Cost-Effectiveness Analysis of: Remote Monitoring of ... · Implantable Cardioverter-Defibrillator (ICD) • Used for primary and secondary prevention of sudden cardiac death, which

Cost-Effectiveness Analysis of:

Remote Monitoring of Implantable Cardioverter Defibrillators and Cardiac Resynchronization Therapy Defibrillators

CADTH SymposiumApril 17, 2018

Man Wah Yeung, Vania Costa, Lindsey Falk,

Nancy Sikich, Irfan Dhalla, Vivian Ng

1

All co-authors have no actual or potential conflict of

interest in relation to this topic or presentation.

Disclosure

2

The Provincial Advisor on the Quality of Health Care in Ontario, Canada

3

Clinical Evidenceclinical evidence review

HTA

Health Services

and Medical

Devices

Health Economicseconomic evidence review

economic evaluation

budget impact analysis

Patient Perspectivepublic and patient

engagement

Ontario Health

Technology Advisory

Committee (OHTAC)

in support of HQO

Funding

Recommendations

Ontario Ministry of

Health and Long-

Term Care

Health Technology Assessment at

Health Quality Ontario

http://www.hqontario.ca/Evidence-to-Improve-Care/Health-Technology-Assessment

4


HTA

Health Services

and Medical

Devices

Health Economicseconomic evidence review

economic evaluation



engagement

Ontario Health

Technology Advisory

Committee (OHTAC)

in support of HQO

Funding

Recommendations

Ontario Ministry of

Health and Long-

Term Care

Health Technology Assessment at

Health Quality Ontario

http://www.hqontario.ca/Evidence-to-Improve-Care/Health-Technology-Assessment

5

Project Team

• Clinical Epidemiologist: Vania Costa

• Health Economist: Man Wah Yeung

• Secondary Health Economist: Lindsey Falk

• Senior Health Economist: Shawn Xie

• Patient Engagement Analyst: Jenny Gilbert and Arshia Ali

• Medical Librarian: Corinne Holubowich

• Business Analyst: Paul Kolodziej

• Project Manager: Claude Soulodre

• Manager, Health Economic Evaluation: Vivian Ng

• Manager, Clinical Reviews: Sarah McDowell

6

Stakeholder Consultations

• CorHealth Ontario (previously Cardiac Care Network)

• Ontario Ministry of Health and Long-Term Care

• Manufacturer representatives

• Clinical experts (nurse, electrophysiologist/ cardiologist)

• Health economic experts from Ontario Health

Technology Advisory Committee, OHTAC

7

Background:Implantable Cardioverter-Defibrillator (ICD)

• Used for primary and secondary prevention of sudden cardiac

death, which is caused by ventricular tachyarrhythmias

• ICD monitors the patient’s heart rhythm, detects irregularities and

corrects them with electrical impulses

• Applies an electrical shock in order to restore normal rhythm when a

life-threatening arrhythmia is detected

Image Source: National Heart, Lung, and Blood Institute; National Institutes of Health; U.S.

Department of Health and Human Services

8

Background:Cardiac Resynchronization Therapy

with Defibrillator (CRT-D)

• Used in patients with advanced heart failure with low ejection

fraction, and at risk for arrhythmias despite optimal medical therapy

• CRT monitors the patient’s heart rhythm, detects irregularities and

corrects them with electrical impulses

• Improves the synchronized contraction of the ventricles

Image Source: National Heart, Lung, and Blood Institute; National Institutes of Health; U.S.

Department of Health and Human Services

9

Usual Care: In-Clinic Visits

• Programmer wand

• Physician reads the data stored in the ICD, CRT-D (e.g.,

arrhythmias, device malfunction)

• Clinical examination, assessment of signs and symptoms,

and medication use

• Usually scheduled every 6 months during maintenance phase

(i.e., 3 months after successful implantation)

• Additional visits may also occur if needed

10

Remote Monitoring of ICD/ CRT-D• Allows transfer of information (data) stored* in the implanted

device to a database that can be accessed by the physician

*Equivalent to information obtained during in-clinic visits

• Data transmitted: Arrhythmias, ICD shocks, device function,

battery depletion

• Data transmission occurs in different ways:

• At scheduled intervals (e.g. daily, weekly, every 3-6 months)

• When triggered by an event, e.g. arrhythmia, ICD shock,

device malfunction = alerts

• In lieu of scheduled, in-clinic visits = remote interrogations

• Does not provide additional therapeutic capabilities to the ICD/

CRT-D

11

Remote Monitoring: Communication

ICD/ CRT-D device

Bedside monitor

Data centre

Internet

(Web application)

Via landline or cellular line

Via MICS (Medical Implant Communication System)

12

Ontario Context

• Between 2010 to 2015, 1,798–2,208 patients implanted

with CRT-Ds and ICDs annually in Ontario

• 5 remote monitoring systems approved by Health

Canada:

– Cardiomessenger® (Biotronik), Carelink® (Medtronic), Latitude®

(Boston Scientific), Merlin-net® (Abbott), Smartview® (Livanova)

• Remote monitoring is not publicly funded, but already

used in Ontario:

– London, Hamilton, Ottawa, Southlake, St Mary’s General

Hospital, Guelph, Kawartha, Kingston, Sudbury, Windsor,

Toronto General, St Michael’s Hospital, among others

• No physician fee code exists for conducting a remote

interrogation in lieu of a scheduled in-clinic visit

13

Clinical ReviewWhat is the effectiveness and safety of remote monitoring

of adult patients implanted with ICDs and CRTs plus in-

clinic visits compared with in-clinic visits?

• 15 open-label RCTs

• Number of patients 151–1,650

• Follow-up: 1–3 years

• Low risk of bias (GRADE approach)

• 8 unpublished RCTs (completed 2012–2017, 1

ongoing)—results could not be included in the review

RCT = randomized controlled trial; GRADE = Grading of

Recommendations Assessment, Development and Evaluation

14

Clinical Review:

Meta-Analysis Results

Compared with in-clinic visits, remote monitoring plus in-

clinic visits:

• Reduced the total number of in-clinic visits, number of

patients with inappropriate ICD shocks, and the time

from medical event onset to both detection by the

physician and clinical action (GRADE: Moderate)

• Showed no difference in adverse events, mortality

(GRADE: Low), stroke (GRADE: Very Low), and

hospitalizations due to heart failure (GRADE: Low)

CRT-D = cardiac resynchronization therapy defibrillator; ICD = implantable cardioverter defibrillator;

GRADE = Grading of Recommendations Assessment, Development and Evaluation

15

Primary Economic Evaluation:

Research Question

Within the context of the Ontario Ministry of Health and

Long-Term Care:

• What is the cost-effectiveness of remote monitoring

(RM) plus in-clinic follow-up in comparison to in-clinic

follow-up alone in adult ICD/CRT-D recipients with heart

failure?

16

Methods

• Analysis type: Cost-utility analysis

• Type of model: Markov model

• Target population: ICD/CRT-D recipients with heart failure

– 65 years, 70% male

– During the maintenance phase (3 months post-

implantation)

• Intervention: Remote monitoring plus in-clinic follow-up, as

per Canadian recommendations1

– 1:1 ratio

– Every 6 months

• Comparator: In-clinic follow-up alone 1 Yee 2013. Can J Cardiology

17

Methods

• Perspective: Public payer (Ontario Ministry of Health

and Long-Term Care)

• Time horizon: 5 years

• Discount rate: 1.5% (CADTH guidelines 4th edition)

• Outcomes:

– Costs (CAD 2017)

– QALYs (quality-adjusted life-year)

– ICER (incremental cost effectiveness ratio)

• Sensitivity & scenario analyses: Structural, parameter,

and methodological uncertainties

18

Markov Model Structure

Events:

• Hospital

admission/

readmission

• Emergency

department visits

• Unscheduled

clinic visits

• Scheduled clinic

visits


NYHA = New York Heart Association functional class

* Post-hospitalized states: increased

risk of hospital readmission and death

19

Model Assumptions

• Devices from different manufacturers are equivalent

• Single- and dual-chamber devices have the same

effectiveness1,2,3

• There is no crossover between RM and non-RM (i.e.,

patients do not switch between intervention arms)

• Excluded events prior to the maintenance phase and

events that are the same between intervention arms

– Implantation, re-implantation attempts, surgical/ device

complications, lead revisions, device recalls, device upgrades/

downgrades, medications and other therapies

1 Hindricks 2014, 2 Peterson 2013, 3 Toff 2005

20

Key Model Inputs and Sources

Clinical Effectiveness

Source: Meta-analysis

No significant difference in:

• Hospital admissions

• Emergency department visits

• Compliance to scheduled visits

• Mortality

• Health utilities

Statistically significant increase in:

• Unscheduled in-clinic visits

Utilities

Source: Various published

literature

• Health states

• Hospitalization (disutility)

• Age (disutility)

Costs

Source: Schedule of

Benefit, Ontario Case

Costing, Ministry, industry

• Professional fees (physician, nurse)

• Hospital costs


• Remote monitoring system

• SA: Northern Health Travel Grant

• SA: Battery replacement

21

Key Model Inputs and Sources

Clinical Effectiveness

Source: Meta-analysis

No significant difference in:

• Hospital admissions


• Compliance to scheduled visits

• Mortality

• Health utilities

Statistically significant increase in:

• Unscheduled in-clinic visits

Utilities

Source: Various published

literature

• Health states

• Hospitalization (disutility)

• Age (disutility)

Costs

Source: Schedule of

Benefit, Ontario Case

Costing, Ministry, industry

• Professional fees (physician, nurse)

• Hospital costs


• Remote monitoring system

• SA: Northern Health Travel Grant

• SA: Battery replacement

22

Cost Parameters:

Direct Medical Costs

• Remote monitoring (beside transmitter via wireless data

transmission [reference case], bedside transmitter via

landline, accessories only, embedded)

• Hospitalization

• Emergency department visit

• Scheduled and unscheduled in-clinic visit

• Reimbursement for remote interrogation

• Ministry Northern Health Travel Grants [scenario

analysis]

• Battery replacements [scenario analysis]

23

Cost Parameters:Reimbursement for Remote Interrogation

• Currently no physician fee code

• Model assumed fee code created after public funding

– Reference case: Payment for remote visit same as that of an in-

clinic visit

– Scenarios: Reduced payments for remote visit

24

Key Cost Parameters

(2017 CAD)

Variable

Cost, 2017

CAD

SD or

Range Source

In-clinic visit

(Physician, Nursing = 12

mins interrogation)

121 Physician:

61 – 180

Nursing: 0.92

Schedule of Benefits,

Elsner et al. 2006

Remote interrogation*

(Physician, Nursing = 1.2

mins interrogation)

110 Physician:

61 – 180

Nursing: 1.02

Schedule of Benefits,

Elsner et al. 2006

ED visit 17,808 8,780 OCC

Hospital visit 32,247 26,503 OCC

ED = emergency department; OCC = Ontario Case Costing; SD = standard deviation

* Additional administrative costs for remote monitoring (nursing time mean =1.9 mins, SD = 10.6 mins

per month per patient)

25

Sensitivity/ Scenario AnalysesUncertainty Parameter

Structural Time Horizon

Functional status after hospitalization

Parameter:

Natural History

Compliance to scheduled, in-clinic visits

NYHA transition probabilities

Time to battery replacement

Parameter:

Remote Monitoring

Impact on Natural

History

Mortality

All-cause hospitalization

Unscheduled, in-clinic visits

ED visits

Compliance to scheduled, in-clinic visits

Optimistic scenario (↓ mortality, hospitalization, unscheduled in-clinic visits, ED visits, ↑ compliance)

Parameter:

Cost and Resource

Use

Northern Health Travel Grant

Nursing time required for remote interrogation

Physician fee code

Model of payment

Methodological Discount rate

26

Results: Reference Case

Strategy Mean

Costs, $

∆

Costs, $

Mean

QALYs

∆

QALYs

ICER,

$/QALY

In-clinic

only55,138

(15,342; 146,719)

2.38(1.58; 2.97)

Remote

monitoring59,492

(18,124; 154,325) 4,3542.56

(1.69; 3.28) 0.19 23,374

∆ = incremental; CRT-D = cardiac resynchronization therapy defibrillator; ICD = implantable cardioverter

defibrillator; ICER = incremental cost-effectiveness ratio

Note: Results may appear incorrect because of rounding.

27

Reference Case Results: Probabilistic Analysis

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

0 10000 20000 30000 40000 50000

Pro

babili

ty I

nte

rvention is

Cost-

Effe

ctive

Willingness to Pay Threshold ($/QALY)

Remote monitoring + In-clinic visits

In-clinic visits only

28

Sensitivity and Scenario

Analysis Results

Strategy

Mean

Costs ∆ Cost

Mean

QALYs ∆ QALYs

ICER,

$/QALY

Structural: Time horizon extended to 10 years

In-clinic 99,977 3.57

RM 107,663 7,686 3.89 0.32 23,697

Impact on natural history: Optimistic scenario

RM 44,130 3.20

In-clinic 62,172 -18,042 2.58 0.61 Dominant

Payment model for RM: Accessories only

In-clinic 45,780 2.38

RM 49,251 3,471 2.56 0.19 18,628

Physician fee code: 50% of in-clinic visit reimbursement

In-clinic 55,138 2.38

RM 59,294 4,156 2.56 0.18 23,313

29

Strengths & Limitations

Strengths:

• Effectiveness based on substantial body of RCT

evidence. Estimates based on meta-analyses (GRADE:

mostly moderate quality)

• Patient pathway and costs specific to Canadian/ Ontario

Limitations:

• Only modelled maintenance phase (3 months post-

implantation)

• Did not include downstream costs of battery replacement

(i.e., complications)

30

Primary Economic Evaluation:

Conclusions

• Remote monitoring provides good value for money in

ICD/ CRT-D recipients compared to in-clinic visits alone

– Provides health gains at an additional cost

– Best estimate ICER = $23,374 per QALY

– 71% probability of being cost-effective

31


Health Economicseconomic evidence

review

economic evaluation



engagement

Decision-Making for Funding

Recommendation

Criteria Sub-Criteria

Overall Clinical

Benefit

Effectiveness

Safety

Burden of Illness

Need

Consistency with

Expected Societal

and Ethical Values

Societal Values

Ethical Values

Value for Money Economic Evaluation

Feasibility of

Adoption

Economic Feasibility

Organizational Feasibility

32

OHTAC Draft Recommendation

• Health Quality Ontario, under the guidance of the Ontario Health

Technology Advisory Committee, recommends publicly funding

remote monitoring for patients with implantable cardioverter-

defibrillators, cardiac resynchronization therapy devices with or

without a defibrillator ….

• Currently open for public comment until April 26th

• Final recommendation pending public comments

Link to Full Report:

http://www.hqontario.ca/Portals/0/Documents/evidence/open-comment/hta-remote-monitoring-of-

implantable-cardioverters-1804.pdf

Link to Submitting Public Comments:

http://www.hqontario.ca/Evidence-to-Improve-Care/Health-Technology-Assessment/Reviews-And-

Recommendations/Remote-Monitoring-of-Implantable-Cardioverter-Defibrillators-Cardiac-

Resynchronization-Therapy-and-Permanent-Pacemakers

33

Questions?

Man Wah Yeung, Health Economist, Health Quality Ontario

[email protected]

mailto:[email protected]

34

Additional Information: Decision Determinants

35

Decision DeterminantsCriteria Sub-Criteria Decision Determinants Considerations

Overall Clinical Benefit Effectiveness Compared to in-clinic visits, remote monitoring plus in-clinic visits reduced the number of in-

clinic visits, the time from medical event onset to both detection by the physician and clinical

action, number of patients with inappropriate ICD shocks (ICDs/CRT-Ds) and the arrhythmia

burden (Pacemakers)

Safety No difference in major adverse events (death, hospitalizations, etc.) between remote

monitoring plus in-clinic visits and in-clinic visits

Burden of

Illness

Burden of illness is expected to be large

Approximately 2,000 patients receive new or replacement ICDs/CRT-Ds and about 6,000

patients receive new or replacement permanent pacemakers each year in Ontario. Unclear

how many are using remote monitoring

Need Remote monitoring may replace some in-clinic visits and may alert the physician’s office of

important medical events and device issues earlier compared to in-clinic visits

Consistency with

Expected Societal and

Ethical Values

Societal Values Patients and their family members reported positive experiences with remote cardiac

monitoring

Participants perceived that these devices provided important medical and safety benefits in

managing their heart condition

Ethical Values Use of the treatments evaluated is expected to be congruent with ethical values

Value for Money Economic

Evaluation

Remote monitoring plus in-clinic visits is good value for money for both ICD/ CRT-D and

pacemaker recipients

Feasibility of Adoption Economic

Feasibility

Remote monitoring plus in-clinic visits may be cost saving over time depending on the current

uptake

Organizational

Feasibility

Remote monitoring is already being used to a certain extent in some hospitals across Ontario

36

Additional Information: Primary Economic Evaluation

Methods

37

Clinical Effectiveness Parameters –

Reference CaseParameter

ICD/CRT-D

(95% CI)

PM

(95% CI)Source

Hospitalizations, RRAll cause: 1.03

(0.97 to 1.09)

All-cause: 0.97

(0.72 to 1.31)

Stroke: 0.82

(0.3 to 2.25)

(i) Meta-analysis

(ii) Meta-analysis

ED visit, IRR 1 (N/A) 1 (N/A)(i) Clinical review

(ii) Assumption

Unscheduled In-

clinic visit, IRR2.8 (2.16 to 3.63) 1 (N/A)

(i) MORE-CARE trial

(ii) Assumption

Compliance, RR 1.06 (0.69 to 1.58) 1 (N/A)(i) MORE-CARE trial

(ii) Assumption

Mortality, RRCardiac: 0.89

(0.75 to 1.06)

All-cause: 1.29

(0.78 to 2.13)

(i) Meta-analysis

(ii) Meta-analysis

Health utilities, MD0.043

(-0.043 to 0.128)

0.12

(-0.04 to 0.27)

(i) EVOLVO trial

(ii) Comoretto 2017

Reported as Mean (95% confidence interval). Measures compare RM versus In-Clinic Only

ED = emergency department; IRR = incidence rate ratio; MD = mean difference; RR = risk ratio

38

Scenario Analysis

CI = confidence interval; ED = emergency department; HR = hazard ratio;

IRR = incidence rate ratio; MD = mean difference; RR = risk ratio; SD = standard deviation

ICD/CRT-D (95% CI) Pacemaker (95% CI)

Reduced

hospitalizations

RR = 0.54 (0.41 to 0.71) RR = 0.60 (0.18 to 2.02)

Reduced ED visits IRR = 0.72 (0.53 to 0.98) Same as ICD/ CRT-D

Reduced scheduled, in-

clinic visits

IRR = 0.65 (0.49 to 0.88) Same as ICD/ CRT-D

Increased compliance RR = 1.26, SD = 0.06 N/A

Reduced mortality HR = 0.36 (0.17 to 0.74) N/A

Improved health utilities N/A MD = 0.12 (-0.04 to 0.27)

• Optimistic scenario (simultaneously modelled alternate

clinical effectiveness parameters in favour of RM)

• Extended 5-year time horizon to 10 years (to include

battery replacement costs)

39

Key Cost Parameters

(2017 CAD)

Variable Cost, $ SD or Range Source

Professional fees (per visit):

Physician 109 61.25 – 179.80 Schedule of

Benefits

Nursing (in-clinic

interrogation = 12 mins)

12 0.92 Elsner et al.

2006

Nursing (remote

interrogation = 1.2 mins)

1 1.02 Elsner et al.

2006

ED visit 17,808 8,780 OCC

Hospital visit 32,247 26,503 OCC

Remote monitoring line item-

wireless transmitter

450 250 – 1400 Manufacturers

Ministry Northern Health

Travel Grants

267 N/A Ministry

ED = emergency department; OCC = Ontario Case Costing; SD = standard deviation

40

Utility ParametersHealth States Mean Utility Source

ICD/ CRT-D:

NYHA class I 0.815 Yao 2007

NYHA class II 0.720

NYHA class III 0.590

NYHA class IV 0.508

Post-hospitalized, Year 1, NYHA I-IV NYHA class – hospital disutility

Pacemaker:

Stable arrhythmia 0.795 Caro 2006

Post-hospitalized, Year 1 Stable arrhythmia – hospital disutility

Post-stroke, Years ≥ 1 0.41 Post 2001

Disutilities

Hospitalization, ICD/ CRT-D

NYHA class I 0.07 Griffiths 2014

NYHA class II 0.03

NYHA class III 0.08

NYHA class IV 0.21

Hospitalization, pacemaker 0.04 Reynolds 2010

Age-related, per year

60 to 69 years 0.004 Berg 2015

≥ 70 years 0.005 Berg 2015

41

Additional Information: Primary Economic Evaluation

Results

42

Results: Probabilistic Analysis

ICD/ CRT-DWillingness-to-pay =

$50,000/ QALY

Cost-effective 71%

Not cost-effective 15%

Inferior 14%

Incremental Effectiveness (QALYs)

Incre

me

nta

l C

ost ($

)

43

Results: One-Way Sensitivity Analysis

ICD/ CRT-D:

EV = expected value; ICER = incremental cost-effectiveness ratio; NHTG = Northern Health Travel Grant;

QALY = quality-adjusted life-year

% using NHTG (0 - 100%):

% reduction in remote interrogation

fee code (100 - 0%)

ICER ($/ QALY)

44

Results: Scenario Analysis

ICD/ CRT-DStrategy Mean

Costs, $

Incremental

Cost, $

Mean

QALYs

Incremental

QALYs

ICER,

$/QALY

Scenario: Optimistic (RM reduced mortality, hospitalizations, and

unscheduled in-clinic and emergency visits; Increased compliance)

Blended RM44130 3.20

RM is

DominantIn-clinic alone 62,173 -18,042 2.58 0.61

Scenario: Ten year time horizon (incl. battery replacement costs)

Blended RM 107,664 3.89

In-clinic alone 99,977 7,686 3.57 0.32 24,019


RM = remote monitoring

45

Additional Information: Clinical Review

46

ICDs/CRTs: ICD Shocks• 5 RCTs (mean follow-up 12–37 months)

• No statistically significant difference between groups in patients with

total (appropriate + inappropriate) or appropriate ICD shocks

Inappropriate shocks (not due to ventricular tachyarrhythmia)

• Fewer patients in the remote vs. non-remote monitoring group had

inappropriate ICD shocks, 4.4% vs. 8.2%

(Risk difference −0.04 [−0.07 to −0.01]; Risk Ratio 0.53 [0.32–0.89]),

4 RCTs

47

ICDs/CRTs: Time From Event Onset to

Data Review or Clinical Decision

Author, Year

N (RM/no RM)

Follow-up, mean months (SD)

Time From Alert/Onset to Event

Review

Median days (IQR)

Time From Event Onset to Clinical

Decision

Median days (IQR)

Boriani et al, 201732

12 months

N = 148 (76/72)

RM: 3 (1–10)

No RM: 37 (14–71)

P < .001

RM: 2 (1–4)

No RM: 29 (3–51)

P = .004

Landolina et al, 2012

N = 200 (99/101)

Up to 16 months

RM: 1.4 (0.8–7.3)

No RM: 24.8 (9.5–48.8)

P < .001

Not reported

Crossley et al, 201143

N = 1,997 (1,014/983)

15 months

Not reported RM: 4.6

No RM: 22.0

P < .001

Varma et al, 2010

N = 1,339 (908/431)

11.5 (2.6)

RM: 1.0

No RM: 35.5

P < .001

Not reported

• 4 RCTs demonstrated a shorter time from event onset to detection

or to clinical decision of approximately 17–34 days

• Types of events detected: arrhythmias, ICD shocks, device (ICD)

malfunction

48

ICDs/CRTs: In-Clinic Visits

• 10 RCTs (mean follow-up 12–37 months)

• Mean total in-clinic visits (scheduled + unscheduled) per patient-

year was 0.9–3.9 and 1.7–6.3 in the remote and non-remote

monitoring groups, respectively

• 26% to 50% reduction in in-clinic visits with remote vs. non remote

monitoring (rate ratio 0.50–0.74)

3 studies did not report rates of visits and could not be included; their results were consistent with the ones

included in the meta-analysis

49

ICDs/CRTs: Patients with ≥ 1

Hospitalization Due to Heart Failure

• 5 RCTs (mean follow-up 11-33 months)

• No statistically significant difference in the number of patients

with ≥ 1 heart failure hospitalization

• Similar results for all-cause hospitalizations

50

ICDs/CRTs: Emergency Department (ED)

Visits

• 4 RCTs (mean follow-up 15-24 months)

• One study showed a lower rate of ED visits in the remote vs. non-

remote monitoring group

• In another study there was a lower rate of a combination of ED visits

or urgent in-office visits with remote monitoring but not ED visits

alone

• Two studies did not show a statistically significant difference in ED

visits between the two groups

• No statistically significant difference between groups in the % of

patients with an ED visit

51

• No statistically significant difference between the two groups

1 RCT not included as the number of deaths were not provided—the authors reported that the difference between the groups

was not statistically significant

ICDs/CRT-Ds: All-Cause Mortality

52

Summary of Findings: Remote Monitoring

of ICDs/CRT-DsNo. of

Studies

(Design)

Risk of Bias Inconsistency Indirectness Imprecision Publication

Bias

Group 1 Group 2 Effect Estimate

(95% CI)

Quality

Patients with Inappropriate ICD shocks

4 RCTs No serious

limitations

No serious

limitations

No serious

limitations

No serious

limitations

Likely (−1)a 4.4% 8.2% RD: −0.04 (−0.07,

−0.1)

⊕⊕⊕Moderate

Number of In-Clinic Visits

7 RCTs No serious

limitations

No serious

limitations

No serious

limitations

No serious

limitations

Likely (−1)a 0.9–3.9

/p-yr

1.7–3.6

/p-yr

Rate ratio: 0.5–

0.7

⊕⊕⊕Moderate

Time To Event Detection or Clinical Decision

4 RCTs No serious

limitations

No serious

limitations

No serious

limitations

No serious

limitations

Likely (−1)a 1–5

days

9–42

days

Shorter (remote

monitoring)⊕⊕⊕Moderate

Patients with Hospitalizations Due to Heart Failure

5 RCTs No serious

limitations

No serious

limitations

No serious

limitations

Serious

limitations (−1)b

Likely (−1)a 54.8% 53.2% RD: 0.0 (−0.02,

0.03)

⊕⊕ Low

Emergency Department Visits

4 RCTs No serious

limitations

Serious

limitations (−1)c

No serious

limitations

Serious

limitations (−1)d

Likely (−1)a 0.05–

0.24/p-yr

0.08–

0.73/p-y

Inconsistent

results

⊕ Very

Low

CI, confidence interval; p-yr, person-years; RD, risk differenceaWe are aware of seven unpublished completed RCTs and we considered that this could affect the results of our meta-analysis.bAs determined by both the lack of statistically significant results and the fact that the minimal information size was not achieved.cResults were inconsistent across studies.dThree studies did not show a statistically significant difference in the rate of emergency department visits between the two groups.

53

Summary of Findings: Remote Monitoring

of ICDs/CRT-Ds (Cont.)No. of

Studies

(Design)

Risk of Bias Inconsistency Indirectness Imprecision Publication

Bias

Group 1 Group 2 Effect Estimate

(95% CI)

Quality

Stroke

2 RCTs No serious

limitations

No serious

limitations

No serious

limitations

Very serious

limitations (−2)a

Likely (−1)b 0.6% 1.7% RD: −0.01 (−0.02,

0)

⊕ Very

Low

Major Adverse Events (Composite of mortality, cardiovascular, procedural, or device-related adverse events)

3 RCTs No serious

limitations

No serious

limitations

Serious

limitations

(−1)c

No serious

limitationsd

Likely (−1)b No difference, satisfied non-inferiority

hypothesis in two trials

⊕⊕ Low

All-Cause Mortality

CI, confidence interval; RD, risk differenceaThe power to detect a difference between the two groups was very low and the number of events reported in the study was very small.bWe are aware of seven unpublished completed RCTs and we considered that this could affect the results of our meta-analysis.cThe fact that a composite endpoint was used affects the generalizability of the results to clinical practice as it is difficult to separate the contribution of each

individual component and the types of events included may have different implications to the patient.dA statistically significant difference in major adverse events was not observed in any of the 3 studies, however, since two of the three studies were non-

inferiority trials and the non-inferiority hypothesis was satisfied in both, we decided not to downgrade for imprecision.eAs determined by both the lack of statistically significant results and the fact that the minimal information size was not achieved.fTwo studies found a statistically significant difference in quality of life and two studies did not find a difference.gNo statistically significant difference in quality of life was reported in two studies, and in one study, a statistically significant difference was only

demonstrated with one of the quality of life measures but not with the other.

4 RCTs No serious

limitations

Serious

limitations (−1)f

No serious

limitations

Serious

limitations (−1)g

Likely (−1)b Different scales used Inconsistent

results

⊕ Very

Low

13 RCTs No serious

limitations

No serious

limitations

No serious

limitations

Serious

limitations (−1)e

Likely (−1)b 6.8% 8.4% RD: −0.01 (−0.02,

0.01)

⊕⊕ Low

Quality of Life

54

Additional Information: Budget Impact Analysis

55

Research Question

• What is the five-year budget impact (2017–2021) of

implementing remote monitoring for adult patients newly

implanted with cardiac electronic devices from the

perspective of the Ontario Ministry of Health and Long-

term Care?

– Implantable cardioverter defibrillators (ICDs)

– Cardiac resynchronization therapy defibrillators (CRT-Ds)

– Cardiac resynchronization therapy pacemaker (CRT-Ps)

– Pacemakers

56

Target PopulationTotal recipients in Ontario (new devices

and replacements) (CIHI)

N= 10,071

(Year 1)

CRT-D = cardiac resynchronization therapy-defibrillator; CRT-P = cardiac resynchronization therapy-

pacemaker; ICD = implantable cardioverter defibrillator; PM = pacemaker; SA = sensitivity analysis

ICD/ CRT-D/ CRT-P (CIHI) Pacemaker (CIHI)

N= 2,049 N= 5,927

New Devices: 79% of total (Ontario ICD Registry) N= 7,976

Current remote monitoring uptake: 15%

(SA: ICD/ CRT-D/ CRT-P: 50%, PM: 4%) (Expert opinion, manufacturers)

Increase in uptake given funding: 10% per year (SA: 15%)(Expert opinion)

Cap in uptake: 47% (SA: ICD/ CRT-D/ CRT-P: 71%, PM: 22%) (Akar, 2013, Mairesse, 2015)

57

Target Population- Reference Case

Baseline RM Uptake at 15% (all devices)


pacemaker; ICD = implantable cardioverter defibrillator; PM = pacemaker; SA = sensitivity analysis

RM Uptake (%) Year 1 Year 2 Year 3 Year 4 Year 5

Current Scenario: Not Funded

Any device 15 15 15 15 15

New Scenario 1: Funded (10% increase per year, cap at 47%)

Any device 25 35 45 47 47

New Scenario 2: Funded (15% increase per year, cap at 71%)

Any device 30 45 60 71 71

New Scenario 3: Funded (10% increase per year, cap at 71% and

22% for ICD/CRT-D/CRT-P and pacemakers, respectively)

ICD/ CRT-P/ CRT-D 25 35 45 55 65

Pacemakers 22 22 22 22 22

58

Target Population- ScenarioBaseline RM Uptake at 50% (ICD/ CRT-P/ CRT-D)

and 4% (Pacemaker)

RM Uptake (%) Year 1 Year 2 Year 3 Year 4 Year 5

Current Scenario: Not Funded (50% and 4% uptake at baseline for

ICD/CRT-D/CRT-P and pacemakers, respectively)

ICD/ CRT-P/ CRT-D 50 50 50 50 50


New Scenario: Funded (10% increase per year, cap at 71%)

ICD/ CRT-P/ CRT-D 60 70 71 71 71



pacemaker; ICD = implantable cardioverter defibrillator

59

Resources and CostsPer-Patient Cost to Ministry, CAD 2017

Year 1 Year 2 Year 3 Year 4 Year 5

Remote monitoring:

ICD/ CRT-P/ CRT-D recipients

under public funding16,371 13,613 11,930 10,449 9,163

Pacemaker recipients under

public funding6,921 6,962 6,456 5,845 5,295

ICD/ CRT-P/ CRT-D recipients

without public funding a14,572 12,918 11,219 9,744 8,476

Pacemaker recipients without

public funding a6,684 7,526 7,164 6,684 6,196

In-clinic alone:

ICD/ CRT-P/ CRT-D recipients 15,120 13,524 11,852 10,380 9,101

Pacemaker recipients 6,471 6,874 6,456 5,773 5,295

CRT-D = cardiac resynchronization therapy-defibrillator; CRT-P = cardiac resynchronization therapy-pacemaker; ICD =

implantable cardioverter defibrillator; RM = remote monitoring

a No fee code for remote interrogation, and no payment for RM components

60

Results: Reference Case


Total Cost to Ministry, $ million

Total Budget Impact Year 1 Year 2 Year 3 Year 4 Year 5 5 Years

Current Scenario: Not Funded69.45 144.75 216.48 284.04 347.84 1,062.57

New Scenario 1: Funded

(10% increase, cap at 47%) 70.75 144.23 214.03 279.15 340.23 1,048.39

Net Budget Impact 1.29 -0.52 -2.45 -4.90 -7.60 -14.18

61

Results: Reference Case (continued)


Total Cost to Ministry, $ million

Net Budget Impact Year 1 Year 2 Year 3 Year 4 Year 5 5 Years

New Scenario 2:

Funded (15%

increase, cap at

71%) 1.55 -0.09 -1.96 -4.30 -7.67 -12.47

New Scenario 3:

Funded (10%

increase, cap at

71% and 22%) 1.25 -0.61 -2.22 -3.43 -4.34 -9.34

62

Net Cost to Ministry, $ million

Year 1 Year 2 Year 3 Year 4 Year 5 5 Years

Total Budget Impact

Current Scenario:

Not Funded 69.41 141.49 210.08 274.29 334.88 1,030.15

New Scenario 4:

Funded (10%

increase, cap at

71%) 72.53 145.41 213.80 277.77 337.77 1,047.28

Net Budget Impact3.12 3.92 3.72 3.48 2.89 17.13

Results- Scenario AnalysisBaseline RM Uptake at 50% (ICD/ CRT-P/ CRT-D)

and 4% (Pacemaker)