Guide to Interpreting your Clinical Indicator Trend Report – Q&As

Report
This presentation provides some guidance on how to interpret the tables and graphs
provided in your Clinical Indicator Trend Report.
Most indicators have a desirable outcome that is specified as being either high (H) or
low (L). The remainder are unspecified (N). Indicators are allocated a Not Specified
(N) level where the impact of this measure on clinical outcomes remains unknown.
For example:
 5.1 Pain intensity scores recorded for surgical patients with LOS ≥1 day (H)
 3.5 Unplanned stay in recovery room >2 hours for medical reasons (L)
 3.1 Malignancies diagnosed at colonoscopy (N)
The following symbols are used to flag unusually high or low rate results and to flag
trends in rates.
A rate is considered unusual when the cumulative excess is more than 5 standard
deviations from the expected cumulative excess, that is, from zero.
Desirable level high
H Rates high.
Desirable level low
L
Rates low.
Desirable level low
H
Rates high.
Desirable level high
L
Rates low.
Desirable level
unspecified.
H
L
Rates high
Rates low
Better
performance
 Rates increasing
Poorer
performance
 Rates increasing
Neither
 Rates decreasing
 Rates decreasing
 Rates increasing
 Rates decreasing
Rates
improving
Rates
deteriorating
Neither
H means the desirable rate is high and
this organisation’s rate is significantly
higher (better) than the average rate
H means the desirable rate is low and
this organisation’s rate is significantly
higher (poorer) than the average rate
L means the desirable rate is high and
this organisation’s rate is significantly
lower (poorer) than the average rate
  Means an improving trend
  Means a deteriorating trend
Arrows (, , , , , ) indicate a statistically significant trend
The absence of a trend arrow means that the change over time (as indicated by a line of best fit) does not reach
statistical significance.
Symbol if the
trend is
statistically
significant
Number of
submissions that
were in the best or
poorest 20% of
rates submitted for
this indicator
Total number of
submissions
from 2009 to
2012
Your organisation’s
rate
20th centile rate - 20% of
HCOs lie below this rate
80th centile rate -20% of HCOs lie
above this rate
Average
rate
Excess is the number of
events in excess of
expected. (The number
that would not have
occurred if the HCO had
the average rate).
Cumulative excess is the
excess over successive
periods. Adding the data
reveals underlying
differences that may not
be apparent in a single
data submission.
Data plotted in a graph
Where there were four or more submissions the data are plotted
in a graph. The fitted line is based on a logistic model, so it may
be curved. It is possible that in the earlier years of reporting
when our data cleaning process was not as effective as now, some
aberrant data points may be apparent in the plot.
Guide to Interpreting your Clinical Indicator
Trend Report – Q&As
Refer to Clinical Indicator 1.4 (ED ATS Category 1.4) above
Question. What do the symbols L and  mean? (in combination)
Answer. It means that for this indicator, this organisation’s rate is significantly poorer than
the average but it has been improving over time.
8
The desirable rate for this indicator
is Low.
This organisation’s rate is
significantly poorer than the
average and it is improving.
Q. What do the overall, 20th and
80th values imply?
A. The overall is relatively constant
however there is a large degree of
variation among HCOs highlighted
by the gap between the 20th and
80th centiles
Text from this organisation’s report :
There were 11 data submissions between 2007 and 2012. The rate was in the higher 20% of rates in 5 of the
submissions. There was a deterioration in the fitted rate from 0.41 in 2007 to 4.2 in 2012, an increase of 3.8
per 100 admissions.
Q. Please comment on the trend
A. This organisation’s rate was
initially high and had suddenly and
dramatically deteriorated since
2009.
Consider what has happened in
the last four years to cause a
deterioration in performance. The
figures and data collection
processes should be reviewed to
ascertain the causes.
Q. What do these
figures tell you?
A. This organisation has
scored a perfect 100%
rate for each data
submission. It is
unlikely that this is the
case in reality and data
collection processes
should be reviewed.

similar documents