Ted E' Palen, PhD, MD, MSPH David Price, MD, FAAFP Colorado Permanente Medical Group Kristin Wallace

1
Ted E. Palen, PhD, MD, MSPHDavid Price, MD,
FAAFPColorado Permanente Medical GroupKristin
Wallace, MPHSusan Shetterly, MSInstitute for
Health Research, Kaiser Permanente Colorado
Decision Support Clinician Education using
Virtual Consults in Kaiser Permanente HeathConnect
April 27, 2009
2
Project Team

• Daniel Winn
• Marian Bailey
• Jannelle Briggs
• Mary Kershner, RN, BC, BSN

3
Background

• Health care requires coordination of complex
  information across multiple settings of care
• Traditionally Face-to-Face patient contact
• Generated referrals to specialists
• Hallway consults
• phone calls, faxes, and letters to/from
  consultants
• Now virtual practice era is rapidly arriving
• Email visits with patients
• Consults via
• Secure message consults
• Video/digital imaging conferencing
• Real-time
• asynchronous
• Skype
• Twitter?

4
Purpose

• Evaluate virtual consults use w/in KPHC to see if
  it
• improves (adult) primary care-specialty
  communication
• meets physician long-term education needs
• improves physician satisfaction
• improves patient satisfaction
• fosters efficient, safe, cost-effective,
  evidence-based care during transitions between
  primary specialty care

5
Project Aims

• Quantify of specialty consultations made VC
  TC (i.e. face-to-face) stratified by specific
  conditions.
• Classify compare the types of VCs c/w TCs.
• Determine the outcomes of virtual consultation
  requests.
• Was advice consistent with evidence-based
  practice guidelines?
• Did consulting provider follow the advice?
• Was a TC eventually needed?
• Cost analysis of virtual consults versus
  traditional consults
• Assess requesting clinician perspective on
  utility of VCs.
• Was the advice clinically useful?
• Did it take the place of a phone or in-person
  consult?
• Was the advice educationally useful?
• Did the advice change clinical practice?
• Determine patient satisfaction with virtual
  consults

6
Project Design

• Phase 1 Identify the cases (virtual and
  traditional consultations)
• Phase 2 Clinician Survey
• Phase 3 Patient Satisfaction Survey
• Phase 4 Analyze and disseminate results

7
Phase 1 Identify the CasesReferral Types June
1- Nov 22, 2008
8
Phase 1 Identify the Cases Referral Types June
1- Nov 22, 2008
9
Phase 1 Identify the CasesReferrals by Provider
and Department
p lt 0.001
10
Phase 1 Identify the cases

• 33,390 Referrals from all Sources in 25 weeks
• 27,932 unique adult patients
• 93.6 Traditional Consults (TC)
• 6.4 Virtual Consults (VC)

11
Phase 1 Identify the cases

• 22,391 Referrals from 294 Primary Care Physicians
  (25 weeks)
• 19,441 unique adult patients (59.5 female)
• 93.0 Traditional Consults (TC)
• 6.9 Virtual Consults
• 86.6 of these persons had a one referral
• 11.4 had 2
• 1.5 had 3
• 0.3 had 4

12
PCP (n 294) Referral Patterns over 25 weeks

• Not adjusted for PCP FTE, case mix, or panel size
• Some patients may have had both TC VC for same
  problem

13
Distribution of PCP VC Frequency During Study
Period
14
Referral to Department

• Top VC Departments
• Dermatology (18.3)
• Neurosurgery (13.5)
• Endocrinology (11.0)
• Neurology (7.9)
• Orthopedic Surgery (6.1)
• 19 other departments

• Top TC Departments
• Physical Therapy (28.7)
• Orthopedics Surgery (12.7)
• Cardiology (10.2)
• Dermatology (10.0)
• General Surgery (8.8)
• 20 other departments

15
Number of People Involved in Processing Referral

p value 0.003
16
Phase 1 Identify the cases

• 540 consult orders were randomly selected
  (equally divided between TC and VC) for the
  clinician survey, patient satisfaction survey and
  chart abstraction

17
Phase 2 Clinician Survey

• 89 of 205 unique physicians (43.4) responded on
  71 VCs and 58 TCs
• FP 54.6 vs IM 37.4 (x2 p .04)
• Return rate was 34.3 of 207 VCs and 27.1 of 214
  TCs (x2 p0.11)

18
Phase 2 Clinician Survey
19
Top 5 Departments Referred to
Rheumatology was 4 (7.4) and GI was tied with
endocrine and audiology for 5 (6.3) in the
overall n540 sample
20
Number of Reasons for Consult
21
Reason for Consult(multiple reasons possible)
22
Reasons for Consult(multiple reasons possible)
23
(No Transcript)
24
Reasons for Consult after Re-allotment of Other
25
Reasons for Consult after Re-allotment of Other
26
Workflow Issues

• Called 1st before making referral?
• TC 10.3 vs VC 4.2 (p ns)
• 2 nonresponses (1 each TC and VC)
• Only 2.4 overall said referral process
  disruptive to workflow, no diff. b/t VC TC
• n 123, 2 VC and 4 TC non-responses

27
Soliciting Patient Preferences

• 2/3 of physicians did not solicit patient
  preference for type of referral
• No difference between FP vs IM or TC vs VC
• Physicians said 2/3 of patients did not express
  preference for TC vs VC, 1/5 did (no difference
  b/t type of referral ultimately ordered.
• 14 nonresponse or missing
• N 123 (6 no response or dont remember)

28
Info received from consultant by time of survey
(2-3 wks)

• x2 p 0.0006 (including missing values)

29
Info received by time of survey use of that info
x2 p 0.80 TC vs VC n 68
30
Use of information from index referral in care of
subsequent patients

• 34 of 68 had either used info or at least
  somewhat likely to use info from index referral
  in subsequent patients
• x2 ns TC vs VC

31
Usefulness satisfaction w/ referral info
1 Heard back n33 (usefulness), n36
(satisfaction) remaining 35-32 missing 2 -
Heard back n14 (usefulness), n34
(satisfaction) remaining 42-22 missing
32
Usefulness satisfaction w/ referral info
33
Conclusions Physician Survey

• Consult reasons were similar for TCs VCs
• ?difference with larger N
• Both TCs VCs created minimal disruption in
  physician workflow
• Most physicians didnt ask patient preference for
  consult type didnt recall patient expressing
  preference

34
Conclusions (2) Physician Survey

• Referring physicians reported hearing back from
  VCs earlier than TCs
• Most info from both TCs VCs was used in care of
  index patient
• Half of referring physicians had already, or
  planned to, use referral info in subsequent
  patients (no difference TC vs VC)

35
Conclusions (3) Physician Survey

• Referring physicians viewed information from both
  TCs and VCs as useful
• Referring physicians were generally satisfied
  with the quality of the referral

36
Phase 3 Patient Survey

• 540 encounters identified to receive a patient
  satisfaction survey over the phone
• 419 patients were called after 121 opted out
• 267 completed the patient survey resulting in an
  overall response rate of 49.4
• 47.57 surveyed had a virtual consult
• 52.43 surveyed had a traditional referral
• p value 0.26

37
Patient Survey Results
38
Patient Survey Results
39
Patient Survey Results
40
Patient Survey ResultsHave you seen the
specialist yet?
Note Missing data N86, 32.21
41
Chart Review How many referrals required a
Face-to-Face encounter with specialist?
42
Patient Survey Care ratings (0-10 Likert scale)
43
Patient Survey Care ratings (0-10 Likert scale)
44
Conclusions Patient Survey

• Over 85 of all referrals by patients PCP
• Referral condition present gt3 months ¾ of the
  time
• Nearly 50 of patients had already seen a
  specialist for the condition in the last 12
  months
• Less than 60 of VC required conversion to a
  Face-to-Face specialist visit.

45
Conclusions (2) Patient Survey

• Rating of
• PCP
• Medical Care
• Advice/Recommendations
• High (over 8.5 on 10 point Likert scale)
• No difference between VC and TC

46
Phase 4 Data Analysis in process

• Describe referral patterns by VC vs TC
• Graphs over time
• Use of VC vs TC by departments, specific
  conditions, and selected demographics
• Frequencies
• Logistic models (random effects to account for
  physician/clinic clustering)

47
Phase 4 Data Analysis in process

• Chart review contrasts VC vs TC on frequencies
  and means of items including
• Number of persons needed to process
• Time to consult completion
• Face-to-face needed
• Physician and Patient surveys contrasting
• Utility of consult
• Satisfaction

48
Next Steps Disseminate Results

• Publications
• Patient perception of virtual care
• Provider education (point of care learning)
• Understanding the virtual care environment
• Cost Analysis
• Partner with network providers for delivery of
  care in novel ways (Southern Colorado expansion)
• Explore and continue development of virtual
  patient care including secure messaging,
  tele-health, patient email, etc

49
Questions

• Ted Palen ted.e.palen_at_kp.org
• David Price david.price_at_kp.org
• Kristin Wallace kristin.b.wallace_at_kp.org