«Observational Epidemiological Study Designs» - Free Essay Paper

Introduction

To excel in handling Public health requires one to blend science, convictions and skills aimed at preserving and improving the health of all people, preferably through preventive measures rather than curative measures. Epidemiology is a comprehensive discipline in public health, its literal translation from Greek means “the study of people.” Epidemiology involves studying the frequency at which disease occurs in different groups of the population and why diseases occur. Information gathered from epidemiological studies can be used to evaluate and plan strategies to prevent illnesses and provide guidance in management of patients already infected. Epidemiology combines health sciences, medicine, statistics and social science, among other disciplines in various cases, which make it a varied and interesting profession. Epidemiological studies are classified into two main groups: experimental and observational. This paper only focuses on observational study designs. The discussion will explore various observational epidemiological study designs; their strengths, weaknesses and their suitable application circumstances. Conclusion will summarize the major findings of this discussion.

Background Information and Definitions

The term epidemiology derives from three Greek terms; “epi”, “demos” and “logos”. “Epi” means upon, “demos” translates people, and “logos” translates as “the study.” Many people have defined epidemiology in different ways, but irrespective of the definitions, there are fundamental ideologies that guide or define the public health spirit of epidemiology. Epidemiology can be termed as the study of factors of health related events or states and their circulation in identified populations, as well as the use of the study findings to tackle the health problem at hand (Woodward, 2014).

There have been several definitions put forward to describe epidemiology. Epidemiology has three main purposes. These objectives are describing sickness patterns in societal populations, identifying the causes of sicknesses (etiology), and lastly to deliver data and information needed for illness management. Epidemiology can also be expressed as a method of causative reasoning founded on creating and applying of hypotheses relating to incidence and avoidance of death and sickness. Lastly, epidemiology is an instrument in public health action, which is used to safeguard and uphold the health of the people by using causative reasoning, science, and applied common sense (Fos, 2011).

Under ideal conditions, the findings from and epidemiologic study are sufficient to direct an effective and fast sequence of actions to control and prevent diseases in the population. There are terminologies used in epidemiology that one needs to understand as they represent the essential principals and components of the discipline (Rothman et al., 2008). These components include:

• Study; epidemiology is a based on comprehensive methods of scientific inquiries and research as it is itself a scientific discipline.
• Health-related events or state; epidemiology involves researching rampant contagious and non-infectious sicknesses. These include injuries, chronic diseases, birth defects, environment health, work-related health, and behaviors related to well-being and vigor. Therefore, Health related events refer to a wide range of events pertaining to health and well-being of the people of interest.
• Determinants; these are the factors and causes that contribute to the prevalence of the health related event of interest. Epidemiological studies always seek to establish how and why the health-related events occur. This research is done by examining various groups with different disease levels or rates, and with different demographic features, ecological experiences, immunologic or hereditary make-up, among other potential risk factors.
• Distribution; epidemiology is always concerned with the patterns and frequencies of the health related events in the population of interest. These frequencies include not only the numbers of these events, but also the risks, and rates of disease in the population of interest. Patterns show the occurrence of these health-related events in relation to place, time, and personal characteristics among other factors of interest. Time characteristics refer to; annual, seasonal, weekly, daily or hourly occurrences. Place characteristics refer to geographic location and variations, such as location of a school or work sites. Personal characteristics refer to demographic factors (such as marital status, age, gender, race), as well as socioeconomic status, behaviors (such as risk-taking activities and occupation) which may lead to exposures (Gerstman, 2013).
• Outcomes; these are the results of the occurrences of a health-related event of interest. They may represent a disease, injuries, and syndromes among others, depending on the health-related event of interest.
• Populations; epidemiology is concerned with the control of disease in both large and small communities. Epidemiological studies are interested, and normally cover a specified area or community. For instance, in case of a diarrhea outbreak, epidemiologist will focus on the exposures (causes or actions) that led to the outbreak of the illness in the specific area. The people living in this particular area with the disease occurrence encompass the population of the epidemiology.

Observational Epidemiological Study Designs

Study designs are specific protocols or plans outlined for the purpose of conducting the study. These designs allow the researcher to translate his theoretical hypotheses into operational ones. In observational study designs, the researcher observes and systematically collects data and information without interfering with the subjects being observed. The researcher must not try to change the people, animals, or whatever reagents being observed. In observational epidemiological study designs, there should be strictly no intervention.

Illustrations of Observational Epidemiological Studies:

• A survey on alcohol and substance abuse among students.
• A researcher joining a biker gang to study their way of living.
• Taking blood samples from students every Monday morning to test for alcohol levels.
• Taking blood samples from students to test for HIV and AIDs.

Reasons for Observational Studies:

• When it is not acceptable to perform an experiment, for example, to what extent does using a condom reduce the risk HIV infection?
• Observational studies are done when one is interested in collecting descriptive information. For instance, are the incidences of HIV and AIDs among high school students rising?
• When the researcher wants to determine the cause of a problem (or disease) without interfering with the natural setting.
• Where the researcher cannot perform an experiment, for instance, why do biker gang members suffer from respiratory illnesses?

Observational Designs

There are several observational study designs, the main ones include; cohort studies, case-control studies, cross-sectional studies and ecological (population based) studies. Observational studies can be either population-based or individual-based.

Case-Control Studies

This method is also described as an individual-based observational study. In case-control studies, a collection of persons who have the outcome in question already (cases) are selected from the definite population. Another collection of persons without the outcome in question is then selected from the very population from which the cases were selected to act as a control collection. The frequency of past exposure is then analyzed in relation to the controls and the cases. In cases where the levels of exposure become higher than in controls, the exposure is considered a major risk factor for the infection. In cases where the rates of exposure is high in cases than in controls, this should mean that exposure is a shielding element of the sickness (Carneiro & Howard, 2011).

Case-control studies are reflective studies because they study initiates at the end (outcomes). The researcher begins with the outcomes (the disease) and works backwards to find the likely causes (exposures). In a brain cancer study example, the investigator may hypothesize that prolonged use of cell phones is an exposure factor to brain cancer. The researcher will identify a collection of brain cancer patients for his study (these represent the cases). He then has to identify a collection of people without brain cancer to act as the control. Great carefulness is essential in selecting of the control collection since numerous factors should be considered. The controls should at least have similar living conditions as the cases, in order to minimize the effects of other external factors not considered in the study. The researcher then collects the information on their previous cell phone use, dating back as long as he can manage. Exposure statistics from their phone use history is then analyzed in both the cases and control collections. The hypothesis should state something like: cell phone use will be considerably greater in the cases collection than in the control collections. The information collected is subjected to statistical tests (Aschengrau & Seage, 2008).

Strengths.

• Outcomes and exposures get measured at the same time, and this shows that the studies can be done and completed quickly without having to wait for outcomes. It is, therefore, cheaper as compared to cohort studies.
• Case-control studies provide a good way to study outcomes with long initiation period and rare outcomes because the study initiates with selection of people already with the outcome in question.

Weaknesses.

• Cases and controls may differ on several characteristics such as sex, wealth, age among others. These differences may present significant contribution to the outcomes, yet they may not all get included in the study of potential causes.
• Measuring outcomes and exposures all at once makes it difficult to perceive whether the sickness followed the exposure or vice versa.
• Case-control studies are not suitable or cannot be used to estimate or gather information about prevalence or incidence as the cases are selected in the beginning of the study.
• This study may fail in collecting information on past exposures. There could be other exposures that contributed to the outcome, but the researcher may fail to include them in his hypothesis.

Suitable application.

A study was carried out to examine for the association between severe maternal anemia in the final three months of pregnancy and low birth weights. All low birth weight babies born in a rural hospital during the year 2013 were identified. A representative group of non-low birth weights were selected from the one hospital during the same period. For all the babies (non-low and low weights at births), the hospital antenatal notes were used to identify the hemoglobin levels of the mothers during their final three months of pregnancy.

Cohort Studies

Cohort studies are also known as “prospective” or “longitudinal” studies. These are individual-based observational studies. They follow a certain group of a population over time. Cohort studies are similar to surveys that extend over a long period. Cohort studies start with selecting the population to be studied referred to as the cohort (Aschengrau & Seage, 2008). Facts and figures are then collected to find out which participants in the group have been exposed to the health-related event of interest. The cohorts are then followed up and studied over time. Studying and observing the target subjects over a long period enables the researcher to study the changes and establish the time sequence in which things occur. The researcher can, therefore, study causes. For example, the investigator draws a sample from the target population without the outcome, in this case, from a sample of high school students. The investigator seeks to find out whether cell phone use is a risk factor in brain cancer. The researcher then hypothesizes about the cause and risk factors of the disease in question (brain cancer). Then researcher collects information about his selected sample’s use of cell phones during a long period, for instance, 20 years (Aschengrau & Seage, 2008).

The researcher might get information on how many minutes the subjects spend on calls from their cell phone company bills. He then collects information on who among the subjects gets brain cancer. The researcher then compares the information to see whether people with brain cancer fall in the group of people with long calling times or not. The researcher records the incidences of brain cancer among the subjects who used their cell phones (exposure) for more than a pre-determined amount; then they compare the information to that of brain cancer (outcome) incidences in those who did not use cell phone for long periods. In some cases, the degree of exposure can be identified or even used to classify the subjects into different groups. From this information, the researcher may calculate the relative risk. It is fundamental in cohort studies that the groups being compared are as similar as possible with regard to all factors that relate to the disease of interest (Gordis, 2009).

There are four major steps in conducting a cohort study, as follows:

1. Selecting a group of individuals who not have the outcomes of interest, for instance, diabetes.
2. Following the group of individuals selected in the previous step over time.

• Classifying the different individuals into different groups with regards to their level of exposure. For example, in studying breastfeeding modes, the groups can be classified as exposed (exclusive breastfeeding) and unexposed (mixed feeding).

1. Comparing the rates of disease infections in the exposed group and the unexposed group.

Strengths.

• This study allows for precise collection of exposure data that is collected over time.
• The main strength of cohort studies is that one can be sure that exposure comes before the outcome.
• The researcher has an opportunity to study and collect detailed information on the study population from the start, when they are not infected, to exposure - until the final outcome.
• Examining the disease over time presents a unique opportunity for studying different aspects of the disease in question.
• The information collected during a cohort study on exposures provides an opportunity for studying dose-response relationships (Gordis, 2009).

Weaknesses.

• Some occurrences (such as brain cancer) are rare. It may, therefore, require the researcher to study large groups of cohort sample.
• The first weakness in a cohort study is time consumption. The study takes time; in some cases, years for the outcome to become apparent, which is normally expensive.
• During the period of study, there are losses of information since the study is hard to follow up. Moreover, participants may move away from the area of study for varied reasons, such as discontinuing the study, for example. Such variations and instabilities will create some incomplete cases in the study.
• There is the possibility of information bias when one identifies the exposure to something he already knows as the outcome.
• Cohort studies are not suitable for studying rare outcomes that take a long time to in terms of their development.

Suitable application example.

Cohort studies were applied in a study to determine how decisions regarding feeding infants may affect the mother, the infant, and the household, including possible impacts on future health, demographic, nutritional and economic outcomes. In this study, African women, who gave birth between the 1^st February 1990 and 1^st June 1991, were randomly selected. The selected women, their children and their siblings were followed up until 2014. The research compared the effects of different prenatal and early childhood nutrition and health exposures on later adult outcomes, such as education, work outcomes, and the development of chronic disease risk factor (Silman & Macfarlane, 2002).

Cross-Sectional Studies

This is the simplest study design, and it mainly deals with measuring the frequency of particular exposures and outcomes in a defined target population at a time. In cross-sectional studies, a sample of individuals is randomly selected from a previously determined population. The sample is examined at a particular point for both outcomes and exposures of interest (Gerstman, 2013). These studies are useful in determining the prevalence of a disease of interest of reasonably long period. Cross-sectional studies can be descriptive or analytical. Descriptive study describes the frequency of exposures and outcomes in a defined population. For instance, the study might deal with describing the prevalence of diabetes in a population living in the ghetto. The study might report the findings as follows: 42 per cent of women and 39 percent of men living in the ghetto are diabetic. Analytical study collects and documents information about exposures and outcomes simultaneously (Silman & Macfarlane, 2002). The frequency of the outcome in the population exposed to the risk factor is compared with in the frequency in those that are not exposed. For example, in a study to establish if body weight affects the prevalence of diabetes in a population living in the ghetto. The results of the study might be reported as follows: 68 percent of overweight and obese people living in the ghetto are at risk of becoming diabetic, whereas 32 percent of this population is not at risk of becoming diabetic.

In cross section studies, the researcher uses a random sample of a population to record data about their health in a systematic way. For example, the researcher could be looking to determine the prevalence rate of diabetes in a certain community. He will record information from the selected sample on areas such as how many people are diabetic and how many are not. He may also compare diabetes prevalence in the population with characteristics such as body weight, and the possible association of body weight with the disease.

Strengths.

• Cross sectional studies are particularly useful in defining the health needs of a population at a certain point in time. The study provides important information on the distribution and burden of exposure among other outcomes relatively easily and quickly.
• Cross-sectional surveys are particularly handy in cancer epidemiology in examining the contributing factor and spreading of conjoint conditions that are linked with cancer. These high-risk behaviors include smoking and using a sunbed regularly.
• Cross-sectional studies are relatively simple to conduct and only take a short time to complete because no follow-up is required on the study subjects.

Weaknesses.

• Cross-sectional studies are based on existing or prevalent cases rather than new or incident cases which make the study less useful in investigating causal relationships.
• It is also hard to detect temporality in association and causation. It is, therefore, hard to establish whether the outcome or exposure emanated first since both are examined concurrently in analytical cross-sectional study.

Suitable application.

Cross- sectional studies are suitable for application in cases where the researcher seeks to describe the health status of his target population. For example, the researcher can seek to establish the incidences of diabetes amongst men and women in the age group 40 – 50 years who live in the suburbs.