Use of Information Technology for Precision Performance Measurement an

AHRQ 2008 Annual Conference

Slide presentation from the AHRQ 2008 conference showcasing Agency research and projects.

Slide Presentation from the AHRQ 2008 Annual Conference

On September 9, 2008, David W. Baker, made this presentation at the 2008 Annual Conference. Select to access the PowerPoint® presentation (2.3 MB; Plugin Software Help).

Slide 1

Use of Information Technology for Precision Performance Measurement and Focused Quality Improvement

David W. Baker, MD, MPH
Chief, General Internal Medicine
Feinberg School of Medicine, Northwestern University.

AHRQ Annual Conference
September 9th, 2008

Slide 2

The Problem

  • We want to routinely measure quality of care for dozens of measures in outpatient practice and use this information to improve care.
  • Cost of chart abstraction problematic.
  • Administrative (claims) data inaccurate.
    • Need to capture medical and patient reasons for not achieving a quality measure.

Slide 3

The Solution?

  • Electronic Health Record (EHR) systems have the potential to routinely measure quality with a high accuracy.
    • Denominator (if diagnoses entered.).
    • Numerator (e.g., satisfied measure): meds, screening tests, blood pressure, etc.
    • Exceptions: diagnoses, allergies, lab abnormalities.
      • But most EHR systems (EHRS) do not have adequate tools to routinely capture medical and patient reasons.

Slide 4

EHR Facilitates Quality Measurement

The slide shows four bar graphs measuring the percentages of "Preventive Services," "Diabetes," "Cardiovascular Disease 1," and Cardiovascular Disease 2" from Q1 2006 through GIM Q2 2006. Results show:

  • Preventive Services:
    • Pap: increased slightly, then decreased.
    • Mam: stayed even, then increased.
    • Crc: decreased slightly, then increased slightly.
    • Pvx: decreased.
  • Diabetes:
    • Hba1c: increased slightly, then decreased.
    • Bp: decreased, then increased.
    • Ldl: decreased.
    • Asp: stayed even, then decreased slightly.
  • Cardiovascular Disease 1:
    • Bp: increased steadily.
    • Ldl: decreased.
    • Asp: increased slightly, then decreased.
  • Cardiovascular Disease 2:
    • Antilipid: increased steadily.
    • Mibeta: stayed even, then decreased.
    • Afibwarf: increased, then decreased.

Slide 5

Accuracy of Quality Measurement Using Only EHRS Data

Compared to Physician Review

  • Note: Persell SD, et al, Arch Intern Med 2006.
    Baker DW et al, Ann Intern Med 2007.

Slide 6

Automated Measurement vs. Hybrid Measurement

The table presents the results for "Automated," "After MD Review Percent," and "Percent Change" for various "Quality Measures." The results show:

  • Antiplatelet drug: 82%; 96%; + 14
  • Lipid lowering drug: 93%; 97%; + 4
  • Beta blocker: 83%; 90%; + 7
  • Blood pressure (BP) measured: 97%; 99%; + 2
  • Lipid measurement: 82%; 88%; + 6
  • Low density lipoprotein (LDL) control: 85%; 87%; + 2
  • Angiotensin-II converting enzyme (ACE) inhibitor: 85%; 89%; + 4

Slide 7

Conclusions

  • Overall, good agreement between quality measured by EHR data compared to MD notes.
  • Several factors limit accuracy of EHR measures.
    • Many pts did not actually have heart failure (HF), coronary artery disease (CAD).
    • Medications were not always documented, but especially problematic for aspirin.
    • Exclusion criteria less well captured.

Slide 8

Implications for Quality Improvement [QI]

  • As quality of care improves and specificity of "failure to comply" declines:
    • Differences in performance more likely due to differences in documentation than to true differences in quality of care.
    • Point-of-care alerts for individual patients are usually incorrect: MDs ignore alerts.
    • List of patients need outreach are mostly wrong: outreach expensive, inefficient.

Slide 9

UPQUAL: Utilizing Precision Performance Measurement to Improve Quality

  • Implement multi-component quality improvement intervention.
  • Aim to achieve ultra-high level of performance through more accurate performance measurement.
  • Use quality measurement system to drive focused quality improvement.
  • Note: Funded by the Agency for Healthcare Research and Quality: 1R18HS017163.

Slide 10

UPQUAL Study Team

  • Dave Baker, Steve Persell, Janu Khandekar, Russell Robertson, Tom Gavagan, Nancy Dolan.
  • Darren Kaiser, Dale Sanders, Tom Smith, Steve Smith, Sue Levi, et al from ENH IT.
  • Jason Thompson.
  • Elisha Friesema.
  • Note: The slide shows a logo for ENH: Evanston Northwestern Healthcare and the seal for Northwestern University Feinberg School of Medicine.

Slide 11

UPQUAL—Components

  • Audit and feedback to physicians.
  • Point of care alerts for quality measures which are not satisfied.
    • Allows easy review and ordering.
    • Allows documentation of medical and patient reasons for not ordering.
  • Medical and patient reasons sent to care manager and member of quality committee.
  • Monthly feedback on individual patients not receiving essential medications.

Slide 12

Quality Measures (18).

  • Coronary heart disease (CHD):
    • Antiplatelet therapy.
    • Lipid lowering.
    • Beta blocker-Myocardial Infarction (MI).
    • ACE/ARB-CHD+DM.
  • Heart failure:
    • Beta blocker-left ventricular systolic dysfunction (LVSD).
    • ACE/ARB-LVSD.
    • Anticoagulation-AFIB.
  • Hypertension control.
  • Diabetes:
    • HbA1c control.
    • LDL control.
    • Blood pressure control.
    • Nephropathy screen/treat.
    • Aspirin primary prevention.
  • Preventive care:
    • Mammography.
    • Cervical cancer screen.
    • Colon cancer screen.
    • Pneumonia vaccine ≥65 y.
    • Osteoporosis screen/treat.

Slide 13

Best Practice Alert

Slide 14

The slide shows a screen shot of Epic's Web site with a sub screen opened to a patient's chart review. Highlighted and flagged on the menu under "Physician," is "Best Practice."

Slide 15

Physician Sees Patient Who Needs Testing or Treatment

Slide 16

The slide shows a screen shot of Epic's Web site with a sub screen opened to "Best Practice." The screen shows four flagged items: consider cervical cancer screening, testing HBA1c, checking lipids in diabetes, and consider mammography. Under each of these is a date when these tests were last administered to the patient and a check box to open "SmartSet."

Slide 17

The slide shows the duplicate screen shot from the previous slide with four red arrows pointing to the checked check boxes for opening "SmartSet."

Slide 18

The slide shows the duplicate screen shot from the previous slides with a red arrow pointing to the "Accept" button.

Slide 19

The slide shows a screen shot of the "SmartSet" page. The page shows checks in check boxes for "Diabetes Mellitus" and "Screening mammogram [V76.12B]" under Diagnosis, and "Lipid Risk Panel" and "Mammogram Screening" under Orders.

Slide 20

The slide shows a duplicate screen shot from the previous slide with a red arrow pointing to the "Accept/Pend" button.

Slide 21

The slide shows a screen shot of the "Order Entry" page with a red box around the order for Lipid Risk Panel and Mammogram Screening and around the encounter diagnoses for Diabetes Mellitus and Screening Mammogram.

Slide 22

Physician Sees Patient Who Cannot Afford Medication

Slide 23

The slide shows a screen shot of a patient's chart review with "Best Practice" highlighted. A red arrow points to "Best Practice Alerts."

Slide 24

The slide shows a duplicate screen shot from the previous slide with a blue arrow pointing to "Not Done-Patient Reason, Cost" under the flagged BestPractice Alert, "Consider antiplatelet drug for CHD."

Slide 25

The slide shows a duplicate screen shot from the previous slide with a red arrow pointing to the "New page" icon next to the entry for "Not Done-Patient Reason, Cost."

Slide 26

The slide shows a duplicate screen shot from the previous slide with a red arrow pointing to a box where "Intolerant to aspirin. Cannot afford clopidogrel" is written.

Slide 27

  • Each week, care manager receives list of patients who refuse or cannot afford a recommended test or procedure which leads to outreach.

Slide 28

Physician Sees Patient Who S/he Thinks Has Contraindication to Medication

Slide 29

The slide shows a screen shot of a "BestPractice Alert" page. A red arrow points to "Not Done-Medical Reason" under "Consider beta blocker for HF with LVSD."

Slide 30

The slide shows a duplicate screen shot from the previous slide with a red arrow pointing to a box where "Symtomatic bradycard" has been typed.

Slide 31

  • Each week, physician reviewer receives list of patients who had a medical exception entered and reviews the chart.

Slide 32

Display of Medical and Patient Reasons for Not Meeting Goals for Chronic Conditions

Slide 33

The slide shows a screen shot of the "SnapShot" page where two, red arrows point to "Colon CA [cancer] Screening Q 10 Yrs, Modifier" and "Hepatitis B Vaccine." A red box surrounds the "Exceptions to Best Practice Alerts" where it shows under "Consider Beta Blocker for HF with LVSD" that it was "Not Done-Patient Reason, Non-Cost" and "Felt fatigued. Advised to take but declined." It also shows the "Due On" date and "Due Soon" date.

Slide 34

Preserving Physician Judgment: Removing Patients from QI Registries with "Global Exeptions"

Slide 35

The slide shows both a screen shot of the "Health Maintenance" page with red arrows pointing to "Due dates" and a sub screen shot of the "Category Select" page with a red box around "Stop All Reminders-Medical Reason."

Slide 36

Improving Quality for the Unseen Patient

Slide 37

Monthly List of Patients Sent to MD

Provider: Marcus Welby, M. D.

  • Name: DOE, JANE
    MRN: 123919
    DOB: 2/1/54
    Consider antiplatelet drug for CHD.
  • Name: JUAN, DON
    MRN: 999660
    DOB: 4/4/37
    Consider beta blocker for prior MI
    Consider ACE/ARB for CHD with DM.
  • Name: SMITH, ZORRO
    MRN: 139784
    DOB: 7/3/24
    Consider antiplatelet drug for CHD.

Slide 38

Preliminary Results from First Three Months of UPQUAL

Slide 39

The line graph presents "Aspirin for Primary Prevention in Diabetes."
The vertical axis, percent, goes from 0 to 90 and the horizontal axis, months, goes from 1 to 19. The results show that at month 14, after remaining steady, "Satisfied" began to increase, "Exceptions" continued to hold steady, and "Deficiencies" began to decrease.

Slide 40

The line graph presents "Anticoagulation in Heart Failure and Atrial Fibrillation."
The vertical axis, percent, goes from 0 to 80 and the horizontal axis, months, goes from 1 to 19. The results show that at month 14, after remaining steady, "Satisfied" continued to hold steady, "Exceptions" began to increase, and "Deficiencies" began to decrease.

Slide 41

Summary

  • Advanced quality measurement can be built into physician work flow.
  • Exceptions to quality measures can be used to drive focused QI activities.
  • Accurate quality measurement can inform the care of an entire panel of patients (both seen and unseen).

Current as of January 2009

Internet Citation:

Use of Information Technology for Precision Performance Measurement and Focused Quality Improvement. Slide Presentation from the AHRQ 2008 Annual Conference (Text Version). January 2009. Agency for Healthcare Research and Quality, Rockville, MD. http://www.ahrq.gov/about/annualmtg08/090908slides/Baker.htm

 

Current as of February 2009
Internet Citation: Use of Information Technology for Precision Performance Measurement an: AHRQ 2008 Annual Conference. February 2009. Agency for Healthcare Research and Quality, Rockville, MD. http://www.ahrq.gov/news/events/conference/2008/Baker.html