Clinical Assessment & Intervention - MPK4001Y - University of Toronto PDF

MPK4001Y Clinical Assessment & Intervention Introduction & Course Outline I am Privileged White, male, cis-gender, straight, educated, happily-married, employed, healthy, two sons and four grandkids all of whom are well & provided for I acknowledge this privilege and pledge to try to m...

MPK4001Y Clinical Assessment & Intervention Introduction & Course Outline I am Privileged White, male, cis-gender, straight, educated, happily-married, employed, healthy, two sons and four grandkids all of whom are well & provided for I acknowledge this privilege and pledge to try to maintain awareness of how this position can affect my perception of the world lewisparker.ca  The Truth Comes First Learning Objectives What we study How we study it Why we study it www.condenaststore.com Outline Course Outline Clinical Assessment Clinical Intervention Why? Assessment of tibial torsion Course Outline You must READ THE SYLLABUS You are responsible for knowledge of every detail within it This intro will not cover all the details, just the big picture READ THE SYLLABUS Evaluation Mid-Term Test – 30% In-Person written test during normal class time on October 8 Includes / covers subjects in Modules 1-5 Short answer questions drawn from and/or similar to practice quiz questions Lab Quizzes – 30% Online quizzes after each of 10 weekly labs September 19 – December 6 – worth 3% each Submission via Quercus due by midnight Friday the week of the related lab Final Exam – 40% In-Person written test during normal class time on December 10 Includes / covers subjects in Modules 6-12 Short answer questions drawn from and/or similar to practice quiz questions Learning Outcomes Understand the paradigm and methods of clinical assessment Develop initial competence in some common clinical assessment techniques Develop competence in design and implementation of kinesiological intervention for individuals with health disorder Part 1 - Principles Modules 1-5 the clinical paradigm principles and methods of clinical assessment and intervention Tested on Mid-Term Test Part 2 - Application Modules 6-12 application of principles and methods to specific body regions or systems Tested on Final Exam Labs 10 weekly sessions from Week 3 (Sep 18-19) to Week 12 (Dec 4-5) Opportunity to practice a few of the most common techniques taught in the lecture that week Class divided into 3 groups Labs are taught by Sandy Heming, HBK, BScPT, Sport Cert PT, FCAMPT, RPT (see Quercus) – different times Head Therapist, MacIntosh Clinic, U of T Contacting the Instructors In Person – after classes, Doug’s Office Hour, or by appointment Quercus Discussions: About the Course – course structure Module Discussions – course content Quercus Inbox – confidential messaging system for personal issues Do NOT send email / text or phone The Clinical Paradigm Clinical Context MPK4001 Lecture 01 and Reasoning Outline The Clinical Context Clinical Assessment and Intervention Health Disorders and Manifestations of Disorder Clinical Reasoning the Bayesian view and the Confusion Matrix Theory vs. Practice livingwithmigraines.com Clinical Context Clinic / Clinical ≡ a “bedside” setting in health care [Latin clinicus = reclining < Greek kline = bed < Greek klinein = slope, lean or recline] Clinicians typically deal with individual clients sometimes families or other groups Clients seek advice on behaviour or other “interventions” Clinicians perform assessments and offer advice / interventions Clinical encounters & interventions may be preventive or reactive Clinical Assessment Assessment of individual health* / performance capacity* typically in a health care setting – but can be anywhere often in response to complaint of disorder – but not necessarily * I view these as equivalent, but recognize that not everyone agrees  Most often an assessment of “movement” capacity rather than “exercise” capacity – what does that mean? Historically involves observation using our senses, and does not involve much technology (that is changing!) Why Clinical Assessment? Several different situations or kinesiological contexts in which some amount of “clinical” assessment is required: Client cannot perform a movement or task – why not? Client has individual characteristics (e.g. - alignment) that warrant them performing an exercise differently. Client has NMSK pathology for which kinesiological intervention / rehabilitation may help. Note that the last of these tends to occur in a “clinic”, but not the first two! Why Clinical Assessment? Assessment of movement / exercise capacity typically starts with “functional” movements (common movement patterns involving many elements of our kinetic chains) or specific tasks What do you do if a person cannot? e.g. - cannot squat “properly” To know WHY they cannot perform a complex movement or task, clinical assessment is required PRoM of ankle dorsiflexion Assesses the components of kinesiological capacity Why Clinical Assessment? Perhaps the most common reason for a kinesiologist to perform a clinical assessment is in the context of kinesiological rehabilitation of neuro-musculo-skeletal (NMSK) pathology While clients may be referred with a diagnosis made by an MD, as an independent professional, you are required to perform your own assessment of that pathology Re-assessment used to monitor / ensure progress as expected Clinical Intervention Intervention via movement / exercise for individuals with impaired function / capacity Typically secondary (partly or completely curative of pathology) or tertiary (improve function without altering pathology) note that primary (entirely preventive) interventions may be delivered in clinical settings, but this is not typically the case This aligns with the MPK “Exercise As Medicine” area of concentration Health Disorders Definitions, classification, and types Health Disorders WHO definition of health = “a state of complete physical, mental, and social well-being; not merely the absence of disease or infirmity” In this context, a health disorder is anything that detracts from such a state of holistic well-being Typically, we consider a phenomenon to be a health disorder if someone complains that something is not normal (i.e. - “disordered”) complaint may be about themselves, or others any such complaint is of a manifestation of disorder Manifestations of Disorder Three types – names based on method of observation: Symptoms (observed by the “subject”) Signs (observed by the health care provider) “Lab” Tests (observed by inanimate technology)  Systematized Nomenclature of Pathology College of American Pathologists (1965) suggested that complete description of a health disorder requires specification of four qualities: Anatomy (structures involved) Pathology (morphology of disorder) Etiology (agents / cause of disorder) Function (states / manifestations) Anatomic Specification The anatomic structures involved in a disorder can be specified at various levels of anatomic hierarchy: body regions > specific structures > component structures, etc. e.g. – upper body > upper extremity > shoulder > gleno-humeral joint > glenoid labrum > antero-inferior labrum > labral enthes Diagnostic labels typically begin with less anatomic specificity and evolve to more specificity as more information is gathered over time Pathologic Specification Based on type / character of structural abnormality (i.e. - pathological anatomy), sometimes observable with just our senses, but often microscopic (e.g. - pathological cellular morphology) tissue failure (structural rupture / fracture / disintegration / dehiscence, etc.) inflammation (@ micro level = infiltration of WBC, chemical changes) tissue degeneration (changes in ultrastructural elements, e.g. - collagen) tumor / neoplasia / changes in cell morphology infection (presence of abnormal micro-organisms in or on tissue) et cetera Patho-Anatomic Labels Knowledge of anatomy and pathology involved is often combined into a single patho-anatomic label: e.g. – torn glenoid labrum, degenerate knee meniscus, cancer of thyroid, infected olecranon bursa, etc. These sorts of patho-anatomic diagnostic labels are the “holy grail” of medical diagnosis Additional specification of etiology and function completes the picture Etiologic Specification Specifies causal agents Physical agents thermal energy electromagnetic energy mechanical energy almost all injuries Chemical agents Biologic agents Functional Specification Specifies an individual’s functional state Typically includes abnormal values for some state variables; i.e. – manifestations of disorder Clusters of manifestations that commonly co-exist are often called clinical patterns or syndromes Clinical Labels of Health Disorders While SNOP requires all four axes of specification, clinical diagnostic labels can involve any number of them e.g. – “concussion” specifies etiology and one or more manifestations, but does not specify anatomy or pathology Summary clinical labels of a client’s situation evolve over time as more information is gathered e.g. – “anterior knee pain” of unknown pathology or etiology specifies anatomy (at a high level) and one manifestation, may later be re-labelled as patellofemoral arthrosis caused by lateral patellar tracking after further tests and analysis Types of Health Disorder In both common and technical language, we distinguish (to some extent) between several types of disorder: Injury Disease Dysfunction These words imply something about etiology or function  Injury Injury is a disorder caused by violence That violence may be: mechanical – virtually all of what we deal with electrical (i.e. electrocution) thermal (i.e. burns) chemical (i.e. poisoning, chemical burns) Further classified by temporal profile and severity Acute Trauma Single-repetition mechanical overload Causes sudden tissue failure Classified by severity of tissue failure old systems used Grade 1-3, or 1st-3rd degree now typically referred to simply as low-grade or high-grade Grades of Acute Trauma High-grade significant tissue failure with significant mechanical insufficiency (laxity, weakness, etc.) Low-grade very little failure not healed by spontaneous recovery no significant mechanical insufficiency Causes of Acute Trauma Acute mechanical violence causing failure of MSK tissues is almost always the result of excessive loads caused by one of two phenomena: Collision trauma – high external forces applied to the body by sudden contact with an external object Deceleration trauma – high internal forces generated by excessively rapid deceleration of the body or its segments Common Acute Trauma Dense Tissue Injuries (bones, teeth, hyaline cartilage) fractures, partial fractures caused by variety of loads - forces (compression, tension) or moments (bending, twisting) Soft Tissue Injuries Tensile – Strains (muscle-tendon), Sprains (ligaments), and Tears / Ruptures (both if complete or nearly so) Compressive - Contusions “Overuse” Injuries Cumulative plastic deformation caused by repetitive micro-trauma accompanied by insufficient spontaneous healing May (not) be accompanied by inflammation initially (usually) and/or ongoing (usually not) Reduced tissue strength predisposes to further injury at lower loads vicious cycle / toboggan ride downhill Disease “A disordered or incorrectly functioning organ, part, structure, or system of the body resulting from the effect of genetic or developmental errors, infection, poisons, nutritional deficiency or imbalance, toxicity, or unfavorable environmental factors; illness; sickness; ailment” (from dictionary.com) This is virtually all disorders other than injuries. We will also distinguish it from dysfunction without the external causative factors above. “Degenerative Diseases” of MSK Tissues Tissue that is dysfunctional because its ultrastructure (microstructure) has become deranged is said to be “degenerate” Tissue degeneration may be caused by: poor response to trauma (lousy tissue, continued overload, …) other factors: genetics, infection, toxins, ischemia, … some combination of these These are the most common NMSK disorders! e.g. – osteoarthrosis, tendonopathies, IV disc degeneration, etc. Dysfunction Dysfunction is a state of functional limitation (∴ a manifestation of disorder) often, but not necessarily co-existing with patho-anatomic (structural) disorder may have dysfunction without patho-anatomic disorder may have patho-anatomic disorder without dysfunction Examples? Take a break, Doug It’s a long class  Sitting still is bad for you – get up and move! Clinical Reasoning Estimating the probability of disorder based on manifestations (or vise versa) Differential Diagnosis Dia-gnosis [< Greek: to “know between”] – distinguish between possibilities Each clinical manifestation has a set of health disorders that can cause it Each health disorder has a set of clinical manifestations that it can cause The art / science of diagnosis involves navigating this two-way street to generate or dispel belief about presence of health disorders in individual clients “Diagnosis” is a controlled act; kinesiologists “assess” What if your assessment differs with your client’s MD’s diagnosis? …MPK4007 A Bayesian View of Diagnosis Each manifestation of disorder is associated with potential causes Probability of a health disorder given presence of a manifestation Written as P(HD|M), spoken as “probability of HD given M” Each health disorder has potential manifestations Probability of a manifestation given presence of a health disorder Written as P(M|HD), read as “probability of M given HD” Thomas Bayes (1701-1761) en.wikipedia.org These probabilities are called a posteriori conditional probabilities Manifestations of disorder Health Disorders P ( HD1 | M1 ) M1 HD1 P ( M1 | HD1 ) M2 P ( HD2 | M1 ) P ( M2 | HD1 ) HD2 P ( HD3 | M1 ) M3 P ( M3 | HD1 ) HD3 M4 P ( M4 | HD1 ) Bidirectional conditional probabilities Example: Low Back Disorders Manifestations Health Disorders 1 Back pain 1 Herniated disc 2 Leg pain 2 End plate fracture P(HD|M) HD1: HD2: P(M|HD) Disc EPF M1:.9.01 Back pain.6.95 M2:.03.00001 Leg pain.1.0005 The numbers are just guesstimates based on my experience  Example: Knee pain + Instability Suppose a person presented with knee pain and giving away of the knee. Possible causes of knee pain: Possible causes of knee instability: degenerative arthrosis torn meniscus torn meniscus torn ligament torn ligament loose body rheumatoid arthritis et cetera et cetera What is the intersection of these two sets of possibilities? Now look for other manifestations that will differentiate remaining possibilities, and repeat the process with new info. Diagnostic Confusion and Utility The table used in our “Low Back Disorders” example is based on a set of “diagnostic confusion matrices” (one for each manifestation-disorder pair) 2 x 2 matrix based on assumed knowledge of 2 things: existence of a specified manifestation in each person of a population existence of a specified disorder in each person of a population both scales are binary, i.e. true or false Several different statistics emerge from these 4 numbers; they reflect the diagnostic utility of a manifestation as a test for disorder Diagnostic Confusion Matrix Consider the number of people in a population of interest who have a manifestation of disorder (M+) or not (M-), and who have a particular health disorder (HD+) or not (H-) Four possible states ≡ “true or false positives or negatives” HD + HD - M+ TP FP (type II error) M- FN TN (type I error) Prevalence (PREV) Prevalence is the proportion of a population that has a disorder Ratio of the sum of the first column TP + FN PREV = (TP + FN) + (TN + FP) to the sum of both columns Accuracy (ACC) Proportion of test results that are correct for the entire population Ratio of the sum of true results TP + TN ACC = to the sum of both columns TP + FP + TN + FN Sensitivity (SN) test accuracy for those with disorder unaffected by PREV TP SN = TP + FN Specificity (SP) test accuracy for those without disorder unaffected by PREV TN SP = TN + FP Positive Predictive Value (PPV) test accuracy for those with manifestation ability of presence of manifestation to rule in disorder TP PPV = TP + FP strongly affected by SP and PREV Negative Predictive Value (NPV) test accuracy for those without manifestation ability of absence of manifestation to rule out disorder strongly affected by SN and PREV TN NPV = TN + FN Effects of Prevalence on DUS Prevalence does not affect SN or SP (each based on only one column) Prevalence is the strongest determinant of NPV and PPV effects of SN and SP on PPV and NPV depend on and can be largely over-ridden by effects of prevalence on those stats Prev  PPV, NPV Prev  PPV, NPV Relationship of SN, SP and PVs In general, highly SN tests have higher NPV high SN  low number of FN  most negative results are true this works best for rare disorders (even fewer FN because few PE+) need extreme SN (~1) to rule out prevalent disorders In general, highly SP tests have higher PPV high SP low number of FP  most positive results are true this works best for prevalent disorders (even fewer FP because few PE-) need extreme SP (~1) to rule in rare disorders Ruling Out Disorder To reduce the probability estimate of a health disorder to a very low value (effectively “ruling out” the presence of that disorder), we seek absence of negatively predictive manifestations NPV is increased by SN and reduced by PREV High NPV results from high SN and low PREV For very prevalent disorders, SN needs to be close to 1 for absence of manifestation to rule out disorder – hard to do! Ruling In Disorder To increase the probability estimate of a health disorder to a very high value (effectively “ruling in” the presence of that disorder), we seek presence of positively predictive manifestations PPV is increased by SP and PREV High NPV results from high SP and high PREV For very uncommon disorders, SP needs to be close to 1 for absence of manifestation to rule in disorder – hard to do! Beyond the Binary Confusion Matrix Aside from these 5 simplest measures, the confusion matrix yields a number of other more advanced measures of diagnostic utility (Fawcett 2006, Powers 2011) What if a manifestation of disorder is not a binary variable? Consider the simple diagnostic utility statistics (SN, SP, PPV, NPV, ACC) of a particular manifestation at different “cut-off scores”… Curve showing SN and SP for different cut-off scores ≡ Receiver Operating Characteristic (ROC) curve for this manifestation and disorder (Fawcett 2006) Example of an ROC Curve Consider the use of the Trails B test (standard neuropsych test) in the diagnosis of concussion: Confusion Matrices for 4 Cut-off Scores in Trails B 55 sec MTBI CTL 45 sec MTBI CTL + 15 2 + 28 6 - 63 43 - 50 39 35 sec MTBI CTL 25 sec MTBI CTL + 59 20 + 74 42 - 19 25 - 4 3 from Richards, Comper, Mainwaring, Hutchison (2008) Proceedings of the 4th International Meeting on mTBI in Sport Receiver Operating Characteristic Curve Calculate SN and SP for each of the four cut-off scores Plot curve of Sensitivity vs. (1-Specificity) each cut-off score generates a point Area Under Curve (AUC) AUC = 0.5  test is only as good as random guess AUC = 1.0  perfect diagnostic test  ROC Curve for Trails B and Concussion AUC = 0.693 from Richards, Comper, Mainwaring, Hutchison (2008) Proceedings of the 4th International Meeting on mTBI in Sport Combining Diagnostic Tests In the Bayesian view, we considered common clinical patterns, or syndromes, and the conditional probabilities of a PE given the existence of a pattern We can determine the measures of diagnostic utility that arise from the confusion matrix for a particular pattern or syndrome In such quantitative analyses, this typically involves using linear regression analysis to determine the most useful weighted combination of scores on multiple tests The test score based on such a combination is sometimes called a “discriminant function” ROC Curve for ANAM Fatigue and Concussion from Richards, Comper, Mainwaring, Hutchison (2008) Proceedings of the 4th International Meeting on mTBI in Sport ROC Curve for ANAM CodeSub and Concussion from Richards, Comper, Mainwaring, Hutchison (2008) Proceedings of the 4th International Meeting on mTBI in Sport  Example of a DF: ANAM and Concussion We developed several DFs DF3 had the best overall ROC linear combination of 7 tests DF3 is significantly better than any single test (SN,SP) = (.85,.97) or (.91,.86) at two best cut-off score choices from Richards, Comper, Mainwaring, Hutchison (2008) Proceedings of the 4th International Meeting on mTBI in Sport Theory vs. Practice Working in the real world…  “In theory, theory and practice are the same. In practice, they are not.” Albert Einstein wikipedia.org In the Real World / Clinical Settings… The HUGE set of values involved in this process (all the conditional probabilities and diagnostic utility statistics for every possible combination of single or multiple manifestations in relation to every disorder) is very sparsely populated with knowledge Clinicians know specific numeric values of these stats for a limited set of manifestation-disorder combinations in their area of expertise Clinicians have a vague / semi-quantitative notion of many of these values, on a scale from “very common to very rare” Clinicians learn to swim in the Sea of Ignorance, or they drown in it Swimming in the Sea of Ignorance When we don’t (really) know what we are doing (very well), we use non-quantitative techniques to guide our assessments (and interventions – more on this in MPK4007): Heuristics – [< Greek – to find or discover] = simple “rules of thumb” Guidelines – like heuristics, typically not as simple Whether heuristics or guidelines are based on evidence is a different question Consider levels and types of evidence… Levels of Evidence The concept of “evidence-based medicine” originated in Canada “Levels of evidence” (really types and levels) first described by the Canadian Task Force on Periodic Health Examination (CTFPHE) Many systems, all have more-or-less the same hierarchy, but with different numbering or lettering of the levels CTFPHE USA National Institutes of Health Oxford others “Evidence” in Clinical Contexts There are several qualitatively different ways in which the word “evidence” gets batted around in clinical settings “Evidence-based” (dogma based on observations of large groups) Truth claims supported by clinical evidence in large numbers of subjects Truth claims supported by deduction from evidence-based theory These are very different; e.g. – core stabilization exercise Individual evidence (dogma based on observations of a single client) While we can’t do sufficient statistical analysis to make claims of causation with much certainty, we typically allow such conclusions Summary and So What? What you should have learned today Why it matters Summary Clinical health care involves caring for individuals in various settings, often when they have complained of manifestation of disorder Health disorders can be classified multiple ways complete specification involves patho-anatomy, etiology, and (dys)function Clinical reasoning considers the probabilistic relationship between disorders and manifestations of disorder Bayesian reasoning is used to determine the disorder(s) most likely responsible for evident manifestations Bottom Lines Tests that are accurate in individuals with pathology (SN tests) tend to have high negative predictive value (NPV), lessened by prevalence; so we use negative results in these tests to rule out disorder Tests that are accurate in individuals without pathology (SP tests) tend to have high positive predictive value (PPV), lessened by rarity; so we use positive results in these tests to rule in disorder So What? Kinesiologists need to understand the clinical paradigm (health disorders, manifestations, clinical reasoning) Clients may consult for the purpose of caring for a disorder Clients consulting for other reasons may nonetheless have disorders or characteristics about which a Kinesiologist should be aware (in order to modify interventions or refer for other professional care) Clinical assessment begins with an understanding of this paradigm Confused? Study The Matrix ! Then take a red pill… or maybe a blue one?