Research Essay Assessment- Social Epidemiology (Prevalence, Distribution and Social Determinants of Disease)

Part 1b. Can social epidemiology (prevalence, distribution and social determinants of disease) include both quantitative and qualitative methodological approaches?

Traditional epidemiological studies included the use of both quantitative and qualitative study designs.

Epidemiology and qualitative study designs

Qualitative research methods are generally used to collect the data that quantitative methods cannot discover. For example, understanding the act of smoking among the adolescents is more frequent even though a lot of prevention campaigns have been organized to target this specific age group. This result was concluded from an epidemiological cohort of certain adolescents suggesting that the tobacco epidemics do not appear to be under control. Researchers wanted to understand how the information about tobacco and smoking reach the adolescents and they absorb this habit despite constant effort is put into social campaigning. To identify the features which lead to this trend, the investigators carried 30 semi-structured qualitative interviews with adolescents. The main aim was to find out how they perceived their smoking activity even after knowing the ill health implications of smoking. It was observed that they believed it to be a long-term effect that would affect the adulthood and will not have any consequences during their adolescence (Fraga, Ramos, & Barros, 2006).

Another example where the importance of qualitative research in epidemiology can be emphasized is the perception of people about the genome-based knowledge and bioinformatics. A semi-structured interview was conducted with 31 male adult inmates in three prisons of Portugal in 2009. It gave insights into understanding of Portuguese prisoners’ awareness about the methods being used for the identification of suspected individuals using techniques like DNA profiling or DNA fingerprinting (Machado, Santos, & Silva, 2011).

Use of quantitative approach in epidemiological studies

Though qualitative epidemiological approach is gaining popularity among the researchers, quantitative methods are still frequently being used. One such study was conducted to understand the possible mechanism of spread of disease between different locations by the spread of contaminated farm dust. The researchers developed a model which could gauge the quantity of contaminated farm-dust particles which get deposited at various sites down the wind. This information was then used to predict the contribution of the wind-borne route to the transmission of highly pathogenic Avian Influenza virus (HPAI) in Netherland during the 2003 epidemic. This model was generated from a Gaussian Plume Model. It incorporated various factors like dust deposition process, pathogen decay, and a model for the infection process on exposed farms. Comparison between results of transmission risk pattern as obtained from the model and as actually observed during the 2003 epidemic showed that the wind-borne route significantly contributed to the spread of virus over short distances but could not alone account and explain all the observations. It was concluded that only 24% of the transmission was due to wind and that too up to a distance of 25 km (Ssematimba, Hagenaars, & De Jong, 2012).

So, both the approaches have their own strengths and limitations. Thus the choice of appropriate method plays an important role in deciding the fate of the research being conducted.

Part 2a. Describe what you know about observational epidemiology as a research approach and compare it to experimental studies. Describe some of the designs within each, e.g. RCT, case control studies, etc.

Both experimental and observational epidemiological studies aim at examining the underlying causes, methods for preventions, and possible treatments for a disease or an outcome. The only difference between the two approaches is in the way the investigator proceeds with the study. In observational studies, the researcher only perceives the prevailing situation and systematically collects the required information or data. No attempt is made to change or maniplulate the condition of the subjects (animals, humans or reagents) under observation. However, in an experiment the researcher actively mediates and initiates some changes (e.g. gives some treatment to the patient in form of a drug) and then observes the consequences. So there is no intervention in observational studies (Aschengrau & Seage, 2013).

For an example, consider a randomized, controlled trial (RCT) in which the investigator is interested in finding out a relation between smoking and lung cancer. In a RCT, the participating subjects are randomly allocated to different groups- one receiving the treatment being tested or investigated, or to a group receiving standard treatment or no treatment (referred to as control group). The people or animals under each group are then observed to record the changes or consequences. The researcher may randomly sample and form a group of people representing the general population. These people can be made to smoke a fixed number of cigarettes over a certain period of time. There can be another group consisting of people who do not smoke. The researcher may then record the number of subjects who develop lung cancer after a certain number of years. Because of the randomized nature of the experiment the other factors which may control or influence the occurrence of lung cancer can be taken care of. Whereas, in observational study, the researcher can only collect the statistics related to smoking and lung cancer and perform some analysis to correlate these two variables.

There are some ethical issues associated with experiment method. Giving a treatment which is not fully understood so far to a group of people to just check its effect on them may not be ethically correct. Another ethical issue that arises is concerned with the placebo experiment. For the purpose of a study, the control group is kept away from the potential benefits of an intervention or treatment which may save lives. There are some logistics issues as well. For example RCTs are generally more expensive and may take more time to observe the effect of treatment despite of great need.Therefore, many health issues concerning public health interest cannot be studied through experiments and hence observational studies are used.

Part 2b. Why must researchers be cautions to say that variable A cause variables B, even though there is evidence to suggest they have been found to be strongly correlated/associated? What an epidemiologist do in terms of providing evidence that a particular cause may be associated with a specific disease?

The quip, “Correlation does not imply causation”implies to the fact that the event which happen to coincide statistically may not necessarily related casually. The cause or effect on one event due to other can be indirect. It may be due to a third unknown factor (confounding variable), or completelycoincidentalandrandom. Claiming one variable to be a cause of another just on the basis of simple correlation is not at all acceptable. To provecausative relation, there is always a need to perform a well-designed controlled experiment.

For example, in some epidemiological studies a lower incidence of coronary heart diseases (CHD) was observed in women patients taking combined hormone replacement therapy (HRT). This led the doctors to propose that HRT helps in protection against CHD. Further, in randomized controlled trials, HRT was shown to slightly increase the risk of CHD. It was proved to be statistically significant. Then the data from those epidemiological studies was re-analyzed. It was then showed the patients undertaking HRT were more likely from higher socio-economic groups and had access to a better diet and exercise routines. Thus this earlier coincidence was just because of the benefits associated with a better living style rather than a direct cause and effect as it was supposed to be (Lawlor, Smith, & Ebrahim, 2004).

Therefore, a causal relationship between variables can only be established if the correlation is tremendously prominent and there is no possible explanation to challenge this relation.

Epidemiological approach to causality

There are four major types of factors for causation of disease (Bonita, Beaglehole, & Kjellström, 2006):

1. Prejudicing factors, such as age, gender and genetic characteristics.
2. Indirect economic factors like low income, poor nutrition and hygiene and inadequate medical facilities.
3. Exposure to a causative agent of a disease which may trigger the onset of the disease.
4. Fortifying factors involving recurrent exposure and environmental conditions.

Various studies design in epidemiology are used to evidently establish the causal relationships and some of these are as following (Bonita et al., 2006):

1. Experimental studies: A well-designed randomized controlled trials can best prove the relationship between the variables under study. Field and community trials are also used sometimes to study the reason of cause.
2. Cohort and case control studies: Cohort studies are well conducted and involve minimum bias but are not always available. Although, case-control studies are somehow biased, large sample size provides good confirmation for establishing the causal nature.
3. Cross-sectional studies: Questionnaires are generally used to collect the data about past and exposure to potential causes can be recorded.
4. Ecological studies: For the exposures which cannot be measured like the effect of air pollution, water pollution, effect of residual pesticide in food, ecological studies are very important.