Evidence-Based Practice Framework in Speech-Language Pathology

Overview of Evidence-Based Practice (EBP) in Speech-Language Pathology

• Conceptual Framework: The framework discussed is based on the Gillum and Gillum article, which provides a process for making clinical decisions driven by data and research critique.
• Origins: EBP was originally born in the field of medicine. It is the process physicians and surgeons employ to make decisions based on a review of information, clinical expertise, and the individual client.
• Historical Context:
  • Pre-Internet Era: Prior to the internet, research was difficult to access, requiring physical visits to university libraries to find bound hard copies of journals.
  • Old Decision-Making Model: Clinicians primarily integrated only two factors: the client’s desires/values and the SLP’s clinical experience.
  • Evolution: Focus shifted toward EBP due to ethical and legal standpoints. ASHA (American Speech-Language-Hearing Association) began publishing more evaluations of research designs and intervention effectiveness in their journals.
• The Four Pillars of EBP:
  1. Best Available Research: External evidence from peer-reviewed studies.
  2. Professional Judgment: The clinician's expertise and theoretical orientation.
  3. Client Values: Individual patient preferences, culture, and values.
  4. Employer Policy: A smaller but present factor. This includes workplace mandates for specific assessment protocols or expensive equipment (like specific speech-generating devices). Clinicians must ethicaly oppose policies that mandate non-evidence-based ‘fads’ or the ‘cheapest’ options over effective ones.

ASHA Resources for Evidence-Based Practice

• National Center for Evidence-Based Practice: Created by ASHA to address the void in efficient clinical decision-making resources.
• Evidence-Based Practice Compendium: An A-to-Z resource list. For example, under the letter ‘V’, clinicians can find resources for Vestibular and Voice disorders.
• ASHA Practice Portal: An evolving resource covering specific areas such as Traumatic Brain Injury (pediatric and adult), Aphasia, Autism, and Dementia. It is categorized into:
  • Assessment: Diagnostic methods and gathering treatment data.
  • Intervention/Treatment: Strategies for specific deficit areas.
  • Service Delivery: Documentation of intensity, duration, and caregiver training.

The PICO Process for Clinical Questions

• Origin and Utility: Adapted by Gillum and Gillum for Speech-Language Pathologists (SLPs), specifically for school-based SLPs treating child language impairments, though applicable to all populations. It is designed to be a fast and efficient method.
• Step 1: Creating the PICO Question: The goal is to ask the right clinical question to identify correct search terms.
  • P (Patient/Population): Specificity is key (e.g., "boys with mild autism and Down Syndrome" vs. "children").
    • Exception: If a population is too specific (e.g., Angelman Syndrome) and yields no research, the clinician should move to a broader categorical search (e.g., "intellectual disabilities").
    • Setting Context: In acute care, the population may be someone who just suffered a stroke, focusing on spontaneous recovery. A year post-stroke, the focus shifts to compensatory approaches.
  • I (Intervention): The strategy, technique, or Independent Variable (IV) being tested.
  • C (Comparison): Comparing two interventions (e.g., Vocabulary Strategy A vs. Vocabulary Strategy B).
    • Class Requirement: For the specific article critique assignment mentioned, the comparison can be a "nothing" or a "control group."
  • O (Outcome): The Dependent Variable (DV). This is the specific goal of therapy, such as increased vocabulary, swallow safety, or speech intelligibility.

Examples of PICO Questions

• Vocabulary Example: "Which intervention—direct instruction or computer-assisted instruction (I/C)—provided to school-age children with language impairment (P), results in the greatest improvement in receptive vocabulary (O)?"
• Narrative Language Example: "For school-age children with language impairment (P), does direct teaching of story grammar (I) or indirect teaching through listening to stories (C) result in greater improvement in narrative production (O)?"
• Refinement of Terms: Clinicians must use varied keywords/Boolean terms for language outcomes (e.g., discourse, narrative, story retell, listening comprehension, semantic maps, semantic webs, graphic organizers) because terminology is not standardized across professions.

External Evidence Search and Evidence Grading

• Search Databases: Academic Search Premier, Academic Search Complete, ASHA Journals, ERIC, and EBSCO. Community libraries (e.g., OhioLINK) are options for those not at universities.
• Macro-Level Critique (Evidence Grading): Categorizing the research design quality.
  • A++: Meta-analysis containing at least one randomized controlled trial (RCT).
  • A/B Range: Includes quasi-experimental studies, which are most common in the field.
  • F: Expert opinion or committee reports with no empirical data.

Critical Appraisal of Research (The ‘Meat’ of the Study)

Clinicians must evaluate five key areas to ensure objectivity, reliability, and validity:

1. Comparison/Control Group: Did the study compare the treatment group to a control group receiving nothing, or to a different intervention? This measures change over baseline.
2. Random Assignment: Participants should be randomly assigned to ensure groups are equivalent at the start. This prevents results from being skewed by "fluky" differences like differing intelligence or compliance levels between groups.
3. Participant Description: Authors must provide details on age, gender, race, ethnicity, socioeconomic status (SES), and severity of speech/cognitive skills. This allows the clinician to determine if the study participants match their specific caseload.
4. Baseline Equivalence: The treatment and control groups should be the same in all important ways except for the treatment itself.
5. Blinding:
   • Assessors: Should they know who is in which group?
   • Analysts: Should they know the source of the data?
   • Providers: To prevent bias in delivery (trying "harder" with the experimental group).

Statistical and Practical Significance

• Statistical Significance: Determines if the change from pre-test to post-test was real and not due to chance or sampling error.
  • The p-value: Represented by an italicized $p$.
  • Threshold: The standard is $p \le 0.05$. A value like $p = 0.00001$ indicates extreme significance.
  • Inference: If a study is statistically significant, the clinician can infer the same benefits would likely happen with their own caseload.
• Practical Significance (Clinical Significance / Effect Size): Measures the magnitude of change.
  • Purpose: Clinicians want the most "bang for the buck." A difference might exist (statistical significance), but it might be too small to be useful.
  • Measurements: Often reported as Eta squared or Cohen’s $d$.
  • Threshold: A value of $0.4$ or higher is generally sought. A value of $2.5$ is considered an excellent effect size.

Internal Evidence: Client and Clinician Factors

• Cultural Beliefs and Values: These can "trump everything."
  • Example: A clinician cannot recommend a speech-generating device or a "Speech Easy" to an Old Order Amish client whose culture prohibits electricity.
• Motivation: Important, but SLPs are generally skilled at motivating clients toward participation (e.g., choosing a tablet over Picture Exchange Communication).
• Financial Resources: Clinicians must consider family finances and insurance. In schools, the Individuals with Disabilities Education Act (IDEA) ensures Free Appropriate Public Education (FAPE).
• Education Level/Training: Clinicians must be trained in the EBP they choose to implement.
• Theoretical Orientation: Does the intervention align with the clinician's professional beliefs?

Data Collection and Implementation

• Implementation as a Mini-Research Study: Once an EBP is chosen, the clinician collects data in therapy to confirm the intervention is working for that specific client.
• The Standardized Test Debate: Standardized tests are designed to diagnose a disorder and establish a baseline, not necessarily to show progress over time.
  • Speaker’s Stance: Standardized tests may not be sensitive enough to detect monthly changes. Progress is better proven through therapy tallies and materials used in daily sessions.
• Final Decision: The clinician integrates external research, professional judgment, client values, and agency policy to make the best decision, then validates it through treatment data.