Department of Health Informatics
Health Information Management Program
BINF 5520 Health Analytics
Agenda
Understanding the Need for Preoperative Risk Assessment
Applying a “Bedside” Model of Open Heart Risk Assessment
Implementing the “Bedside” Model in a Second Hospital
Open Heart Risk Assessment Today: The Society for Thoracic Surgery (STS) Model
Implications for Health Analytics
Understanding the Need for Preoperative
Risk Assessment and Stratification: The New York Experience
NYS Among First to Implement Cardiac Risk Model
Model Based on Earlier Work in New Jersey
Model Applied to All non-Federal Hospitals in NYS
Model Compared Both Hospitals and Providers
Model Calculates a Risk Adjusted Mortality Rate (RAMR)
Model Equalizes Results Based on a Hypothetical Statewide Case Mix
Health.ny.gov/statistics/diseases/cardiovascular/heart_disease/docs/2011-2013_adult_cardiac_surgery.pdf
Understanding the Need for Preoperative
Risk Assessment and Stratification: The New York Experience
NYS Department of Health Report Summarizes:
Creation of RAMR Model
Data Collection Methods
Case Mix Assumptions
Description of Patient Population
Discussion of Critical Metrics
Impact on Quality Improvement
Health.ny.gov/statistics/diseases/cardiovascular/heart_disease/docs/2011-2013_adult_cardiac_surgery.pdf
Understanding the Need for Preoperative
Risk Assessment and Stratification: The New York Experience
Table 1 compares both Observed and Risk-Adjusted Mortality Rates for Isolated CABG Surgery in NYS for 2013 discharges.
RAMR=Risk Adjusted Mortality Rate: the Provider’s Mortality Rate if the Provider’s case mix was identical to a hypothetical statewide case mix.
Health.ny.gov/statistics/diseases/cardiovascular/heart_disease/docs/2011-2013_adult_cardiac_surgery.pdf
Understanding the Need for Preoperative
Risk Assessment and Stratification: The New York Experience
Table 6 presents the data by both Hospital and Provider.
Care was taken to collapse data when insufficient individual performance metrics were available.
This report was publically available via the NYS Department of Health website, and it can be found at the link below.
How did Cardiac Surgeons begin considering these issues?
These efforts actually started in the mid-1980s at a hospital in New Jersey.
Health.ny.gov/statistics/diseases/cardiovascular/heart_disease/docs/2011-2013_adult_cardiac_surgery.pdf
Developing and Implementing a “Bedside Estimation of Risk”
Model of Open Heart Risk Stratification
This work, which was begun in the mid-1980s, discussed the need for the development of a clinical model which helps surgeons when discussing Open Heart Risk with patients.
The authors conclusively demonstrate the need for a “bedside scoring system” which facilitates provider-patient dialogue.
Many of the subsequent risk models were, in some part, based on this work.
Implementing the “Bedside” Model in a Second Hospital
The Canadian authors implement the model ...
Presiding Officer Training module 2024 lok sabha elections
Department of Health InformaticsHealth Information Managemen
1. Department of Health Informatics
Health Information Management Program
BINF 5520 Health Analytics
Agenda
Understanding the Need for Preoperative Risk Assessment
Applying a “Bedside” Model of Open Heart Risk Assessment
Implementing the “Bedside” Model in a Second Hospital
Open Heart Risk Assessment Today: The Society for Thoracic
Surgery (STS) Model
Implications for Health Analytics
Understanding the Need for Preoperative
Risk Assessment and Stratification: The New York Experience
NYS Among First to Implement Cardiac Risk Model
Model Based on Earlier Work in New Jersey
Model Applied to All non-Federal Hospitals in NYS
Model Compared Both Hospitals and Providers
2. Model Calculates a Risk Adjusted Mortality Rate (RAMR)
Model Equalizes Results Based on a Hypothetical Statewide
Case Mix
Health.ny.gov/statistics/diseases/cardiovascular/heart_disease/d
ocs/2011-2013_adult_cardiac_surgery.pdf
Understanding the Need for Preoperative
Risk Assessment and Stratification: The New York Experience
NYS Department of Health Report Summarizes:
Creation of RAMR Model
Data Collection Methods
Case Mix Assumptions
Description of Patient Population
Discussion of Critical Metrics
Impact on Quality Improvement
Health.ny.gov/statistics/diseases/cardiovascular/heart_disease/d
ocs/2011-2013_adult_cardiac_surgery.pdf
Understanding the Need for Preoperative
Risk Assessment and Stratification: The New York Experience
Table 1 compares both Observed and Risk-Adjusted Mortality
Rates for Isolated CABG Surgery in NYS for 2013 discharges.
RAMR=Risk Adjusted Mortality Rate: the Provider’s Mortality
Rate if the Provider’s case mix was identical to a hypothetical
statewide case mix.
Health.ny.gov/statistics/diseases/cardiovascular/heart_disease/d
3. ocs/2011-2013_adult_cardiac_surgery.pdf
Understanding the Need for Preoperative
Risk Assessment and Stratification: The New York Experience
Table 6 presents the data by both Hospital and Provider.
Care was taken to collapse data when insufficient individual
performance metrics were available.
This report was publically available via the NYS Department of
Health website, and it can be found at the link below.
How did Cardiac Surgeons begin considering these issues?
These efforts actually started in the mid-1980s at a hospital in
New Jersey.
Health.ny.gov/statistics/diseases/cardiovascular/heart_disease/d
ocs/2011-2013_adult_cardiac_surgery.pdf
Developing and Implementing a “Bedside Estimation of Risk”
Model of Open Heart Risk Stratification
This work, which was begun in the mid-1980s, discussed the
need for the development of a clinical model which helps
surgeons when discussing Open Heart Risk with patients.
The authors conclusively demonstrate the need for a “bedside
scoring system” which facilitates provider-patient dialogue.
Many of the subsequent risk models were, in some part, based
on this work.
4. Implementing the “Bedside” Model in a Second Hospital
The Canadian authors implement the model developed earlier in
NJ to a population of cardiac patients at greater risk.
They discuss both the need for the development of a model as
well as the factors that governed their selection of the
appropriate risk model for their institution.
Ncbi.nlm.nih.gov/pmc/articles/PMC2651930
Implementing the “Bedside” Model in a Second Hospital
The Canadian authors describe the results that they obtained
when applying the model developed in NJ to their patient
cohort.
The results at the Canadian hospital compared favorably when
applying the model developed in NJ.
They also discuss the need for adjusting baseline data based on
clinical risk factors.
Ncbi.nlm.nih.gov/pmc/articles/PMC2651930
Where Are We Today With Regard To
Open Heart Risk Assessment ?
Many efforts worldwide have been implemented in pursuit of a
better understanding of Preoperative Risk Assessment
In the United States, there has been a greater interest in this
subject for clinical, administrative, and financial reasons.
Preoperative Risk Assessment has been accepted as normal part
of administering a surgical program.
5. Preoperative Risk Assessment has been accepted by many non-
cardiac disciplines, including orthopedic surgery, neurosurgical,
and obstetrics.
There are many competing models of risk assessment within the
cardiac surgery area. One that has met with widespread
acceptance was developed by the Society for Thoracic Surgery
(STS).
Components of the STS model are available on the web at this
link:
riskcalc.sts.org/stswebriskcalc/#/
Open Heart Risk Assessment Today:
The Society for Thoracic Surgery (STS) Model
In the intervening years, many groups associated with Open
Heart surgery have developed various models of Risk
Assessment.
One of the most widely used models was developed by the
Society for Thoracic Surgery (STS) and was based somewhat on
earlier models, including the NJ and NY efforts.
The STS Model, which is accessible at the link below, enables
the provider to input key risk factors and then observe the
predicted risk mortality for that individual patient based on data
used in a national model.
riskcalc.sts.org/stswebriskcalc/#/
Open Heart Risk Assessment Today:
The Society for Thoracic Surgery (STS) Model
6. In the intervening years, many groups associated with Open
Heart surgery have developed various models of Risk
Assessment.
This model is easy to use and for members of the STS, it comes
with comprehensive user documentation and training.
riskcalc.sts.org/stswebriskcalc/#/
Open Heart Risk Assessment Today:
The Society for Thoracic Surgery (STS) Model
riskcalc.sts.org/stswebriskcalc/#/
The model requires the user to enter a number of different
factors which are then compared to the STS National Database.
Over the years, the STS has expanded their model to incorporate
both isolated CABG surgery as well as a wide range of the
cardiac surgical procedures.
Open Heart Risk Assessment Today:
The Society for Thoracic Surgery (STS) Model
In addition to providing the surgeon with risk mortality data
based on procedure, the model predicts other indicators which
are helpful to the surgeon when discussing the prognosis with
the patient.
Some aspects of these risk assessment models have now been
incorporated into the Quality Resource Management discipline
within many healthcare enterprises.
Open Heart Risk Assessment Today:
7. The Society for Thoracic Surgery (STS) Model
riskcalc.sts.org/stswebriskcalc/#/
An important aspect to the development of any risk assessment
model is whether or not it can change over time to become
better at predicting surgical outcome.
All models must have the capacity to incorporate new
technology. In the case of Cardiac Surgery, the development of
a greater number of minimally invasive procedures has
necessitate these changes.
Open Heart Risk Assessment Today:
The Society for Thoracic Surgery (STS) Model
riskcalc.sts.org/stswebriskcalc/#/
Once the model have been concluded for a given patient, the
model permits the user to begin another patient.
Some Practical Analytical Implications
for the Health Informaticist
What are the data collection challenges that face the Health
Informaticist?
How do we establish a “baseline” which will be used for
comparative purposes?
What metrics are used to determine “Success”, “Failure”, or
other designations when considering the assessment or surgical
risk?
What are the challenges associated with transporting and/or
applying a model between multiple institutions?
8. Assessing the effectiveness of fire prevention strategies
Mark Taylora, Deb Appletonb, Guy Keenc and John Fieldingd
aDepartment of Computer Science, Liverpool John Moores
University, Liverpool UK; bStrategy and Performance,
Merseyside Fire and Rescue
Service, Liverpool UK; cCommunity Risk Management,
Merseyside Fire and Rescue Service, Liverpool UK; dBusiness
Intelligence Manager,
Merseyside Fire and Rescue Service, Liverpool UK
ABSTRACT
The authors examined the process of assessing the effectiveness
of fire prevention within a Fire
and Rescue Service in north west England, demonstrating how
this was applied in practice. The
approach to fire prevention strategy assessment included overall
fire and rescue service
performance, performance in relation to different population
segments, cost per head
performance, and multi-agency collaboration to support an
effectiveness-based rather than
outcomes-based performance assessment.
IMPACT
This case study examination of the effectiveness of fire
prevention strategies shows that fire
prevention activities have a clear impact on the reduction of the
number of fires. Major
lessons for other fire and rescue services include that referrals
to partner agencies were
9. appropriate to assist in reducing factors associated with fire
risk. In addition, concentrating
fire prevention on the elderly was important given the
proportion of fire fatalities and
injuries involving this group. Finally, analysis of the
effectiveness of fire prevention activities
can inform fire prevention strategies.
KEYWORDS
Effectiveness; fire; fire and
rescue services; prevention;
strategy
Introduction
McKevitt (2015) said that defining and measuring effec-
tiveness at the organizational level is complex. Typi -
cally, public sector organizations are assessed by
output measures, rather than by effectiveness
measures (Ammons & Roenigk, 2015). In terms of
public management theory, there is a need for public
sector organizations to move towards performance
measures that assess effectiveness rather than just
outputs—since, ultimately, it is important that public
sector organizations are actually effective in what
they do. English fire and rescue services are mainly
assessed through outcome measures (FRNF, 2018). UK
fire and rescue services are currently expanding their
scope of operations through the ‘Safe and Well initiat-
ive’ (Safe and Well, 2017), which is an evolution of the
Home Fire Safety Check initiative (HFSC, 2017). The
initiative involves health and social care prevention
initiatives, as well as just fire prevention. This has the
potential to improve fire prevention effectiveness by
not only advising potentially vulnerable individuals of
how to avoid house fires, but also overall to reduce
10. the impact of factors such as alcohol consumption
and smoking that have been linked with house fire inci-
dence (Higgins, Taylor, Jones, & Lisboa, 2013). An
important aspect of fire prevention is fitting and check-
ing smoke alarms. Operational smoke alarms prevent
or reduce many fires and the associated injuries or
fatalities (Tannous et al., 2017). Costs associated with
home fire safety checks can include internal costs
such as staff salaries, and on-costs, marketing costs,
and travel costs, and external costs such as home fire
safety material (batteries, smoke alarms, information
booklets, calling cards, and fire blankets) (Tannous
et al., 2107).
Puolokainen (2017) stated that the desired fire and
rescue service outcomes that include the emergencies
prevented or suppressed and, ultimately, the human
life and property preserved are not easy to measure.
Evaluations of the UK Home Fire Safety Check initiative
in Wales identified a link between the installation of
smoke alarms and reductions in dwelling fires and
non-fatal casualties. The benefits of the Home Fire
Safety Check initiative were found to far outweigh its
costs (Welsh Government Social Research, 2013).
Tannous et al. (2017), in a study in New South Wales,
Australia, commented that analysis of the cost of fire
within the region, combined with measurements of
the success of similar home fire safety check pro-
grammes internationally, demonstrated the home fire
safety check programme’s cost-effectiveness in New
South Wales. Weinholt and Andersson Granberg
(2015) commented that a common tool for evaluating
policy initiatives in the public sector is cost-benefit
analysis (CBA), where all the benefits and costs are
monetized and compared to each other, and that this
12. Literature review
Fire prevention approaches
UK fire and rescue services generally adopt an inte-
grated risk management planning approach that
involves the development of operational strategies to
reducing risk within the community. This is typically
achieved by combining prevention, protection and
emergency response, on a risk-assessed basis in order
to improve the safety of the community (MFRSIRM,
2018). Jennings (2013) concluded that research into
fire in the community residential context has remained
fragmented and isolated by discipline. Various fire pre-
vention initiatives have been introduced in different
communities around the world in an attempt to
reduce fire incidence, morbidity and mortality. These
initiatives have included education and training of chil -
dren, parents, and certain high-risk populations such as
the elderly (Jonsson et al., 2013; Wuschke, Clare, &
Garis, 2013) in community-based programmes (Lehna
et al., 2015; Warda et al., 1999). Andrews, Ashworth,
and Meier (2014) commented that on the community
service side, firefighters seek to build commitment to
fire prevention. This is an area with fewer rules and
much more discretion on how to approach the task,
it is also an area in which the relationship between
the service provider and client may hold the key to
success. However, the research conducted by
Andrews et al. (2014) focused more on equality and
diversity in fire and rescue services rather than actual
performance management.
Typically, the fitting or checking of smoke alarms is a
major fire prevention initiative (Tannous et al., 2017)
undertaken by fire and rescue services. Fires detected
13. by smoke alarms tend to be discovered more rapidly
and are associated with a reduced risk of death and
property damage (Saramago et al., 2014). However,
Jennings (2013) warned that when developing fire pre-
vention strategies based upon risk levels, the dynamic
of fire risk may vary depending on the level of analysis.
Care needs to be taken to avoid ecological fallacy
(failure in reasoning that arises when an inference is
made about an individual based on aggregate data
for a group) when moving from higher-level statistical
data to making conclusions about household or indi-
vidual behaviour or risk.
The UK fire and rescue services’ safe and well initiat-
ive involves not only fire prevention, but also health
and social care prevention initiatives involving referrals
to other agencies such as smoking cessation, and
alcohol management. In terms of assessing the useful-
ness of referrals to other agencies (such as the UK NHS
or local council) via a home fire safety check or safe and
well visit, the quality-adjusted life year (QALY) measure
can be beneficial (Prieto & Sacristan, 2003). The QALY is
a measure of the state of health of a person or group in
which the benefits, in terms of length of life, are
adjusted to reflect the quality of life. One QALY is
equal to one year of life in perfect health. QALYs are cal-
culated by estimating the years of life remaining for a
patient following a particular treatment or intervention,
and weighting each year with a quality-of-life score (on
a 0 to 1 scale). It is often measured in terms of the
person’s ability to carry out the activities of daily life,
and freedom from pain and mental disturbance. This
can be used to assess generic (and banded) levels of
the person’s ability to carry out the activities of daily
life (for example with regard to dementia patients).
14. An important aspect of fire prevention approaches,
such as the home fire safety check or safe and well
visit, is targeting of resources towards those most at
risk of fire. This includes vulnerable groups, such as
the elderly, those with disabilities, and those with life-
styles and behaviours (for example binge drinking
and smoking) that can increase the risk of fire
(Chainey, 2013; Corcoran, Higgs, & Anderson, 2013;
Higgins et al., 2013; Taylor, Higgins, Lisboa, Jarman, &
Hussain, 2016; Turner et al., 2017).
Fire prevention effectiveness
Devlin and Parkin (2004) stated that evidence on cost-
effectiveness is used in many countries to inform
decisions about the allocation of public funds to
public services and products. Saramago et al. (2014)
commented that assessing cost-effectiveness, as well
as effectiveness, is important in a public sector system
operating under a fixed budget restraint. Abdel-
Maksoud, Elbanna, Mahama, and Pollanen (2015)
stated that there can be issues regarding whether the
development of performance measures is properly
aligned with the objectives of public organizations.
Kloot (2009) commented that fire and rescue service
PUBLIC MONEY & MANAGEMENT 419
performance assessments need to examine both oper-
ational and financial performance outcomes.
Hastie and Searle (2016) commented that there are
striking inequalities in the way in which dwelling fire
incidence is distributed through society. Jennings
15. (2013) found that fires can be comparable in cost to
crimes and therefore warrant the significant attention
of the public and governments at all levels. Warda,
Tenenbein, and Moffact (1999) concluded that the
effectiveness of fire prevention approaches in redu-
cing fire incidence, injuries and deaths needs further
study. Educational programmes, for example, are
widespread, but their relative effectiveness has not
been properly established. Clark and Smith (2015)
commented that further research is required into
how fire risk is perceived and mitigated against by
individuals, in particular with regard to the impor-
tance of clear communication in fire prevention strat-
egies, including communicating the causes of fire.
Studies of fire prevention interventions have reported
variable success in reducing fire injury. Typically, this is
measured in terms of the rates of fire incidence, injury
and fatality before and after the fire prevention inter -
vention. Typically, such measurements do not take
into account socio-economic groupings, and do not
include the use of control groups. Jaaskelainen et al.
(2012) commented that, in general, measurement of
service operations has proven to be a significant
challenge.
McNamara (2017) discussed the difficulties associ-
ated with determining fire prevention strategy priori-
ties. Previous studies of fire prevention effectiveness
have not really analysed whether targeting high-risk
age groups achieves a greater reduction in injuries or
produces a larger increase in the implementation of
fire safety practices than alternative age prioritizati ons.
Typically, studies of individual fire prevention pro-
grammes may be unreliable due to inconsistent
design or performance measures, lack of randomness
or controls, and inadequate sample size or follow -up
16. time. Murphy and Greenhalgh (2013) commented
upon the changing nature of performance manage-
ment regimes adopted by the UK government for fire
and rescue services. Murphy and Greenhalgh (2013)
found that a particular focus of performance review
for English fire and rescue services was the integrated
risk management plan, however, they also advocated
that further assessment of performance management
regarding efficiency and effectiveness was required.
This is an important area of research, given that UK
fire and rescue services need to meet the challenges
of continuously improving their services. Carvalho,
Fernandes, Lambert, and Lapsley (2006) and Adcroft
and Willis (2005) stated that the delivery of public
services can be undermined by the time, effort and
management expertise deployed to cope with ever
more elaborate forms of performance measures.
A study by the Welsh Government concluded that
targeting was essential for successful home fire safety
check initiatives. To reduce the number of dwelling
fires, home fire safety checks need to be targeted at
those most at risk, such as vulnerable populations
(Welsh Government Social Research, 2013). Clare,
Garis, Plecas, and Jennings (2012) commented that
public education, and targeted home visits by fire
and rescue services, have produced promising results
via a range of outcome measures, from reduction in
rates of fires and fire-related casualty, through to
increased presence of working smoke alarms when
residences were audited. Clare et al. (2012) used pre-
and post-home visit comparisons via percentages and
ANOVA to analyse for significant difference in fire inci -
dence rates following home visits in British Columbia,
Canada, by fire and rescue service staff.
17. Overall, previous research (Jaaskelainen et al., 2012;
Murphy & Greenhalgh, 2013) has indicated that it can
be difficult to conclusively deem any given fire preven-
tion strategy to be effective, given the general absence
of appropriate randomized controls across studies,
without which positive results have to be interpreted
with caution. The contribution of the research reported
in this paper is the development of an approach to
assess the effectiveness of fire prevention strategies,
in order to inform future fire prevention approaches.
Research method
Merseyside fire and rescue service staff and staff from a
university in the north west region of the UK were
involved in the assessment of fire prevention strategy
effectiveness over a six-month period. Fire and rescue
service staff involved in the assessment of fire preven-
tion strategy effectiveness included the community risk
manager, the director of strategy and performance, the
corporate information and systems manager, and the
business intelligence manager.
The business intelligence manager performed the
primary data collection. University staff and the
business intelligence analyst undertook the analysis
of the data following discussions regarding the analysis
approach to be adopted involving the community risk
manager, the director of strategy and performance, the
corporate information and systems manager, and the
business intelligence manager within the fire and
rescue service studied. The research was internally
driven by the fire and rescue service’s director of strat-
egy and performance, and the community risk
manager. Collection of data was mainly an internal
exercise for the organization managed by the corpor-
18. ate information and systems manager, and undertaken
by the business intelligence manager, apart from the
population data that was gathered by the university
staff from the UK Office for National Statistics (ONS).
In order to address issues of ethics, subjective bias
420 M. TAYLOR ET AL.
and objectivity to preserve the validity and sanctity of
the data gathered, all the work undertaken conformed
to the codes of ethics and professional conduct in oper-
ation at Merseyside fire and rescue service and the uni -
versity involved in the research.
Fire prevention within the area covered by the fire
and rescue service was mostly conducted through the
home fire safety check initiative (HFSC, 2017). Referrals
to other agencies carried out during home fire safety
checks (and later expanded) during safe and well (Safe
and Well Visit Principles, 2017) visits included: smoking
cessation referrals, alcohol reduction referrals, bowel
cancer referrals, and falls risk assessment referrals.
Barton and Valero-Silva (2013) had commented upon
the positive impact on service users of successful
multi-agency working.
The research questions posed by the work reported
in this paper were:
. How can the effectiveness of overall fire and rescue
service provision be assessed?
. How can the effectiveness of a given fire prevention
strategy be assessed?
19. . How can the effectiveness and of the targeting of
fire prevention be assessed?
The research questions arose from the need of the
fire and rescue service studied to assess effectiveness
of operations. Previously English fire and rescue ser-
vices were mainly assessed through outcome measures
(FRNF, 2018); however, a new inspection approach by
HM Inspectorate of Constabularies, Fire and Rescue Ser-
vices (HMICFRS, 2018) for the 45 fire and rescue ser -
vices in England places greater emphasis on
measures of efficiency and effectiveness. During the
period of the study, Merseyside fire and rescue
service had not had an inspection under the new
inspectorate. In addition, given the reductions in the
rates of fire in England during the last 10 years
(during a period of reducing budgets), fire prevention,
and in particular the targeting of fire prevention are
increasingly important aspects of fire and rescue
service operations.
Our research questions are important since the con-
tinuing budget reductions experienced by fire and
rescue services in the UK necessitates a deeper under-
standing of the effectiveness of current fire prevention
activities, in order for future fire prevention activities to
be as effective and efficient as possible given the
decreasing level of funding available. Funding for Mer -
seyside fire and rescue service decreased to £59.490
million for 2017/2018, from £73.576 million for 2004/
2005. Merseyside has had the largest budget
reductions of any English fire and rescue service
(MFRS, 2018). The main objective of the research was
to develop an approach that could assess effectiveness
of fire and rescue service operations at different levels:
20. at an overall operational level, and at the level of oper -
ational fire prevention.
Study of the outcomes from an intervention might
typically use random samples from intervention and
control groups. However, in the case of fire prevention,
due to duty-of-care requirements, it would not be
appropriate, having identified vulnerable groups, to
apply such fire prevention in some areas within the
region and not others, in order to assess the relative
impact on the number of fire incidents, injuries and
fatalities. In addition, in the case of fire prevention, typi -
cally home fire safety checks and safe and well visits
would be conducted across the areas in a region over
a relatively long period of time (due to the number of
visits involved, and the limited numbers of fire safety
officers). For these reasons, analysis of effectiveness
was more appropriate over a longer time period, than
shorter term analyses. The research approach
adopted was appropriate given the time periods over
which fire prevention activities are conducted, and
the time taken for the impact of fire prevention
approaches to be measurable in terms of the frequency
of production of fire statistics. Data for the analyses
undertaken in order to answer the research questions
posed was obtained from Merseyside fire and rescue
service and the ONS.
The conceptual contribution of the research is the
development of an approach to assessing the effective-
ness of fire prevention strategies, based upon analysis of
the approach to targeting fire prevention, and multi-
agency collaboration. A limitation of the research pre-
sented here is that the results are only relevant to one
fire and rescue service serving a relatively small popu-
21. lation, and therefore the applicability and generalizabil-
ity to a wider population may be limited. However, due
to the variable nature of the populations served by
different fire and rescue services and the different area
geographies, and the different funding levels, meaning-
ful comparisons of effectiveness between different fire
and rescue services would be difficult to interpret.
Research results
Assessing the effectiveness of overall fire and
rescue service provision
The first stage in the examination of fire prevention
strategy effectiveness was to determine what quantifi-
able measures could potentially be used to assess the
effectiveness of fire and rescue service provision. Dis -
cussions with the relevant staff within Merseyside fire
and rescue service identified the following potential
measures:
. Overall number of fire incidences, injuries and fatal -
ities per year within the area covered by the fire and
rescue service.
PUBLIC MONEY & MANAGEMENT 421
. Number of fire incidences, injuries and fatalities per
year within the different population segments within
the area covered by the fire and rescue service.
. Proportion of individuals within each different popu-
lation segment per 100,000 of population within the
area covered by the fire and rescue service involved
22. in a fire incidence, injury and fatality.
These measures could be used to assess overall fire
and rescue service performance, performance in
relation to different population segments, and per-
formance in relation to vulnerable population seg-
ments. These measures address the research question
about how to assess the effectiveness of overall fire
and rescue service provision. Table 1 shows the
overall number of fire incidences, injuries and fatalities
in the period 2006 to 2016 within the area covered by
Merseyside fire and rescue service.
In terms of overall fire and rescue service effective-
ness indicated by how many fire incidences, injuries
and fatalities occurred over the 10-year period 2006
to 2016, there was a clear downward trend in the
number of fire incidences. In the other 44 fire and
rescue services in England, the number of fire inci-
dents had also been on a general downward trend
since 2004 (HO, 2017). This data also answers our
research question about how to assess the effective-
ness of overall fire and rescue service provision
because, by comparing fire-incidence rates over time,
a significant reduction in fire incidences was identified.
In terms of the pattern of fire injuries and fatalities, due
to their relatively small numbers, a few fires could sig-
nificantly affect the figures, which makes interpreting
the data more difficult. Table 2 shows the overall
number of fire incidences, injuries and fatalities by
population segments within the Merseyside area
from 2006 to 2016. Merseyside fire and rescue
service had identified the elderly (those aged 65+) as
being the most vulnerable population segment in
terms of fire risk (as had other previous research: for
23. Table 2. Overall number of fire incidences, injuries and
fatalities by population segments in the Merseyside area: 2006
to 2016.
(Data regarding dwelling occupants involved in a fire incidence
is typically only recorded for fire injuries and fatalities.)
2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016
Fire incidences
(all ages)
1384 1318 1315 1337 1248 1183 1090 1221 1023 1084 1049
Fire injuries
(age 65+)
35 31 35 24 25 43 36 52 40 35 25
Fire injuries
(age < 65)
94 68 70 91 101 102 71 93 77 73 75
Fire fatalities
(age 65+)
4 5 5 3 3 4 1 4 6 10 6
Fire fatalities
(age < 65)
2 6 5 3 4 1 2 6 3 5 4
Figure 1. Fire instances in the Merseyside area: 2006 to 2016.
24. Table 1. Overall number of fire incidences, injuries and
fatalities in Merseyside: 2006 to 2016.
2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016
Fire incidences 1384 1318 1315 1337 1248 1183 1090 1221
1023 1084 1049
Fire injuries 129 100 106 115 126 145 110 146 118 110 104
Fire fatalities 6 11 10 6 7 5 3 10 9 15 10
422 M. TAYLOR ET AL.
example Mulvaney et al., 2009; and Corcoran, Higgs,
Rohde, & Chhetri, 2011). In the period 2006 to 2016
for the Merseyside area, 56% of fire fatality victims
were over the age of 65.
This data also answers our research question about
how to assess the effectiveness of overall fire and
rescue service provision because, by examini ng fire fatal-
ity and injury rates in greater detail via different ident-
ified population segments, it was possible to examine
the differences in frequency between different popu-
lation segments over time. Table 3 shows the proportion
of individuals within each different population segment
involved in fire incidences, injuries and fatalities by
population segment per 100,000 of population in the
Merseyside area in the period 2006 to 2016.
Data regarding dwelling occupants involved in a fire
is typically only recorded for fire injuries and fatalities.
This data also answers our research question about
how to assess the effectiveness of overall fire and
rescue service provision by examining the proportions
of different identified population groups involved in
25. fire incidences. This allows an examination of the rela-
tive rates of fire injuries and fatalities between different
population segments over time. In particular, the pro-
portion of elderly individuals involved in fire injuries
and fatalities was significantly higher than the pro-
portion of the non-elderly population.
For the purposes of calculating population pro-
portions, the population estimate figures for Mersey-
side were as shown in table 4 (Nomis, 2017).
The population estimate figures for Merseyside resi-
dents aged 65+ were as shown in table 5 (Nomis, 2017).
There appeared to be a clear downward trend in the
overall proportion of the Merseyside population
involved in a fire incidence over the 10-year period
studied. The small proportions of those involved in a
fire injury and fatality meant that these figures were
less clear to interpret. However, there appeared to be
a clear indication that those aged 65+ were proportion-
ately more likely to be injured or die in a dwelling fire
than those aged less than 65.
Assessing fire incidences, injuries and fatalities
compared to spend per head
The next effectiveness measures assessed the number
of fire incidences, injuries and fatalities compared to
the relevant spend per head:
. Fire incidences, injuries and fatalities versus overall
spend per head of overall population in the area.
. Fire incidences, injuries and fatalities versus overall
fire prevention spend per head of overall population
26. in the area.
. Fire injuries and fatalities versus overall fire preven-
tion spend per head of different population seg-
ments in the area.
Table 6 shows the overall fire and rescue service
spend per head (in £s) and the overall number of fire
incidences, injuries and fatalities in the Merseyside
area in the period 2006 to 2016.
Table 7 shows the ratio of overall fire and rescue
service spend (or cost) (in £000s) to fire incidents/
Table 5. Population figures for Merseyside residents aged 65 +
2006 to 2016 (000s).
2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016
230 231 232 234 236 239 247 251 256 259 263
Table 3. Proportion of individuals within each different
population segment involved in fire incidences, injuries and
fatalities in the
Merseyside area: 2006 to 2016.
2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016
Fire incidences
proportion
(all ages)
101.1 96.4 96.1 97.5 90.7 85.7 78.7 88.1 73.5 77.5 74.6
Fire injuries
proportion
(age 65+)
27. 15.2 13.4 15.1 10.3 10.6 18.0 14.6 20.7 15.6 13.5 9.5
Fire injuries
proportion
(age < 65)
8.3 6.0 6.2 8.0 8.9 8.9 6.2 8.2 6.8 6.4 6.6
Fire fatalities
proportion
(age 65+)
1.7 2.2 2.2 1.3 1.3 1.7 0.4 1.6 2.3 3.9 2.3
Fire fatalities
proportion
(age < 65)
0.2 0.5 0.4 0.3 0.4 0.1 0.2 0.5 0.3 0.4 0.3
Table 4. Population figures for Merseyside 2006 to 2016 (000s).
2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016
1,368 1,367 1,369 1,372 1,376 1,381 1,386 1,387 1,391 1,398
1,406
PUBLIC MONEY & MANAGEMENT 423
injuries/fatalities. This data also answers the research
question: How can the effectiveness of overall fire
and rescue service provision be assessed, by examining
the spending associated with fire incidents, injuries and
fatalities over time, in terms of the cost per fire incident,
28. injury and fatality.
Assessing the effectiveness of fire prevention
strategies
Table 8 shows the overall fire prevention spend per head
(in £s) and the number of fire incidences, fire incidents
involving injuries, and fire incidents involving fatalities
in the Merseyside area in the period 2006 to 2016.
This appeared to indicate that overall the spend per
head on fire prevention has been effective in reducing
the number of fire instances over the 10-year period
studied. However, the impact upon fire injuries and
fatalities was more difficult to interpret. Table 9
shows the ratio of overall fire prevention spend (or
cost) (in £000s) to fire incidents/fire incidents involving
injuries/fire incidents involving fatalities.
This data answers our research question about how
to assess the effectiveness of a given fire prevention
strategy by examining the spending associated with
fire prevention compared with the number of fire inci-
dents, injuries and fatalities. However, due to fluctu-
ations in the spend per head of population, and in
the overall spend on fire prevention per year over the
period studied, it was difficult to discern an overall
trend or pattern. For example, whether spending
more on fire prevention was associated with fewer
fire injuries and fatalities, or spending less on fire pre-
vention was associated with fewer fire injuries and
fatalities year on year.
Assessing the effectiveness of the targeting of
fire prevention
29. Table 10 shows the fire prevention spend per head (in
£s) and the number of fire injuries and fatalities by
population segment in the Merseyside area in the
period 2006 to 2016. (Data regarding dwelling occu-
pants involved in a fire incidence is typically only
recorded for fire injuries and fatalities.)
This data answers our research question about how
to assess the effectiveness of the targeting of fire pre-
vention by examining the spending associated with
fire prevention for different population groups com-
pared with the number of fire incidents, injuries and
fatalities for those groups. Overall, fire prevention
activities appear to have had a clear impact on the
number of fires, which steadily reduced during the
period 2006 to 2016, even though the spend per
head of the population on fire prevention fluctuated
between £13.50 and £16.99 during this period.
The impact of fire prevention on fire injuries and fire
fatalities over the period studied was less clear. The
number of fire incidents involving injuries fluctuated
between 77 and 114 with a mean value of 92.6, and
Table 6. Overall fire and rescue service spend per head (in £s)
and overall number of fire incidences, injuries and fatalities in
the
Merseyside area, 2006 to 2016.
2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016
Spend per
head of
population
30. 48.72 50.23 51.80 51.07 52.24 51.14 52.63 53.64 52.07 43.37
46.73
Fire incidences 1384 1318 1315 1337 1248 1183 1090 1221
1023 1084 1049
Fire injuries 129 100 106 115 126 145 110 146 118 110 104
Fire fatalities 6 11 10 6 7 5 3 10 9 15 10
Table 8. Overall fire prevention spend per head (in £s) and
overall number of fire incidences, fire incidents involving
injuries, and
fire incidents involving fatalities in the Merseyside area, 2006
to 2016.
2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016
Fire prevention
spend per
head of
population
16.07 15.11 13.72 14.26 13.50 14.73 14.00 15.62 16.99 13.86
16.11
Fire incidences 1384 1318 1315 1337 1248 1183 1090 1221
1023 1084 1049
Fire incidents involving
injuries
109 84 85 90 84 112 86 114 91 86 77
Fire incidents involving
fatalities
6 11 10 6 7 5 3 9 9 14 7
31. Table 7. Ratio of overall fire and rescue service spend (or cost)
(in £000s) to fire incidents/injuries/fatalities in the Merseyside
area,
2006 to 2016.
2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016
Cost/fire incidence ratio 48 52 54 52 58 60 67 61 71 56 63
Cost/fire injuries ratio 517 687 669 609 570 487 663 510 614
551 632
Cost/fire fatalities ratio 11108 6242 7091 11678 10269 14125
24315 7440 8048 4042 6570
424 M. TAYLOR ET AL.
the number of fire incidents involving fatalities fluctu-
ated between 3 and 14 with a mean value of 7.9.
There did not appear to be an overall trend with
either the numbers of fire incidents involving injuries,
or the numbers of fire incidents involving fatalities.
Nor did there appear to be a link between the spend
per head of the population on fire prevention in a
given year and the numbers of fire incidents involving
injuries, or fire incidents involving fatalities in that year
(or the following year).
However, analysis of the factors associated with fire
fatalities over the period 2006 to 2016 in the Mersey-
side area revealed that:
. 50.5% of fire fatalities involved smoker’s materials,
indicating that the referrals to NHS smoking cessa-
tion services that are an integral part of the home
fire safety checks carried out by Merseyside fire
32. and rescue service were an appropriate and necess-
ary aspect of current and future fire prevention.
. 39.8% of fire fatalities involved alcohol consump-
tion, indicating that the referrals to NHS alcohol
reduction services that are an integral part of the
home fire safety checks carried out by Merseyside
fire and rescue service were an appropriate and
necessary aspect of current and future fire
prevention.
. 55.9% of fire fatalities and 21.5% of fire injuries
involved those aged over 65, indicating that this
age group was the most in need of fire prevention
support, which is the current main focus of fire pre-
vention activities carried out by Merseyside Fire and
Rescue Services.
. 64.5% of the fire fatalities involved those living
alone, indicating that this social group was also
the most in need of fire prevention support.
This data also answers the research question about
how to assess the effectiveness of the targeting of
fire prevention by examining the different population
groups and lifestyles and behaviours associated with
fire fatalities. Overall, the effectiveness measures
described in this research combined what was done
by the fire and rescue service studied, the value to
the users of the service (the population of Merseyside),
and the manner in which the services were delivered
(Fryer, Antony, & Ogden, 2009).
Effectiveness of communication strategy
33. A particular issue with regard to assessing the effective-
ness of fire prevention approaches was attempting to
understand the communication aspect of fire preven-
tion (Andrews et al., 2014). Behavioural insights can
potentially aid in determining how best to contact/
influence people with regard to fire prevention accord-
ing to the behaviours they display (Clark & Smith,
2015). Typically, communication methods include face-
to-face via home fire safety checks or safe and well
visits, and leaflets distributed to local residents as part
Table 9. Ratio of overall fire prevention spend per head of
population (or cost) (in £000s) to fire incidents/fire incidents
involving
injuries/fire incidents involving fatalities in the Merseyside
area, 2006 to 2016.
2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016
Fire prevention cost/incidence ratio 16 16 14 15 15 17 18 18 23
18 22
Fire prevention cost/incidences involving
injuries ratio
202 246 221 217 221 182 226 190 260 225 294
Fire prevention cost/incidences involving
fatalities ratio
3664 1878 1878 3261 2654 4068 6468 2407 2626 1384 3236
Table 10. Fire prevention spend per head (in £s) and the number
of fire injuries and fatalities by population segment in the
Merseyside area, 2006 to 2016.
2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016
34. Fire prevention
spend per
head of
population
(Age 65+)
2.70 2.55 2.32 2.43 2.32 2.55 2.49 2.83 3.12 2.57 3.01
Fire prevention
spend per
head of
population
(Age < 65)
13.37 12.56 11.39 11.83 11.18 12.18 11.51 12.49 13.86 11.30
13.11
Fire injuries
(Age 65+)
35 31 35 24 25 43 36 52 40 35 25
Fire injuries
(Age < 65)
94 68 70 91 101 102 71 93 77 73 75
Fire fatalities
(Age 65+)
4 5 5 3 3 4 1 4 6 10 6
Fire fatalities
(Age < 65)
35. 2 6 5 3 4 1 2 6 3 5 4
PUBLIC MONEY & MANAGEMENT 425
of a publication scheme under the UK Freedom of Infor-
mation Act 2000. In addition, referral to other agencies,
such as NHS smoking cessation or alcohol reduction, can
provide an effective communication approach.
Conclusions
Evidence on cost-effectiveness is used in many
countries to inform decisions about the allocation of
public funds to public services and products. In this
paper we have examined a fire prevention strategy
assessment approach that supports moving from
output-based performance measures to effectiveness-
based performance measures. The fire prevention strat-
egy assessment approach included overall fire and
rescue service performance, performance in relation
to different population segments (especially the most
vulnerable groups such as the elderly), and fire inci -
dences, injuries and fatalities compared to spend per
head of the population covered by the fire and
rescue service concerned.
The key contribution from the research reported in
this paper is the development of an approach to
assessing the effectiveness of fire prevention strat-
egies, based on analysis of the approach to targeting
fire prevention, and multi-agency collaboration. Our
approach measures overall fire and rescue service per -
formance, performance in relation to different popu-
lation segments, and performance in relation to
36. vulnerable population segments using fire and
rescue service data and UK national statistics and
NHS data.
In terms of overall fire and rescue service effective-
ness indicated by how many fire incidences, injuries
and fatalities occurred over the 10-year period
studied, there was a clear downward trend in the
number of fire incidences within the region. There
also appeared to be a clear downward trend in the
overall proportion of the Merseyside population
involved in a fire incidence over the 10-year period.
The small proportions of those involved in a fire
injury or fatality meant that these figures were less
easy to interpret. Over the 10-year period studied, the
fire and rescue service spend per head on fire preven-
tion appeared to have been effective in reducing the
number of fire instances.
Given that a significant proportion of the fire fatal -
ities over the time period studied involved smoker’s
materials and alcohol consumption, this indicated
that referrals to NHS smoking cessation and alcohol
reduction services were an appropriate and necessary
aspect of current and future fire prevention.
The implications for theory and practice of the
research reported in this paper are the need to move
from output-based performance measures to effective-
ness-based performance measures. In this paper we
have examined how this can be achieved via detailed
analysis of fire incidences, injuries and fatalities in
terms of the different population segments within
the area covered by a given fire and rescue service. In
addition, it is necessary to analyse the costs associated
37. with fire prevention for the different population seg-
ments and, finally, it is important to recognize the
need for fire and rescue services to work closely with
other public sector agencies.
A limitation of the research presented is that due to
the variable nature of the populations served by
different fire and rescue services and the different area
geographies, and the different funding levels, meaning-
ful comparisons of effectiveness between different fire
and rescue services would be difficult to interpret.
Future research could, however, examine the use of
the fire prevention effectiveness assessment measures
in other fire and rescue service contexts. It is hoped
that the examination of fire prevention strategy assess-
ment reported in this paper will be of benefit to other
fire and rescue services, both in the UK and elsewhere.
References
Abdel-Maksoud, A., Elbanna, S., Mahama, H., & Pollanen, R.
(2015). The use of performance information in strategic
decision making in public organizations. International
Journal of Public Sector Management, 28(7), 528–549.
Adcroft, A., & Willis, R. (2005). The (un)intended outcome of
public sector performance measurement. International
Journal of Public Sector Management, 18(5), 386–400.
Ammons, D., & Roenigk, D. (2015). Performance management
in local government: is practice influenced by doctrine?
Public Performance and Management Review, 38(3), 514–
541.
Andrews, R., Ashworth, R., & Meier, K. (2014). Representative
bureaucracy and fire service performance. International
38. Public Management Journal, 17(1), 1–24.
Barton, H., & Valero-Silva, N. (2013). Policing in partnership:
a
case study in crime prevention. International Journal of
Public Sector Management, 26(7), 543–553.
Carvalho, J., Fernandes, M., Lambert, V., & Lapsley, I. (2006).
Measuring fire service performance: a comparative study.
International Journal of Public Sector Management, 19(2),
165–179.
Chainey, S. (2013). Using the vulnerable localities index to
identify priority areas for targeting fire safety services.
Fire Safety Journal, 62, 30–36.
Clare, J., Garis, L., Plecas, D., & Jennings, C. (2012). Reduced
fre-
quency and severity of residential fires following delivery
of fire prevention education by on-duty fire fighters:
Cluster randomized controlled study. Journal of Safety
Research, 43, 123–128.
Clark, A., & Smith, J. (2015). Experiencing a domestic fire: an
overview of key findings from a post incident research pro-
gramme. Safer Communities, 14(2), 95–103.
Corcoran, J., Higgs, G., Rohde, D., & Chhetri, P. (2011).
Investigating the association between weather conditions,
calendar events and socio-economic patterns with trends
in fire incidence: an Australian case study. Journal of
Geographic Systems, 13, 193–226.
Corcoran, J., Higgs, G., & Anderson, T. (2013). Examining the
use of a geodemographic classification in an exploratory
39. 426 M. TAYLOR ET AL.
analysis of variations in fire incidence in South Wales, UK.
Fire Safety Journal, 62, 37–48.
Devlin, N., & Parkin, D. (2004). Does NICE have a cost-
effective-
ness threshold and what other factors influence its
decisions? A binary choice analysis. Health economics, 13
(5), 437–452.
FRRF (2018) Fire and Rescue National Framework for England,
UK Home Office. Retrieved from https://www.gov.uk/
government/publications/fire-and-rescue-national-frame
work-for-england–2
Fryer, K., Antony, J., & Ogden, S. (2009). Performance
manage-
ment in the public sector. International Journal of Public
Sector Management, 22(6), 478–498.
Hastie, C., & Searle, R. (2016). Socio-economic and demo-
graphic predictors of accidental dwelling fire rates. Fire
Safety Journal, 84, 50–56.
HFSC (2017) Home fire safety check. Retrieved from https://
www.merseyfire.gov.uk/aspx/pages/prevention/campaign.
aspx
Higgins, E., Taylor, M., Jones, M., & Lisboa, P. (2013).
Understanding Community Fire Risk – A Spatial Model
for Targeting Fire Prevention Activities. Fire Safety
Journal, 62, 20–29.
40. HMICFRS (2018) HM Inspectorate of Constabularies, Fire and
Rescue Services, How we inspect fire and rescue services.
Retrieved from https://www.justiceinspectorates.gov.uk/
hmicfrs/fire-and-rescue-services/how-we-inspect-fire-and-
rescue-services/
HO (2017) UK Home Office, Fire Statistics: England April 2015
to March 2016. Retrieved from https://assets.publishing.
service.gov.uk/government/uploads/system/uploads/
attachment_data/file/611182/fire-statistics-england-
1516-hosb0517.pdf
Jaaskelainen, A., Laihonen, H., Lonnqvist, A., Palvalin, M.,
Sillanpaa, V., Pekkola, S., & Ukko, J. (2012). A contingency
approach to performance measurement in service oper-
ations. Measuring Business Excellence, 16(1), 43–52.
Jennings, C. (2013). Social and economic characteristics as
determinants of residential fire risk in urban neighbor-
hoods: A review of the literature. Fire Safety Journal, 62,
13–19.
Jonsson, A., Runefors, M., Sardqvist, S., & Nilson, F. (2016).
Fire-
related mortality in Sweden: temporal trends 1952 to 2013.
Fire technology, 52(6), 1697–1707.
Kloot, L. (2009). Performance measurement and accountabil-
ity in an Australian fire service. International Journal of
Public Sector Management, 22(2), 128–145.
Lehna, C., Coty, M., Fahey, E., Williams, J., Scrivener, D.,
Wishnia, G., & Myers, J. (2015). Intervention study for
changes in home fire safety knowledge in urban older
adults. Burns, 41(6), 1205–1211.
41. McKevitt, D. (2015). Debate: Value for money - in search of a
definition. Public Money & Management, 35(2), 99–100.
McNamara. (2017). Questioning risk-based fire and
life safety education age priorities. Injury Prevention, 23,
73–74.
MFRS (2018) Merseyside Fire and Rescue Authority Budget
and Medium Term Financial Plan Resolution 2015/2016 -
2019/20. Retrieved from http://mfra.merseyfire.gov.uk/
documents/s6640/Appendix%20A%20Budget%
20Resolution.pdf
MFRSIRM (2018) Merseyside fire and rescue service Integrated
Risk Management Plan. Retrieved from https://www.
merseyfire.gov.uk/aspx/pages/IRMP/pdf/IRMP_2017-20_
Update_2017.pdf
Mulvaney, C., Kendrick, D., Towner, E., Brussoni, M., Hayes,
M.,
Powell, J., … Ward, H. (2009). Fatal and non-fatal fire inju-
ries in England 1995-2004: time trends and inequalities
by age, sex and area deprivation. Journal of Public Health,
31(1), 154–161.
Murphy, P., & Greenhalgh, K. (2013). Performance manage-
ment in fire and rescue services. Public Money &
Management, 33(3), 225–232.
NAO (2015) Financial sustainability of fire and rescue services,
UK National Audit Office HC 491. Retrieved from https://
www.nao.org.uk/report/financial-sustainability-of-fire-and-
rescue-services/
Nomis, Official labour market statistics, UK Office for National
Statistics. Retrieved from https://www.nomisweb.co.uk/
42. Prieto, L., & Sacristan, J. (2003). Problems and solutions in
calcu-
lating quality-adjusted life years (QALYs). Health and Quality
of Life Outcomes, 1(80), doi: 10.1186/1477-7525-1-80
Puolokainen, T. (2017). Reforming fire and rescue services: a
comparative study of Estonia and Georgia. International
Journal of Public Sector Management, 30(3), 227–240.
Saar, I. (2017). The effects of the lower ignition propensity
cigarettes standard in Estonia: time-series analysis. Injury
prevention, DOI: 10.1136/injuryprev-2016-042187
Safe and Well (2017) Fire Safety in the Home, Safe and Well
visit. Retrieved from http://www.fireservice.co.uk/safety/
Safe and Well Visit Principles (2017) Principles for a Safe and
Well visit by a Fire and Rescue Service, Public Health
England. Retrieved from https://www.england.nhs.uk/wp-
content/uploads/2015/09/safe-well-visit-pinciples.pdf
Saramago, P., Cooper, N., Sutton, A., Hayes, M., Dunn, K.,
Manca, A., & Kendrick, D. (2014). Cost-effectiveness of inter-
ventions for increasing the possession of functioning
smoke alarms in households with pre-school children: a
modelling study. BMC public health, 14(1), 459–471.
Tannous, W., Whybro, M., Lewis, C., Broomhall, S.,
Ollerenshaw, M., Watson, G., … Franks, E. (2017). Home
Fire Safety Checks in New South Wales: an economic evalu-
ation of the pilot program. Journal of Risk Research, DOI: 10.
1080/13669877.2017.1281336
Taylor, M., Higgins, E., Lisboa, P., Jarman, I., & Hussain, A.
(2016). Community fire prevention via population segmen-
43. tation modelling. Community Development Journal, 51(2),
229–247.
Turner, S., Johnson, R., Weightman, L., Rodgers, S., Arthur, G.,
Bailey, R., & Lyons, R. (2017). Risk factors associated with
unintentional house fire incidents, injuries and deaths in
high-income countries: a systematic review. Injury
Prevention, 23, 131–137.
Warda, L., Tenenbein, M., & Moffact, M. (1999). House fire
injury prevention update. Part II. A review of the effective-
ness of preventive interventions. Injury Prevention, 5, 217–
225.
Welsh Government Social Research (2013) Evaluation of the
Home Fire Safety Check, Knowledge and Analytical
Services, Welsh Government, Rhydycar, Merthyr Tydfil,
CF48 1UZ.
Weinholt, A., & Andersson Granberg, T. (2015). New collabor -
ations in daily emergency response: Applying cost-benefit
analysis to new first response initiatives in the Swedish fire
and rescue service. International Journal of Emergency
Services, 4(2), 177–193.
Wuschke, K., Clare, J., & Garis, L. (2013). Temporal and geo-
graphic clustering of residential structure fires: A theoreti -
cal platform for targeted fire prevention. Fire Safety Journal,
62, 3–12.
PUBLIC MONEY & MANAGEMENT 427
https://www.gov.uk/government/publications/fire-and-rescue-
national-framework-for-england%E2%80%932
https://www.gov.uk/government/publications/fire-and-rescue-
national-framework-for-england%E2%80%932
46. Instructions
For this assignment, choose a journal article from the CSU
Online Library that relates to one of the following topics:
Article for this assignment is reference at the bottom.
· fire prevention,
The article chosen for the review must be from a professional
(scholarly) journal and published within the past five years. The
Academic One File, Academic Search Ultimate, and Business
Source Ultimate databases are good sources of journals for
safety-related articles. If you are not sure if an article you have
chosen meets the assignment requirements, be sure to contact
your professor for additional guidance.
Prepare a review and critique of the chosen journal article that
includes the following elements:
· a brief introduction to the article,
· a summary and analysis of the key points in the article,
· how the article relates to the concepts presented in the course
textbook,
· whether or not the article provides support for the relevant
Occupational Safety and Health Administration (OSHA)
standards surrounding the topic, and
· how the article could be used by the safety professional to
improve accident prevention efforts.
The assignment submission must be a minimum of two pages in
length, not including title and reference pages. APA formatting
is required. All sources used, including the course textbook,
must be referenced. Paraphrased and quoted material must have
accompanying in-text citations. For assistance in the search for
professional journal articles, this resource from the CSU Online
Library may be helpful.
References
Taylor, M., Appleton, D., Keen, G., & Fielding, J. (2019).
Assessing the effectiveness of fire prevention strategies. Public
47. Money & Management, 39(6), 418–427. https://doi-
org.libraryresources.columbiasouthern.edu/10.1080/09540962.2
019.1579439
Rutgers Department of Health
Informatics
School of Health Professions Health
Information Management Program
BINF 5520 Health Data
Analytics
Analyzing Cardiac Surgery Data New York State 2011 - 2013.
The purpose of this assignment is to provide students with an
opportunity to analyze data that was collected by the New York
State Department of Health for all Isolated Coronary Artery
Bypass Graft (CABG) procedures that occurred during 2011
thru 2013. To complete this assignment, do the following:
1. Read the article that is on Canvas as “Adult Cardiac Surgery
in New York State 2011-2013”. Pay particular attention to
Table 6 (Page 25) of this article where the data for each surgeon
is presented.
2. After you have completed reading the article mentioned
above, open up the Excel file that is on Canvas called
“Analyzing Cardiac Surgery Data”. This is essentially the date
that you see in the article. I have removed all surgeons who did
less than 10 procedures at any hospital. I have also removed the
category of surgeons called “All Others” which you see in Table
6 of the article referenced above.
3. Based on the data that is presented in “Analyzing Cardiac
Surgery Data”, please answer the following questions:
1) How many surgeons who performed 200 or more procedures
during the time-period in Question had five or fewer deaths
48. within 30 days of surgery?
2) Please create a Pivot Table, which summarizes the data,
using the Hospital Name as the Row Label, the Region as the
Column Label and the Sum of Cases as the Value. Which
Region, Upstate or Downstate, performed more Open Heart
Procedures?
3) Please create a Pivot Table, which summarizes the data using
the Surgeon Name as the Row Label, the Hospital Name as the
second Row Label and the Sum of Cases as the Value. You do
not need to provide a Column Label. How many surgeons
performed Open Heart Procedures in more than one hospital?
Analyzing Open Heart Risk Stratification Data (Continued)
4) Does the definition of the Risk Adjusted Mortality Rate
(RARM), that is discussed on page 4 of the reading assignment
make intuitive sense to you as an observer of this data?
5) Does the use of a 30-day window for calculating the Risk
Adjusted Mortality Rate seem to you to be a useful metric for
rating surgeons and/or hospitals?
49. 6) Based on the data presented here, which surgeon and hospital
would you advise your 65 year old family member, who had a
number of different pre-operative risk factors (see page 13 for
discussion of the calculation of a risk-adjusted mortality rate) to
perform the isolated CABG procedure?
ADULT
CARDIAC
SURGERY
in New York State
2011-2013
Department
of Health
September 2016
Members of the New York State Cardiac Advisory Committee
Chair
Spencer King III, M.D.
Professor of Medicine, Emeritus
50. Emory University School of Medicine
Atlanta, GA
Vice Chair
Gary Walford, M.D.
Associate Professor of Medicine
Johns Hopkins Medical Center
Baltimore, MD
Members
M. Hashmat Ashraf, M.D., FRCS
Chief, Department of Cardiothoracic Surgery
Kaleida Health
Buffalo, NY
Peter B. Berger, M.D.
Senior Vice President of Clinical Research and
Interventional Cardiologist
Professor of Cardiology and Medicine
Northwell Health
Hofstra - Northwell School of Medicine
New Hyde Park, NY
Frederick Bierman, M.D.
Director of Graduate Medical Education
Westchester Medical Center
Valhalla, NY
Jeptha Curtis, M.D.
Asst. Professor, Dept. of Internal Medicine
Director, Center for Outcomes Research & Evaluation Data
Analytic Center
Yale University School of Medicine
New Haven, CT
Jeffrey P. Gold, M.D.
51. Chancellor
University of Nebraska Medical Center
Omaha, NE
Alice Jacobs, M.D.
Director, Cardiac Catheterization Laboratory &
Interventional Cardiology
Boston Medical Center
Boston, MA
Desmond Jordan, M.D.
Associate Professor of Clinical Anesthesiology in
Biomedical Informatics
NY Presbyterian Hospital – Columbia
New York, NY
Thomas Kulik, M.D.
Director, Pulmonary Hypertension Program
Children’s Hospital Boston
Boston, MA
Stephen Lahey, M.D.
Chief, Division of Cardiothoracic Surgery
University of Connecticut Health Center
Farmington, CT
Frederick S. Ling, MD
Professor in Medicine (Cardiology)
University of Rochester Medical Center
Rochester, NY
Ralph Mosca, M.D.
Vice Chairman, Department of Cardiac Surgery
Director, Congenital Cardiac Surgery
NYU Medical Center
52. New York, NY
Carlos E. Ruiz, M.D., Ph.D.
Professor of Cardiology in Pediatrics and Medicine
Director, Structural and Congenital Heart Disease
Hackensack University Medical Center Heart
and Vascular Hospital
The Joseph M. Sanzari Children’s Hospital
Hackensack, NJ
Samin K. Sharma, M.D.
Director, Cardiac Catheterization Laboratory
Mount Sinai Medical Center
New York, NY
Craig Smith, M.D.
Chairman, Department of Surgery
NY Presbyterian Hospital – Columbia
New York, NY
Nicholas Stamato, M.D.
Co-Director, Cardiology
Campbell County Memorial Hospital
Gilette, WY
Thoralf Sundt, III, M.D.
Chief, Cardiac Surgical Division
Co-Director, Heart Center and Institute for Heart,
Vascular and Stroke
Massachusetts General Hospital
Boston, MA
James Tweddell, M.D.
Surgical Director and Executive Co-Director
The Heart Institute
Professor of Surgery
53. Cincinnati Children’s Hospital Medical Center
Cincinnati, OH
Ferdinand Venditti, Jr., M.D.
Executive Vice President for System Care Delivery
Hospital General Director
Vice Dean for Clinical Affairs
Professor of Medicine
Albany Medical Center
Albany, NY
Andrew S. Wechsler, M.D.
Professor and Chair, Department of Cardiothoracic Surgery
Drexel University College of Medicine
Philadelphia, PA
Consultant
Edward L. Hannan, Ph.D.
Distinguished Professor Emeritus
Department of Health Policy, Management & Behavior
Associate Dean Emeritus
University at Albany, School of Public Health
Rensselaer, NY
Cardiac Surgery Reporting System Subcommittee
Members & Consultants
Craig Smith, M.D. (Chair)
Chairman, Department of Surgery
NY Presbyterian Hospital – Columbia
M. Hashmat Ashraf, M.D., FRCS
Chief, Department of Cardiothoracic Surgery
Kaleida Health
54. Jeffrey P. Gold, M.D.
Chancellor
University of Nebraska Medical Center
Edward L. Hannan, Ph.D.
Distinguished Professor Emeritus
Department of Health Policy, Management & Behavior
Associate Dean Emeritus
University at Albany, School of Public Health
Desmond Jordan, M.D.
Associate Professor of Clinical Anesthesiology in
Biomedical Informatics
NY Presbyterian Hospital – Columbia
Stephen Lahey, M.D.
Chief, Division of Cardiothoracic Surgery
University of Connecticut Health Center
Ralph Mosca, M.D.
Vice Chairman, Department of Cardiac Surgery
Director, Congenital Cardiac Surgery
NYU Medical Center
Carlos E. Ruiz, M.D., Ph.D.
Professor of Cardiology in Pediatrics and Medicine
Director, Structural and Congenital Heart Disease
Hackensack University Medical Center Heart
and Vascular Hospital
The Joseph M. Sanzari Children’s Hospital
Nicholas J. Stamato, M.D.
Co-Director, Cardiology
Campbell County Memorial Hospital
55. Thoralf Sundt, III, M.D.
Chief, Cardiac Surgical Division
Co-Director, Heart Center and Institute for Heart, Vascular
and Stroke
Massachusetts General Hospital
James Tweddell, M.D.
Surgical Director and Executive Co-Director
The Heart Institute
Professor of Surgery
Cincinnati Children’s Hospital Medical Center
Andrew S. Wechsler, M.D.
Professor and Chair, Department of Cardiothoracic Surgery
Drexel University College of Medicine
Staff to CSRS Analysis Workgroup –
New York State Department of Health
Foster C. Gesten, MD
Chief Medical Officer
Office of Quality and Patient Safety, NYSDOH
Kimberly S. Cozzens, M.A.
Program Manager
Cardiac Services Program
Ashraf Al-Hamadani, MD, MPH
Clinical Record Reviewer
Cardiac Services Program
Lori Frazier
Project Assistant
Cardiac Services Program
Jessica Kincaid
Quality Improvement Project Coordinator
59. Table 6 In-Hospital/30-Day Observed, Expected and Risk-
Adjusted Mortality Rates
by Surgeon for Isolated CABG and Valve Surgery (done in
combination
with or without CABG) in New York State, 2011-2013
Discharges . . . . . . . . . . . . . . . . 25
Table 7 Summary Information for Surgeons Practicing at More
Than One Hospital, 2011-2013. . . . 33
6
SURGEON AND HOSPITAL VOLUMES FOR TOTAL ADULT
CARDIAC SURGERY, 2011-2013 . . . . . . . 36
Table 8 Surgeon and Hospital Volume for Isolated CABG,
Valve or Valve/CABG,
Other Cardiac Surgery and Total Adult Cardiac Surgery, 2011-
2013 . . . . . . . . . . . . . . 36
CRITERIA USED IN REPORTING SIGNIFICANT RISK
FACTORS (2013) . . . . . . . . . . . . . . . . . . . 43
MEDICAL TERMINOLOGY . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . 45
APPENDIX 1 Risk Factors for CABG In-Hospital / 30-Day
Deaths
in New York State in 2013 . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . 47
APPENDIX 2 Risk Factors for CABG 30-Day Readmissions
in New York State in 2013 . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . 49
60. APPENDIX 3 Risk Factors for Valve Surgery In-Hospital/30-
Day Mortality
in New York State in 2011-2013 . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . 51
APPENDIX 4 Risk Factors for Valve and CABG Surgery In-
Hospital/30-Day Mortality
in New York State in 2011-2013 . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . 53
APPENDIX 5 Risk Factors for Isolated CABG In-Hospital/30-
Day Mortality
in New York State 2011-2013 . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . 55
NEW YORK STATE CARDIAC SURGERY CENTERS . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . 56
7
INTRODUCTION
For over twenty years, the NYS Cardiac Data Reporting System
has been a powerful resource for quality
improvement in the areas of cardiac surgery and percutaneous
coronary interventions (PCI). Building on
this strong foundation, we are pleased to include in one report
information on mortality after coronary
artery bypass graft (CABG) surgery and valve repair or
replacement surgery, readmissions after CABG, and
preliminary information on transcatheter aortic valve
replacement (TAVR) in NYS.
New York State (NYS) has taken a leadership role in setting
61. standards for cardiac services, monitoring
outcomes and sharing performance data with patients, hospitals
and physicians. Hospitals and doctors
involved in cardiac care have worked in cooperation with the
NYS Department of Health (Department of
Health) and the NYS Cardiac Advisory Committee (Cardiac
Advisory Committee) to compile accurate and
meaningful data that can and have been used to enhance quality
of care. We believe that this process
has been instrumental in achieving the excellent outcomes that
are evidenced in this report for centers
across NYS.
The information contained in this report is intended for health
care providers, patients and families of
patients who are considering cardiac surgery. It includes:
• Mortality rates, adjusted for patient severity of illness, for
CABG surgery and valve repair or
replacement surgery at NYS hospitals.
• Readmission rates, adjusted for patient severity of illness,
following CABG at NYS hospitals.
• Mortality rates, adjusted for patient severity of illness,
following CABG for surgeons performing
the procedure.
• Volume (number of cases) of all cardiac surgery for NYS
hospitals and surgeons.
• Volume (number of cases) of TAVR at NYS hospitals.
• Description of the patient risk factors associated with
mortality for CABG and valve surgery and those
associated with readmissions after CABG surgery.
The data that serve as the basis for this report are collected by
62. the NYS Department of Health
cooperatively with hospitals throughout the state. Careful
auditing and rigorous analysis assure that these
reports represent meaningful outcome assessments. The report
was developed with clinical guidance
from the NYS Cardiac Advisory Committee, an advisory body
to the Commissioner of Health consisting
of nationally recognized cardiac surgeons, cardiologists and
others from related disciplines working both
in New York State and elsewhere. The Cardiac Advisory
Committee is to be commended for sustained
leadership in these efforts.
As they develop treatment plans, we encourage doctors to
discuss this information with their patients and
colleagues. While these statistics are an important tool in
making informed health care choices, individual
treatment plans must be made by doctors and patients together
after careful consideration of all pertinent
factors. It is important to recognize that many factors can
influence the outcome of cardiac surgery. These
include the patient’s health before the procedure, the skill of the
operating team and general after-care.
In addition, keep in mind that the information in this booklet
does not include data after 2013. Important
changes may have taken place in hospitals during that time
period.
It is important that patients and physicians alike give careful
consideration to the importance of healthy
lifestyles for all those affected by heart disease. While some
risk factors, such as heredity, gender and age
cannot be controlled, others certainly can. Controllable risk
factors that contribute to a higher likelihood
of developing coronary artery disease are high cholesterol
levels, cigarette smoking, high blood pressure,
63. obesity and sedentary lifestyle. Careful attention to these risk
factors after surgery will continue to be
important in promoting good health and preventing recurrence
of disease.
Hospitals and physicians in NYS can take pride in the excellent
patient care provided and in their role
in contributing to this unique collaborative quality improvement
system. The Department of Health will
continue to work in partnership with hospitals and physicians to
ensure that continued high-quality cardiac
surgery is available to NYS residents.
8
CORONARY ARTERY BYPASS GRAFT SURGERY (CABG)
Heart disease is the leading cause of death
in NYS, and the most common form of heart
disease is atherosclerotic coronary artery
disease. Different treatments are recommended
for patients with coronary artery disease. For
some people, changes in lifestyle, such as
dietary changes, not smoking and regular
exercise, can result in great improvements in
health. In other cases, medication prescribed
for high blood pressure or other conditions can
make a significant difference.
Sometimes, however, an interventional
procedure is recommended. The two common
procedures performed on patients with
coronary artery disease are CABG surgery and
percutaneous coronary intervention (PCI).
64. CABG surgery is an operation in which a vein
or artery from another part of the body is used
to create an alternate path for blood to flow
to the heart muscle, bypassing the arterial
blockage. Typically, a section of one of the large
(saphenous) veins in the leg, the radial artery
in the arm or the mammary artery in the chest
is used to construct the bypass. One or more
bypasses may be performed during a single
operation, since providing several routes for the
blood supply to travel is believed to improve
long-term success for the procedure. CABG
surgery is one of the most common, successful
major operations currently performed in the
United States.
As is true of all major surgery, risks must be
considered. The patient is totally anesthetized
and there is generally a substantial recovery
period in the hospital followed by several weeks
of recuperation at home. Even in successful
cases, there is a risk of relapse causing the need
for another operation.
Those who have CABG surgery are not cured
of coronary artery disease; the disease can
still occur in the grafted blood vessels or other
coronary arteries. In order to minimize new
blockages, patients should continue to reduce
their risk factors for heart disease.
CARDIAC VALVE PROCEDURES
Heart valves control the flow of blood as
it enters the heart and is pumped from
the chambers of the heart to the lungs for
65. oxygenation and back to the body. There are
four valves: the tricuspid, mitral, pulmonary
and aortic valves. Heart valve disease occurs
when a valve cannot open all the way because
of disease or injury, thus causing a decrease in
blood flow to the next heart chamber. Another
type of valve problem occurs when the valve
does not close completely, which leads to blood
leaking backward into the previous chamber.
Either of these problems causes the heart to
work harder to pump blood or causes blood to
back up in the lungs or lower body.
When a valve is stenotic (too narrow to allow
enough blood to flow through the valve
opening) or incompetent (cannot close tightly
enough to prevent the backflow of blood), one
of the treatment options is to repair the valve.
Repair of a stenotic valve typically involves
widening the valve opening, whereas repair
of an incompetent valve is typically achieved
by narrowing or tightening the supporting
structures of the valve. The mitral valve is
particularly amenable to valve repairs because
its parts can frequently be repaired without
having to be replaced.
In many cases, defective valves are replaced
rather than repaired, using either a mechanical
or biological valve. Mechanical valves are built
using durable materials that generally last a
lifetime. Biological valves are made from tissue
taken from pigs, cows or humans. Mechanical
and biological valves each have advantages
and disadvantages that can be discussed with
66. referring physicians.
The most common heart valve surgeries involve
the aortic and mitral valves. Patients undergoing
heart surgery are totally anesthetized and
are usually placed on a heart-lung machine,
whereby the heart is stopped for a short period
of time using special drugs. As is the case for
CABG surgery, there is a recovery period of
9
several weeks at home after being discharged
from the hospital. Some patients require
replacement of more than one valve and some
patients with both coronary artery disease and
valve disease require valve replacement and
CABG surgery. This report contains outcomes
for the following valve surgeries when done
alone or in combination with CABG: Aortic Valve
Replacement, Mitral Valve Repair, Mitral Valve
Replacement and Multiple Valve Surgery.
In recent years, a new technique for
replacement of the aortic valve has been tested
and approved for use in the United States under
certain circumstances. This procedure, known
as Transcatheter Aortic Valve Replacement
(TAVR, also sometimes called Transcatheter
Aortic Valve Implantation or TAVI), differs
from traditional surgical valve replacement in
that the replacement valve is delivered to the
heart through a catheter rather than through
67. a standard surgical incision. The procedure is
performed collaboratively by cardiologists and
cardiac surgeons.
During the time-period included in this report
(2011-2013), the procedure was performed
relatively infrequently at a limited number of
hospitals. The volume of cases in this time
period is not large enough to adequately risk-
adjust outcomes at the hospital or physician
level. In the interest of providing some
information about this emerging procedure
while avoiding the risk of drawing invalid
conclusions based on limited data, this report
contains the volume of TAVR cases in 2013 for
each hospital as well as the volume of cases for
NYS and in-hospital / 30-Day mortality rate for all
TAVR procedures performed in NYS in Table 5.
THE DEPARTMENT OF HEALTH PROGRAM
For many years, the Department of Health
has been studying the effects of patient and
treatment characteristics (called risk factors)
on outcomes for patients with heart disease.
Detailed statistical analyses of the information
received from the study have been conducted
under the guidance of the Cardiac Advisory
Committee, a group of independent practicing
cardiac surgeons, cardiologists and other
professionals in related fields.
The results have been used to create a cardiac
profile system which assesses the performance
of hospitals and surgeons over time,
independent of the severity of each individual
68. patient’s pre-operative conditions.
Designed to improve health in people with heart
disease, this program is aimed at:
• understanding the health risks of patients that
adversely affect how they will fare in coronary
artery bypass surgery and/or valve surgery;
• improving the results of different treatments of
heart disease;
• improving cardiac care; and
• providing information to help patients make
better decisions about their own care.
PATIENT POPULATION
This report is based on data for patients
discharged between December 1, 2010, and
November 30, 2013, provided by all non-federal
hospitals in NYS where cardiac surgery is
performed. The analysis period for this report
includes patients discharged in December
2010 but not those discharged in December
2013. This strategy allows for more timely
report publication by eliminating the need to
track patients for 30-day mortality into the
following calendar year. Inclusion of cases from
the previous December allows for meaningful
comparison of 12-month volume as found in
previous reports. The single year analysis for
2013 cases includes patients discharged from
December 1, 2012 through November 30, 2013.
In total there were 57,623 cardiac surgical
69. procedures performed during this time period.
For various reasons, some of these cases
are excluded from analysis in this report. The
reasons for exclusion and number of cases
affected are described below.
10
Records for 118 patients residing outside the
United States were excluded because these
patients could not be followed after hospital
discharge. There were 12 cases excluded from
analysis because each 30-day mortality can only
be associated with a single cardiac surgery.
Beginning with patients discharged in 2006,
the Department of Health, with the advice of
the Cardiac Advisory Committee, began a trial
period of excluding from publicly released
reports any patients meeting the Cardiac Data
System definition of pre-operative cardiogenic
shock. Cardiogenic shock is a condition
associated with severe hypotension (very low
blood pressure). [The technical definition used
in this report can be found on page 43.] Patients
in cardiogenic shock are extremely high-risk,
but for some, cardiac surgery may be their best
chance for survival. Furthermore, the magnitude
of the risk is not always easily determined
using registry data. These cases were excluded
after careful deliberation and input from NYS
providers and others in an effort to ensure that
physicians could accept these cases where
70. appropriate without concern over a detrimental
impact on their reported outcomes. In total, 569
cases with cardiogenic shock were removed
from the data. This accounts for 0.99 percent
of all cardiac surgeries (CABG, valve surgery
and other cardiac surgery reported in this data
system) in the three years.
After all of the above exclusions, there were
56,924 cardiac surgeries analyzed in this report.
Isolated CABG surgery represented 43.81
percent of all adult cardiac surgery included
in this report. Valve or combined valve/CABG
surgery represented 39.02 percent of all adult
cardiac surgery for the same period. TAVR
represented 2.37 percent of all cardiac surgeries
reported. Total cardiac surgery, isolated CABG,
valve surgery and other cardiac surgery volumes
are tabulated in Table 8 by hospital and surgeon
for the period 2011 through 2013.
While there were 8,168 CABG cases included
in the mortality analysis for 2013 discharges,
some additional exclusions were required for
the readmission analysis. Records belonging to
patients residing outside NYS were excluded
because there is no reliable way to track out-
of state readmissions. This accounted for 287
cases. Another 127 patients were excluded
because they died in the same admission
as their index CABG, so readmission was
impossible. Two patients were transferred after
CABG and discharged from the second hospital
in December, making them ineligible for
30-day follow-up.
71. In total, the number of exclusions was 416,
leaving 7,755 cases to be examined for 30-day
readmission rates.
Note on Hospitals Not Performing Cardiac
Surgery During Entire 2011 – 2013 Period
Southside Hospital began performing cardiac
surgery in February 2011. Millard Fillmore
Hospital closed in 2013 and performed the last
cardiac surgery in March of that year.
11
RISK ADJUSTMENT
FOR ASSESSING PROVIDER PERFORMANCE
Provider performance is directly related to
patient outcomes. Whether patients recover
quickly, experience complications, require
another hospitalization, or die following a
procedure is, in part, a result of the kind of
medical care they receive. It is difficult, however,
to compare outcomes across hospitals when
assessing provider performance because
different hospitals treat different types of
patients. Hospitals with sicker patients may
have higher rates of death and readmission
than other hospitals in the state. The following
describes how the Department of Health adjusts
for patient risk in assessing provider outcomes.
Data Collection, Data Validation and
Identifying In-Hospital/30-Day Deaths and
72. 30-Day Readmission
As part of the risk-adjustment process, NYS
hospitals where cardiac surgery is performed
provide information to the Department of Health
for each patient undergoing that procedure.
Cardiac surgery departments collect data
concerning patients’ demographic and clinical
characteristics. Approximately 40 of these
characteristics (called risk factors) are collected
for each patient. Along with information about
the procedure, physician and the patient’s status
at discharge, these data are entered into a
computer and sent to the Department of Health
for analysis.
Data are verified through review of unusual
reporting frequencies, cross-matching of cardiac
surgery data with other Department of Health
databases and a review of medical records for
a selected sample of cases. These activities
are extremely helpful in ensuring consistent
interpretation of data elements across hospitals.
The analyses in this report base mortality on
deaths occurring during the same hospital stay
in which a patient underwent cardiac surgery
and on deaths that occur after discharge but
within 30 days of surgery.
An in-hospital death is defined as a patient who
died subsequent to CABG or valve surgery
during the same admission or was discharged to
hospice care and expired within 30 days.
Deaths that occur after hospital discharge but
73. within 30 days of surgery are also counted in
the risk-adjusted mortality analyses. This is
done because hospital length of stay has been
decreasing and, in the opinion of the Cardiac
Advisory Committee, most deaths that occur
after hospital discharge but within 30 days of
surgery are related to complications of surgery.
Data on deaths occurring after discharge from
the hospital are obtained from the Department
of Health, the New York City Department of
Health and Mental Hygiene Bureau of Vital
Statistics, and the National Death Index.
Data on readmissions are obtained from the
Department of Health’s acute care hospital
dataset, the Statewide Planning and Research
Cooperative System (SPARCS), which contains
data pertaining to all acute care hospital
discharges in the state.
Thirty-day readmission is defined as admission
to a NYS non-Federal hospital within 30 days
of discharge from the index hospitalization. For
patients whose index hospitalization ends in
transfer to another acute care facility, the 30
day period begins upon discharge from the
second hospital.
Assessing Patient Risk
Each person who develops heart disease
has a unique health history. A cardiac profile
system has been developed to evaluate the
risk of treatment for each individual patient
based on his or her history, weighing the
74. important health factors for that person based
on the experiences of thousands of patients
who have undergone the same procedures in
recent years. All important risk factors for each
patient are combined to create a risk profile.
For example, an 80-year-old patient with renal
failure requiring dialysis has a very different risk
profile than a 40-year-old with no renal failure.
The statistical analyses conducted by the
Department of Health consist of determining
which of the risk factors collected are
significantly related to death or readmission
following CABG and/or valve surgery and
12
determining how to weigh the significant risk
factors to predict the chance each patient will
have of dying or being readmitted, given his or
her specific characteristics.
Doctors and patients should review individual
risk profiles together. Treatment decisions must
be made by doctors and patients together after
consideration of all the information.
Predicting Patient Mortality Rates for Providers
The statistical methods used to predict mortality
on the basis of the significant risk factors are
tested to determine whether they are sufficiently
accurate in predicting mortality for patients
who are extremely ill prior to undergoing the
75. procedure as well as for patients who are
relatively healthy. These tests have confirmed
that the models are reasonably accurate in
predicting how patients of all different risk levels
will fare when undergoing cardiac surgery.
The mortality rate for each hospital and surgeon
is also predicted using the relevant statistical
models. This is accomplished by summing
the predicted probabilities of death for each
of the provider’s patients and dividing by the
number of patients. The resulting rate is an
estimate of what the provider’s mortality rate
would have been if the provider’s performance
were identical to the state performance.
The percentage is called the predicted or
expected mortality rate (EMR). A hospital’s EMR
is contrasted with its observed mortality rate
(OMR), which is the number of patients who died
divided by the total number of patients.
Computing the Risk-Adjusted Mortality Rate
The risk-adjusted mortality rate (RAMR)
represents the best estimate, based on the
associated statistical model, of what the
provider’s mortality rate would have been if
the provider had a mix of patients identical to
the statewide mix. Thus, the RAMR has, to the
extent possible, ironed out differences among
providers in patient severity of illness, since
it arrives at a mortality rate for each provider
for an identical group of patients. To calculate
the RAMR, the OMR is first divided by the
provider’s EMR. If the resulting ratio is larger
than one, the provider has a higher mortality
76. rate than expected on the basis of its patient
mix; if it is smaller than one, the provider has
a lower mortality rate than expected from its
patient mix. For isolated CABG patients the
ratio is then multiplied by the overall statewide
mortality rate of 1.84 percent (in-hospital/30-day
mortality in 2013) to obtain the provider’s RAMR.
For the three-year period 2011-2013, the ratio
is multiplied by 1.51 percent (in-hospital/30-day
mortality rate) for isolated CABG patients or 3.45
percent (in-hospital/30-day mortality rate) for
valve or valve/CABG patients.
There is no Statewide EMR or RAMR, because
the statewide data is not risk-adjusted. The
Statewide OMR (number of total cases divided
by number of total deaths) serves as the basis for
comparison for each hospital’s EMR and RAMR.
Interpreting the Risk-Adjusted Mortality Rate
If the RAMR is significantly lower than the
statewide mortality rate, the provider has a
significantly better performance than the state
as a whole; if the RAMR is significantly higher
than the statewide mortality rate, the provider
has a significantly worse performance than the
state as a whole.
The RAMR is used in this report as a measure
of quality of care provided by hospitals and
surgeons. However, there are reasons that a
provider’s RAMR may not be indicative of its
true quality. For example, extreme outcome
rates may occur due to chance alone. This
77. is particularly true for low-volume providers,
for whom very high or very low mortality
rates are more likely to occur than for high-
volume providers. To prevent misinterpretation
of differences caused by chance variation,
confidence intervals are reported in the results.
The interpretations of those terms are provided
later when the data are presented.
Differences in hospital coding of risk factors
could be an additional reason that a provider’s
RAMR may not be reflective of quality of care.
The Department of Health monitors the quality
of coded data by reviewing samples of patients’
medical records to ascertain the presence
of key risk factors. When significant coding
problems are discovered, hospitals are required
to correct these data and are subjected to
subsequent monitoring.
13
Although there are reasons that RAMRs
presented here may not be a perfect reflection
of quality of care, the Department of Health feels
that this information is a valuable aid in choosing
providers for cardiac surgery.
Predicting Patient Readmission and
Computing and Interpreting Risk-Adjusted
Readmission Rates
Patient risk of 30-day readmission is assessed
using the same methods used for assessing