AHRQ PSIs and IQIs in National Pay for Reporting (Text Version)

Slide presentation from the AHRQ 2009 conference.

On September 14, 2009, Shaheen Halim, Ph.D. made this presentation at the 2009 Annual Conference. Select to access the PowerPoint® presentation (355 KB). Plugin Software Help.


Slide 1

AHRQ PSIs and IQIs in National Pay for Reporting

September 14, 2009
AHRQ QI Conference

Shaheen Halim, Ph.D.
Centers for Medicare & Medicaid Services

Slide 2

CMS' Office of Clinical Standards and Quality

  • Lead on quality and clinical issues and policies for the Agency's programs. Coordinates with external organizations and Agencies.
  • Promote and monitor quality and quality improvement for the Agency's programs. Evaluates the success of interventions.
  • Develop, evaluate, adopt and support performance measurement systems (quality indicators) to evaluate care provided to CMS beneficiaries.

Slide 3

OCSQ's Quality Measurement and Health Assessment Group

Lead for measure development and public reporting of quality measures:

  • Hospital Inpatient and Outpatient.
  • Physician.
  • Nursing Home.
  • Home Health.
  • ESRD.

Websites available on http://www.Medicare.gov

Slide 4

Hospital Pay for Reporting Programs

Reporting Hospital Quality for Annual Payment Update (RHQDAPU)

  • MMA 2003: .4% APU to report 10 measure starter set.
  • DRA 2005: 2% APU to report expanded measure set . 44 measures for the 2010 payment determination.

Hospital Outpatient Quality Data Reporting Program (HOP QDRP)

  • TRHCA 2006: 2% 2009 APU for 7 measures.
  • 11 measures for 2010 APU, including 4 claims-based measures on Imaging Efficiency.

Slide 5

RHQDAPU Over the Years

  • 2004: 10 process measures (AMI, HF, PN, SCIP).
  • 2006: 21 process measures (11 added).
  • 2007: 30-day mortality measures, HCAHPS.
  • 2008: 30-day readmission, AHRQ measures, structural measure.
  • 2009: 2 structural measures.

Currently 46 measures in RHQDAPU Program.

  • 27 process, 15 claims-based outcome measures, 3 structural, HCAHPS.

Desire to expand outcomes measurement in RHQDAPU.

Slide 6

AHRQ PSIs and IQIs in RHQDAPU

  • 9 AHRQ Indicators were adopted into CMS' Reporting Hospital Quality Data for Annual Payment Update Program (RHQDAPU) in FY 2009 IPPS Rule.
  • Allows expansion of RHQDAPU program topics to include Patient Safety, Complications, and In-hospital mortality.

Slide 7

Patient Safety Indicators

AHRQ Patient Safety Indicators adopted:

  • PSI 4: Death among surgical patients with serious, treatable complications.
  • PSI 6: Iatrogenic Pneumothorax.
  • PSI 14: Postoperative Wound Dehiscence.
  • PSI 15: Accidental Puncture or Laceration.
  • PSI Composite: Complications/Patient Safety for Selected Indicators.

Slide 8

Complication/patient safety for selected indicators

  • PSI #03 Decubitus Ulcer.
  • PSI #06 Iatrogenic Pneumothorax.
  • PSI #07 Infection Due To Medical Care.
  • PSI #08 Postop Hip Fracture.
  • PSI #09 Postop Hemorrhage or Hematoma.
  • PSI #10 Postop Physio Metabol Derangmt.
  • PSI #11 Postop Respiratory Failure.
  • PSI #12 Postop PE Or DVT.
  • PSI #13 Postop Sepsis.
  • PSI #14 Postop Wound Dehiscence.
  • PSI #15 Accidental Puncture/Laceration.

Slide 9

Inpatient Quality Indicators

AHRQ Inpatient Quality Indicators adopted:

  • IQI 11: Abdominal Aortic Aneurysm (AAA) Repair Mortality Rate.
  • IQI 19: Hip Fracture Mortality Rate.
  • IQI Composite: Mortality for Selected Procedures.
  • IQI Composite: Mortality for Selected Conditions.

Slide 10

Mortality for Selected Procedures

  • IQI #08 In-Hosp Mort Esophageal Resection.
  • IQI #09 In-Hosp Mort Pancreatic Resection.
  • IQI #11 In-Hosp Mort AAA Repair.
  • IQI #12 In-Hosp Mort CABG.
  • IQI #13 In-Hosp Mort Craniotomy.
  • IQI #14 In-Hosp Mort Hip Replacement.
  • IQI #30 In-Hosp Mort PTCA.
  • IQI #31 In-Hosp Mort Carotid Endarterectomy.

Slide 11

Mortality for Selected Conditions

  • IQI #15 In-Hosp Mort AMI.
  • IQI #16 In-Hosp Mort CHF.
  • IQI #17 In-Hosp Mort Stroke.
  • IQI #18 In-Hosp Mort GI Hemorrhage.
  • IQI #19 In-Hosp Mort Hip Fracture.
  • IQI #20 In-Hosp Mort Pneumonia.

Slide 12

CMS 2009 Dry Run

  • CMS conducts "Dry run" for claims-based measures to provide methodology information about the measures prior to formal implementation.
  • Provide hospitals with an opportunity to provide CMS with feedback to inform implementation.
  • Opportunity to answer questions regarding measure methodology and calculations.

Slide 13

CMS 2009 Dry Run

  • Hospital-Specific Reports were generated and released to hospitals via their QualityNet accounts on February 27, 2009.
    • Hospital Specific Performance.
    • National, State, and Regional summary statistics.
  • Mock-report and Summary Statistics made available for download on QualityNet.
  • Began 30-day question and comment period to end April 2, 2009.
  • Webinars to provide further information and to answer frequently asked questions about the dry run.

Slide 14

CMS 2009 Dry Run

  • 2006 Inpatient Medicare claims (100% file)
    • Data obtained from Dartmouth Medical School.
    • 10 diagnostic and 6 procedural codes were reported.
    • No age restriction.
  • AHRQ PSI and IQI software v3.2
    • Excludes claims missing Age or Sex from all analyses.
    • Excludes claims missing other variables (e.g. Admission Source, Admission Type, Disposition Status, DRG, LOS, etc.) from the denominators of specific measures.
  • 3M™ APR™-DRG V3.2 Limited License Grouper software
    • AHRQ PSI and IQI software use this for risk adjustment.
    • APR-DRG software downloaded from the AHRQ website.

Slide 15

CMS 2009 Dry Run

  • Modifications to our claims data were required because the AHRQ software was designed for use with HCUP, not Medicare claims data
    • The levels for some categorical variables required reassignment (e.g. Admission Source, Race, etc.)
      • For example, Hispanic = 5 in our data, AHRQ software specifies that Hispanic = 3.
  • MDC (Major Diagnostic Category) was assigned using the CMS DRG version 23 and 24 Relative Weights files

Slide 16

CMS 2009 Dry Run

  • AHRQ PSI and IQI software
    • Defines the inclusion and exclusion criteria for each indicator.
    • Generates the numerator, denominator, observed, expected, risk-adjusted and smoothed rates for each indicator.
  • Indicators were reported as rates per 1000
    • Rate = 200 (per 1000).
    • Rate calculated by multiplying x 1000 = 200 (per 1000).

Slide 17

CMS 2009 Dry Run

  • Population reference: national rates based upon the HCUP State Inpatient Database (SID).
    • Includes 90 million discharges in 2002, 2003, and 2004 from the 38 states participating in the HCUP SID.
  • Equal weight option applied for the Composite Scores
    • In this case, each component indicator is assigned an identical weight based on the number of indicators. That is, the weight equals 1 divided by the number of indicators in the composite.
    • For example, 1/8 = 0.1250

Slide 18

CMS 2009 Dry Run

  • Hospital Specific Reports provided Observed, Expected, Risk Adjusted and Smoothed Rates.
  • Provided definitions for each of the 4 rates and guidance on how to interpret and use them.
  • State, National and Regional (HHS Region) summary statistics provided for comparison.

Slide 19

2010 Implementation

  • Initial display on www.cms.hhs.gov tentatively scheduled for January 2010.
  • Will be calculated using Medicare claims spanning July 1, 2007 to June 30, 2008 for the FY 2010 payment determination.
  • Hospital preview reports tentatively scheduled for November/December 2009.
  • Reporting on Hospital Compare tentatively scheduled for June 2010 using calculations spanning July 1, 2008 to June 30, 2009.

Slide 20

Future Implementation

  • Include Observed Numerator and Denominator in hospital previews.
  • Include Confidence Intervals for rates.
  • Consumer testing to inform future display and language for the AHRQ PSIs and IQIs on Hospital Compare.
  • Possible display of composites similar to current bucket approach for 30-day mortality and 30-day readmission.

Slide 21

Example Display

Image: An image of an updated readmissions table is shown.

Slide 22

Future Issues

  • Examine POA for future use.
  • Small N threshold.
  • Which rate (predicted or smoothed) for consumer display.
Current as of December 2009
Internet Citation: AHRQ PSIs and IQIs in National Pay for Reporting (Text Version). December 2009. Agency for Healthcare Research and Quality, Rockville, MD. http://www.ahrq.gov/news/events/conference/2009/halim/index.html