6/17/2019

1

Digitizing Peer Review

Using the EHR for Chart Review

Charles Kitzman, MS/MBA

Chief Information Officer – Shasta Community Health Center

Redding CA

Background Information

History/Vitals

Problems/Goals

Implementation

Lessons Learned/Process Improvements

Current State/Feedback/Analysis

Questions and Answers

Agenda

[Use this space to place an image that

best captures the main idea of the

speech]

Located in Redding, CA Shasta County Est. 1992

ACGME Teaching Health Center

Primary Care Residency/NP/PA Fellowship

40K Unduplicated Patients/Year ¼ of Shasta County

93% of Patients live below Federal Poverty level

Homeless, Behavioral Health, Dental, Pediatrics, Ryan

White, Specialty Care, Maternity, ISAP, MAT.

Innovation Hub – Center of Care Innovations

PCMH Level 3 Certified in All Sites

2015 HRSA OSV 19/19 score

Vitals – About SCHC

6/17/2019

2

Mission and Staff

Shasta Community Health Center’s mission is to provide quality

health care services to the medically underserved populations we serve and to improve the overall health of our community.

The Paper Process

[use this space to place an image that best

captures the main idea of point #1]

Lack of engagement

Time consuming

Low volume

Infrequent results

Often difficult to assess care with just one note

Goals Stay in line with Quarterly

schedule

Improve data collection

Increase engagement

Decrease “burden”

Provide “longitudinal” look [use this space to place an image that best

captures the main idea of point #2]

6/17/2019

3

Some Considerations

Sharing more data than ever

Our data has “legs” (ED, PH, Pharmacy, HIE)

Coordination of care increasingly important

Process becomes transparent

Implied bias?

Regional HIE efforts

SacValley MedShare

19+ Live Data Contributors

750K/2.2M lives in Service Area

EDIE/VA/Public Health

The Survey Itself

[Use this space to place an image that best captures the main idea of the action step]

6/17/2019

4

Questions

Identify the Correct Encounter

Identify the Correct Type of Visit

The System highlights/selects the

encounter for the end-user to make

things easier

Questions (Continued)

Evaluate the Chronic Problem List

Evaluate the Completeness and

accuracy of the Medication List

Questions

Assess Functional Status

Evaluate Physical Exam

6/17/2019

5

Questions

Lab selection and timing

Diagnosis, Integration of

Clinical Information

Questions

Development/Execution of

Treatment plans

Questions

Communication, Education and

Access to Care

Clinician/Patient

Clinician/Consultants

6/17/2019

6

Questions

Overall Quality/Coordination

Dental? Integrated Behavioral

Health? Specialty Consults?

Some early changes….

Eliminated a question that asked “Would

you recommend this provider for a family

member?”

Added comment boxes to all questions

Added rubrics for several questions

Force comment if a score of poor is given

Other than that, no substantive changes

to the process.

The Process/Tools

NextGen EHR platform

System Practice Templates for Configuration

SQL Server Reporting Services (Enterprise)

SQL Jobs

Survey Monkey (Premier Plan)

STATA analysis tool

Chart Review M&M Committee

6/17/2019

7

System Practice Templates

Designed to give clients real-time control

over template features and processes

Part of the core software design

System allows for creative additional

functionality

Benefits/Liabilities

We have vigorously exploited this

tremendous opportunity

Manage Peer Review Template

3 Panels for 3 Processes to

control:

Who will be reviewed?

Which charts should be

selected/Removed?

Who will do the reviewing?

Who will review?

We cannot expect all providers in our practice to participate.

Telemedicine, Specialists, Moonlighters, Per Diem etc.

Most of the reviewing is done by FT/PT staff.

Requires Coordination/Process with HR to add/remove users.

Employee Number Field/Sys Admin is leveraged for categorization.

6/17/2019

8

What charts should we use?

Used to remove

confidential charts

VIP – Board Members,

Senior Leadership, etc

Can use this to also

evaluate scribes to a

degree

Schedule

Each provider/reviewer gets one task sent

every Monday.

They can open it and complete it whenever

they like so long as all Peer Review tasks are

done prior to the end of each quarter.

We track for compliance and the CMO

receives a quarterly report.

Reports are usually emailed to providers and

medical directors within two weeks after the

end of the quarter.

6/17/2019

9

Early Lessons Learned

74% Completion Rate

AVG time to complete 6 minutes

This has helped us identify systemic

issues in our processes. (Ex. “Not

taking” med reconciliation)

Dental integration means even

more eyes on the chart than usual

– mutual professional grace

Rubrics help – changes are

coming

6/17/2019

10

Med Reconciliation

Poor Scripting by MA staff

MA’s can’t stop meds

Document Not taking

No follow-up reason

Difficult to erase

Identified as a systemic issue

Resolved by new workflow

Informatics built new tools

Results

The Road ahead…

Change is hard – questions and weighting can be

compromised

Aggregate reporting for administrative purposes/trending

More emphasis on coordination of care

6/17/2019

11

Breakout Activity 1

Review the existing questions in the SCHC Chart Review.

Work together to find a suitable replacement question for the “Communication between

providers” question.

The question must have the same 6 options for response. Excellent, Good, Adequate, Poor, Very

poor and N/A.

Create a rubric for the question. What does excellent look like? Poor? Etc.

Anecdotes

“It definitely saves time as far as the reviews. Knowing who did the encounter does

influence my review a bit, even though I try not to be biased. I could usually figure out

when it was redacted though anyway.”

“I think the digital chart review is an excellent system. Once I learned the system, I can

quickly scan the chart looking for the pertinent information and then fill out the

questionnaire. It has tremendously increased my efficiency and I think it has helped us to

obtain good information in the most efficient and “pain free” way possible. I’m very

thankful for it.”

“Easy and fairly efficient way to incorporate into my weekly workflow.”

“I like it, it’s easier than looking at the paper charts and less wasteful.“

“Accessible, Fits into the provider’s workflow, Need two screens for convenience, Great

addition to our process, So glad to be rid of paper in this process

6/17/2019

12

Assessing Quality from Review

of Medical Records

Robert Moore MD MPH MBA

Chief Medical Officer, Partnership HealthPlan of California

Objectives

Understand the underlying causes of diagnostic error

Identify ways electronic Peer Review Tools can detect these errors

6/17/2019

13

Six Domains of Quality

Safe

Effective

Patient Centered

Timely

Efficient

Equitable

Crossing the Quality Chasm, 1999,

Institute of MedicineWhat is missing?

Diagnostic Accuracy

Up to 50% of clinical diagnoses are inaccurate, depending on the

type/prevalence of the problem, the degree of workup, the time

since initial presentation.

Autopsy series of patients with serious conditions and secret

shopper visits to primary care clinicians both find fundamental errors

10-20% of the time

Source: Mark Graber: “The Incidence of

Diagnostic Error in Medicine,” BMJ Oct 2013

Cause of Misdiagnosis

The human brain, dealing

with uncertainty!

6/17/2019

14

Intuition vs. Data in decision making

The other side of the coin: over-ordering diagnostic tests

Choosing Wisely campaign

Most Common Errors in Health Records

1. Inadequate history and physical examination (Fast thinking)

2. Prematurely anchoring on a common diagnosis, ignoring data not

consistent with this

3. Missing key findings generated by support staff

4. Illogical electronic medical record notes

6/17/2019

15

Common Errors I: Inadequate history and physical exam

Reason for Visit: Back Pain

History: 28 year old former construction worker on disability with a history of chronic lower back pain for the past 10 years, here to establish care with a new doctor. Pain is present constantly, worse with activity, helped by taking hydrocodone.

Review of systems: No cough, headaches, chest pain

Vitals: BP 130/90 Pulse: 80 Weight 240 lb.

Chest: clear to auscultation

CV: regular rate

Abd: Soft

Assessment: Back Pain

Plan: CT Scan of Lumbar Spine

Common Errors II: Premature Anchoring

Reason for Visit: Painful urination

History: Othewise healthy 26 year old woman with 1 week history of burning with urination. Denies frequency, urgency, abdominal pain. Had UTI diagnosed 1 month ago.

ROS: No nausea, vomiting, diarrhea, fever, no history of kidney stones

Exam: BP 130/90 P: 80 Wt. 170 lb

Chest: clear to auscultation

CV: Regular rate, no murmur

Abd: Soft, non-tender, no hepatomegaly. Flank: No pain.

Urinalysis: Neg WBC, RBC: 1+, Ketones 1+, Urobilinogen negative, Nitrite negative

Assessment: Urinary tract infection

Plan: Ciprofloxacin 400mg PO qday for 3 days

Drink plenty of fluids

Common Errors III: Ignoring Data Collected by Medical Assistants

Reason for Visit: Chronic Pain

History: 30 year old man with history of chronic pain, taking MS ER 15 mg twice a day, here to get his quarterly visit to pick up a prescription for his medication. Taking Medication on schedule, uses acetaminophen for breakthrough pain. No tobacco, drinks 1-2 glasses of wine per day. Depression under control with SSRI. CURES report shows only MS ER prescribed by PCP

Exam: BP: 210/130 P: 100 Weight: 180

PHQ9 (done by medical assistant): 20

Exam: No acute distress, ambulates well, normal affect

Chest: CTA CV: Regular rate

Back: mild paraspinous muscle tenderness

Lab: Urine toxicology screen positive for morphine only

Assessment: Chronic Pain, Depression

Plan: Refill MSER 15mg #56, x 3 months (3 triplicates given). Refill 25 mg Sertraline #90, refill x3

Medication use agreement updated.

6/17/2019

16

Quality of Documentation: Worse with Electronic Health Records

Human Decision Making: Heuristics and Biases

Heuristics (Mental rules of thumb):

1. Affect heuristic

2. Anchoring heuristic

3. Availability heuristic

4. Representativeness heuristic

5. Commitment heuristic

1. Belief bias

2. Confirmation bias

3. Optimism bias

4. Hindsight bias

5. Framing effect

6. Loss aversion

7. Narrative fallacy

8. Regression fallacy

9. Planning Fallacy

10. Halo Effect

11. The law of small

numbers

12. “What you see is all

there is” bias

Biases

Breakout Activity II

Four groups

Each assigned one of the 4 most common problems encountered in medical records

Discuss what your health centers have done to address the issue (if anything)

How can an automated peer review process be leveraged to address this:

Chart selection

What question asked of the reviewer; how is question phrased?

6/17/2019

17

Disease Focused Peer Review

Example:

Diagnosis: Acute Bronchitis (with no history of COPD or immune compromise)

Standard of Care: No antibiotics prescribed if no COPD or other co-morbidity.

Screen for: Incorrect Diagnosis used (should have use COPD exacerbation or pneumonia, for example)

If diagnosis correct, and no co-morbidities: screen for prescription of antibiotics

Breakout Activity III: Disease Specific Peer Review

Same exercise for all tables

Work together to answer the following:

1. List conditions in which there is clear evidence that a specific action should be done or should not be done. (Hint: consider both the most common reasons

for outpatient visits and rarer conditions where failure to perform an action could have catastrophic consequences.)

2. For each condition, list one or more parameters that could be objectively evaluated by a peer reviewer.