Embed Size (px)
Citation preview
15/05/2012
1
IT to improve the safetyof the medication process
at the hospital
Prof. Pascal BONNABRY
Berner Fachhochschule
Geneva, May 25, 2012
Agenda
13h45 Theoretical introduction14h30 Di i14h30 Discussion14h45 Break15h00 Visit in 2 groups
- Pharmacy- Central warehouseCentral warehouse
16h30 Discussion17h00 End
15/05/2012
2
Geneva university hospitals(HUG)
Geneva university hospitals(HUG)
Consortium of public hospitals in the Geneva county
One of the 5 swiss university hospitals
1900 beds
≈ 50’000 hospitalisations/year
≈ 800’000 outpatientsconsultations/year
www.hug-ge.ch
15/05/2012
3
Pharmacy: Missions
• Supply drugs to the whole hospital(purchase stock management distribution) (purchase, stock management, distribution)
• Implement a full traceability of drugs until the administration to the patient
• Ensure a production and a personnalisedpreparation of drugs, when they are not available on the market
• Contribute to a safe and rational use of drugs
Services - Education - Research
Organisation
Q li d fPascal BonnabryHead of
pharmacy
Quality and safetylink with services
Laurence CingriaQuality management
N th li V F hid S d hi S d i Fl C li F
ResearchEducation
Pharmaceutical assistance
Quality controlLogistics Production
Clinical pharmacyCollaboration with
clinical pharmacology
Nathalie Vernaz Farshid Sadeghipour Sandrine Fleury Caroline Fonzo
15/05/2012
4
Drug medication process
• Support the teaching of drug
Drug selectionInterests of a restricted list
Support the teaching of drugpharmacotherapy to the physicians
• Limit the risk of errors by reducingthe number of differentdrugs available
• Favour an optimal Favour an optimal stock management
• Negociate attractive purchase conditions
15/05/2012
5
Process organisations
• Global distributionThe pharmacy delivers boxes of drugs and nurses dispense individualtreatments from the ward stock
• Nominal or individual distribution Drug dispensing is performedat the pharmacy, for each patient,based on the prescription
Ordering ofdrug boxes
Global distribution
Prescriptiondrug boxes
Distribution of boxes
Administration to patient Ward
stock
Deliveryof boxes
Drugdispensing
15/05/2012
6
Individual distribution
Prescription
Pharmaceutical validation
Administration to patient ( )
Drug dispensing
Deliveryof doses
Global or individual ?
• Individual distribution is more convenient in some conditions– Few prescription modifications (chronic care)
– Pharmacy close to the wards
• This doesn’t fit to HUG situation– Acute care in majority– Long distance between the pharmacy and some
wards (multi-sites hospital)
15/05/2012
7
The medication processIndustry stock
Prescription
CytosTPN
Pharmacy stock Ward stock Dispensation Production
EPR
MP
Production stock
Raw-materialsanalysisProductionEnd-product
analysisAdministration
to patients
Safety problems ?
15/05/2012
8
Human reliability
O h 6 h d G d d « On the 6th day, God created man … »
… but God was tired, and histi t f tcreation was not perfect …
The addition of two errors
Commission error AND Control failure
SelectionCalculationCounting
CheckDouble-check
Check-listElectronic
15/05/2012
9
• Error rate= 1 %
Distribution errors(real life)
24%
20%
56%
Gschwind L, Carrez L, François O, HUG, 2006-11
20%
Counting Omission Selection
Dispensing errors(experimental)
• Error rate= 3 %
74%
20%6%
Selection error
Repartition error
Counting errorGarnerin P, Eur J Clin Pharmacol 2007;63:769
15/05/2012
10
Administration errors
• Error rate= 19 %
Observation study in 36 institutionsBarker KN, Arch Intern Med 2002;162:1897
Limited performance of controls
• Introduction of errors during unit dose dispensingdispensing
• Detection ability during human-performedcontrol:
• Pharmacists: 87.7%• Nurses: 82.1%
Facchinetti NJ, Med Care 1999;37:39-43
Efficiency ≈ 85%(known value in the industry)
Do not be too confident with the double-checks!
15/05/2012
11
Implementation of IT in the medication process
Diogène 1 - 1978
Potential interests of IT
• To improve– The safety
by reducing the rate of errors and improving the reliability of controls
– The traceabilityby facilitating the registration of logs
– The efficiencyby increasing the working performance
– The communicationby connecting the different steps of the processes
15/05/2012
12
Many questions before to start
• Positive impacts ?• New risks ?• New risks ?• Return on investment ?• System selection ?• Commercial or homemade ?• Interoperability ?• User’s training strategy ?• User s training strategy ?• Acceptability ?
Electronic systemsto catch errors
Industry stock
Prescription
CytosTPNDelivery
Pharmacy stock Ward stock Dispensation Production
EPR
TPNMP
( )
Delivery
Production stock
Raw-materialsanalysisProductionEnd-product
analysisAdministration
to patients
15/05/2012
13
Electronic systemsto avoid errors
Industry stock
Prescription
CytosTPN
CPOE / CDSS
Pharmacy stock Ward stock Dispensation Production
EPR
TPNMP
EDI
CDSS
Production stock
Raw-materialsanalysisProductionEnd-product
analysisAdministration
to patients
Robotisation of drug distributionSafety
1 % 0.5 % 0.2 %Phase I Phase II
n=5805 n=4365 n=1497
François O et al, HUG, 2012
Selection 0 % 0%Convoy 0.44% 0.20%Manual finalisation 0.04% 0%
n=5805 n=4365 n=1497
software
15/05/2012
14
Robotisation of drug distribution Efficiency
• Hypothesisd f kl d ti t d t 2 FTE– decrease of workload estimated to 2 FTE
• Results– Reality seems to confirm this estimation– Real workload decrease is actually measured…
• Return on investment (ROI)2– 2 FTE suppressed
– 3.1 year, without the operating costs– 5.4 year, with the operating costs
(amortization, maintenance contract)
• Experimental
Automation of drug dispensingSafety
3
0
0.5
1
1.5
2
2.5
3
Erro
r rat
e [%
]
without
with
Du Pasquier C, Riberdy L, HUG, 2003
15/05/2012
15
Automation of drug dispensingEfficiency
• Real life (digestive surgery ward)600
Min
utes
/wee
k
- 8h +4h +1h
200
300
400
500
600sans armoireau débuten routine
without
beginning
routineM
François O et al, HUG, 2011
0
100
200
3AL Assistante PharmacienNurse Technician Pharmacist
Administration to patientsBedside scanning
• CytostaticsDrugNurse g
Physician
Database
Patient
15/05/2012
16
Administration to patientsSafety
• Benefit of bedside scanning
• Wrong drug - 75%
• Wrong dose - 62%
• Wrong patient - 93%
• Wrong administration time - 87%
Gl b ll 80%Globally - 80%
Johnson, J Healthcare Inf Manag 2002;16:1
Automated dispensing systemRobotized
distribution
The medication processFinal perspective
Clinicalinformation
system
EDI
Pharmacystock
Manufacturerstock
Wardstock
CPOE
Logisticinformation
systemBedside scanning Distribution
with scanning
15/05/2012
17
Computer-assisted production management
• ObjectivesT t t f ti b IT – To support any type of preparation by IT
• Batches• Cytostatics• TPN• Other individualized prescriptions
– To implement in-process electronic controls duringthe most critical stepsthe most critical steps
– To link individualized preparation to theirprescription and their administration
Production managementAutomation
Nutrition (Baxa)
Cytostatics(CytoCare)
CIVAS (Smartfiller)
PharmaHelp (Medical Dispensing Systems), Riva, …
15/05/2012
18
Production managementGravimetry
• Cytostatics(Cato Cypro )
• Batch production(Cato, Cypro )
Production management Efficiency
• Production – Parenteral nutrition (10/day)
- 60%
Manual2 x 3h = 6 h
Automated (BAXA)1 x 2.5h = 2.5 h
15/05/2012
19
Prerequisite to successfulIT implementation
• Electronic management of processesElectronic management of processes(CPOE, stocks, …)
• Technical infrastructure (hard-, soft-)
• Actors identification (caregivers, patients, drugs)
• Acceptability (patients, caregivers)
• Adaptation to processes• Adaptation to processes• Project leadership• Financing
Actors identificationThe patient
The caregiver The drug
15/05/2012
20
Drug identification
Reconditionedby the pharmacy
Identified
?Identified
by the industry
Drug identification
Unit dose• Hierarchy
Unit dose
Secondary package
Hospital package
Box
Pallet
= international standard
15/05/2012
21
GS1 codification of pharmaceuticals at HUG
• Product ID (cytostatics)y
01 07613167000009 7003 1103161400 21 cyt-11198499GTIN - cytos EXP (date and time) Serial
GS1 codification of pharmaceuticals at HUG
• Product ID (batch production)p
01 17613167001249 17 120831 10 PDS-11289663GTIN BatchEXP date
15/05/2012
22
Adaptation to processes• Reasons for workaround
– Process– Process• Training requirements• Process flow (administration of drug
before scanning, shortage of time)
– Technology• Hardware (performance of scanners)• Software (delays in response)• Barcode (difficulties in reading)
– Resistance• Communication• Changing role• Negative perception of IT
Van Onzenoort HA, Am J Health-Syst Pharm 2008;65:644Nanji KC, J Am Med Inform Assoc 2009;16:645
How to progress?
• Determine an institutional strategy and an implementation schedule taking into accountimplementation schedule, taking into account– the local organisation– the local culture– the expected return on investment
• Involve the different partners• Re-think the process organisation
(re-engineering)• Manage each projet independently, without
loosing the global vision
15/05/2012
23
Conclusion
• The medication process at the hospital iscomplex and involves many differentcomplex and involves many differentprofessionals
• A clever organisation contributes to improvethe safety, the efficiency, the communication and the traceability
• IT take more and more importance in the i h h iprocess improvement approaches: their
implementation is necessary but is a challenge• Each hospital must determine a strategy, based
on the local context
