Describe your study population and the study setting (as detailed as relevant). The level of detail should be based on the notion to which patients your study results can be generalized (gender, age, disease, type of healthcare instution etc.).
The study determinant is the variable(s) under study. This can be a (set of) patient characteristic(s), diagnostic test or treatment, dependent on your research type. If applicable, also think about the variables that you adjust for, reference diagnostic test or comparative treatment.
Describe the primary outcome under study as specifically as possible. Include the time of follow-up, if applicable. Avoid non-specific descriptors such as diagnostic value. Secondary study outcomes do not need to be included in the research question.
Here's an example:
Does apalutamide [determinant] increase metastasis-free survival [outcome] in patients with nonmetastatic castration resistant prostate cancer [domain] compared to placebo?
The aim of etiological research is to determine causes/risk factors for disease. Typically a cohort study is used to explore patient characteristics, habits or environmental factors that may cause disease, making sure to adjust for confounding variables. Characteristics: Observational (non-interventional), causal, follow-up. A special type of etiological research is the nested case-control study. Etiological research is closely related to therapeutic research, since both types of research are causal. The difference is that etiological research is observational and not interventional.
The aim of diagnostic research is to determine the ability of a diagnostic test to estimate disease probability. Typically, a cross-sectional cohort study is used to explore whether a diagnostic test is able to predict whether patients have/don’t have the disease under study at the time of testing. Characteristics: Observational (non-interventional), predictive, cross-sectional. Well-designed studies look at the added value of a diagnostic test in relation to known information (baseline characteristics and other investigations). Diagnostic research is closely related to prognostic research, since both types of research are prediction research. However, diagnostic research is cross-sectional, i.e. tries to predict the presence or absence of disease at the time of testing. In prognostic research, time of follow-up is an important element.
The aim of prognostic research is to determine the ability of (a set of) variable(s) to predict a certain outcome. Typically, a cohort study is used to explore whether a set of variables is able to predict whether patients have a certain outcome after follow-up, for example survival after 5 years. Characteristics: Observational (non-interventional), predictive, follow-up. A special type of prognostic research is survival studies (time-to-event analyses). Prognostic research is closely related to ethiological and diagnostic research. The difference with ethiological research is that a causal relationship does not need to proved. Instead, the goal is to select the best predictors for the studied outcome. The difference with diagnostic research is that time of follow-up is an important element.
The aim of intervention research is to determine the effects of an intervention. Usually a type of treatment is compared with another, as for example in a randomized controlled trial. Characteristics: Interventional (non-observational), causal, comparative, follow-up Special types of intervention research include before-after studies (using a within-patient control), factorial RCT’s (comparison of multiple interventions, non-inferiority trials and propensity-matched studies. Note: A different kind of research can be conducted on a therapeutic topic. For example, a prognostic study in a cohort of patients treated with anti-hypertensive drugs.
The aim of reliability and agreement studies is to provide information about the amount of error inherent in any diagnosis, score, or measurement. Typically, the interpretation of diagnostic tests is compared across different observers or across different time points in the same observer, while adjusting for chance, to determine the inter- and intra-observer reliability. Other types of studies measure the equality of different instruments to measure a variable. ‘Reproducibility’ is the umbrella term for reliability and agreement.
Before you get started, do yourself a favour and take care of the following items. This will help you publish your research and gain acceptance in the research community.
Go to the GCR Statistics Academy