Table Of ContentDesigning future ambulance
transport for patient safety:
Research undertaken
Contents
Executive summary 2
1. Introduction 4
2. Literature review 6
3. Method 16
4. Results 23
5. Discussion 40
6. Conclusion 43
7. References 44
Research team/authors
Helen Hamlyn Research Centre, Royal College of Art, London
Prof Roger Coleman Principal Investigator
Prof Dale Harrow Co-Investigator
Owen Evans Senior Research Associate
Merih Kunur Senior Research Associate
Sally Halls Research Associate
Daniel Kafka MA designer
Healthcare Ergonomics and Patient Safety Unit,
University of Loughborough
Dr Sue Hignett Principal Investigator
Emma Crumpton Research Fellow
Anna Jones Research Associate
1
Executive summary
BACKGROUND METHOD
In March 2005, the NPSA prioritisation panel strongly supported Three types of data were collected: archival incident reports,
a project on ambulance design to be taken forward in the coming research literature and empirical data from workshops. The
year, in response to concerns relating to the design of vehicles archival data were collected from three sources about reported
and equipment that impact on patient safety. In June 2005, the incidents relating to ambulance, ambulance equipment design
Department of Health (DH) set out a vision for the provision of and use, and patient and staff safety. The research literature
future ambulance services by 2010. This included providing an review was used to not only set out the background context but
increasing range of quality mobile healthcare services for patients also to develop the conceptual framework for the analysis of
with urgent and emergency care needs (e.g. ‘see and treat’). The the workshop data. Empirical data were collected from four user
overarching aims are that patients will receive improved care by workshops.
consistently receiving the right response, first time, in time, and
that more patients will be treated in the community, resulting in
more effective and efficient use of NHS resources. It seems likely
that these changes will require different vehicles and equipment
for ambulance services.
AIM
This first scoping study aims to investigate the developing models
of service provision and the short and long term requirements
of vehicles and equipment that will be needed to address
the concerns of patient and staff safety in the future in the
Ambulance Service.
2
RESULTS DISCUSSION AND CONCLUSION
A dataset of 1,52 incidents was received from the National Each of the nine design challenges were reviewed to look at the
Reporting and Learning System (NRLS) database and 1,259 individual datasets (literature, incident reports and workshop
were retrieved from the Manufacturer and User facility Device themes). There was found to be a divergence between the
Experience (MAUDE) database. Ten ambulance trusts responded NRLS data and the incidents reported by the individual trusts for
to a request for information (from the 2 trusts contacted). some of the design challenges, for example securing people and
The incidents were scrutinised individually and initially coded to equipment in transit, and equipment. There are a number of
provide a framework for discussion at the workshops. After the possible causes for this divergence, including the design of the
analysis of the workshop data, the incident reports were reviewed NRLS interface and subsequent data input (this is recognised and
and coded into the nine design challenges. expected by the NPSA), through to local reporting cultures and
screening by trusts before sending incident data to the project
The data from the workbooks at the strategic workshop were
team. Issues that were raised in the workshops were not always
analysed thematically to identify six core areas of service provision.
reported in the incident reports. The sparsity of the literature
These areas of service provision were used as the discussion
relating to many of the design challenges is of concern. Particular
framework at the manufacturer and operational workshops. The
areas needing further research are communication, hygiene and
data from the operational workshops were coded in two stages
the patient experience. For future research in this area we would
to allow for iterative analysis and further exploration of codes and
recommend that the literature search is widened to domains
themes. The coding by Roger Coleman/Merih Kunur resulted in
other than ambulance, for example primary, secondary and
two distinct design outputs for (1) design issues and (2) problems/
community care, and possibly other emergency services, e.g. fire
features. These codes were then scrutinised by Emma Crumpton,
and police.
resulting in the 1 codes. At this stage a detailed secondary
coding was conducted within the codes to identify nine higher
level codes and address duplication between codes (Emma
Crumpton/Sally Halls). These design challenges were further
checked against the primary coding by Sally Halls to confirm
inclusiveness.
1. Introduction
1.1 DRIVERS FOR CHANGE IN THE PROVISION OF In 2002/0, the Ambulance Service provided over 4.8 million
AMBULANCE SERVICES emergency responses using ,481 emergency ambulances
(Ambulance Service Association, 2006). Each accident and
In March 2005, the NPSA prioritisation panel strongly supported
emergency (A&E) ambulance costs in excess of £100,000 and has
a project on ambulance design to be taken forward in the coming
a service life of approximately five years. The current situation
year in response to concerns relating to the design of vehicles and
is that most NHS ambulance trusts produce their own vehicle
equipment that impact on patient safety.
specification; resulting in over 40 different designs. This presents
In June 2005, the DH set out a vision for the provision of future an increased risk to patient safety as the location of equipment
ambulance services by 2010 (2005). This included providing an and consumables, as well as interior layout, varies in each vehicle,
increasing range of quality mobile healthcare services for patients which impacts on safe systems of work and the efficiency of
with urgent and emergency care needs (e.g. ‘see and treat’). The clinical care.
overarching aims are that patients will receive improved care by
Vehicles used by the Ambulance Services have to be licensed by
consistently receiving the right response, first time, in time, and
the Driver and Vehicle Licensing Agency (DVLA) and have to meet
that more patients will be treated in the community, resulting in
a type approval requirement from the Vehicle Certification Agency
more effective and efficient use of NHS resources. It seems likely
(Vehicle Certification Agency, 2004). As part of this approval,
that these changes will require different vehicles and equipment for
the vehicles and equipment must comply with two specific
ambulance services.
European/British Standards (British Standards Institution, 2000a &
Previous research considered two aspects of patient safety in b). These standards provide a baseline for safety in the design of
ambulance design by looking at the clinical working environment emergency vehicles, but mostly relate to: electrical requirements;
when either taking the patient to the ambulance or the vehicle performance requirements; medical devices; fixation
ambulance to the patient. Ferreira and Hignett (2005) looked of the equipment in the patient’s compartment; emergency
at the efficiency of the patient compartment design for both exits; minimum seating dimensions; braking requirements; glass
frequency and safety-critical clinical tasks in the ambulance requirements; interior lighting; and sound (Vehicle Certification
(taking the patient to the ambulance). They found that the high Agency, 2005). Although patient safety is considered for crash
frequency tasks (e.g. oxygen administration and heart monitoring) protection (e.g. stretcher seat belt design), there are no guidelines
included 40% of high musculoskeletal risk postures for the to assist vehicle designers and ambulance fleet managers with
paramedics. Redden and Hignett (200) looked at responder clinical or workforce safety issues. As the range of services,
bag design for providing clinical care and treatment away from equipment, drugs etc. is likely to increase over the next five years,
the ambulance. A user trial for a chest pain response was used it is timely that the NPSA and Ambulance Service Association
to test the design and use of the bags. They found that the lack (ASA) have supported a design project to address these issues.
of standardisation introduced delays in treatment provision. An
ongoing research project on stretcher loading systems has elicited
the views of ambulance staff (managers and operational staff),
manufacturers and others on the importance of patient safety for
stretcher loading systems (Jones & Hignett, 2005). It was found
that patient and staff safety was the most important design issue,
ahead of manual handling, mechanical/electrical reliability, time
and infection control.
4
1.1.1 Aim
This first scoping study aims to investigate the developing models
of service provision and the short and longer term requirements
of vehicles and equipment that will be needed to address
the concerns of patient and staff safety in the future in the
Ambulance Service.
1.1.2 Objectives
The objectives were:
1. To review international literature on ambulance service
delivery systems and vehicle/equipment design;
2. To review the NRLS database for patient safety incidents
relating to ambulance services;
. To hold a strategic stakeholders meeting to discuss future
provision of ambulance services;
4. To hold two operational stakeholder meetings to investigate
problems and possible solutions;
5. To present models/scenarios of future ambulance transport
at the national ASA conference (AMBEX 2006) to a second
series of stakeholder meetings to provide a basis for an
adaptation strategy for existing vehicles and equipment;
6. To communicate the findings through a final report
submitted to the NPSA and ASA in order to provide a future
vision for a safe system of emergency transport.
5
2. Literature review
An extensive literature search was carried out using the following 2.2 ACCESS/EGRESS (LOADING/UNLOADING)
sources:
Two higher quality papers were found relating to access and
• Medline (1960–2005);
egress issues. The first (Petzäll, 1995) looked at the dimensions of
• EMBASE;
the entrances of cars used in a taxi service for disabled people. A
• CINAHL;
mock-up car was used with methods including video, observation
• Ergonomics Abstract;
and interviews. It was found that much the same dimensions
• personal collections;
were required for people confined to wheelchairs and ambulant
• grey literature.
disabled people, and ideal measurements were specified.
The key words ‘ambulance’ and ‘ambulance design’
The second paper (Boocock et al., 2000) called for redesign to
were used and the results are presented below. The
minimise the physical effort involved in loading and unloading
review is narrative rather than systematic, and so the inclusion/
stretchers. This paper examined the interface between the
exclusion criteria err towards inclusiveness for treatment and care
ambulance crew and the equipment from an ergonomic
in ambulances, at home and in non-healthcare premises.
perspective, in particular loading the trolley cot into the
The review is presented as brief descriptions of relevant papers ambulance. Biomechanical analyses of simulated tasks led to the
given under the nine design challenges which were derived
main recommendations relating to optimising operator posture
from the workshop data, incorporating the literature and the
and strength capabilities. Long term recommendations included
incident reports (see chapter , Method). More detail is given
the need for redesigned, lower vehicles to minimise physical effort
where papers are of a higher quality. Higher quality, for the
in loading and unloading trolley cots.
purpose of this review, is deemed to be those papers that are
not simply professional opinion, that is, some intervention has Other papers were found to include elements pertaining
been described. However the quality of the intervention and
to access and egress, but are listed under the other
methodology has not been critically appraised.
design challenges.
2.1 NINE DESIGN CHALLENGES A previous detailed project, funded by Engineering and Physical
Sciences Research Council (EPSRC) project grant no. GR/S56078/01,
The nine design challenges were as follows:
looked at ambulance loading systems in detail. The findings are
1. access/egress (loading/unloading);
summarised in the following section.
2. space/layout;
3. securing people and equipment in transit; 2.2.1 EPSRC project grant no. GR/S56078/01
4. communication;
5. security, violence and aggression; Loading/unloading of patients is generally achieved using a
mobility device such as a stretcher (emergency cot) that is loaded
6. hygiene;
into the patient compartment of the vehicle.
7. equipment;
8. vehicle engineering; There are three main types of loading systems: easi-loader, ramp
9. patient experience (safety, comfort and dignity). and winch, and tail lift. The easi-loader is the most commonly
found system internationally but in the European Union (EU) usage
has significantly decreased since 2000 following the introduction
of new EU standards (British Standards Institution 2000a & b). In
200 the Medical Devices Agency (Medical Devices Agency, 200)
reported a number of adverse incidents with easi-loader stretchers,
6
with over 50 relating to stretcher collapse and others including The human:machine interface was investigated using task
limb entrapment, mechanism jamming, component failure, analysis methods in the field and postural analysis methods in
unstable load, operator ergonomic mismatch and receiving the laboratory as shown in figure 1. The field-based methods
tray incompatibility. included critical incident technique (CIT), link analysis (LA) and
There has been relatively little international research looking at hierarchical task analysis (HTA). The laboratory-based methods
emergency care services (Levick & Mener, 2006) but this project included rapid entire body assessment (REBA), the NIOSH lifting
adds valuable data to a growing body of literature (Kluth & equation, kinematic analysis (KA) and vibrational analysis (VA).
Strasser, 2006) by analysing the human:machine interface and
Over 300 hours of field data were collected during 12-hour shifts
performance using a range of task analysis and postural analysis
tools. The data have been used to develop recommendations to with frontline A&E crews at the three participating ambulance
both improve the safety of current systems and specify the key trusts (East Midlands, East Anglia and Two Shires). Data were
design features for future systems. collected for loading and unloading patients from nine stations
selected by geographical location (rural, urban and intermediate
Figure 1. EPSRC project methodology
stations) at different times over nine months to observe the
three systems in a range of environments and climates. These
CIT data were analysed using LA and HTA. During shift breaks, staff
were interviewed about critical incidents. Video-recording during
observations was not permitted due to ethical constraints.
Field LA Taxonomy
Laboratory-based postural data were collected using multi-
directional filming and analysed with REBA and NIOSH. Force
HTA measurements (for KA) were taken with all systems for a stretcher
load of 75kg using a force handle designed to be as similar as
possible in geometry to the existing stretcher handles. Equations
Ranking to give were derived to calculate the force exerted by the ambulance
recommendations for: workers for each phase. The British Standards Institution
• technical specifications;
recommendations for force limits in design and use of ambulance
• safe systems of work.
stretchers were used as a comparator for the results (British
Standards Institution, 2000b).
KA
The data from the Critical Incidents Techniques (CIT), LA and
HTA were analysed independently before triangulation. The
REBA empirical data from the interviews were entered into NVivo
Postural for detailed analysis. Primary coding of the 10 CITs elicited five
Laboratory
analysis
first level generic codes across all three systems: system failure,
NIOSH
environment, patient-related, equipment, and coping strategies
and adaptation. These codes were used as the framework
VA taxonomy. A more detailed thematic analysis of the full interview
data produced 5 codes. These were separated into the three
systems and then further analysed to produce the first taxonomies
for the individual three stretcher loading systems. An HTA was
7
developed for each observed loading and unloading task (170 A total of 662 postures were analysed to give the average REBA
datasets). These were combined into 2 generic task summaries postural analysis score for each system. The easi-loader stretcher
indicating the type of issue and where it occurred. There were had the highest REBA score (8.1) followed by the tail lift (5.8) and
many more issues observed with the tail lift than with the easi- the ramp and winch (5.7). An analysis of variance (ANOVA) post
loader and ramp and winch. hoc test was carried out to examine the difference in REBA scores
between the different systems. It was found that the easi-loader
LA was used to highlight task complexity and redefine the layout
scored significantly higher than all of the other systems (p<0.05).
of the stretcher loading system. The LA data were mapped
The risk level using the easi-loader was high, with an action
individually for each loading and unloading task of the stretcher
category of three (action is necessary soon). For the other systems
loading system. The 170 datasets were then summarised into
the risk level was medium with an action category of two (action is
18 generic tasks. These data were then analysed to determine
necessary). NIOSH was used to validate the REBA analysis. Sixteen
the average number of links per task. The relative links for each
postures were analysed using both analysis tools, and a Friedman
system were analysed. The tail lift was a more complex task for
test concluded that there was no significant difference in the results
both loading and unloading a stretcher, with an average of 19.
(Lenton, 2005).
links for loading compared with 11.6 and 10.5 for the easi-
loader and hydraulic ramp respectively and 9.7 for unloading To assist with the development of design guidelines, a further
compared with 7.4 and .9 for the hydraulic ramp and easi- stage was added. Sixteen design issues were identified in the
loader respectively. Finally the LA and HTA data were added to the summary taxonomy. To define management priorities, a list of
taxonomy for each system with 16 design issues identified. 14 design issues (table 1) from the comparative findings were
used in a national ranking exercise (through the ASA and AMBEX)
KA models were developed for each phase of the task using free
with ambulance staff and manufacturers. Two issues were
body diagrams. Equations were derived to calculate the force
omitted from the questionnaire (stretcher/control location and
exerted by the paramedics to produce the observed movement
obstacles) as they related to the stretcher design/use rather than
for loading/unloading stretchers. The tail lift produced forces
the loading system.
(peaks around 100–200N) significantly less than the ramp and
winch (00–400N) and easi-loader (600N for one operator) The ranking positions for the remaining 14 factors are shown
systems. Force limits (British Standards Institution, 2002) were in table 1. It was found that ‘patient and operator safety’ and
exceeded in all the easi-loader and ramp loading and leading ‘manual handling’ were ranked as the most important issues for
measurements in some part of the system operation, suggesting loading and unloading patients. This was followed by system
that they present a risk of injury. design with respect to mechanical/electrical reliability. Three
factors received a very close ranking: time to operate, carry chair
VA data were collected using an accelerometer in three locations
access and vehicle layout. The tail lift was found to perform best
on the stretcher (head, chest and pelvis). The readings were
for patient and operator safety. There were manual handling risks
passed through a charge amplifier into a SigLab box and analysed
in all the systems, with more problems identified for the ramp
in MATLAB to produce acceleration time curves, converted to
and winch and the easi-loader systems. The easi-loader stretcher
frequency-time domain. The vibration level for the ramp and
created the most problems for patient and operator safety.
winch was considerably higher than the other two systems but all
were within the comfort level of a healthy person. No data were From the KA, three design changes were recommended to reduce:
collected to model the experience for an ‘unhealthy’ person. (1) stretcher mass; (2) maximum force application to stretcher
handle to less than 25°; and (3) wheel friction. The VA results
identified the ramp and winch system as the poorest performer.
8
Table 1. Design features The postural analysis found that the:
• tail lift had significantly lower forces for the lower
Average Ranking
Factor back (KA);
ranking position
• tail lift and ramp and winch had significantly lower risk than the
easi-loader (REBA).
Patient and operator safety 1.7 1
The vibrational analysis found that the:
Manual handling 2 • ramp and winch system had a significantly higher level than the
other systems.
Mechanical/electrical reliability 4.9
It was concluded that the tail lift is the best option currently
Time to operate 5.4 4= available but that there is scope for further future design
improvements as all systems still have risks. The tail lift operation
Carry chair access 5.4 4= needs to be simplified and the stretcher mass and wheel friction
need to be reduced.
Vehicle layout 5.7 6
2.3 SPACE/LAYOUT
Task complication 6.1 7
Several papers were found, including two higher quality papers
Weather/environment 8.0 8 about layout and general equipment issues and the effects of layout
on operator posture.
Clearance 8.1 9
Ferreira and Hignett (Ferreira & Hignett, 2005) reported on a review
of the layout of the patient compartment in a UK ambulance.
Effect of camber 8. 10
The methods used included link analysis and postural analysis.
Paramedics were observed over 16 shifts. The results showed that
Security 8.8 11=
24% of a typical shift was spent treating patients in the patient
compartment. An average of 29% of this was in a stationary
Infection control 8.8 11=
ambulance with the rest in transit. The most frequently occurring
Equipment misuse 9.0 1 clinical tasks were checking blood oxygen saturation, oxygen
administration, and monitoring the heart and blood pressure.
Need for sensors 11.0 14 The paramedics preferred to sit alongside the patient, so reach
distances were increased because the ambulance was designed
with the paramedic sitting at the head end for access to equipment
and consumables. High frequency tasks included 40% of postures
The HTA, LA and CIT analysis found that the:
requiring corrective measures. The study concluded that future
• tail lift performed best for patient and operator safety and
design should be based on ergonomic analysis of clinical activities.
had lowest manual handling risks.
Doormaal et al. (Doormaal et al., 1995) produced a more
equipment-based high quality paper, finding that harmful postures
are a feature of working in an ambulance. Tasks were assessed
using the Ovako Working posture Analysis System (OWAS) and
9
Description:ahead of manual handling, mechanical/electrical reliability, time and infection control A previous detailed project, funded by Engineering and Physical. Sciences . KA models were developed for each phase of the task using free . patients is a problem (Tintinalli & McCoy 1993; Corbett et al. 1998).