2016 Mutu Sesi 11 TL Statistics For QI
2016 Mutu Sesi 11 TL Statistics For QI
Peningkatan Mutu
Trisasi Lestari
2016
Seandainya praktek klinis bisa
sesederhana ini
X Y
Pasien Pasien
konsultasi sembuh
dg Dokter dan puas
Banyak faktor yang mempengaruhi
X1 Usia X4 Koordinasi
pelayanan
kesehatan
Y Outcome
pasien (sembuh,
X2 Gender perbaikan
fungsional atau
kepuasan)
X3 Status
kesehatan X5 Komunikasi
Time 1 Time 2 Time 3
Kenyataannya seperti ini
R1 R4
X1 Usia X4 Koordinasi
pelayanan
kesehatan
R2 Y Outcome
pasien (sembuh,
X2 Gender perbaikan
R3 fungsional atau
kepuasan)
X3 Status
kesehatan
R5
X5 Komunikasi
Time 1 Time 2 Time 3
The Quality Gurus
1. Deskriptif Statistik
2. Percentage change
Vital Few
Trivial many
4. Check sheet
Check Sheet
• A means of recording historical data
– on causes of nonconformities
– On nonconforming units
• Source of data for Pareto Chart
• no uniform design
• notes on raw materials, machine
performance, or operator changes.
5. Scatter Plot
• Simplest: bivariate scatter plot
• Vertical axis: dependent variable
• Perhatikan scaling axis
• Ada dua jenis:
– a time sequence plot
• (minute, hours, days, weeks, years)
– A probability plot
• Horizontal Axis: observation
• Vertical axis: probability distribution
A time sequence scatter plot
of patient priority score vs time waited for surgery (n=50)
A probability scatter plot
6. Control Chart
• a time sequence plot with “decision lines”
added:
• to determine whether or not a process is in control
• can indicate when a process is “out of
control.”
• Similar to hipothesis testing (balance between
type 1 and type 2 error)
Basic form of a control chart
7. Cause and Effect Diagram
Cause and Effect Diagram
• Introduced by Prof Ishikawa in 1943
• Ishikawa diagram/fishbone chart
• list as many possible or probable causes as we
can think of without making the diagram too
cluttered
• maintaining the correct relationship between
the causes.
Cause and Effect Diagram
• When a relationship between the quality
characteristic and a cause can be shown
quantitatively using numerical information,
the cause should be enclosed in a box
• When it is known that a relationship between
a cause and an effect does exist, but the
relationship cannot be supported with data,
the cause should be underlined
• No relationship : only be labeled
Negative Ishikawa diagram
8. Defect Concentration Diagram
• a schematic diagram that shows the various
sides of a unit of production
• the positions where nonconformities occur
pinpointed
• Defect location check sheet (Ishikawa)
• information about the potential causes
Control Chart
Control chart representing nosocomial
infections in the ED
Performance improvement Data
Chest Pain in Emergency
Department. Slide courtesy of IHI
Average CABG Mortality
Before and After implementation of a new Protocol
(Slide courtesy of IHI)
A second look at the Data
7%
2%
Angka rata-rata tidak menggambarkan
situasi sesungguhnya
Χ (CL)
Waktu
Bagaimana variasi dalam sebuah
sistem dengan berjalannya waktu?
X (Median)
Waktu
Non-random rules for run chart
Shewhart’s Control Chart
Sigma Limit
UCL
Upper Control Limit
Χ (CL)
LCL
Lower Control Limit
Biasanya diperlukan 15-20 data points
Waktu
Pembagian Zona dalam Control Chart
Zone A +3 SL
UCL
Zone B +2 SL Upper Control Limit
Zone C +1 SL Χ (CL)
Zone C -1 SL
Zone B -2 SL
LCL
Lower Control Limit
Zone A -3 SL
Aturan Control Chart untuk
mengidentifikasi penyebab variasi
Time to surfactant
administration of premature infants
Jenis-jenis control chart
“If I had to reduce
my message for
management to just
a few words, I’d say
it all had to do with
reducing variation”.
(W.Edwards Deming)
you don’t have to be a
statistician to be successful in
quality improvement.
Terima Kasih