e33

J Popul Ther Clin Pharmacol Vol 25(2):e33-e62; October 26, 2018.

This article is distributed under the terms of the Creative Commons Attribution-Non Commercial 4.0 International License.

DOI: 10.22374/1710-6222.25.2.6

ABSTRACTS CAPT / ACTP ANNUAL MEETING 2018

The Canadian Association for Population Therapeutics /

Association Canadienne pour la Thérapeutique des

Populations presents:

Taking Action on Real World Evidence:

from Analysis to Impact

ABSTRACTS

October 22

– 23

, 2018

MaRS Discovery District

101 College Street

Toronto, Ontario

Abstracts are published on-line in

The Journal of Population Therapeutics and Clinical Pharmacology

www.jptcp.com

e34

ORAL PRESENTATIONS

1. Inﬂuence of opioid prescribing standards on

drug utilization among patients with chronic

opioid use: a retrospective cohort study

Morrow RL

, Bassett, K

1,2

, Wright JM

1,3

, Carney G

Dormuth CR

Department of Anesthesiology, Pharmacology and

Therapeutics at the University of British Columbia,

Department of Family Practice, University of British

Columbia,

Department of Medicine, University of

British Columbia

Address: Therapeutics Initiative, University of British

Columbia

richard.morrow@ti.ubc.ca

Background: In mid-2016, the College of Physi-

cians and Surgeons of British Columbia (CPSBC)

issued prescribing standards and guidelines relating

to opioid drugs, and a policy on prescribing of the

opioid substitution therapy buprenorphine/naloxone.

We evaluated whether the Colleges policies inﬂu-

enced prescription drug utilization.

Methods: We used a longitudinal cohort study de-

sign with monthly repeated outcome measures.

Patients with chronic prescription opioid use during

a 6-month identiﬁcation period were followed for

a 24-month pre-policy period and 12-month post-

policy period, and were compared to equivalent his-

torical controls who were followed one year earlier

and were therefore not exposed the policies. We

estimated changes in the utilization of opioids, high-

dose opioids (>90 milligrams of morphine equiva-

lents (MME)/day), opioids with sedatives/hypnotics,

opioid discontinuation, and opioid substitution ther-

apy (methadone or buprenorphine/naloxone).

Results: The study included 68,113 patients in the

main cohort and 68,429 historical controls; 47,416

patients were in both cohorts. Following the opioid

policies, we observed reductions in average monthly

utilization levels of opioids (adjusted difference -63

MME; 95% CI -81 to -45), high-dose opioids (ad-

justed difference 0.2 days; 95% CI -0.3 to -0.2), and

opioids with sedatives/hypnotics (adjusted differ-

ence -0.2 days; 95% CI -0.2 to -0.1). There were

increases in opioid discontinuation and opioid sub-

stitution therapy. The study did not evaluate the im-

pact of the opioid policies on health outcomes.

Conclusion: The CPSBC opioid policies modestly

reduced utilization of opioids, high-dose opioids and

opioids with sedatives/hypnotics among patients

with chronic use of prescribed opioids while increas-

ing use of opioid substitution therapy.

2. Estimating the Impact of Treatment Evolution

in Non-Small Cell Lung Cancer (NSCLC): the iTEN

model

Moldaver D

, Hurry M

, Tran D

, Grima D

Cornerstone Research Group, Burlington, Canada,

AstraZeneca Canada, Mississauga, Canada

manjusha.hurry@astrazeneca.com

Objectives: Treatments options for patients with

advanced NSCLC (aNSCLC) are rapidly expanding.

The iTEN model was developed to support medical

decision-making by predicting aNSCLC patient sur-

vival and associated costs, from a Canadian health-

care system perspective.

Methods: A discrete event simulation of aNSCLC

treatment patterns was developed. Treatment se-

quencing and eligibility were derived from a modi-

ﬁed Delphi process with Canadian clinical experts.

Published Kaplan Meier progression free and over-

all survival data were ﬁt with parametric functions to

estimate treatment efﬁcacy and guide the determi-

nation of progression and death events. Treatment

history was assumed to have no impact on the efﬁ-

cacy of subsequent therapies. Costs included were:

drug acquisition and administration, monitoring, im-

aging, physician visits, end-of-life, best supportive

care and adverse events. Model survival predictions

were validated against published real-world es-

timates from the Ontario Cancer and TYROL reg-

istries and a US medical records analysis (Nadler,

2018).

Results: The Ontario Cancer Registry, TYROL

registry, and US medical records study reported

overall survival for patients receiving two-lines of

J Popul Ther Clin Pharmacol Vol 25(2):e33-e62; October 26, 2018.

This article is distributed under the terms of the Creative Commons Attribution-Non Commercial 4.0 International License.

ABSTRACTS CAPT / ACTP ANNUAL MEETING 2018

e35

chemotherapy, up to ﬁve-lines of treatment with

chemotherapy/targeted therapies or up to three-

lines of treatment with immunotherapies/chemo-

therapy/targeted therapies, respectively. One-year

survival rates in these studies were 67%, 38% and

49%; corresponding iTEN estimates, after restrict-

ing the treatment pattern to match each study, were

61%, 39% and 60%. Based on current Canadian

practice patterns, the estimated one-year survival

and life-time cost of treating aNSCLC were 46% and

$89,899.

Conclusions: Further validation is planned; how-

ever, the model may reasonably estimate the clinical

and cost consequences of treating aNSCLC.

3. Patient characteristics, treatment patterns and

survival for unresectable stage III non-small cell

lung cancer in Ontario, Canada

Seung SJ

, Hurry M

, Walton R

, Evans WK

HOPE Research Centre, Sunnybrook Research In-

stitute;

AstraZeneca Canada,

McMaster University

soojin.seung@sunnybrook.ca

Objectives: In anticipation of new treatment strat-

egies for unresectable stage III NSCLC, we under-

took a retrospective study to determine how these

patients have been managed in Ontario, Canada

and their survival by treatment approach.

Methods: Individuals diagnosed between April 1,

2010 and March 31, 2015 were identiﬁed in the

Ontario Cancer Registry. Patients were unresecta-

ble if no surgery was undertaken within 3 months

of diagnosis. Initial treatments included: radiation

(RT), chemotherapy, concurrent and sequential

chemo+RT (cCRT, sCRT). Survival was calculated

from date of diagnosis to death.

Results: There were 24,729 individuals diagnosed

with NSCLC; 5,243 (21.2%) were stage III and 4,542

(18.4%) were stage III unresectable. Mean age of

the latter group was 69.7 Â± 10.3 years; 54.2%

were male. 64.2% of patients were treated within 3

months of diagnosis. The frequency of treatment ap-

proach was: cCRT (21.6%), palliative RT (21.3%),

curative RT (20.2%), no treatment (19.6%), chemo-

therapy (11.6%), sCRT (4.9%) and targeted therapy

(0.7%). Median survival (IQR) was 2.9 yrs (1.7-4.8)

for targeted therapy, 2.0 yrs (1.0-5.5) for cCRT,

1.4 yrs (0.7-3.4) for curative RT, 1.4 yrs (0.7-3.1)

for chemotherapy, 1.2 yr (0.6-2.9) for sCRT, 0.6 yrs

(0.3-1.2) for palliative RT and 0.5 yrs (0.2-1.2) for no

treatment.

Conclusion: Although cCRT is generally consid-

ered standard of care for stage III unresectable NS-

CLC, patients in Ontario receive various treatment

approaches. Survival outcomes vary widely.

4. Time trends among new users of osteoporosis

drugs over 20 years: considerations for

pharmacoepidemiologic study design

Hayes KN

, Ban JK

, Burden AM

, Athanasiadis G

Cadarette SM

1,2

University of Toronto Dalla Lana School of Public

Health, Department of Public Health Sciences, Ep-

idemiology Division,

University of Toronto Leslie

Dan Faculty of Pharmacy, Department of Pharma-

ceutical Sciences

k.hayes@mail.utoronto.ca

Background: Oral bisphosphonates entered the

market in 1996 and are the main osteoporosis

drugs prescribed in Canada. Introduction of new

therapeutic options, updates to practice guidelines,

and healthcare policy changes may make studies

examining new-users of bisphosphonates suscep-

tible to time-varying biases. Objective: To compare

characteristics of ﬁrst-time bisphosphonate users in

Ontario, Canada over time.

Methods: The Ontario Drug Beneﬁt (ODB) program

covers prescription drugs listed on the ODB formu-

lary for residents aged 65 or more years. We iden-

tiﬁed ﬁrst dispensation of an oral bisphosphonate

among Ontario residents from 1996/04 to 2015/03.

We excluded patients aged younger than 66 years,

taking other osteoporosis drugs, with health condi-

tions known to impact bone, and long-term care fa-

cility residents. Medical and pharmacy claims within

the year prior to bisphosphonate dispensation were

used to characterize patients. Descriptive statistics

were used to summariz e and compare patient char-

acteristics by ﬁscal year.

J Popul Ther Clin Pharmacol Vol 25(2):e33-e62; October 26, 2018.

This article is distributed under the terms of the Creative Commons Attribution-Non Commercial 4.0 International License.

ABSTRACTS CAPT / ACTP ANNUAL MEETING 2018

e36

Results: We identiﬁed 523,210 eligible seniors (mean

age 75 years). A larger proportion of men (6.2% to

29.0%) and diabetics (8.7% to 17.8%) initiated ther-

apy over time. History of benzodiazepines decreased

(26.6% to 11.5%), while prior statin use (9.5% to

42.8%), oral corticosteroid use (10.6% to 14.7%), and

prior bone mineral density testing (46.7% to 68.0%)

increased. A shift in prescriptions from etidronate to

alendronate and risedronate was documented.

Conclusions: Characteristics of new initiators

of oral bisphosphonates among Ontario seniors

changed over time, reﬂecting changes in healthcare

delivery and osteoporosis management. Consider-

ation must be given to time-trends when designing

pharmacoepidemiologic studies in this population.

5. Interrupted time series analysis of long-acting

injectable antipsychotic use

Janzen D

, Kuo I

, Leong C

, Bolton JM

, Alessi-

Severini S

College of Pharmacy, Rady Faculty of Health

Sciences, University of Manitoba, Winnipeg,

De-

partment of Psychiatry, Max Rady College of Medi-

cine, Rady Faculty of Health Sciences, University of

Manitoba, Winnipeg

janzend6@myumanitoba.ca

Background: Long-acting injectable (LAI) antipsy-

chotics are recommended to improve adherence

to maintenance antipsychotic therapy; however,

injectable antipsychotic therapy may be negatively

perceived and subsequently underutilized. We hy-

pothesized LAI use increased after market entry of

the second-generation antipsychotic (SGA) risperi-

done in LAI formulation in 2004.

Methods:Administrative health databases at the

Manitoba Centre for Health Policy (MCHP) were

accessed to identify and describe LAI antipsychotic

users. Interrupted time series analysis with Poisson

regression was used to model LAI user counts be-

tween 1998 and 2016. Predictor variables included

time in ﬁscal quarters, a dummy variable indicating

quarters pre- or post-LAI risperidone market entry,

and time since LAI risperidone market entry. The

Manitoba population was included as an offset

variable and the model was adjusted for autocorrela-

tion. Analysis was conducted with SASÂ® software.

Results: A downward trend in LAI use was ob-

served prior to LAI risperidone market entry (RR:

0.976, p<0.0001). Post-entry, the trend reversed

(RR: 1.038, p<0.0001), resulting in a net increase of

1.4% per quarter after 2004 (RR: 1.014, p<0.0001).

Median time to LAI initiation following schizophrenia

diagnosis decreased from 3.3 years (IQR 0.7-6) in

2004 to 0.3 years (IQR 0.1-0.9) in 2015 (p=0.003).

Conclusions: The introduction of LAI risperidone in

2004 had a positive impact on LAI use. Furthermore,

initiation of LAIs occurred earlier after schizophrenia

diagnosis in 2015 compared with 2004. Acknowl-

edgements Results and conclusions are those of

the authors; no ofﬁcial endorsement by Manitoba

Health, Seniors and Healthy Living or MCHP is in-

tended or should be inferred.

6. Outcomes of metastasectomy in metastatic

renal cell carcinoma patients: The Canadian

Kidney Cancer information system experience

Nazha S

, Dragomir A

,Wood LA

,Rendon RA

Finelli A

, Hansen A

,So AI

,Kollmannsberger C

Basappa N

, Pouliot F

, Soulieres D

,Heng D

Kapoor A

,Tanguay S

McGill University Health Center, McGill Universit,

Queen Elizabeth II Health Sciences Centre, Nova

Scotia,

Princess Margaret Cancer Centre, Uni-

versity of Toronto,

BC Cancer Agency Vancouver

Cancer Centre, BC Cancer Agency,

Cross Can-

cer Institute, University of Alberta,

Centre Hospi-

talier Universitaire de Quebec, University of Laval,

Centre Hospitalier de L‘Universite de Montreal,

University of Montreal,

Tom Baker Cancer Center,

University of Calgary,

Juravinski Cancer Centre,

McMaster University

sara.nazha@hotmail.com

Objective: Surgical resection of metastasis can be

integrated in the management of metastatic renal

cell carcinoma (mRCC) as it can contribute to de-

lay disease progression and improve survival. This

study assessed the impact of metastasectomy in

mRCC patients using real-world pan-Canadian data.

ABSTRACTS CAPT / ACTP ANNUAL MEETING 2018

J Popul Ther Clin Pharmacol Vol 25(2):e33-e62; October 26, 2018.

This article is distributed under the terms of the Creative Commons Attribution-Non Commercial 4.0 International License.

e37

Methods: The Canadian Kidney Cancer information

system (CKCis) database was used to select pa-

tients who were diagnosed with mRCC between Jan

2011 and Dec 2017.All patients having had a com-

plete or incomplete metastasectomy and no metas-

tasectomy during follow-up where included in the

study cohort. Each patient having received a com-

plete or incomplete metastasectomy was matched

with up to 10 patients with no metastasectomy by

several potential confounding factors, such as: age,

time from RCC diagnosis until metastasis, clear cell

histology and use of targeted treatment before me-

tastasectomy and having had a nephrectomy.

Results: A total of 329 patients had complete (221

patients) and incomplete (10 8 patients) metastasec-

tomy, while 1,318 mRCC patients did not undergo

a metastasectomy. At 12 months, 99.1%, 88% and

76.8% of patients were alive in the complete metas-

tasectomy, incomplete metastasectomy and no me-

tastasectomy group, respectively (p<0.001). A total of

298 patients receiving metastasectomy (101 incom-

plete metastasectomy and 197 receiving complete)

have been matched to a control group of patients

who did not have metastasectomy. Finally, patients

receiving metastasectomy show an increased sur-

vival compared to patients not having had a metas-

tasectomy HR: 0.48 (95%CI 0.37-0.63), p<0.001.

Conclusion: Our study revealed the positive ef-

fect of metastasectomy performed in mRCC with

an improved OS compared to patients with no

metastasectomy.

7. A non-opioid alternative to intravenous (IV)

opioids in emergency departments (ED) would

signiﬁcantly reduce the economic burden on the

healthcare system. A societal perspective cost-

consequence analysis.

Dao S, Nguyen TTY, Muhimpundu CV

Synergyx Consulting Inc.

sebastien@synergyxconsulting.ca

Background: IV opioids are commonly used for the

management of pain in ED. This drug utilization re-

sults in signiﬁcant costs to the healthcare system.

The objective was to evaluate the cost-effectiveness

of methoxyﬂurane (MXF) in the treatment of adult

patients requiring analgesia for moderate to severe

acute pain associated with minor trauma.

Methods: A retrospective analysis of medical re-

cords, as well as primary and secondary researches

were conducted. A cost-consequence analysis

(CCA) was chosen to compare MFX to IV opioids

(morphine, fentanyl, hydromorphone). The cost per

consequence avoided and total cost of treatment

per patient were evaluated in the CCA. The conse-

quences included IV opioid’s side effects/complica-

tions and the use of healthcare resources.

Results: The total cost of treatment per patient were

$165.85 for MFX, $316.54 for morphine, $320.41 for

fentanyl and $315.33 for hydromorphone. MFX in ED

can generate savings of $151.58 per patient by reducing

nursing/physician time for IV administration and patient

monitoring. MFX has an improved safety proﬁle with

less nausea (2.0% for MFX versus 12.1% morphine,

21.1% fentanyl and 19.0% hydromorphone), vomiting

(2.0% versus 5.8%, 9.5% and 10.7% respectively), hy-

potension (1.0% versus 2.1%, 1.6% and 1.1% respec-

tively), respiratory depression (0.0% versus 1.1%, 1.1%

and 1.1% respectively) and no complication (phlebitis,

extravasation and harmful IV errors) associated with

IV opioids. Which translates into additional savings of

$90.66 per patient for avoided consequences.

Conclusion: MFX is a cost-effective alternative to

IV opioids with an improved safety proﬁle and a re-

duced economic burden on the healthcare system.

8. Real-world safety of second-generation direct-

acting antivirals in Hepatitis C

Machado MA

, Moura CS

, Klein M

, Winthrop K

Carleton B

, Abrahamowicz M

, Feld J

, Curtis JR

Bernatsky S

Research Institute of the McGill University Health

Centre.

Oregon Health & Science University.

University of British Columbia.

Toronto Western

Hospital Liver Center, University Health Network.

University of Alabama at Birmingham.

marina.machado@mail.mcgill.ca

Objective: To compare the safety of direct-acting

antivirals (DAA) hepatitis C virus (HCV).

J Popul Ther Clin Pharmacol Vol 25(2):e33-e62; October 26, 2018.

This article is distributed under the terms of the Creative Commons Attribution-Non Commercial 4.0 International License.

ABSTRACTS CAPT / ACTP ANNUAL MEETING 2018

e38

HOPE Research Centre, Sunnybrook Research

Institute

shazia.hassan@sunnybrook.ca

Objectives:The Build Better Bones with Exercise

(B3E) pilot trial evaluated the effects of home ex-

ercise on community-dwelling older women with

vertebral fractures. The objective of this exploratory

economic analysis was to examine the health re-

source utilization and cost-effectiveness of the B3E

study.

Methods: The B3E study participants were rand-

omized in a 1:1 ratio to a 12-month home exercise

program or equal attention control group. Clini-

cal and health system resources such as adverse

events (AEs), allied health visits, caregiver time, lost

productivity, medical equipment, medications, out-

of-pocket costs, and physician visits and tests were

collected during monthly phone calls by a blinded

research assistant and from daily diaries completed

by participants. Program costs for both groups were

included as well. Quality of life (QoL) information

was collected via EQ-5D-5L at baseline, 6 and 12

months. Unit costs (2018 CAD) were applied to

health system resources to calculate total, medi-

an±standard deviation (SD) and minimum-maxi-

mum (min-max) costs of the intervention and control

groups, as well as the median cost per patient. The

incremental cost-effectiveness ratio (ICER) between

the two groups was also calculated.

Results: The study included 141 women (mean

age=76±6.40y for intervention, 77±7.28y for control)

and overall total costs were $664,923 and $614,033,

respectively. The top three cost drivers from the

resource utilization collected were caregiver time

($250,269 and $240,811), medications ($151,000

and $122,145) and adverse events ($58,807 and

$71,981). The mean cost per intervention patient

was $9,365±$9,988 (median=$5,054, min–max=

$1,210–$57,774), and the mean cost per control

patient was $8,772±$9,718 (median=$5,203, min–

max = $931–$51,327). The mean EQ-5D index

score per patient was 0.81±0.11 (median=0.81,

min–max =0.54–1) and 0.79±0.13 (median=0.80,

min–max= 0.33–1) respectively. The difference in

Methods: We performed a retrospective cohort

study of adults with HCV using MarketScan data-

bases. Cohort entry was the date of ﬁrst prescription

of simeprevir/sofosbuvir (SIM/SOF), ledipasvir/so-

fosbuvir (LDV/SOF), ombitasvir/paritaprevir/ritona-

vir+dasabuvir (OPrD), sofosbuvir/velpatasvir (SOF/

VEL), or elbasvir/grazoprevir (EBR/GZR) between

Jan/2014-Nov/2016. Adverse events were deﬁned

as rashes, anaemia, and hepatic decompensation

recorded in hospital or outpatient claims between

cohort entry and therapy completion or discontinu-

ation. Patients with any of these diagnoses before

cohort entry were excluded. Cox regression was

performed to estimate hazard ratios (HR) and 95%

conﬁdence intervals (CI) for a composite outcome

(at least one event) adjusted for demographics,

co-morbidities, past HCV drugs, concomitant ribavi-

rin, and baseline health care use.

Results: We identiﬁed 15,480 HCV patients treated

with DAAs (72.9% LDV/SOF, 13.3% OPrD, 9.4%

SIM/SOF, 2.9% SOF/VEL, and 1.5% EBR/GZR);

11.8% used ribavirin concomitantly. Median age was

58 years (interquartile range 53-62) and 61.6% were

male. The frequency of skin events ranged from

0.2% (95%CI 0.0-0.7.) in SOF/VEL to 2.2% (95%CI

0.3-4.1) in EBR/GZR. The frequency of anaemia

ranged from 0.7% (95%CI 0.5-0.9) in LDV/SOF to

2.7% (95%CI 2.0-3.4) in OPrD. Hepatic decompen-

sation was rare (<0.6% overall). Multivariate analy-

ses were unable to establish differences between

drugs. Ribavirin use (HR 1.89, 1.36-2.62), baseline

cirrhosis (HR 2.05, 1.65-2.54) and previous hospi-

talization (HR 1.21, 1.01-1.45) were associated with

shorter time to an adverse event.

Conclusion: We found similar safety proﬁles for dif-

ferent DAAs agents. Reasons for altered risk should

be further explored.

POSTER PRESENTATIONS

9. Assessing the cost-effectiveness of the build

better bones with exercise pilot trial

Hassan S, Seung SJ, Templeton JA, Adachi JD,

Ashe MC, Clark R, Gibbs JC, Kendler D, Mitt-

mann N, Papaioannou A, Thabane L, Wark JD,

Giangregorio LM.

ABSTRACTS CAPT / ACTP ANNUAL MEETING 2018

J Popul Ther Clin Pharmacol Vol 25(2):e33-e62; October 26, 2018.

This article is distributed under the terms of the Creative Commons Attribution-Non Commercial 4.0 International License.©

2018 CAPT.

e39

quality adjusted life year (QALY) (0.02) was used to

calculate the ICER, and the ICER was determined

to be $29,650.

Conclusions:Results from this preliminary study of

141 women in the B3E study indicate a variety of

health resources are being utilized over a 12-month

period.

10. Time to Discontinuation of Biologic Therapy

by Mechanism of Action in Rheumatoid Arthritis:

Results from The Ontario Best Practice Research

Initiative (OBRI) Cohort

Movahedi M

, Hepworth E, Mirza R

, Cesta

, Larche MG

, Bombardier C

1,4,5,

and OBRI

investigators

Toronto General Hospital Research Institute, Uni-

versity Health Network, Toronto, ON;

Department

of Internal Medicine, McMaster University, Hamil-

ton, ON;

Divisions of Rheumatology and Clinical

Immunology and Allergy, Department of Medicine,

Hamilton, ON,

Department of Medicine (DOM) and

Institute of Health Policy, Management, and Evalu-

ation (IHPME), University of Toronto, Toronto, ON;

Division of Rheumatology, Mount Sinai Hospital,

Toronto, ON

mmovahed@uhnres.utoronto.ca

Background: Time to discontinuation of biologic

therapy may be related to mechanism of action. We

aimed to compare drug survival of tumor necrosis

factor inhibitors (TNFi) versus non-TNFi in an obser-

vational rheumatoid arthritis cohort.

Methods: Patients in the Ontario Best Practice Re-

search Initiative (OBRI) who started their ﬁrst bio-

logics on or any time after enrolment were included.

Time to discontinuation due to (1) any reason,

(2) non-response, physician, and patient decision,

(3) non-response, and (4) adverse events (AEs)

were assessed using Kaplan-Meier survival and

Cox proportional hazards regression analysis.

Results: A total of 796 patients were included of

whom 130 (16.3%) received non-TNFi and 756

(83.7%) received TNFi. Over a mean follow-up of

2.4 years, biologic discontinuation was reported

for 291 (36.6%) due to any reason, 229 (28.8%)

due to non-response, AEs, physician, and patient

decision, 110 (13.8%) due to non-response, and

81 (10.2%) due to AEs, respectively. There was a

signiﬁcant difference in time to discontinuation due

to any reason (Logrank p=0.0002); non-response,

AEs, physician, and patient decision (Logrank

p=0.04) between groups. After adjusting for poten-

tial confounders, difference remained signiﬁcant

for any reason [HR: 0.62 (0.46-0.84)] and non-

response, AEs, physician, and patient decision

[HR: 0.67 (0.47-0.94)].

Conclusions: The analysis demonstrates that pa-

tients initially started on non-TNFi are signiﬁcantly

more likely to discontinue their therapy earlier for

any reason and due to non-response, AEs, physi-

cian and patient decision compared to TNFi therapy.

Lack of response is likely not driving this, whereas

patient and physician preference likely inﬂuenced

the results.

11. Collection of anti-rheumatic medication data

from both patients and rheumatologists shows

strong agreement in a real world clinical cohort:

results from the Ontario best practices research

initiative (OBRI)

Movahedi M

, Cesta A

, Li X

, Bombardier C

1,2,3

behalf of Other OBRI Investigators

Ontario Best Practices Research Initiative, Toronto

General Research Institute, University Health Net-

work,

Department of Medicine (DOM) and Insti-

tute of Health Policy, Management, and Evaluation

(IHPME), University of Toronto,

Division of Rheu-

matology, Mount Sinai Hospital, Toronto, Canada

mmovahed@uhnres.utoronto.ca

Background: Collection of Anti-Rheumatic Medica-

tion (ARM) information from both patients and rheu-

matologists is considered a strength for Rheumatoid

Arthritis (RA) registries. However, it is important to

assess the agreement between these two data

sources.

Objectives: To examine the agreement of ARM use

between self-reports and rheumatologist reports

in the Ontario Best Practices Research Initiative

(OBRI).

ABSTRACTS CAPT / ACTP ANNUAL MEETING 2018

J Popul Ther Clin Pharmacol Vol 25(2):e33-e62; October 26, 2018.

This article is distributed under the terms of the Creative Commons Attribution-Non Commercial 4.0 International License.

e40

Methods: Patients enrolled in the OBRI on or after

Sep 1st 2010 with the ARM use reports from two

sources within 60 days of each other were included.

ARM included conventional synthetic Disease-

Modifying Antirheumatic Drugs (csDMARDs) and

biologic DMARDs (bDMARDs). Cohen’s Kappa sta-

tistics of agreement between the two data sources

were calculated. A multivariate backward stepwise

logistic regression was also used to exam the im-

pact of different factors on ARM use agreement be-

tween two data sources.

Results: 2,154 patients (78.7% female) were

included with a mean (SD) age of 57.8 (12.6)

year. There was a good agreement (Cohen’s

Kappa=0.61-0.80) between two sources. Increased

HAQ-pain index (OR: 0.66; 95% CI: 0.60-0.73) and

physician global score (OR: 0.95; 95% CI: 0.92-

0.98) were signiﬁcantly associated with the lower

agreement. By contrast, post-secondary education

(OR: 1.20; 95% CI: 1.02-1.40), and seeing an ac-

ademic rheumatologist (OR: 1.47; 95% CI: 1.25-

1.73) were signiﬁcantly associated with the higher

agreement.

Conclusions: The results of this analysis suggest

that ARM reports from the two data sources have

strong agreement in the OBRI. This agreement is

even better for patients who have post-secondary

education and are being treated by an academic

rheumatologist.

12. Physician characteristics associated with

opioid overdose hospitalization and death among

patients with long-term opioid use: a retrospective

cohort study

Morrow RL

, Carney G

, Dormuth CR

Department of Anesthesiology, Pharmacology and

Therapeutics at the University of British Columbia

richard.morrow@ti.ubc.ca

Background: We investigated whether the phy-

sician characteristics of sex, medical school grad-

uation year, high practice volume, or prescribing

behaviour were associated with hospitalization and

death involving accidental opioid overdose among

patients with long-term prescription opioid use.

Methods: We conducted a longitudinal cohort study

with monthly repeated outcome measures, using

administrative health data in British Columbia. The

study included patients with >=180 days of continu-

ous prescription opioid use at the start of any month

during 2013-2015, excluding patients with a history

of long-term care, palliative care or cancer. We used

multivariable Poisson regression to estimate the

association of physician characteristics and patient

outcomes.

Results: Our analyses included 136,565 patients

and 8,721 physicians. The baseline rates of hospital

admissions and deaths involving accidental over-

dose were 4.0 (95% CI 2.2 to 7.2) per 10,000 per-

son-years and 1.9 (95% CI 0.6 to 6 .1) per 10,000

person-years, respectively. Graduation during a pe-

riod of increasing promotion and use of opioids for

chronic non-cancer pain (1996-2010) was associ-

ated with opioid overdose hospitalization (rate ratio

(RR) 1.42; 95% CI 1.06 1.91) and opioid overdose

death (RR 3.02; 95% CI 1.61 5.68); graduation in

1975 or earlier was associated with opioid overdose

hospitalization (RR 1.52; 95% CI 1.072.15).

Conclusions: Physician graduation during a pe-

riod of increasing promotion and use of opioids for

chronic non-cancer pain may have increased risk of

hospitalization and death involving accidental opioid

overdose among patients with long-term opioid use.

Longer time in practice may also be a risk factor for

opioid overdose hospitalization.

13. Canadian Patients Attitudes and Preferences

for Original and Similar Biologic Medicines: 3-Year

Comparison

Wong-Rieger D

Consumer Advocare Network

durhane@sympatico.ca

Objectives: In Canada, similar biologic medicines

are available across many conditions, recently in-

cluding diabetes and cancer. For the past three

years, the Consumer Advocare Network has con-

ducted an annual survey of patient knowledge,

attitudes and usage preferences. Issues include:

perceived effectiveness, adverse effects, switching,

ABSTRACTS CAPT / ACTP ANNUAL MEETING 2018

J Popul Ther Clin Pharmacol Vol 25(2):e33-e62; October 26, 2018.

This article is distributed under the terms of the Creative Commons Attribution-Non Commercial 4.0 International License.

e41

interchangeability, pharmacovigilance, cost, and in-

formed decision making. Over the years, differences

among patient groups have emerged as important

inﬂuences, including continuous/episodic use, pri-

mary/secondary use, and variability of condition. Bi-

ologics require patient engagement for appropriate

use; as Canadian drug plans consider mandating

policies like switching, patient attitudes and behav-

ioural responses must be considered.

Methods: In 2016, an electronic survey with forced

choice, rating, and open-ended questions was sent

to about 2000 Canadian patients across diseases,

asking about biosimilar knowledge, attitudes and

preferences. In 2017 an updated version was sent

to an expanded pool, and in 2018, a revised survey

was distributed. We compared factors inﬂuencing

acceptabiity and usage of biosimilars from 200 re-

sponses (2016), 370 (2017), and 300 (2018).

Results: The trend over three years has been in-

creased knowledge and more positive attitudes

about biosimilars. Patients using biologics continu-

ously versus occasionally are more negative about

switching (70% vs.20%). Diabetes and gastrointesti-

nal patients are more concerned than arthritis. Can-

cer patients are most concerned about safety and

effectiveness.

Conclusions: Patients remain resistant to forced

switching to biosimilars, with implications for adher-

ence to and conﬁdence in their use.

14. Case Study: Real World Evidence Supporting

Reimbursement Decision Making using the

Canadian Kidney Cancer information system

(CKCis).

Robert Bick

Kidney Cancer Canada

robertfbick@gmail.com

Background: On April 18, 2017, the pan Cana-

dian Drug Review (pCODR) posted a Request for

Advice (RFA) for axitinib (Inlyta) for metastatic re-

nal cell carcinoma (MRCC): Is there evidence to

fund axitinib as an alternative to everolimus for the

second-line treatment of metastatic clear cell renal

carcinoma? In 2009, the Kidney Cancer Research

Network of Canada (KCRNC), the research arm of

Kidney Cancer Canada (KCC) established the Ca-

nadian Kidney Cancer information system (CKCis),

a centralized database to collect data from medical

centers across the country. CKCis now contains

retrospective and prospective de-identiﬁed patient

data from over 9500 consented patients who have

been diagnosed and treated for RCC.

Methods: KCC requested that CKCis investigators

make as a research priority the RFA question. A

cohort of patients who were pretreated with either

sunitinib or pazopanib were identiﬁed where axitinib

was given second line in 108 patients while everoli-

mus was used in 229 patients.

Results: Time to treatment failure (TTF) was found

to be longer in the axitinib group with Overall Sur-

vival (OS) similar in both groups. This Real World

Evidence was submitted for review to pCODR by

KCC.

Conclusions: Conclusion of CKCis Investigators:

Axitinib should be considered an option for all pa-

tients in Canada post 1stL VEGF-Targeted Therapy

without the limitations of the existing pCODR recom-

mendation. CODR Conclusions: The Clinical Guid-

ance Panel is of the opinion that there is appropriate

real world evidence and expert judgment to justify

axitinib as an equal alternative to everolimus in the

second line setting.

15. Real world treatment patterns and survival of

stage IV non-small cell lung cancer (NSCLC) in

Ontario, Canada

Seung SJ

, Hurry M

, Walton R

, Evans WK

HOPE Research Centre, Sunnybrook Research In-

stitute;

AstraZeneca Canada;

McMaster University

soojin.seung@sunnybrook.ca

Objectives: The majority of NSCLC patients are

diagnosed with Stage IV so it is important to under-

stand which patients are treated with systemic ther-

apies and to what beneﬁt.

Methods: This longitudinal, population-level study

determined treatment patterns and survival in Stage

IV NSCLC patients diagnosed between April 1,

2010 and March 31, 2015 from the Ontario Cancer

ABSTRACTS CAPT / ACTP ANNUAL MEETING 2018

J Popul Ther Clin Pharmacol Vol 25(2):e33-e62; October 26, 2018.

This article is distributed under the terms of the Creative Commons Attribution-Non Commercial 4.0 International License.

e42

Registry. Individuals were further identiﬁed as hav-

ing non-squamous disease, and those who received

an EGFR-TKI (afatinib, erlotinib, geﬁtinib) were as-

sumed to be EGFR mutation-positive (EGFR+). Sur-

vival was calculated from date of diagnosis to death.

Results: There were 24,729 individuals diag-

nosed with NSCLC. Approximately half (12,159;

49.2%) had stage IV disease, including 10,103 with

non-squamous disease, of whom 508 were cat-

egorized as EGFR+. The mean age for the stage

IV non-squamous and EGFR+ cohorts were 68.7

± 11.0 years and 69.1 ± 10.4 years, respectively;

49.3% and 60.8% were female, respectively. The

most frequent treatments for stage IV non- squamous

patients were palliative radiotherapy (RT) (46.7%)

and systemic therapy (14.9%) while 26.7% received

no treatment. 75.6% of the EGFR+ cohort received

geﬁtinib, with the majority receiving no subsequent

treatment (44.6%). Mean and median survival times

(IQR) for the stage IV non-squamous patients were

0.9 ± 0.0 years and 0.4 (0.2-1.0) years, respec-

tively. Mean and median survival times (IQR) for the

EGFR+ cohort were 1.9 ± 0.1 years and 1.5 (0.9-

3.0) years, respectively.

Conclusion: Few patients with Stage IV non-squa-

mous NSCLC received systemic therapy. Survival

was generally very poor, but better in the subgroup

of EGFR+ patients.

16. Evaluation of the Fentanyl Patch-for-Patch

Program in Ontario, Canada

Tadrous M, Greaves S, Martins D, Nadeem K, Singh

S, Auger C, Gomes T

St. Michael’s Hospital

tadrousm@smh.ca

Background: The rising impact of opioid use is a

major public-health concern, especially for fentanyl

given its high potency and potential for overdose. To

curb the misuse and diversion of fentanyl patches,

an early Patch-for-Patch (P4P) program was im-

plemented in Ontario between 2012 and 2015. The

P4P program requires that patients must return their

used fentanyl patches to a pharmacy before receiv-

ing a reﬁll.

Methods: We conducted a cross-sectional time-se-

ries analysis among counties that implemented the

P4P program using Ontario claims-data. We ze-

roed all intervention months and looked at outcome

rates in the 5 years prior and 12 months following

the launch of the P4P program. Outcomes included

monthly rates of prescriptions dispensed for fentanyl

and non-fentanyl opioids, and opioid toxicity-related

hospital emergency department visits and hospi-

tal admissions. We modeled each outcome using

ARIMA models and tested the impact of the P4P

program using a ramp function.

Results: We analyzed 16 counties that imple-

mented the early P4P program. Introduction of the

P4P program resulted in a signiﬁcant decline in the

number of fentanyl patches dispensed (from 1,277

to 888 patches per 10,000 population; p=0.04).

There was no signiﬁcant change in the rate of

non-fentanyl opioids dispensed (p=0.32) or opioid

toxicity related hospitalizations and emergency de-

partment visits (p=0.4) following implementation of

the program.

Conclusions: Implementation of a P4P program in

select counties in Ontario reduced the number of

fentanyl patches dispensed, but did not have any

measurable impact on rates of opioid toxicity-related

hospitalizations and emergency department visits.

17. Aromatase Inhibitors and Risk of

Cardiovascular Outcomes in Post-Menopausal

Women with Breast Cancer

Khosrow-Khavar F

1,2

, Filion KB

1,2,3

, Bouganim N

4,5

Suissa S

1,2

, Azoulay L

1,2,4

Centre for Clinical Epidemiology, Lady Davis Insti-

tute, Jewish General Hospital, Montreal, Canada,

Department of Epidemiology, Biostatistics, and

Occupational Health, McGill University, Montreal,

Canada,

Division of Clinical Epidemiology, Depart-

ment of Medicine, McGill University, Montreal, Can-

ada,

Gerald Bronfman Department of Oncology,

McGill University, Montreal, Canada,

Department

of Oncology, Cedar Cancer Center, McGill Univer-

sity Health Center, Montreal, Canada

farzin.khosrow-khavar@mail.mcgill.ca

ABSTRACTS CAPT / ACTP ANNUAL MEETING 2018

J Popul Ther Clin Pharmacol Vol 25(2):e33-e62; October 26, 2018.

This article is distributed under the terms of the Creative Commons Attribution-Non Commercial 4.0 International License.

e43

Background: Aromatase inhibitors (AIs) and

tamoxifen are common treatments for estrogen-

receptor positive breast cancer. However, data

from trials suggest that AIs may increase the risk

of cardiovascular outcomes, when compared with

tamoxifen. Thus, to address this safety concern, we

assessed whether use of AIs is associated with an

increased risk of cardiovascular outcomes in the

setting of real-world clinical practice.

Methods: Using the United Kingdom Clinical Prac-

tice Research Datalink linked to the Hospital Episodes

Statistics and Ofﬁce for National Statistics databases,

we identiﬁed patients newly-diagnosed with post-

menopausal breast cancer who were newly-treated

with either AIs or tamoxifen between 1998 and 2015,

and followed until 2016. Cox proportional hazards mod-

els using inverse-probability-of-treatment and censoring

weighting were used to estimate hazard ratios (HRs)

with 95% conﬁdence intervals (CIs) of fatal and non-fa-

tal myocardial infarction (MI) or ischemic stroke asso-

ciated with use of AIs when compared with tamoxifen.

Results: A total of 7,813 and 9,170 patients were

newly-treated with AIs and tamoxifen during the

study period, respectively. The use of AIs was

associated with a 54% increased risk of MI, when

compared with tamoxifen (4.1/1000 vs 1.9/1000

person-years; HR: 1.54, 95% CI: 1.04-2.27). In

contrast, use of AIs was not associated with an

increased risk of ischemic stroke, when compared

with tamoxifen (5.1/1000 vs 3.3/1000 person-years;

HR: 1.16, 95% CI: 0.83-1.61).

Conclusion: In this study, use of AIs was associated

with an increased risk of MI but not ischemic stroke,

when compared with tamoxifen in post- menopausal

women diagnosed with breast cancer.

18. Does Active Identiﬁcation of Patients Nearing

End of Life Change Use of Palliative Care and

Home Care Services?

Mittmann N, Liu N, MacKinnon M, Seung SJ, Hong

NJL, Earle CC, Gradin S, Sati S, Buchman S, Wright F

HOPE Research Centre, Sunnybrook Research

Institute

soojin.seung@sunnybrook.ca

Objective: To evaluate whether active identiﬁcation of

patients who could beneﬁt from a palliative approach

changes the use of palliative and home care services.

Methods: From 2014 to 2017, Cancer Care Ontario

implemented the INTEGRATE study among 4 can-

cer and 4 primary care centres. Physicians identi-

ﬁed patients likely to die within 1-year and beneﬁt

from palliative care. INTEGRATE patients were 1:1

matched to non-intervention controls selected from

provincial healthcare administrative data using pro-

pensity score-matching. Palliative and home care

services utilization was evaluated within 1-year af-

ter the index date, censoring on death, or March

31, 2017. Cumulative incidence function was used

to estimate the probability of having used services,

with death as a competing event. Rate of service

use per 360 patient-days was calculated. Palliative

and home care was analyzed.

Results: Of the 1,187 INTEGRATE patients, 1,185

were matched to a control. The intervention and

control groups were well-balanced on demograph-

ics, diagnosis, comorbidities, and death status. The

probability of using palliative services in the inter-

vention group was 81.3% (95% CI: 78.9% to 83.5%),

signiﬁcantly higher than controls (63.5%, 95% CI:

60.6% to 66.2%). The intervention group had a

statistically higher number palliative visits (29.7 vs.

19.6 per 360 patient-days) and greater probabil-

ity of receiving home care (82.5% vs. 56.8%) than

controls. Rate of palliative visits in the intervention

group (67.0 per 360 patient-days) was doubled than

controls (33.2 per 360 patient-days).

Conclusion: Physicians actively identifying patients

beneﬁting from palliative care resulted in increased

use of palliative and home care services.

19. Prioritization of Oncology Drugs: What factors

lead to priority status?

Sabarre KA

, Ellis KB

, Haynes AE

, Assasi N

, Denburg

2,3

, Cheung MC

2,4

, Moltzan CJ

2,5

, Trudeau M

2,6

CADTH pan-Canadian Oncology Drug Review,

pCODR Expert Review Committee,

Division of Hae-

matology/Oncology, The Hospital for Sick Children

and University of Toronto,

Division of Hematology/

Oncology, Sunnybrook Health Sciences Centre and

ABSTRACTS CAPT / ACTP ANNUAL MEETING 2018

J Popul Ther Clin Pharmacol Vol 25(2):e33-e62; October 26, 2018.

This article is distributed under the terms of the Creative Commons Attribution-Non Commercial 4.0 International License.

e44

University of Toronto,

Department of Medical On-

cology and Hematology CancerCare Manitoba and

University of Manitoba,

Division of Medical Oncol-

ogy and Hematology, Sunnybrook Health Sciences

Centre and University of Toronto

kelley-annes@cadth.ca

Background: At the time of their pan-Canadian On-

cology Drug Review (pCODR) submission, Submit-

ters may request a priority review (PR). This request

has an impact on the order of review, and placement

on the pCODR Expert Review Committee (pERC)

meeting agenda. PR requests are assessed by a

panel of pERC members and priority status is either

granted or not.

Objective: To provide a qualitative assessment of

PR requests submitted to pCODR, with the aim of

understanding what factors lead to prioritization.

Methods: Data from 53 pCODR submissions that

have requested PR were explored using qualitative

content analysis.

Results: 32 submissions received priority status,

21 submissions did not. Factors related to clini-

cally meaningful outcomes (e.g., overall survival

[OS], toxicity, progression-free survival [PFS]),

promising evidence (e.g., OS, PFS, quality of life

[QoL]), availability, and novel drugs were com-

monly mentioned in the panel’s conclusion for

granting priority status. Deterrents to prioritization

were not having clinically meaningful outcomes or

an urgent unmet need, availability of other treat-

ment options, and uncertainty of evidence (related

OS, QoL, safety). The decision to grant priority

status does not automatically translate to a pos-

itive funding recommendation; ﬁve submissions

with priority status did not receive positive funding

recommendations.

Conclusions: A range of factors contributes to

priority status being granted, premised on a lim-

ited set of values deemed relevant to technology

prioritization. Future research will 1) characterize

the determinants of technology prioritization using

quantitative measures and 2) explore the normative

basis for prioritization through comparative analyses

of priority-setting criteria among HTA institutions.

20. Factors affecting ototoxicity in head and neck

squamous cell carcinoma (HNSCC) patients treated

curatively with cisplatin-based chemoradiation.

Hueniken K, Zhang Y, Falls C, Brown MC, Mirshams

M, Rahimi M, Spreaﬁco A, Jang R, Hansen A, Siu L,

Waldron J, Bayley A, Giuliani M, Goldstein D, Wang J,

Hope A, Xu W

Princess Margaret Cancer Centre

k.hueniken@mail.utoronto.ca

Purpose: To estimate prevalence of hearing loss

among HNSCC patients treated with cisplatin-based

chemoradiation, and assesses need for baseline

and follow-up hearing tests.

Methods: At Princess Margaret Cancer Centre, a

prospective observational study assessed HNSCC

patients receiving cisplatin-based chemoradiation.

Hearing tests was tested at baseline and follow-up

(0.7-18.5 months after treatment initiation) at physi-

cian’s discretion. Signiﬁcant ototoxicity was deﬁned

as a grade 2 audiometric change from baseline to

post-treatment (CTCAE v.4.02) in one or both ears.

Results: Of 642 eligible patients (years 2008-2015)

enrolled, 549 received deﬁnitive chemoradiation and

93 post-operative chemoradiation. Median age was

57.5 years; 114(18%) were female; sites included

oropharyngeal (OPC; 421(66%)), oral cavity (OC;

105(16%)), laryngeal (56(9%)), hypopharyngeal

(34(5%)) and unknown primary treated as HNSCC

(26(4%)). 63(10%) were stage I-III and 576(90%)

were Stage IV. Only 246 patients received both

baseline and follow-up hearing tests, of which 142

(22% of 642) exhibited signiﬁcant ototoxicity (94

bilateral/48 unilateral). None received prolonged

courses of other ototoxic agents to explain hearing

loss. Hearing loss was signiﬁcantly higher (each

p<0.001) for OPCs (27% of 421) compared to all

others (13% of 221); high dose (24% of 557) vs.

low-dose weekly cisplatin (6% of 78); and deﬁnitive

(24%) vs. post-op chemoradiation (9%).

Conclusions: Hearing loss in 22% of HNSCC pa-

tients is clinically signiﬁcant. The true proportion of

hearing loss will be higher, as only 38% had base-

line and follow-up hearing tests, often triggered by

symptoms. Baseline and follow-up hearing tests

ABSTRACTS CAPT / ACTP ANNUAL MEETING 2018

J Popul Ther Clin Pharmacol Vol 25(2):e33-e62; October 26, 2018.

This article is distributed under the terms of the Creative Commons Attribution-Non Commercial 4.0 International License.

e45

should be routine, and primary preventive hearing

loss strategies should be emphasized.

21. Migraine treatment patterns and opioid

utilization among chronic and episodic migraine

patients identiﬁed by a clinician-administered semi-

structured diagnostic interview

Yu JS

, Pavlovic JM

, Silberstein SD

, Reed ML

Kawahara SH

, Cowan RP

, Dabbous F

, Campbell

, Pulicharam R

, Viswanathan HN

, Lipton RB

Allergan plc, Irvine, CA;

Monteﬁore Medical

Center, Bronx, NY;

Jefferson Headache Center,

Philadelphia, PA;

Vedanta Research, Chapel Hill,

NC;

DaVita Medical Group, El Segundo, CA;

Stan-

ford University School of Medicine, Stanford, CA;

Independent Consultant, La Jolla, CA;

Monteﬁore

Medical Center/Albert Einstein College of Medicine,

Bronx, NY

monica.elmore@allergan.com

Background: The objective was to describe

migraine treatment patterns and opioid utilization in

chronic migraine (CM) and episodic migraine (EM)

patients.

Methods: Eligible patients were enrolled from a

large medical group, 18 years, and had 1 claim with a

migraine diagnosis in the 12 months prior to screen-

ing. A physician-administered Semi- Structured

Diagnostic Interview, which included questions

related to headache symptoms, frequency, disability,

medication use, and diagnosis, was used to classify

patients as CM or EM. Acute treatment for migraine,

preventive treatment for migraine, opioid treatment,

and baseline characteristics were assessed in the

12 months prior to the study enrollment date.

Results: Of the 192 patients, 129 had CM and 63

had EM. The CM cohort had a mean age of 49.4

(SD=12.6) years and was 93.8% female. The EM

cohort had a mean age of 48.9 (SD=15.4) years

and was 82.5% female. 67.4% of CM and 55.6%

of EM patients had 1 claim for both acute and pre-

ventive medications. 53.5% of CM and 36.5% of

EM patients had 1 opioid claim (P<0.05); the mean

number of opioid claims was 4.0 (SD=7.1) among

CM and 2.8 (SD=8.2) among EM patients (P<0.05).

33.3% of CM and 15.9% of EM patients had 3 opioid

claims. 13.2% of CM and 7.9% of EM patients had a

pain diagnosis code.

Conclusions: Over half of CM patients and about

a third of EM patients received an opioid prescrip-

tion in the past year. Treatment patterns, particu-

larly in CM patients, indicate opportunities for better

management.

22. Pharmaceutical outcome-based pricing and

reimbursement agreements: lessons learned from

key international markets in Australia, Canada,

Europe and the United States.

Grubert N

, Gran-Ruaz S

, Hughes A

, Tims K

, Aly-

ward, B

, Mani A

, O’Quinn S

Independent global market access consultant

MORSE Consulting Inc

avery@morseconsulting.ca

Background: Health technology assessments are

used by many jurisdictions worldwide to measure

the value of new medicines. However, even after

undertaking such reviews, many questions remain

at the forefront of payers™ minds: How can uncer-

tainty at launch regarding a new drug’s real value be

overcome? What is needed to ensure a medicine is

used by the right patients in the real world? How can

spending on the drug be contained for the patient

population? Payers in various markets are currently

exploring the use of outcomes-based pricing and

reimbursement arrangements (OPRAs) to mitigate

risks at launch and expedite patient access.

Methods: A comprehensive literature review of the

trends, challenges, and opportunities in key inter-

national markets was conducted by the authors to

provide clarity on the feasibility of such innovative

outcomes-based agreements. Speciﬁcally, the payer

reimbursement environment in Australia, Canada,

Europe, and the United States was explored for this

analysis.

Results: Although there was congruence as to the

therapies that are appropriate for negotiation of such

agreements, jurisdictions vary signiﬁcantly on lev-

els of experience and methodologies/data sources

used. Challenges include system limitations,

ABSTRACTS CAPT / ACTP ANNUAL MEETING 2018

J Popul Ther Clin Pharmacol Vol 25(2):e33-e62; October 26, 2018.

This article is distributed under the terms of the Creative Commons Attribution-Non Commercial 4.0 International License.

e46

resource availability, administrative burden, and the

presence of existing agreements.

Conclusions: Given the variability in health sys-

tems, resources, and priorities, it is difﬁcult to gen-

eralize learnings from one market and apply them

to another. The authors have, therefore, developed

a checklist of questions to support payer decision

making in identifying when the use of OPRAs is

most appropriate.

23. Clinical atherosclerotic cardiovascular disease

(ASCVD): patient attitudes and awareness in

Canada

Rogoza RM

, Tai M-H

, Shepherd J

, Bailey H

Fairbairn M

Amgen Canada Inc., Toronto, Ontario, Canada,

Amgen Inc. Thousand Oaks, California, United

States,

Adelphi Real World, Macclesﬁeld, UK

rrogoza@amgen.com

Background: As a leading cause of morbidity and

mortality in Canada, cardiovascular (CV) disease

has a profound effect on the lives of patients. The

purpose of this study was to describe patient charac-

teristics, treatment patterns and disease awareness

of Canadian patients with dyslipidemia receiving

lipid-lowering therapy (LLT).

Methods: A physician and patient (pt) survey was

conducted in Canada from January to April 2017

through the Adelphi Dyslipidemia Disease Speciﬁc

Programme. Participating physicians provided infor-

mation on treated dyslipidemia pts, with the same pt

completing Pt-reported forms containing questions

on their condition, its impact and disease awareness.

Pts with ASCVD were systematically over-sampled.

Key outcome measures included: pt characteristics,

treatment patterns and pt disease awareness.

Results: The study included 67 physicians, pro-

viding data on 230 ASCVD and 367 non-ASCVD

pts. ASCVD pts were older and were more often

diagnosed by a cardiologist. ASCVD pts had more

comorbidities, and were taking higher dose statins.

LDL-C was not optimized in all patients (mean

2.38mmol/L). Despite a high disease burden,

disease awareness was low; only two thirds of pts

recalled discussing CV risk with their doctor and

one ﬁfth knew their LDL-C values. Finally, ASCVD

pts were more likely to have their work productiv-

ity, daily activities and quality of life (QoL) impacted

than non-ASCVD pts.

Conclusion: Canadian pts with dyslipidemia may

not be optimally managed or educated about their

disease, and ASCVD pts have worse QoL than

non-ASCVD pts. Further efforts are needed to

improve patient education and disease awareness.

24. Innovation for whom? A systematic literature

review on incidence and prevalence of severe

migraine subtypes in North America to understand

the market for emerging migraine therapies

Ackroyd T, Avramioti D, Syed I, and Hancock-

Howard R.

Amaris Consulting Ltd, Toronto, Ontario, Canada

rebecca.hancock-howard@amaris.com

Objectives: Migraine is the second most burden-

some disease globally. The only Health Canada

approved therapy with an indication for migraine

is Botox; additional treatment options are needed.

Monoclonal antibodies (MABs) are potential prophy-

lactic treatments for migraine, currently in phase III

development, speciﬁcally for severe migraineurs

with high frequency episodic (HFEM) and chronic

migraine (CM) subtypes. An understanding of

the epidemiology of severe migraine subtypes is

needed to estimate the patient populations poten-

tially eligible for these new treatments.

Methods: A systematic literature review was per-

formed with the following strategy (run in EMBASE

on 09/01/2018): terms for HFEM and CM popula-

tions, observational study design ﬁlter, outcomes of

interest (i.e. incidence and prevalence), and publi-

cation date restriction 2008-2018. Epidemiological

studies presenting incidence and prevalence data

for HEFM and CM in North America were included.

Results: Of 371 unique hits, inclusion criteria was

met by four cross-sectional survey studies from the

United States reporting prevalence data for CM. The

total number of returned surveys across the four

studies was 249,277. The prevalence of CM in the

ABSTRACTS CAPT / ACTP ANNUAL MEETING 2018

J Popul Ther Clin Pharmacol Vol 25(2):e33-e62; October 26, 2018.

This article is distributed under the terms of the Creative Commons Attribution-Non Commercial 4.0 International License.

e47

general population varied among included studies,

ranging from 0.025% to 1.79%. Where reported,

the prevalence of CM was higher in females and

peaked at midlife. No epidemiologic data on HFEM

were identiﬁed in North America.

Conclusion: Limited data exists regarding the inci-

dence and prevalence of severe migraine subtypes,

especially related to HFEM. No publications on

the epidemiology of HFEM or CM in Canada were

identiﬁed. Further research is needed to assess the

epidemiology of the migraine subtypes.

25. Population-based study of orchiectomy

treatment for prostate cancer in Quebec

Hu J

, Aprikian A

, Dragomir A

Division of Urology, McGill University

jason.hu@mail.mcgill.ca

Background: Androgen deprivation therapy (ADT)

can be delivered either surgically by orchiectomy

or medically via luteinizing hormone-releasing hor-

mone (LHRH) drugs. They are considered thera-

peutically equivalent in the treatment of advanced

prostate cancer (PCa). The objective is to describe

the use of orchiectomy for PCa and analyze factors

associated with orchiectomy treatment over medical

castration in Quebec.

Methods: The cohort consists of PCa patients from

2004-2012 and treated by ADT (either by LHRH

drugs or by orchiectomy), extracted from a random

sample from Quebec public healthcare insurance

databases. The primary study outcome was the

use of orchiectomy. Multivariable logistic regression

analysis was performed to identify variables associ-

ated with orchiectomy treatment.

Results: We identiﬁed 7096 patients treated by ADT,

of which 106 (1.5%) underwent orchiectomy. Fol-

lowing multivariable analyses, age (odds ratio [OR]

1.60, 95% conﬁdence interval (CI), 1.06 -2.43) and

Charlson comorbidity score (OR 1.09, 95%CI 1.01-

1.17) were associated with increased odds of orchi-

ectomy. Conversely, local radical treatment prior to

ADT initiation (OR 0.18, 95%CI 0.07-0.46) and res-

idence in a region with university-afﬁliated hospitals

(OR 0.39, 95%CI 0.26-0.58) were associated with

lower odds. Also, year of ADT initiation was asso-

ciated with lower odds of orchiectomy (OR 0.88 for

each increasing year, 95%CI 0.81-0.95).

Conclusion: A minority of PCa patients are treated

by orchiectomy in Quebec. These men were likely to

be older, more comorbid, not treated by local radical

treatment, living in a region not serviced by a uni-

versity-afﬁliated hospital, and initiated ADT in ear-

lier years compared to patients treated by medical

castration.

26. Systematic review and critical assessment

of the literature on methodologies of indirect

comparison of health technologies: Matching-

Adjusted Indirect Comparisons (MAIC)

Bini C

, Marcellusi A

1,2

, Mennini F

1,2

, Hojjati M

, Ver-

nia M

Economic Evaluation and HTA (CEIS-EEHTA) Fac-

ulty of Economics, University of Rome,

Institute for

Leadership and Management in Health, Kingstone

University,

Takeda Canada Inc.

Takeda Italia

S.p.A.

michelle.hojjati@takeda.com

As more therapeutic options emerge within the same

therapeutic area, there is often a lack of evidence

from head-to-head clinical trials that directly com-

pare the safety and efﬁcacy of a drug with the

standard of care. Several statistical methods have

been developed to allow for indirect comparison be-

tween different therapeutic alternatives that affect

the same therapeutic area. Among these, one of the

most commonly considered by regulatory bodies is

the method of Matching-Adjusted Indirect Compari-

sons (MAIC). The goal of this study was to evaluate

the applications of the MAIC method by HTA in order

to understand its usefulness to health care decision

making. The research was performed through a sys-

tematic review of the literature using the MEDLINE

electronic database (Pubmed) and included data

from HTA databases (speciﬁcally CADTH, NICE,

SMC, and PBAC). Publications that reported the

details of the MAIC methodology for various thera-

peutic areas were captured. Per PRISMA 2 guideli

ne methodology, 50 studies identiﬁed and 11 were

ABSTRACTS CAPT / ACTP ANNUAL MEETING 2018

J Popul Ther Clin Pharmacol Vol 25(2):e33-e62; October 26, 2018.

This article is distributed under the terms of the Creative Commons Attribution-Non Commercial 4.0 International License.

e48

included following screening. Of these, 2 were

CADTH-speciﬁc assessments: Intuniv XR (ADHD)

and Daklinza (Hepatitis C). Indirect comparisons

represent a valuable tool for comparing technol-

ogies in the absence of head-to-head trials. Their

quality will depend on several factors including the

method of analysis, and the validity of the underlying

assumptions. The MAIC method can provide useful

information to support HTA decision-making when a

head-to-head trial compared to standard of care is

unavailable.

27. Clinical and Economic Burden of

Atherosclerotic Cardiovascular Disease (ASCVD)

in Ontario, Canada

Pericleous L

, Goodman SG

, Oh P

, Rogoza RM

Colgan S

, Hume M

, and Goeree R

Amgen Canada Inc.,

St. Michael’s Hospital,

University of Toronto,

University Health Network,

Toronto, Ontario, Canada,

Goeree Consulting Ltd

louisa.pericleous@amgen.com

Objectives: Atherosclerotic cardiovascular disease

(ASCVD) is a leading cause of morbidity, mortality,

and economic burden in Canada. The objectives

of this study were to characterize patients who ex-

perienced a primary ASCVD event and to quantify

subsequent events and associated costs within a

10-year period.

Methods: Institute for Clinical Evaluative Sciences

(ICES) datasets were used to identify adult patients

with a primary ASCVD event using ICD-diagnostic

codes for hospital admissions in Ontario. Patients

were included in the study cohort if they had a pri-

mary event from April 1st 2005 to March 31st 2014.

Data were collected for second and subsequent

hospitalization and outpatient event(s) (2005-2015).

Speciﬁc ASCVD events of interest were myocardial

infarction (MI), stroke, and transient ischemic at-

tack (TIA).

Results: The mean age of the cohort was 65 years

and 54% of patients were male. Hypertension

(64%) and diabetes (27%) were the most common

comorbidities. The annua l incidence of a primary

MI, stroke or TIA event generally decreased or

remained stable over time, whereas the incidence

of subsequent events trended upwards. The to-

tal mean annual cost per patient in the ﬁrst year

post-primary event was $14,179, $36,899, and

$51,892 for TIA, MI, and stroke, respectively, and

generally decreased over time.

Conclusion: Patients in Ontario who experience a

primary ASCVD event remain at high risk for future

episodes. Further analyses to identify populations

who may beneﬁt from intensiﬁed risk reduction strat-

egies are warranted.

28. Moving towards universal coverage of direct-

acting antiviral therapies for hepatitis C infection

in Canada: an environmental scan of provincial

and international jurisdictions

Myers S, Khosa G, Kuo I, Alessi-Severini S

University of Manitoba, College of Pharmacy

myerss34@myumanitoba.ca

Background: Direct-acting antivirals (DAAs) have

become the standard treatment for patients with

chronic hepatitis C infections because of their high

cure rates and favourable side effect proﬁles; how-

ever, access to this new class of agents has been

limited because of its high cost. While public pay-

ers across Canada have implemented strict criteria

for drug coverage in order to contain expenditures,

efforts have been made to provide treatment cov-

erage as to improve access to medication for this

high-burden condition. This environmental scan

compares recent coverage criteria across national

and international jurisdictions.

Methods: Coverage criteria for Daklinza, Epclusa,

Sunvepra, Galexos, Harvoni, Sovaldi, Holkira Pak,

Zepatier, Maviret, Technivie, and Vosevi were

reviewed by accessing Canadian provincial drug

formularies. International coverage (e.g., Europe,

Australia, United States, Egypt, India) was reviewed

by searching available literature.

Results: Coverage criteria vary across Canada but

all provincial payers (except PEI) provide coverage

for Daklinza and Zepatier. By April 2018, all Cana-

dian jurisdictions, except Nova Scotia, New Brun-

swick, and Newfoundland & Labrador, had removed

ABSTRACTS CAPT / ACTP ANNUAL MEETING 2018

J Popul Ther Clin Pharmacol Vol 25(2):e33-e62; October 26, 2018.

This article is distributed under the terms of the Creative Commons Attribution-Non Commercial 4.0 International License.

e49

the stage 2 liver ﬁbrosis requirement for patients to

be eligible for coverage. Internationally, patient’s

access to DAAs differs signiﬁcantly. Many jurisdic-

tions restrict DAA prescribing authority to specialists

and request documentation of chronic hepatitis C.

In Australia all patients appear to have unrestricted

access to DAAs. In the US, considerable gaps of

coverage are identiﬁable and patients might face

signiﬁcant ﬁnancial burden to receive treatment.

Conclusion: DAAs appear to be generally accessi-

ble through public drug plans in Canada compared

to other countries.

29. Longitudinal changes in relative market share

proportions of biologic and targeted synthetic

disease-modifying anti-rheumatic drugs for

treatment of rheumatoid arthritis: descriptive data

from the Ontario best-practice research initiative

database

Hepworth E

, Movahedi M

, Rampakakis E

, Mirza

, Lau A

, Cesta A

Pope J

, Sampalis J.S

, Bom-

bardier C

2,5,6

and OBRI investigators

Department of Internal Medicine, McMaster Uni-

versity, Hamilton, ON;

Toronto General Hospi-

tal Research Institute, University Health Network,

Toronto, ON;

Divisions of Rheumatology, De-

partment of Medicine, Hamilton, ON;

Divisions of

Rheumatology, Epidemiology and Biostatistics, De-

partment of medicine, Western University, London,

ON;

Department of Medicine (DOM) and Institute

of Health Policy, Management, and Evaluation (IH-

PME), University of Toronto, Toronto, ON; 6Division

of Rheumatology, Mount Sinai Hospital, Toronto, ON.

elliot.hepworth@medportal.ca

Background/Purpose: For patients with Rheuma-

toid Arthritis (RA) without adequate clinical response

to conventional synthetic disease modifying

anti-rheumatic drugs (csDMARDs), the next therapy

is either biologic DMARDs (bDMARDs) or targeted

synthetic DMARDs (tsDMARDs). bDMARDs include

tumour-necrosis factor inhibitors (TNFi) or non-TNFi

classes. Since inception of Ontario Best Practice

Research Initiative (OBRI), new treatment options

have become available. We describe evolving use

of non-TNFi vs. TNFi in Ontario practices from

2008-2017.

Methods: Adult patients with RA enrolled in OBRI

who started bDMARDs/tsDMARDs anytime during,

or up to 30 days before, enrollment were included.

The yearly proportion of the population treated with

TNFi and non-TNFi therapy was measured for (i) all

patients and (ii) those initiating their ﬁrst bDMARD/

tsDMARD. TNFi included: Etanercept, Adalimumab,

Certolizumab, Golimumab, and Inﬂiximab. Non-

TNFi included: Abatacept, Rituximab, Tocilizumab,

and Tofacitinib.

Results: A total of 1,057 patients were included of

whom 653 were bDMARD/tsDMARD naive. In 2008,

the relative non-TNFi use was 3/56 (5.4%) in all

patients and 0/31 (0%) in treatment-naive patients.

By 2016, relative use was 224/679 (33.0%) in all

patients and 17/56 (30.4%) in treatment-naive. This

was followed by 144/426 (33.8%) and 4/15 (26.7%),

respectively in 2017.

Conclusion: This descriptive analysis shows an

increase in non-TNFi therapy use. The overall

trend towards greater use of non-TNFi therapies

as ﬁrst line advanced therapeutics may be partially

explained by recent guidelines allowing clinicians to

select either class as ﬁrst line advanced therapies.

Future analyses evaluating patient-, disease- and

concomitant drug use-speciﬁc determinants of phy-

sician decision-making will be conducted.

30. Higher drug cost for pregabalin/gabapentin

shouldn’t dissuade clinicians from prescribing

this intervention in spinal cord injured individuals

for neuropathic pain

Chan B

, Wodchis W

1,2

, Craven C

1,3

Toronto Rehabilitation Institute - University Health

Network,

Institute for Health Policy Management

and Evaluation - University of Toronto,

Department

of Medicine - University of Toronto

brian.chan@uhn.ca

Background: Canadian guidelines for treatment

of neuropathic pain in spinal cord injury (SCI) rec-

ommend pregabalin or gabapentin as ﬁrst-line

therapy followed by amitriptyline. To understand the

ABSTRACTS CAPT / ACTP ANNUAL MEETING 2018

J Popul Ther Clin Pharmacol Vol 25(2):e33-e62; October 26, 2018.

This article is distributed under the terms of the Creative Commons Attribution-Non Commercial 4.0 International License.

e50

economic impact of this recommendation we eval-

uated the health care costs of SCI individuals with

neuropathic pain prescribed pregabalin/gabapentin

compared to amitriptyline.

Methods: Former patients of a tertiary SCI rehabil-

itation facility in Ontario, Canada were recruited to

participate in a one year prospective study evalu-

ating health care utilization, costs and health out-

comes. Participants were over 18 years of age,

C1-T12, AIS A-D, injury duration greater than 1 year

with neuropathic pain and attended quarterly phone

follow-up. Information collected include: hospitaliza-

tions, physician visits, other health care practitioner

visits, drugs prescribed and equipment. Data was

converted to 2016 Canadian costs using publicly

available sources.

Results: Seventeen individuals were prescribed

pregabalin/gabapentin and 7 amitriptyline. Median

monthly cost for pain-related prescription drugs was

$102 for pregabalin/gabapentin and $66 for amitrip-

tyline. However, total median monthly healthcare

costs were $1,987 for pregabalin/gabapentin and

$3,357 for amitriptyline. Greatest cost differential

was observed in emergency department ($1,211

for amitriptyline and $0 for pregabalin/gabapentin)

and mobility equipment/device ($207 for amitripty-

line and $0 for pregabalin/gabapentin). There was

no difference in physiotherapy and occupational

therapy costs. Results were limited to self-report of

costs for a small case series in Ontario.

Conclusions: Higher prescription drug costs for

pregabalin/gabapentin were not the primary cost

driver for monthly healthcare costs. Higher total

healthcare costs for amitriptyline were driven by

higher emergency department visits and mobility

equipment/device purchases.

31. Health Utilities Index Mark 3 scores for major

chronic conditions: population norms for canada

based on the 2013-2014 Canadian Community

Health Survey

Guertin JR

1,2

, Humphries B

, Feeny D

4,5,6

, Tarride

3,4,6,7

Department of Social and Preventive Medicine,

Faculty of Medicine, Université Laval, Quebec City,

Quebec, Canada,

Centre de recherche du CHU de

Québec – Université Laval, Axe Santé des Popula-

tions et Pratiques Optimales en Santé, Hôpital du

Saint-Sacrement, Quebec City, Quebec, Canada,

Department of Health Research Methods, Evidence

and Impact, Faculty of Health Sciences, McMaster

University, Hamilton, Ontario, Canada

Department

of Economics, Faculty of Social Sciences, McMas-

ter University, Hamilton, Ontario, Canada ,

Health

Utilities Incorporated, Dundas, Ontario, Canada,

Centre for Health Economics and Policy Analysis

(CHEPA), McMaster University, Hamilton, Ontario,

Canada,

Programs for Assessment of Technol-

ogy in Health (PATH), The Research Institute of St.

Joe’s Hamilton, St. Joseph’s Healthcare Hamilton,

Ontario, Canada

jason.guertin@fmed.ulaval.ca

Background: Utility scores are frequently used as

preference weights when estimating quality-adjusted

life-years within cost-utility analyses or health-

adjusted life expectancies. Though previous Cana-

dian estimates for speciﬁc chronic conditions have

been produced, these may no longer reﬂect current

patient populations.

Methods: Data from the 2013-2014 Canadian

Community Health Survey were used to provide

Canadian utility score norms for seventeen chronic

conditions. Utility scores were estimated using the

Health Utilities Index Mark 3 (HUI3) instrument and

were reported as weighted average (95% conﬁ-

dence intervals [95% CI]) values. In addition to age

and sex-stratiﬁed analyses, results were also strat-

iﬁed according to the number of reported chronic

conditions (i.e., “none” to “≥5”). All results were

weighted using sampling and bootstrapped weights

provided by Statistics Canada.

Results: Utility scores were estimated for 123,654

(97.2%) respondents (weighted frequency =

29,337,370 [97.7%]). Of the chronic conditions that

were examined, “Asthma” had the least detrimental

effect (weighted average utility score = 0.803 [95%CI

0.795 – 0.811]) on respondents’ utility scores and

“Alzheimer’s disease or any other dementia” had

the worst (weighted average utility score = 0.374

ABSTRACTS CAPT / ACTP ANNUAL MEETING 2018

J Popul Ther Clin Pharmacol Vol 25(2):e33-e62; October 26, 2018.

This article is distributed under the terms of the Creative Commons Attribution-Non Commercial 4.0 International License.

e51

[95%CI 0.323 – 0.426]). Respondents who reported

suffering from no chronic conditions had, on aver-

age, the highest utility scores (weighted average util-

ity score = 0.928 [95%CI 0.926 – 0.930]); estimates

dropped as a function of the number of reported

chronic conditions.

Conclusion: Utility score differed between various

chronic conditions and as a function of the number

of reported chronic conditions. Results also highlight

several differences with previously published Cana-

dian utility norms

32. The development of a chronic disease

incidence prediction model for Ontario public

health planning

Ng R

, Sutradhar R

, Wodchis WP

2,3

, Rosella LC

1,2

Dalla Lana School of Public Health, University of

Toronto,

Institute for Clinical Evaluative Sciences,

Institute of Health Policy, Management and Evalua-

tion,

University of Toronto

ry.ng@mail.utoronto.ca

Background: Informed decision-making by pub-

lic health ofﬁcials is needed when planning for the

future burden of chronic disease in the population.

A prediction model that accurately predicts the inci-

dence of chronic disease in a jurisdiction based on

current population characteristics would be invalua-

ble towards decision-making.

Methods: The Ontario version of the Canadian

Community Health Survey was linked to health ad-

ministrative data to predict the incidence of chronic

disease over a ﬁfteen-year period. Sixteen varia-

bles, including: modiﬁable lifestyle behaviors (e.g.,

alcohol, smoking, poor diet and physical inactivity),

sociodemographic factors (e.g., age, ethnicity, in-

come) and other health-related factors (e.g., body

mass index) were used as predictors. Sex-speciﬁc

prediction models were developed using Weibull re-

gression models for males and females separately.

The performances of the prediction models were

evaluated based on measures of overall predictive

accuracy, discrimination and calibration.

Results: The derivation cohort consisted of 47,960

females and 38,267 males. For both sexes, the

overall predictive performance and discrimina-

tion improved when sociodemographic and other

health-related predictors were added to the base

model consisting of only modiﬁable lifestyle be-

haviors. Using the male model as an example,

Nagelkerke™s R2 (0.036 to 0.152) and the Harrell’s

c-statistic (0.627 to 0.778) improved with more

predictors; however, calibration as measured with

the Greenwood-Nam-D’Agostino statistic (p-value

<.0001) and calibration plots indicated underﬁtting

in lower-risk groups.

Conclusions: Overall, the derivation models pre-

dicted the incidence of chronic disease well for both

sexes. The next steps are to further improve the

predictive accuracy of the model followed by model

validation.

33. Patient preferences among individuals

with type 2 diabetes for attributes of diabetes

medications: a discrete choice experiment

Donnan J MSC

, Marra CA PharMD, PHD

1,2

, Bassler

KA MD

, Johnston K PHD

, Najafzadeh M PHD

Swab M

, Nguyen H PHD

, Gamble JM PHD

1,5

School of Pharmacy, Memorial University, St. John’s,

Newfoundland and Labrador, Canada,

School of

Pharmacy, University of Otago, Dunedin, New Zea-

land,

Faculty of Medicine, Memorial University,

St. John’s, Newfoundland and Labrador, Canada,

Department of Medicine, Harvard Medical School,

Boston, Massachusetts, USA,

School of Pharmacy,

University of Waterloo, Kitchener, Ontario, Canada

jennifer.donnan@mun.ca

Objective: To estimate the trade-offs that patients

with type 2 diabetes (T2D) make between attributes

of glucose-lowering medications using a discrete

choice experiment.

Methods: Eight relevant attributes were identiﬁed

using a literature review, focus groups and interviews

with patients and clinicians. A sample of Canadians

with T2D, recruited through a survey company, com-

pleted 14 choice tasks. A multinomial logit model

was used to estimate the weights of each attribute.

Willingness-to-pay was used to assess trade-offs

between attributes.

ABSTRACTS CAPT / ACTP ANNUAL MEETING 2018

J Popul Ther Clin Pharmacol Vol 25(2):e33-e62; October 26, 2018.

This article is distributed under the terms of the Creative Commons Attribution-Non Commercial 4.0 International License.

e52

Results: A total of 513 individuals completed the

survey. Average age was 59 years (SD=12), 59%

were male, 55% had tried two or more oral medi-

cations, and 62% had diabetes for at least 6 years.

All attributes were found to signiﬁcantly inﬂuence

choice. On average patients were willing-to-pay a

monthly cost for their therapy of: $135 to achieve

three additional years of life; $48 and $36 for a 20%

reduction in their risk of macrovascular and micro-

vascular events respectively; $37 for a 1% drop in

HbA1C; $32 for a 50% less risk of severe hypogly-

cemia over 10 years; $28 for a 50% less risk of a

minor side effect; and $19 for a 50% less risk of a

rare but serious side effect over 10 years.

Conclusion: All eight examined attributes were

shown to signiﬁcantly inﬂuence choice with cost and

life expectancy carrying the most weight, and seri-

ous and minor side effects carrying the least weight.

34. Predictors of hip fracture among patients

with inﬂammatory arthritis receiving chronic oral

glucocorticoid therapy

Shakeri A

, Amine Amiche M

, Levesque LE

Gomes T

, Adachi JD

, Cadarette SM

University of

Toronto,

McMaster University

ahmad.shakeri@mail.utoronto.ca

Objective: Oral glucocorticoids (GCs) rapidly in-

duce bone loss and increase fracture risk. Predic-

tors of fracture among oral GC users have not been

examined by underlying indication. We aimed to

identify predictors of hip fracture risk among patients

with inﬂammatory arthritis receiving chronic oral GC

therapy.

Method: We identiﬁed patients with inﬂammatory

arthritis (gout, lupus, polymyalgia rheumatica, rheu-

matoid arthritis) aged 66 or more years taking oral

GCs using Ontario public healthcare administrative

data, 1998/01/01-2014/09/30. We included patients

who dispensed 2 oral GC prescriptions totalling

450 mg prednisone equivalent using a 6-month as-

certainment window. GC exposure was quantiﬁed

during the ascertainment window by dose (mean

daily, cumulative) and pattern (continuous, intermit-

tent or sporadic). Index date was deﬁned at the end

of the ascertainment window. Medical and pharmacy

claim were used to deﬁne fracture risk factors within

the year prior to index. We used Cox proportional

hazard models to identify predictors of hip fracture

within 1 year following the index date.

Results: We identiﬁed 23,065 eligible patients

(mean age=74.6 years, SD=6.4; 64% women).

Average daily dose, cumulative dose and pattern

of GC use were not associated with fracture risk.

Rather, older age, rheumatoid arthritis, fracture his-

tory, stroke, antidepressant use and opioid use were

associated with increased hip fracture risk and male

sex was protective.

Conclusion: Controlling for GC exposure, risk fac-

tors for hip fracture parallel those commonly reported

among patients without inﬂammatory arthritis. Of

interest, among patients with inﬂammatory arthritis,

patients with rheumatoid arthritis are at highest risk

of hip fracture.

35. Disinvestment decision-making using value of

information analysis: a hypothetical case study

Ball G, O’Reilly D

McMaster University

ballga@mcmaster.ca

Background: Prudent health technology manage-

ment involves allocation of ﬁnancial resources to-

ward technologies with high assessed value and

away from those of low assessed value. Value of

information (VoI) analysis may be practical for pri-

oritizing cost-ineffective technologies for disinvest-

ment.This research illustrates an application of VoI

analysis to select the most appropriate disinvest-

ment candidate using cost-effectiveness models of

everolimus for advanced breast cancer (ABC) and

bevacizumab for platinum-resistant ovarian cancer

(PROC) in Canada.

Methods: The following comparative per-patient

and population-level VoI analyses were conducted.

Expected value of perfect information (EVPI) was

estimated to assess the value of reducing cost-

effectiveness uncertainty. Expected value of perfect

information for parameters (EVPPI) was assessed

to identify the type of additional research required

ABSTRACTS CAPT / ACTP ANNUAL MEETING 2018

J Popul Ther Clin Pharmacol Vol 25(2):e33-e62; October 26, 2018.

This article is distributed under the terms of the Creative Commons Attribution-Non Commercial 4.0 International License.

e53

to reduce uncertainty. Expected net beneﬁt of 1

sampling (ENBS) was calculated to estimate the

overall value of disinvesting each health technology,

respectively. The technology with highest ENBS

should be prioritized for disinvestment.

Results: Per-patient EVPI of everolimus for ABC

($9,378) was higher than that of bevacizumab for

PROC ($2,531). Analyses of EVPPI indicated that

clinical trials were the most appropriate method for

reducing uncertainty. The population-level ENBS

was found to be substantially higher for everolimus

($1,615,253,171) compared with bevacizumab

($10,320,497).

Conclusion: Results suggested that bevacizumab

for PROC should be prioritized for disinvestment

over everolimus for ABC. Decision-making uncer-

tainty may be minimized by prioritizing disinvestment

candidate technologies according to the likelihood

that additional information changes cost-ineffec-

tiveness status. In this way, VoI analysis may be a

useful analytical guide for informing healthcare tech-

nology management.

36. Disease-speciﬁc costs of non- metastatic and

metastatic castration-resistant prostate cancer in

Quebec

Yanev I

, Aprikian A

, Nablsi E

, Primiani J

, Vaillan-

court Z

, Bremner K

, Carcone S

, Mitsakakis N

Ryan W

, Krahn M

, & Dragomir A

McGill University, Montreal, QC, Canada,

Toronto

Health Economics and Technology Assessment

(THETA) Collaborative, Toronto General Hospital

Research Institute, University Health Network

yanev.ivan@mail.mcgill.ca

Objectives: Estimate the disease speciﬁc costs of

prostate cancer patients during the health states

of nonmetastatic castration-resistant prostate can-

cer (NM-CRPC) and metastatic castration-resistant

prostate cancer (M-CRPC).

Methods: This cohort analysis contains 211 pros-

tate cancer patients from the MUHC. An algorithm

of detecting NM-CRPC and M-CRPC was based on

increases of prostate speciﬁc antigen (PSA) levels

after castration and the detection of metastasis. The

usage of imagery tests, hospital visits and treat-

ments was extracted from the patients ﬁles and

the mean usage per resource was calculated for

30 days in the given health state. Resources price

were obtained from the RAMQ List of Medications

when available, when unavailable prices were ob-

tained from the MUHC internal prices lists. This

cost analysis was performed by health care system

perspective.

Results: Mean duration of NM-CRPC was 26.07

months while duration of M-CRPC was 20.79

months, with 62 and 6 8 patients per health

state respectfully. The average disease speciﬁc

resource utilisation per patient for 30 days was

$786 for NM-CRPC and $2,210 for M-CRPC

health states with the cost driver being chemo-

therapy or prescription drugs different than ADT.

The total average cost for NM-CRPC health state

was $20,457 compared to $45,956 for M-CRPC

health state.

Conclusions: The disease speciﬁc resource utilisa-

tion costs for a patient in M-CRPC are signiﬁcantly

higher than the costs of NM-CRPC. This being said

it would be interesting to study the total healthcare

costs from a societal perspective in order to improve

the cost management of PCa.

37. The use and effects of ehealth tools for patient

self-monitoring and reporting of outcomes

following medication use: a systematic review

Lancaster K

, Abuzour A

, Khaira M

2,4

, Mathers A

Chan A

, Bui V

3,4

, Lok A

, Thabane L

, Dolovich L

1,2,4

McMaster University,

University of Toronto,

Sun-

nybrook Health Sciences,

University of Waterloo

mkhaira@edu.uwaterloo.ca

Background: ehealth tools are becoming increas-

ingly popular for helping patients self-manage

chronic conditions. Little research has examined

the effect of ehealth tools for patient self-report-

ing on medication management. This review aims

to evaluate whether ehealth tools featuring patient

self-reporting of symptoms and adverse effects

are effective at promoting medication changes and

improving patient outcomes.

ABSTRACTS CAPT / ACTP ANNUAL MEETING 2018

J Popul Ther Clin Pharmacol Vol 25(2):e33-e62; October 26, 2018.

This article is distributed under the terms of the Creative Commons Attribution-Non Commercial 4.0 International License.

e54

Methods: MEDLINE, EMBASE and CINAHL

were searched from Jan 1, 2000 through to April

25, 2018. References were also searched. Title,

abstract and full text review, as well as data abstrac-

tion and risk of bias assessment were performed in

duplicate. Due to high heterogeneity, results were

not meta-analyzed, and instead presented as a

narrative synthesis.

Results: Thirteen randomized controlled trials

(RCTs) and one open-label intervention were

included, from which eleven unique ehealth tools

were identiﬁed. Four RCTs found statistically signif-

icant increases in positive medication changes as

a result of using ehealth tools. Eight RCTs found

improvement in patient symptoms following ehealth

tool use, especially in adolescent asthma patients.

Three RCTs showed that ehealth tools may improve

patient self-efﬁcacy and self-management of chronic

disease. Little or no evidence was found to sup-

port the effectiveness of ehealth tools at improving

medication recommendations and reconciliation by

clinicians, medication-use behaviour, health service

utilization, quality of life, or patient satisfaction.

Conclusions: Initial evidence showing ehealth tools

may improve patient symptoms and lead to medi-

cation changes is promising, however more high-

quality research is needed to explore how ehealth

tools can be used to effectively manage use of med-

ications to improve patient outcomes.

38. The pan-Canadian Pharmaceutical Alliance:

how policy, governance and process changes are

affecting public drug plan reimbursement

Chambers J, Dempster W

Sixty Public Affairs Inc.

wdempster@3sixtypublicaffairs.com

Increasingly, the pan-Canadian Pharmaceutical

Alliance (pCPA) has become the key forum for in-

terjurisdictional dialogue and action on pharma-

ceutical policy issues. Since its generic pricing

agreement achieved earlier this year, the alliance

has turned more of its attention to single-source

medicines, leading to the launch of new brand pro-

cess guidelines in June 2018. This presentation will:

1) explore how the evolving pCPA is affecting the

Canadian pharmaceutical marketplace; 2) examine

what recent developments tell us about the future

state of public reimbursement for medicines; and

3) consider what else may lie ahead for the pCPA.

In addition, we will identify issues arising from the

process guidelines, including gaps and opportuni-

ties. This presentation will update CAPT members

on the latest data regarding pCPA’s negotiations

and their impact on the public reimbursement of

new innovative medicines in Canada. It will build on

3Sixtyâ€™s previous analyses that were presented

at CA PT 2016. Among the relevant trends we will

explore are the impact of Quebec’s participation in

pCPA negotiations on the availability of new inno-

vative medicines in that province, therapeutic class

negotiations, the growing number of negotiations

that lead to closed ﬁles without agreement and how

long new medicines wait in the queue before negoti-

ations are undertaken. We will also review the politi-

cal, health policy, economic and legal aspects of the

pCPA and provide a number of considerations for

how pCPA can continue to evolve in the context of

other national issues, such as national pharmacare.

39. Prevalence and characteristics of hospitalized

patients on opioid therapy and risk of re-

admissions and emergency department visits

associated with opioid use in the 90 days post-

discharge using Cox proportional hazards model

Kurteva S

1,2

, Weir DL

1,2

, Lee TC

3,4

, Tamblyn R

1,2,4

Department of Epidemiology, Biostatistics and Oc-

cupational Health, McGill University, Montreal, QC,

Canada.

Clinical and Health Informatics Research

Group, McGill University, Montreal, QC, Canada.

Division of Infectious Diseases and Department

of Medicine, McGill University Health Centre, Mon-

treal, QC, Canada.

Department of Medicine, McGill

University, Montreal, QC, Canada

siyana.kurteva@mail.mcgill.ca

Background: Canada is the world second largest

consumer of opioids. As the number of opioid

prescriptions has increased over the past two

decades, the country has witnessed an increase

ABSTRACTS CAPT / ACTP ANNUAL MEETING 2018

J Popul Ther Clin Pharmacol Vol 25(2):e33-e62; October 26, 2018.

This article is distributed under the terms of the Creative Commons Attribution-Non Commercial 4.0 International License.

e55

in hospitalizations resulting from opioid poison-

ings. Our main objective was to evaluate opioid

use for patients admitted to surgical and hospital

units at two tertiary hospitals in Montreal and their

associated risk of emergency department (ED)

visits and hospital re-admissions in the 90 days

post-discharge.

Methods: Multiple sources of data were assembled

and linked including patient demographics, medical

services and prescription claims from the Quebec

provincial healthcare databases to describe the

clinical proﬁle of patients on opioids. Time- varying

opioid use was measured by using dispensing phar-

macy records of ﬁlled opioid prescriptions post-

discharge and was modeled as current, continuous

and cumulative duration of use using Cox propor-

tional hazards models. All analyses were adjusted

for age, chronic conditions, concomitant medication

use, and history of opioid use.

Results: Overall, half of the patients received an

opioid prescription at discharge, the majority of

whom had no history of opioid use in the one year

prior to admission (67%). We found that opioid use

was associated with a 19% increase in the risk of

ED visits and hospitalizations and the risk increases

when examining longer cumulative or continuous

duration of use.

Conclusion: Our ﬁndings suggest that long-term

use of opioids after hospitalization may increase the

risk of re-admissions and ED visits and physicians

may need to reassess the optimal duration of treat-

ment with these drugs.

40. Calcium channel blockers and diuretics: A

retrospective cohort study exploring a prescribing

cascade

Visentin JD

, Savage RD

1,2

, Gatley JM

, Stall N

1,3,5

Herrmann, N

3,6

, Gruneir A

1,2,4

, Rochon PA

1-3

, Bron-

skill SE

1-3

, McCarthy LM

1,3

Women’s College Hospital,

Institute for Clinical

and Evaluative Sciences,

University of Toronto,

University of Alberta,

Mount Sinai Hospital,

Sun-

nybrook Health Sciences Centre

jessica.visentin@mail.utoronto.ca

Background: Calcium channel blockers (CCBs)

are commonly prescribed for hypertension. While

their safety proﬁle is generally favourable, CCBs

are known to cause peripheral edema, which can be

distressing for patients and may result in prescrip-

tion of a diuretic. Older adults are particularly sus-

ceptible to adverse events associated with diuretics

(e.g., falls, hypokalemia, acute kidney injury). The

extent to which diuretics are prescribed with CCBs

at a population-level is presently unknown.

This study seeks to: (1) characterize the association

between CCB use and the subsequent receipt of a di-

uretic and (2) identify drug, patient and provider factors

that may be associated with this prescribing cascade.

Methods: A retrospective, population based cohort

study was conducted using health administrative da-

tabases from Ontario, Canada between September

2011 and September 2016 to identify adults aged 66

and older with hypertension and/or coronary artery

disease. Individuals newly dispensed a CCB were

compared to prevalent users of angiotensin-convert-

ing-enzyme inhibitors or angiotensin receptor block-

ers. Individuals were followed for 90 days post-index

date to identify receipt of a loop diuretic. Hazard ra-

tios were estimated using Cox proportional hazard

models adjusting for individual-level characteristics.

Results: Data analysis is ongoing but will be com-

plete before the conference.

Conclusions: Speciﬁc conclusions will be submit-

ted subsequently. Due to widespread use of CCBs

and the frequency of CCB-induced peripheral

edema, we expect a large proportion of the popula-

tion may be at risk of experiencing the CCB-diuretic

cascade. Knowledge of the pervasiveness of this

cascade and contributing factors can inform clinical

and policy strategies to improve medication safety.

41. Productivity losses after acute coronary

syndrome events in Canada: an interim analysis

Pandey AS

, Gupta M

, Mancini J

, Fischer A

Sidelnikov E

, Pericleous L

Department of Medicine, McMaster University,

IQVIA,

Amgen

varun.myageri@iqvia.com

ABSTRACTS CAPT / ACTP ANNUAL MEETING 2018

J Popul Ther Clin Pharmacol Vol 25(2):e33-e62; October 26, 2018.

This article is distributed under the terms of the Creative Commons Attribution-Non Commercial 4.0 International License.

e56

Background: The impact of an acute coronary syn-

drome (ACS) event on a patient’s productivity is not

well characterized. This study describes the produc-

tivity loss following hospitalization for an ACS event

in Canada. This interim analysis was conducted to

investigate early trends in the study ﬁndings.

Methods: Patients were enrolled across three sites

in Canada during a routine ambulatory cardiologist

visit 4-18 months post-index ACS hospitalization (my-

ocardial infarction or unstable angina). At enrollment,

study candidates were < 65 years old, had returned

to work 4 weeks following their index ACS hospitali-

zation, and were on a lipid lowering therapy prior to

or at index hospitalization. Baseline characteristics

were collected from medical records, while produc-

tivity loss was collected using a validated survey.

Results: An interim analysis of the study was per-

formed on 44 patients (91% myocardial infarction,

9% unstable angina). Enrolled patients were almost

exclusively male (96%), on average 53.6 years old,

and had an average BMI of 29.1 kg/m2. The total

annual lost productivity from a societal perspective

was 57.5 (SD 56.3) workdays. This was composed

of 45.8 (SD 48.8) days due to missed work, 8.8 (SD

24.8) days due to presenteeism, and 3.0 (SD 19.6)

days attributed to caregiver help.

Conclusion: The results suggest that days taken off

work after discharge are signiﬁcant for patients who

have experienced an ACS event, and the societal

impact of ACS is high. This study is likely to have

underestimated the impact, as the study population

is of working-age and relatively healthy.

42. Persistence of pharmaceutically-derived

cannabinoid use in Manitoba

Alkabbani W

, Marrie RA

, Bugden S

1,3

, Alessi-

Severini S

, Daeninck P

, Leon C

College of Pharmacy, Rady Faculty of Health

Sciences, Apotex Centre, University of Manitoba,

Winnipeg, MB, Canada,

Max Rady College of Med-

icine, Rady Faculty of Health Sciences, University

of Manitoba, Winnipeg, MB, Canada,

School of

Pharmacy, Memorial University of Newfoundland,

St John’s, NL, Canada

alkabbaw@myumanitoba.ca

Aim: This study aims to assess the persistence of

use of pharmaceutically-derived cannabinoid agents

and assess the potential socio-demographic char-

acteristics and medical conditions associated with

discontinuation of use.

Methods: This study used administrative data from

April/1st 2004 to March/31st 2017, from the Mani-

toba Population Research Data Repository located

at the Manitoba Centre for Health Policy (MCHP),

University of Manitoba. Incident users were included

and followed for one year from the date of ﬁrst pre-

scription dispensation. Data were analyzed, using

a competing risk regression model (Cause-speciﬁc

hazards model), with death from any cause as the

competing risk, to assess factors that may affect dis-

continuation rates. Time to discontinuation, in days,

was the dependent variable.

Results: Among 7050 pharmaceutical cannabinoid

users, 6835 were incident users. The mean (SD) age

of users was 52.5(15.2) and 59.3% were females.

Only 15.2% of incident users continued using can-

nabinoids after one year. The ﬁnal regression model

showed that age and income status had a signiﬁcant

effect on persistence of cannabinoid use. Among all

medical conditions included, ﬁbromyalgia, osteo-

arthritis, and substance abuse disorder had a sig-

niﬁcant effect on discontinuation rates with hazard

ratios (95%CI) of 0.86 (0.81-0.92), 0.85 (0.76-0.94),

0.91 (0.84-0.98), respectively.

Conclusion: In a naturalistic setting, the persis-

tence of prescription cannabinoid use was affected

by age, income, and speciﬁc medical conditions

of the incident user. The reason for these ob-

served differences is uncertain and warrants further

investigation.

43. Prevalence and risk factors associated

with persistent use of opioids after surgery: a

meta-analysis

Rehman Y, Guyatt G, Couban R, Busse J

National Pain Center, McMaster University

dry_rehman@yahoo.ca

Objectives: Opioids are commonly prescribed

to manage acute pain after the surgery; however,

ABSTRACTS CAPT / ACTP ANNUAL MEETING 2018

J Popul Ther Clin Pharmacol Vol 25(2):e33-e62; October 26, 2018.

This article is distributed under the terms of the Creative Commons Attribution-Non Commercial 4.0 International License.

e57

about 10% continue to consume opioids beyond 90

after the surgery. We will conduct a systematic re-

view and meta-analysis of observational studies to

establish the rate of persistent opioid use after sur-

gery, and explore risk factors associated with opioid

use beyond 90 days after surgery.

Methods: We will report our systematic review ac-

cording to the PRISMA guidelines. We will develop

comprehensive search strategies with no language

restriction, for MEDLINE, PsychINFO, EMBASE and

CINAHL in collaboration with an expert librarian. Liter-

ature screening and data extraction will be performed

by teams of reviewers, independently and in dupli-

cate. When possible, we will pool the association of

all predictors reported by two or more studies with

persistent opioid use after surgery. We will report our

results as odds ratios, absolute risk increase, and as-

sociated 95% conﬁdence interval s (CIs). We will ex-

plore heterogeneity and publication bias by the visual

inspection of forest plot and funnel plots, respectively.

We will explore heterogeneity with a priori subgroup

analyses (e.g. risk of bias, length of follow-up, type

of surgery). The risk of bias will be assessed with cri-

teria proposed by the Users Guides to the Medical

Literature. The quality of evidence will be determined

with the GRADE approach for each predictor.

Clinical Implications: Pain is important sequelae

of surgery. Acute post-operative pain intensity and

prolong opioid use correlates with PPSP; therefore

ﬁndings of our study will enhance understanding to

reduce the prevalence PPSP.

44. Leveraging CADTH’s Scientiﬁc Advice Program

for early feedback: from phase 2 studies to RWE

Lees M

, Selby R

, Hojjati M

Takeda Pharmaceuticals International AG,

Takeda

Pharmaceuticals International Co.,

Takeda Canada

Inc.

michelle.hojjati@takeda.com

Randomized controlled trials are considered the

gold standard for evidence of clinical efﬁcacy and

safety for regulatory and HTA bodies and use in clin-

ical practice. Designing the outright comparative de-

velopment programs that maintain clinical relevance

in rapidly advancing therapeutic areas is becoming

increasingly challenging. CADTH’s Scientiﬁc Advice

program engages scientiﬁc experts, health econ-

omists, HTA, and patients early in the drug devel-

opment process in order to provide an opportunity

to adjust drug development plans based on advice

from multiple stakeholders. Takeda is currently

seeking scientiﬁc advice from the program for a drug

in an area of high unmet need, currently in the pro-

tocol planning stage of phase II study development.

Advice is being sought on the clinical development

program, in order to increase the likelihood that this

innovative medicine will be available for patients as

soon as possible. Although CADTH’s scientiﬁc ad-

vice process is non-binding, it has the potential to

provide insight into possible areas of concern in the

proposed study protocol and RWE collection when

modiﬁcations are still possible. Takeda’s experience

with the program will be highlighted.

45. Management of post-traumatic stress disorder:

a protocol for a multiple treatment comparison

meta-analysis of randomized controlled trials

Rehman Y, Guyatt G, McKinnon MC, McCabe RE,

Lanius RA, Richardson JD, Couban R, Busse J

Health Research Methodology, McMaster University

dry_rehman@yahoo.ca

Objectives: We will conduct a network meta-anal-

ysis of all randomized controlled trials (RCTs)

evaluating therapies for PTSD to determine which

therapies show evidence of promoting functional re-

covery (e.g. return to work), and the relative effec-

tiveness of these treatments.

Methods: We will identify eligible trials, in any lan-

guage, by a systematic search of PILOTS.

46. Proposed changes to the federal patented

medicine pricing rules in canada: case study of a

manufacturer’s decision-making about regulatory

submission

rawson nsb

Eastlake Research Group, Oakville, Ontario

EastlakeRG@gmail.com

ABSTRACTS CAPT / ACTP ANNUAL MEETING 2018

J Popul Ther Clin Pharmacol Vol 25(2):e33-e62; October 26, 2018.

This article is distributed under the terms of the Creative Commons Attribution-Non Commercial 4.0 International License.

e58

Objective: To examine the proposed changes to the

regulations and guidelines of the Patented Medicine

Prices Review Board (PMPRB) and apply them to a

case study of the decision-making process that the

manufacturer of a new rare disorder medication is

likely to go through when assessing whether to seek

regulatory approval in Canada.

Methods: The current and proposed PMPRB guide-

lines are summarized and, using a hypothetical new

rare disorder drug as an example, the potential man-

ufacturer’s decision-making regarding the impact of

the proposed guidelines on its launch is examined.

Results: We assume the new drug will require an

average price reduction to all payers to bring the ef-

fective price down by 80-90% due to the application

of a new basket of comparator countries, pharma-

coeconomic criteria and market size.

Conclusion: What should the manufacturer do

about the drug’s price? Some possible decisions

are: (1) make a huge price reduction at regula-

tory approval, risking pricing in other countries,

(2) launch at the desired price and face a PMPRB

price investigation, (3) delay the launch in Canada

until after other countries, and (4) not to seek reg-

ulatory approval in Canada. The high level of un-

certainty being generated by the proposed changes

in the PMPRB’s guidelines will imperil the launch of

all new medicines in Canada because it will signiﬁ-

cantly decrease the attractiveness of the country as

a priority jurisdiction in which pharmaceutical com-

panies seek regulatory approval for innovative new

products.

47. Assessing the emotional, physical, and

ﬁnancial burden of caregivers for persons with

dementia

Baker AR, Tran F, Pisterzi LF, Mittmann N, Lam B,

Black SE.

Sunnybrook Research Institute

Luca.pisterzi@sunnybrook.ca

Objectives: Unpaid or informal caregivers play a

key role in supporting persons experiencing cog-

nitive decline; presently, literature on the burden of

informal caregiving is limited.This study aims to ex-

amine emotional, physical and ﬁnancial burden in

informal caregivers of persons with dementia.

Methods: Participants are recruited as they ac-

company a patient for which they are the primary

caregiver to a visit with a cognitive neurologist at

Sunnybrook Health Sciences Centre.A convenience

sample of 100-150 participants is currently being

sought.Participants complete a questionnaire cap-

turing demographics, the Zarit Burden Interview,

and the Sunnybrook Costing Tool (SBT), designed

to assess ﬁnancial toxicity.In the absence of an

appropriate tool to assess the ﬁnancial burden of

caregivers supporting persons with dementia, the

Comprehensive Scale for Financial Toxicity from

the FACIT Measurement System was modiﬁed to

address issues in this population in a manner that

is relevant.The patient’s performance on the Mi-

ni-Mental State Exam (MMSE) was also recorded.

Upon completion of recruitment, descriptive statis-

tics will be applied to the data.

Results: At the time of writing 29 participants had

been recruited to the study, over 60% of which are

spouses of the patients and are female.A prelimi-

nary assessment of the score on the Zarit Burden

Interview and the patient’s MMSE score reveals a

low level of burden in patients with greater cognition,

as is expected.

Conclusion: Upon completion, this study aims to

shed light on the burden of informal caregivers, and

to better inform policies developed to support them.

ABSTRACTS CAPT / ACTP ANNUAL MEETING 2018

J Popul Ther Clin Pharmacol Vol 25(2):e33-e62; October 26, 2018.

This article is distributed under the terms of the Creative Commons Attribution-Non Commercial 4.0 International License.

e59

Burden AM 4

Busse J 43, 45

Cadarette SM 4, 34

Campbell KL 21

Carcone S 36

Carleton B 8

Carney G 1, 12

Cesta A 10, 11, 29

Chambers J 38

Chan A 37

Chan B 30

Cheung MC 19

Clark R 9

Colgan S 27

Couban R 43, 45

Cowan RP 21

Craven C 30

Curtis JR 8

Dabbous F 21

Daeninck P 42

Dao S 7

Dempster W 38

Denburg AE 19

Dolovich L 37

Donnan J 33

Dormuth CR 1, 12

Dragomir A 6, 25, 36

Earle CC 18

Ellis KB 19

Evans WK 3, 15

Falls C 20

Fairbairn M 23

Feeny D 31

Feld J 8

Filion KB 17

Finelli A 6

Fischer A 41

Author Index #

Abrahamowicz M 8

Abuzour A 37

Ackroyd T 24

Adachi JD 9, 34

Alkabbani W 42

Alessi-Severini S 5, 28, 42

Alyward B 22

Amine Amiche M 34

Aprikian A 25, 36

Assasi N 19

Ashe MC 9

Athanasiadis G 4

Auger C 16

Avramioti D 24

Azoulay L 17

Bailey H 23

Baker AR 47

Ball G

Ban JK 4

Basappa N 6

Bassett K 1

Bassler KA 33

Bayley A 20

Bernatsky S 8

Bick R 14

Bini C 26

Black SE 47

Bolton JM 5

Bombardier C 10,11, 29

Bouganim N 17

Bremner K 36

Bronskill SE 40

Brown MC 20

Buchman S 18

Bugden S 42

Bui V 37

AUTHOR INDEX

ABSTRACTS CAPT / ACTP ANNUAL MEETING 2018

J Popul Ther Clin Pharmacol Vol 25(2):e33-e62; October 26, 2018.

This article is distributed under the terms of the Creative Commons Attribution-Non Commercial 4.0 International License.

e60

Gamble JM 33

Gradin S 18

Gatley JM 40

Giangregorio LM 9

Gibbs JC 9

Giuliani M 20

Goldstein D 20

Goodman SG 27

Goeree R 27

Gomes T 16, 34

Gran-Ruaz S 22

Greaves S 16

Grima D 2

Grubert N 22

Gruneir A 40

Guertin JR 31

Gupta M 41

Guyatt G 43, 45

Hancock-Howard R 24

Hansen A 6, 20

Hassan S 9

Hayes KN 4

Haynes AE 19

Hepworth E 10, 29

Heng D 6

Herrmann, N 40

Hojjati M 26, 44

Hong NJL 18

Hope A 20

Hueniken K 20

Hughes A 22

Hume M 27

Humphries B 31

Hurry M 2, 3, 15

Hu J 25

Jang R 20

Janzen D 5

Johnston K 33

Kapoor A 6

Kawahara SH 21

Kendler D 9

Khaira M 37

Khosa G 28

Khosrow-Khavar F 17

Klein M 8

Kollmannsberger C 6

Krahn M 36

Kuo I 5, 28

Kurteva S 39

Lam B 47

Lancaster K 37

Lanius RA 45

Larche MG 10

Lau A 29

Lee TC 39

Lees M 44

Leon C 42

Leong C 5

Levesque LE 34

Li X 11

Lipton RB 21

Liu N 18

Lok A 37

Machado MA 8

MacKinnon M 18

Mani A 22

Mancini J 41

Marcellusi A 26

Marra CA 33

Marrie RA 42

Mathers A 37

Martins D 16

McCabe RE 45

McCarthy LM 40

McKinnon MC 45

Mennini F 26

Mirshams M 20

Mirza R 10, 30

ABSTRACTS CAPT / ACTP ANNUAL MEETING 2018

J Popul Ther Clin Pharmacol Vol 25(2):e33-e62; October 26, 2018.

This article is distributed under the terms of the Creative Commons Attribution-Non Commercial 4.0 International License.

e61

Mitsakakis N 36

Mittmann N 9, 18, 47

Moldaver D 2

Moltzan CJ 19

Morrow RL 1, 12

Moura CS 8

Movahedi M 10, 11, 29

Muhimpundu CV 7

Myers S 28

Nablsi E 36

Nadeem K 16

Najafzadeh M 33

Nazha S 6

Ng R 32

Nguyen H 33

Nguyen TTY 7

Oh P 27

O’Quinn S 22

O’Reilly D 35

Pandey AS 41

Papaioannou A 9

Pavlovic JM 21

Pericleous L 27, 41

Pisterzi LF 47

Pope J 29

Pouliot F 6

Primiani J 36

Pulicharam R 21

Rahimi M 20

Rampakakis E

Rawson NSB 46

Reed ML 21

Rehman Y 43, 45

Rendon RA 6

Richardson JD 45

Rochon PA 40

Rogoza RM 23, 27

Rosella LC 32

Ryan W 36

Sabarre KA 19

Sampalis J3 29

Sati S 18

Savage RD 40

Seung SJ 3, 9, 15, 18

Selby R 44

Shakeri A 34

Shepherd J 23

Sidelnikov E 41

Silberstein SD 21

Singh S 16

Siu L 20

So AI 6

Soulieres D 6

Spreaﬁco A 20

Stall N 40

Suissa S 17

Sutradhar R 32

Swab M 33

Syed I 24

Tadrous M 16

Tai M-H 23

Tamblyn R 39

Tanguay S 6

Tarride JE 31

Templeton JA 9

Thabane L 9, 37

Tims K 22

Tran D 2

Tran F 48

Trudeau M 19

Vaillancourt Z 36

Vernia M 26

Visentin JD 40

Viswanathan HN 21

Waldron J 20

Walton R 3, 15

Wang J 20

Wark JD 9

ABSTRACTS CAPT / ACTP ANNUAL MEETING 2018

J Popul Ther Clin Pharmacol Vol 25(2):e33-e62; October 26, 2018.

This article is distributed under the terms of the Creative Commons Attribution-Non Commercial 4.0 International License.

e62

Weir DL 39

Winthrop K 8

Wodchis W 30, 32

Wong-Rieger D 13

Wood LA 6

Wright F 18

Wright JM 1

Xu W 20

Yanev I 36

Yu JS 21

Zhang Y 20

ABSTRACTS CAPT / ACTP ANNUAL MEETING 2018

J Popul Ther Clin Pharmacol Vol 25(2):e33-e62; October 26, 2018.

This article is distributed under the terms of the Creative Commons Attribution-Non Commercial 4.0 International License.