Quality Improvement Practice Flashcards

Vocabulary-style flashcards covering the basics of medical quality, QI methods (PDSA, Six Sigma, Lean), process mapping tools, and data presentation techniques based on the lecture notes by Brent W. Bost, MD.

Quality of Care (IOM Definition)

The degree to which health services increase the likelihood of desired health outcomes and are consistent with current professional knowledge.

Healthcare Value Equation

$V = \frac{Q + S}{C}$ where $V$ is Value, $Q$ is Quality, $S$ is Service, and $C$ is Cost.

To Err is Human (1999)

An IOM report stating medical errors caused $45,000$ to $98,000$ deaths annually, exceeding deaths from Breast Cancer or Motor Vehicle Accidents.

Crossing the Quality Chasm (2001)

A report calling for action by DHHS to improve patient safety, cost, care fragmentation, waste/efficiency, and patient waiting times.

Six Aims for Improvement (IOM)

Safe, Effective, Patient-centered, Timely, Efficient, and Equitable.

Safe (National Agenda Aim)

Primum non nocere; meaning First, do no harm.

Shewhart Cycle

The PDSA cycle consisting of Plan, Do, Study (Control), and Act developed at Bell Labs to improve quality.

PDSA: Plan

The phase where objectives are stated, predictions are made, and a plan is developed to carry out the cycle.

PDSA: Do

The phase involving carrying out the test and documenting problems or unexpected observations.

Avedis Donabedian Framework

A theoretical framework for patient care evaluation categorized into Structure, Process, and Outcomes.

Structure (Donabedian)

Variables involving equipment, training, facilities, licensure, and people.

Process (Donabedian)

Actions involved in caring for patients such as clinical pathways, workflow, safety checklists, and timeouts.

Outcomes (Donabedian)

The results that occur as a consequence of healthcare delivery, including clinical outcomes and patient experience/satisfaction.

W. Edwards Deming's Systems Approach

A philosophy that the system accounts for $85\%$ of problems while people account for $15\%$, advocating to change the system but keep the people.

Triple Aim

Introduced by Berwick while at CMS to drive improvement in healthcare.

Six Sigma

A process seeking to achieve less than $6$ standard deviations of normal, resulting in an error rate of $0.003\%$ or $3.4$ defects per $1$ million opportunities.

DMAIC

The Six Sigma method: Define the problem, Measure the steps, Analyze data, Improve the process, and Control.

Lean Methodology

Modelled by Toyota; focuses on accelerating velocity and reducing cost by removing any activity that absorbs resources yet creates no value.

Kaizen Blitz

An intense process to effect change by bringing everyone involved together to analyze and resolve an issue.

7 Step Model for Process Mapping

1. Identify target, 2. Define Optimal Practice, 3. Define Current Practice, 4. Identify discrepancies, 5. Develop Improvement Strategy, 6. Assess effectiveness/cost, 7. Determine implementation.

Ishikawa Diagram

Also known as a Fishbone Diagram; a visual tool used to collect and organize potential underlying causes to a specific problem.

Affinity Diagram

A tool used after brainstorming to organize and summarize ideas into related groups of five to ten items to understand the essence of a problem.

Pareto Principle

$20\%$ of the cause (events) accounts for $80\%$ of the problem.

Pareto Chart

A graph showing the relative contribution of each cause to a problem, ordered from largest to smallest to help prioritize improvements.

Benchmarking

Comparing performance to an ideal external standard or comparing internal performance over time.

Bundled Measures

A grouping of interventions that should all be done to measure an outcome, such as SCIP (Surgical Care Improvement Project) or ERAS (Enhanced Recovery After Surgery).

Balancing Measures

Measures used to ensure quality isn't compromised elsewhere by a QI initiative, such as measuring $30$-day readmissions when seeking to decrease length of stay (LOS).

Run Chart

A graph of data plotted over time with the data set's median as the center line to help determine if a process shows non-random patterns.

Shift (Run Chart Rule)

Indicated by six or more consecutive points above or below the median.

Trend (Run Chart Rule)

Indicated by five or more consecutive points moving in one direction.

Control Charts

Sensitive process measures with Upper Control Limits (UCL) and Lower Control Limits (LCL) usually set at $3$ standard deviations above and below the mean.

Common Cause Variation

Variations that occur naturally and commonly; data remains within the control limits.

Special Cause Variation

Variations not expected to happen naturally; data points outside control limits are true outliers requiring investigation.

Big Data (Three Vs)

Characterized by high Volume, Variety (measures, wearables), and Velocity (accumulated at high speed).

Dark Data

Unstructured data such as scanned text, images, audio recordings, or biometric data.

Deep Data

Large amounts of data on a single individual, including genomic, lifestyle, and patient-generated data; key to Precision Medicine.