The choice of study type is an important part of the design of medical research studies. The study design and subsequent study type greatly influence the scientific quality and clinical importance of a study.
This article describes the structured separation of medical research studies into primary and secondary categories, as well as a further subcategorization of primary studies. This is done on the basis of a chosen literature search relevant to medical research study types, in addition to the authors’ personal experience.
There are three main types of medical research: basic (experimental), clinical, and epidemiological. Clinical and epidemiological research may also be classified as interventional or noninterventional.
In addition to considering available financial resources, personnel, and practical feasibility (organisation, medical requirements, number of patients, etc.), it is critical to select the study type that will allow the researcher to address the specific research topic at hand most effectively.
The efficacy, validity, and chance of publishing a study are all affected by the research design used. The study type must be decided before the inquiry starts since it is part of the research design (see the article “Study Design in Medical Research”). The research kind is determined by the subject to be addressed, which also defines how relevant and readily interpretable a scientific study is. If the wrong study type was selected, it is difficult to modify it after the research has started.
Following on from a previous article that discussed study design components, this article examines study types in primary and secondary research. The article focuses on many types of primary medical research studies. A separate page will discuss the many types of secondary research investigations, such as meta-analyses and reviews. This article examines how various study types are classified. Examples are used to highlight the concept, execution, advantages, limitations, and prospective applications of different types of research. The writers’ experiences, as well as a survey of the relevant literature on different study designs used in medical research, are utilised to write the article.
Methodological Categorization Of Research
Medical research is divided into two broad categories: primary and secondary research. Primary research conducts the actual medical research studies, while secondary research summaries previous results in the form of reviews and meta-analyses. The three principal areas are defined as basic medical research, clinical research, and epidemiological research. In certain cases, assigning particular study to one of these three basic categories or subcategories may be difficult. For the purpose of clarity and avoiding needless length, the writers will avoid discussing specific areas of study such as clinical epidemiology, quality assurance, or health services research. Figure 1 depicts an overview of the many study types utilised in medical research.
Experimental research, often known as basic medical research, includes investigations into the properties of pharmaceuticals and materials, as well as animal experiments, cell studies, biochemical, genetic, and physiological research. Almost all studies investigate the impact of modifying at least one independent variable on the dependent variable. The methodology and experimental design may be properly defined and carried out (1). For example, the population, number of groups, case numbers, treatments, and dosages may all be specified accurately.
Confounding factors must also be adequately accounted for or reduced. Experiments are used to test certain hypotheses and make causal assertions. Standardised experimental conditions with low change in the units of observation (such as cells, animals, or materials) result in high internal validity (= unambiguity). The issue of external validity is more complicated. Because isolated cell or animal processes are not exactly the same as those in humans, laboratory conditions are not always easily transferrable to normal clinical practises (2).
Basic research also includes the development and advancement of analytical techniques, such as the analytical determination of enzymes, markers, or genes, imaging techniques, such as computed tomography or magnetic resonance imaging, and gene sequencing, such as the relationship between eye colour and specific gene sequences. This involves developing biometric procedures such as statistical test methods, modelling, and statistical evaluation approaches.
Clinical research includes both experimental (or interventional) and noninterventional (or observational) studies. A clinical drug study is an interventional clinical study that is defined as “any study performed on man with the purpose of studying or demonstrating the clinical or pharmacological effects of drugs, to establish side effects, or to investigate absorption, distribution, metabolism, or elimination, with the aim of providing clear evidence of the efficacy or safety of the drug.”
Interventional studies include those that investigate medical equipment as well as those that investigate surgical, physical, or psychotherapy procedures. Noninterventional studies, as opposed to clinical studies, are defined in 4 paragraph 23 of the AMG as follows: “A noninterventional study is a study in which knowledge from the treatment of persons with drugs in accordance with the instructions for use specified in their registration is analysed using epidemiological methods; the diagnosis, treatment, and monitoring are not carried out according to a previously specified study protocol, but exclusively according to the medical practice.”
The goal of an interventional clinical study is to evaluate various therapeutic modalities among a patient group that should differ little from one another internally apart from the intervention (4, e1). This is to be accomplished by appropriate actions, primarily through the allocation of patients to groups at random, preventing bias in the outcome. A medication, surgery, the therapeutic use of a medical device such as a stent, or physiotherapy, acupuncture, psychological intervention, or other interventions.
The Medicines Act and the Law on Medical Devices, among other laws and regulations, place restrictions on interventional clinical investigations. The appropriate authorities must register research using medical equipment and must also authorise trials involving pharmaceuticals. Approval from the relevant ethical commission is also required for drug investigations.
Randomization, in which patients are randomly assigned to treatment arms, should ideally be used in clinical research (6-8). Randomization ensures that patients are divided evenly across the groups and that any potential confounding variables, such as risk factors, comorbid conditions, and genetic variations, are distributed randomly among the groups (structural equivalency) (9, 10).
Blinding is another effective strategy for avoiding prejudice. While the patient is unaware of the therapy he is receiving when single blinding is used, neither the patient nor the researcher is aware of the intended treatment when double blinding is used. Blinding the patient and the researcher eliminates any potential subjective (even subconscious) influences on the assessment of a particular treatment.
Case number planning is a critical component of a well-planned clinical trial because it ensures that the expected treatment effect can be recognised with the statistical power (4, 6, 12) that was previously determined.
The success of a clinical trial is dependent on thorough planning and the precise stipulation of all relevant clinical information and procedures in the research protocol (13). It is also critical to keep an eye on how the research is carried out in accordance with the protocol and how data is collected. Data quality must be guaranteed for top-notch research by double data input, programming plausibility checks, and review by a biometrician.
For all of the methodological characteristics mentioned above, as well as for ethical reasons, the randomised controlled and blinded clinical trial with case number planning is considered as the gold standard for assessing the efficacy and safety of therapies or drugs (4, e1, 15).
Contrarily, patients are offered an individually tailored treatment in noninterventional clinical studies (NIS), which are patient-related observational studies. Based on the patient’s desires and the medical diagnosis, the responsible doctor sets the course of treatment. NIS include case series and single case analysis, prognosis studies, observational drug studies, secondary data analyses, and noninterventional therapeutic trials (13, 16). Noninterventional therapy studies, like clinical studies, compare various treatments; however, the choice of treatment is solely at the discretion of the doctor. Frequently, the review is done in the past. Studies on prognosis look at how prognostic variables, including body mass index or tumour stage, may affect how a disease develops over time. Another type of observational study is the diagnostic study, which compares the effectiveness of a diagnostic technique to an established technique (ideally a gold standard), or an investigator to one or more other investigators (inter-rater comparison), or to himself at various time points (intra-rater comparison) (e1). A single-case study or a case series are options if an occurrence is very uncommon (such as a rare illness or a particular course of therapy). A study of a bigger patient population with a particular condition is known as a case series. For instance, the USA’s Centre for Disease Control (CDC) assembled a case series of 1000 individuals after the identification of the AIDS virus in order to research the infection’s common sequelae. A drawback of case series is the absence of a control group. Because of this, case series are often employed to describe things (3).
The main purpose of epidemiological investigations is to investigate the distribution of diseases, their historical trends in frequency, and their causes; similarly to clinical investigations, experimental and observational epidemiological research are differentiated (16, 17).
Group studies (sample from a specific group, such as a certain social or ethnic group) and field studies (sample from an area, such as a large region or a nation) are additional divisions of interventional studies, which are experimental in nature. For example, the study into adding iodine to cooking salt to prevent cretinism in an iodine-deficient area.
Other types of observational epidemiological research include cohort studies (follow-up studies), case control studies, cross-sectional studies (prevalence studies), and ecological studies (correlation studies or studies utilising aggregated data).
In studies that exclusively use descriptive assessment, the frequency (incidence and prevalence) and distribution of a disease within a community may only be shown in basic terms; the frequent recording of information (monitoring, surveillance) may also be the goal of the description. Registry data are suitable for describing prevalence and incidence; for example, registry data are used in German national health reports.
Cohort studies, in their most basic form, entail the long-term monitoring of two healthy subject groups. One group (for instance, employees in a chemical industry) is exposed to a certain material, whilst the other is not. The frequency of a particular condition, such as lung cancer, in the two groups, is noted prospectively (in the future) (figure 2a). It is possible to calculate the disease’s incidence for both populations. Additionally, cohort studies allow for the calculation of the relative risk, which is a crucial statistical measure. The general populace may be utilized as controls for uncommon exposure types (e6). The age and gender distributions in the appropriate cohorts are automatically taken into account in all analyses. Cohort studies’ main goal is to collect comprehensive data on exposure and confounding variables such as job length, maximum exposure, and cumulative exposure. The British physician’s research is a renowned cohort research that looked at the impact of smoking on death among British physicians across a number of decades (e7). Cohort studies are ideal for identifying causal links between exposure and disease development. Cohort studies, on the other hand, can require a significant investment of time, planning, and cash. The so-called historical cohort studies are an exception. In this instance, all exposure and effect (sickness) data were already available before the research began, and they were all analyzed in retrospect. Studies of this kind, for instance, are utilized to look at cancer types that may occur at work. Typically, they cost less (16).
Cases and controls are contrasted in case-control studies. Cases are individuals who get the condition under consideration. Controls are individuals who are healthy but otherwise similar to the cases. To determine the amount to which members of the case and control groups were exposed, a retrospective study is conducted (figure 2b). Pollutant load, diet, and smoking are examples of potential exposure variables. It is important to take care to analyze the exposure’s strength and duration as thoroughly and in-depth as is humanly feasible. If it is shown that sick individuals are exposed more often than healthy ones, it may be assumed that there is a connection between the risk factor and the sickness. The odds ratio is the most crucial statistical factor in case-control studies.
Compared to cohort studies, case-control studies often need less time and money (16). Case-control studies have the drawback that the incidence rate (rate of new cases) cannot be determined. Additionally, there is a high danger of bias due to poor selection of the research population (“selection bias”) and poor recollection (“recall bias”) (also see the article “Avoiding Bias in Observational Studies”). An overview of potential epidemiological research types is provided in Table 1 (e8). The benefits and drawbacks of observational studies are shown in Table 2 (16).
Choosing the appropriate research type is a critical component of study design (for more information, see “Study Design in Medical Research” in issue 11/2009). However, if the research is organized and carried out at a high level of quality, the scientific questions can only be appropriately addressed (e9). Considering or even removing any confounding variables is critical because the outcome cannot be properly understood without them.
The choice of study design for a given main research topic must take into account not only scientific factors but also practical concerns, hospital capacity, and resource (financial) difficulties. Access to registry data is a need for the implementation of epidemiological research. Observational studies should be held to the same standards as experimental investigations in terms of design, execution, and statistical assessment. For the development, execution, and assessment of clinical research, there are especially stringent standards and legally-based rules (such as the Medicines Act and Good Clinical Practice). Both interventional and noninterventional investigations need the preparation of a research protocol (6, 13). The study protocol must include details about the study’s conditions, the question to be answered (both the objective and the subjective), the measurement techniques, the implementation, organization, study population, data management, case number planning, biometric evaluation, and the question’s clinical applicability (13).
Even medical research studies with the best scientific and statistical characteristics may be limited by significant and valid ethical issues. For example, it is impossible and unethical to conduct randomised intervention research under tightly controlled settings to determine the impact of exposure to risky substances (such as smoking, radiation, or a fatty diet). Observational studies, while less trustworthy and difficult to manage than interventional trials, offer a potential substitute.
A peer-reviewed journal should always publish medical research. Depending on the type of research, there are guidelines and checklists for presenting the findings, which could include a summary of the population, instructions on how to handle missing data and confounding variables, and details on statistical parameters. For clinical studies (14, 20, e10, e11), diagnostic studies (21, 22, e12), and epidemiological research (23, e13), recommendations and suggestions.
When determining the research kind and study design for medical investigations, it is essential to collaborate with a trained biometrician; if all critical aspects are prepared simultaneously, the study’s quality and reliability may be greatly raised (12, 25).