Health and Epidemiology Literature and Systematic Reviews

Health: Epidemiology

Question 1

Literature and systematic reviews are quite different in terms of method and value. A literature review is generally conducted to explore a research question or a phenomenon, while a systematic review aims to answer the research question as relevant as possible (Griffith University, 2022). For example, it has specific inclusion and exclusion criteria for the selected articles, which are not seen in a general literature review. The reliability of the variables and results is taken into account, while the systematic review adds to the value of the authenticity of the research. This is not the case in the literature review, which only serves to investigate the previous studies related to the current research, irrespective of the research design.

Systematic reviews are considered more reasonable than literature reviews because they are aimed at a specific phenomenon with a specific consideration of scope. The clarity for the researcher is pre-defined by what he is looking for and how he wants the results to be arranged, looking for only a particular research category/design. Various reviews are included to ensure that selection and population bias are avoided so that several opinions from previous researchers give genuineness to the results.

Question 2

Randomization means when the participants in a study/experiment are selected by equal chance or even assigned to a different group with an equal chance of selection (National Cancer Institute, n.d.-a). It is especially carried out in clinical experiments where the treatment of one group is to be compared with non-treatment or dissimilar treatment for true experimental results. This sampling method strengthens experimental studies since the study remains free of selection and accidental bias and the comparison of two groups becomes valid (Suresh, 2011).

When randomization is conducted in research, no previous knowledge of sample participants is known, so each person has an equal chance of being represented. The selection bias is limited, and the tendency to produce inaccurate results that might create risks for the research’s internal and external validity is eliminated.

Question 3

A major difference between descriptive and analytic studies is that descriptive studies tend to describe the relationship between two variables only (Ranganathan & Aggarwal, 2019). It could involve an occasional population for a selected sample. Analytic studies do the same; however, quantification of variables occurs before reaching the outcome or findings related to detecting the relationship between variables (Ranganathan & Aggarwal, 2019).

A descriptive study is less complex, while an analytic study is more complex and comprises a larger study area (Lakna, 2019). The descriptive study involves case reports and case studies, while true experiments and observational studies are part of analytic studies (Lakna, 2019). Further, the descriptive study analyzes the what, when, and where of the diseases under consideration, whereas the epidemiology studies analyze the why and how of the illness (Lakna, 2019).

Question 4

Population-based controls are deemed more appropriate than hospital-based due to minimized risk of selection bias. It is due to the representativeness that population-based control offers more in comparison with hospital-based (Ruano-Ravina et al., 2008). Certain characteristics are presented in the larger population rather than in hospital-based samples such as for a study that aims to explore tobacco use and the prevalence of lung cancer in low socio-economic class in the entire population rather than a specific hospital only (Ruano-Ravina et al. 2008).

Evidence shows that each study design has advantages and disadvantages (Ruano-Ravina et al., 2008). However, a population-based design is deemed more appropriate due to the representativeness factor. The differential characteristic is a serious consideration when using either of them as demographic factors, and the prevalence of diseases in the population has a larger impact (Ruano-Ravina et al., 2008).

Question 5

Non-differential misclassification is when the probability or risk of all study groups is equal for being misclassified (Catalog of Bias, n.d.). Differential misclassification occurs when the same probability differs in all study groups (Catalog of Bias, n.d.). The misclassification can cause great bias in the results since the group that should belong to a certain category or classification is done incorrectly, creating errors in outcomes of interests. The effect of non-differential misclassification is when certain ratios, like the risk ratio, the odd ratio, etc., move toward the null hypothesis (Alexander et al., n.d.). Also, the effect of differential misclassification is the opposite, showing the results of such ratios moving away from the null hypothesis (Alexander et al., n.d.).

Question 6

It is necessary to understand confounding in research first. Confounding occurs when a mixed exposure effect is observed (Skelly et al., 2012). For example, the given outcome has a mixed effect that is considered an ‘additional’ factor to influence the relationship between variables (Skelly et al., 2012). Similarly, an association between smoking cigarettes and motor vehicle accidents was noticed in a study as an association between alcohol drinking, and accidents were also discerned (Tsai et al., 1990). It was associated subtly with the driving style of smokers, as smokers are more prone to take risks and break traffic rules, resulting in a high incidence of road accidents (Tsai et al., 1990).

Question 7

The case-control studies indicate higher risks that show a bias as they are collected previously (Health Knowledge, n.d.). Moreover, selection bias and the researcher’s knowledge causing bias in the research are also great risks for validating the final results (Health Knowledge, n.d.). A non-comparison factor emerges from this bias as a controlled and case group is impossible. Some characteristics of control groups cannot be studied as exposure is complicated, and study control does not show representativeness of the population, causing considerable bias (Health Knowledge, n.d.).

Question 8

Loss to follow-up occurs when the study’s participants were actively involved at the research time but now have been lost or not located because, for example, several years have passed (LaMorte, 2016a). If the systematic tracking system is absent, it is impossible to follow up with them, and meaningful clinical results are hard to determine (LaMorte, 2016a).

A threat to cohort studies is implied by loss to follow-up since the likelihood of closeness of exposure, and outcomes are not evident (LaMorte, 2016a). Internal validity is missing as comparisons cannot be made between then and now results. The systematic relatedness is missing as bias is obvious.

Question 9

Recall bias is apparent when the participants of an interview, for example, find it hard to recollect information that they delivered in the past (Pandis, 2014). The recall bias causes problems in case-control studies as under or over-reporting of responses is evident, and under or over-reporting of exposures or outcomes is noticeable (Pandis, 2014). Comparing the control group is difficult, and the potential generalized results for an entire population, for instance, joint pain and negative psychological effects, would be challenging (Pandis, 2014).

Question 10

Two ways confounding could be controlled during the study design phase are matching and restriction (LaMorte, 2021). The matching method allows the researchers to control confounding in the design and analysis phases as the matching of two groups is done (LaMorte, 2021). For instance, age and gender groups for smoking in males and females under 30 years could be formulated for b better comparisons. Similarly, in the restriction method, the limitation of confounding factors in the study is ensured (LaMorte, 2021). It could be done by restricting the form of subjects or terms of variables, such as age or gender.

Question 11

A cohort study is one where sample participants are exposed to certain variables over time to which the exposure of interest is measured (Setia, 2016). They are not initially set for the exposure; however, their exposure over time from the beginning phase is measured to evaluate their changes or occurrences of interest (Setia, 2016).

The advantages of cohort studies include a sequential occurrence of exposure of interest with temporal clarity, gauging the occurrence of exposures over time that might not be possible otherwise as rare occurrences are also evident, calculation of risk over time for the exposure group with the inclusion of risk difference, relative rive and attributable risk, etc. (LaMorte, 2016b). The disadvantages include an ongoing measuring of participants and their exposure over a long period that might make the study vulnerable to loss of follow-up risk, making it time and cost-consuming and certainly not favorable for rare diseases (LaMorte, 2016b).

Question 12

A case-control study design incorporates a comparison of two groups; one group undergoes a specific health condition, for example, and the other is the one who has similar symptoms but not the exact condition, the control group (National Cancer Institute, n.d.-b). To see what effect a certain variable has on one group while the same variable has an effect on the other with relevance to particular demographic features, for instance, would be a suitable research design for a case-control study.