Care expert assistant for Medicare system using Machine learning
Articulo de telemedicina en urgencia
1. Journal of Plastic, Reconstructive & Aesthetic Surgery (2008) 61, 31e36
Telemedicine for acute plastic surgical
trauma and burns
D.L. Wallace a,b,d, S.M. Jones a,c,e, C. Milroy a, M.A. Pickford a,*
a
Plastic Surgery Department, Queen Victoria Hospital, Holtye Road,
East Grinstead, West Sussex RH19 3DZ, UK
Received 20 December 2005; accepted 17 March 2006
KEYWORDS Summary Visual images can enhance communication over a distance. In the UK,
Telemedicine; plastic surgery provides services over large distances by a ‘hub and spoke’ model.
Store-and-forward; Telemedicine could help to increase the efficiency of service for plastic surgery pa-
Digital; tients. Telemedicine, along with the impending Electronic Patient Record system
Trauma; could combine to improve communication, patient triage, record keeping, audit
Triage and could lead to a better quality of clinical care. Another benefit could be signif-
icant cost savings.
We report our experience of the introduction of telemedicine to a Regional Plas-
tic Surgery Service. Our first study compared assessments from images and patient
examinations, which gave us confidence in the use of images [Jones SM, Milroy C,
Pickford MA. Telemedicine in acute plastic surgical trauma and burns. Ann R Coll
Surg Engl 2004;86:239e42]. We proceeded to a 10-week evaluation of all 973 refer-
rals to our unit. We found that the system was used for a wide variety of injuries
and for 42% of the 452 patients where the system was available. Initial resistance
was overcome by the ease of use of the system, with both receiving and referring
clinicians reporting benefits. The third phase was a 12-week prospective cohort
study of 996 patients comparing the referrals with and without the telemedicine
system. The system was available for 389 patients, and used for 243 patients
(63%). The groups were analysed by a chi squared test and confidence interval cal-
culation. We demonstrated a significant difference in the initial management of pa-
tients, with 10% more being booked directly to our Day Surgery Unit. There was
a decrease in number of occasions when we were unable to accept a patient due
* Corresponding author. Tel.: þ44 1342 414035; fax: þ44 1342 414121.
E-mail addresses: davidandvix@aol.com (D.L. Wallace), mark.pickford@qvh.nhs.uk (M.A. Pickford).
b
DL Wallace presented a part of the submission at the 3rd Annual International Conference of Telecare and Telehealth in Brisbane,
Australia, August 2003.
c
SM Jones presented a part of the submission at the British Association of Plastic Surgeons Winter Meeting, December 2001.
d
Present address: 21 Huddesford Drive, Balsall Common, Coventry CV7 7RR, UK
e
Present address: Salisbury District Hospital, Odstock Road, Salisbury, Wiltshire SP2 8BJ, UK
1748-6815/$ - see front matter ª 2006 British Association of Plastic, Reconstructive and Aesthetic Surgeons. Published by Elsevier Ltd. All rights reserved.
doi:10.1016/j.bjps.2006.03.045
2. 32 D.L. Wallace et al.
to a lack of capacity. We found no change in the patients being managed with tele-
phone only advice.
We found that telemedicine is a valuable method of providing useful preliminary
information in the referral process for injured patients and often significantly mod-
ifies their treatment and/or management plan. This has implications for the use of
Information Technology resources and potentially the delivery of healthcare in
relation to the management of injured patients.
ª 2006 British Association of Plastic, Reconstructive and Aesthetic Surgeons.
Published by Elsevier Ltd. All rights reserved.
The simplest definition of telemedicine is ‘the Information Technology experts, Consultants and
practise of medicine at a distance’.2 There are Junior Staff of the surgical specialties based at the
two main types of visual telemedicine; store-and- QVH (Plastic Surgery and Maxillofacial Surgery).
forward or real-time videoconferencing systems. Specialist software was commissioned (DISTAR
Videoconferencing allows synchronous consulta- Telemedicine, Digital Image ST-orage And Re-
tions that report high patient and clinician satis- trieval, and TM Clientª) to provide secure and
faction rates,3 and is successfully used in many technically simple data acquisition and transfer.
Minor Injury Units to provide a real-time link to The software consists of five simple steps for im-
a larger Accident and Emergency Department.4 age location, selection of the correct images with
However, real-time technology can be expensive a small preview check, demographic data entry,
to install and maintain, the quality may be vari- recording the level of consent and transmission
able, the software is complex and often restric- of the encrypted email. The average file size was
tive, and the clinician requires technical skills 700 KB, for 24-bit colour images at a resolution of
and usually a designated room. Store-and-forward 1600 Â 1200 pixels. This was manageable for the
systems provide high resolution images but are hospital networks and allowed rapid transmission
asynchronous; they have been used for many years over the Internet. The encrypted email could be
in specialties such as Dermatology and Radiology. sent from other hospital sites (external) or net-
The agreement in diagnosis between face-to-face worked computers within the QVH (internal).
consultations and real-time is 51e80% and for Upon arrival at the QVH server (Pentium 4 256MB
store-and-forward systems 70e95%.5 RAM and RAID-1 mirrored 14 GB disks) emails
The Queen Victoria Hospital (QVH) provides were decrypted and were available for viewing at
plastic surgical services to the South East of England any networked computer. The hospital computers
(population 4.5 million). Approximately 450 trauma operated Windows NT and were a minimum of Pen-
referrals are received each month from 24 different tium II, 64 RAM; but most commonly used were
NHS Trusts in the region. Audit in our department Pentium III, 128 RAM. Images were viewed on
has shown that referrals have doubled over the last a screen resolution of 800 Â 600 pixels and re-
five years and are still increasing. Capacity to treat corded onto a database, which was backed up
hand injuries in our geographical region is under daily.
considerable strain with 17% of units being unable to The referring clinicians obtained consent for
accept referrals.6 We were keen to assess the po- photographs from their patients and emailed
tential of telemedicine to improve the triage and digital images via the NHS intranet, in addition to
management of trauma referrals to our unit. the usual telephone referral. A plastic surgery
The choice of telemedicine system is best made trainee (SMJ) was employed to drive the project
by the clinical users.7 We chose a store-and-for- in the early stages. Over a six-month period the
ward system because high resolution images could system was introduced to 10 hospitals that
be easily and cheaply accessed at any networked regularly referred injuries to QVH. During this six-
hospital computer, even in theatre while surgeons month period we undertook a 10-week retrospec-
are scrubbed. tive evaluation study. The Telemedicine team
reviewed the initial results, and the installation
of the telemedicine software and education
Methods and materials programme in referring units was continued.
A subsequent prospective 12-week trial was
A Telemedicine team was created consisting of carried out to investigate changes in patient
nursing staff, medical illustrators, audit personnel, management from the telemedicine assisted
3. Telemedicine for acute plastic surgical trauma and burns 33
referrals compared to telephone only referrals.
From 01 March 2003 to 23 May 2003 the receiving
clinician recorded the demographic and clinical
details of all trauma referrals. Statistical analysis
was performed with the chi squared test and
confidence interval calculation.
Results
Ten-week retrospective evaluation of the
telemedicine system
We received 973 referrals over 10 weeks from 53
different sites. There were 644 male and 329 Figure 2 Radiograph of a crushed finger fracture.
female patients; 730 adults and 174 children. A
wide variety of injuries were assessed (Figs. 1e7).
Use of the system was seen to steadily increase A significant difference was noted (P ¼ 0.004) in
over the initial six-month period (Fig. 8). From the management of patients with and without the
the sites with the telemedicine installed there availability of the telemedicine system (Table 1).
were 452 patients referred; telemedicine was Significantly fewer patients needed to come for
used for 42% of these patients. After initial resis- further assessment or review and more patients
tance, referring clinicians were generally pleased could be directly booked for definitive care on an
with the introduction of the system, finding it operating list in our Day Surgery Unit (10.5%)
easy to use and helpful in the referral process. Re- when the telemedicine system was available. We
ceiving clinicians also expressed an improvement also observed a decrease in number of occasions
in the clarity of the information. when the QVH was unable to accept a referral
due to a lack of capacity. There was no increase
Twelve-week prospective cohort study or decrease in patients being managed with only
telephone advice, nor for patients admitted to
Over 12 weeks 996 referrals were received from their local hospital to await transfer to the QVH
over 60 different sites. There were 607 male and (outlier).
389 female referrals; 725 were adults and 271
were children. The telemedicine system was avail-
able for 389 referrals, and used for 243 referrals Discussion
(63%). Analysis was undertaken of telemedicine
system availability, not only when used, to mini- Telemedicine is a rapidly growing tool in modern
mise bias from the referring clinician’s choice. medicine, giving the promise of efficiency and
250
With Telemedicine
200 Without Telemedicine
Number of referrals
150
100
50
0
ry ury s ip rn n
tur
e n
nju inj los ert Bu ec
tio tio
ni d kin ing Inf Fr
ac
lanta
lea n ate e/s d/F ep
C
mi su ilb
e R
nta Tis Na
Co
Injury Type
Figure 1 The distribution of injuries and the use of
telemedicine during the 10-week evaluation study. Figure 3 Clinical image of finger in Fig. 2.
4. 34 D.L. Wallace et al.
Figure 4 Flame burn to fingers.
better communication between clinicians.8 Tele-
medicine systems have proved useful in reducing
unnecessary transfers in Neurosurgical emergen-
cies,9 and in reducing mortality, complications
and costs in Intensive Care.10 In the UK the na-
tional telemedicine database has 260 projects cur-
rently listed,11 though only a minority have been
reported moving into routine or mainstream use,
even with successful pilots. There are numerous
Figure 6 Paediatric mixed depth burn.
barriers to the implementation of telemedicine
including limited financial support or reimburse-
ment,12,13 legal uncertainties14 and an inadequate adequate, technical difficulties in the viewing of
evidence base.15 images led us to commission specialist software.
Image resolution in telemedicine is crucial. On This facilitated the ease of loading, sending, re-
viewing a series of images of different resolutions viewing, record keeping and security of images.
clinicians can lose confidence in their diagnosis There have been feasibility studies assessing the
before they realise that the image resolution has role of telemedicine for wound assessment,19 burn
changed.16 Our initial assessment demonstrated injuries,18,20 fractures21 and replantations.22
that 800 Â 600 pixels provided good enough resolu- There are no previous studies demonstrating supe-
tion for diagnosis,1 confirming a resolution level riority to the traditional telephone referral for
shown to be effective for dermatology17 and burn acute plastic surgery, nor reporting the routine
injuries.18 Although this image resolution was
Figure 5 Flank and buttock post-debridement of
necrotising fascitis. Figure 7 Pre-tibial laceration.
5. Telemedicine for acute plastic surgical trauma and burns 35
180
160
140
120
100
80
60
40
20
0
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun Jul
January 2002 to July 2003
Figure 8 The use of telemedicine at the Queen Victoria Hospital since the introduction of the specialist software TM
Clientª in January 2002.
use of telemedicine in the usual referral process. the receiving clinician with greater confidence in
Our initial results confirmed the feasibility of the planning the delivery of the surgical service, e.g.
telemedicine system for assessing all types of to book directly to an urgent day case appoint-
acute plastic surgical referrals, with both referring ment, a next day clinic review or organise an
and receiving clinicians reporting benefits.1 immediate transfer.
Our results demonstrated significant differences In addition to the improved clarity of commu-
in the management of referrals when the tele- nication we also observed improved access, earlier
medicine system was used (Table 1). Triage of and more frequent senior clinician involvement.
referrals was improved with greater use of day sur- We believe this is because the telemedicine
gery and a decreased need for face-to-face system images provide enough information to
assessment. An audit of London region Hand sur- make some clinical decisions immediately and
gery units (including the QVH) showed that nine involve senior clinicians prior to the face-to-face
departments are working so near to maximum consultation.
capacity that on 17% of occasions they are unable The evidence for direct cost savings to the
to accept referrals.6 We observed a decrease in providers of telemedicine is poor.23 Only in a few
the frequency with which patients were not prison and military populations where the patient
accepted at the QVH with the use of telemedicine. transportation costs are high is there evidence of
The changes observed could be because the cost savings.24 We found no evidence of cost sav-
telemedicine image provided more objective in- ings for the QVH trust, and only anecdotal patient
formation to the receiving clinician compared to cost savings. The capital outlay was significant
telephone referrals alone. This may have allowed (£70K) for the installation of the computer
Table 1 Difference in the management of referrals with and without telemedicine availability
Telemedicine assisted referral Telephone only referral
% (n ¼ 389) Confidence intervals % (n ¼ 607) Confidence intervals
Admit 29.6 (25.2e34.3) 28.3 (24.9e32.0)
Review 15.4 (12.2e19.3) 22.1 (19.0e25.5)
Dressing clinic 0.5 (0.1e1.9) 1.2 (0.6e2.4)
Day surgery 27.5 (23.3e32.1) 17 (14.2e20.2)
Full e unable 8.7 (6.3e12.0) 12.5 (10.1e15.4)
to accept
Outpatients 7.7 (5.5e10.8) 8.1 (6.2e10.5)
Outlier 4.6 (2.9e7.2) 4.4 (3.1e6.4)
Telephone advice 5.7 (3.8e8.4) 5.9 (4.3e8.1)
Inappropriate referral 0.3 (0.0e1.4) 0.5 (0.2e1.4)
Total 100 100
6. 36 D.L. Wallace et al.
network lines, equipment and software. Since in- 3. Mair F, Whitten P. Systematic review of studies of patient
ception we have had a single plastic surgery satisfaction with telemedicine. BMJ 2000;320:1517e20.
4. Benger J. A review of minor injuries telemedicine. J Tel-
trainee responsible for the project, which has fa- emed Telecare 1999;5(Suppl. 3):S5e13.
cilitated the coordination of the many clinicians 5. Whitten P. Teledermatology delivery modalities: real time ver-
and IT personnel using the system at numerous sus store and forward. Curr Probl Dermatol 2003;32:24e31.
sites.7 This telemedicine clinician also ensures sys- 6. Skillman JM, et al. Audit of pattern of closures to acute
tem availability, guides maintenance and up- hand services in Pan Thames area. J Hand Surg [Br] 2003;
28:381e3.
grades, and continues the training of staff. 7. Yellowlees P. Successful development of telemedicine sys-
The QVH trust developed a ‘Patient Photo- tems e seven core principles. J Telemed Telecare 1997;3:
graphic and Video Recording Policy’; to ensure 215e22.
images recorded complied with the Data Protec- 8. Wyatt JC, Keen J. The new NHS information technology
tion Acts (1984 and 1998) and the Computer Misuse strategy. Technology will change practice. BMJ 2001;322:
1378e9.
Act 1990. This ensured clinicians had clear and 9. Goh KY, Lam CK, Poon WS. The impact of teleradiology on
legally robust guidelines on digital imaging.25 The the inter-hospital transfer of neurosurgical patients. Br J
telemedicine system was designed with a simple Neurosurg 1997;11:52e6.
tick box page to confirm consent from the patient 10. Rosenfeld BA, et al. Intensive care unit telemedicine: alter-
without the need for extra paperwork. There is an nate paradigm for providing continuous intensivist care.
Crit Care Med 2000;28:3925e31.
audit trail for each image. Confidential Patient In- 11. Telemedicine Information Service <http://www.tis.port.ac.
formation can be protected by either using anony- uk/tm/owa/projects.allUK>. Internet [28.11.2004, 1.12.2004].
mous data, an approach adopted by the British 12. Cabana MD, et al. Why don’t physicians follow clinical prac-
army,26 or by the use of encryption techniques. tice guidelines? A framework for improvement. JAMA 1999;
We use publiceprivate key encryption to allow se- 282:1458e65.
13. Finch TL, May CR, Mair FS, et al. Integrating service devel-
cure information transfer. Anonymous data are opment with evaluation in telehealthcare: an ethnographic
inappropriate for a network with high volumes of study. BMJ 2003;327:1205e8.
similar referrals, as the risk of confusion is likely 14. Brahams D. The medicolegal implications of teleconsulting
to be unacceptably high. in the UK. J Telemed Telecare 1995;1:196e201.
The introduction of telemedicine has led to 15. Hailey D, Roine R, Ohinmaa A. Systematic review of evi-
dence for the benefits of telemedicine. J Telemed Telecare
proven benefits in the management of acute plastic 2002;8(Suppl. 1):1e30.
surgery referrals at the QVH. This was achieved by 16. Briggs R, Bailey JE, Eddy C, et al. A methodologic issue for
using specially designed software, with integral IT ophthalmic telemedicine: image quality and its effect on
support. Telemedicine may be an important com- diagnostic accuracy and confidence. J Am Optom Assoc
ponent of the Electronic Patient Record, which is 1998;69:601e5.
17. Bittorf A, Fartasch M, Schuler G, et al. Resolution require-
due for implementation within the next few years. ments for digital images in dermatology. J Am Acad Derma-
tol 1997;37:195e8.
18. Jones OC, Wilson DI, Andrews S. The reliability of digital im-
Acknowledgements ages when used to assess burn wounds. J Telemed Telecare
2003;9:22e4.
19. Houghton PE, Kincaid CB, Campbell KE, et al. Photographic
We would like to thank S. Reynolds of Godalming
assessment of the appearance of chronic pressure and leg
Computer Products for the development of the ulcers. Ostomy Wound Manage 2000;46:20e30.
telemedicine software and the Information Tech- 20. Roa L, Gomez-Cia T, Acha B, et al. Digital imaging in remote
nology Department at the QVH for their knowledge diagnosis of burns. Burns 1999;25:617e23.
and ongoing support. We also thank M. Hankins at 21. Ricci WM, Borrelli J. Teleradiology in orthopaedic surgery:
impact on clinical decision making for acute fracture man-
the Centre for Healthcare Research at Brighton
agement. J Orthop Trauma 2002;16:1e6.
University for his assistance with the statistics. 22. Buntic RF, et al. Using the Internet for rapid exchange of
The authors declare that they have no financial photographs and X-ray images to evaluate potential ex-
interest in the telemedicine software or any of tremity replantation candidates. J Trauma 1997;43:342e4.
hardware used in this study. 23. Whitten PS, et al. Systematic review of cost effectiveness
studies of telemedicine interventions. BMJ 2002;324:
1434e7.
24. Hammack GG. Telemedicine in corrections. Curr Probl Der-
References matol 2003;32:148e52.
25. Stanberry B. The legal and ethical aspects of telemedicine.
1. Jones SM, Milroy C, Pickford MA. Telemedicine in acute 1: confidentiality and the patient’s rights of access. J Tel-
plastic surgical trauma and burns. Ann R Coll Surg Engl emed Telecare 1997;3:179e87.
2004;86:239e42. 26. Scerri GV, Vassallo DJ. Initial plastic surgery experience
2. Wootton R. Recent advances: telemedicine. BMJ 2001;323: with the first telemedicine links for the British Forces. Br
557e60. J Plast Surg 1999;52:294e8.