Diagnostiek op maat: de balans tussen snelheid en patiëntvriendelijkheid Ger Koole PICA: VU-Vumc kenniscentrum voor zorglogistiek Congres Architectuur in de zorg Nieuwegein, 21 juni 2012
Zorgpaden • van verticaal naar horizontaal • parallel met produktieprocessen • tbv kwaliteit (evidence-based, data verzamelen) • tbv logistiek – one-stop shop – hele keten in 1 keer plannen
• some alleen deel zorgtraject (diagnostiek) • over muren heen (vgl supply chains) • veel toegepast voor “planbare” zorg
Planbare zorg • Volledig voorspelbaar – vaste instroom – vaste routering – (vrijwel) constante behandeltijden
• instroom fluctueert meestal – oplossing: alleen deel patiënten in zorgpad
• resultaat: weinig patiënten in zorgpaden
Minder planbare zorg • Niets is niet planbaar • Hoe onvoorspelbaarder, hoe slechter afweging kwaliteit/efficientie • Resultaat: slecht gebruik apparatuur, lange wachttijden, doorverwijzingen, … • Kan ook in zorgpaden • Hoe plan je dat?
Zorgpadoptimalisatie • weghalen wachttijden/overprocessing (lean) • capaciteit reserveren voor zorgpaden – slimme oplossingen voor maximale benutting
• opnieuw inrichten proces – bijv. eerst onderzoek dan dokter
• vaak grote reducties doorlooptijden mogelijk – consequenties andere patiëntgroepen beperkt
Doelstellingen • Onderzoek (in progress): patiënttevredenheid neemt nauwelijks toe bij korte doorlooptijden • Oorzaken: – korte doorlooptijden worden (redelijk) gewaardeerd – niet gelijk naar dokter veel minder – spreiding in patiënten
• Wat wil je bereiken? – Efficiënt en snel proces of klanttevredenheid?
Casus • Increasing competition between hospitals in NL • VU university medical center (VUmc) got funding for new oncology clinic • Concepts: – healing environment – multi-disciplinary – rapid
Cancer diagnostics • Diagnostics is perhaps most complicated part of treatment • Requires many different departments/specializations • Can take weeks • From referral (by GP/other hospital) to treatment plan + patient informed
Rapid diagnostics • Implemented in several hospitals • Usual model: – “Diagnostic day” for clearly defined cancer type and number of patients – Completely predictable
• VUmc model: – For all cancer types and all patients – Main challenge: variability in diagnostic path and number of patients
VUmc goals • Initial goal: all patients diagnozed in 48h – general idea: actual time required for diagnosis << 2 days
• However: – quality requires sometimes extra time (academic function, SL should not be leading) – patients do not always wish diagnosis in 48h (for physical and emotional reasons)
• “Patient-centered” diagnostics – Rapid when required or desired
Approach • Step 1: make flowcharts – high diversity (between and inside clinical pathways) – hard to get data
• How to get started?
Model for diagnostics • Diagnostics is “netwerk” of activities (visits to doctor, scan, etc.) • Different type of capacity needed for activities (doctor, multi-disciplinary team meeting, PA, MRI, etc.) • Every clinical pathway has a very different diagnostical “network” • Focus on 80% that is equal • Focus on critical path
Example FVC
MRI
punction PA
scopie
MDT
Other clinical pathways
• Mamma: FVC
PA
MDT
MRI
• Colon: scopie
if strong suspicion
PA
MDT
MRI/CT
• Prostate: FVC
PA
MDT
MRI
• Hematology: FVC
• Lung:
PA FVC
MRI/PET-CT PET-CT CTguided punction
MDT MDT
PA
Critical path Activities that often occur on critical path: • FVC: first visit to clinic • PA=pathology: long sojourn time • MRI: long waiting time • MDT=multi-disciplinary team meeting: once a week
Critical path • Crucial for rapid diagnostics: – PA has short sojourn time – MRI has short waiting time
• On the basis of SL agreements for PA and MRI: – Fix new moments for FVC and MDT – Usually: count backwards from MDT
PA • Consists of multiple subactivities • Of which 1 batch step (every night) • Result: – waiting during whole day – high workload in the morning
• Much better: continuous process • Requires: new machine
Batch vs continuous • Other advantages: – smoother employees schedule – leading to longer opening hours – more focus on customer wishes – better service
• Other logistic improvements – priority to cancer diagnostics – priority only is not sufficient for SL targets – sojourn time on other PA has not increased
Results PA • Graph with sojourn times cancer diagnosis before/after implementation rapid diagnosis
average 80% percentile
2009 2009 2009 2009 2010 2010 2010 2010 2011 2011 Q1
Q2
Q3
Q4
Q1
Q2
Q3
Q4
Q1
Q2
Radiology • 2 models: – on appointment – walk-in
• Walk-ins possible for CT scans – not a bottleneck (anymore) – short scan time
• MRI on appointment – – – –
bottleneck can take as long as 60 minutes waiting time too long Solution: slots for semi-urgent/diagnostics
• PET/CT – on appointment because of medication
Slots • Challenge: random demand • Risk of empty slots or shortage • Scale (sharing slots between clinical pathways) reduces these risks (relatively) – not always possible – length of slot depends on type of MRI
• Solution: reallocate slots if not occupied – for example for clinical patients – only possible for “next-day” scans
Occupation before and after reallocation
How customer-friendly is rapid? • Initial definition: within 24 hours • Our experience: – one-stop shop too much a burden – patient needs to get organized (family) – patient needs to get used to the idea – wishes depend on patient
• Patient-centered diagnostics
