This study evaluates the effect of Iodine-impregnated self-adhesive plastic drapes (IIAD) on the quantity of skin bacteria, which is a known factor in the development of surgical site infections in orthopedic surgery. 16 patients scheduled for total knee replacement surgery (TKA) are recruited for participation in an intervention performed at the orthopedic ward of Odense University Hospital, Denmark.

Right or left patient knee is randomized to receive draping with IIAD, while the other knee is not draped. Local standard preoperative procedure for knee surgery is then simulated.

After 75 minutes both knee are sampled for bacterial quantity using the cup scrub method and samples are aerobically incubated. The bacterial growth of the IIAD and no-drape group are then compared using nonparametric statistics.

Background

Surgical site infection (SSI) is a serious complication to orthopaedic surgery, resulting in re-operation and prolonged recovery (1, 2)

To prevent SSI, prophylactic methods have been introduced as part of procedure in many types of surgery. Prophylactic strategies such as preoperative disinfection using chlorhexidine or iodine solutions with alcohol have been shown to decrease the number of skin bacteria prior to surgical incision or intravenous catheter insertion. (3-7). Creating a completely aseptic environment for the duration of surgery is however impossible, due to the inability of preoperative disinfection to completely sterilize the skin, combined with re-colonization of the patients endogenous flora from hair follicles, sebaceous glands and sweat glands.(6) (8, 9)

The use of self-adhesive plastic drapes was introduced to prevent contamination of the surgical site due to re-colonizing skin flora, by blocking the translocation of bacteria from the skin adjacent to the surgical site into the surgical wound. Self-adhesive plastic incision drapes are in this way designed to work as a barrier, containing skin flora during surgery.

These types of drapes have however been found to be inferior in preventing bacterial SSI, compared to newer drape types that incorporates an antiseptic iodine-complex into the drape adhesive, facing the skin (10, 11) This outcome might be explained by a drape-induced "greenhouse effect" supported by evidence of increased bacterial colonization-rate with the intraoperative use of non-iodine self-adhesive plastic drapes (12, 13). This effect may be a result of relative occlusion of the skin leading to an increase in skin hydration.(14) This increase in bacterial growth has never been demonstrated in iodine impregnated self-adhesive incision drapes (IIAD) and has been contradicted in studies reporting a reduction in re-colonization rate of skin bacteria on the abdomen when IIAD was compared to no draping and non-iodine drapes while other authors reported decrease in surgical wound contamination when IIAD was used preoperatively in abdominal surgery (3, 15)

Regardless of the superiority of IIAD to non-iodine drapes, it has recently been concluded by a meta-analysis, that the use of IIAD does not reduce SSI incidence compared to no draping (10). This conclusion has been criticized for being due to type 2 error resulting from the relatively low rate of SSI in clean surgery (16) The review does, however, contain minimal data from the use of IIAD in orthopaedic surgery. The transferability of its conclusions to an orthopaedic setting is therefore questionable.

Skin flora, predominately staphylococci, is the primary course of SSI in orthopaedic surgery (17, 18) and has also been shown to constitute the majority of the recolonizing species on the skin (8, 19).

IIAD has therefore become a part of standard preoperative procedure in many orthopaedic surgical setting in the hospitals of Southern Denmark, including Odense University Hospital and Svendborg Hospital. This decision may be justified by in-vitro studies concluding marked bactericidal effect of IIAD on a broad range of bacterial species, especially staphylococci. (16) The bactericidal effect of IIAD has also been evaluated in-vivo in hip surgery, where bactericidal effect and significant reduction in wound contamination was reported when IIAD was applied 24 hours prior to skin sampling (20). The bactericidal effect has however not been evaluated when IIAD is applied at the surgical start prior to incision, as is normal practice in orthopaedic surgery.

Re-colonization of skin bacteria following disinfection has been shown to occur completely or partially within 7 days when the skin is disinfected with a chlorhexidine or iodine solution and covered with a polyurethane dressing, not containing antimicrobials.(21-24) Re-colonization may however already be detectable within few hours following disinfection. This has however not been associated with a significant increase in SSI rates.(3, 13) It is worth noting that some evidence suggests that skin preparation solutions containing chlorhexidine, may reside on the skin and provide residual antiseptic action on its own. (25-27)

IIAD has not been shown to improve the incidence of SSI in non-orthopaedic surgery and some retrospective evidence suggest similar findings in hip surgery (28). It is therefore of great relevance to shine further light on the effect of IIAD in an orthopaedic surgical setting.

Motivation

Iodine impregnated self-adhesive plastic drapes have been used in many types of surgery for years. Despite of this, its use has not been proven to decrease the incidence of SSI.

SSI resulting in intraarticular infections with skin bacteria is a well-known and serious complication to orthopaedic surgery. The evaluation of prophylactic measures such as IIAD is therefore desperately needed. In spite is this, orthopaedic surgery is poorly represented in the current literature on the subject.

It is the goal of our study to add to the knowledge on the effect of the use of IIAD on dermal bacterial growth in an orthopaedic surgical setting.

Objectives

The objective of this study is to evaluate the effects of intraoperative application of iodine impregnated self-adhesive drapes on the bacterial quantity (colony forming units) of the skin in patients scheduled for total knee replacement surgery (TKA). We wish to investigate if intraoperative draping with an IIAD increases re-colonization with skin microbes compared to no drape use during simulated TKA surgery. Primary outcome is microbial quantity, while secondary outcome is identification of viable organisms on the skin. The hypothesis is that IIAD increases re-colonization of skin microbes compared to no drape use.

Method

Study design

The study is designed as an experimental randomized controlled trial performed in a parallel setting. Patients will serve as their own control and allocation ratio is therefore equal.

Participants

Patients are recruited in a prospective manner through the outpatient clinic for osteoarthritis of the knee, Odense University Hospital. The attending physician screens patients for participation.

Patients are eligible for inclusion if they are scheduled for total knee replacement surgery through Odense University Hospital. Reasons for exclusion are listed below.

Sample size

A difference in bacterial quantity of a factor 10 (1.0 log10) between intervention and control group is judged to be of clinical significance in the development of SSI. To detect a difference of 1.0 log in bacterial quantity with a power level of 80%, a significance level of 5% and a standard deviation 1.0 log10, a sample size of 16 knees in each group is required, necessitating the inclusion of 16 patients . To recruit this number of patients, an inclusion period of 6 months is anticipated.

Randomization

Knee allocation to the draping vs. non-draping group will be randomized. Randomization ensures that difference in bacterial flora between the patients knees has minimal effect on outcome. Potential allocation bias is minimized.

In this study the allocation will be assigned using envelopes containing numbered paper sheets. The quality of the envelopes is assessed to ensure that allocation cannot be determined without opening the envelopes. 30 envelopes will be prepared, determining allocation. An external party will prepare the envelopes.

Even numbers allocates the participant to draping of the right knee, while odd numbers allocate to left-side draping. The envelopes will contain even quantity of even and odd numbers.

We do not suspect that knowledge of knee allocation is important for study outcome. It is therefore not deemed necessary to blind participants or researchers from the intervention.

Intervention

The procedure will take place in the surgical section of the orthopaedic ward of Odense University hospital. Patients are greeted and asked to change to local patient clothing.

Patients are positioned supine in a standard operating theatre, with active laminar flow and surgical lamps . An operating drape will be set up at the level of the patient's waist to ensure minimal contamination by the patients breathing. Both knees are disinfected using a curcumin coloured solution of 0,5% chlorhexidine in 96% ethanol applied in a painting motion in 2 sessions of 10 seconds separated by a 30 second drying period. The disinfected area is then left to completely dry. Disinfection is followed by standard covering of the areas adjacent to the surgical site using surgical drapes. One knee is subsequently draped with an iodine plastic adhesive drape (Ioban 2), while the other knee is left uncovered. Application of the drape is done in accordance with 3M guidelines.(11) If the participant has long hair on the legs, these are trimmed using electrical clippers prior to the intervention. The patient is then asked to remain supine on the operating table for 75 minutes. Following disinfection and subsequent draping of both knees, the operating table is placed the beach-chair position and the patient is offered a pillow to support their knees. This is placed under the surgical drapes. The beach-chair position minimizes the discomfort of having to stay supine by elevating the patient's head and ensuring some flexion of the knees. The patient is allowed to move their legs in a controlled manner, so as not to contaminate the contralateral knee. Both knees must be moved equally to ensure equal stress on the skin of the sampling site during intervention. The patient´s torso is covered with a "Bair Hugger" blanket to prevent hypothermia.

The researcher and nurse will be wearing hairnet, sterile operation garments, sterile gloves and facemask during disinfection and application of draping. The project nurse, with experience in applying the draping, will do so in conformity with standard hospital procedure.

Skin sampling

Many methods of bacteriological skin sampling have been used in earlier studies. Since the outcome of this study is to compare mean bacterial quantity, we decide to use a modified version of the cup-scrub technique/cylinder sampling method, which is a standard method for testing preoperative skin preparation.(21, 29, 30)

This method samples the skin using a cylinder restricting the area of sampling to a small area covered by the drape. 1 ml of the sampling solution is applied to the sampling cylinder. The sample site is then rubbed with moderate pressure in an even grid formation with at sterile, plastic inoculation loop for 1 minute. After rubbing, the sampling solution is collected in a sterile test tube using a sterile pipette. The sampling process is then repeated, yielding another sample, which is pooled with the former.

By using two scrubs instead of one, the percentage of organisms removed from the sample site is increased, while scrubbing the cutaneous surface appears to be a vital part of the sampling process. Even though the majority of organisms are located on or near the surface of the skin, some have adherent properties, while others are situated in deeper parts of the horny layer and hair follicles. This might explain why washing of the sample site without scrubbing is inferior to the scrub-method, only yielding under half of the organisms sampled by scrubbing (30). The efficiency of this method has been evaluated by other authors and is seen as an appropriate approach to skin sampling when the goal is to evaluate both superficial and resident/"hidden" flora.(5, 6, 31)

We acknowledge that sampling of skin bacteria will only yield a fraction of the total flora due to the relative inaccessibility of the deep residing flora. (6)

Skin samples are taken bilaterally from the distal anterior thigh prior to disinfection to assess normal bacterial quantity. Following disinfection and complete evaporation of any alcohols, baseline samples will be taken from both knees in order to document equal bacterial quantity prior to draping. After the duration of the intervention (75 minutes), post-intervention samples will be taken from the no-drape knee, followed by opening of the Ioban draping over he sampling areas using sterile a scissor and surgical forceps. The skin is then sampled. Samples will be taken from a suprapatellar skin area, corresponding to the skin covering the borders of the quadriceps femoris muscle tendon and from an infrapatellar area of skin covering borders of the patellar ligament. These areas correspond to the skin bordering the incision line of the medial para-patellar approach commonly used in TKA surgery. The precise area for sampling in these regions is chosen at sampling time to enable minimal leakage between the skin and the sampling cylinder.

3 plates containing 5% blood agar are placed on a table in a distance of maximum 1 m from the knees. These plates will be uncovered and open to inoculation by the microorganisms in the air of the operating theatre at the start of the intervention time of 75 minutes following draping. The plates are covered at the end of the intervention, following final knee sampling. An unused sample of the sampling fluid is plated in duplicate and will serve as a negative control.

The same researcher will perform all sampling, thereby ensuring consistency in the sampling routine. A nurse aids in the sampling process. The control knee will be sampled first to prevent pre-sampling contamination of the contralateral knee, when the self-adhesive draping is removed.

The researcher and nurse will be wearing hairnet, sterile operation garments, sterile gloves and a facemask during sampling.

Sample solution and neutralizer system

The skin is sampled using a sampling solution consisting of phosphate buffer solution (PBS) containing 10% polysorbate 80 (Tween 80), 1% lecithin, 0,5% Sodium thiosulfate pentahydrate and 0,1% Triton-x 100 with pH adjusted to 7,9. (21, 32-35)

The neutralizer system is to be validated in accordance to the American Society for Testing and Materials (ASTM) standard E-1054-08 "Standard test methods for evaluation of in-activators of antimicrobial agents" using Staph. epidermidis as the challenge organism with an exposure time of 30 minutes.

Sample handling

Following sampling, pooled samples are immediately plated in the operating theater. This ensures optimal bacterial viability in the presence of triton-x and carry-over contamination of the skin antiseptics, which has been shown to inhibit bacterial growth even in very small concentrations(33, 36-38). A validation study on properties of the sampling fluid showed no inhibition of microorganism growth when plating was performed within 30 minutes of sampling.

Inoculated plates are then transported to the microbiological department of OUH for incubation after the end of the intervention.

Knee samples will be serially diluted in a sterile phosphate buffer in ten-fold dilutions to 10-3 for pre-disinfection samples and 10-1 for post-disinfection samples. Aliquots of 100-200 microliters of each diluted and undiluted specimen is then spread in duplicate on 5% blood agar plates. All plates are then aerobically incubated for 36-48 hours at 35 degrees Celsius. The air-inoculated plates and negative control plates are incubated along with the knee samples for 36-48 hours at 35 degrees Celsius.

Samples will be marked in accordance to the coding system described in the data management section of this protocol.

The number of colony forming units, CFUs, in the samples is then determined. N. Milandt will perform CFU counting with assistance from staff of the microbiological department of OUH. All plates will be photographed so that subsequent recounting can be performed if necessary.

Facilities

The Orthopaedic ward of The Department of Orthopaedic Surgery, Odense University Hospital (OUH), will supply operating theatres for the project.

The Department of Clinical Microbiology, OUH, supplies laboratory facilities for microbial analysis.

Ethics

We anticipate no complications or side effect to the intervention of this study. Being positioned in a supine position for 75 minutes may produce minor discomfort and can result in back pain for some patients. Removal of the Ioban draping can be associated with discomfort caused by adherence to the hair of the lower extremity. Application of disinfection solution containing curcumin, may result in staining on the skin that can be difficult to remove completely immediately following the intervention. This discolouration will however disappear within a few days with normal shedding of the epidermis.

Information regarding the project and its intervention will be presented to the participants verbally and in writing. Patients will first be contacted verbally and handed written information about the project. Patients expressing interest in participating will later receive verbal information, given by the main researcher. If the patient needs additional information regarding the project, they are supplied with the contact information of the project nurse and main researcher.

Undergraduate N. Milandt has taken the initiative for the project in cooperation with Professor, MD, S. Overgaard.

No members of this researcher group have any financial ties to any enterprises that sell or produce Iodine-impregnated self-adhesive draping.

Participants of this study will not receive any compensation for their participation. Participants will however be offered fruit, snacks and a non-alcoholic beverage following the intervention and will be eligible for reimbursement of transportation expenses for transportation up to 100 km round-trip in conformity with current regional agreements.

This protocol has been submitted to The Regional Scientific Ethical Committees for Southern Denmark, and was accepted in June 2014.

Results

Data management

Data registration

All patient data is registered and saved using Epidata. One document contains all patient information linked to a unique identification (ID) number. This key-file will contain the social security number and name of all participants along other variables of interest.

Results from the microbiological analysis are recorded in a separate document only containing ID numbers. The analysis result will be assigned a code that designate patient ID nr (1,2,3...), Sample time (Pre- (0), 10 minutes (10) or 75 minutes (75) following disinfection.). The code will also contain information about leg randomization ((Drape (D) / No-drape (N)) and sampling site (Suprapatellar (S) / Infrapatellar (I)). All samples are plated in duplicate and an additional number will therefore be assign to identify the unique plate and the dilution step for the plate. The unique code is assigned to corresponding plate with sticker labels and will have a corresponding variable in the data entry form.

All data will by recorded without arithmetic alteration.

All entry variables are documented in a codebook.

Encryption

All data is encrypted using the integrated Epidata encryption module based on Advanced Encryption Standard encryption protocols. The key-file linking ID numbers to social security numbers will be encrypted using this standard. Plate photographs are saved only containing the patients assigned code and cannot be linked to an identifiable individual without unlocking the main data files. The sample data file will not contain any person sensitive information and will therefore not be encrypted.

Backup

Back up of all project documents and files are performed weekly to a secure server of the computer system of Odense University Hospital. N. Milandt will have sole access to the backed up data.

N. Milandt will serve as contact person and will have the main responsibility for the project data. All Data will be stored and published in accordance with the decisions of The Danish Data Protection Agency.

Outcomes and statistical inference

Cleaning of data

In order to ensure valid outcomes when data is log transformed, the minimal value of measurement is set to 1 resulting in 0 log10 following transformation. A duplicate mean of <1 colony forming units (CFU)/plate is set to 0 log10 as a result. The sensitivity of the spread plate analysis may be lacking in distinguishing between 0 and 1 CFU and merging these values as a common flooring value of log10 = 0 may therefore be appropriate.

Bacterial density of the samples performed after disinfection is not expected to exceed the sample density prior to disinfection. If this is the case the observation is likely due to contamination and is excluded from analysis.

Primary outcomes

The primary outcome of this study is the number of CFU grown on the agar plates. This number includes all viable organisms. Raw data will be converted to CFU/cm2 of skin and then log10 transformed in order to allow assumption of linearity. A mean, representing the entire knee will be calculated from the infrapatellar and suprapatellar samples.

The mean surviving CFU/cm2 are then compared. Comparison will be performed within the same knee and between knees in the same patient. Samples from one knee are compared at different sampling times (baseline, 10 minutes after disinfection and 75 minutes after start of simulated surgery). Analysis between knees will be performed between the drape group and the non-drape group at all sampling times. Finally, a mean bacterial regrowth will be calculated as a variable representing the increase in bacterial density between sampling performed before and after the use of Ioban drape. The regrowth of bacteria in the two settings are then compared.

Data is examined using an exploratory data analysis to justify the assumption of Gaussian distribution. If this is the case, the analysis will be performed using a two-tailed students t-test.

If the data does not approximate normal distribution, the nonparametric Mann-Whitney U test will be used. P ≤0,05 is interpreted as significant. Statistical analysis will be performed using the data analysis and statistical software, STATA.

All analytical processes are documented in a data management logbook and a STATA do-file

Secondary outcome

The type of organism present on the skin following the disinfection and the use of Ioban draping is also of interest in clinical practise. It is, however, not the scope of this study to produce at detailed quantitative analysis of the frequency of various organisms on the skin. We will however perform a minor descriptive examination of the skin flora following the use of Ioban draping.

The undiluted growth plates from the 75 min samples of every participant will be macroscopically examined for unique colonies. Both the draped knee and undraped knee is analysed. The colonies are then identified using Matrix-assisted laser desorption/ionization Time of Flight analysis (MALDI-TOF). Findings will be listed as a prevalence of growth of specific organisms in the study population.

Time schedule

Planned project start is September 1, 2014. The project will be initiated by a data collection period of 9 months. During this period a minimum of 16 patients will be recruited and subjected to project intervention. The last 3 months of the project period will be dedicated the analysis of data and to production of an article and preparation of a project presentation. The project is scheduled to end August 31, 2015.

Comment from the Authors:

We acknowledge that revisions has been made to this trial protocol following primary trial registration. We declare that the changes to the protocol was made prior to study recruitment and data analysis. No change was made to the criteria for inclusion/exclusion, nor the outcomes measures of the study.

Inclusion Criteria:

• Patient is scheduled for primary knee replacement surgery (TKA).

Exclusion Criteria:

• Patient age under 50 year
• Patient is allergic to iodine
• Patient has received antibiotic treatment within 4 weeks of intervention.
• Patient suffers from a chronic skin disease with lesions in the knee area
• Patient has lesions of the skin due to trauma suffered within 1 week of intervention.

The intervention and control groups are the knees of the same patients assigned to either draping or no draping (split-person design). The avoid double-counting we report each unique participant here (each receiving both interventions). However in outcomes, each knee is considered as separate in the analysis.

5 months