Aug. 21: Clinical Questions and Hypothesis Testing (Use objectives for test study)

evidence, patient, and clinical experience are 3 legs of EBP

The Process…

  1. Ask a question

  2. Find best evidence

  3. Evaluate evidence

  4. Apply information in combination w/ clinical experience and patient values

  5. Evaluate outcomes

Formulate relevant clinical questions

  • v

Asking Answerable Clinical Questions

  • begins and ends w/ the patient

Types of Questions

Background Questions: (no format used to create these questions)

  • intended to increase understanding of situation (GENERALLY ANSWERING QUESTIONS ABOUT GENERAL THINGS)

  • ask for general information about disorder

  • often focus on medical aspects rather than PT components

  • help understand clinical context of diagnosis

  • EXAMPLES:

    • “What are common side effects of Oxycodone?”

    • “How long will it take for the total knee arthroplasty incision to heal?”

    • “What are the characteristic symptoms of cerebral palsy?”

    • “Will it be possible to play baseball again after elbow surgery?”

Foreground Questions: (uses the pico format to pose questions)

  • help clinicians and patients/clients make decisions

  • arise from any aspect of PT management model

P-I-C-O: (used to help formulate a well-defined searchable question)

breaking question into components in order to identify relevant information for answering it

  • relates to specific aspect of patient/client management:

    • Diagnosis; Clinical Measurement/Examination ; Prognosis; Intervention; Outcomes

  • P: Patient, Population, or Problem:

    • What are the most important characteristics of the patient and their health status?

  • I: Intervention/Exposure: Iis THIS (intervention) more better than THAT (comparison)|

    • What main intervention are you considering (medical, surgical, preventative)?

  • C: Comparison:

    • What are the alternative bencmark or gold standards being considered, if any?

  • O: Outcome:

    • What is the estimated likelihood of a clinical outcome attributable to a specific disease, condition, or injury?

PICO- INTERVENTION?

  • P= Patient/problem

  • I= Intervention

  • C= Comparison or Control group

  • O= Outcome or Effect of Intervention

*********Example:

Research question: In obese older adults, is walking a more effective exercise than swimming for weight loss ?

P: Obese older adults

I: Walking

C: Swimming

O: Weight loss

  • Aspect of patient/client management model: Intervention question

Null Hypothesis (Ho)

  • statistical hypothesis

  • there wil be NO STATISTICALLY SIGNFICANT DIFFERENCE (or relationship) b/w groups in the study

  • any differences (relationships) observed are due to chance

    • If the null hypothesis is rejected, that means the study’s results are not likely to be due to chance

Example:********

Research question: In obese older adults, is walking a more effective exercise than swimming for weight loss ?

Null hypothesis: There will be no (statistically significant) difference in weight loss in obese older adults who complete a walking program versus a swimming program.

Aspect of patient/client management model: Intervention question.

Research (alternate) Hypothesis (Ha)

  • investigator expects to find difference (or relationship)

  • once committed to this expectation, you cannot change your mind

  • if research hypothesis rejected, that means that the study results are likely due to chance and not due to experimental condition

Hypotheses

Research hypothesis

  • The statement of the predicted relationship or difference between groups

    • Non-directional research hypothesis (not greater or less, just change)

    • Directional research hypothesis (defined change)

Example:******

Research question: In obese older adults, is walking a more effective exercise than swimming for weight loss ?

Non- directional research hypothesis: Amounts of weight loss will be different in obese older adults who complete a walking program versus a swimming program.

Directional research hypothesis: Weight loss will be greater in obese older adults who complete a walking program versus a swimming program.

Other examples of Research Questions

(related to different aspects of the patient client management model)

Example: Clinical Measure Question

  • P= Patient and /or problem

  • I= Clinical Measure

  • C= Comparison- usually a gold or reference standard

  • O= Outcome/ Test Accuracy/ Measurement validity, reliability, and responsiveness

Example: Clinical Measure Question

Is the Fitbit or the Apple watch activity monitor a more reliable measure of steps taken in overweight adults?

P: Overweight adults

I: Fitbit activity monitor

C: Apple watch activity monitor

O: Reliable measure of steps taken

Example: Clinical Measure Question

Is the Fitbit or the Apple watch activity monitor a more reliable measure of steps taken in overweight adults?

Null: Both activity monitors are reliable measures of steps taken in overweight adults.

Research:

  • Non-directional: There will be a difference in the reliability of the Fitbit versus the Apple watch in measuring steps.

  • Directional: The Fitbit will be a more reliable measure of steps than the Apple watch.

PICO- Diagnosis?

P= Patient and /or problem

I= Diagnostic Test

C= Comparison, gold standard or reference test

O= Outcome, accuracy of diagnosing (or classifying) the condition

Example: Diagnosis/Classification Question

  • Is the treadmill stress test or cardiac ultrasound a more accurate test to detect heart disease in middle aged adults?

P: Middle aged adults

I: Treadmill stress test

C: Cardiac Ultrasound test

O: Accuracy in detecting heart disease

Example: Diagnosis/Classification Question

  • Is the treadmill stress test or a cardiac ultrasound a more accurate test to detect heart disease in middle aged adults?

Null: There will be no difference in accuracy between the treadmill stress test and a cardiac ultrasound in detecting heart disease in middle aged adults.

Research:

  • Non-directional research hypothesis: The treadmill stress test and a cardiac ultrasound will demonstrate differences in accuracy in detecting...

  • Directional research hypothesis: The treadmill stress test will be less accurate than a cardiac ultrasound in detecting...

PICO- Prognosis?

P= Patient and /or problem

I= Predictive Factors

C= Comparison, often not present in Prognosis ?

O= Outcome/ What we are trying to predict

Example: Prognosis Question****

  • Does the # of steps a person walks per day predict the risk of a cardiac arrest (heart attack) in individuals with heart disease?

P: Individuals with heart disease

I: # steps walked per day

C: None

O: Accuracy of prediction of cardiac arrest

Example: Prognosis Question

  • Does the # of steps a person walks per day predict the risk of a cardiac arrest in individuals with heart disease?

Null: The # of steps a person walks per day does not predict the risk of a cardiac arrest in individuals with heart disease.

Research:

  • Non-directional: The # of steps a person walks per day does predict the risk of a cardiac arrest.

  • Directional: The more steps a person walks per day, the risk of a cardiac arrest decreases.

Foreground Questions: Practice

Diagnosis: which tests & measures provide most precise & accurate info with least amount of risk, cost, etc.

  • Example: Is the hip scour test more accurate than the figure-4 test for detecting hip osteoarthritis in a 55-year-old female accountant with hip pain?

P: adults w/ Hip pain

I: hip scour test

C: figure-4 test

O: accuracy in detecting hip osteoarthritis

Clinical Measures: assessment of measurement reliability, measurement validity, and responsiveness.

  • Example: Is a manual muscle strength testing a more reliable, valid, and responsive measure of quadriceps strength than a force gauge in a 43-year-old female with multiple sclerosis?

P: adult w/ multiple sclerosis

I: manual muscle strength testing

C: force gague

O: most reliable, valid, and responsive measurement of gradriceps strength

Clinical Measure: ability to capture relevant information or responsiveness to change in status

  • Example: Will the Minnesota Living with Heart Failure Questionnaire detect change following rehab in an 83-year-old female following congestive heart failure?

P: 83 y/o following congestive heart failure

I: Minnesota Living with Heart Failure Questionnaire

C:

O: detect change following rehab

Prognostic Factors: predictions about future health status of patients with particular indicators, predictors or factors.

  • Example: Is lower extremity muscle weakness a predictor of fall risk in a 76-year-old female with diabetes?

P: 76 y/o female w/ diabetes

I: lower extremity muscle weakness

C:

O: predictiveness of fall risk

Interventions: benefits and risks of particular treatment

  • Example: Is Yoga a more effective treatment technique than Mat Pilates for restoring core trunk stability in a 77-year-old male with right hemiparesis (paralysis) due to stroke?

P: 77 y/o male w/ right hemiparesis (paralysis) POST-STROKE

I: Yoga

C: Mat Pilates

O: Effectiveness of restoring core trunk stability

EXAMPLES PICO PRACTICE:

Is the five times sit to stand test a more valid and reliable measure of functional mobility than the timed up and go test in a 36-year-old female with traumatic brain injury?

P:

I:

C:

O:

  • Pt management model? Clinical Measure (comparing tests)

In a 24-year-old soccer player, is the anterior drawer test a more accurate test in detecting an anterior cruciate ligament injury (ACL) tear than a Lachman’s test?

P:

I:

C:

O:

  • Pt management model? Diagnosis

In older adults with Parkinson’s disease, does the Berg Balance Test predict falls more accurately than Timed Up & Go test?

P:

I:

C:

O:

  • Pt management model? Prognosis (prediction: looking at something in future is clue for prognosis)

Which is a better predictor of fall risk, muscle weakness or balance deficits, in a 76-year-old female with diabetes?

P:

I:

C:

O:

  • Pt management model? Prognosis (predictor)

Literature Search Strategies: