Outcomes Modeling

Meeting Aug 21, 2012

Endpoints

• Outcomes
  • VAS-Back Pain, VAS-Leg Pain
  • ODI Score
  • EQ-5D Score
  • Return to Work
  • Mortality: Hospital, 30-day, 3-month, 12-month
• Quality
  • Length of Hospital Stay
  • Estimated Blood Loss
  • Surgery Time
  • Hospital Readmission (all causes), 30 days, 3 month
  • Revision Surgery (all causes), 3 month, 12 month
  • Patient Satisfaction (1 vs. 2~4, or 1~2 vs. 3~4)
  • Morbidities
    • Major
      • Pulmonary Embolism (PE)
      • Stroke
      • Myocardial Infarction (MI)
      • Surgical Site Infection (SSI)
      • Surgical Site Hematoma
      • Motor Deficit
    • Minor
      • Deep Venous Thrombosis (DVT)
      • Urinary Tract Infection (UTI)
• Utilization
  • Arthrodesis per diagnosis

Covariates

• Demographics
  • Gender
  • Age
  • Race (6-levels)
  • Hispanic (Yes/No)
  • Education (5 levels)
  • Employment & Activity
• Practice Group
• Surgeon
• Risk Factors
  • Workers Compensation (Yes/No)
  • Disability Insurance (Yes/No)
  • Motor Vehicle Injury (Yes/No)
  • Employed (Yes/No)
  • Return to Work Plans (Yes/No)
  • Major Surgery in Past Year (Yes/No)
  • Comorbidity (Diabetes, etc, 5 in total)
  • BMI
  • Smoking (6 level)
• Clinical Variables
  • Principal Diagnosis (one of 5 diagnoses)
  • Predominant Symptom (Back/Leg Pain or Both)
  • Motor Deficits (Yes/No)
  • Ambulation (3-level)
  • Duration (<3m, >3m or Unknown)
  • Modic Endplate Changes (Yes/No)
  • ASA Grade (1~5)
  • Surgical Approach (Posterior or Anterior Only)

Reports Format

• Types of Reports
  • Overall report
  • Practice group report
  • Surgeon report
  • Decision aids
• Subgroup Strata
  • By Diagnosis
  • By Procedure
  • Employment and Type of Occupation
  • Practice Group
• Requests from ??
  • The first year report needs not to be risk adjusted.
  • Each site-specific endpoint should have a mean and distribution of the aggregate national data next to for context.
  • No crude site ranking needed.
  • Only the primary endpoints need graphic display.
  • Types of Reports
    • Surgeon Centered Report (for use with internal practice improvement)
      • Graphic displays of primary endpoints for the entire practice group and per diagnosis with national norms
      • Descriptive statistics on patient population compared to national norms
      • Time trends of primary endpoints
    • Practice Administrator Centered Format (for use with external entities)
      • Graphic displays of primary endpoints for the entire practice group and per diagnosis WITH and WITHOUT national norms
      • Treatment type per diagnosis with national norms
    • Patient Centered Format (for use with informed/shared decision making)
      • Easy to understand graphic display of endpoints per diagnosis for each site (to be shared with patients in clinic)
    • Comprehensive descriptive analysis of all variables (for detailed referencing)
      • Include detailed mean (SD) range of each variable recorded for the site's data
      • Include mean (SD) range of each variable recorded for the aggregate national data