Does Wearing a Face Mask Reduce Bacterial Wound Infection? A Literature Review* Fiona McCluskey RGN * In view of the nature of this article the Harvard referencing system has been retained in place of the Vancouver system normally used in this journal. Current practices of operating room management and sterile technique are direct descendants of the elaborate principles of antisepsis and asepsis set down by Lister. The surgical face mask has become an integral part of the uniform for theatre personnel since its introduction in an attempt to reduce the rate of clinical wound infections (von Mikulicz 1897). ‘The Medical Research Council’s (1968) recommendations on aseptic procedure advise donning a new mask for each patient and changing the mask part-way through long procedures (four hours or more). In most hospitals no one is allowed to enter an operating room without wearing a face mask. Anecdotal evidence in the author’s own area of practice suggests that surgeons and other theatre personnel consider wearing face masks a nuisance but feel it is the “thing to do”. Increasing costs of medical services is now a real problem. In one major teaching hospital in 1990, £10,000 was spent on masks for theatre use (Leyland and McCloy 1993). Although the available clinical data suggests that the present generation of masks does not protect staff either from airborne bacteria or Hepatitis B virus (Ransjo 1986, Reingold 1988), theatre personnel may adopt self- protection as a reason for wearing a mask. It is not the intention of this literature review to examine self protection as a reason for wearing a face mask in theatre but to investigate whether masks do reduce bacterial infections in the postoperative patient. The literature gathered for this review was accessed through Medline 1982-95 and Cinahl 1982-95 on CD ROM. The Hawthorne checklist (1983) was used to assist in the evaluation of the research reports. Research studies in this literature review cite personnel (Gillespie 1959, Duguid 1946), mask efficiency (Greene and Vesley 1962, Quesnel 1975, Letts et al 1983) and the environment (Mitchell and Hunt 1991, Ritter et al 1975) as determinants of bacterial infection, Research to demonstrate a relationship between mask wearing and postoperative infection is also reviewed (Orr 1981, Chamberlain and Houang 1984, Tunevall 1991, Berger et al 1993). Research on Personnel When the theory of droplet infection was introduced, Meleney (1927) and Walker (1930) both advocated the wearing of masks in operating theatres to reduce the risk of haemolytic streptococcus. Early studies in the 1940's and 1950's demonstrated that a few organisms were dispersed from the respiratory tract during talking and that staphylococcus aureus was mainly disseminated on the clothing rather than directly from the nose (Duguid 1946 & 1948, Hare & Thomas 1956). In 1959 Gillespie et al reported that the strains of Staphylococcus aureus in the noses of ward staff were usually different, both in antibiotic resistance and phage type from staphylococci isolated from wounds with a ward acquired infection, Although no controlled trials were carried out it is now generally accepted that the wearing of masks for ward procedures is unnecessary (Taylor 1980). However most surgeons continue to wear masks in the theatre, arguing that open wounds are exposed for a much longer period during an operation. People expel large numbers of particles of saliva from their mouth when they sneeze but much less when they talk, cough and breathe (Duguid 1946). Duguid found that on average 39,000 bacteria containing particles are produced from a sneeze, 710 from a cough and 36 from speaking 100 words loudly. The number of bacteria British Journal of Theatre Nursing. Vol 6 No 5 August 1996carrying particles from normal breathing, however,was not reported, Duguid also found that particles less than about 100 microns reduce when dry to between one quarter to one seventh of their original size, or to the size of bacteria. Research on Mask Efficiency Greene and Vesley (1962) and Quesnel (1975) investigated the efficiency of different types of masks by collecting and sizing contaminated particles escaping through or around the mask during speech by volunteers placed in dispersal chambers. These experimental studies found that most of the particles expelled by the mouth are of sufficient size and inertia to be projected onto a mask. Greene and Vesley devised a special phrase (‘spit and chew’) to be uttered to obtain significant numbers of bacteria. Quesnel (1975) commented that Greene and Vesley’s phrase ‘spit and chew’ was not adequate and he advocated a repeating of the word ‘chew’ 120 times with the unmasked face to produce several thousand bacteria. It appears from these experiments that if the mask is made of a thin paper type then the efficiency can be low (about 50%). Effective masks appear to he the soft and pleated type made from laminated fabries which wore found to have efficiencies of over 95%. Quesnel reported that the loss of efficiency in a mask could be due to particles passing around the side of the mask. The limitation of Quesnel’s study was the short sampling period of five minutes in which the subject spoke which is much shorter than an actual operation. Although this study identified which style of mask was the most efficient in an experimental situation no clinical trial was carried out so correlation toa clinical situation cannot be made. Letts et al (1983) studied the role of mask efficiency during conversation by measuring both microbial contamination of a simulated wound and operating room air. It appeared that air contamination was increased by the presence of operating room personnel which varied according to the density of the traffic with a significant increase in contamination in a simulated wound during conversation. Results highlighted that only one colony cultured from 5,595 was staphylococcus aureus, all other organisms being commensal bacteria which rarely cause infection in healthy patients. The study recommended reducing conversation and wearing a mask below a hood to reduce bacterial fallout from oral and nasal cavities. Mask Wearing and the Environment ‘An experimental study carried out by Mitchell and Hunt (1991) was primarily concerned with reviewing air flows on airborne micro-organisms in an operating theatre with forced ventilation. Results showed that during breathing and quiet talking bacterial contamination of settle plates appeared to be reduced by two to seven fold. However, no airborne bacteria counts were taken in the empty operating room which would have given a baseline recording. The author states that “few nasal bacteria were expelled into the air despite heavy colonisation in the nose” but as no nasal swabs were taken of volunteers this assumption cannot be made, The sample of subjects is not stated and it is not clear from the results or the discussion if it was the same subjects that were used throughout the experiments which would affect the reliability of the data. The study concludes that the wearing of face masks by non-scrubbed staff in theatres with forced ventilation appears to be unnecessary and advocates whispering or quiet talking and mouth breathing during an operation. These recommendations would be difficult to enforce in a clinical situation. Ritter at al (1975) demonstrated during simulated operations that the use of clinical masks did not make any difference to the number of bacteria deposited on settle plates in an operating room. What did appear to make a difference was whether the door to the room was open and the amount of traffic entering and leaving the room. Ritter found a 33-fold increase in the counts of airborne bacteria when five people entered an empty operating room, It would appear from this study that the difference in bacterial fall out between a masked and unmasked person drowns in the fallout from the rest of the body. Surgical Masks and Wound Infection Rates Orr’s study (1981) directly contrasts the results of Lett’s (1983) study, Orr’s study was designed to determine whether surgical mask wearing reduces the wound infection rate in general surgery. After an initial pilot study of one month during which no masks were worn, there was found to be no rise in the incidence of wound infections. Masks were abandoned in approximately 1000 operations during the six month period of the study. The wound infection rate was found to be 1.8% in ‘the unmasked period. Retrospective review of infection British Journal of Theatre Nursing. Vol 6 No 5 August 1996rates were 5.4%, 4.2%, 4.5% and 3.7% in the masked periods using a historical control group from the previous five years. The bacteria isolated from those wounds that became septic appeared to be associated with endogenous infection. A 50% fall in infection rate is dramatic but the length of postoperative stay which might influence the number of infections recorded was not included. ‘Tools for collection of the data and analysis are not stated so reliability and validity of the data cannot be verified. Chamberlain and Houang (1984) replicated Orr's study to see if his recommendations were applicable in gynaecological surgery. A randomly controlled trial was designed to measure bacterial contamination of the theatre environment using centrifugal air samplers and blood agar plates. A sample of forty-one women undergoing gynaecological surgery were randomly assigned to ‘masked’ or ‘unmasked’ teams of theatre staff. Although a greater number of streptococci were found on the settle plates on the operating table during the unmasked operations, all the masked groups had a higher concentration of organisms found per litre of air. ‘The trial was discontinued after the third case of postoperative infection in the unmasked group was diagnosed although none were infected with strains corresponding to those isolated from the staff. Discrepancy in the results may have been due to the small sample size, the variability in the size of the theatre teams and the difference in the air sampling techniques. The infection rate of those in the ‘masked’ group is not stated to give a comparison of the results. Berger et al (1993) carried out a quasi-experimental study to determine the relationship between mask usage and bacterial shedding. ‘Thirty cardiac catheterisation procedures were carried out with patients randomly assigned to unmasked, subnasal masked or fully masked groups. Blood agar plates and masks used all came from the same batch to ensure reliability. Nasal swabs were obtained from the group of physicians prior to each procedure which was carried out using the same technique. All bacterial colonies were enumerated and identified by an independent technician using a standard technique. Using inferential statistics (Students T+test and Wilcoxon signed rank test), mean colony counts were found to be 6.4% in the unmasked group compared to 3.2% in the fully masked group. There was no statistical difference between the unmasked and subnasal masked group. The bacterial species isolated (staphylococci, enterobacteriaceae) were assumed to represent, contamination from personnel but the exact sources of these organisms is unclear because no correlation was found between bacterial species cultures and physician nasal cultures. Although no postoperative infections were encountered, the study was discontinued “in the interests of patient safety”. This study was limited to a small sample size which may have led to the results obtained. ‘Tunevalll (1991) carried out a prospective study on the effect of wearing face masks on the surgical infection rate of 3088 patients during a two year period in acute and elective general surgery. The study design was randomised into weeks during which staff were ‘masked’ or ‘unmasked’ 1537 operations were ‘masked’ with an infection rate of 4.7%. After 1551 ‘ unmasked’ operations the infection rate was 3.5% . Results were analysed using the 2-tailed chi-square test which found that there was no statistical significance between the masked and unmasked group. The bacterial species cultured did not differ in any way between the two groups, supporting the conclusion that masks have no effect on rates of wound infection. Conclusion Modern day techniques in surgery, combined with shorter anaesthetics and hospital stays, allow surgeons to perform operations with the full expectation that the surgical wound will heal primarily without infection, The incidence of wound infection is usually not a cause of death but it does usually increase length of stay in hospital, cost of care and morbidity. Its less clear which of the often ritualistic practices add to the incidence of postoperative surgical wound infection. Some studies suggest that surgical face masks might actually increase the incidence of surgical wound infection by inereasing the shedding of facial skin (Letts 1983). Another hypothesis may be that by discarding masks individual nasal and oral droplets might be more likely to atomise and remain airborne (Quesnel 1975). Further research with carefully designed clinical studies is required in specialised areas, such as cardiac or orthopaedic surgery where the use of masks is traditionally defended, to ensure that their use is beneficial is required. But surgical principle, in the absence of hard data, will make it unlikely that such studies will be considered ethical. References Berger, S et al. 1993. Effect of surgical mask position on bacterial contamination of the operative field. Journal of Hospital Infection, 23. 51-54. Chamberlain, G., Houang, E. 1984. Trial of the use of masks in the gynaecological operating theatre. Annals of the Royal College of Surgeons of England 66. Continued on page 29 British Journal of Theatre Nursing. Vol 6 No 5 August 1996DP Continued from page 20 Duguid, J. P. 1946. The Size and Duration of Air Carriage of Respiratory Droplets and Droplet Nuclei. Journal of Hygiene (Cambridge) 44, 471-479, Duguid, J. P. & Wallace, A. 1948, Air infection with dust liberated from clothing. Lancet 2. 845-849, Gillespie, W. et al. 1959. Staphylococcal cross infection in surgery. Effects of some preventative measures. Lancet 2. 781-784. Greene, W. & Vesley, D. 1962. Method for evaluating effectiveness of surgical masks, Journal of Bacteriology. 83. 6, 63-667. Hare, R. & Thomas, C. G. A, 1956. The transmission of Staphylococcus aureus. British Medical Journal. 2. 840-844. checklist. Hawthorne, P, J. 1983. Principles of resear Nursing Times. 79. 23. 41-43. Letts, R. M. & Doermer, E. 1983. Conversation in the Operating Theater as a Cause of Airborne Bacterial Contamination. The Journal of Bone and Joint Surgery. 65. 3. 357-262. Leyland, M. & McCloy, R. 1993. Surgical face masks: protection of the self or patient? 75, 1-2. Medical Research Council. 1968. Aseptie methods in the operating suite. Lancet 1, 704-709. Meleney, F. L. 1927. Seasonal Incidence of Hemolytic Streptococcus in the Nose and Throat. Journal of the American Medical Association. 88, 1392. Mikuliez, J. 1897. Das operiren in sterilisirten zwimhandschuhen und mit mundbinde. Centrlblatt fur Chirurgie. 26, 714. Mitchell, N. J. & Hunt, 8. 1991. Surgical face masks in ‘modern operating rooms ~ a costly and unnecessary ritual? Journal of Hospital Infection, 18, 239-242. Orr, N. W. M. 1981. Is a mask necessary in the operating theatre? Annals of the Royal College of Surgeons of England. 63. 390-391. Quesnel, L. B. 1975. The efficiency of surgical masks of varying design and composition. British Journal of Surgery. 62, 936-940. Ransjo, U. 1986. Masks: a ward investigation and review of the literature. Journal of Hospital Infection. 7. 289-294. Reingold, A. L., Kane, M. A., Hightow. 1988, Failure of gloves and other protective devices to prevent transmission of Hepatitis B virus to oral surgeons. JAMA. 259, 255-260. Ritter, M.A. et al. 1975. The operating room environment as affected by people and the surgical face mask. Clinical Orthopaedics and Related Research. 111. 147-160, ‘Taylor, L. J. 1980. Are face masks necessary in operating theatres and wards? Journal of Hospital Infection. 1. 173- 174, ‘Tunevall, T. H, G. 1991, Postoperative Wound Infections and Surgical Face masks: A Controlled study. World Journal of Surgery. 15. 383-388. Walker, I. J. 1930, How Can We Determine the Efficiency of the Surgical Mask? Surgery Gynaecology and Obstetrics, 50. 266. Bibliography Berridge, D., et al, 1993, Eye protection for the vascular surgeon. British Journal of Surgery. 80. 1379-1380. Bethune, D., et al. 1965. Dispersal of Staphylococcus Aurous by patients and surgical staff. Lancet. February 27. 480-483. Charnley, J. & Eftekhar, N, 1969, Postoperative infection in total prosthetic replacement arthroplasty of the with special reference to the bacterial content of the air of the operating room. British Journal of Surgery. 56. 9. 641- 649, Courington, B, Patterson, 8. L., Howard, R. J. 1991. Universal Precautions are not Universally Followed. Archives of Surgery. 126. 93-96. Davies, R. R., Noble, W. C. 1962, Dispersal of bacteria on desquamated skin. Lancet December. 1295-1297. British Journal of Theatre Nursing. Vol 6 No 5 August 1996