You are on page 1of 4

Occupational Medicine 2003;53:147–150

DOI: 10.1093/occmed/kqg032


Investigation of eye splash and needlestick

incidents from an HIV-positive donor on an
intensive care unit using root cause analysis
L. Batty, K. Holland-Elliott and D. Rosenfeld

Background Two doctors working on a busy intensive care unit sustained injuries whilst removing
a chest drain from an HIV-positive patient. One doctor had a needlestick injury into
his finger whilst the other sustained an eyesplash when the chest drain was pulled
Methods Following Department of Health format 'Doing less harm', a root cause and human
factor analysis of the incident was carried out. The aim was to explore the underlying
Results and Training, cultural and organizational issues were exposed, and are now being
conclusions addressed. This approach has led to a far more effective dialogue with the National
Health Trust concerned than was previously experienced, and there is early evidence
of progress on important aspects of health and safety management at organizational
level. Lack of health and safety training of doctors at undergraduate and post-
graduate level needs to be addressed.
Key words Accident investigation; eye splash; HIV; human factors; intensive care; needlestick;
occupational injury; root cause analysis.
Received 25 June 2002
Revised 3 December 2002
Accepted 7 January 2003

King’s College Hospital Occupational Health and Safety ‘incident form’ and later to the occupational health and
Department provides services and advice to many safety department for follow-up. The incident was
different National Health Service (NHS), public sector assessed as being notifiable to the Health & Safety
and private sector organizations. In April 2002, it was Executive (HSE), and the appropriate form F2508 sent
alerted to an incident on an intensive care unit (ICU) by the Trust health and safety advisor in accordance
in an acute NHS Trust in London. Two doctors were with the Reporting of Injuries, Diseases and Dangerous
engaged in removing a chest drain from an HIV-positive Occurrences Regulations (RIDDOR) 1995. The head of
patient during which one doctor received a splash of nursing for the area had performed an initial investigation
pleural fluid to the eye and the other doctor sustained a into the facts and identified that the doctor who suffered
needlestick injury to the thumb. Both doctors were seen the eye splash had not worn eye protection. Upon receiv-
at the time in the hospital accident and emergency ing the RIDDOR report, the HSE requested a formal
department, assessed and treated in accordance with investigation report.
Department of Health guidelines [1]. It was decided that a root cause analysis approach to
In accordance with that hospital’s procedure, the the accident investigation was appropriate. This format
incident was reported to management on a specific follows the Department of Health guidance issued in
2001, which encourages a systematic analysis of the
Department of Occupational Health and Safety, Kings College Hospital, underlying causes of incidents [2]. Further analysis seek-
Denmark Hill, London SE5 9RS, UK.
ing root causes was started using a human factors analysis
Correspondence to: Dr L. Batty, Department of Occupational Health and
Safety, Kings College Hospital, Denmark Hill, London SE5 9RS, UK.
[3] (Figure 1) that was later expanded to a causal chain
e-mail: diagram (Figure 2) in order to gain as full an under-

Occupational Medicine, Vol. 53 No. 2

© Society of Occupational Medicine; all rights reserved 147

Figure 1. Protective eyewear—human factors.

Figure 2. No eye splash protection—root causes.

standing as possible of the underlying causes of the protection against blood-borne viruses were present at
incident. It has long been understood that workplace the Trust level. Written guidance on safety procedures was
safety behaviour is the result of different influences within also available. The initial report indicated that there was
an organization [3], and so the context of the work, the discretional use of eyewear. It was accepted that constant
doctors’ training arrangements and working practices, use of eyewear for clinicians in intensive care is neither
and the culture in that workplace were specifically practicable nor necessary. Protection against exposure to
examined. blood-borne viruses therefore relies on the use of eyewear
in specific circumstances. The decision whether or not to
wear eyewear is delegated to the individual doctor. This
Investigation approach decision would be reliant on the risk perception of the
Appropriate policies regarding health and safety as well as individual doctor, which in turn would be affected by a

number of factors. These factors could include any they had not. Direct local learning from reported safety
training the doctors received, together with workplace incidents of any kind appeared to be uncommon. Sharing
reinforcement by colleagues and supervisors of safe incidents informally was not encouraged and doctors
systems of work. Two perspectives were applied to this perceived stigmatization if they did.
decision-making process by the doctor [4]. The next area of attention was the nature of the
First, human factors analysis divides failures into two supervision. The junior doctor’s line of supervision for
groups: (i) error and (ii) deliberate action of the individual safety purposes appeared confusing. It was unclear to all
(Figure 1). In this case, it was clear that there was a delib- where the doctors were based with regard to the health
erate decision not to use personal protective equipment and safety management structure, and whether the
(PPE) in the form of eye protection. On further interview, doctors involved in the incident were counted in the ICU
it became clear that the use of eyewear was routinely safety management returns. The doctor’s appraisals were
ignored, based on the belief that no significant risk factor carried out by the consultant but concentrated on aca-
applied. The second method of analysis was to work back demic assessment. Supervision lacked a personal safety
to the underlying causes by laying out diagrammatically orientation. The investigators reviewed knowledge of and
a chain of causes leading to the event. The HSE has access to the written safety procedures. It was accepted
identified that the general safety behaviour of colleagues that, by the nature of ICU doctors’ work, few daily tasks
and supervisors can be a significant underlying cause of involved even momentary referral to written procedures,
incidents [5]. The aim of causal chain analysis (Figure 2) even though these were available. The doctors appeared
was to identify immediate causes and underlying causes. unaware of the infection control policies and the module
The underlying causes can then be subdivided into that was available throughout the Trust intranet system.
general and management causes of failure to use eye There was information available on the use of gloves and
protection. eye protection within the ICU. The ICU also provided
written posters, guidance policies on the unit and access
to eye protection. It was specifically noted that ICU safety
arrangements were generally clearly written in terms
Analysis of immediate causes
of both policy and protocol. There was a blood-borne
The doctors were interviewed about their health and virus checklist, which was reviewed regularly by unit
safety training. At undergraduate level, they felt that any managers. The checklist had sections on risk assessment,
advice about splash protection and personal safety information to staff, provision of equipment and facilities,
training was obscure and not emphasized. They recalled periodic recorded checks on staff practice, and follow-up
being given warnings and advice regarding prevention of of each incident. Any negative answer to that manage-
needlesticks and eye splashes by a senior doctor when ment checklist developed an action plan. The provision
they were inducted into the employing Trust. The and maintenance of PPE were specifically allocated to an
personal presentation was recalled, though they did not ICU technician as a lead responsibility, and as a result a
refer to written materials that were also provided. wide range of protective equipment was readily available
Within a crowded corporate induction programme, and properly maintained.
there is reliance upon some safety information delivered
locally. Both doctors had ‘a guided tour’ of the clinical
area, but by a non-clinical staff member, so personal
safety precautions were not discussed. This briefing also Analysis of underlying causes
omitted information about local availability of written or The investigators went on to consider underlying causes.
computer-based policies and procedures, such as infec- General causes were considered first. The investigators
tion control. In contrast, nursing colleagues working in found very little consideration of personal safety by the
the same area received a longer briefing from their nurse doctors. This appeared to reflect a wider medical culture
manager and completed a mandatory induction checklist. in which health and safety has a low priority in training
These checklists confirmed that information on manage- [6], with the patient’s safety coming first. The doctors
ment responsibilities, location of safety equipment, and reported that not only did they not routinely wear eye
other important clinical and safety information was given. protection, but neither did their medical colleagues. This
Interview highlighted that where a medical speciality was noted to be in direct contrast to the nurses, who were
team has patients on different wards, there may be no trained and inducted as a separate professional group and
local unit briefing for those doctors. The doctors referred who did wear eye protection in similar circumstances.
to ‘on-the-job training’ as an important method of learn- The investigators noted that role-model doctors were
ing. They observe procedures before practising. The apparently not using health and safety equipment
doctor who showed them how to remove chest drains had when teaching. This would potentially reinforce non-
not worn eye protection. They were asked whether they compliance and weaken attempts to persuade more
had heard of any similar incidents and confirmed that junior medical staff to adopt good safety practice.

The investigators then moved into examining possible cations relating to the safety culture of doctors and that its
management causes using the National Patient Safety basis in undergraduate and postgraduate teaching war-
Agency hierarchy of cause categories [7]. Institutional rants further attention. The investigators understand that
influences were considered first. The investigators noted various bodies, including the Health Services Advisory
that, in some organizations, senior managers have safety Committee to the Health and Safety Commission [9],
objectives as part of their appraisal process. It was noted have attempted to encourage the further training and
that the NHS consultant’s appraisal handbook that was engagement of doctors in health and safety. Training
implemented in that Trust made little or no reference to modules for students have been developed to persuade
personal safety and related health and safety issues. It was trainee doctors to think and act more proactively in
noted that doctors’ training is managed separately from considering their own and others’ safety [10], but these
that of other staff, and frequently involves rotation appear not to be widely implemented yet. The General
through different work areas. Doctors in training were Medical Council has a role to play, and this is being
unclear as to who their line manager was, as opposed to developed with ‘tomorrow’s doctors’, though they may be
the consultant as academic supervisor. Therefore, the in a position to do more [11]. The investigators would be
individual roles and responsibilities for the health and keen to hear from other occupational health and safety
safety of medical staff were also unclear in practice. As a departments that have performed similar analysis to see
result, in contrast to the hierarchical nursing structure, whether their incident findings were similar, and to see
direction to health and safety compliance and appropriate whether other Trusts have found this approach as useful.
training was absent or reduced.
Although infection control policies identifying the
range of risks and precautions necessary for safe practice References
were present, the lack of examples of specific at-risk 1. Department of Health. Guidance from UK Chief Medical
clinical procedures requiring goggles was considered an officers’ Expert Advisory Group on AIDS: HIV Post-exposure
underlying factor in this case. The doctors were unclear Prophylaxis. July 2000; 8/2.1.
when they had to wear goggles, but the clear perception 2. Department of Health National Patient Safety Agency.
was that their department compared no less favourably Doing Less Harm, version 1.0a. August 2001. http://www.
than other departments in the hospital, or other hospitals
3. HSE. Reducing Error and Influencing Behaviour, 2nd edn.
they had worked in. Supervision and learning by example
1999. ISBN 0717624528. HSG 48.
were also identified as possible underlying causes, as
uk/hid/land/comah/level3/5c72542.htm; http://www.mike.
discussed earlier. It was interesting to note that although
eye protection was not used, protective gloves and aprons 4. Reason J. Human error: models and management.
were used when they removed the chest drain ‘because Br Med J 2000;320:768–770.
they knew the patient was HIV positive’. This reasoning is content/full/320/7237/768
contrary to infection control policy, as well as Depart- 5. Jeffe DB, Mutha S, Kim LE. Does clinical experience affect
ment of Health guidelines, in that precautions should be medical students’ knowledge, attitudes and compliance
used universally [8]. with universal precautions? Infect Control Hosp Epidemiol
6. Vincent CH, Taylor-Adams S, Stanhope N. Framework for
Discussion analysing risk and safety in clinical medicine. Br Med J
The Trust in question has an annual incidence of 1000 full/316/7138/1154/Fu5
needlestick injuries and eye splash incidents per annum 7. Department of Health National Patient Safety Agency.
reported on its staff survey. Approximately 200 of these Reporting Incidents.
report to the occupational health department per year. reporting.asp
The background rate of HIV infection in the population 8. Department of Health. Guidance for Clinical Health Care
that the hospital serves is unknown, but is likely to be Workers: Protection against Infection with Blood-borne
higher in London than in many other parts of the UK. Of Viruses, 2.19:8; 3.4:9.
more concern is the unknown background prevalence of doh/chcguid1.pdf
more infective agents, such as hepatitis viruses. Using this 9. Davies R. Health and Safety Commission. Heath Services
Advisory Committee. Ensuring Doctors are Engaged in
approach has enabled the occupational health and safety
Health and Safety.
department to progress preventative strategies within the 10. Agius R, Smith I. Risk Assessment and Risk Reduction in
Trust in a way that the initial factual report did not facili- Occupational Medicine and Environmental Health—Vision
tate. In particular, the local consultants identified that 2000 Year3.
there were opportunities for improved safety training and year3.html
they were developing these as a result of this investigation. 11. Henderson M, Holland-Elliott K. Good medical practice
The investigators believe that there are wider impli- for occupational health. Lancet 2002;359:895.