Journal of Hospital Infection (2007) 66, 232e236
www.elsevierhealth.com/journals/jhin
Post-discharge surgical site infection
surveillance by automated telephony
J. McNeish a,*, D. Lyle b, M. McCowan a, S. Emmerson a,
S. McAuley a, J. Reilly c
a
Infection Control Department, Golden Jubilee National Hospital, Glasgow G81 4HX, UK
Clinical Analyst, Golden Jubilee National Hospital, Glasgow, UK
c
Health Protection Scotland, Glasgow, UK
b
Received 5 April 2006; accepted 6 April 2007
Available online 4 June 2007
KEYWORDS
Surgical site infection;
Post-discharge
surveillance;
Automated telephony
system
Summary Surgical site infection (SSI) is an important outcome indicator
after surgery and is part of a national programme of surveillance in Scotland. Post-discharge surveillance has important cost implications for both
primary and acute areas of the NHS and it is therefore important to establish a robust method to obtain these data. This study used an automated
telephony system to gain information on SSI developing post-discharge.
The patients included were those who had inpatient surveillance carried
out following hip and knee replacements, and cardiac surgery. A recorded
message was used to ask the patient questions concerning the state of their
wound. These questions were answered by pressing numbers on their
phone. The study population was 104 and there were 18 patients who
thought they had a wound infection but after clarification only nine
patients suffered a post-discharge SSI. The number of patients who responded with at least one call to the system was 62. Although this pilot
study was carried out with small numbers, it was felt that with certain
refinements it should continue to be used for post-discharge surgical site
surveillance of infection.
ª 2007 The Hospital Infection Society. Published by Elsevier Ltd. All rights
reserved.
Introduction
* Corresponding author. Address: Infection Control Department, Golden Jubilee National Hospital, Beardmore Street,
Clydebank, Glasgow G81 4HX, UK. Tel.: þ44 0141 951 5441;
fax: þ44 0141 951 5042.
E-mail address: jane.mcneish@gjnh.scot.nhs.uk
Surgical site infection (SSI) is an important outcome
indicator after surgery. The costs for the patient,
hospital and community care team are considerable.1 The cost of SSI to the NHS in England has
0195-6701/$ - see front matter ª 2007 The Hospital Infection Society. Published by Elsevier Ltd. All rights reserved.
doi:10.1016/j.jhin.2007.04.003
Post-discharge surgical site infection surveillance
been estimated to be around £61 million per annum
with the costs of treating a single infection estimated to be around £3200.1
The mandatory requirement from the Departments of Health’s National Surveillance systems in
the UK has been established using a common
definition for SSI2 and only rates of SSI identified
in inpatients are published. This measure underestimates the true costs of SSI, as it is estimated
that around 70% of post-operative infections
present after discharge.3
Post-discharge surgical site infection surveillance presents a challenge to develop methods
for data capture as the patient is no longer under
direct medical supervision. A Department of Health
expert working group in the UK concluded that
post-discharge surveillance (PDS) was an important
consideration in SSI surveillance and that priority
procedures should be those with a short length of
stay.4 It also concluded that methods should be
sought where direct observation of the surgical
site was possible by a healthcare worker in the
normal patient pathway, to minimise additional
cost associated with surveillance activities.
National surveillance of SSI in Scotland is being
facilitated and coordinated by Health Protection
Scotland (HPS) within the Scottish Surveillance
of Healthcare Associated Infection Programme.
HDL(57)2001 required all acute NHS Health Boards
to collect data on SSI following two categories of
surgical procedure from a list of nine, one of which
should be orthopaedic. The protocol for this
programme encourages hospitals to develop
programmes of post-discharge surveillance of SSI
utilizing a method of direct observation of the
patient’s wound as this is recognised to be the
most reliable and valid way of detecting infection.3
Staff seeing the patient routinely in the course
of normal clinical care could carry out direct
observation of the wound after discharge. This is
easily achieved with procedures such as Caesarean
section and breast surgery following which community midwives and breast care nurse specialists
respectively routinely see the patients for the 30
days following surgery. However, patients having
other procedures, such as orthopaedic implant and
cardiac procedures, are not seen post-operatively
within 30 days and follow-up of all patients by
healthcare workers for surveillance purposes is
considered too costly.5 Alternative robust methods
of data capture need to be explored.
Recent literature on post-discharge surveillance
has focused on patients’ ability to detect SSI.
Postal questionnaires returned from the patient
to the hospital have been utilised but in recognition of the number of people in the UK who now
233
have telephones, a few studies have explored this
method as a possibility for collecting surveillance
data.6e11 A further development of these studies is
to test automated telephony for the purpose of
post-discharge surveillance.
Excelicare Direct is an interactive telephonebased, patient home-monitoring module linked
directly to the electronic patient record. By integrating telephony technology with the Excelicare electronic patient record, the breadth and
depth of the patient record is expanded to include
home episodes as well as primary, secondary and
tertiary encounters, making it suitable for home
monitoring, e.g. PDS, chronic disease management
and organization of a managed clinical network.
The use of Excelicare Direct in the patients’
homes empowers them to become more actively
involved in their own health and treatment,
allowing closer and more accurate monitoring
within their homes with minimal inconvenience,
cost and time expenditure incurred for both the
patient and the healthcare provider(s), whilst
improving the overall outcomes of treatment.
Patients access the system using their touch-tone
phone. Once identified (by the use of a personal
identification number, PIN) the system delivers
a customised script over the phone. Patients listen
and respond to script questions by entering the
corresponding keys on the touch phone. Each
touch-tone response from the patient is interpreted by the system and the next appropriate
reply or question is triggered and delivered to the
patient dynamically. Excelicare Direct allows the
recording of script questions and personalised
voicemail messages by the healthcare provider.
This study aimed to test the feasibility of automated telephony as an approach to post-discharge
surgical site infection surveillance. It also had
a number of further objectives, which were:
e to assess the practicalities of this approach;
e to evaluate whether sufficient information
could be obtained using this method;
e to analyse the cost implications of this
approach.
The Golden Jubilee National Hospital is the
National Waiting Time Centre for Scotland. A large
and increasing part of this work is for orthopaedic
and cardiac surgery. The patient population is
spread throughout all Scottish Health Board areas
and therefore are discharged and followed-up by
their own health board services. The hospital
participates in the national mandatory surveillance
of inpatient SSI surveillance for hip replacements,
knee replacements and cardiac operations. The
234
patients included in this study of post-discharge SSI
were those who had operations between 1 February
2005 and 4 March 2005, lived in Scotland, who had
touch-tone telephones and were able to use the
telephone.
The definitions used for diagnosis of SSI were
those of National Nosocomial Infection Surveillance system (NNIS, 1992) and are those used in
the Scottish Surveillance of Healthcare Associated
Infection Programme for inpatient surveillance.
The criteria for classifying an infection for the
telephone crib sheet were those proposed by the
Centers for Disease Control and Prevention (CDC)
and have been previously validated.2
Methods
During the inpatient stay, the project leader
interviewed all patients who met the study inclusion criteria, explained the aims of the pilot
study and obtained consent. The process was
explained and PINs were issued to allow access to
the automated telephony system (ATS) on a freephone number. In addition, the system required use
of the patient’s eight-digit date of birth. Patients
were asked to phone on three occasions, at days
10, 20 and 30 following their operation date. A credit
card-sized reminder with these dates was provided.
To assist patients prepare for their calls, they were
also issued with a printed copy of the recorded
questions. Patients were assured that if a problem
developed with their wound healing, this would be
identified by their answers. A member of the postdischarge team would return their call to ensure
that appropriate treatment had been given.
On each occasion that the patients used the ATS,
pre-recorded instructions were played, followed by
the questions. Replies were a choice between
‘yes’, ‘no’ or ‘not sure’ and were answered by
pressing ‘1’, ‘2’ or ‘3’ on their telephone.
Defined alerts were set-up within the automated
system if patients answered positively or were
unsure whether they had developed a wound infection. All these patients were contacted by telephone, a detailed clinical history was taken and
development of infection categorised in accordance
with CDC criteria by an infection control surveillance
nurse. The system has the facility to notify defined
users of the arrival of patient alerts on screen, as
well as by e-mail or pager. An electronic form was
devised to allow the infection control nurse to
document any communication with patients.
Data analysis was carried out and reports were
created by extracting information using the database. All data items entered either manually or via
J. McNeish et al.
the telephony system could be fully utilised for
audit purposes.
At the end of the surveillance period, patients
were asked to evaluate the user friendliness and
practicality of the system by completing a written
questionnaire. This consisted of 13 Likert score
questions with five graded possible answers from
‘strongly agree’ to ‘strongly disagree’ concerning
use of the system, perception of care, clarity of
the system, and patient’s preference.
Results
The pilot study of orthopaedic and cardiac patients
included a total population of 109 patients. Five
patients were excluded from the study as three
were unable to respond appropriately to commands,
one did not have access to a touch-tone telephone
and one patient was blind. Of the remaining 104
patients, 53 (51%) were male and 51 (49%) female.
Thirty-six percent of the patients had undergone total knee replacement surgery, 34% total hip
replacement, and one bilateral hip replacement.
There were also 27% with coronary artery bypass
grafts, and 2% valve replacements.
Sixty-two patients made at least one telephone
call to the ATS, giving a response rate of 60%. Of
these, 43 patients (69%) completed the three calls,
10 (16%) called twice and nine (14%) only called once.
Among orthopaedic patients (N ¼ 74), 61% used
the system, 44% calling three times. Fifty-seven
percent of cardiac patients (N ¼ 30) used the system, with 33% calling three times.
Eighteen patients (29%) self-reported ‘yes’ or
‘unsure’ to the question, ‘Do you think you have
developed a wound infection?’ However, following
clinical assessment by telephone, only nine (14.5%)
fulfilled the CDC criteria for development of infection. Of these, seven (16%) were orthopaedic
patients and two (12%) cardiac patients. All infections were superficial in nature with no deep
tissue or organ space involved.
Eleven patients answered positively when asked
if they were taking antibiotics. Of these, one
patient was referring to a prescribed cream and
one had answered positively in error. The remaining nine patients were those who had infections
confirmed by the study.
During the same period, inpatient SSI surveillance identified one infection (N ¼ 154, 0.6%).
Results of exit questionnaire
The response rate was 56% (N ¼ 104). Overall, 81%
responded positively to the system. Of those who
Post-discharge surgical site infection surveillance
responded (N ¼ 58), when asked if they would have
preferred the hospital to call them 91% responded
negatively; when asked if the telephone system
was simple to use, 95% of patients answered positively; 96% answered positively when asked
whether they appreciated that their concerns
were being addressed; and 88% responded positively when asked if they liked this way of keeping
in contact.
Discussion
This report describes a pilot study of the implementation of post-discharge SSI surveillance using
an ATS.
It was encouraging to find that 60% of patients
used the system. Other studies using telephone
surveillance found similar response rates.9 When
studies used postal questionnaires the response
was similar or at times higher.6,7 Reilly et al. found
a higher response rate by the active telephoning of
patients by a research nurse. The passive nature of
the automated telephony may affect the response
rate as the onus rests with the patient to initiate
the call.8
Most patients found the system easy to use.
However, two patients informed the project leader
that they had been unable to access the system. It
is unknown how many others may have had similar
difficulties.
Many patients were very positive at the interview stage and two apologies were received
on the exit questionnaires concerning omitting one
of the call dates. However, 40% of patients initially
agreed to carry out the calls but failed to do so.
Among these were two patients who, after agreeing to carry out the calls, refused to take the
paperwork home. This limited the reliability of the
deduced infection rates and it is not known
whether the infection rate among responders was
representative of the whole patient group.
Although information concerning treatment
with antibiotics was obtained, it may have been
helpful to enquire whether a swab was examined.
During the pilot study the team was able to obtain
this information when responding by telephone to
the alert generated by the computer system. This
information was then typed into a note within the
patient’s computer record.
The advantage of being able to contact patients
if any of their answers caused concern was
especially valuable when they thought they had,
or may have, developed a wound infection. Having
spoken to all these patients, any false positives
were eliminated. This supports the findings of
235
Whitby et al. that patients are able to recognise
the absence of infection.11 We therefore concentrated our resources on those replies that suggested a potential post-discharge infection.
A ‘G’ grade registered nurse led the pilot study.
Wide clinical experience was considered essential
to enable the nurse to assess the patients accurately and categorise them accordingly. It was also
considered that one individual should take responsibility for the whole project, giving consistency to the initial interviews, and ensured that
during follow-up telephone calls, patients spoke to
a nurse that they had met.
The total recurrent cost of the pilot study
included the cost of freephone telephone calls
from the patients (£31.80), the return calls when
required (£32.00) and the cost of the nurse’s time
(£828) giving a total cost of £891 for the 104
patients included in the surveillance, i.e. £8.56
per patient. The information technology infrastructure was in place and so no additional cost
was incurred. However, considerable capital
expenditure for the software was required which
the Board provided.
Due to the diffuse geographical location of the
hospital’s patients, it is not possible for them to be
monitored in person. This system enables the
hospital to comply with the recommendations of
HPS. This pilot study does, however, suffer from
a number of limitations as it was carried out in one
hospital and with small numbers of patients in
a limited number of specialties. The use of the
system on a larger scale would have cost implications, as additional user licences would have
to be purchased. Furthermore, the number of
infections identified is a minimum estimate as
non-responding patients were not included in the
infection rate calculations.
Conclusions
This pilot study indicates that automated telephony is a feasible approach for post-discharge SSI
surveillance. However, the system requires formal
validation by further research. Once validated, the
system would have the potential to be extended
and could provide a means of standardizing postdischarge surgical site infection surveillance.
References
1. Plowman R, Graves N, Griffin MA, et al. The rate and cost of
hospital-acquired infections occurring in patients admitted
236
2.
3.
4.
5.
to selected specialties of a district general hospital in
England and the national burden imposed. J Hosp Infect
2001;47:198e209.
Mangram AJ, Horan TC, Pearson ML, Silver LC, Jarvis WR,
the Hospital Infection Control Practice Advisory Committee.
Guidelines for prevention of surgical site infection. Infect
Control Hosp Epidemiol 1999;20:247e278.
Bruce J, Russell EM, Mollison J, Krukowski ZH. The quality of
measurement of surgical wound infection as the basis for
monitoring: a systematic review. J Hosp Infect 2001;49:
99e108.
Reilly J, Ward V, Crawshaw S, et al. A UK approach to
post discharge surveillance: exploring the decision making
process. In: Poster presented at 5th international conference of the Hospital Infection Society, September 2002,
London.
Glenister HM, Taylor LJ, Bartlett CL, et al. An evaluation of
surveillance methods for detecting infections in hospital
inpatients. J Hosp Infect 1993;23:229e242.
J. McNeish et al.
6. Manian FA, Meyer L. Comparison of patient telephone
survey with traditional surveillance and monthly physician
questionnaires in monitoring surgical wound infections.
Infect Control Hosp Epidemiol 1993;14:216e218.
7. Noel I, Hollyoak V, Galloway A. A survey of the incidence
and care of postoperative wound infections in the community. J Hosp Infect 1997;36:267e273.
8. Reilly J, Noone A, Clift L, et al. A study of telephone screening and direct observation of surgical wound infections after
discharge from hospital. J Bone Joint Surg 2005;7:997e999.
9. Taylor EW, Duffy K, Lee K, et al. Telephone call contact
for post-discharge surveillance of surgical site infections.
A pilot, methodological study. J Hosp Infect 2003;55:8e13.
10. Fanning C, Johnston BL, MacDonald S, LeFort-Jost S,
Dockerty E. Postdischarge surgical site infection surveillance. Can J Infect Control 1995;10:75e79.
11. Whitby M, McLaws M-L, Collopy B, et al. Post-discharge surveillance: can patients reliably diagnose surgical wound
infections? J Hosp Infect 2002;52:155e160.