Page | 1

Frequency (2) and Risk Research Methods Lecture # 11 Prevalence

Today we will continue the topic Prevalence, but before continuing this topic just commenting on the issue that we raise last time, about the denominator in the calculation of prevalence.. To keep things more clear
Remember : Prevalence is the proportion of a group of people with a clinical condition at a given point in time.

We discussed something important last time, we said if you wont to calculate prevalence at the beginning of 1992 (this example about the incidence of cancer among population studied between 1991 and 1993, slide#8).
-

So we dont have any problem in calculate prevalence here at the beginning of the study, Because simply here we have 100 subjects, those 100 subjects at the beginning 4 of them had cancer so Prevalence = 4/100 (its simple)

But here if we want to calculate the prevalence at this slide which is the beginning of 1992, we have to exclude those who died, For example this case develop cancer and continuo maybe until two months and this case die !! So thats why because this case die, we cant include this case in the denominator (.)

Page | 2

This depends on what we mean by the stopping of the cancer, I mean here if we consider that this case died then of course it shouldnt be incorporator in the denominator. But if we consider that the stopping of the cancer means that the patient was cured. Can we incorporate this patient again in the study? Or do you think its fair to include in the study people who develop cancer and they are cured? I think its not fair or its better not to include them but actually its not a concept!! So ..

If the case die 100% it will not included in the study, so if we mean by stopping the cancer the death for sure we will excluded from the denominator. If we consider that the stopping of the cancer means that the patient was cured .. here we have a question mark !! ( If I am a researcher I dont include people with previous experience of cancer )

But for the EXAM, if we tell you that the case dies dont include it (and there will be no question means that the case was cured)
-

Or sometime the case drop from the study (it wont contributes) also in this case you have to exclude those cases who cant be included in your study. ------------------------------------------------

Incidence
In Incidence study we usually study incidence by cohort study, this means we identify a group of people that have something in common and then we follow this group up for a period of time. In prevalence we start the study by cross sectional study, thats mean we come to group and examinants at once (single point of time) and then we calculate the cases that we want.
Page | 3

Remember: Incidence is the proportion of a group of people initially free of the condition that develops the condition over a given period of time

Thats why we cohort study all people should be the same, but in prevalence study not all the people should be necessarily the same. For example if I want to study cancer in my students ( ) in prevalence study at this single point of time I just calculate those with cancer so this is the nominator ( )then the denominator is the total number of my students . But if I want to study the incidence of cancer in this group of people, I have to determine the beginning of the study which today for example and the study have to end after 5 years. During this period of time I count new cases of cancer, this mean at the beginning of study anybody with cancer should be excluded. Only we should include non-cases, because its not logical at the beginning of my study I have people with cancer. I want to study during period of time to know who develop a cancer, so thats why in my study with incidence I need a group of people with very much similar characteristics. But in prevalence people should not be the same.

And I want to clarify that if its very impossible for your subjects to develop the disease or to develop the outcome, dont include in your study. I gave you an example lets assume we want to study cervical cancer and we have women who underwent hysterectomy its not logical to include those women in my study because they have removed their uterus, thats why we should have a base line examination, we have to exclude cases that are impossible to develop the outcome, another example How can you study caries on a patient that is edentulous?
Page | 4

Suppose I want to study the incidence of caries on maxillary first premolar now I need to exclude the patients who had removed that tooth mostly due to orthodontic treatment. This rule is applicable for both incidence & prevalence, also in prevalence if I want to study prevalence of tuberculated marginal ridge on max. first premolar I have to exclude people who removed that tooth. Again in incidence the denominator must have similar people but in prevalence it isnt necessary for people to be similar but they should have equal chances to get the outcome or not.

Measuring prevalence
Prevalence study or cross- sectional study is the most common study in literature because its very easy to do it because it needs less time, less resources, less statistical analysis and less complication. Example is the research that the doctor did on 3,000 children for the timing of tooth eruption and it only took him 3 months, meanwhile if I want to make a longitudinal study or cohort study I have to follow up with these people for at least 10 years. However the number should be less than that maybe 100 so I can follow them easily plus the cohort study is expensive because I have to follow them like every 3 months for the next 10 years, also it takes a long time so maybe a case will drop out because he/she doesnt want to participate any more.

Measuring incidence
Incidence study or cohort study should have a baseline examination, because we have to exclude the ones with the case who already have the outcome . In prevalence we do baseline examination but in prevalence study the baseline

Page | 5

examination is the examination itself we examine them once those with problems have to be excluded. In the doctor research he excluded those with high BMI (Body Mass Index) too fat or too thin because obesity have some effect on tooth eruption and people with bone disease have to be excluded, also people with orthodontic treatment because all of these affect timing of tooth emergence, however I may include them on a study on their own like the effect of body mass index on timing of tooth eruption. Identifying cases free from the condition, I have to identify people who are free from the condition if we have a case with a condition we have to exclude it, but in prevalence I have to include it, following up to identify new cases with the condition.

Prevalence and incidence Vs. duration

Increasing duration of disease increases the chance that the patient will be included in a prevalence study, if the duration of the disease is long for example some cancers lead to death and others not, for example hepatic pancreatic cancer will lead to death in short period of time, while patients with Hodgkins lymphoma are curable and live for long time. Now if he have long survival rate he has a great percentage in being included in the study because anytime you want to do that study you will find that patient is alive or still with the condition, diseases of short duration are more likely to be missed by a future prevalence study. Crohns disease is a chronic disease so any time you want to make a study you will find that case ready but congestive heart disease (CHD) has short duration due to congestive heart failure. Prevalence is better with diseases of long duration
Page | 6

 Incidence is better with diseases of short duration Caries which is a short duration disease within 1-2 years the study will end but Crohns needs about 20-40 years.

Case fatality rate

Other frequency rate that we have in addition to incidence and prevalence we have what we call "case fatality rate" which is mean the proportion of people with condition who die of it. For example, I follow up 100 cases of women having breast cancer for 5 years and 5 of them died because of breast cancer than 5% will be case fatality rate. Follow up duration should be long enough for potential death to have occurred. Seems not logical to study breast cancer for 3 months only! May be none of those women die during this short period! At least I must study them for at least 5 years to count these cases that have been died due to disease.

Case cure rate

Let us assume that we follow 100 women with breast cancer for 5 years, I count the number of cases who are totally cured from breast cancer so this is case cure rate.

Case complication rate

Those cases who developed complications!

Infant mortality rate

Infant that dies after birth, there is a common syndrome called "infant sudden death syndrome "is the sudden and unexplained
Page | 7

death of an infant who is younger than 1 year old. It's a frightening prospect because it can strike without warning, usually in seemingly healthy babies. Most SIDS deaths are associated with sleep (hence the common reference to "crib death") and infants who die of SIDS show no signs of suffering. I want to study infant mortality rate, I study children from birth to 6 months then I count the cases that have mortality.

Prenatal mortality
The number of cases that havent born yet and died <fetus in uterus >. All of these frequencies are incidence rate, not prevalence rate because I follow them for a specific duration of time.

Slide 16 & 17 NOT included

Interpreting temporal sequence

In incidence study the population should be free of the disease and follow up to measure the development of that disease. BUT in prevalence study, population of existing cases and also non cases of the disease these are studied at once (cross section the time) to measure cases with no reference to when they have been occurred. The problem in the prevalence is that you study the case and you have no idea when the disease has been occurred. Let us assume that we want to study caries in students, I account the students with caries and I found that they were 5% BUT I didnt know when the caries started to develop.

Page | 8

In incidence I have the case and he develop the disease under the study, I can follow up the case and determine exactly when the case start to develop the disease < increase, decrease,> so the incidence give u a better idea about the duration that the disease has taken, also what happens during the development of disease in terms of <complications, remission,>.

Old and new cases

Prevalence studies include old and new cases of disease. Cases included are only the cases available at the time of examination. The cases in the numerator are different between prevalence and incidence studies.

I have the old and new cases, if I want to study incidence, I have to exclude those old cases BUT in the prevalence study the old cases that already have the outcome will be included.

In slide # 20 we have a longitudinal study, if you want to study the incidence some of cases may enter the study after a period of time (after a month for example), another is cured, another left the study, so in the incidence I have a possibility to some people to get in or get out BUT in the prevalence you examine the patients once so those will be present in a point of time exactly.

Determines of disease distribution:

Time Place
Page | 9

Person

Distribution by the time: Epidemic

When the concentration of new cases in a period of time substantially exceeds what is normally expected based on recent experience / research .(locally) Let us assume the level of the flu in Jordanian people at any point of time is 5%. If the flu increased suddenly to 25% or 30% at once in this case we say flu is epidemic. Note that substantially exceeds not from 5% to 10%. It is from 5% to 30% for example.

Pandemic

Disease widespread across a large region such as a whole continent or globally (H1N1 influenza)

Now what are prevalence studies good for?

Cross-sectional studies are the basis of diagnostic testing. Planning health services related to common diseases such as diabetes and hypertension in Jordan.

Let us assume that 30% of Jordanians have diabetes and 40% of the Jordanians have hypertension THEN WHAT!! Ministry of health will improve the services to reduce these percentages.

Slide 24 & 25 NOT included

Page | 10

---------------------------------------------------------------------------

Risk
Risk is the probability of some untoward effect.

Definition
The probability that people who are exposed to certain risk factors will subsequently develop the disease more often than similar unexposed people. For example you want to study the effect of smoking on lung cancer. Here the smoking is the risk factor. Why? Because people who are exposed to this risk factor (smoker) they have higher possibility to have cancer than other. - Factors associated with an increased risk of becoming diseased these are called Risk factor. This lecture describes how estimates of risk are obtained by observing the relationship between exposure to possible risk and the subsequent development of the disease: Sometimes we backwards. look forwards and sometimes we look

Sometime in the study I begin with group of people then the study over period of time in this case I go forward called prospective study () Sometime in my study I can go backward: I go to the record or files of the patient and this called retrospective study ( )for example I go back to the files of patients in the years from 2000 to 2010 and I see who was a smoker and developed lung cancer.
Page | 11

But the problem in retrospective studies are not very efficient in people with bad filing system. Like here in Jordan unfortunately we dont have good filing system a lot of time the files are missing or the doctor doesnt describe the case very well in the file so in developing countries we cant rely on the retrospective study but in developed countries we can efficiently depend on retrospective study.

Risk factors can be:

Physical environment factors: Toxin, infectious agents, gas, pollutants. Social environment factors: Emotional illness, stress, loss of family members, culture. Behavioral factors: Smoking, driving without seat belts, inactivity. Inherited factors: Diabetes, cholesterol, triglyceride.

Exposure to risk factors

-The exposed person:
Has come in contact with risk factor Or has manifested the factor in question

You must study him Before becoming ill

-Duration of exposure:
-The risk factor affect at a single point in time, example nuclear bomb in Hiroshima affect people at that time.
Page | 12

- Or over a long period of time likes the effect of smoking.

Amount of exposure
We ask relevant questions:

Have you ever been exposed? What is the current dose? Largest dose taken. Total cumulative dose. Years of exposure. Years since first exposure.

These questions we ask them about any risk factor. In Dentistry the risk factor that we most commonly ask about is smoking. We ask the patient when you started smoking , how many times do you smoke in a day? And so on .. The Doctor skipped a group of slides .. and then :

Cohort study
The other synonyms of the cohort study are: Incidence studies, longitudinal studies, Prospective studies. Cohort: People assembled have not experienced the outcome (didnt have the disease before) but have equal susceptibility to develop the outcome. Thats why you cant include women with hysterectomy (removal of the uterus) in study the incidence of the cervical cancer. So all the assembled people must have equal susceptibility to develop the disease but no one have the disease in the beginning of the study.

Page | 13

People are then observed over a period of time and we examine which people experience the outcome. -People at risk: -Exposed to risk factor: yes or no -Not exposed to risk factor: yes or no Thats why in Cohort study you can include a control group. you choose people exposed to risk factor like smoker and nonsmoker and study them over long period of time, and you see who develop the disease to know if the risk factor strongly associated with the disease or not. Suppose I study smokers and nonsmokers people for 10 years and I found that the risk of lung cancer is 10 % in both groups so in this case I can say that smoking have effect on the lung cancer thats why sometimes I put people in control group and other in case group.

Types of cohort studies

Concurrent (prospective) Historical (retrospective) In slide # 33 if you study from the past to the present this is historical cohort study. If you study from present to the future this is concurrent cohort study. The last three slides (advantages & disadvantages of cohort study & the table which is a summary) I will not talk about it but they are required for the exam.

Good luck dentist

Done by : Alaa Khalaf & Sanaa Qasem
Page | 14

Big thanks go to : Ruba Ghanim and Noor Bdeir

Page | 15