Midterm Study Guide EBP PDF

 Week 1 Evidence-Based Practice (EBP) is a method of clinical decision-making that combines the best available scientific research with clinical expertise and a patient's unique values and circumstances. The goal of EBP is to ensure that patient care is informed by the best available evidence...

 Week 1 Evidence-Based Practice (EBP) is a method of clinical decision-making that combines the best available scientific research with clinical expertise and a patient's unique values and circumstances. The goal of EBP is to ensure that patient care is informed by the best available evidence, which optimizes the benefits patients receive from therapy. Key vocabulary in EBP includes: Scientific Research: Empirical evidence acquired through systematic testing of a hypothesis. This includes both clinical research, which involves human subjects, and nonclinical research, which may involve healthy humans, animals, or cell specimens. Clinical Expertise: Knowledge gained from cumulative years of caring for patients and working to improve that care. This expertise is shared through formal academic settings, post-professional education, mentorship, and informally between colleagues. Individual therapists also develop expertise through reflective practice. Patient Values and Circumstances: This includes a patient's beliefs, preferences, expectations, cultural identification, medical history, access to medical services, and family environment. Shared Informed Decision: A choice that is generated through a partnership between the therapist and patient and that is informed by the best evidence. The EBP Process The EBP process involves five fundamental steps: 1 Identify a question: Collect information from the patient and develop a focused, searchable clinical question. 2 Search for evidence: Use electronic databases to search for research evidence that answers the clinical question. 3 Appraise the evidence: Determine if the research is applicable and of sufficient quality to guide clinical decision-making. 4 Integrate the evidence: Combine research evidence with clinical expertise and patient values and circumstances. 5 Evaluate the outcomes: Assess the effectiveness of the process and identify areas for improvement. Barriers to EBP There are several barriers to using EBP in the real world: Time: Clinicians have limited time to spend reviewing research. Lack of generalizability: Research often addresses groups of patients rather than individuals. Lack of research skills: Clinicians may not have been trained to conduct or interpret research. Lack of understanding of statistics: Statistical concepts can be intimidating. Lack of search and appraisal skills: Finding and evaluating research takes practice. Lack of information resources: Accessing research may be difficult or costly Inconsistent culture of EBP: Not all therapists value or implement EBP. Importance of EBP EBP is important because it leads to enhanced confidence and the ability to assist patients in choosing the best options for care. It moves the profession away from habit-based practice toward a careful and systematic assessment of the best available evidence. Patients expect and demand that their care is based on the best available evidence. Applying EBP When using EBP, it's important to remember that the best evidence may not be the best intervention for every patient, due to individual circumstances and barriers. When there is a lack of evidence for a particular intervention, you should consider the scientific rationale behind it, as well as the potential advantages and disadvantages. Clinicians should strive to make the best decisions possible based on the information and resources available. Clinical expertise should be appraised for quality, just like research evidence. Patient values and circumstances are a critical pillar of evidence in the decision-making process. The five steps of the Evidence-Based Practice (EBP) model are designed to facilitate a structured approach to EBP. They are: Step 1: Identify a need for information, and construct a focused and searchable clinical question. This involves collecting information from the patient to identify what additional information is needed for clinical decision-making and then developing a specific and focused question. Step 2: Conduct a search to find the best possible research evidence to answer the question. This step involves using electronic databases to search for research evidence. Step 3: Critically appraise the research evidence for validity and applicability. This involves determining whether a research article is applicable to the clinical question and if it is of sufficient quality to help guide clinical decision-making. Step 4: Integrate the critically appraised research evidence with clinical expertise and patient values and circumstances. This step involves combining the three pillars of EBP and is completed in partnership with the patient. Step 5: Evaluate the effectiveness and efficacy of the efforts in Steps 1-4 and identify ways to improve them in the future. This involves evaluating the process at both the individual patient level and the level of overall practice habits and skills. The three pillars of evidence that support optimal outcomes for patients are: Scientific Research: This is empirical evidence acquired through systematic testing of a hypothesis. It includes both clinical research, which involves human subjects, and nonclinical research, which may involve healthy humans, animals, cadavers, or cell specimens. Clinical Expertise: This refers to the implicit and explicit knowledge about physical therapy gained from cumulative years of caring for patients and working to improve that care. This also includes the therapist's own values and preferences. Patient Values and Circumstances: This includes the patient’s beliefs, preferences, expectations, cultural identification, medical history, access to medical services, and family environment. The three sources of evidence in EBP form a foundation from which therapists and patients work together to determine the best course of care. The three sources of evidence are: Scientific Research Clinical Expertise Patient Values and Circumstances The evidence pyramid illustrates a hierarchy of the best sources of evidence for searchable clinical questions. At the top of the pyramid are systematic reviews (SRs), which combine other studies and represent the highest level of research evidence. Below SRs are individual studies, such as randomized clinical trials (RCTs) and cohort studies. Evidence-based summaries, including clinical practice guidelines (CPGs), are considered "secondary" research studies and are placed slightly above SRs. CPGs include direct recommendations for practice that combine the best available clinical research, clinical expertise, and, ideally, patient perspectives.  Week 2 Here's a briefing document with key vocabulary words related to searching for research evidence, based on the provided sources: **Key Concepts & Vocabulary** * **Searchable Clinical Question:** A foreground question about a patient that is structured to help you find the best available research evidence as efficiently as possible. These questions contain three elements: * **Patient characteristics:** The most important patient characteristics related to their health condition. * **Patient management:** The component of interest, such as intervention, diagnosis, or prognosis. * **Outcome:** Specifies whether the therapist is looking for information about body structures and functions, activity, or participation. * **Background Questions:** Questions that ask for general information about a condition. They are typically answered using general resources such as textbooks. * **Foreground Questions:** Specific questions related to a particular patient, condition, and clinical outcome of interest. These questions are typically answered using a research study or evidence-based clinical practice guideline. * **Boolean Search:** A search method using operators such as "AND", "OR", and "NOT" to combine search terms and refine results. * **AND:** Narrows search results to articles containing all specified terms. * **OR:** Broadens search results to articles containing at least one of the specified terms. * **Keywords:** Important words from a searchable clinical question or synonyms of those words. * **MeSH (Medical Subject Headings):** A controlled vocabulary thesaurus used to index articles in the MEDLINE database, providing a common language across published articles. PubMed has a MeSH database to help find the best MeSH term for a topic. However, sometimes keywords are better to use than MeSH terms. * **Search Engine:** A user interface that allows you to efficiently search a database. * **Database:** A collection of research articles and guidelines. * **MEDLINE:** A comprehensive database of biomedical research evidence accessed through the PubMed search engine. * **Clinical Queries:** A PubMed tool that uses pre-defined filters to help find clinically relevant articles. * **My NCBI:** A tool within PubMed that allows you to create a personal account to save searches, set up alerts, and customize filters. * **Cited By:** A Google Scholar feature that shows publications that have cited or referenced a particular study. * **PICO:** An acronym used to structure a searchable clinical question, especially for questions about intervention. The elements are: * **P**atient/Population * **I**ntervention * **C**omparison * **O**utcome **Databases & Search Engines** * **Google Scholar:** A search engine that provides access to a wide range of scholarly literature. It is noted for having helpful search tools and cited by feature. * **PubMed:** A freely available search engine that accesses the MEDLINE database. It is the most comprehensive and powerful search engine for finding research evidence in healthcare. PubMed is useful for narrowing searches using techniques such as MeSH terms. * **TRIP Database:** A database that helps clinicians find relevant research and assesses the risk of bias using AI. It also helps to identify different types of research such as randomized clinical trials, systemic reviews, and clinical practice guidelines. TRIP also allows for PICO searches. * **PEDro (Physiotherapy Evidence Database):** A free database and search engine specific to physical therapy literature. * **APTA (American Physical Therapy Association):** A resource that provides clinical practice guidelines and clinical summaries for members. * **AHRQ (Agency for Healthcare Research and Quality):** A resource that provides clinical practice guidelines, some of which are not indexed elsewhere. * **Cochrane Library:** A resource containing systematic reviews. **Types of Research** * **Systematic Review:** A study that synthesizes information from multiple primary studies. * **Randomized Controlled Trial (RCT):** A study design where participants are randomly assigned to different intervention groups. * **Clinical Practice Guidelines (CPGs):** Recommendations based on the best available evidence, often developed by experts. * **Descriptive Research:** Seeks to describe a situation. These questions can be answered using cross-sectional or longitudinal designs. * **Predictive Research**: Aims to determine how well a set of measures can predict an outcome. * **Causal Research**: Aims to determine if an intervention causes a change in an outcome. **Additional Concepts** * **Efficiency:** Using search strategies to quickly locate relevant information. * **Full Text:** The complete version of a research article. * **Filters:** Tools to narrow search results by study type, language, or time since publication. * **Synonyms:** Different words that have the same or similar meaning and can be used to broaden a search. * **Impairments**: Problems at the level of body functions and structures. * **Activity limitations**: Problems with actions like walking or climbing stairs. * **Participation restrictions**: Problems with work, school, or community involvement. **Steps for Searching** * Identify a need for information and develop a focused, searchable clinical question. * List potential keywords and synonyms. * Reorder terms by importance. * Use the MeSH database in PubMed to refine your terms. * Enter terms individually into the search engine. * Combine terms using "OR" for similar concepts. * Combine concepts using "AND" to narrow search results. * Use filters as needed to refine search. * Save article links and search strategy. This briefing document should provide a useful overview of key concepts and vocabulary related to searching for research evidence. Week 3 To create a briefing document with key vocabulary, here's a breakdown of important terms and concepts from the sources: Key Concepts in Research Applicability & Quality Applicability: This refers to evaluating a study to see if its findings can benefit specific people or groups of patients with similar issues. It involves determining if a study is relevant to a particular patient or clinical question. "Close Enough": It is not always possible to find the perfect match between a patient and the evidence, so sometimes a study that is "close enough" is sufficient. Quality: This refers to the rigor of a study and if it has been conducted with sufficient controls and methods to produce valid results that can be confidently applied to inform patient care. Evidence-Based Practice (EBP): The process of using research evidence to inform care for patients. Study Design and Bias Randomization: Randomly assigning participants to different treatment groups to reduce bias and ensure groups are similar at the beginning of the study. Stratification: Dividing subjects into subgroups based on characteristics that might affect outcomes, and then randomizing within those subgroups. Blinding: Concealing the treatment assignment from participants (single-blind), clinicians (single-blind), or both (double-blind) to minimize bias. A sham group may be used to blind participants in studies where blinding is difficult. Sham Group: A control group that receives a fake treatment that appears like the real intervention, but it is not a real treatment. Placebo Effect: A phenomenon where a patient's belief in a treatment can produce a positive effect, even if the treatment has no real physiological effect. Intention-to-Treat Analysis (ITT): A statistical analysis where all participants are analyzed in the groups to which they were initially assigned, even if they did not complete their assigned treatment. Participant Attrition: Participants dropping out of a study before completion. A good study will report how many participants started, finished, and why people dropped out. Preservation of Original Groups: Keeping participants in their originally assigned groups, even if they switch treatments mid-study. Control Group: A group in a study that does not receive the treatment, or receives a standard treatment to compare against a new intervention. Comparison Group: A group in a study that receives an alternate treatment or intervention to compare against a new intervention. Study Factors Intervention: The treatment or procedure being studied. It should be clinically realistic. Outcome Measures: The tools used to assess the results of a study. They should be relevant to the clinical question and the patient's goals. Surrogate Outcomes: Indirect measures of a more useful clinical outcome. For example, cholesterol levels might be measured as a surrogate outcome for cardiovascular disease. Clinically Meaningful Outcome Variables: Outcomes that are important and relevant to the patient's goals and quality of life. Follow-up Duration: The length of time participants are followed in a study. It's important to consider both short-term and long-term outcomes. Realistic Intervention: Considering whether the intervention and control groups are receiving treatments that are feasible and practical in a real-world clinical setting. Sampling Consecutive Sampling: Includes all patients referred to a clinic with a specific problem. Selective Sampling: Recruiting participants based on referrals or other non-random methods. Sample Size: The number of participants in a study. A study needs a sufficient sample size to have enough power. Power: The likelihood that a study will detect a difference between groups if a difference exists. Measurement Construct: What you are interested in measuring Measure: How a construct is measured Subjective vs. Objective Measures: Measures are on a continuum of subjectivity depending on the extent of personal judgement involved rather than being wholly one or the other. Patient-Reported Measures: Ratings provided by the patient, often used to assess symptoms, impacts of a condition or psychological constructs. Observer-Rated Measures: Ratings from observations made by a clinician, such as physical capacity measures. Other Key Terms Bias: Any factor that makes you less certain that the results of a study are due to the intervention. Conflict of Interest: A situation where an author might have personal gain from a treatment. This vocabulary should be helpful in understanding and critically appraising research in the context of evidence-based practice. Here is a table summarizing different types of bias, as discussed in the sources: Type of Description How to Reduce Bias Occurs when participants are not randomly assigned to groups, or Selection Randomly assign participants to groups. Use a computerized when there is a selection process that favors one group over another, Bias randomization process with concealed group assignment. potentially skewing results. Arises when there are differences in how treatments or interventions Ensure treatments are equivalent in length and intensity. Make Performanc are administered, or when groups receive unequal care apart from sure all groups have the same experience other than the intended e Bias the treatment being studied. intervention. Mask or blind the evaluators who are measuring the outcomes to Detection Occurs when the person measuring outcomes is aware of the which treatment group each participant was assigned. The "treaters" Bias treatment group and this knowledge affects how they assess results. should not be the "testers". Report the number of participants who started and finished the Attrition Happens when there is a loss of participants during the study and study. Include reasons why participants dropped out. Use intention- Bias these dropouts are not accounted for or analyzed properly. to-treat (ITT) analysis. Occurs when the researchers present the data or interpret the results Reporting Authors should declare any conflicts of interest. Consider the in a way that is not truthful or accurate due to a conflict of interest or Bias funding sources. other incentives. Arises from differences in accuracy of reporting prior events due to Employ consistent methods of measuring and gathering information Recall Bias memory issues. across all groups. This can result from using a small sample size or using a selective Use consecutive sampling where all eligible patients are invited to Sampling sampling strategy where participants are not representative of the participate. Use a sample size sufficient to ensure adequate study Bias population of interest. power. Placebo A phenomenon where a patient's belief in a treatment can produce a Include a sham group where participants believe they are receiving Effect positive effect, even if the treatment has no real physiological effect. the treatment in order to control for this effect. Additional points regarding bias: Randomization is a key method to reduce bias. If the sample size is large enough, randomization ensures that groups will be similar in characteristics that could affect outcomes before treatment begins. Blinding is used to reduce bias by concealing group assignment. If participants are not blinded, there is a risk of a placebo effect. Intention-to-treat analysis is used to account for participant attrition and helps to maintain the integrity of the original group assignment. Sample size affects the power of a study. Studies with small sample sizes might not detect a true difference between groups. Conflicts of interest do not necessarily mean a study is bad, but they should be declared and considered when interpreting the results. Moving participants between groups after a study begins can bias results. Studies that use a control or comparison group are higher quality than studies with only one treatment group. Both subjective and objective measures are susceptible to bias; it is important to use reliable and valid measures. By being aware of these different types of bias, it is possible to critically evaluate the quality of research and make more informed decisions about patient care.  Week 4 A briefing document on key vocabulary related to data types and statistical analysis in research: Data Types Nominal Data: Represents categories or groups with no inherent order. Examples include treatment groups (A, B, control), biological sex (male, female), or presence of a condition (hip replacement vs. no hip replacement). Nominal data is often coded into numbers for statistical programs, but these numbers are arbitrary and do not indicate order or value. Ordinal Data: Categorical data with a meaningful order or ranking, but the intervals between categories are not uniform or meaningful. Likert scales, which use a range of choices (e.g., strongly disagree to strongly agree) are a common form of ordinal data. The focus is on the rank order, rather than the numerical value. Discrete Data: Data that can only take on specific, separate values, with a limited number of possibilities. The number of times someone has had COVID (e.g., 0, 1, 2, 3, 4, or 5) is an example of discrete data. Continuous Data: Data that can take on any value within a given range. There is an infinite number of possibilities between any two values. ◦ Ratio scales are continuous scales with equal intervals and a meaningful zero value, such as range of motion measured with a goniometer. ◦ Interval scales are also continuous with equal intervals, but the zero point is arbitrary, such as temperature in Fahrenheit or Celsius. Descriptive Statistics Descriptive statistics summarize and describe the main features of a dataset. They provide an overview of typical values and variability within and between groups. Measures of central tendency describe the "average" or "most typical" value of a dataset. ◦ Mean: The arithmetic average of a set of observations. It is calculated by summing all values and dividing by the number of values. The mean is usually reported for normally distributed data. ◦ Median: The midpoint of a dataset, where half of the observations fall above and half fall below. The median is usually reported for skewed data. ◦ Mode: The most frequently occurring observation in a dataset. Measures of variability (or dispersion) reflect how spread out or dispersed the data are. ◦ Range: The difference between the highest and lowest scores in a dataset. ◦ Standard Deviation (SD): The average amount that each individual score varies from the mean of the set of scores. Normal Distribution: A bell-shaped curve that represents the distribution of many variables when measured in a large sample. Skewed Distribution: A non-normal distribution where the data is not symmetrical. Inferential Statistics Inferential statistics use probability to interpret differences observed in research studies and draw conclusions about a larger population based on a sample. Null Hypothesis: The assumption that there is no difference between groups being compared. P-value: The probability that the observed difference between groups is due to chance. A small p-value indicates that the results are unlikely due to chance. Alpha Level: The threshold of probability that researchers are willing to accept for the results of a study being due to chance; typically set at 0.05. If a p-value is less than the alpha level, the null hypothesis is rejected. Confidence Interval (CI): A range of values that is likely to contain the true population mean. A narrow CI indicates that the sample mean is a more precise estimate of the true population mean. T-test: A statistical method used to compare the means of two groups. Analysis of Variance (ANOVA): A statistical method used to compare the means of three or more groups. Chi-Square (X²) A statistic used to analyze nominal data. It compares the observed frequency of a particular category with the expected frequency of that category. Clinical Relevance Effect Size: A measure of the magnitude of the difference between two groups. Cohen's d is a common measure of effect size. ◦ Small effect size: 0.2-0.5 ◦ Medium effect size: 0.5-0.8 ◦ Large effect size: greater than 0.8 Number Needed to Treat (NNT): The number of patients who need to be treated with an intervention to achieve one additional desired outcome compared to a control group. NNT is used for dichotomous outcomes. Minimal Clinically Important Difference (MCID): The smallest difference in a measured outcome that a patient perceives as beneficial. Study Design Blinding: The practice of ensuring that participants, therapists, and/or assessors are unaware of who is receiving the treatment or the control to minimize bias. ◦ Participant blinding: When the patients in the trial don't know whether they are in the treatment or control group. ◦ Therapist blinding: When the therapists delivering the interventions don't know which group they are in. ◦ Assessor blinding: When the researchers measuring the outcomes don't know which group the participants are in. Intention-to-Treat (ITT) Analysis: An analysis in which participants are analyzed in the groups to which they were originally assigned, regardless of whether they completed the study or adhered to the treatment. Key Points The type of data dictates the appropriate statistical analyses that can be used. Using the wrong statistical test can lead to invalid conclusions. Descriptive statistics are useful for summarizing data, but inferential statistics are needed to make conclusions about differences between groups. Statistical significance (p-value less than 0.05) does not necessarily mean that the results are clinically important. Clinical importance is determined by measures such as effect size and NNT, as well as patient values and preferences. Blinding in research helps to reduce bias. It is important to evaluate the reliability and validity of outcome measures used in research. When interpreting study results, consider whether the groups were similar at baseline, whether confidence intervals were reported, whether descriptive and inferential statistics were applied, and whether there was a clinically relevant treatment effect. Carefully review all p-values and consider the relevance of values between 0.05 and 0.15 as a pattern of results may be relevant even if the p values are above the conventional alpha of 0.05. This briefing document provides a foundation for understanding key statistical concepts. Remember that the best way to learn these concepts is to apply them to real-world research studies.  Week 5 Here is a briefing document with key vocabulary and concepts from the sources: Key Statistical Concepts Descriptive Statistics: Used to describe data, not to make comparisons. ◦ Mean: The average of a normally distributed data set. ◦ Median: The halfway point of a data set, used when data is skewed. ◦ Standard Deviation: Used to measure the variability of data in a normally distributed data set. ◦ Interquartile Range (IQR): Equivalent to standard deviation, but used for non-normally distributed data. ◦ Confidence Intervals: A range of values that you can be 95% confident contains the true population mean. Inferential Statistics: Used to make inferences about a larger population based on a sample. ◦ Null Hypothesis Significance Testing: A process to determine if there are differences between groups. ▪ Null Hypothesis: The assumption that there is no difference between groups. ▪ Alternative Hypothesis: The hypothesis that there is a difference between groups. ▪ P-Value: The probability of obtaining the study results if the null hypothesis were true. A p-value of less than or equal to 0.05 is often set as the threshold for statistical significance. ▪ Statistical Significance: When a p-value is less than a set threshold (often 0.05), indicating that the results are unlikely due to chance, but not necessarily clinically meaningful. ◦ Type I Error: A false positive, rejecting the null hypothesis when it is true. ◦ Type II Error: A false negative, accepting the null hypothesis when it is false. Often due to an underpowered study (not enough participants). Study Design & Analysis Sample vs. Population: A sample is a subset of a larger population, intended to be representative of that population. Randomization: Randomly assigning participants to treatment groups to ensure baseline characteristics are similar. Baseline Characteristics: The characteristics of participants at the start of a study; these should be similar between groups. Power Analysis: Used to determine the number of participants needed to detect a statistically significant effect. Pilot Studies: Small scale studies that provide preliminary data to inform power analysis and larger studies T-tests: Statistical test used to compare the means of two groups. ◦ Paired Samples T-test: Compares the change within an individual over time. ANOVA (Analysis of Variance): Statistical test used to compare the means of multiple groups. Correlation: A measure of the relationship between two continuous variables. Linear Regression: A statistical method to examine the relationship between multiple variables and a continuous outcome variable. Logistic Regression: A statistical method used to predict group membership based on multiple predictor variables. ◦ Binary Logistic Regression: Used to predict membership in one of two groups. ◦ Multinomial Logistic Regression: Used to predict membership in multiple groups Interpreting Results Change: The difference in a score on an outcome measure within a person or group over time. Difference: The difference in mean scores on an outcome measure between two groups. This can be called the treatment effect if the study is well designed. Clinical Meaningfulness: The size of a change or difference that is considered important from the patient's point of view. Treatment Effect: A comparative effect between groups based on an outcome measure. Responder: A patient with a large change in outcome but this language is misleading because they may have responded similarly to other treatments. Generalizability: The extent to which the results of a study can be applied to other populations. Important Considerations Cronbach's Alpha: Measures the correlation between multiple outcome variables. It's important that they are related, but not too closely related, which would be redundant. The 0.05 threshold for statistical significance is arbitrary. It's useful to look at the p-value itself rather than just deeming it significant or not. A statistically significant difference does not always mean a clinically meaningful difference. Within-group change is not the same as the treatment effect, as it can include natural history, regression to the mean, and nonspecific effects. Recruiting participants for research studies can be challenging. This can lead to smaller samples, and selection bias. Researchers should report how the sample was recruited, and the study's inclusion/exclusion criteria. These factors influence generalizability of study results. Week 6 Here's a briefing document with key vocabulary related to outcome measures, based on the provided sources: I. Outcome Measures An outcome measure is a characteristic or quality that is measured to assess a patient's status. They are used to track patient progress during physical therapy. Outcome measures can be divided into two main categories: ◦ Questionnaire-based measures: require the therapist to interview the patient or the patient to complete the questionnaire independently. ◦ Performance-based measures: require the patient to perform a set of movements or tasks. Outcome measures can be classified by the components of the International Classification of Functioning, Disability and Health (ICF), including body structures and functions, activity, and/or participation. II. Psychometric Properties Psychometric properties are the intrinsic qualities of an outcome measure, which include reliability, validity, and clinical meaningfulness. Reliability refers to the consistency of an outcome measure in producing the same score. ◦ Internal consistency: measures the consistency of results across items within the same test. ◦ Test-retest reliability: assesses the consistency of a measure when it is repeated on a patient who has not changed. ◦ Intra-rater reliability: the consistency of scores when one rater measures the same subject multiple times. ◦ Inter-rater reliability: the consistency of scores when multiple raters measure the same subject. Validity is an outcome measure's ability to measure what it is intended to measure. ◦ Content validity: establishes that a measure includes all aspects of the characteristic it intends to measure. ▪ Face validity is a less formal evaluation of content validity, based on the opinion of experts that a measure appears to measure what it intends to. ◦ Criterion validity: assesses the validity of a measure by comparing it to a more established measure. ▪ Gold standard: a measure with irrefutable validity. ▪ Reference standard: a reasonable comparison when a gold standard is not available. ◦ Construct validity: establishes a measure's ability to assess an abstract characteristic or concept. It includes: ▪ Convergent validity: performance on a measure converges with other measures that represent the same construct. ▪ Discriminant validity: performance on the measure diverges from measures that do not represent the same construct. Clinical Meaningfulness relates to whether a measure provides useful information in a clinical setting and if the change is important to the patient ◦ Floor and ceiling effects: occur when scores cluster at the lowest or highest end of a scale, limiting the measure's ability to detect change. ◦ Minimal Detectable Change (MDC): the minimum amount of change on a measure that exceeds expected measurement error. ◦ Responsiveness: an outcome measure's ability to detect change over time. ◦ Minimal Clinically Important Difference (MCID): the smallest change on a measure that is considered meaningful to patients. III. Study Appraisal Applicability: Determining if a study's purpose, participants, and methods are similar enough to your clinical question. Quality: Evaluating if the study design and methods are appropriate for the type of psychometric property being assessed. Interpreting Results: Understanding the statistical methods used and the clinical implications of findings. Clinical Bottom Line: Determining if a measure is sufficiently reliable, valid, and clinically meaningful for use in practice. IV. Key Concepts Related to Change Change refers to a variation in a patient's score on an outcome measure over time, often within the same individual or group. Difference refers to a variation in scores between two or more distinct groups of patients. The term 'effect' should be reserved for between-group difference and not used when talking about within group or person change. The term "minimally important change (MIC)" and "minimal clinically important change (MCIC)" are different from "minimal clinically important difference (MCID)". MCID refers to differences between groups. Cohen's effect sizes are proposed thresholds (0.2, 0.5, and 0.8) for small, medium, and large effects respectively. These are multiples of the standard deviation. Anchor-based methods compare changes in a measure to a patient's rating of their overall change. Distribution-based methods are based on the spread of data and the reliability of the measurement. V. Additional Points It is important to avoid using measures with unknown reliability and validity. A measure can be reliable without being valid, but a measure cannot be valid if it is not reliable. There are no perfectly reliable and valid measures. Cronbach's alpha is used to determine if items in a test are closely related and not redundant. DEXA (dual energy X-ray absorptiometry) measures bone density and body fat percentage. Spearman's rho correlation (r) is a common statistical method used in studies of outcome measure validity. Receiver operator characteristic (ROC) curve is used to determine the MCID. PubMed and Rehabilitation Measures Database are resources for finding research on outcome measures. This document should help you understand the key concepts and vocabulary related to outcome measures discussed in the sources.