Comparison between visual analysis and
microscope assessment of surgical instrument
cleanliness from sterile service departments
I.P. Lipscomb*, A.K. Sihota, C.W. Keevil
Received 29 September 2006; accepted 3 August 2007
Available online 17 October 2007
ity of this method. Surgical instrument sets were obtained from nine anony-
mous National Health Service (NHS) Primary Care Trust SSDs to investigate
the efficacy of ‘in-place’ cleaning procedures. The instruments were first in-
spected visually, followed by a novel technique called episcopic differential
on the medical devices. The application of a Contamination Index (0e4) for
both proteinaceous and non-proteinaceous deposits on the surface allowed
quantitative assessment. Close correlation was seen for simple instruments
between visual assessment and microscopic analysis. For more complex in-
struments, however, there was a marked difference between the two assess-
ment techniques and the microscopy procedure showed areas of
proteinaceous and non-proteinaceous crystalline soiling that was difficult or
tected using such antiquated methods. The new methodology applied in the
assessment of surface contamination is rapid and generally applicable and
could be used more widely for routine monitoring of instrument cleanliness.
ª 2007 The Hospital Infection Society. Published by Elsevier Ltd. All rights
Modern hospital sterile service departments (SSDs) routinely in-
* Corresponding author. Address: Environmental Healthcare Unit, School of Biological Sciences, University of Southampton, Rm
11003, Faraday Building, Highfield, Southampton SO16 7PX, UK. Tel.: þ44 (0)2389 593242; fax: þ44 (0)2380 597322.
E-mail address: firstname.lastname@example.org
0195-6701/$ - see front matter ª 2007 The Hospital Infection Society. Published by Elsevier Ltd. All rights reserved.
Journal of Hospital Infection (2008) 68, 52e58
Available online at www.sciencedirect.com
There are more than 6.5 million operations a year
performed in England alone.1These procedures
produce approximately 9.2 million surgical trays
that require decontamination.2With an average
of 12 instruments per set, this means that approx-
imately 110 million instruments require decontam-
ination per annum, or in real terms 2 million
instruments per week spread over the 249 hospi-
tals with sterile service departments (SSDs).3
Therefore the average SSD, which handles approx-
imately 50 000 trays per annum, must cope with
more than 1500 instruments per day.3
The functions of an SSD are five-fold. Once the
instruments have been taken into the SSD either
post-surgery or new, they are cleaned and dis-
infected in a mechanical washer/disinfector (WD)
with the application of an enzymatic or alkaline
detergent, and sometimes sonication. All the
cleaning cycles have to be validated in accordance
with both British Government (HTM2030) and
European Union (ISO EN15883) guidelines and this
validation includes the chemical testing for protein
residues with approved methods such as the
Ninhydrin or Biuret test. Recent research has cast
doubt over the sensitivity of such test procedures.4
Once dry, the instruments tend to be visibly
inspected and passed for any residue soiling or
mechanical failure before being packaged up and
sent for sterilisation. It has been recommended
that these visual inspections are performed daily to
verify the efficacy of cleaning, and it has been
reported that the haem pigment found in haemo-
globin can easily be detected at levels as low as
10 mg/cm2.5It was also noted in the same report
that bodily fluids without pigments are extremely
difficult to visualise, even in large quantities. Hu-
man cerebral spinal fluid (CSF) is both colourless
and odourless but has been shown to be a carrier
of infection of prion diseases e a group of neuro-
degenerative and invariably fatal conditions that
include variant CreutzfeldteJakob disease (vCJD).6
Prion diseases commonly display symptoms of
rapidly progressive dementia accompanied by cere-
bellar ataxia and myoclonus.7The prion-causing
agent is highly robust and has been shown to remain
inactivation regimes such as autoclaving (121?C,
30 min) or chemical methods, e.g. formaldehyde
gas, have been implemented.6,8
Such persistence, in conjunction with the evi-
dence of iatrogenic transmission, has increased the
concern that relatively large numbers of instru-
ments may be passing through SSDs without
sufficient processing.9,10Consequently the Depart-
ment of Health issued revised guidelines on the
decontamination of instruments, and the British
Government announced a £200 million investment
(NHS) decontamination/sterilisation facilities.11e13
Episcopic differential interference contrast/
epifluorescence microscopy (EDIC/EF microscopy)
has been applied previously to visualise very small
proteinaceous deposits upon surgical steel sub-
strata.14e16In the present research the EDIC/EF
microscope was applied to confirm the ability of
visual analysis in preventing poorly reprocessed
and possibly infectious instruments from entering
the theatre environment and to offer a possible al-
ternative assessment technique. The research also
aimed to look at any subjective visual inspection
bias placed on the instruments by an inspector by
subdividing the instruments into simple and hinged
devices and further categories of these devices.
Nine surgical instrument sets were received from
the Department of Health and all identification
marks had been removed before delivery. The nine
sets consisted of over 350 individual instruments,
with an average of 40 instruments per set. The
instruments were identified by type and size; all
single or double instruments were tested, but in
cases where the instrument types consisted of
more than two, a representative selection (>50%)
of that instrument’s type was examined. In total,
260 instruments were assessed for contamination.
The surgical instruments ranged in size and shape.
All had passed through traditional machine WD
cleaning procedures and had been deemed clean.
All the instruments were visually inspected on
arrival by a researcher familiar with the different
instrument layouts (A.S.) and were scored for their
cleanliness between 0 (no visible contamination)
and 2 (large amounts of visible contamination) in
all of the defined sample regions.
The instruments were stained with SYPRO Ruby
and visualised using episcopic differential inter-
ference contrast/epifluorescence (EDIC/EF) light
microscopy in a manner previously described.14
All of the instruments were examined at mul-
tiple sample points over their surface and scored by
applying a previously defined Contamination Index
(Table I) of between 0 and 4, with 4a being gross
contamination but not of a proteinaceous nature,
i.e. readily observable using EDIC microscopy but
not staining with SYPRO Ruby.14As the handling
Assessment of surgical instrument contamination 53
of devices post-cleaning is often carried out with-
out gloves, scores of ?1 were considered to be
low and possibly due to non-clinical contamination.
The defined sample areas of instruments were
assessed and scored by comparing the visualised
contamination with previously obtained represen-
tative images of known Contamination Index
(Figure 2). This enabled the rapid assessment of
the degree of contamination apparent for each
region of interest and multi-regional sampling
was performed on all instruments.
For this investigation the sets were analysed and
subdivided into instrument class (i.e. hinged or
simple). Hinged instruments (N ¼ 119) were classi-
fied as those instruments which possessed a box
joint, e.g. artery forceps (Figure 1). Simple instru-
ments (N ¼ 83) were those without a box joint and
with three distinct regions of assessment, e.g.
tongue plates or British Pharmacopeia (B.P.)
The instruments examined had a wide variation in
both size and complexity. Of the instruments
inspected visually 37% (N ¼ 75/202) showed a low
level of contamination by visual assessment with
a mean score of <1 and only 4% (N ¼ 8/202)
showed high levels of contamination (mean score
These findings were in contrast to those found
by applying the previously defined microscope
contamination index. The results from the EDIC/
EF microscopy analysis indicated that 66% of all
the instruments inspected showed severe (classes
3 and 4; >4.2 mg protein per mm2) contamination,
27% were moderately contaminated (classes 2 and
3; 0.42e4.2 mg protein per mm2), and only 7% dis-
played low-level soiling (classes 0e2; 0e420 ng
protein per mm2). In addition, regions of crystal-
line deposits that were not identified by visual
inspection were clearly visualised under EDIC illu-
mination. These deposits, although co-localised
with proteinaceous soiling, did not test positive
using SYPRO Ruby staining. Image analysis of a com-
posite of the EDIC and EF images clearly showed
the SR-positive proteinaceous material adhering
on the surface of the non-proteinaceous crystal-
line deposits. Scores of the adjacent areas dis-
played relatively little SR-positive material.
The average Contamination Index per instrument
set varied among the nine trays (range 2.4e3.6),
with the overall mean Contamination Index value
for all the instruments being 3.2 (Figure 3).
Within the instrument subgroups examined, 41%
(N ¼ 49) of the hinged and 31% (N ¼ 26) of the
Parameters and equivalent protein concentration for the Contamination Indexa
>4.2 mg PE
FOV, field of view (0.36 mm2); PE, protein equivalent soil that did not stain with SYPRO Ruby.
aProtein per mm2, based on the calculation that a 1 mm diameter area of protein with an average molecular weight of 30 kDa
and 3 mm in height is approximately equivalent to 1 pg (data not shown).
examination regions tested within the experiment.
Hinged device showing the four different
54I.P. Lipscomb et al.
simple instruments had visual assessment scores of
<1. The results of both visual assessment and mi-
croscope analysis are shown in Figure 4. A compar-
instruments (Figure 4a) shows a consistency with
the microscopic soiling levels detected. However,
the visual findings for test areas of the hinged in-
struments (Figure 4b) vary much more, even
though the microscopic levels observed remain rel-
atively constant, and from these findings it ap-
pears that a subjective bias is present. Indeed
certain instrument areas such as the inner hinge
were perceived as possessing high levels of con-
tamination when compared with other regions of
the same instrument.
For simple instruments, the KruskaleWallis
analysis demonstrates no significant difference
between the different regions of the instruments
for both visual assessment
(P < 0.05) between the visual scores from the in-
ner hinge region and all of the other three re-
(P < 0.05) between the scores for the outer
hinge and the arm.
There are approximately 6.5 million surgical pro-
cedures performed within England each year.1
These procedures are spread across the 182 acute
NHS trusts which themselves cover the 249 hospi-
tals with sterile service departments (SSDs) in
England.17The emergence of evidence that highly
robust infectious agents such as the prion protein,
a characteristic of variant and sporadic Creuzt-
feldteJakob disease, may remain viable following
standard hospital decontaminating procedures,
led the Department of Health to issue revised
guidelines on the decontamination of instruments
1999.6,11,12,18e20However, it is clear that subse-
quent and ongoing monitoring of cleaning stan-
dards must be maintained in order to ensure that
reached and maintained, to reduce any possibility
of nosocomial infection.
The efficacy of the cleaning process is tradi-
tionally assessed using visual inspection of the
instruments. Time constraints and the sheer num-
ber of instruments involved must cast doubt on
4 (gross proteinaceous contamination) and 4a (gross contamination), as seen by episcopic differential interference
contrast microscopy, but not proteinaceous in nature (Bar ¼75 mm).
Examples of Contamination Index (CI) assessment. CI ranges from 0 (no proteinaceous contamination) to
Assessment of surgical instrument contamination 55
both the ability and the validity of visual inspec-
tion to ensure that effective cleaning is achieved.
In 1995 a study found that although more than 90%
(29/32) of the visually inspected instruments
looked clean, more than 84% possessed some
residual soiling.21However, that study did not at-
tempt to identify the origin of the debris. The find-
ings from our study are in broad agreement; in
addition, our method enables clear differentiation
between proteinaceous and non-proteinaceous
The present investigation has examined 260
instruments obtained anonymously
Primary Care Trusts in England and Wales. They
were assessed using a combination of a novel
microscopy technique, sensitive protein staining
anda previously described
Average visual inspection score
Average contamination index score
Tip MidHandle Outside hinge Arm Inside hinge Blade
Average visual inspection score
Average contamination index score
subgroups that were classified as having low amounts of contamination. A comparison of the inspection regions for
simple instruments (a) shows consistency with the microscopic soiling levels. The hinged instruments (b) also show
little variation in the contamination levels under microscope analysis but visual scores vary significantly between
the instrument test regions. *Significant difference between inside hinge region and outside hinge, arm and blade
(P<0.05). **Significant difference between outside hinge region and the arm (P< 0.05).
Visual assessment (white bars) and microscope analysis (black bars) results from the surgical instrument
Instrument set number
Average visual inspection score (0-2)
Average contamination index score obtained (0-4)
National Health Service trusts using both visual assessment (white bars) and microscope analysis (black bars).
Mean Contamination Index scores of the different instrument sets obtained from the nine anonymous
56 I.P. Lipscomb et al.
The inter-set results showed significant differ-
ences in cleaning efficacy between instrument
sets gained from different sources; the overall
trend of the results between microscopic and
visual assessment, however, appear to be in
general agreement with each other.
The results obtained from the simple instruments
indicated that there was no significant difference
in the visibly distinguished soiling levels between
the different sample regions of the instrument. In
contrast, the hinged instruments displayed a sig-
nificant variation between the different regions
with the inner hinge showing the highest visual
scores. This can be considered to occur for two
main reasons: (i) discolouring occurs more readily
within this region due to a combination of pro-
teinaceous contaminants, corrosive by-products,
detergent deposits and general scoring of the
surface through normal use; (ii) subjective bias e
the preconception that a particular area is most
likely to harbour contamination, resulting in closer
scrutiny of this area than any other.
Some of the instruments appeared heavily
soiled when observed using EDIC microscopy; this
soiling was not proteinaceous as it did not stain
with SYPRO Ruby. The soiling frequently appeared
crystalline in nature and may have consisted of
deposits remaining from the use of detergent or
enzymatic cleansers in the WDs. These deposits
could not be observed by eye, perhaps suggesting
that their occurrence may have been more wide-
spread than hitherto realised. Image analysis of
a composite of the EDIC and EF images clearly
showed the SR-positive proteinaceous material
adhering on the surface of these non-proteina-
ceous crystalline deposits. Many of the adjacent
areas displayed relatively little SR-positive ma-
terial, suggesting that either: (i) protein binds
preferentially to the crystalline deposits compared
to stainless steel; or (ii) the protein is more
difficult to remove from the crystalline deposits
during the cleaning process; or (iii) crystalline
deposition may occur in inaccessible regions of
the instrument that are difficult to clean.
Indeed, if they are found to occur commonly
then consideration must be given to the formation
of cleaning agents used in WDs and to the way in
which WDs are operated with respect to temper-
ature and rinse cycles.
In conclusion, we have compared direct visual-
isation with an in-situ description of proteinaceous
and non-proteinaceous contamination on surgical
steel instruments. Our method overcomes the
drawbacks inherent with traditional methods that
employ sampling techniques to assess contamina-
tion on a surface.14The method has been shown to
allow sensitive quantification of the contamination
and the results indicate that complex instruments
may need a more stringent assessment of day-
to-day cleaning than direct visualisation can
Although this technique requires further de-
velopment before it can be applied to every
instrument passing through an SSD, its simplicity
and speed would allow it to be used as a ‘spot
check’ to confirm that standards of cleanliness are
being maintained. In addition, since the SYPRO
Ruby fluorescent probe binds only to protein, the
removal of such contamination would also remove
this marker and have no effect on subsequent
This work may help to reduce iatrogenic trans-
mission of robust infectious agents such as the
prion protein. It can therefore offer an increase in
public confidence towards healthcare cleaning and
decontamination procedures worldwide. Although
it is worth bearing in mind that the age and history
of the instruments were unknown, it is clear that
visual inspection lacks the sensitivity required to
ensure effective decontamination.
Conflict of interest statement
This work was funded by the Department of
Health (contract DH 0070073). The views ex-
pressed in this paper are not necessarily those
of the Department of Health.
1. Government Statistical Service. Hospital Episode Statistics
England: financial year 2003e04. Department of Health;
2. NHS Decontamination Project. NHS Decontamination pro-
ject, background and expectations. NHS Estates; 2003.
3. NHS Purchasing and Supply Agency. Paper on individual in-
strument and surgical tray identification. NHS PASA; 2004.
4. Lipscomb IP, Pinchin HE, Collin R, Harris K, Keevil CW. The
sensitivity of approved Ninhydrin and Biuret tests in the
assessment of protein contamination on surgical steel as
an aid to prevent iatrogenic prion transmission. J Hosp
5. Quality task group. Cleaning (part 2) e validation of clean-
ing efficacy. Zentral Steril 2002;10:60e66.
agents: safe working and the prevention of infection. The
Stationery Office; 2003.
Assessment of surgical instrument contamination57
7. Collins S, Boyd A, Fletcher A, et al. CreutzfeldteJakob Download full-text
disease: diagnostic utility of 14-3-3 protein immunode-
tection in cerebrospinal fluid. J Clin Neurosci 2000;7:
8. Taylor DM. Inactivation of transmissible degenerative en-
cephalopathy agents: a review. Vet J 2000;159:10e17.
9. Bernoulli C, Siegfried J, Baumgartner G, et al. Danger of
accidental person-to-person transmission of Creutzfeldte
Jakob disease by surgery. Lancet 1977;1:478e479.
10. Collins S, Masters CL. Iatrogenic and zoonotic Creutzfeldte
Jakob disease: the Australian perspective. Med J Aust 1996;
11. Department of Health. Variant CreutzfeldteJakob disease
(vCJD): minimising the risk of transmission. Health Service
Circular 1999, HSC 1999/178.
12. Department of Health. Controls assurance in infection con-
trol: decontamination of medical devices. Health Service
Circular 1999, HSC 1999/179.
13. Anonymous. v-CJD in the future. The Parliamentary Office
of Science and Technology 2002;171.
14. Lipscomb IP, Sihota AK, Botham M, Harris KL, Keevil CW.
Rapid method for the sensitive detection of protein con-
tamination on surgical instruments. J Hosp Infect 2006;62:
15. Lipscomb IP, Sihota AK, Keevil CW. Diathermy forceps and
pencils: reservoirs for protein and prion contamination?
J Hosp Infect 2006;64:193e194.
16. Lipscomb IP, Sihota AK, Keevil CW. Comparative study
of surgical instruments from sterile-service departments
for presence of residual gram-negative endotoxin and pro-
teinaceous deposits. J Clin Microbiol 2006;44:3728e3733.
17. NHS Estates. Comprehensive report: a review of the decon-
tamination of surgical instruments in the NHS in England.
NHS Estates; 2001.
18. Taylor DM, Fernie K, McConnell I, Steele PJ. Observations on
thermostable subpopulations of the unconventional agents
that cause transmissible degenerative encephalopathies.
Vet Microbiol 1998;64:33e38.
19. Taylor DM. Inactivation of prions by physical and chemical
means. J Hosp Infect 1999;43(Suppl.):S69eS76.
20. Brown P, Rau EH, Johnson BK, Bacote AE, Gibbs Jr CJ,
Gajdusek DC. New studies on the heat resistance of ham-
ster-adapted scrapie agent: threshold survival after ashing
at 600 degrees C suggests an inorganic template of replica-
tion. Proc Natl Acad Sci USA 2000;97:3418e3421.
21. DesCoteaux JG, Poulin EC, Julien M, Guidoin R. Residual
organic debris on processed surgical instruments. AORN J
58I.P. Lipscomb et al.