types of studies for research

An official website of the United States government

The .gov means it's official. Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you're on a federal government site.

The site is secure. The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

• Publications
• Account settings
• Browse Titles

NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.

InformedHealth.org [Internet]. Cologne, Germany: Institute for Quality and Efficiency in Health Care (IQWiG); 2006-.

InformedHealth.org [Internet].

In brief: what types of studies are there.

Last Update: March 25, 2020 ; Next update: 2024.

There are various types of scientific studies such as experiments and comparative analyses, observational studies, surveys, or interviews. The choice of study type will mainly depend on the research question being asked.

When making decisions, patients and doctors need reliable answers to a number of questions. Depending on the medical condition and patient's personal situation, the following questions may be asked:

• What is the cause of the condition?
• What is the natural course of the disease if left untreated?
• What will change because of the treatment?
• How many other people have the same condition?
• How do other people cope with it?

Each of these questions can best be answered by a different type of study.

In order to get reliable results, a study has to be carefully planned right from the start. One thing that is especially important to consider is which type of study is best suited to the research question. A study protocol should be written and complete documentation of the study's process should also be done. This is vital in order for other scientists to be able to reproduce and check the results afterwards.

The main types of studies are randomized controlled trials (RCTs), cohort studies, case-control studies and qualitative studies.

• Randomized controlled trials

If you want to know how effective a treatment or diagnostic test is, randomized trials provide the most reliable answers. Because the effect of the treatment is often compared with "no treatment" (or a different treatment), they can also show what happens if you opt to not have the treatment or diagnostic test.

When planning this type of study, a research question is stipulated first. This involves deciding what exactly should be tested and in what group of people. In order to be able to reliably assess how effective the treatment is, the following things also need to be determined before the study is started:

• How long the study should last
• How many participants are needed
• How the effect of the treatment should be measured

For instance, a medication used to treat menopause symptoms needs to be tested on a different group of people than a flu medicine. And a study on treatment for a stuffy nose may be much shorter than a study on a drug taken to prevent strokes .

“Randomized” means divided into groups by chance. In RCTs participants are randomly assigned to one of two or more groups. Then one group receives the new drug A, for example, while the other group receives the conventional drug B or a placebo (dummy drug). Things like the appearance and taste of the drug and the placebo should be as similar as possible. Ideally, the assignment to the various groups is done "double blinded," meaning that neither the participants nor their doctors know who is in which group.

The assignment to groups has to be random in order to make sure that only the effects of the medications are compared, and no other factors influence the results. If doctors decided themselves which patients should receive which treatment, they might – for instance – give the more promising drug to patients who have better chances of recovery. This would distort the results. Random allocation ensures that differences between the results of the two groups at the end of the study are actually due to the treatment and not something else.

Randomized controlled trials provide the best results when trying to find out if there is a causal relationship. That means finding out whether a certain effect is actually due to the medication being tested. RCTs can answer questions such as these:

• Is the new drug A better than the standard treatment for medical condition X?
• Does regular physical activity speed up recovery after a slipped disc when compared to passive waiting?
• Cohort studies

A cohort is a group of people who are observed frequently over a period of many years – for instance, to determine how often a certain disease occurs. In a cohort study, two (or more) groups that are exposed to different things are compared with each other: For example, one group might smoke while the other doesn't. Or one group may be exposed to a hazardous substance at work, while the comparison group isn't. The researchers then observe how the health of the people in both groups develops over the course of several years, whether they become ill, and how many of them pass away. Cohort studies often include people who are healthy at the start of the study. Cohort studies can have a prospective (forward-looking) design or a retrospective (backward-looking) design. In a prospective study, the result that the researchers are interested in (such as a specific illness) has not yet occurred by the time the study starts. But the outcomes that they want to measure and other possible influential factors can be precisely defined beforehand. In a retrospective study, the result (the illness) has already occurred before the study starts, and the researchers look at the patient's history to find risk factors.

Cohort studies are especially useful if you want to find out how common a medical condition is and which factors increase the risk of developing it. They can answer questions such as:

• How does high blood pressure affect heart health?
• Does smoking increase your risk of lung cancer?

For example, one famous long-term cohort study observed a group of 40,000 British doctors, many of whom smoked. It tracked how many doctors died over the years, and what they died of. The study showed that smoking caused a lot of deaths, and that people who smoked more were more likely to get ill and die.

• Case-control studies

Case-control studies compare people who have a certain medical condition with people who do not have the medical condition, but who are otherwise as similar as possible, for example in terms of their sex and age. Then the two groups are interviewed, or their medical files are analyzed, to find anything that might be risk factors for the disease. So case-control studies are generally retrospective.

Case-control studies are one way to gain knowledge about rare diseases. They are also not as expensive or time-consuming as RCTs or cohort studies. But it is often difficult to tell which people are the most similar to each other and should therefore be compared with each other. Because the researchers usually ask about past events, they are dependent on the participants’ memories. But the people they interview might no longer remember whether they were, for instance, exposed to certain risk factors in the past.

Still, case-control studies can help to investigate the causes of a specific disease, and answer questions like these:

• Do HPV infections increase the risk of cervical cancer ?
• Is the risk of sudden infant death syndrome (“cot death”) increased by parents smoking at home?

Cohort studies and case-control studies are types of "observational studies."

• Cross-sectional studies

Many people will be familiar with this kind of study. The classic type of cross-sectional study is the survey: A representative group of people – usually a random sample – are interviewed or examined in order to find out their opinions or facts. Because this data is collected only once, cross-sectional studies are relatively quick and inexpensive. They can provide information on things like the prevalence of a particular disease (how common it is). But they can't tell us anything about the cause of a disease or what the best treatment might be.

Cross-sectional studies can answer questions such as these:

• How tall are German men and women at age 20?
• How many people have cancer screening?
• Qualitative studies

This type of study helps us understand, for instance, what it is like for people to live with a certain disease. Unlike other kinds of research, qualitative research does not rely on numbers and data. Instead, it is based on information collected by talking to people who have a particular medical condition and people close to them. Written documents and observations are used too. The information that is obtained is then analyzed and interpreted using a number of methods.

Qualitative studies can answer questions such as these:

• How do women experience a Cesarean section?
• What aspects of treatment are especially important to men who have prostate cancer ?
• How reliable are the different types of studies?

Each type of study has its advantages and disadvantages. It is always important to find out the following: Did the researchers select a study type that will actually allow them to find the answers they are looking for? You can’t use a survey to find out what is causing a particular disease, for instance.

It is really only possible to draw reliable conclusions about cause and effect by using randomized controlled trials. Other types of studies usually only allow us to establish correlations (relationships where it isn’t clear whether one thing is causing the other). For instance, data from a cohort study may show that people who eat more red meat develop bowel cancer more often than people who don't. This might suggest that eating red meat can increase your risk of getting bowel cancer. But people who eat a lot of red meat might also smoke more, drink more alcohol , or tend to be overweight . The influence of these and other possible risk factors can only be determined by comparing two equal-sized groups made up of randomly assigned participants.

That is why randomized controlled trials are usually the only suitable way to find out how effective a treatment is. Systematic reviews, which summarize multiple RCTs , are even better. In order to be good-quality, though, all studies and systematic reviews need to be designed properly and eliminate as many potential sources of error as possible.

• Greenhalgh T. Einführung in die Evidence-based Medicine: kritische Beurteilung klinischer Studien als Basis einer rationalen Medizin. Bern: Huber; 2003.
• Institute for Quality and Efficiency in Health Care (IQWiG, Germany). General methods . Version 5.0. Cologne: IQWiG; 2017.
• Klug SJ, Bender R, Blettner M, Lange S. Wichtige epidemiologische Studientypen . Dtsch Med Wochenschr 2004; 129: T7-T10. [ PubMed : 17530597 ]
• Schäfer T. Kritische Bewertung von Studien zur Ätiologie. In: Kunz R, Ollenschläger G, Raspe H, Jonitz G, Donner-Banzhoff N (Ed). Lehrbuch evidenzbasierte Medizin in Klinik und Praxis. Cologne: Deutscher Ärzte-Verlag; 2007.

IQWiG health information is written with the aim of helping people understand the advantages and disadvantages of the main treatment options and health care services.

Because IQWiG is a German institute, some of the information provided here is specific to the German health care system. The suitability of any of the described options in an individual case can be determined by talking to a doctor. informedhealth.org can provide support for talks with doctors and other medical professionals, but cannot replace them. We do not offer individual consultations.

Our information is based on the results of good-quality studies. It is written by a team of health care professionals, scientists and editors, and reviewed by external experts. You can find a detailed description of how our health information is produced and updated in our methods.

• Cite this Page InformedHealth.org [Internet]. Cologne, Germany: Institute for Quality and Efficiency in Health Care (IQWiG); 2006-. In brief: What types of studies are there? [Updated 2020 Mar 25].

In this Page

Informed health links, related information.

• PubMed Links to PubMed

Recent Activity

• In brief: What types of studies are there? - InformedHealth.org In brief: What types of studies are there? - InformedHealth.org

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

Connect with NLM

National Library of Medicine 8600 Rockville Pike Bethesda, MD 20894

Web Policies FOIA HHS Vulnerability Disclosure

Help Accessibility Careers

Evidence-Based Medicine: Types of Studies

• What is Evidence-Based Practice?
• Question Types and Corresponding Resources
• Types of Studies
• Practice Guidelines
• Step 3: Appraise This link opens in a new window
• Steps 4-5: Apply & Assess

Experimental vs. Observational Studies

An observational study is a study in which the investigator cannot control the assignment of treatment to subjects because the participants or conditions are not directly assigned by the researcher.

• Examines predetermined treatments, interventions, policies, and their effects
• Four main types: case series , case-control studies , cross-sectional studies , and cohort studies

In an experimental study , the investigators directly manipulate or assign participants to different interventions or environments

Experimental studies that involve humans are called clinical trials . They fall into two categories: those with controls, and those without controls.

• Controlled trials - studies in which the experimental drug or procedure is compared with another drug or procedure
• Uncontrolled trials - studies in which the investigators' experience with the experimental drug or procedure is described, but the treatment is not compared with another treatment

Definitions taken from: Dawson B, Trapp R.G. (2004). Chapter 2. Study Designs in Medical Research. In Dawson B, Trapp R.G. (Eds), Basic & Clinical Biostatistics, 4e . Retrieved September 15, 2014 from  https://accessmedicine.mhmedical.com/book.aspx?bookid=2724

Levels of Evidence Pyramid

Levels of Evidence Pyramid created by Andy Puro, September 2014

Additional Study Design Resources

Study Design 101 : Himmelfarb's tutorial on study types and how to find them

Study Designs  (Centre for Evidence Based Medicine, University of Oxford)

Learn about Clinical Studies  (ClinicalTrials.gov, National Institutes of Health)

Study Designs Guide  (Deakin University)

• << Previous: Step 2: Acquire
• Next: Practice Guidelines >>

• Last Updated: Jul 1, 2024 9:31 AM
• URL: https://guides.himmelfarb.gwu.edu/ebm

• Himmelfarb Intranet
• Privacy Notice
• Terms of Use
• GW is committed to digital accessibility. If you experience a barrier that affects your ability to access content on this page, let us know via the Accessibility Feedback Form .
• Himmelfarb Health Sciences Library
• 2300 Eye St., NW, Washington, DC 20037
• Phone: (202) 994-2962
• [email protected]
• https://himmelfarb.gwu.edu

Cookies on this website

We use cookies to ensure that we give you the best experience on our website. If you click 'Accept all cookies' we'll assume that you are happy to receive all cookies and you won't see this message again. If you click 'Reject all non-essential cookies' only necessary cookies providing core functionality such as security, network management, and accessibility will be enabled. Click 'Find out more' for information on how to change your cookie settings.

Study designs

This short article gives a brief guide to the different study types and a comparison of the advantages and disadvantages.

See also  Levels of Evidence  

These study designs all have similar components (as we’d expect from the PICO):

• A defined population (P) from which groups of subjects are studied
• Outcomes (O) that are measured

And for experimental and analytic observational studies:

• Interventions (I) or exposures (E) that are applied to different groups of subjects

Overview of the design tree

Figure 1 shows the tree of possible designs, branching into subgroups of study designs by whether the studies are descriptive or analytic and by whether the analytic studies are experimental or observational. The list is not completely exhaustive but covers most basics designs.

Figure: Tree of different types of studies (Q1, 2, and 3 refer to the three questions below)

> Download a PDF by Jeremy Howick about study designs

Our first distinction is whether the study is analytic or non-analytic. A  non-analytic  or  descriptive  study does not try to quantify the relationship but tries to give us a picture of what is happening in a population, e.g., the prevalence, incidence, or experience of a group. Descriptive studies include case reports, case-series, qualitative studies and surveys (cross-sectional) studies, which measure the frequency of several factors, and hence the size of the problem. They may sometimes also include analytic work (comparing factors “” see below).

An  analytic  study attempts to quantify the relationship between two factors, that is, the effect of an intervention (I) or exposure (E) on an outcome (O). To quantify the effect we will need to know the rate of outcomes in a comparison (C) group as well as the intervention or exposed group. Whether the researcher actively changes a factor or imposes uses an intervention determines whether the study is considered to be observational (passive involvement of researcher), or experimental (active involvement of researcher).

In  experimental  studies, the researcher manipulates the exposure, that is he or she allocates subjects to the intervention or exposure group. Experimental studies, or randomised controlled trials (RCTs), are similar to experiments in other areas of science. That is, subjects are allocated to two or more groups to receive an intervention or exposure and then followed up under carefully controlled conditions. Such studies controlled trials, particularly if randomised and blinded, have the potential to control for most of the biases that can occur in scientific studies but whether this actually occurs depends on the quality of the study design and implementation.

In  analytic observational  studies, the researcher simply measures the exposure or treatments of the groups. Analytical observational studies include case””control studies, cohort studies and some population (cross-sectional) studies. These studies all include matched groups of subjects and assess of associations between exposures and outcomes.

Observational studies investigate and record exposures (such as interventions or risk factors) and observe outcomes (such as disease) as they occur. Such studies may be purely descriptive or more analytical.

We should finally note that studies can incorporate several design elements. For example, a the control arm of a randomised trial may also be used as a cohort study; and the baseline measures of a cohort study may be used as a cross-sectional study.

Spotting the study design

The type of study can generally be worked at by looking at three issues (as per the Tree of design in Figure 1):

Q1. What was the aim of the study?

• To simply describe a population (PO questions) descriptive
• To quantify the relationship between factors (PICO questions) analytic.

Q2. If analytic, was the intervention randomly allocated?

• No? Observational study

For observational study the main types will then depend on the timing of the measurement of outcome, so our third question is:

Q3. When were the outcomes determined?

• Some time after the exposure or intervention? cohort study (‘prospective study’)
• At the same time as the exposure or intervention? cross sectional study or survey
• Before the exposure was determined? case-control study (‘retrospective study’ based on recall of the exposure)

Advantages and Disadvantages of the Designs

Randomised Controlled Trial

An experimental comparison study in which participants are allocated to treatment/intervention or control/placebo groups using a random mechanism (see randomisation). Best for study the effect of an intervention.

Advantages:

• unbiased distribution of confounders;
• blinding more likely;
• randomisation facilitates statistical analysis.

Disadvantages:

• expensive: time and money;
• volunteer bias;
• ethically problematic at times.

Crossover Design

A controlled trial where each study participant has both therapies, e.g, is randomised to treatment A first, at the crossover point they then start treatment B. Only relevant if the outcome is reversible with time, e.g, symptoms.

• all subjects serve as own controls and error variance is reduced thus reducing sample size needed;
• all subjects receive treatment (at least some of the time);
• statistical tests assuming randomisation can be used;
• blinding can be maintained.
• all subjects receive placebo or alternative treatment at some point;
• washout period lengthy or unknown;
• cannot be used for treatments with permanent effects

Cohort Study

Data are obtained from groups who have been exposed, or not exposed, to the new technology or factor of interest (eg from databases). No allocation of exposure is made by the researcher. Best for study the effect of predictive risk factors on an outcome.

• ethically safe;
• subjects can be matched;
• can establish timing and directionality of events;
• eligibility criteria and outcome assessments can be standardised;
• administratively easier and cheaper than RCT.
• controls may be difficult to identify;
• exposure may be linked to a hidden confounder;
• blinding is difficult;
• randomisation not present;
• for rare disease, large sample sizes or long follow-up necessary.

Case-Control Studies

Patients with a certain outcome or disease and an appropriate group of controls without the outcome or disease are selected (usually with careful consideration of appropriate choice of controls, matching, etc) and then information is obtained on whether the subjects have been exposed to the factor under investigation.

• quick and cheap;
• only feasible method for very rare disorders or those with long lag between exposure and outcome;
• fewer subjects needed than cross-sectional studies.
• reliance on recall or records to determine exposure status;
• confounders;
• selection of control groups is difficult;
• potential bias: recall, selection.

Cross-Sectional Survey

A study that examines the relationship between diseases (or other health-related characteristics) and other variables of interest as they exist in a defined population at one particular time (ie exposure and outcomes are both measured at the same time). Best for quantifying the prevalence of a disease or risk factor, and for quantifying the accuracy of a diagnostic test.

• cheap and simple;
• ethically safe.
• establishes association at most, not causality;
• recall bias susceptibility;
• confounders may be unequally distributed;
• Neyman bias;
• group sizes may be unequal.

Community Blog

Keep up-to-date on postgraduate related issues with our quick reads written by students, postdocs, professors and industry leaders.

Types of Research – Explained with Examples

• By DiscoverPhDs
• October 2, 2020

Types of Research

Research is about using established methods to investigate a problem or question in detail with the aim of generating new knowledge about it.

It is a vital tool for scientific advancement because it allows researchers to prove or refute hypotheses based on clearly defined parameters, environments and assumptions. Due to this, it enables us to confidently contribute to knowledge as it allows research to be verified and replicated.

Knowing the types of research and what each of them focuses on will allow you to better plan your project, utilises the most appropriate methodologies and techniques and better communicate your findings to other researchers and supervisors.

Classification of Types of Research

There are various types of research that are classified according to their objective, depth of study, analysed data, time required to study the phenomenon and other factors. It’s important to note that a research project will not be limited to one type of research, but will likely use several.

According to its Purpose

Theoretical research.

Theoretical research, also referred to as pure or basic research, focuses on generating knowledge , regardless of its practical application. Here, data collection is used to generate new general concepts for a better understanding of a particular field or to answer a theoretical research question.

Results of this kind are usually oriented towards the formulation of theories and are usually based on documentary analysis, the development of mathematical formulas and the reflection of high-level researchers.

Applied Research

Here, the goal is to find strategies that can be used to address a specific research problem. Applied research draws on theory to generate practical scientific knowledge, and its use is very common in STEM fields such as engineering, computer science and medicine.

This type of research is subdivided into two types:

• Technological applied research : looks towards improving efficiency in a particular productive sector through the improvement of processes or machinery related to said productive processes.
• Scientific applied research : has predictive purposes. Through this type of research design, we can measure certain variables to predict behaviours useful to the goods and services sector, such as consumption patterns and viability of commercial projects.

According to your Depth of Scope

Exploratory research.

Exploratory research is used for the preliminary investigation of a subject that is not yet well understood or sufficiently researched. It serves to establish a frame of reference and a hypothesis from which an in-depth study can be developed that will enable conclusive results to be generated.

Because exploratory research is based on the study of little-studied phenomena, it relies less on theory and more on the collection of data to identify patterns that explain these phenomena.

Descriptive Research

The primary objective of descriptive research is to define the characteristics of a particular phenomenon without necessarily investigating the causes that produce it.

In this type of research, the researcher must take particular care not to intervene in the observed object or phenomenon, as its behaviour may change if an external factor is involved.

Explanatory Research

Explanatory research is the most common type of research method and is responsible for establishing cause-and-effect relationships that allow generalisations to be extended to similar realities. It is closely related to descriptive research, although it provides additional information about the observed object and its interactions with the environment.

Correlational Research

The purpose of this type of scientific research is to identify the relationship between two or more variables. A correlational study aims to determine whether a variable changes, how much the other elements of the observed system change.

According to the Type of Data Used

Qualitative research.

Qualitative methods are often used in the social sciences to collect, compare and interpret information, has a linguistic-semiotic basis and is used in techniques such as discourse analysis, interviews, surveys, records and participant observations.

In order to use statistical methods to validate their results, the observations collected must be evaluated numerically. Qualitative research, however, tends to be subjective, since not all data can be fully controlled. Therefore, this type of research design is better suited to extracting meaning from an event or phenomenon (the ‘why’) than its cause (the ‘how’).

Quantitative Research

Quantitative research study delves into a phenomena through quantitative data collection and using mathematical, statistical and computer-aided tools to measure them . This allows generalised conclusions to be projected over time.

According to the Degree of Manipulation of Variables

Experimental research.

It is about designing or replicating a phenomenon whose variables are manipulated under strictly controlled conditions in order to identify or discover its effect on another independent variable or object. The phenomenon to be studied is measured through study and control groups, and according to the guidelines of the scientific method.

Non-Experimental Research

Also known as an observational study, it focuses on the analysis of a phenomenon in its natural context. As such, the researcher does not intervene directly, but limits their involvement to measuring the variables required for the study. Due to its observational nature, it is often used in descriptive research.

Quasi-Experimental Research

It controls only some variables of the phenomenon under investigation and is therefore not entirely experimental. In this case, the study and the focus group cannot be randomly selected, but are chosen from existing groups or populations . This is to ensure the collected data is relevant and that the knowledge, perspectives and opinions of the population can be incorporated into the study.

According to the Type of Inference

Deductive investigation.

In this type of research, reality is explained by general laws that point to certain conclusions; conclusions are expected to be part of the premise of the research problem and considered correct if the premise is valid and the inductive method is applied correctly.

Inductive Research

In this type of research, knowledge is generated from an observation to achieve a generalisation. It is based on the collection of specific data to develop new theories.

Hypothetical-Deductive Investigation

It is based on observing reality to make a hypothesis, then use deduction to obtain a conclusion and finally verify or reject it through experience.

According to the Time in Which it is Carried Out

Longitudinal study (also referred to as diachronic research).

It is the monitoring of the same event, individual or group over a defined period of time. It aims to track changes in a number of variables and see how they evolve over time. It is often used in medical, psychological and social areas .

Cross-Sectional Study (also referred to as Synchronous Research)

Cross-sectional research design is used to observe phenomena, an individual or a group of research subjects at a given time.

According to The Sources of Information

Primary research.

This fundamental research type is defined by the fact that the data is collected directly from the source, that is, it consists of primary, first-hand information.

Secondary research

Unlike primary research, secondary research is developed with information from secondary sources, which are generally based on scientific literature and other documents compiled by another researcher.

According to How the Data is Obtained

Documentary (cabinet).

Documentary research, or secondary sources, is based on a systematic review of existing sources of information on a particular subject. This type of scientific research is commonly used when undertaking literature reviews or producing a case study.

Field research study involves the direct collection of information at the location where the observed phenomenon occurs.

From Laboratory

Laboratory research is carried out in a controlled environment in order to isolate a dependent variable and establish its relationship with other variables through scientific methods.

Mixed-Method: Documentary, Field and/or Laboratory

Mixed research methodologies combine results from both secondary (documentary) sources and primary sources through field or laboratory research.

Being a new graduate teaching assistant can be a scary but rewarding undertaking – our 7 tips will help make your teaching journey as smooth as possible.

Learn more about using cloud storage effectively, video conferencing calling, good note-taking solutions and online calendar and task management options.

Learn about defining your workspace, having a list of daily tasks and using technology to stay connected, all whilst working from home as a research student.

Join thousands of other students and stay up to date with the latest PhD programmes, funding opportunities and advice.

Browse PhDs Now

The term research instrument refers to any tool that you may use to collect, measure and analyse research data.

When you should and shouldn’t capitalise the names of chemical compounds and their abbreviations is not always clear.

Ellie is a final year PhD student at the University of Hertfordshire, investigating a protein which is implicated in pancreatic cancer; this work can improve the efficacy of cancer drug treatments.

Chris is a third (and final) year PhD student at Ulster University. His project aims to develop a novel method of delivering antibiofilm compounds directly to an infected wound bed in patients.

Join Thousands of Students

An official website of the United States government

Official websites use .gov A .gov website belongs to an official government organization in the United States.

Secure .gov websites use HTTPS A lock ( Lock Locked padlock icon ) or https:// means you've safely connected to the .gov website. Share sensitive information only on official, secure websites.

• Publications
• Account settings
• Advanced Search
• Journal List

Epidemiology and Clinical Research Design, Part 1: Study Types

Veena manja , md, satyan lakshminrusimha , md.

• Author information
• Copyright and License information

Selecting the best available preventive and therapeutic measures to avoid disability and death is an important goal for all health care practitioners. To achieve this goal, we need to perform studies that determine the value of these measures. In this article, we discuss the possible study designs that can be used for evaluating new approaches to prevention and treatment. The gold standard study design is a randomized, controlled, double-blind trial. In many instances, a randomized controlled trial may not be ethically or practically feasible. Other study types, such as case series, case-control studies, cohort studies, cross-sectional studies, crossover designs, and open-label studies, may be required to hypothesize and evaluate the link between an exposure or predictor variable and an outcome variable. Various study types pertaining to neonatal-perinatal medicine are reviewed in this article.

After completing this article the readers should be able to:

Describe various study designs and their strengths and limitations.

Introduction

This article provides a brief overview of principles of epidemiology and clinical research design and covers all the topics required by the American Board of Pediatrics content outline pertaining to study types (and uses the same alphabetical numbering in the content outline) and systematic reviews. The reader is referred to other review books listed in the Suggested Reading section for a complete understanding of study types and epidemiology. ( 1 )( 2 )( 3 )( 4 )( 5 )

1) Study Types

Clinical trials ( Figure 1 ; also an infographic at http://www.fda.gov/downloads/Drugs/ResourcesForYou/Consumers/UCM284393.pdf ):

Preclinical: The first step in development of a new drug is bench research using tissue cultures or animal models. Information on mechanism of action, efficacy, toxicity, pharmacokinetics, and pharmacodynamics is obtained from these studies.

The drug sponsor (eg, a pharmaceutical company) applies an Investigational New Drug Application to the Food and Drug Administration (FDA). After approval is obtained from the FDA, the sponsor and principal investigator plan human trials.

Phase I trial: This phase emphasizes safety. It typically involves 20 to 80 healthy volunteers. Occasionally, drugs that cause adverse effects may be evaluated in patients with end-stage disease (eg, anticancer drugs). Information on the drug’s most frequent adverse effects, drug metabolism, and excretion are obtained from phase I studies.

Phase II trial: The goal of phase II trials is to obtain preliminary data on whether the drug works in patients who have a certain disease or condition. It typically involves hundreds of patients. In controlled trials, patients receiving the drug are compared with patients receiving a different treatment (usually a placebo or a different drug). Safety continues to be evaluated, and short-term adverse effects are studied.

Phase III trial: Phase III trials typically involve hundreds or thousands of patients and gather more information on safety and efficacy.

New Drug Application review: If the phase III trial is successful, the sponsor applies for an NewDrug Application to the FDA. This process includes a review of the proposed professional labeling and inspection of the manufacturing facility. If the review is favorable, the FDA may approve the drug for marketing.

Phase IV or postmarketing surveillance: This surveillance is performed by the sponsor (typically the manufacturer), who submits periodic safety updates to the FDA. The MedWatch voluntary system enables physicians and consumers to report adverse events. If important risks are uncovered, risks are added to prescribing information, and drug use may be limited, or in rare instances the drug may be withdrawn from the market.

Retrospective study : A retrospective study uses existing data that have been recorded for nonresearch (such as a clinical database) purposes. The baseline measurements and follow-up, including the exposure and the outcome, have all occurred in the past. The investigator starts at the time of exposure and selects a cohort of patients with the exposure and a comparable cohort without the exposure. Available medical records are used to follow up these patients to evaluate the probability of outcome in the cohort with exposure compared with those without. Patients in the 2 groups are matched based on baseline characteristics so that the risk of outcome is the same in the exposed and control groups except for the exposure of interest.

Strengths and limitations of retrospective studies : The main advantages of studies using existing data are speed and cost. A retrospective analysis of an association between caffeine and necrotizing enterocolitis (NEC) can be conducted in a few months using a neonatal intensive care unit (NICU) database with minimal expense. ( 6 )

Answers can be obtained rapidly (speed).

Relatively inexpensive (cost).

Weaknesses:

The investigator has no control over data collected. The existing data may be incomplete, inaccurate, or measured in ways that are not ideal for answering a research question. If the clinician has erroneously diagnosed a case of focal, spontaneous intestinal perforation as NEC in the NICU database, the investigator conducting a retrospective analysis of this database may assume the patient had NEC.

It is difficult to ensure that the exposed and control groups have the same risk of outcome (other than the exposure). In addition to known baseline characteristics for which the exposed and control groups are matched, there may be unknown prognostic factors that are unevenly distributed in the 2 groups contributing to the outcome.

Bias may be due to many factors. For example, selection bias or detection bias may lead to a spurious association between the predictor variable and outcome in the study sample that does not exist in the population.

Confounding identifies real associations in the population, but these associations are not causal in the direction of interest.

Loss to follow-up may influence results.

Case series : Case reports and series are helpful in recognizing and describing new disease processes or rare manifestations and identify emerging health conditions. ( 7 ) A case series of the first 1000 patients with AIDS reported that 727 were homosexual or bisexual males and 236 were injection drug users. ( 8 ) It did not require a formal control group to conclude that these groups were at higher risk. ( 3 )

Strengths and limitations of case series :

Help describe rare manifestations and new diseases.

Identify emerging health conditions.

Better suited to describing the characteristics of the disease than identify causality.

Limitations:

Purely descriptive and considered to be the weakest form of evidence.

Misleading and may suggest a plausible causal relationship where none exists in real population.

Cross-sectional study : The investigator makes all of his/her measurements at a single point in time or within a short period. ( 3 ) For example, the association between systemic blood pressure and postmenstrual age can be evaluated in a cross-sectional study in the NICU.

Strengths and limitations of cross-sectional studies ( 9 ):

Prevalence estimation is the proportion of patients who have a disease or condition at one point in time. The relative prevalence of feeding intolerance among human milk–fed preterm infants can be compared with formula-fed preterm infants.

There is no waiting around for the outcome to occur (fast and inexpensive).

Avoids the problem of loss to follow-up.

A cross-sectional study conducted at the beginning of a cohort or clinical trial provides demographic and clinical characteristics at baseline.

Cannot estimate incidence (the proportion who develop a disease or condition over time).

Difficult to establish causal relationship (risk of possible spurious associations).

Impractical for the study of rare diseases and requires a large sample size.

Case-control study : The investigator works backwards and begins by choosing a sample of patients with the outcome (eg, NEC patients or cases) and another group of individuals without the outcome (controls; Figure 2 ). The outcome variable is chosen first, and the exposure (levels of predictor variable; –eg, human milk vs formula feeding) is evaluated in cases vs controls to see whether there is an association between exposure and outcome.

Strengths and limitations of case-control study ( 10 ):

Relatively inexpensive because of smaller sample sizes.

Because the cases are chosen at the beginning of the study, rare outcomes can be studied using a relatively small sample size (eg, prone vs supine sleeping and sudden infant death syndrome).

Multiple etiologic factors (antibiotic use, histamine 2 -blocker use, and human milk intake) and predictor variables can be studied. Case-control studies are useful for generating hypotheses about the causes of an outcome variable.

Low internal validity (data collection is based on event recall).

High chance of bias because of separate sampling of cases and controls and retrospective measurement of various predictor variables.

Cannot estimate incidence or prevalence.

Only one outcome can be studied, whereas cohort and cross-sectional studies can evaluate multiple outcome variables.

Longitudinal study ( 9 ): In longitudinal studies, a group of individuals are identified at the outset and are followed up over time. Data collection occurs at repeated predetermined intervals. Cohort studies and repeated cross-sectional studies are longitudinal studies.

Strengths and limitations of longitudinal studies :

Determine causality.

Monitor trends.

Time-consuming.

Cohort study : In a prospective cohort study, patients are identified at the beginning of the study based on their exposure status (exposed or not exposed) and followed up for a period from exposure to outcome ( Figure 3 ). ( 11 ) Preterm infants exposed to formula and not exposed to formula (exclusively human milk fed) are followed up over time to detect the prevalence of asthma during the first 10 years (outcome). The exposure is self-selected or determined (mother’s decision and/or availability of human milk or formula feed). This is in contrast to randomized trials in which the investigator randomly allocates the exposure ( Figure 4 ). Cohort studies can be prospective or retrospective (see below).

Strengths and limitations of a prospective cohort study :

Can measure incidence.

Establish temporal association (and strengthens the basis of inferring a causal basis of an association).

Good for common diseases.

Good for rare exposures.

Possible to study associations of an exposure (human milk feeds) with several outcome variables (eg, asthma, neurodevelopmental outcome, and obesity). ( 12 )

Expensive because of large sample size and long follow-up.

Not good for rare diseases.

Confounding factors (factors other than the exposure that might influence the outcome).

A number of potential biases must be taken into account in conducting cohort studies. ( 12 )

Bias in assessment of the outcome: the pediatrician making the diagnosis of asthma in an infant born preterm knows the history of exposure to infant formula and is aware of the hypothesis being tested.

Information bias or recall bias: the quality of information obtained from exposed (formula fed) and unexposed (exclusively breastfed) preterm infants may be different. A difference in educational background of the mother may alter ability to recall and provide accurate information.

Biases from lack of response due to loss to follow-up.

Analytic bias: epidemiologists and statisticians may have a strong preconception that breast milk is protective against asthma.

Retrospective cohort studies ( Figure 3 ): Some of the above limitations, such as duration of follow-up and cost, associated with prospective cohort studies may be reduced by conducting retrospective cohort studies. Instead of prospectively following up preterm infants born in 2014 for 10 years to evaluate risk of developing asthma, investigators may decide to evaluate preterm infants born in 2004. The study design for a cohort remains the same: preterm infants exposed and unexposed to formula based on historical data are combined with outcome (asthma) from present data.

The advantage of the retrospective cohort design is that it is less expensive and fast. The baseline measurements are already made, and the follow-up period is complete.

The main disadvantages are limited control over the approach to sampling and follow up of the population and over the nature and quality of baseline measurements. The existing data may be incomplete, inaccurate, or measured in ways that are not ideal for answering the research question. ( 3 )

Randomized controlled study : In a clinical trial, the investigator applies an intervention and observes the effect on one of more outcomes ( Figure 4 ).

Random assignment is expected to equally allocate all known and unknown predictor variables in the 2 groups, creating 2 groups with similar prognosis on average. The patients are placed into 2 categories: intervention group (receiving the active treatment, eg, donor human milk) and control group (receiving usual care or inactive treatment, eg, formula feeds). This example illustrates the fact that randomization is not possible in all circumstances. It is not ethically and practically possible to randomize breastfeeding and formula feeding among preterm infants.

Blinding (also known as masking, especially in ophthalmologic trials) is the process of ensuring that patients and/or investigators (double-blind) are unaware of the group. In addition to data collectors, outcome adjudicators and data analysts may also be blinded to treatment allocation. Blinding is achieved by use of placebo pills or infusions. Blinding reduces the chance of confounding by postrandomization, co-intervention variables (such as rapid rate of feed volume advancement in donor milk group compared with formula group). In some instances, blinding may not be possible. Two such examples are as follows:

Fetal tracheal occlusion is an intervention to improve lung size in congenital diaphragmatic hernia. However, fetal sham surgery increases the risk of preterm labor and is not ethical. Hence, the control group may not undergo a sham surgery.

In some cases, blinding is not technically possible. In the cooling trial evaluating infants with moderate to severe encephalopathy, allocation is not blinded because the clinical team and parents can easily assess the infant’s temperature.

Inclusion criteria should produce a sufficient number of participants (large sample size of very low birth weight [VLBW] infants) who have a high enough rate of primary outcome (death or neurodevelopmental disability) to achieve adequate power to find a clinically significant effect of the intervention on outcome.

Blocked randomization is a technique to ensure that the number of participants is equally distributed among the study groups. In a study that involves 60 participants, the block size may be 6. Randomization proceeds normally within each block of 6 until 3 VLBW infants are randomized to donor milk or formula group, after which participants are automatically assigned to the other group until the block of 6 is completed. This way the maximum disproportion of allotment of infants to the donor milk and formula groups at a given time will be 3 or less (eg, after enrolling 33 patients, 18:15).

Stratification of participants by a characteristic (such as gestational age or birth weight group, eg, ≤1,000 g vs 1,001–1,500 g) allows investigators to enroll a desired number of participants with a characteristic that may have an influence on the effect of the treatment or its generalizability.

Stratified blocked randomization ensures that an important predictor of the outcome (birth weight and gestational age influence the combined outcome of death and neurodevelopmental disability) is more evenly distributed between the study groups than chance alone would dictate.

Pragmatic clinical trials are randomized controlled trials that are designed to determine the risks, benefits, and costs of an intervention as they would occur in routine clinical practice. They include less restrictive inclusion criteria, a broader range of patients, and many clinical sites (including nonacademic sites) to simulate real-world settings.

Strengths and limitations of randomized controlled trials ( 3 ):

Ability to demonstrate causality—a properly designed randomized, blinded trial can provide the most definitive causal inference of all study designs.

Minimizes the influence of confounding variables.

More than one outcome can be measured —usually a single primary outcome and a few secondary outcomes can be evaluated.

Time-consuming and expensive.

May expose participants to potential harm.

Not every research question is amenable to randomized study design. For example, as mentioned above, the study depicted in Figure 3 cannot be performed because breastfeeding is a personal choice. A similar study can be performed among preterm infants without access to maternal breast milk, randomizing these infants to donor human milk or formula ( Figure 4 ).

To ethically conduct a randomized controlled trial, equipoise must exist between the investigative arm and the control arm. Equipoise is a situation in which it is not known which of the 2 possibilities (eg, donor milk or formula) is more likely to achieve the outcome variable (neurodevelopmental outcome at 2 years).

The patients enrolled in a randomized controlled trial may not fully represent the target population for the intervention.

Before-after study design (historical controls) ( 12 ): If randomization is not possible or will not be used, one possible study design is to compare people who received care before a program was established (donor milk availability for VLBW infants in a NICU) with those who received care after the program or health care measure became available on the outcome variable (incidence of NEC).

Strengths and limitations of before-after studies :

Inexpensive and fast.

Provide suggestion (although not conclusive) demonstrating the effectiveness of a health care intervention.

Data obtained in each of the 2 periods are frequently not comparable in terms of quality or completeness. Often data collected after implementing the program is complete and of research quality. However, data collected before program implementation is from health care records designed and used only for clinical purposes.

Other factors may have changed over time. The NICU may have instituted a bronchopulmonary dysplasia (BPD) prevention bundle or a central catheter–associated bloodstream infection bundle that may have an effect on primary outcome variable (NEC).

Crossover study : Study designs typically are between group (parallel) comparisons ( Figure 4 ) or within group comparison (eg, incidence of feeding intolerance before and after starting fortification). The crossover design has features of within- and between-group designs. In addition to planned crossover design, patients may also cross over in parallel design trials (unplanned crossover).

Unplanned crossover and intention to treat ( 12 ): Figure 5 shows the design of a randomized trial of Nissen fundoplication, comparing it with medical therapy for gastroesophageal reflux disease. Randomization is performed after informed consent is obtained. Some parents of infants assigned to Nissen fundoplication may have second thoughts after randomization and refuse surgery. In addition, some patients assigned to medical therapy may deteriorate and require Nissen fundoplication. This is called unplanned crossover. Ideally, analysis is performed on the basis of intent to treat: patients assigned to the surgical arm at randomization are analyzed as the surgical group (although some patients did not get surgery), and patients randomized to medical arm are analyzed as the medical group (although some patients had surgery).

Planned crossover study ( Figure 6 ): Preterm VLBW infants receiving full expressed breast milk feeds are randomized to standard fortification to 24 kcal/oz with liquid human milk fortifier (group A) or fortification with medium chain triglyceride (MCT) oil to same caloric density (group B). The primary outcome variable is daily weight gain. After 1 week, fortification is stopped in both groups for a week (washout period so that carryover effects of one fortifier to the other group are avoided). After 1 week of the washout period, group B is fortified with liquid fortifier, and group A is fortified with MCT oil (crossover period). Weight gain is compared between groups (A vs B) and also within each group ( Figure 6 ).

Strengths and limitations of crossover studies :

Planned crossover studies are attractive and useful because they evaluate within-group and between-group comparisons ( Figure 6 ). Crossover design removes between-patient variation and requires fewer patients.

Limitations of planned crossover:

Valid only if there is no residual carryover from the first therapy.

Not applicable if the therapy is surgical or if it cures the disease.

Washout period may deprive the patient of a useful therapy (such as fortification).

Order in which therapy is given may elicit psychological responses and difference in physiologic maturity with increasing postnatal age may influence response.

Open-label study is a type of clinical trial in which the researchers and participants (or parents) know which treatment is being administered. This contrasts with single-blind and double-blind designs. An open-label study may still be randomized.

Strengths and limitations of open-label studies :

Strengths: A blinded trial is not possible in certain circumstances involving surgery (abdominal drain vs laparotomy for NEC or intestinal perforation) or physical intervention (optimizing cooling trial for hypoxic ischemic encephalopathy [randomization to 33.5°C or 32°C for 72 or 120 hours]; – duration of therapy and infant’s body temperature are known to caregivers, investigators, and parents).

Limitations: A blinded trial is regarded as being less subject to bias than an open trial because it minimizes the effect of knowledge of treatment allocation on reporting of outcomes.

Post hoc analysis (from Latin post hoc meaning “after this”): Post hoc analysis consists of looking at the data after the experiment has concluded for patterns that were not specified a priori. If the hypothesis was formulated after the data were analyzed, it is known as a post hoc hypothesis. Because spurious associations may be present just by chance, post hoc analysis should only be hypothesis generating and should be tested in future trials to confirm the effect seen.

Strengths and limitations of post hoc analysis :

Strengths: A clinically relevant association may be detected during post hoc analysis. In a study evaluating inhaled nitric oxide in pulmonary hypertension, a finding associating respiratory alkalosis with later-onset sensorineural deafness may be detected by post hoc analysis. As noted above, however, this hypothesis ideally should be tested in a future trial before an association can be confirmed. However, such a trial may not be ethically appropriate.

Limitations: In addition to caution regarding interpretation of a finding on post hoc analysis, if the number of analyses increases, some positive results may be due to chance. As the number of hypotheses increase, the α value should be adjusted (Bonferroni correction). So, if 2 hypotheses are being tested in the same sample, the α to assign significance may be .025. The risk of erroneous conclusion increases with post hoc analysis.

Subgroup analyses are defined as comparisons between randomized groups in a subset of the trial cohort. The main reason for performing these analyses is to discover effect modification (interaction) in subgroups, for example, whether inhaled nitric oxide is more effective in reducing the incidence of death or BPD in preterm infants with birth weights greater than 1,000 g compared with infants with birth weights of 1,000 g or less. To preserve the value of randomization, subgroups should be defined by measurements that were made before randomization. The subgroup effect should be one of a small number of hypotheses tested. If a large number of hypotheses are tested, some of the statistically significant findings may be due to chance alone.

Strengths and limitations of subgroup analysis :

To discover whether the effect of treatment is different based on sex, gestational age, or birth weight.

Easy to misuse and can lead to wrong conclusions.

Being smaller than the entire trial population, there may not be sufficient power to find important differences. ( 3 )

Systematic review and meta-analysis :

A systematic review is the assembly, critical appraisal, and synthesis of all relevant studies addressing a clearly formulated clinical question and incorporating strategies to minimize bias. A comprehensive search of the literature is performed to identify, select, and critically evaluate all relevant research using systematic and explicit methods and to collect and analyze data from all relevant studies included in the review. Statistical methods (meta-analysis) may or may not be used to combine the data in the included studies. Systematic reviews are retrospective analyses and, similar to other retrospective research, are subject to bias. To limit bias, the clinical question should be clearly stated and explicit and systematic methods that can be reproduced by others should be used throughout the process. The inclusion and exclusion criteria for the search and selection of studies, the subgroup analyses planned (including the anticipated direction of subgroup effect), and sensitivity analyses planned should all be specified a priori.

The purpose of a systematic review: Systematic reviews summarize the available evidence relating to a specific clinical question. In addition to providing an overall estimate of treatment effect to guide clinical decisions, systematic reviews can also help to inform research by identifying the areas of uncertainty requiring further study and guide policy decisions based on the entire body of evidence.

Advantages of adding a meta-analysis to a systematic review and interpreting the results of a meta-analysis.

A meta-analysis is a statistical method to pool effect estimates of individual studies and provide an overall estimate of treatment effect. The results of a meta-analysis are presented in a forest plot ( Figure 7 ). The forest plot has a horizontal scale that displays the possible values of treatment effect. If the effect estimate is a ratio (eg, odds ratio or risk ratio), the scale is logarithmic. Alternatively, if the effect size is presented as a difference (eg, risk difference or difference in means), the scale is linear. On the logarithmic scale, the value of 1 lies on the line of no effect, whereas on the linear scale, the value of 0 lies on the line of no effect. The vertical line in a forest plot is called the line of no effect, and if the confidence interval (CI) crosses this line, it indicates no difference in outcomes in the treatment and comparator arms.

On a forest plot, the treatment effect of each individual study is represented by a square. The point at the center of the square is the point estimate. The relative weight given to each study is represented by the size of the square (usually determined by the study size), the CI is represented by a horizontal line that runs through the center of the block. The pooled effect estimate is represented by a diamond at the end of all individual study estimates. The point estimate of the pooled effect is at the center of the diamond, and the CI is represented by the width of the diamond. If the diamond crosses the line of no effect, this indicates that after combining all relevant studies, there is no significant difference in effect in the treatment vs comparator.

It is not always appropriate to pool results of individual studies in a meta-analysis. Data from different studies can be combined when the studies address a common question and measure and report outcomes similarly. Alternatively, if the studies are very dissimilar or if the studies are at a high risk of bias, leading to a low confidence in the estimate of effect, the data should not be statistically pooled.

The advantages of adding a meta-analysis to a systematic review can be summarized in the following points:

The combination provides a pooled quantitative estimate of the effects of the intervention, and the uncertainty associated with the result can be inferred using the CI. Combining studies results in greater power to detect treatment effects and decreases uncertainty (narrower CI and greater precision).

The differences in included studies can be analyzed to explore differences in treatment effect in different study populations and settings.

The comparison of different studies may lead to new ideas and hypotheses for future trials.

Limitations of a systematic review :

Because a systematic review is a retrospective review, similar to other retrospective studies, it is at risk of bias.

Use of explicit criteria and critical appraisal of the literature reduce the likelihood of a biased review.

Not recognizing publication bias and bias in the conduct of the studies included in the review may lead to unreliable results. A thorough evaluation of the risk of bias of included studies and assessment of publication bias will limit this possibility. Publication bias is a distortion of the published literature that occurs when published studies are not representative of all studies that have been performed. This bias is secondary to a tendency to submit and publish positive results more often than negative results. ( 3 )

Limitations of a meta-analysis :

Careful consideration should be given to the studies that are included in the meta-analysis. The results are a direct reflection of the studies included in the analysis. If the studies included are at a high risk of bias, and the results of the individual studies do not represent the true effect, combining these studies may result in increased precision of the wrong results, giving biased results credibility (garbage in, garbage out). Analyzing the data by different statistical methods may give different results with the same set of studies. Heterogeneity should be considered and explored in the results of the meta-analysis.

Distinguishing among preclinical and phase I, II, III, and IV clinical trials.

IND = Investigational New Drug Application, FDA = Food and Drug Administration, NDA = New Drug Application, PI = principal investigator.

Case-control study. Cases (patients with necrotizing enterocolitis [NEC]) are identified and are matched with controls (eg, preterm infants of the same gestational age and sex born in the same month in the same neonatal intensive care unit [NICU]). Working backwards, investigators can evaluate multiple predictor variables.

PDA = patent ductus arteriosus, PICC = peripherally inserted central catheter, SGA = small for gestational age, TPN = total parenteral nutrition.

Cohort study. Preterm infants exposed to formula and not exposed to formula born in 2014 form the cohort for a prospective cohort study. They are followed up longitudinally for development of asthma by age 10 years. In a retrospective cohort study, preterm infants born in 2004 are evaluated over time to the current presence of disease (2014). Confounding factors (variables other than the one of interest, eg, feeding) may influence the primary outcome (eg, family history of atopy, socioeconomic status, and severity of bronchopulmonary dysplasia [BPD]). Losses to follow-up may bias the results of a cohort study.

Randomized controlled and double-blinded clinical trial (modified from the Eunice Kennedy Shriver National Institute of Child Health and Human Development Donor Human Milk and Neurodevelopmental Outcomes in Very Low Birth Weight [VLBW] Infants trial; http://clinicaltrials.gov/ct2/show/NCT01232725 ). Access to maternal breast milk (MBM) cannot be randomized. Hence, infants with access to MBM are excluded from the trial. Preterm infants without access to MBM form the study population and are randomized to donor human milk and formula. Because the color of these 2 products is different, it is delivered in amber syringes and bottles to blind the caregivers and parents. A double-blind approach indicates that both investigators or clinical team and parents or patients are unaware of the allocation. The primary outcome variable is neurodevelopmental outcome at age 2 years. Patients may be lost to follow-up. A primary outcome and several secondary outcomes may be evaluated. A blinded randomized controlled trial is the gold standard and provides strong direct evidence between exposure and outcome but is time-consuming and expensive.

Unplanned crossover study in a randomized study of Nissen fundoplication vs lansoprazole medical management. Analysis by treatment would analyze patients who underwent Nissen fundoplication vs patients who were treated with lansoprazole alone. However, intent-to-treat analysis would compare patients randomized to fundoplication (all infants with light yellow onesies) to patients randomized to lansoprazole only (all infants with dark red onesies).

Planned crossover design. Preterm infants (<1,500-g birth weight [very low birth weight (VLBW)]) on exclusive maternal breast milk feeds were randomized to standard fortification to 24 kcal/oz and experimental fortification with medium chain triglyceride (MCT) oil for 1 week. At the end of the week, outcome variable (mean weight gain per day) was measured and compared between the 2 groups. After a washout period of 1 week to avoid residual carryover effect, group A was fortified with MCT oil and group B was fortified with liquid human milk fortifier. Within-group comparisons can be performed between 2 fortifications.

Characteristics of a forest plot. This plot can be shown on a logarithmic scale (when the line of no effect is at 1) or on a linear scale (when the line of no effect is at 0). The squares represent individual treatment effects, and the diamond represents pooled effect. The width of the diamond represents the 95% confidence interval. If the diamond crosses the line of no effect, the meta-analysis does not significantly favor experimental or control group.

The second part of this review covers bias and confounding, causation, incidence and prevalence, screening, sensitivity analysis, and measurement and will appear in a subsequent issue of Neoreviews .

American Board of Pediatrics Neonatal-Perinatal Content Specification.

Understand the purpose of a systematic review.

Understand the advantages of adding a meta-analysis to a systematic review.

Interpret the results of a meta-analysis.

Identify the limitations of a systematic review.

Identify the limitations of a meta-analysis.

Distinguish phase I, II, III, and IV clinical trials.

Recognize a retrospective study.

Understand the strengths and limitations of retrospective studies.

Recognize a case series.

Understand the strengths and limitations of case series.

Recognize a cross-sectional study.

Understand the strengths and limitations of cross-sectional studies.

Recognize a case-control study.

Understand the strengths and limitations of case-control studies.

Recognize a longitudinal study.

Understand the strengths and limitations of longitudinal studies.

Recognize a cohort study.

Understand the strengths and limitations of cohort studies.

Recognize a randomized controlled study.

Understand the strengths and limitations of randomized controlled studies.

Recognize a before-after study.

Understand the strengths and limitations of before-after studies.

Recognize a crossover study.

Understand the strengths and limitations of crossover studies.

Recognize an open-label study.

Understand the strengths and limitations of open-label studies.

Recognize a post-hoc analysis.

Understand the strengths and limitations of post hoc analyses.

Recognize a subgroup analysis.

Understand the strengths and limitations of subgroup analyses.

Abbreviations

bronchopulmonary dysplasia

confidence interval

medium chain triglyceride

necrotizing enterocolitis

neonatal intensive care unit

very low birth weight

Author Disclosure

Drs Manja and Lakshminrusimha have disclosed funding from grant 1R01HD072929-0 (S.L.) and that they are consultants and on the speaker bureau of Ikaria, Inc. This commentary does not contain a discussion of an unapproved/ investigative use of a commercial product/device.

• 1. Brodsky D, Martin C. Neonatology Review. Philadelphia, PA: Hanley & Belfus; 2003. [ Google Scholar ]
• 2. Hermansen M. Biostatistics: Some Basic Concepts. Gainesville, FL: Caduceus Medical Publishers; 1990. [ Google Scholar ]
• 3. Hulley SB, Cummings SR, Browner WS, Grady DG, Newman TB. Designing Clinical Research. Philadelphia, PA: Wolters Kluwer Health; 2013. [ Google Scholar ]
• 4. Norman GR, Streiner DL. Biostatistics: The Bare Essentials. Hamilton, Ontario: BC Decker; 2008. [ Google Scholar ]
• 5. Pagano M, Gauvreau K. Principles of Biostatistics. 2. Boston, MA: Duxbury/Thomson Learning; 2000. [ Google Scholar ]
• 6. Lampkin SJ, Turner AM, Lakshminrusimha S, et al. Association between caffeine citrate exposure and necrotizing enterocolitis in preterm infants. Am J Health Syst Pharm. 2013;70(7):603–608. doi: 10.2146/ajhp120457. [ DOI ] [ PubMed ] [ Google Scholar ]
• 7. Moore EM, Johnson SB. Research and statistics: case reports, anecdotal evidence, and descriptive epidemiologic studies in pediatric practice. Pediatr Rev. 2009;30(8):323–324. doi: 10.1542/pir.30-8-323. [ DOI ] [ PubMed ] [ Google Scholar ]
• 8. Jaffe HW, Bregman DJ, Selik RM. Acquired immune deficiency syndrome in the United States: the first 1,000 cases. J Infect Dis. 1983;148(2):339–345. doi: 10.1093/infdis/148.2.339. [ DOI ] [ PubMed ] [ Google Scholar ]
• 9. Johnson SL. Research and statistics: a question of time: crosssectional versus longitudinal study designs. Pediatr Rev. 2010;31 (6):250–251. doi: 10.1542/pir.31-6-250. [ DOI ] [ PubMed ] [ Google Scholar ]
• 10. Upadhya K, Rowe P. Research and statistics: case-control studies. Pediatr Rev. 2010;31(2):70–71. doi: 10.1542/pir.31-2-70. [ DOI ] [ PubMed ] [ Google Scholar ]
• 11. Hernandez RG, Rowe PC. Research and statistics: cohort studies. Acad Pediatr Rev. 2009;30(9):364–365. doi: 10.1542/pir.30-9-364. [ DOI ] [ PubMed ] [ Google Scholar ]
• 12. Gordis L. Epidemiology: With STUDENT CONSULT Online Access. Philadelphia, PA: Elsevier Health Sciences; 2013. [ Google Scholar ]
• View on publisher site
• PDF (1.7 MB)
• Collections

Similar articles

Cited by other articles, links to ncbi databases.

• Download .nbib .nbib
• Format: AMA APA MLA NLM

Add to Collections

• Access Provided by: Googlebot
• Sign in or Create a Free Access Profile
• Remote Access
• Save figures into PowerPoint
• Download tables as PDFs

Chapter 3:  Types of Studies in Clinical Research—Part I: Observational Studies

• Download Chapter PDF

Disclaimer: These citations have been automatically generated based on the information we have and it may not be 100% accurate. Please consult the latest official manual style if you have any questions regarding the format accuracy.

Download citation file:

• Search Book

Jump to a Section

Introduction, research design and studies.

• STUDIES OF RISK ASSESSMENT: OBSERVATIONAL STUDIES
• Full Chapter
• Supplementary Content

When you have completed this chapter, you will be able to understand:

Types of studies and their relation to the research objectives

The different types of studies

The difference between primary and secondary studies

The different types of primary studies

Descriptive and analytical observational studies

Various descriptive observational studies and their functions

Various analytical observational studies and their functions

The advantages and disadvantages of observational studies

The previous chapter described the various steps of planning and conducting a research study. This chapter briefly introduces the reader to the different types of studies and then elaborates on the observational studies. In observational studies, the researcher observes the involvement of the participants and collects data by simply observing events as they happen, without playing an active part in what takes place. In interventional, or experimental, studies, the investigator exposes the participants to some kind of intervention and tries to find a relation between the intervention and the outcome. Observational studies can be descriptive, like the case studies and case series, but are more commonly analytical (cross-sectional, case–control, and cohort studies). Descriptive observational studies describe characteristics of a population and usually do not have a hypothesis; they are sometimes hypothesis-generating studies. An analytical observational study, in addition, tries to find a causal relationship between two or more comparable groups (variables) and has a hypothesis to prove.

A study design is a road map or blueprint based on the type of research to be carried out. It starts with development of the research question, formulating a hypothesis and research objectives, and subsequent planning for carrying out the research. The research objectives of the proposed study determine the type of study to be undertaken.

Types of Studies

Type of studies in medical research can be broadly classified into primary and secondary studies. Primary studies are those that are actually performed by the investigators, while secondary studies summarize the results of different primary studies in the form of systematic reviews and meta-analyses without actually performing the studies. 1 Primary studies can be put into three groups based on the type of research undertaken: basic medical or experimental studies, epidemiologic studies, and clinical studies. Basic medical studies include research in animal experiments, cell studies, biochemical, genetic and physiologic investigations, and studies on the properties of drugs and materials. Epidemiologic studies investigate the distribution and historical changes in the frequency of diseases and the causes for these diseases, while clinical studies involve research in human subjects. However, it may be difficult to classify individual studies into one of these three main categories. 1 A more practical way to classify the types of research studies based on their function is to group them into observational and interventional (experimental) studies; the former can be further subclassified into descriptive and analytical studies ( Figure 3-1 ).

Sign in or create a free Access profile below to access even more exclusive content.

With an Access profile, you can save and manage favorites from your personal dashboard, complete case quizzes, review Q&A, and take these feature on the go with our Access app.

Download the Access App: iOS | Android

Pop-up div Successfully Displayed

This div only appears when the trigger link is hovered over. Otherwise it is hidden from view.

Please Wait

• Recommended
• Recently Viewed
• Collections
• Review Questions
• Create a Free Profile