Should eye protection be worn during dermatological surgery: prospective observational study

Background  There is a potential risk of infection with blood‐borne viruses if a doctor receives a blood splash to a mucous membrane. The quantification of facial contamination with blood has never been documented in the context of dermatological surgery.


Summary
Background There is a potential risk of infection with blood-borne viruses if a doctor receives a blood splash to a mucous membrane. The quantification of facial contamination with blood has never been documented in the context of dermatological surgery. Objectives (i) To identify the number of facial blood splashes that occur during skin surgery and to identify the procedures that present higher risks for the operator and assistant. (ii) To assess the provision of eye protection and attitudes to its use in dermatological surgery in the U.K. Methods (i) Prospective, observational study in the skin surgery suite of a U.K. teaching hospital assessing 100 consecutive dermatological surgery procedures, plus 100 consecutive operations in which an assistant was present. Primary outcome: number of face-mask visors with at least one blood splash. Secondary outcomes: to identify if any of the following variables influenced the occurrence of a blood splash: grade of operator, site and type of procedure, and the use of electrocautery. (ii) A postal survey of all U.K.-based members of the British Society of Dermatological Surgery (BSDS) was conducted assessing facilities available and the attitudes of U.K.-based clinicians to the use of face masks during surgery. Results (i) In 33% of all surgical procedures there was at least one facial splash to the operator (range 1-75) and in 15% of procedures the assistant received at least one splash (range 1-11). Use of monopolar electrocautery was significantly less likely to result in splashes to the mask compared with bipolar electrocautery [odds ratio (OR) 0AE04; 95% confidence interval (CI) 0AE01-0AE19]. Compared with the head/neck, operations on the body were significantly more likely to result in splashes to the mask (OR 6AE52) (95% CI 1AE7-25AE07). The type of procedure and the status of the operator did not have a bearing on the likelihood of receiving a splash to the mask. (ii) From the survey, 33 of 159 (20AE8%) of BSDS members had no face masks available and 54 of 159 (34AE0%) did not wear any facial protection while operating. The majority (53AE5%) thought they received a splash in £ 1% of procedures. Conclusions There is a substantial risk of a splash of blood coming into contact with the face during dermatological surgery for both the operator and assistant, regardless of the procedure. The risk of receiving a blood splash to the face may be substantially underestimated by U.K.-based dermatologists. The use of protective eyewear is advisable at all times, but particularly when using bipolar electrocautery, or when operating on high-risk individuals.
There is a potential risk of infection with blood-borne viruses if a doctor receives a blood splash to a mucous membrane. The quantification of facial contamination with blood has been documented in both orthopaedic, 1 and obstetrics and gynaecological surgery, 2 but never in the context of dermatological surgery. Therefore, we aimed to document the number of facial splashes that occur, to both the operator and assistant, during skin surgery and identify the higher risk procedures. In addition, the facilities available to and the attitudes of the U.K.-based dermatological surgeons to the use of facial protection during surgery were sought.

Materials and methods
One hundred consecutive operations performed in the skin surgery unit of a major teaching hospital were assessed prospectively. The operator and their assistant, if they had one, used a new surgical mask with clear plastic visor for each procedure. After the operation the mask was placed in an envelope on which the grade of operator, site and nature of the procedure, and whether any electrocautery was used was noted. The mask was later examined (by A.J.B.) using a magnifying glass, and the number of blood splashes on the visor was recorded. These were identified by their red-brown colour; clear or light yellow spots were ignored. A logistic regression model was created including the following four variables: (i) procedure type; (ii) status of operator; (iii) type of electrocautery; and (iv) body site involved. The binary outcome of splashed (yes/no) was used as the dependent variable. All variables were entered together initially, with entry criteria set at 0AE05 and exclusion at 1AE0. The model was then re-run including the two variables that were identified as being significant in the initial model.
In addition to the main study a postal questionnaire ( Fig. 1) was sent to all U.K.-based members of the British Society for Dermatological Surgery (BSDS) asking whether they were provided with face masks and whether they used protective equipment while performing dermatological surgery. In addition they were asked in what proportion of their operations they perceive a blood splash to the face.

Results
Of all the surgical procedures conducted 33% resulted in at least one visor splash to the operator (range 1-75) and 15% resulted in at least one splash to the assistant (range 1-11). The most striking predisposing factor appeared to be the use of bipolar electrocautery with 27 of 57 (47%) procedures involving this instrument resulting in blood splashes. Having adjusted for the other factors using the logistic regression model, use of the Hyfrecator (CONMED Corporation, Utica, NY, U.S.A.) (monopolar electrocautery) was significantly less likely to result in visor splashes compared with bipolar electrocautery [odds ratio (OR) 0AE04; 95% confidence interval (CI) 0AE01-0AE19]. Compared with the head/neck, operations on the body were significantly more likely to result in visor splashes (OR 6AE52; 95% CI 1AE7-25AE07). The type of procedure and the status of the operator did not have a bearing on the likelihood of receiving a visor splash. Tables 1 and 2 provide the individual data.
Questionnaires were mailed to all 193 members of the BSDS. There were 159 responses (82AE4%): 33 of 159 (20AE8%) do not have any face masks available and only 48 of 159 (30AE2%) have access to face masks with visors; 54 of 159 (34AE0%) do not wear any facial protection while operating (this includes eye protection-even basic prescription spectacles). The responses to the question asking the doctor's opinion as to the proportion of procedures in which they receive a splash to the face were positively skewed with only 12 of 159 (7AE5%) thinking that they received splashes in more than 10% of operations; both the median and mode figures were splashes in 1% of operations.

Discussion
Of all surgical procedures, 33% resulted in at least one visor splash to the operator and in 15% of procedures the assistant received at least one splash. It is also significant that the use of bipolar electrocautery results in more blood splashes. The majority of BSDS members thought that they received a blood splash in only £ 1% of surgical procedures.
We sought to avoid bias by documenting the number of blood splashes in consecutive procedures and thus a wide range of procedures was performed by all members of the department. It could be argued that the number of splashes recorded in this study may be higher than expected in other departments as the procedures performed in Nottingham may be of higher complexity and a greater emphasis is placed on training in skin surgery. However, one might expect that with better technique fewer splashes may occur (which appears not to be the case as consultants received more splashes than registrars) and that those questioned in the survey are dermatologists with a special interest in dermatological surgery and thus may perform more complex procedures than the national average. We did not correlate the results of the assistant with those of the main operator as only a proportion of cases required assistance and it was felt that a higher number of recorded procedures was more useful than identifying whether a splash occurred at the same time to the assistant and the operator.
This study is the first of its kind in dermatological practice in documenting the actual number of blood splashes occurring during skin surgery. Blood contamination has been identified in 86% of visor masks used during orthopaedic procedures, 1 in 50% of caesarean sections, and in 32% of vaginal deliveries. 2 Our results suggest that dermatological surgery is less traumatic than these specialties, which seems intuitively correct.
BSDS members may considerably underestimate the number of facial blood splashes they receive during dermatological surgery, and it would seem reasonable to extrapolate this to anyone undertaking skin surgery. The actual figure of 33% of procedures causing facial splashes suggests that many departments may be putting their employees at risk by not providing them with protection as there is a potential risk of infection with blood-borne viruses if a doctor receives a blood splash to a mucous membrane. Indeed, this can be extended to the assistant operators-15% of all procedures (which includes simple biopsies). Conjunctival transmission has been reported for both human immunodeficiency and hepatitis B viruses. 3 The risk of transmission of these viruses by this route must be low as there have been so few reports of infection despite the large number of procedures performed; however, in today's risk-averse society one would expect surgeons to protect themselves as much as possible.
Intuitively, one may have expected there to be more visor splashes resulting from operations on the head as this is a more vascular area; however, the results did not support this and indeed there was a significantly greater chance of receiving a splash when operating on the body. One possible explanation for this is that one tends to experience a more 'explosive' response when using bipolar electrocautery on sites with a lot of fat. The trunk is often an area with increased fat and surgery in these areas may also be deeper due to wide local excisions for malignant melanoma removal.
These results show that there is a substantial risk of a splash of blood coming into contact with the face for both the operator and assistant regardless of the procedure. Indeed, although only eight of 100 procedures in this study were punch biopsies, there was still one visor splash resulting from a punch biopsy. Likewise, three of seven procedures performed without any electrocautery resulted in the visor receiving a blood splash. One possible reason for this finding may be that when tying a suture some blood may splash as the knot is pulled together, a situation which the authors (A.J.B. and S.V.) have encountered.
One may be able to extrapolate these results to other surgical procedures, as most will involve passing through the skin, and in particular to general practice where an increasing number of minor operations are being performed. The subject of face and in particular eye protection is given little space, but 2 nd November 2005

Dear
We are currently conducting a study at Queen's Medical Centre in Nottingham to document the rate of occurrence of facial splashes during skin surgery. As part of this study we would like to record the use of surgical masks and eye protection amongst dermatologists who perform any surgical procedures.
We would be extremely grateful if you could complete the attached questionnaire and return it to us, at the dermatology department, in the reply paid envelope provided.  recommended, in dermatology and dermatological surgery textbooks. [4][5][6] The use of protective eyewear is advisable at all times, but particularly when using bipolar electrocautery and operating on high-risk patients. Thus, this evidence could be presented to hospital infection control and managers when planning surgical services. The question as to why more cases of conjunctival transmission have not been reported needs to be considered. Perhaps the heat of electrocautery sterilizes the blood before it splashes the operator. Skin surgery is a major specialty in Nottingham and the results here may not be reflected in smaller units performing less complex procedures. Although visor splashes were significantly increased when performing operations on the body, rather than on the face, the 95% confidence intervals were wide and it would be useful to be able to confirm this finding. A repeat of this study in another centre would help to answer these questions. One has to ask why surgeons do not protect themselves more. Could it be that they find the mask uncomfortable, or suffer from steaming up of the visor? Perhaps they are not concerned if they receive a splash of blood to the face, or, alternatively, surgeons are unaware of the amount of blood that splashes towards the face during dermatological surgery. If properly applied then the visor does not steam up and can be worn by those who wear spectacles, so with further education and after publication of this manuscript it would be interesting to repeat the questionnaire study to see if attitudes have changed.