When things get messy: QI on the front lines David Frost MD FRCPC - - PowerPoint PPT Presentation

When things get messy: QI on the front lines

David Frost MD FRCPC Site Director, CTU Director, Toronto Western Hospital University Health Network, University of Toronto

CSIM Annual Meeting 2017

The following presentation represents the views of the speaker at the time of the presentation. This information is meant for educational purposes, and should not replace other sources

• f information or your medical judgment.

Learning Objectives:

• To illustrate QI principles and change management using real examples
• (Hopefully), you can learn from our experiences- positive and negative
• To appreciate the pitfalls of certain cognitive biases as they relate to QI
• To generate discussion and share experiences around real life QI

Disclosures

I have no conflicts to declare

Actually just one disclosure…

I am a clinician teacher with some local hospital administrative responsibility as site lead and CTU director for GIM. I have some medical leadership training, but zero quality improvement training. Fortunately I have colleagues who do…

Unintended consequences

https://youtu.be/vYglvcLwkK4?t=7s

4 projects from Toronto Western Hospital

• For each initiative,
• 1. Motivation for change
• 2. Stakeholder engagement and design of intervention
• 3. Intervention
• 4. Outcome measures and anticipated result
• 5. What really happened
• 6. What we didn’t expect
• 7. Course correction
• 8. Current status

1) GIM Ambulatory Strategy

• Replacement of longstanding,

model of 2 halfday clinics per weeks with residents and multiple attendings in each clinic, often simultaneously covering ward patients, accepting referrals from community family MDs

2) Acute Care of the Elderly GIM Unit

• Building ACE unit from ground

up, integrating with existing GIM service

3) General Internist embedded in Family Practice

• Situating a general internist in

the hospital’s Family Health Team to provide consultations 1/2 day per week

4) Non-Teaching GIM ‘team’

• Building a faculty-only
• verflow service to limit CTU

team size

Motivation for change

GIM Ambulatory Strategy

• GIM at UHN has had sustained increases in inpatient volumes on order of

~5-6% per year for about 5 years

• Pressure to decrease admissions
• Pressure to decrease length of stay without increasing readmissions
• Educational goals (Royal College) not being met
• Recognition that existing ambulatory structure did not meet needs

ACE Unit

• Increasing number of elderly GIM inpatients with complex

needs

• Evidence for decreased ALC days, decreased length of

stay, improved patient satisfaction elsewhere

• Variability between hospital wards in ability to respond to

behavioural and other issues in the elderly

GIM clinic in Family Practice

• Unnecessarily complex process for conventional referrals

between family practice and GIM

• Many patients with complex comorbidities discharged

from hospital; frequent readmissions

• Funding opportunity; sessional fee for specialist halfday at

family practice

Traditional workflow

Frost DW, Toubassi D, Detsky, AS. Canadian Family Physician 2012; 58: 825-28

Referral Form Appointment info given to FMD

• ffice

FMD office contacts patient with appointment details (FMD staff might be unfamiliar with details of consultant’s office)  sends patient to attend specialist appointment (probably unfamiliar

• ffice)

Consult letter (maybe) with plan that may o may not be what the FMD had been seeking

Family Physician

Clarification or more info requested

3 1 5 2 4 6 Consultant

Non-teaching GIM ‘team’

• Increasing CTU team census
• Concern re: service to education ratio
• Number of patients covered overnight
• Residency program to impose cap in team size

Stakeholder Engagement and Design of Intervention

GIM Ambulatory Strategy

• Hospital administration
• GIM physician staff (survey then retreat)
• Referring physicians (rounds attendance)
• Working group with physician and hospital administrative

representation

ACE Unit

• Allied health personnel, nursing, hospital administration,

physician leadership

• Patient representatives
• Site visits to other facilities with ACE units
• Spot audit of proportion on GIM patients who meet

proposed criteria (~30%)

• Steering committee with broad representation

(administrators, allied health personnel, nursing, physician)

GIM clinic in Family Practice

• Survey of family physicians at the site, soliciting

feedback on patient population, preferences around logistics, communications

• Informal discussion with family physician colleagues,

department head

Non-teaching GIM ‘team’

• GIM physician staff input
• Hospital administration

Interventions

GIM Ambulatory Strategy

• Discontinue previous ambulatory clinic structure
• Clinics 5d per week
• “Rapid Referral” from ED 3 x ½ days (no triaging, liberal

referral criteria- safe for discharge, internal medicine problem requiring urgent assessment, not better served by existing specialty clinic)

GIM Ambulatory Strategy

• Post-discharge clinic 2 ½ days per week- referrals from ward
• HTN clinic weekly
• Longitudinal GIM clinic weekly
• Dedicated attending for 2 or 4 week block in ambulatory clinic

ACE Unit

• Patients meeting (rather liberal) criteria are identified on admission

and assigned to designated wards; different admission order set used

• Automatic PT, OT, and SW consults
• Specific interdisciplinary planning meeting (above usual IP rounds)
• All nursing and allied health working on these 2 wards have

specialized geriatrics training

• Communal dining area, music therapy, recreation therapy

GIM clinic in Family Practice

• Patients with multiple medical problems who do not need more

subspecialty care, but more ‘complex care’, diagnostic dilemmas , undifferentiated presentations, patients in whom a ‘second opinion’ is requested, patients at high risk of presenting to ER with a general medical issue

• General internist is given an examination room and existing

clinic administration and nursing support (1/2 day per week)

GIM clinic in Family Practice

• Family MDs refer via EMR, specialist reviews request with access

to patients’ whole electronic chart

• Patient is booked in usual family practice, seen on site,

immediately dictate letter

• Direct communication by email with ‘bottom line’ from consultation

Non-teaching GIM ‘team’

• Pilot during 3 traditionally busiest months (Jan-Mar)
• An attending will be MRP for a 2 week period, and will provide all

necessary medical care to the patients admitted to the service

• Anticipate caring for around 16 patients on such a service, which

would decrease CTU volumes by around 4 patients per team

Outcome measures, anticipated results

GIM Ambulatory Strategy

• Number of referrals and patients seen, wait times, report from

ED physicians whether referral avoided, CTU team sizes, readmission rate

• Anticipated result: Diversion of stable patients away from

ward because of guaranteed prompt internal medicine

• assessment. Better followup leading to decreased

readmission rate

ACE Unit

• Improvement in several parameters- falls, catheter use,

ALC days, length of stay, patient experience, LTC as destination

GIM clinic in Family Practice

• Improved communication, better patient experience, more

prompt GIM assessments, increased referring MD satisfaction with consultation process

Non-teaching GIM ‘team’

• Number of patients diverted from CTU teams, attending physician

experience

What really happened…

GIM Ambulatory Strategy

• Worked nearly as anticipated (patients diverted from ED

and ward, high trainee satisfaction)

• Referrals (intended to be scanned and sent to clinic) often

never arrived; patients would arrive at clinic, and staff would not have reason for referral

• Billings not quite as high as anticipated

GIM Ambulatory Strategy

• Rapid referral patients needed more followup visits

than anticipated

• Lack of faculty longitudinality was at times a serious

problem

ACE Unit

• Admitting order set (paper) was 6 pages long and nobody used it
• Admitting order set was reduced to 2 pages long and nobody used it
• Admission criteria included the ISAR score- nobody knows what this

is

• The (few) patients who were admitted had decreased falls,

decreased catheter use, positive experience

ACE Unit

• We did not expect so few patients to be admitted to ACE
• Orientation to ACE unit for rotating residents was more difficult

than expected

GIM clinic in Family Practice

• High volume of referrals
• High degree of provider satisfaction
• Far more patients with single symptom for workup, diagnostic

dilemma than anticipated. In a consecutive sample of 100 referrals, NONE were referred for comanagment of complex comorbidities

• Charts in this practice became fully electronic

19/26 (73.1%) FHTs with specialists responded Perception of the FHT-specialist partnership compared to perception of other models of specialist consultation used (i.e. regular consultation of unaffiliated external specialists)

68% 26% Timeliness of consultation Quality of communication Quality of initial consult letter Quality of follow-up notes Access to specialist for questions/clarification Patient experience of care Engagement of allied health Much worse Slightly worse No difference Somewhat better Much better 53% 53% 61% 56% 58% 47% 47%

FHT Physician Perspective

Slide courtesy of Dr. K. Ng

Non-teaching GIM ‘team’

• Staff agreed to try this; 6 signed up to provide service
• Hospital was unwilling to provide resource in way of ‘physician

extender’

• Ours and other hospitals in same network approached

administration to fund similar initiative permanently

Course corrections

GIM Ambulatory Strategy

• Referral process revamped
• Clinic schedule changed to include more time for followups
• Dedicated longitudinal clinic removed
• Currently advertising for a faculty position to focus on

ambulatory GIM (plan for at least half of the clinic blocks to be covered by this individual)

ACE Unit

• Admission order set changed (did not help)
• Increased emphasis at orientation, signage
• Competition between teams to admit to unit with incentive
• Ability to transfer missed patients to ACE unit next morning

GIM clinic in Family Practice

• Consultation record is now typed directly into EMR as soon as

patient seen (rather than dictated)

• More e-consults via email
• Did not try to alter the patient population; just reminded referring

MDs that option to see a different population exists

Non-teaching GIM ‘team’

• At a staff meeting, expected benefit of pilot and risks in light of request

for resources analyzed in detail

• Pilot project abandoned

Current status

GIM Ambulatory Strategy

• Still going strong

ACE Unit

• Work in progress

GIM clinic in Family Practice

• Still going strong, but with different patient population / focus from

anticipated

Non-teaching GIM ‘team’

• Shelved for now

A word on psychology of change management

Loss aversion

• AKA “Endowment effect”
• “a good is worth more when it is considered as

something that could be lost or given up than when it is evaluated as a potential gain”- D.Kahneman (psychologist)

• Can be impetus to maintaining the status quo (cf.

“status quo bias”)

Confirmation bias

• Tendency to accept evidence that

confirms an established viewpoint (or hypothesis) and reject evidence that suggests the contrary

Sunk Cost Fallacy

• The tendency to continue an

endeavour, even to ongoing detriment, based on costs previously incurred

Sunk Cost Fallacy

Regression to the mean

• After an extreme measurement, subsequent

measurements tend to return to baseline

• Can lead to false attribution of a given result to

an intervention

Regression to the mean

Halliday R. Department of Health Analysis of Hospital Episode Statistics for England

Lessons learned

• The “status quo bias” is strong (and sometimes appropriate!)
• The relationship between hospitals and physicians is complex and

fragile

• Introducing operational changes in a teaching hospital that are

dependent on transient players to initiate are doomed to, if not fail, be extremely difficult to implement

Lessons learned

• Even with meticulous planning, expect a major surprise;

introducing changes to a highly complex system will produce an unanticipated result

• Expect that you will need to course correct, or to modify the goals
• f the program
• Don’t be “faked out” by regression to the mean
• Sometimes the correct move is abandonment, and the ‘sunk cost

fallacy’ is just that

Discussion, questions