QAware
The Futureware Company
We are there when success is non-negotiable.
We deliver. Guaranteed.
Process Transparency
Through Data Mining
A Hospital Case Study
Harald Störrle, QAware
DataGeeks Meetup 2025-12-16
Process Transparency Through Data Mining
A Hospital Case Study
Procedures in hospitals require complex logistics involving dozens of
stakeholders and multiple IT systems. Despite best efforts by
experienced experts, deviations from the operation schedule occur
time and again resulting in delays. If push comes to shove, surgeries
have to be postponed which impacts both patients and medical staff,
and may results in significant costs for the hospital, stress for medical
staff, and worse medical outcomes.
Using a case study from the University Hospital Munich (Klinikum der
LMU), we demonstrate how such complex processes can be analyzed
through a smart combination of quantitative and qualitative methods. In
particular, we apply business process modeling, process mining, data
analysis, and interactive visualizations to achieve sustainable process
improvements.
Dr. Harald Störrle is Lead IT Consultant at QAware. He has been
working in the field of requirements, processes, and modeling since
2000. He collaborates with stakeholders and business departments,
translating their needs and goals into a form that enables developers
and architects to build great software. In a parallel universe, he is a
scientist focusing on empirical methods to foster evidence based
software engineering.
Agenda
Background The perioperative process in hospitals
Methodology Mixed Methods Process Discovery
Data gathering Process Mining is a useful tool
Insights Substantial improvements are possible
Communication Empowering process stakeholders is magic
Data analysis Sometimes, it’s not so hard
We are there where success is a must.
We deliver. Guaranteed.
Key data
250
employees
35 M€
Turnover
Munich,
Germany
Mainz
Darmstadt
Rosenheim
Successful in the
most demanding
projects for 20
years
Top
provider:
NPS 100
Top employer:
97% say:
"QAware is a
a very good
place to work"
VC subsidiary
company
Who we are
We are an independent consulting and
project house for customized software
solutions and transformative
technologies
We develop and revitalize futureware:
software that moves.
1. We take responsibility and offer
a contractually fixed guarantee
of success
2. We reduce complexity by acting
comprehensively: We advise,
manage and develop
3. We don't just develop software,
we transform companies with
software
Business critical
software
Differentiating
solutions
Transformative
Technology
6
Process Transparency
Through Data Mining
A Hospital Case Study
Harald Störrle, QAware
DataGeeks Meetup 2025-12-16
https://www.bpb.de/themen/gesundheit/gesundheitspolitik/549730/ausgaben-und-finanzierung-des-gesundheitssystems/
https://www.destatis.de/DE/Themen/Gesellschaft-Umwelt/Gesundheit/Gesundheitsausgaben/_inhalt.html
Health Care Expenditures
8
500 bn €
(Germany, 2024)
https://www.dkgev.de/fileadmin/default/Mediapool/1_DKG/1.7_Presse/1.7.1_Pressemitteilungen/2025/2025-02-17-DKI-Studie_Europaeische_Gesundheits-_und_Krankenhauskosten_im_Vergleich.pdf
https://www.vdek.com/content/dam/vdeksite/vdek/daten/d_ausgaben_krankenhaus/gkv_ausgaben_krankenhaus_mrd_eur_2019_2024_endgueltig.jpg
Health Care Expenditures
9
● The LMU Klinikum is one of the largest hospitals in Germany
(5,500 staff in 31 departments, >2k beds with >300k cases
p.a.).
● Surgical procedures account for 40% of expenses and 60% of
revenue.
● The Perioperative Process stretches from beginning to end of
anesthesia; existing duration predictions are poor.
● Sqior medical provides IT infrastructure and
applications for vital hospital logistics, supporting
several of largest hospitals in Germany.
● QAware has a stake in Sqior, and has provided them
with process mining software and data analysis. sqior medical
Context
IT Improvement
Automation
Augmentation
Business Process Management
● Focus on people and tasks
● Document planned process
● Process improvement, BPR
Process Mining,
…Observability, Analytics, Transparency
Process Mining
● Focus on Event Logs
● one-time analysis
● historic data
Process Observability
● continuous analyse
● Real-time data
● interactive visualisation
● empower process stakeholders
Data Science
● Analysis of particular KPIs
● Hypothesis test
Business Processes
Theory… and practice
Complete process transparency requires both angles
– and more…
Anesthesia
Patient logistics Surgery
Observation 1
The perioperative process is structured
Ordering
Transport
Induction
Preparation
Surgical
Procedure
Waking
Cleaning
Ausschleusen
Observation 2
Process steps show up in the data
Process Discovery using pm4py
Folienvarianten
KPIs
Replacing subjective opinions to
decisions based on data
Folienvarianten
Plan stability
Postponed procedures are racing
for patients and staff alike. Not to
mention cost.
Observation 3
Multiple factors may influence operation duration
Folienvarianten
Observation 4
Expert knowledge is crucial
Factor Analysis
11 076 2 081 135
procedures with
anesthesia
distinct
descriptions
clusters of similar
descriptions
15
…including knowledge
of clinicians
● MS & Magensonde
● BDK & Blasendauerkatheter
● SpA &Spinalanästhesie
● DIB & Ischiadicus Block & SchiadikusB & IschB
● FemB & FemBlock & Femoralisblock & Fem & Fem.Block
● SahB & Saphenusblock
● Art (wach) & Wach-Art & Wach,Art & Wach-ART & Wacharterie
● Art & Arterie
● Sedierung & Analgosedierung
● Spiral & Spiraltubus
● Video & Videolaryngoskopie & Videolaryngoskop & Video-ITN
● usw.
● “Sonstige"
7
1. Allgemeine Anästhesie (Narkose)
2. Regionalanästhesie
3. Atemwegssicherung und Beatmung
4. Gefäßzugänge & Katheter
5. Notfallmaßnahmen & Reanimation
6. Blutmanagement & Transfusionen
7. Sonstige Maßnahmen
…actually useful
clusters
Observation 5
Actual and planned durations deviate
Observation 6
Planned durations exhibit artefacts
Observation 7
Planned durations too large
PLANNED
(pre-registered)
ACTUAL
(measured)
Duration Anesthesia (aggregated)
Plan stability
Postponed operations are taxing for
staff and patients, not to mention
cost.
Observation 8
…surprise
PLANNED
(pre-registered)
ACTUAL
(measured)
Duration Anesthesia (aggregated)
Plan stability
Postponed operations are taxing for
staff and patients, not to mention
cost.
Insights
Process Transparency is more than Process Mining
Img: flaticon.com, Authors: Cap Cool, Freepik, Andrejs Kirma, Eucalyp
Extract & enrich
event log data
Fold logs to nets
Extract process chains
Understand
processes and problems
in domain context
Holistic
Prozessketten
extrahieren
Analyze & visualize
data in detail
Statistics & ML
Observe, display &
steer operational
processes
Sustainable
Tool based Automatic
conventional Process Discovery
augmentation,
not automation
● high variation
● strong context dependency
● resists automation
…breaks AI
qualitative approach (hypothesis driven)
complemets
quantitative approach (data driven)
Take Aways
● Process Mining is a great
stepping stone for data
analysis.
● Low hanging fruit are to be
had, still.
● AI is not the answer to every
question.
● Simpler is better. Always.
WTF?! Gaaah! Hang on… I see!
QAware GmbH | Aschauer Straße 30 | 81549 München | GF: Dr. Josef Adersberger, Michael Stehnken, Michael Rohleder,
Mario-Leander Reimer| Niederlassungen in München, Mainz, Rosenheim, Darmstadt | +49 89 232315-0 |
info@qaware.de | www.qaware.de
Intrigued?
Let’s talk
Harald Störrle
Lead IT Consultant
Harald.Stoerrle@QAware.de
+49 151 5092 1851
@haraldstoerrle
Get the latest on tech trends, our meetups,
project insights and everything else
happening at QAware.
Stay updated with the QAware Insights Newsletter!
Scan to Subscribe

Process Transparency Through Data Mining

  • 1.
    QAware The Futureware Company Weare there when success is non-negotiable. We deliver. Guaranteed.
  • 2.
    Process Transparency Through DataMining A Hospital Case Study Harald Störrle, QAware DataGeeks Meetup 2025-12-16
  • 3.
    Process Transparency ThroughData Mining A Hospital Case Study Procedures in hospitals require complex logistics involving dozens of stakeholders and multiple IT systems. Despite best efforts by experienced experts, deviations from the operation schedule occur time and again resulting in delays. If push comes to shove, surgeries have to be postponed which impacts both patients and medical staff, and may results in significant costs for the hospital, stress for medical staff, and worse medical outcomes. Using a case study from the University Hospital Munich (Klinikum der LMU), we demonstrate how such complex processes can be analyzed through a smart combination of quantitative and qualitative methods. In particular, we apply business process modeling, process mining, data analysis, and interactive visualizations to achieve sustainable process improvements. Dr. Harald Störrle is Lead IT Consultant at QAware. He has been working in the field of requirements, processes, and modeling since 2000. He collaborates with stakeholders and business departments, translating their needs and goals into a form that enables developers and architects to build great software. In a parallel universe, he is a scientist focusing on empirical methods to foster evidence based software engineering.
  • 4.
    Agenda Background The perioperativeprocess in hospitals Methodology Mixed Methods Process Discovery Data gathering Process Mining is a useful tool Insights Substantial improvements are possible Communication Empowering process stakeholders is magic Data analysis Sometimes, it’s not so hard
  • 6.
    We are therewhere success is a must. We deliver. Guaranteed. Key data 250 employees 35 M€ Turnover Munich, Germany Mainz Darmstadt Rosenheim Successful in the most demanding projects for 20 years Top provider: NPS 100 Top employer: 97% say: "QAware is a a very good place to work" VC subsidiary company Who we are We are an independent consulting and project house for customized software solutions and transformative technologies We develop and revitalize futureware: software that moves. 1. We take responsibility and offer a contractually fixed guarantee of success 2. We reduce complexity by acting comprehensively: We advise, manage and develop 3. We don't just develop software, we transform companies with software Business critical software Differentiating solutions Transformative Technology 6
  • 7.
    Process Transparency Through DataMining A Hospital Case Study Harald Störrle, QAware DataGeeks Meetup 2025-12-16
  • 8.
  • 9.
  • 10.
    ● The LMUKlinikum is one of the largest hospitals in Germany (5,500 staff in 31 departments, >2k beds with >300k cases p.a.). ● Surgical procedures account for 40% of expenses and 60% of revenue. ● The Perioperative Process stretches from beginning to end of anesthesia; existing duration predictions are poor. ● Sqior medical provides IT infrastructure and applications for vital hospital logistics, supporting several of largest hospitals in Germany. ● QAware has a stake in Sqior, and has provided them with process mining software and data analysis. sqior medical Context
  • 11.
    IT Improvement Automation Augmentation Business ProcessManagement ● Focus on people and tasks ● Document planned process ● Process improvement, BPR Process Mining, …Observability, Analytics, Transparency Process Mining ● Focus on Event Logs ● one-time analysis ● historic data Process Observability ● continuous analyse ● Real-time data ● interactive visualisation ● empower process stakeholders Data Science ● Analysis of particular KPIs ● Hypothesis test
  • 12.
    Business Processes Theory… andpractice Complete process transparency requires both angles – and more…
  • 13.
    Anesthesia Patient logistics Surgery Observation1 The perioperative process is structured Ordering Transport Induction Preparation Surgical Procedure Waking Cleaning Ausschleusen
  • 14.
    Observation 2 Process stepsshow up in the data
  • 15.
  • 17.
  • 18.
    Folienvarianten Plan stability Postponed proceduresare racing for patients and staff alike. Not to mention cost.
  • 19.
    Observation 3 Multiple factorsmay influence operation duration
  • 20.
    Folienvarianten Observation 4 Expert knowledgeis crucial Factor Analysis 11 076 2 081 135 procedures with anesthesia distinct descriptions clusters of similar descriptions 15 …including knowledge of clinicians ● MS & Magensonde ● BDK & Blasendauerkatheter ● SpA &Spinalanästhesie ● DIB & Ischiadicus Block & SchiadikusB & IschB ● FemB & FemBlock & Femoralisblock & Fem & Fem.Block ● SahB & Saphenusblock ● Art (wach) & Wach-Art & Wach,Art & Wach-ART & Wacharterie ● Art & Arterie ● Sedierung & Analgosedierung ● Spiral & Spiraltubus ● Video & Videolaryngoskopie & Videolaryngoskop & Video-ITN ● usw. ● “Sonstige" 7 1. Allgemeine Anästhesie (Narkose) 2. Regionalanästhesie 3. Atemwegssicherung und Beatmung 4. Gefäßzugänge & Katheter 5. Notfallmaßnahmen & Reanimation 6. Blutmanagement & Transfusionen 7. Sonstige Maßnahmen …actually useful clusters
  • 21.
    Observation 5 Actual andplanned durations deviate
  • 22.
  • 23.
    Observation 7 Planned durationstoo large PLANNED (pre-registered) ACTUAL (measured) Duration Anesthesia (aggregated) Plan stability Postponed operations are taxing for staff and patients, not to mention cost.
  • 24.
    Observation 8 …surprise PLANNED (pre-registered) ACTUAL (measured) Duration Anesthesia(aggregated) Plan stability Postponed operations are taxing for staff and patients, not to mention cost.
  • 25.
    Insights Process Transparency ismore than Process Mining Img: flaticon.com, Authors: Cap Cool, Freepik, Andrejs Kirma, Eucalyp Extract & enrich event log data Fold logs to nets Extract process chains Understand processes and problems in domain context Holistic Prozessketten extrahieren Analyze & visualize data in detail Statistics & ML Observe, display & steer operational processes Sustainable Tool based Automatic conventional Process Discovery augmentation, not automation ● high variation ● strong context dependency ● resists automation …breaks AI qualitative approach (hypothesis driven) complemets quantitative approach (data driven)
  • 26.
    Take Aways ● ProcessMining is a great stepping stone for data analysis. ● Low hanging fruit are to be had, still. ● AI is not the answer to every question. ● Simpler is better. Always. WTF?! Gaaah! Hang on… I see!
  • 27.
    QAware GmbH |Aschauer Straße 30 | 81549 München | GF: Dr. Josef Adersberger, Michael Stehnken, Michael Rohleder, Mario-Leander Reimer| Niederlassungen in München, Mainz, Rosenheim, Darmstadt | +49 89 232315-0 | info@qaware.de | www.qaware.de Intrigued? Let’s talk Harald Störrle Lead IT Consultant Harald.Stoerrle@QAware.de +49 151 5092 1851 @haraldstoerrle
  • 28.
    Get the lateston tech trends, our meetups, project insights and everything else happening at QAware. Stay updated with the QAware Insights Newsletter! Scan to Subscribe