Improvement of best practice in

1

• Improvement of best practice in
• early breast cancer
• Actionable surgeon and hospital factors
• Marjan Gort
• Manda Broekhuis, Renée Otter, Jacob Wijngaard,
  Niek Klazinga
• Comprehensive Cancer Centre North Netherlands
• Faculty of Management and Organization,
  University of Groningen
• UKACR NCR Conference 2006

2
Study background

• Despite guidelines, large variation between
  hospitals in surgical treatment of breast cancer
  patients (Baldwin, 2004)
• Variation remains while adjusting for patient
  characteristics (Schaapveld et al, 20051
  Siesling et al, 20052)
• Other causes for variation surgeon and hospital?
• 1 Guideline adherence for early breast cancer
  before and after introduction of the sentinel
  node biopsy. Br J Cancer, 2005
• 2 Trends and variation in breast conserving
  surgery in the southeast and east of the
  Netherlands over the period 1990-2002, NTvG, 2005
  

3
Study objectives

• Identify actionable elements for improving best
  practice for breast cancer
• Relative effects of patient, surgeon and hospital
  factors on surgical treatment
• in early breast cancer
• CCCN region
• 2.1 million inhabitants
• 17 hospitals
• 1800 breast cancer patients/year

4
Methods

• Patient selection
• 2929 early staged breast cancer patients (cT1/
  cT2)
• 89 different surgeons
• 16 hospitals (university hospital excluded) CCCN
  region
• Jan 1998 Jan 2002
• Data sources
• Cancer Registry CCCN
• 2nd visitatie
• Medical specialist registry
• Patient, surgeon and hospital variables in
• multilevel logistic regression

5
Methods

• Surgical treatment
• Breast conserving surgery (BCS) ? best practice
  for early stage
• Modified radical mastectomy (MRM)
• Patient variables
• Age, tumour size (cT), mode of detection, period
  of diagnosis
• Surgeon variables
• Volume (lt30/ 30), member of breast cancer
  working group, years in practice, quality of
  nodal staging (Kingsmore, 2005)
• Hospital variables
• Volume (lt100/ 100), teaching status, trial
  oriented,
• management policy, multidisciplinary care (
  e.g. oncology committee, patient discussion)

6
Results
Differences in BCS use in CCCN region
7
Results

• Null model (empty model, no explanatory
  variables)
• Relative effect of 3 levels on BCS use

Random part B (SE)
Hospital-level 0.123 (0.072) 3.4
Surgeon-level 0.194 (0.061) 5.4
Patient-level (?2/3) 3.290 91.2
Total 3.606 100
the lowest level variance is fixed to the
variance of a logistic distribution (p2/3),
therefore, the surgeon and hospital level
variances can only be interpreted as a proportion
of the total variance
8
Results
Relative effects on BCS use (multivariate)
Final Model OR (95 CI)
Patient Size of tumour (cT) cT1 1.0
cT2 0.37 (0.31-0.44)
Detection mode Non-screen detected 1.0
Screen detected 1.41 (1.16-1.71)
Surgeon Volume lt30 1.0
30 1.04 (0.74-1.46)
Breast cancer WG No 1.0
Yes 1.40 (0.99-1.98)
Hospital Volume lt100 1.0
100 1.38 (0.71-2.66)
Multidisciplinary care 1.30 (1.08-1.58)
p lt 0.05
9
Results

• Variation in BCS use (final model)
• Patient Tumour size and detection mode
• Surgeon membership breast cancer working group
• Hospital Multidisciplinary care (MD meeting,
  oncology committee)
• Volume (surgeon and hospital) no strong predictor
• Variance partly explained, but remaining
  variation largely on surgeon level

10
Conclusions

• Surgeon and hospital ? important starting points
  for improving best practice
• Actionable elements
• Patient level population based screening for
  early detection
• Surgeon multidisciplinary working group
• Hospital multidisciplinary care
• (A minimum) volume as pre-requisite for
  organising multidisciplinary care
• Unexplained surgeon variation conclusive
  factors?

11
Discussion

• BCS in early staged breast cancer ? best practice
• BUT Individual treatment depends on e.g.
• risk factors for local recurrence
• expected cosmetic result (breast- tumour ratio)
• co morbidity
• patient preferences
• Importance of link to outcomes
• Medical complications, survival
• Patient related satisfaction, quality of life

Article published in Breast Cancer Research and
Treatment (online available) M. Gort, M.
Broekhuis, R. Otter, N.S.Klazinga, Improvement
of best practice in breast cancer, actionable
surgeon and hospital factors, 2006
12
Further research for actionable indicators

• Other cancers
• Different aspects
• Structure (organisation, volume)
• Process (compliance to guidelines, throughput
  times..)
• Outcomes (medical, patient related)
• Useful data sources CCCN
• Cancer Registration
• Visitatie oncological care
• Patient satisfaction research
• Guidelines
• Pattern of care studies

13

• Comprehensive Cancer Centre North Netherlands
• Faculty of Management and Organization
• m.gort_at_ikn.nl