comparison intervention(ethical problem allocate people into a smoking group? Make them smoke? Or just to know weather they are already smoking )

Analyses AN - If there were different characteristics in EG and CG, were they adjusted in further analyses?

-

stratified analysis : Confounding can be reduced in the study analyses by dividing participants into, say, older and younger age groups or ‘strata’ (equivalent to dividing the

-

-

-

study participants in the triangle into two circles (older and younger circles) and then analysing the data as if there were two sub-studies). The results of the analyses in the different strata can then be combined, if they give reasonably similar results. If they give very different results, they should be reported separately. This analytical approach is known as  and in the example given here, the stratified analysis adjusts for the confounding caused by allocating more young people to the exposure group (e.g. frequent physical activity) than in the comparison group (infrequent physical activity). This is equivalent to the process of direct age-standardisation, which is commonly done when disease incidence or prevalence in different populations with different age structures are compared. Each population is stratified into comparable age groups (i.e. age strata) and disease incidence or prevalence is calculated for each strata. Then each population’s age structures are standardised (adjusted to be identical to each other) and a standardised disease measure is calculated for the combined age strata. These stratified analyses are also called ‘adjusted’ analyses.

Maintenance - Were most of the participants maintained throughout the study in the groups?

1. maintain their exposure or comparison status throughout the study 2. not be exposed to other factors that could influence the study outcome 3. not drop out of the study -

maintenance error : If some participants’ exposure status changes or some are lost to follow-up

-

In practice participants are seldom perfectly maintained in their allocated groups but as long as any maintenance errors are small and similar in both EG and CG, the error will underestimate the true effect of the exposure on the study outcome(s). This conservative error is usually considered preferable to not knowing whether the error will exaggerate or underestimate the true study effect estimates.

-

The best way to keep the degree of maintenance error similar to EG and CG is to keep the participants and study practitioners ‘blind’ to whether the participant is receiving the study exposure (E) or the comparison exposure (C). This is easier to achieve with drugs than with other interventions like surgery or physiotherapy or diet.

Blind and Objective Measurement

-

measurement errors related to the study Outcomes

-

Were the people who measured the dis-ease outcomes blind to the participants' exposure status? Were these measurements made objectively? (using measurement instruments that were not influenced by subjective factors - human factors)

-

-

Outcome measurement roor can be caused by error from deficiencies of the measurement methods or instruments

-

Measurement of outcomes errors can be reduced in several ways - blinding and objective measurement instrument

-

Knowledge of a participant’s exposure status can influence the participant’s or the practitioner’s perception or interpretation of signs and symptoms of the study outcome.

For example, the results of RCTs of surgery (E) versus physiotherapy (C) for treating reduced range of movement of the knee or pain on movement due to damaged cartilage in the knee joint, can be influenced just by the knowledge of which intervention was used. Participants receiving surgery may report greater improvements in movement and less pain than participants receiving physiotherapy alone, because they may assume that surgery to remove damaged cartilage should be more effective than physiotherapy. As the outcomes being investigated (i.e. range of movement of the knee and pain) are not simple, objective, clear-cut measures (e.g. dead or alive), they are susceptible to measurement error. The practitioner who measures the degree of movement in a participant’s knee and asks about pain may also be influenced by knowledge of the type of treatment received- w  hen the measurement become not objective One way to reduce this problem is to blind the participants or investigators or both to knowledge of which intervention (exposure) participants received. -

Double bind : It is generally not possible to keep information about surgery from participants, there is a famous blinded study of surgery versus physiotherapy for the knee problem described above. To blind the participants, everyone in the study received a local anaesthetic and a small cut in the skin of the knee. However the actual surgery, which used a ‘keyhole’ procedure through the small cut, was only undertaken on participants randomly allocated to the surgical group (EG). The surgeon just pretended to do the surgery on the comparison group (CG). As the procedure was done behind surgical drapes, participants were unable to tell if they had received surgery or not, so they were ‘blind to the exposure.’ The study investigators who m...