Transcript pharmacist-led medication review

Does Pharmacist-Led Medication Review Improve
Health Outcomes?
Evaluation of a Quality Improvement Project
February 19, 2015
Quality Forum
Corinne M Hohl MD FRCP MHSc
Attending Physician, Dept of Emergency Medicine, VGH
Associate Professor, Dept of Emergency Medicine, University of British Columbia
Scientist, Centre for Clinical Epidemiology and Evaluation
CIHR New Investigator
ADE: unintended and harmful effects of medications.
• Most common cause of preventable iatrogenic morbidity.
• Patients with ADEs commonly present to EDs:
– 12% of ED visits amounting to 20,000 visits per year to VCHA
– 35-40% of are not attributed to medication use by MDs
– 50% greater odds of spending additional days in hospital/month.
• Leading cause of unplanned admissions.
Co-funded by BC HSPO & VCHA in 2011
Aim:
To evaluate the effect of pharmacist-led medication
review in high-risk ED patients on health outcomes
compared to usual care (medication reconciliation).
Innovation:
A new model of delivering pharmacy care to ensure that
patients receive optimal medication therapy from when
they arrive.
Design:
– Prospective comparator study, non-blinded that
was nested within a continuous quality
improvement project at 3 sites.
Population:
– High-risk adults presenting to a participating
ED at a time an ADE pharmacist was present
(including evenings and weekends)
Intervention: Medication Review by Clinical
Pharmacists
– Best-possible medication history & initiated med reconciliation
– Review of medications critically to identify ADEs, and
medication-related problems
– Communication with emergency MD/GP/geriatric triage nurse if
discharged or emergency/admitting MD if admitted.
Control:
– Med reconciliation by admitting/consultant services &
medication review on the ward among admitted.
• Reconciliation an accurate medication history from a
variety of sources based on which medications are represcribed.
Aim: to avoid medication errors at care transitions.
• Review an accurate medication history followed by a
structured, critical examination of a patient’s medications
Aim: to reach agreement on the goals of therapy,
minimize errors, optimize medication effectiveness,
identify medication-related problems and ADEs.
Outcomes:
1o
Proportion of days spent in hospital within 30 days of the
index ED visit (encompasses index admission and days
readmitted)
2o
Proportion admitted
Unplanned ED visits within 7d
Mortality (all-cause)
Patient Flow
ED visits (VGH, LGH, RH)
(n=135,323)
High-Risk eligible
(n=10,805)
Excluded (n=124,516)
▪ Low-risk or unknown (n=95,837 )
▪ Age < 19yrs (n=882)
▪ Pharmacist not available (n=22,675)
▪ CTAS=1 unknown (n=236)
▪ Multisystem Trauma (n=302)
▪ Scheduled re-visit (n=1,643)
▪ Sexual assaults (n=4)
▪ Post-operative complications (n=224)
▪ Social problems (n=317)
▪ Pregnancy-related complication (n=18)
▪ Death on arrival/in ED (n=406)
▪ Repeat visits (n=1,964)
▪ Unresolved linkage (n=1)
▪ Left Against Medical Advice/Missing data (n=9)
Systematic
Allocation
Med Rev
(n=6,416)
Control
(n=4,389)
Baseline Characteristics – Overall
Med Rev Control
(n=6,416) (n=4,389)
71
69
Female, %
56.4
55.1
Mean No. Meds
8.1
7.7
Lowest SES quintile, %
11.7
11.2
Ambulance Arrival, %
37.1
32.9
Daytime Arrival, %
61.2
61.3
Weekend Arrival, %
13.2
12.4
Median age, yrs
Baseline Characteristics – Overall
Med Rev Control
(n=6,416) (n=4,389)
71
69
Female, %
56.4
55.1
Mean No. Meds
8.1
7.7
Lowest SES quintile, %
11.7
11.2
Ambulance Arrival, %
37.1
32.9
Daytime Arrival, %
61.2
61.3
Weekend Arrival, %
13.2
12.4
Median age, yrs
Median No. Hospital Days over 30
Days of Follow-Up
All Sites
<80 yrs
Difference in Days
( 95% CI)
p-value
-0.48 days
0.058
(-0.96 to 0.0)
-0.6 days
(-1.17, -0.06)
0.03
Pts in the medication
review group spent a
median of 12h in
hospital less over 30d,
p=0.058
0
3
6
9
12 15 18 21 24 30 days
Days in Hosp
over 30d
All
Sites
Median
Difference
( 95% CI)
pvalue
-0.48
0.05
8
(-0.96 to
0.0)
-0.60
<80
(-1.17, 0.06)
0.03
Admissions
OR
(95% CI)
pvalue
0.98
(0.90,
1.06)
ED Revisits
OR
(95% CI)
0.98
0.54
pvalu
e
Mortality
OR
(95% CI)
pvalue
1.11
(0.85,1.13 0.82 (0.96,1.30 0.16
)
)
% Days in
Hospital
over 30d
All
Sites
8.5%
reduction in
hosp days
over 30d
<80 yrs
11.5%
reduction in
hosp days
over 30d
Admission
Rate
ED
Revisits
Mortality
No
Difference
No
Difference
No
Difference
• At VGH:
– n=1,928 high-risk patients whose medications were
reviewed in the ED & who were admitted.
•
•
•
•
Avoided 1041 hospital days (0.54 days/patient).
Resulted in avoided costs of $468,000 ($450/hospital day).
Costs of maintaining the program: $400,957 in pharmacy salaries.
The cost of nursing salaries & administrative costs (IT, coordinator,
evaluation) removed after implementation period.
– Pharmacist-led medication review in ED was cost avoidant
by reducing hospital days.
We have an opportunity to implement best-evidence
and meet the Accreditation Canada ROP:
1. We have developed evidence-based clinical
decision rules to identify high-risk ED patients, and
have implemented them at VGH, LGH and RH to
identify & select high-risk patients.
1. Our results suggest that if pharmacists review the
medications of high-risk patients, this will lead to
LOS reduction and get patients home sooner.
• Intervention was not fully optimized because of
rotating staffing model: time spent on training new
staff & relationship building can be reduced.
• Calculations do not take into account any other
benefit of the program.
Where next?
Status Quo: Medication Information in BC
Health Care Provider
Outpatient
Medications
Ministry of Health
Hospitals
Community
Based
Others
Outpatient
PharmaNet
Medications
Pharmacies
Computer-based/automated
Paper-based/patient-driven
Vision: Medication & ADE Information
Health Care Provider
Outpatient
Medications
Ministry of Health
Hospitals
Community
Based
Outpatient
Pharmacies
PharmaNet
Others
PharmaNet becomes a “de facto” provincial
medication information “EHR”
Pharmacists and Nurses at all sites
Pharmacy Departments
Nilu Partovi, Fruzsina Pataki, Mark Collins, Terri Betts, Robert
McCollum, Ruth Tsang, Susanne Moadebi, Jane de Lemos
Emergency Physicians, Hospitalists, Internists
VCHA SET: Jeff Coleman, Duncan Campbell
BC HSPO: Les Vertesi