Research exam 1

What is Evidence-Based Medicine?

"Evidence-based medicine is the integration of best research evidence with clinical expertise and patient values" Clinical expertise, Best Research Evidence, Patient Values and Preferences -> EBP (Still solidly centered on research evidence How treatments are likely to work in your community What the patient wants (not just on clinician to make decisions)

The research question should be framed in a manner where....

"The answer to the research question should help an investigator to understand or explain the problem. This answer should also advance our understanding of the nature, laws, and theories underlying the given research problem." The research question should focus on a specific aspect of the problem. It should address a small component, not the whole issue.

What Is Patient-Centered Care?

"The term 'patient-centered medicine' was introduced by Balint and colleagues, who contrasted it with 'illness-centered medicine'. An understanding of the patient's complaints, based on patient-centered thinking, was called 'overall diagnosis', and an understanding based on disease-centered thinking was called 'traditional diagnosis'." Outcomes that patients experience in real-world settings They are patient-oriented rather than disease- or physician-oriented, e.g.: - Health status - Functional status - Quality of life - Quality of death - Symptoms; pain, nausea - Psychosocial well-being (Need to think about outcomes that having meaning to patient Successful for clinician does not equal successful for patient ) (Need to add outcomes to research studies How quickly can get back to games, work Pain can be both clinical and patient centered )

Framing a Research Question

"a well-worded research question is necessary ...... to define a useful study" What is a well worded research question? A good research question should be focused on a concept that is relevant to the problem. A well focused question will clearly identify the core components that will be examined in a research study.

Qualitative Research

'Qualitative Research...involves finding out what people think, and how they feel - or at any rate, what they say they think and how they say they feel. This kind of information is subjective. It involves feelings and impressions, rather than numbers'

Variable

- An issue that can change, that can vary, or that can be expressed as more than one value, or in various values or categories. Example of Variables: age, gender, depression scale scores, muscle strength, range of motion, type of diet, supplements taken daily, and many more

Observational studies examine study subjects in a "non-controlled" fashion

- Exposure levels in a subject are observed or - Types of treatments are observed - The investigator observes the natural course of events - Documents who is exposed or non-exposed - Documents who has or has not developed the outcome of interest (In contrast, research designs that are observational in nature are noted by their methods that 'observe' and record what happens to research subjects. In an observational design, there is no experimental control over how treatments are provided to subjects; rather there is an observation of what treatments individuals actually use. Similarly, the levels of exposure as they exist are identified and recorded by the study investigator. Observational studies are lower in the research hierarchy than experimental studies. This lower ranking is due to the possibility that the measurement of the exposure (or the intervention) may not be as precise as in experimental settings.)

Post-Positivism or post-positivist philosophies emerged in the late 1950s, and are often associated with qualitative approaches to looking at problems

... "reality is not rigid. Reality is influenced by its context, and many constructions of reality are possible" (Objective approach may overlook problems affecting individuals Everything can't be reduced to a simple fact Today trying to do a mixed method of combining both) Post-Positivism emphasises - that the world is too complex to be reduced to a series of law-like generalisations. the uniqueness of people, and circumstances and the constant nature of change. details matter - there are deep layers of reality to scientific knowledge that positivist approaches overlook. subjective reality matters.

Research is defined as....

... the systematic investigation into and study of materials and sources in order to establish facts and reach new conclusions.

Hypotheses in the Real World

1. Discoveries or hypotheses are sometimes resisted because they seem counter-intuitive 2. "Disappointment is when a beautiful hypothesis is destroyed by an ugly fact" ( This introduction to hypotheses has been purposely instructional in its orientation. From this presentation, one might assume that hypotheses are fairly simple constructs.... and they are. However, in real world settings, it is important to point out the following: Many researchers may disagree with your proposed hypothesis. This disagreement is not because your hypothesis is not correctly stated, but rather because other researchers believe that your interpretation of the research literature is not correct (and thus, your hypothesis is not worthwhile). There can be very different interpretations of the research evidence in science. All investigators should be prepared to have their ideas questioned. Particularly if your idea is new or novel. The sad reality is that even the best hypotheses frequently turn out to be false. That is, the research is conducted and the results support the null hypothesis!)

Expand on the Alternative Hypothesis;

1. Predict an outcome for the study... Persons exposed to the factor have a five times higher risk of disease than persons not exposed The computer assisted treatment intervention leads to improved functional outcome compared to standard care (In this we know previous research has proven this - now trying to learn more In framing a testable hypothesis, the alternative hypothesis can be used to express different types of statements regarding the relationship between the independent and dependent variables. In this example, one strategy is to 'predict an outcome' for the study. The purpose of the study, then, will be to identify if the predicted outcome is true or not. In setting a prediction, an investigator can propose that the exposure predicts an outcome (yes/no) or predicts a certain level of the outcome (e.g. to propose a magnitude of effect for the exposure). An investigator can also propose a statistical prediction to indicate a specific value for a population parameter (such as the outcome). Most alternative hypothesis that predict an outcome are used when several prior studies have been conducted on the issue, and the challenge is not to identify if an association exists between an independent and dependent variable....that answer is known from prior research....but rather the challenge is to identify the level of the relationship between these variables.) Expand on the Alternative Hypothesis: 2. Suggest relationship between specific exposure and health-related event... A high cholesterol intake is associated with the development (risk) of coronary heart disease. (In this one: We don't have a solid relationship between these 2 Another strategy to use as an alternative hypothesis is to "suggest" that a relationship exists between the independent and dependent variables. This is a very simple version of an alternative hypothesis and is a very common type of hypothesis formulated in epidemiologic research settings. The purpose of the research study, then, will be to identify if there is any relationship at all. "Suggesting" a relationship as an alternative hypothesis would be used in scenarios where there is not much previous research on the issue, or in scenarios where the current research evidence is conflicting regarding the relationship between the independent and dependent variables.)

Further Guidelines for Framing Hypotheses:

1. State the exposure to be measured as specifically as possible. 2. State the health outcome as specifically as possible. Strive to explain the smallest amount of ignorance ( As the research hypothesis sets the framework for the conduct of the research study, it is important that the key variables in the study are expressed with a specific and narrow focus, rather than with a broad perspective. The exposure variable should be expressed so that it identifies a very narrow measure of the variable. The same thing should be done for the outcome variable.)

The Scientific Method

1. Suspicion that a factor (an exposure or an intervention) may influence occurrence of disease or a noted health outcome Observations in clinical practice Examination of patterns - Do subpopulations have higher or lower rates? - Are rates increased in the presence of certain factors? Observations in laboratory research Theoretical speculation 2. Identify variables you are interested in: • Exposure - (risk factor, protective factor, predictor variable, treatment) • Outcome - (disease, event) 3. Formulate a specific hypothesis Frame a hypothesis which seeks to answer a specific question about the relationship between an exposure and an outcome

Reasoning has been shaped by significant research events in healthcare

1747; Lind identifies citrus as prevention for scurvy 1867; Lister identifies disinfection in surgery 1870; Koch & Pasteur develop of germ theory of disease 1895; Roentgen discovers X-rays 1897; aspirin identified in Germany 1921; Banting and Best discover insulin 1928; Fleming identifies penicillin 1951; link between fluoride and reduced cavities 1952; study linking smoking and lung cancer 1955; Salk vaccine for polio distributed in community 1957; Safar and Elam advance CPR 1963; Starzl performs first human liver transplant 1978; first test-tube baby 1983; HIV identified 2006; HPV vaccine implemented

.. continued

3. Suggest a cause-effect relationship.... The intervention causes an improvement in lung functioning (Another type of strategy for the alternative hypothesis is to propose that the relationship that exists between the key variables is a causal relationship. This type of formulation is an advanced type of hypothesis. It is generally only applied when there is strong prior research evidence of a link between the exposure and the outcome (or an experimental study is being undertaken where the intervention is thought to produce an specific health outcome). The goal of this type of research is to carry out a study to gain stronger (and additional) evidence that an exposure is the likely reason for the outcome. Note that there is a difference between the wording used in this alternative hypothesis and the previous two alternative hypotheses. In this example, the term "causes an outcome" is the strongest type of statement that an investigator can make in a hypothesis. In contrast, the term "the exposure is associated with the outcome" is a neutral statement. The term "the risk is five times higher in the exposed group" is a moderate type of statement. The point of this discussion is to point out that any proposed hypothesis should fit in with the current research evidence on a given topic. A very strong causal relationship statement, for example, should not be used when the research evidence is not clear. A neutral statement suggesting an association would be more appropriate in that scenario.)

Directionality in a Hypothesis:

4. "One-sided" vs. "Two-sided" One-sided example [Directional]: Helicobacter pylori infection is associated with increased risk of stomach ulcer Two-sided example [Non-directional]: Weight-lifting is associated with risk of lower back injury (Saying increased helicobacter pylori is two-sided; its not saying increased risk The alternative hypothesis may or may not identify a direction to the relationship between the independent and dependent variables. This decision is at the discretion of the investigator. However, some guidelines have been proposed to determine if a directional hypothesis is appropriate or not. First, what is the meaning of having a direction in an alternative hypothesis? A hypothesis that is directional (a one-sided hypothesis) is a hypothesis that proposes that there is only one direction to the relationship between the independent and dependent variables. In the example, the direction is that the risk of the outcome is higher. In other words, the exposure produces only an increase in the outcome. In contrast, a hypothesis that is non-directional (a two-sided hypothesis) is a hypothesis that proposes that there can be two possible directions to the relationship; that the exposure can lead to a higher level of the outcome or the exposure can lead to a lower level of the outcome. In the example, weight-lifting (the exposure) may be related to an increased risk of back injury or to a decreased risk of back injury. Now, let's consider the guidelines proposed for when to propose a directional or non-directional alternative hypothesis. The type of direction to assign in a hypothesis will depend upon the research question. If there is only one direction that an exposure or intervention can affect an outcome, then a one-sided hypothesis is appropriate. However, most exposures and interventions may produce impacts in either direction (positive or negative) on the outcome, and therefore, a two-sided hypothesis would be appropriate. The level of scientific understanding of the research topic also shapes the directionality that should be assigned in a hypothesis. If the level of research evidence for an issue is preliminary, then, most investigations should use a two-sided hypothesis to allow for a fuller exploration of the issue. If a great deal of evidence exists for a research topic, much of the evidence will already have identified a direction to the relationship. Thus, a one-sided hypothesis should be adopted for a proposed study.)

A long history underlies the development and use of the scientific method. Core assumptions include....

A cause-and-effect relationship is a phenomenon that is a part of the laws of nature The laws of nature are universal and orderly Relationships can be discovered Findings about relationships are grounded in the objective observation of phenomenon (evidence) Any proposition (such as a cause-and-effect relationship) is open to analysis and critique

The difference between positivist and post-positivist research approaches

A positivist approach - Deduction You identify a theory regarding a health issue and develop a strategy to test a hypotheses regarding the application of theory to the issue. A post-positivist approach - Induction You collect data from subjects about a health issue and analyze the data to develop themes about the issue

Descriptive Study Data

A systematic approach to identify patterns from observations. Use patterns to help formulate hypotheses.

Reality of Scientific Research

Advancement of knowledge through research moves slowly. Many great ideas turn into dead ends. Most published research findings are false. Many researchers cling to false assumptions. Attributes of good science: patience, flexibility of thinking, reliance on proven truths.

Basic Question in Research

Are exposure and disease/outcome linked? Is there an association between them? E (exposure/intervention) -> D (disease/ health outcome)

"Generalization is an act of reasoning that involves drawing broad inferences from particular observations"

Classic: infer from observations in a sample to a population sample -> population The standard in quantitative studies, controversial in qualitative studies

Data from Case Series Studies

Clinical Case Series - a coherent and consecutive set of cases of disease (or health outcome) derived from a clinical practice or defined health care setting. Describes patterns of disease in the clinical practice Population Case Series - collection of cases within a defined geographical area Describes patterns of disease in populations ( a complete set of cases) Usually rely upon routinely collected data from established surveillance systems, such as a population survey or registry. (Patients or individuals may be identified from a single or multiple sources)

There are multiple perspectives in health outcomes

Clinician Focused Physio-logical Caregiver Focused Patient Focused Examples: Global impressions, Observations & tests of function FEV1, HbA1c, Tumor size Dependency, Functional status Functional status, Well-being Symptoms, HRQL, Satisfaction with TX, TX adherence, Utility/preference-based measures (What is indication that clinician is looking for (clinical indicator such as knee flexion) #2 is actually lab tests - breathing, blood tests Can't exclude affect of caregiver Will they follow/stick with treatment?)

What Is Comparative Effectiveness?

Comparative effectiveness research is the conduct and synthesis of research comparing the benefits and harms of different interventions and strategies to prevent, diagnose, treat and monitor health conditions in "real world" settings. The purpose of this research is to improve health outcomes by developing and disseminating evidence-based information to patients, clinicians, and other decision-makers, responding to their expressed needs, about which interventions are most effective for which patients under specific circumstances. Comparative effectiveness research must assess a comprehensive array of health-related outcomes for diverse patient populations and subgroups. Defined interventions compared may include medications, procedures, medical and assistive devices and technologies, diagnostic testing, behavioral change, and delivery system strategies. This research necessitates the development, expansion, and use of a variety of data sources and methods to assess comparative effectiveness and actively disseminate the results.

Comparison

Comparison of Interest Identify the alternative to compare to the intervention or the exposure Examples: New drug compared to placebo New treatment compared to existing treatment No treatment compared to surgery High exposure compared to low exposure

Reasoning has been shaped by the philosophy of research and its evolution

Core Philosophies of Research Positivism Post-positivism (Objective based philosophy is big deal = positivism)

Why Comparative Effectiveness Research?

Currently, evidence of inappropriate use of health care technologies, including over-use, under-use, and improper use Evidence of large variations in practice Lack of evidence on "head-to-head" comparisons of alternative interventions for particular health problems Continued rapid increases in health care costs

Different criteria and philosophical views have been proposed to assess causality over time

DEDUCTION: Conclusions can be drawn from evidence when there is 100% certainty FALSIFICATION: Karl Popper, 1959 Scientific hypotheses can never be proved or established as true. Therefore, science advances by a process of elimination CONSENSUS: Thomas Kuhn, 1962 The consensus of the scientific community determines what is considered accepted and what is refuted BAYESIANISM: Bayes; / Cornfield, 1976 Cause should not be treated as being known with certainty. Reasoning requires assignment of degree of probability (Different philosophical views on science and the meaning of causation have existed over time. A causal philosophy sets the broad framework for thinking about what is important and what is not important in causal arguments. Major philosophical viewpoints are highlighted on this slide, but other viewpoints also exist. For several centuries, causation was typically framed around the deduction philosophy. Under this belief system, causation is only shown with solid proof. More recently, it has been recognized that the philosophy of deductionism is unrealistic with respect to several health issues. Differing viewpoints have gained acceptance, such as falsification, consensus and bayesianism. Today, many scientists would argue that the current prevailing philosophy for causation is a hybrid of these three viewpoints.)

Nature of research - positivist

Deals with positive facts and observations Subscribes to the 'scientific method' Primary goal is not only description but prediction and explanation Classification of substances and events, and observation of these, provide the basis for descriptive laws based on consistencies in patterns and properties Quantitative, as it draws on measurable evidence

Evidence-Based Practice

Defined as: "Conscientious, explicit and judicious use of current best evidence in making decisions about the care of individual patients/clients."

Hypothesis

Definition: A tentative explanation for an observation, phenomenon, or scientific problem that can be tested by further investigation [American Heritage Dictionary] A tentative statement, arrived at by observation or reflection, that characterizes a testable prediction, which may be accepted or refuted through testing [Last, Dictionary of Epidemiology]

Framing a Hypothesis

Definition: A tentative explanation for an observation, phenomenon, or scientific problem that can be tested by further investigation [American Heritage Dictionary] A tentative statement, arrived at by observation or reflection, that characterizes a testable prediction, which may be accepted or refuted through testing [Last, Dictionary of Epidemiology] Hypotheses shape and guide a research study through several mechanisms in terms of: - identification of study sample size - what issues should be involved in data collection - the proper analysis of the data - data interpretation (Hypothesis = what you predict relationship is or what outcome will be Needs to include 2 key variables)

Research Question

Definition: Construction of an interesting question that is relevant to a research topic, which should advance our understanding of the nature, laws, and theories underlying the given topic

History of EBP

Developed by David Sackett, Medical Professor from McMaster University, Canada, 1970 A clear need existed for a better way to make clinical decisions Fueled by developments in the field of clinical epidemiology First textbook based on EBM 1985 (Large scale studies not being introduced to physicians/practitioners in the field -> main motivation of evidencebased practice - get practitioners to actually utilize Large portion of healthcare is hesitant to change)

Main Attributes of CER

Direct ("head-to-head") comparisons of alternative interventions (as opposed to comparison with placebo or indirect comparisons) Applies to all types of interventions - pharma, biotech, devices/equip't, medical and surgical procedures; organization, delivery, financing Effectiveness (in realistic health care settings) rather than efficacy (in ideal circumstances) Emphasizes health care outcomes (e.g., morbidity, mortality, symptoms, QoL, adverse events) rather than intermediate/surrogate endpoints

Assess validity of association

Does the observed association really exist? - Is the association valid? - Are there alternative explanations for the association? -- Chance -- Bias -- Confounding

The Positivist philosophy of research...

Emerged in the 1800s Deals with positive facts and observations Observes events and classifies them based on consistencies in patterns and properties. This knowledge is the basis for descriptive laws. Primary goal is not only description, but also prediction and explanation Draws on measurable evidence for deduction Positivism or positivist philosophies are often associated with quantitative methods and the hard, physical sciences ... "assume things can be studied as hard facts and the relationship between these facts can be established as scientific laws." ....emergence of the scientific method (Quantitative and positivism are synonyms ; focuses on quantitative approach (a large majority in this method are not correct; people gather evidence and make decisions - may not have complete evidence Led to scientific method)

Three fundamental elements underlie the scientific method

Empiricism: the notion that enquiry is conducted through observation and knowledge verified through evidence Determinism: the notion that events occur according to regular laws and causes. The goal of research is to discover these Scepticism: the notion that any proposition is open to analysis and critique (These points highlight previous slide Answer is there - need adequate approach to find it; don't research something that has no chance to exist; can satisfy if make coherent argument Assumption that mechanism is there; research will give more info to understand laws better Don't accept at face value what you read)

Example: Adult Stature and Risk of Cancer in Post-menopausal Women

Environmental factors in may affect growth factors in children and adults. Growth factors may spur cancer growth. Does height affect risk for cancer? Height in a large cohort of women was identified and new cancers over 12 years were documented. Height was positively correlated with risk of all cancers Environmental factors related to growth may play a role in cancer development.

Research Planning; Introduction to Study Designs Hierarchy of Evidence in Research Descriptive Study Designs

Explain the difference between a quantitative and qualitative research design Identify the fundamental research designs and their classification Identify the fundamental time frames that can be used in research Explain how different study designs provide different levels of evidence to support a relationship between an independent and dependent variable

Hypothesis Generation; Identifying a Research Question/ Hypothesis

Explain where research questions and hypotheses fit in the scheme of research Identify strategies which underlie the formation of a research question/hypothesis Identify a research topic/hypothesis from observational data Explain the role of the literature review in identifying research issues and hypotheses Search PubMed on a research topic area

Quantitative and Qualitative Research

Focus on different types of research questions Qualitative research aims at understanding the dynamics of an event. It addresses a "How" or "What" question. Quantitative research aims at a (causal) explanation for events. It addresses a "Why" question. Mixed methods - a research approach that incorporates qualitative and quantitative elements together. (Quantitative and Qualitative research designs focus on gathering different types of information. A qualitative research design is targeted to identify information that describes or explores the common experiences or dynamics related to a health issue. A quantitative research design is targeted to identify information that answers a hypothesis, most often by comparing two groups (those exposed and not exposed; or those with an intervention and those without the intervention). In some situations, both goals are desired, and a mixed methods study design would be appropriate. Most research in the health care sciences is focused on quantitative study designs.)

What is the essence of a good research question?

For a research question/hypothesis, one should consider whether or not it is: Feasible - Study can be accomplished in current environment - Adequate number of study subjects - Available technical expertise - Affordable Interesting - The question intrigues peers and the community Novel - The study is original or can provide original information Ethical - The study protects research subjects, fits community norms The researchers are ethical Relevant - The study question is relevant to current practices, future needs, and/or policy questions = FINER Framework

A hypothesis should be framed to test a specific question about the relationship between an exposure or intervention and an outcome

Framing a Testable Relationship Traditional; offer a statistical hypothesis H0: "Null" hypothesis (assumed) H1: "Alternative" hypothesis (NULL = NO RELATIONSHIP; ALTERNATIVE = RELATIONSHIP ; NULL IS ASSUMED CORRECT UNTIL PROVING OTHERWISE Another core element of a good hypothesis is the proposition of a relationship between the independent variable and dependent variable. Further, this relationship should be testable. Several strategies exist for proposing a testable relationship between the independent and dependent variables. In most epidemiologic research, a statistical hypothesis is proposed. In a statistical hypothesis, a null hypothesis and alternative hypothesis are proposed. Common notation of a null hypothesis is the abbreviation, H (naught). Common notation of an alternative hypothesis is the abbreviation, H1 or Ha.)

Association

From the results of your study, does a statistical relationship exist between two or more events, characteristics, or other variables Is there a statistical relationship, or association, between exposure/intervention and disease/outcome?

Framing a Testable Relationship

H0: There is no association between the exposure and disease of interest H1: There is an association between the exposure and disease of interest (beyond what might be expected from random error alone) (The null hypothesis is a statement that there is no relationship between the independent and dependent variables. The alternative hypothesis is a statement that some type of relationship does exist between the independent and dependent variables. The next set of slides will further explore how the alternative hypothesis can be stated. Note that in a research study of possible determinants for disease, the study is working to answer the broad idea of if the exposure causes the outcome or not. Many describe this philosophy as examining if a cause-effect relationship exists or not. In any research study of this nature, the investigators should enter the study with the default assumption that the null hypothesis is correct (i.e. that there is no relationship between the exposure and the outcome), until the study proves otherwise. Thus, it is "assumed" that a null relationship exists in a research study.)

Why is EBP Important?

Healthcare literature with clinically applicable findings is published at a rate that is impossible for individual clinicians to keep up with and stay current. Using guidelines can save considerable time in identifying relevant evidence

Independent Variable

Hypotheses identify the key variables for the study. One key variable in a hypothesis is the independent variable The independent variable represents the presumed 'exposure' of interest in a study; also a variable that can be used to predict or explain the values of another variable In experimental studies, the independent variable (intervention) is manipulated by a researcher who predicts that the manipulation of this variable will have an affect on another variable (INDEPENDENT VARIABLE = INTERVENTION The key variables in a research study are the exposure and outcome variables. A good hypothesis should include these two variables in its wording. The exposure variable is also often referred to as the "independent" variable. In theory, this represents a variable that occurs with independence in the population. The purpose of the study is then to evaluate if a given level of this "independent" variable is related to the level of a health outcome observed. In clinical epidemiology research, a health care intervention may be the focus of the research in order to identify if the intervention affects a health outcome. In this scenario, the intervention variable is akin to the "independent" variable.)

Identifying a Research Idea.....How?

Ideas in research can be formulated from intuition or suspicions Intuition/suspicion that a factor may influence occurrence of a health event or a noted health outcome Suspicions from: Observations in clinical practice Examination of health/health outcome patterns Do subpopulations have higher or lower rates? Are rates increased in the presence of certain factors? Observations in laboratory research Theory

Research - Where do you get started?

Identify Research Problem Identify Specific Research Question(s) and Aim(s) Specify Hypothesis(es) related to Aim(s)

Using a literature review to identify a research question

Identify a topic area of interest. Search the literature for relevant research. Read the literature, identify methodological approaches and study results. Evaluate (critically appraise) the evidence. Identify areas where clear patterns of knowledge are known. Identify areas where knowledge is not clear

The Scientific Method Evidence-Based Practice Patient Centered Outcomes Research

Identify the basic approaches to the scientific method Identify the framework of evidence-based practice and how it uses research Describe what patient values mean Describe comparative effectiveness research and identify how patient values are included

History and Philosophy Underlying Research

Identify the definition of research Identify the philosophies that underlie the approaches to research

Hypothesis Generation; Framing a Research Question Framing a Hypothesis

Identify the fundamental elements of a good research question Explain the role of a hypothesis in a research study Identify the fundamental elements of a testable hypothesis Identify the difference between a one-sided and a two-sided hypothesis

Key Elements in a Hypothesis

Identify variables you are interested in: - Exposure or Intervention (risk factor, treatment) - Outcome (disease, event) State a relationship between the variables - Identify a relationship between an exposure (or intervention) and the outcome The statement must be testable - Frame the hypothesis to answer a question (e.g. yes/no) about the relationship

EBP argument

If there is a better way to practice, therapists should use it. They should critically evaluate what has already been done to see if it could be improved. EBP does not ignore clinical practice. Knowledgeable clinicians will know best how to implement EBPs findings EBP can be considered to be a: "combination of information from what we know from research, what we have learned from clinical wisdom, and what we learned from information from the clients and their families to make the best use of knowledge." (Guidelines is telling people what to do; Mostly works once insurance etc gives guidelines - then not paid if do otherwise )

Intervention

Intervention (or exposure in observational studies) Identify the intervention under assessment (in experimental studies) Identify the exposure under assessment (in observational studies) Examples: New drug, therapy, technology, or procedure Exposure of interest (e.g. behavior, environmental, social, etc.)

Intuition and Creativity are vital parts in the production of scientific knowledge

Is the following issue an example of knowledge acquisition gained through intuition in the healthcare field? How (Applying intuition in health care research is not easy. Much of it comes from prior experience both in and out of the health care research field.)

Qualitative research provides a means of exploring and understanding the meaning of a health problem

It seeks a subjective assessment to identify a fuller explanation to the problem. It is based upon an inductive style of analysis (moving from specific data to general meaning) Research process: involves emerging questions and procedures, identifies general themes, and the researcher makes interpretations of the meaning of the data.

Quantitative research provides a means for comparing outcomes between variables related to a health problem

It tests objective questions by examining the relationship among variables. It is based upon deductive reasoning ( moving from general understanding to specific meaning) Research process: variables can be measured so that numbered data can be analyzed using statistical procedures. Most widely applied type of research in health care

Research in health care is .......

Knowledge acquisition gained (Systemic investigation) through reasoning through intuition but most importantly through the use of appropriate designs and methods Reasoning underlying research in healthcare has evolved over time

Using Research for Clinical Care

Many paradigms have been proposed to address and improve patient care These include: (among others) - Outcomes research - Evidence-based medicine/practice - Comparative effectiveness - Patient centered outcomes research (Shift in opinion as to what health research should focus on)

Observations from:

Multi-national comparisons Natural experiments Descriptive studies (assessment of patterns) Literature (review PubMed on topic area) Creativity

Research Design Overview Which Design? Quantitative

Observational design: Preliminary to moderate knowledge of issue Small/moderate resources Independent variable may be harmful Want to observe level of independent variable Dependent variable is rare in the population Want to gain evidence to infer causality Quasi-experimental design Moderate knowledge of issue Moderate resources Independent variable is not or may be harmful Want control on level of independent var. Dependent variable is common Want to gain evidence to infer causality Experimental design Advanced knowledge of issue Ample resources Independent variable is not harmful Want control on level of independent var. Dependent variable is common Want strongest support for causality (So which type of design should an investigator select? A brief flow chart is shown here to highlight the criteria that support the use of each type of design. Note that this flow chart illustrates that topic areas with very little prior research should be addressed first in observational designs. Also note that the level of resources available can influence this decision process. Experimental studies usually require a large commitment of human and financial resources. Observational designs can be lower in cost. Many quasi-experimental studies are moderate to expensive in their use of resources. Also highlighted in the chart that exposures that are thought to be harmful are best suited for observational designs. Why is it possible to study harmful exposures in observational, but not experimental studies? Remember that the representations in this flow chart are meant to be general. Information presented in subsequent lectures will provide further details on when specific types of study designs would be appropriate in these scenarios.)

Outcome

Outcome of Interest Identify the expected outcome of the intervention or exposure Examples: Improve performance of everyday life tasks Increase recovery rate Improve morbidity, improve mortality Improve QOL

Core Components of a Research Question

PICO Patient, Population or Problem Intervention or Exposure Comparison Outcomes of Interest Example: In females with carpal tunnel syndrome, what is the efficacy of exercise and ergonomic intervention compared to surgical treatment for decreasing pain and disability? To whom does research apply? Intervention = treatment; how will it effect; still much research does not involve an intervention; in question in previous slide- olive oil is an exposure not an intervention Giving something = intervention; observing something happening is exposure (such as population that is already consuming olive oil) Comparing to normal care? Placebo? In what way does treatment have an effect? State explicitly This question is showing 2 outcomes (decreasing pain, disability); should only show 1

Example Hypotheses:

POOR: Eating junk food is associated with the development of cancer. GOOD: The human papilloma virus (HPV) subtype 16 is associated with the development of cervical cancer. In this context, consider the example in the slide which highlights two hypotheses. (The first hypothesis is an example of a poor hypothesis, because the focus of the key variables is quite broad. The independent variable is 'junk food' and the dependent variable is 'cancer'. Both of these variables are not narrow. First, 'junk food' is a term that can have several different meanings. What is junk food; ice cream? Cookies? Potato crisps? Second, 'cancer' is also non-specific. The term 'cancer' means that all forms of cancer would be the focus of the investigation. There are several forms of cancer! If you consider that the research hypothesis should serve as a guide to how the study is carried out, this designation of the dependent variable would mean that the investigator has to measure every form of cancer in order to test the hypothesis (as written). The second hypothesis is a much better hypothesis. It proposes a relationship between a specific exposure (HPV subtype 16) and a specific outcome (cervical cancer). There are many different sub-types of the HPV agent, but in this research, the measurement will only be focused to identify subtype 16. There are many different types of cancer, but in this research the measurement will only focus on cervical cancer.)

Five Steps to the EBP Process

Patient Focus Step 1: Formulate a well-built question Step 2: Identify articles & other EB resources that answer the question Step 3: Critically appraising the evidence to assess its validity Step 4: Apply the evidence Step 5: Re-evaluate the application of evidence and areas for improvement

What Is Patient-Centered Outcomes Research?

Patient-centered outcomes research (PCOR) helps people make informed health care decisions and allows their voice to be heard in assessing the value of health care options. This research answers patient-focused questions:

In descriptive data, patterns of a health issue are assessed.

Patterns may exist by the characteristics of who is affected: Age, gender, health care behavior Patterns may exist by the characteristics where the problem occurs: location, community clinical practices Patterns may exist by when the problem occurs: changes in health issue/practice over time Remarkable variation exists in rates of disease across racial and ethnic groups in the community Health issues and health outcomes show considerable variation both between countries and within a country

Basic Approaches in Evidence-based Medicine

Population Focus: Use evidence base, clinical norms, and societal norms to develop guidelines for patient care Patient Focus: Use evidence base, clinical expertise, and patient preferences to develop treatment approaches to apply to specific patients (Patient focus is most talked about GUIDELINES - normally evidence base for patient comes from guidelines for population)

Population

Population or [in EBM, the Patient or Problem] Identify the specific population that is the focus of the study Examples: Patients who have sustained a stroke Adults between the ages of 18 and 25 with a mild traumatic brain injury Veterans with a PTSD diagnosis

Example: Can the study of gun violence provide a basis for improvements in safety in city neighborhoods?

Positivist approach: collect data via interview, classify types of incidents, identify key variables related to the incidents, make recommendations to reduce violence based on analysis Post-positivist approach: analyse interviews in depth, seek to draw conclusions about causal factors Questions: Y/N; how many times (to quantify) [scientific/health research is very positivist] [coming into equation is more post-positivist approaches - can lead to prevention strategies etc to improve health] How often guns were involved; characteristics of neighborhoods (population density) Vs. have you been a victim and how has it affected you? More open ended responses, - help find common elements; effects on people can be different - more rich, deep answers

Research ideas are identified from a current problem. These ideas are then refined to investigate a specific element of the problem.

Problem - High rates of stroke in Scandinavia Research idea - Diet is different in Italy than in Scandinavia. Thus, diet may play a role in stroke. Research question: Does olive oil protect against the development of stroke? (Efficient to have narrow question - still very general Who does this question pertain to? Where are we going to carry out this research?)

Timeline to CER

Provided $$ to research that technology was beneficial Different ways to treat patient - evidence into outcomes and which approach is best Comparative - most research tx are compared to placebos; test tx and compare to each other

Delay in Lab-to-Practice of Research Findings

Research findings are often delayed in being implement into clinical practice. It takes an average of 17 years for clinical research to be fully integrated into everyday practice.

Research Process

Research question Hypothesis Identify research design (design, study population) Data collection (exposure method, outcome method) Presentation of data (portraying study results) Data analysis (descriptive statistics, inferential statistics) Interpretation of data (bias, confounding)

One outline of the deductive approach in a quantitative study would look like the following:

Researcher identifies hypothesis Researcher identifies an explicit design to test the hypothesis Researcher defines and operationalizes key variables to assess the hypothesis Researcher measures variables using an appropriate method Researcher examines the relationship of the key variables Researcher draws conclusions on the hypothesis

One outline of the inductive logic in a qualitative study would look like the following:

Researcher poses generalizations, or theories, and compares to past experiences and literature Researcher looks for broad patterns, generalizations, or theories from themes or categories Researcher analyzes data to form themes or categories Researcher asks open-ended questions of participants or records fieldnotes Researcher gathers information

The systematic acquisition of knowledge in health care is gathered largely through the use of the scientific method

SCIENTIFIC METHOD Choose a question to investigate Identify a hypothesis related to the question Make testable predictions in the hypothesis Design an experiment to answer hypothesis question Collect data in experiment Determine results and assess their validity Determine if results support or refute your hypothesis

Opposition to Evidence-Based Practice

Some established practitioners object to EBP because it bases decisions on research evidence (or dictates clinical decisions based on research evidence) instead of recognizing the experience and expertise that professionals have been developing throughout entire careers.

Natural Experiments -- Migrant Studies

Studies taking advantage of the migration of individuals from one country to another country; ..... with the assumption that the physical and biological environments are different between the countries, as well as the cultural background and/or genetic makeup of the populations. The populations also have different morbidity or mortality experiences. Source Population [old country] -> Host Population [new country] (Source population can be Japan population at Japan and Migrant population can be Japanese-American people at California, US. So it is all we need for migrant study_ Genetic hypothesis / environmental hypothesis: Source: 33 per 100,000 pop -> Host: 15 per 100,000 pop (16 per 100,000) (This is example of environmental reason for disease. Everybody has the same risk at host and migration populations . And pure environment at source population. If disease have a genetic reason - the migrant population will have the same risk as parent population.)

4. Design Study

Study Designs ...(not exhaustive) - Case series - Cross-sectional - Case-control - Cohort - Randomized controlled clinical trial

Health care research may focus on either characterizing risk or prognosis

The Course of a Health Problem healthy -> onset of problem = risk onset of problem -> outcomes of tx = prognosis (How to treat; resolution of treatment Risk = chance to develop health problem in first place; exposure is causal After problem develops - how to treat? What happens = prognosis )

Statistical Association

The degree to which the rate of disease or outcome in persons with a specific exposure is either higher or lower than the rate of disease or outcome among those without that exposure.

Dependent Variable

The dependent variable represents the presumed 'outcome' of interest in a study, or the expected effect in a study. The level of this variable 'depends' on the effect of the independent variable in a well validated study DEPENDENT = OUTCOME VARIABLE The outcome variable is another key variable in a research study. It is often also referred to as the "dependent" variable. In other words, the outcome (or the level of the outcome) is dependent upon the presence/absence (or effect) of the independent variable.

Several factors are involved in the decision of which study design to choose

The selection of a research design is based on: 1. the nature of the research question (what information do you want to gather) 2. our prior understanding of the research problem 3. available resources 4. available time (Scientists have many options to consider when identifying a study design to use in a research project. The selection of the most appropriate design for a given project is an important decision. Several factors affect the choice of which study design may be the most appropriate to pursue in research. Broadly, the design to choose will depend upon the type of information that you want to gather, the level of prior research evidence on the topic, the amount of resources available to carry out the study, the time available to conduct the study. An appropriate design is one which can be justified on the basis of these items, as well as other considerations. For example, if goal of the project is to gather information to generate a hypothesis, then a descriptive design would be warranted. If the research problem has been examined several times with an observational design, then it may be justifiable to move to an experimental design. If few financial and human resources are available to carry out the project, then a design that is low in cost would be the target, etc.)

Assessing Prognosis

To evaluate if treatments work, research examines if there is an association between an intervention and an outcome I (intervention) -> O (health outcome) Does the intervention produce an effect? (Whether I or E; still Group of subjects with variability in both I and O)

Assessing Risk

To investigate factors that cause health problems, research examines if there is an association between an exposure and an outcome/health issue E (exposure) -> D (disease/health problem) Does the exposure produce an effect?

Intervention studies or experimental studies examine study subjects in a "controlled" fashion.

What is "controlled"? - Exposure levels in a subject are assigned or - Intervention status is assigned - Assignment is often selected by randomization - Data collection methods may involve masking (Research designs whose intent is to implement an intervention and test if it works, are often conducted through experimentation. Intervention (a.k.a. experimental) studies are noted by the high level of control that an investigator has over the conduct of the research and how information is gathered. What does this mean? It means that study participants are assigned to receive a given intervention (or not), or subjects are assigned to receive a given level of the exposure (or not) during the study. In health care research, the experimental component is often the allocation of a given health care intervention, such as a drug, procedure, program, to one group of study participants and the allocation of a comparison (drug, procedure, etc.) to another group of study participants. In this setting, the type of intervention that subjects receive is controlled by the investigator. The ability to control the level of the exposure or the intervention is regarded as a valuable tool in research, as it allows for a greater understanding of the relationship between and exposure and an outcome. For this reason, intervention studies are often at the top of the research hierarchy of research designs.)

Ideas for research questions can also be identified by conducting a literature review

What is a literature review? A synthesis of research study results (for a given topic) that summarizes key findings in the topic area and identifies gaps in knowledge where questions still remain. This is the traditional manner in which research ideas are identified. (Many include this question is still not known to find topics for future projects ) A literature review involves searching for and reading several scientific papers and reports How do you go about finding those that are relevant? Search for articles in large scientific literature databases search using basic keywords related to the topic Revise search after identifying 1-4 papers that relevant. Revise to use more defined keywords Retrieve articles (most are electronic) Read and synthesize the key findings

1. What basic issue in research do you want to address?

What is the focus of the information you want to gather Description Comparison Exploration (A fundamental question that many researchers face is the question of "Which study design should I use?". One parameter considered in the selection of a research design is the nature of the research problem; or what type of information does the investigator want to gather. Does an investigator want to describe a problem, compare among two alternatives to identify which is best, or explore the meaning of an event.)

1. How do you want to gather the information for your research question?

What is the intent of your research Intervention Observation A mixture of observation and intervention (Another parameter considered in the selection of a research design is the underlying intent planned for gathering information on the research question. Does an investigator gather information through observation or experimentation; or a mixture of both? Different study designs have been developed based upon the intent underlying the research.)

Research Design Overview

Which Design? What is the focus of the research question? address an objective theory or explanation = Quantitative design explore a subjective view or explanation = Qualitative design Examine objective and subjective details = Mixed-method design

The next step in the scientific method is to formally test the identified hypothesis in a research study

hypothesis -> study design The study should follow a specific plan or protocol (the study design) Study designs direct how the investigation is conducted and allows for the translation of a conceptual hypothesis into an operational one (After a research hypothesis has been identified, the next step in research is to develop a research study to test the hypothesis. The research study should follow a plan. The plan for how research is conducted is the study design. The study design outlines the specific processes to follow in the collection of data on the key variables in the study, their analysis, and their interpretation. The remainder of this lecture, and the next several lectures will outline several different types of study designs that exist and are used in epidemiologic research.)

Quasi-Experimental Research

mixture of experiment and observation an experiment in which the investigator lacks full control over the manipulation of the intervention (In some scenarios, the intent is to have a mixture of both experiment and observation; these types of designs are referred to a quasi-experimental research designs. In most circumstances, a quasi-experimental design is a study where the investigator assigns a particular treatment to a research participant, but there is limited control over the treatment, or limited control over the parameters of how the treatment is evaluated.)

Temporal trends

refer to changes in occurrence of a health event over a period of calendar time.

The structure of quantitative research has been likened to follow an hourglass model

start with a broad question narrow down to meaningful hypothesis investigate and observe analyze data reach conclusions generalize result back to prior knowledge and draw new inferences

A study design provides a plan or protocol that directs how a research investigation should be conducted

when to assess exposure when to assess the outcome how to analyze data how to interpret results (A study design provides the foundation for the research project. The study design outlines a plan or structure regarding how a study will be carried out. A study design directs the methods of data collection (when to assess the key study variables), data analysis (the framework in which to analyze the data) and the interpretation of the results.)