Journal of Hospital Infection (2006) 63, 439e444

www.elsevierhealth.com/journals/jhin

Quantitative analysis of residual proteincontamination on reprocessed surgicalinstruments

R.L. Baxter*, H.C. Baxter, G.A. Campbell, K. Grant, A. Jones,P. Richardson, G. Whittaker

School of Chemistry, University of Edinburgh, Edinburgh, UK

Received 17 February 2006; accepted 10 March 2006Available online 12 June 2006

KEYWORDSCleaning; Surgicalinstruments; Priondiseases; Sterileservices; Proteindetermination

Summary ‘Ready-for-use’ instruments from surgical instrument trayswere examined after routine cleaning and sterilization in a blinded study.These reprocessed instruments originated from five National Health Ser-vice hospital trust sterile service departments in England and Wales. Deter-mination of residual protein and peptide contamination was carried out byacid stripping of the instrument surfaces, hydrolysis of the constituentamino acids and quantitative total amino acid analysis. One hundred andtwenty instruments were analysed, and the median levels of residual pro-tein contamination per instrument for the individual trays were 267, 260,163, 456 and 756 mg. Scanning electron microscopy and energy dispersiveX-ray spectroscopic analyses of the instruments showed that tissue de-posits were localized on surfaces, but there was no significant correlationbetween overall protein soiling and instrument complexity. The highestlevels of residual contamination were found on instruments used for tonsil-lectomy and adenoid surgery.ª 2006 The Hospital Infection Society. Published by Elsevier Ltd. All rightsreserved.

Introduction

Decontamination in medical instrument reprocessingis one of the major challenges facing healthcare

* Corresponding author. Address: School of Chemistry, Univer-sity of Edinburgh, West Mains Road, Edinburgh EH9 3JJ, UK.Tel.: þ44 131 6504708; fax: þ44 131 6506453.

E-mail address: r.baxter@ed.ac.uk

0195-6701/$ - see front matter ª 2006 The Hospital Infection Socidoi:10.1016/j.jhin.2006.03.011

facilities. This has been highlighted in several officialreports.1,2 There is now a body of scientific literatureconcerning residual tissue contamination on specificsurgical instruments and cases of nosocomial infec-tions.3e6 Concerns have also been expressed with re-spect to bacterial contamination and pyrogens.7e9 Inthis context, the potential risk of iatrogenic trans-mission of Creutzfeldt Jacob disease is noteworthy,as the infectious agent shows a marked resistance

ety. Published by Elsevier Ltd. All rights reserved.

440 R.L. Baxter et al.

to conventional chemical and thermal decontamina-tion procedures and has been shown to bind to stain-less steel.10e18 Perversely, however, virtually noquantitative data are available on the amount of tis-sue-derived material that typically remains on in-struments after cleaning and sterilization bymodern hospital decontamination procedures.This may derive from the degree of tenacity bywhich some tissue residues bind to surfaces. A crit-icism of the majority of previous studies is thatthey fail to characterize tissue contamination foundby superficial inspection, or they confound the re-sults by removing soil with detergents beforesemi-quantification by filtration and/or assay ofthe cleaning solution.

As part of a survey commissioned by the UKDepartment of Health, trays of sterile reprocessedsurgical instruments, chosen at random from thesterile services departments (SSDs) of severalNational Health Service hospital trusts in Englandand Wales, were removed from service and ana-lysed for contamination. A destructive method wasused to remove all surface contamination from theinstrument surfaces, and the bound protein wasquantified after hydrolysis to its constituent aminoacids. The present study reports on the results fromfive individual instrument trays (comprising a totalof 120 instruments), each of which was processed ina different SSD. This is the first study to providequantification of the total amount of proteinaceouscontamination that persists in adhering to instru-ments after routine hospital cleaning and steriliza-tion, and to provide reliable data on which futurerisk assessments might be based.

Methods

Scanning electron microscopy

Detailed microscopic inspection of eight instru-ments from each tray was conducted using a PhilipsXL30CP instrument, operating at 20 kV, providinga resolution better than 5 nm. Backscatter electronimaging enabled regions with a mean atomic num-ber difference of greater than 0.1 to be resolved.

Energy dispersive X-ray spectroscopicanalysis

Energy dispersive X-ray spectroscopic analysis(EDX) was carried out in a scanning electronmicroscope (SEM) using an Oxford Instruments Isis300 X-ray analyser, capable of detecting elementsof atomic number greater than 6. The imaged

areas (5 mm2) were subjected to elemental analy-sis to a depth of c. 3 mm (V w 6� 10�17 cm3) usingan X-ray fluorimeter capable of detecting 0.1%within the sample volume. No standardizationcould be applied to the EDX spectra, and theseonly indicate the relative proportions of elementspresent in the sample volume.

Protein analysis

Bovine serum albumin (BSA) and chemical reagentswere supplied by Sigma-Aldrich Ltd (Gillingham,UK); gas-plasma cleaning was carried out usinga Plasma-Etch instrument (Carson City, NV, USA)and UV-visible absorptions measured using an ATIUnicam UV4 spectrophotometer (Cambridge, UK).Standards were prepared using radiofrequencygas-plasma cleaned (argon:O2 plasma, 1 Torr, 1 h)316 stainless steel tokens (5 mm diameter,0.5 mm thick). These were preloaded with varyingamounts of BSA (10e1000 mg/disc), applied inaqueous solution (50e100 mL) and dried on to thesurfaces at c. 60 �C. Exposure to vapour over 6 MHCl at 75e80 �C in a sealed reaction vessel com-pletely destroyed the stainless steel surfaceswithin 20 min. The stripping acid and acid washingsof the surfaces were combined, and hydrolysis wascarried out in sealed pressure vessels at 135 �Cfor 16 h. Hydrosylates were evaporated to drynessunder high vacuum, dissolved in water (1 mL),adjusted to pH 12 with 4 M NaOH, and centrifugedat 4200 g to remove precipitated Fe(OH)3. Typi-cally, the iron content of the supernatant was<10 ppm. Amino acid analyses of aliquots of thesamples were carried out using an established nin-hydrin assay.19 This proved reliable over the range5e1000 mg protein with a standard deviation of 3%.

Surgical instruments were treated in an identicalmanner. In a small number of cases (c. 10%), col-oured charge transfer complexes (with dissolvedmetal ions) were formed that interfered with thecolourimetric assay. Treatment of the test solutionswith solid sodium borohydride (0.12 mmol) prior tothe ninhydrin analysis discharged the colour andhad no effect on the assay. This was effective inall but a few cases, and the results for these arenot included in the data. However, trace metalanalyses of these samples revealed no consistentpattern, and the identity of the constituent(s)responsible for the colour is unknown.

Statistical methods

Since the protein measurements within trays hada positively skewed distribution, trays were

Residual protein contamination on reprocessed surgical instruments 441

compared using Kruskal-Wallis tests. Spearmanrank correlation was used to test the associationwith instrument complexity. To test other factors,adjusted for differences between trays, analysis ofcovariance was applied to the logarithms of theprotein values.

Results

The results of complete surface protein analysisfor each tray of instruments are shown in Table I.Results for specific instruments are given in thesupplementary data on ScienceDirect (see Appendix).

Protein levels varied significantly between thefive trays (P< 0.001). Trays 1e3 did not show a sig-nificant difference from one another, while thedifference between Trays 4 and 5 was just signifi-cant (P¼ 0.03). There was no significant correla-tion between instrument complexity (as judgedon a four-point scale) and protein level, and thisremained true after adjusting for trays. The resid-ual variance from the analyses on the logarithms ofthe protein levels translated into a proportionalvariation within trays of the order of 50%, whichfar exceeds the measurement error of around 3%;this indicates that the differences between instru-ments on the same tray are real.

The instruments selected for detailed scrutinyby SEM were examined prior to surface stripping,and no effort was made to choose those of anyparticular design or material. The images shown inFigures 1 and 2 are typical of the SEM/EDX resultsobtained on 33 other instruments.

A detailed SEM examination of the inner surfaceof a pair of Metzenbaum scissors from Tray 2 usingbackscatter electron imaging showed large areasof residual contamination (Figure 1a). EDX analysis

Table I Residual proteina contamination of surgicalinstruments after conventional cleaning and steriliza-tion in a sterile service department

Trayb No. ofinstruments

analysed

Median proteinper instrument (mg)

Proteinrange (mg)

1 8 267 62e5382 27 260 58e7403 48 163 0e960c

4 20 456 37e11185 17 756 36e1173

a Only contamination from proteinaceous material wasquantified in this study.

b Trays 1e3 were basic surgery sets. Trays 4 and 5 compri-sed instruments used for tonsillectomy and adenoid surgery.

c No protein could be detected on one instrument on Tray 3.

of selected areas of the residue (Figure 1b) showedthat this material predominantly contained car-bon, nitrogen oxygen and sulphur. Protein analysisof the stripped scissor surfaces yielded a total pro-tein load of 416� 12 mg for this instrument.

Instruments from the tonsillectomy and adenoidsurgery trays (Trays 4 and 5) were typically moreheavily contaminated than instruments used inother procedures (Table I). Examination of thetips of a pair of tonsil artery forceps from Tray 5using secondary and backscatter electron imagingshowed extensive contamination between theprongs of the forceps (Figure 2a). EDX analysisshowed that this material contained carbon, nitro-gen, oxygen, sulphur, sodium and potassium(Figure 2b). Protein analysis of the stripped instru-ment surface indicated a total protein load for thisinstrument of 1173� 35 mg.

Discussion

Previous studies on evaluation of residual contam-ination on surgical instruments fall into threecategories: (1) visual or microscopy examinationsof instrument surfaces;3,5,6 (2) exhaustive washingand measurement of contaminants in the wash-ings;4 and (3) sampling small areas of the surfacesby swabbing and analysis of contamination on theswab or swab washings.20e22

Visual examination and microscopy can, at best,give an indication of heavily contaminated sites butcan hardly be considered quantitative methods. Anexception is a recent quantitative study on 23reprocessed surgical instruments by Lipscombet al., which used a protein selective fluorescentstain (SYPRO Ruby) and epifluorescence differentialinterference contrast microscopy to evaluate levelsof proteinaceous contamination on the surfaces.23

This technique gives quantitative results for smallareas of contamination but is less satisfactory fordetermination of total instrument contamination.The results were scored on the basis of a four-point‘contamination index’, ranging from uncontami-nated (average of <42 ng protein/mm2) to highlycontaminated (average of >4.2 mg protein/mm2).Since the total surface areas of the instrumentsused in the study were not measured, the totalprotein load/instrument was not determined.

A major criticism of the second ‘washing’approach is that attempts to remove dried proteinresidues that have survived instrument reprocess-ing by treatment with detergents, even if accom-panied by sonication, is inefficient and of variableefficacy.24 Indeed, in a recent study on transmissi-ble spongiform encephalopathy infected brain tis-sue dried on to steel wires, it was found that both

442 R.L. Baxter et al.

Figure 1 (a) Scanning electron microscopy backscatter image of surface contamination on the cutting face of a pairof Metzenbaum scissors from Tray 2. (b) An energy dispersive X-ray spectroscopic analysis spectrum of the surfacecontamination showing the presence of carbon, nitrogen, oxygen and sulphur from protein and the presence of alkalimetal and magnesium salts.

enzymatic digestion and prolonged detergenttreatment were required to reduce infectivitysignificantly.25 Furthermore, total protein colouri-metric assays, such as the Bradford procedure, givevariable results with different proteins and canbe inaccurate in the presence of detergents.26,27

Analysis of contamination by swabbing ‘test’areas of instrument surfaces, followed by colouri-metric tests to detect proteinaceous material onthe swab or swab eluate, is currently applied inclinical practice but affords the least reliabledata. This approach, used in commercial surfacehygiene kits, is intrinsically flawed because of thevariability in sampling areas and the inefficienciesof residue removal from the test sample and theswab itself.21 There is now a body of evidence thatcasts doubt on the sensitivity and reliability of thisapproach.28e30 At best, the technique can give anindication of gross contamination on small areasof a surface.

In this study, to guarantee total removal of allthe tissue-derived soil for analysis, all the instru-ments were sacrificed by acid stripping their entiresurfaces. The proteins were then hydrolysed totheir constituent amino acids and the total aminoacid content was determined. The results showthat while conventional SSD cleaning removesgross tissue contamination effectively, signifi-cant amounts of protein can be detected oninstruments passed for re-use on the basis of being‘visually clean.’ In general, the levels of contam-ination were far lower than could be detected byvisual examination. However, since the quantifi-cation is based on total amino acid detection, onlyproteinaceous materials were measured and thetotal contamination load, including salts, fats,polysaccharides and nucleic acids, may be muchhigher. Indeed, evidence from EDX analyses ofdeposits on selected instruments showed thepresence of metal salts and phosphorus-containing

Figure 2 (a) Scanning electron microscopy backscatter image of heavy surface contamination on a tip of a pair oftonsil artery forceps from Tray 5. (b) An energy dispersive X-ray spectroscopic analysis spectrum of the residue show-ing the presence of carbon, nitrogen, oxygen and sulphur from protein and sodium chloride.

Residual protein contamination on reprocessed surgical instruments 443

materials. It was also evident from preliminary SEMexaminations that contamination was not smearedevenly over the surfaces, but was localized insmall deposits of variable size and thickness.These patterns of contamination are similar tothose observed by Lipscomb et al. in their recentfluorescence microscopy study.23

One surprising fact was that the degree ofprotein contamination did not necessarily reflectthe degree of complexity of the instrument. Oncertain trays, plain scalpel handles or towel hooksshowed the highest levels of residual contamina-tion. It was also clear that instruments used intonsillectomy and adenoid surgery showed thehighest average residual contamination levels.With that in mind, the authors suggest that, ingeneral practice, more attention should be paid toinstruments used in this field of surgery.

Bearing in mind the fact that 1 mg of a typicalprotein residue can contain in the order of 1014 in-dividual protein molecules, this study reinforcesthe need to develop and implement more effectivemethods for instrument cleaning and for sensitivenon-destructive monitoring of the cleaning process.

Acknowledgements

The authors received funding from the Departmentof Health for this study. The views expressed in thepublication are those of the authors and notnecessarily those of the Department of Health.The authors wish to thank John Craven and NicolaCayzer of the SEM facility, Geology and Geophys-ics, University of Edinburgh.

Appendix: supplementary data

Supplementary data associated with this articlecan be found, in the online version, atdoi:10.1016/j.jhin.2006.03.011.

References

1. Pratt RJ, Pellowe C, Loveday HP, et al. The epic project.Developing national evidence-based guidelines for prevent-ing healthcare acquired infections. Phase I. Guidelines forpreventing hospital-acquired infections. Department ofHealth (England). J Hosp Infect 2001;47(Suppl. 1):S3eS82.

2. Scotland NHS. Sterile services provision review group: firstreport e the Glennie framework. Edinburgh: NHS Scotland;2001.

3. Descoteaux JG, Poulin EC, Julien M, Guidoin R. Residualorganic debris on processed surgical instruments. AORN J1995;62:23e30.

4. Alfa MJ, Degagne RT, Olson RT. Worst-case soiling levels forpatient-used flexible endoscopes before and after cleaning.Am J Infect Control 1999;27:392e401.

5. Miller DM, Youkhana I, Kuaruratne WU, Pearce A. Presenceof protein deposits on cleaned re-usable anaesthetic equip-ment. Anaesthesia 2001;56:1069e1072.

6. Dinakaran S, Kayarkar VV. Debris on processed ophthalmicinstruments: a cause for concern. Eye 2002;16:281e284.

7. Kibria SM, Kerr KG, Dave J, et al. Bacterial colonisation ofDoppler probes on vascular surgical wards. Eur J Vasc Endo-vasc Surg 2002;23:241e243.

8. Nurnberg M, Schulz HJ, Ruden H, Vogt K. Do conventionalcleaning and disinfection techniques avoid the risk of endo-scopic Helicobacter pylori transmission? Endoscopy 2003;35:295e299.

9. Williams D. Endotoxins and medical devices: thesignificance of dead bacteria. Med Device Technol 2003;14:8e11.

10. Brown P, Preece M, Brandel J-P, et al. Iatrogenic Creutz-feldt-Jacob disease at the millennium. Neurology 2000;55:1075e1081.

11. Dean M. UK concern over reuse of neurosurgical tools.Lancet 2002;360:1485.

12. Sehulster LM. Prion inactivation and medical instrumentreprocessing: challenges facing healthcare facilities. InfectControl Hosp Epidemiol 2004;25:276e279.

13. Taylor DM. Inactivation of prions by physical and chemicalmeans. J Hosp Infect 1999;43(Suppl):S69eS76.

14. Taylor DM. Resistance of transmissible spongiform encepha-lopathy agents to decontamination. Contrib Microbiol 2004;11:136e145.

15. Zobeley E, Flechsig E, Cozzio A, et al. Infectivity of scrapieprions bound to a stainless steel surface. Mol Med 1999;5:240e243.

16. Flechsig E, Hegyi I, Enari M, et al. Transmission of scrapie bysteel-surface-bound prions. Mol Med 2001;7:679e684.

17. Yan ZX, Stitz L, Heeg P, et al. Infectivity of prion proteinbound to stainless steel wires: a model for testing decon-tamination procedures for transmissible spongiformencephalopathies. Infect Control Hosp Epidemiol 2004;25:280e283.

18. Baxter HC, Campbell GA, Whittaker AG, et al. Eliminationof TSE infectivity and decontamination of surgical instru-ments using Rf gas plasma treatment. J Gen Virol 2005;86:2393e2399.

19. Spies JR. Colorimetric procedures for amino acids. MethodsEnzymol 1957;III:468e471.

20. Yamayoshi T, Doi H, Tatsumi N. Surface sampling using asingle swab method. J Hosp Infect 1984;5:386e390.

21. Moore G, Griffith C, Fielding LM. A comparison of traditionaland recently developed methods for monitoring surfacehygiene within the food industry: a laboratory study.Dairy Food Environ Sanit 2001;21:478e488.

22. Moore G, Griffith C. A comparison of traditional and re-cently developed methods for monitoring surface hygienewithin the food industry: an industrial trial. Int J EnvironHealth Res 2002;12:317e329.

23. Lipscomb IP, Sihota AK, Botham M, et al. Rapid method forthe detection of protein contamination on surgical instru-ments. J Hosp Infect 2006;62:141e148.

24. Bloss R, Hampf G. Test models to determine cleaning effi-cacy with different types of bioburden and its clinical cor-relation. J Hosp Infect 2004;56(Suppl):S44eS48.

25. Jackson GS, McKintosh E, Flechsig E, et al. An enzymee

detergent method for effective prion decontamination ofsurgical steel. J Gen Virol 2005;86:869e878.

26. Sapan C, Lundblad RL, Price NC. Colorimetric protein assaytechniques. Biotechnol Appl Biochem 1999;29:99e108.

27. Marshall T, Williams KM. Interference in the Coumassie bril-liant blue and pyrogallol red protein dye binding assays is

444 R.L. Baxter et al.

increased by the addition of sodium dodecyl sulfate to thedye reagents. Anal Biochem 2004;331:255e259.

28. Davidson CA, Griffith CJ, Peters AC, et al. Evaluation of twomethods for monitoring surface cleanliness e ATP biolumi-nescence and traditional surface swabbing. Luminescence1999;14:33e38.

29. Blais BW. Swab-based enzyme immunoassay system fordetection of meat residues on food contact surfaces as ahygiene monitoring tool. J Food Prot 1999;62:386e389.

30. Sanderson WT, Hein MJ, Taylor L, et al. Surface samplingmethods for Bacillus anthracis spore contamination. EmergInfect Dis 2002;8:1145e1151.