Description of the condition
External fixation is a process by which pins or wires are inserted into bone fragments through small incisions in the skin, and then held together with an external clamp or framework. This means of treating fractures was proposed by Malgaigne in 1853 (cited Sisk 1983) as an alternative to immobilisation in a plaster cast, traction or internal fixation.
Percutaneous orthopaedic pins are metal rods, or wires, used to apply either skeletal traction, or support an external fixator device in the management of orthopaedic fractures or surgical procedures. They penetrate the skin, adjacent muscle and soft tissue, then enter bone and may penetrate through to the other side of a limb. Skeletal traction using a Steinmann pin does not totally immobilise bone fragments, but allows a force to be applied to a limb, thus holding it in a suitable position and allowing fracture healing to occur.
The early advocates of external fixation experienced considerable problems. The skin entry points, known as pin sites, frequently became infected and reported infection rates range between 1% for major infections to 80% for minor infections (Green 1983). Infection is undesirable as it can lead to failure of fixation with consequent loss of alignment of the fracture. It can also lead to osteomyelitis (inflammation of bone due to infection) and systemic infection, which may be both costly and difficult to treat. Therefore, external fixation was frequently criticised during the ensuing century (Sisk 1983), but its advantages, namely early mobilisation, axial loading of the fracture (along the normal line of load for the limb), easy observation of the limb, and access to the skin for wound care, have led to its continued use and development (Behrens 1988). External fixation is now an established treatment modality, and is extensively used to treat fractures either alone, or in combination with internal fixation (where screws and plates are used to hold bone fragments together beneath a surgical wound), traction or plaster cast (Blaiser 1997).
There is no uniformly accepted definition of pin site infection, which makes comparing infection rates difficult. Consensus opinion from experts on pin site management have attempted to differentiate between reaction, colonisation and infection (Lee-Smith 2001). Reaction was defined as the normal changes that occur at the pin site after pin insertion (i.e. changes in normal skin colour, skin warmth and drainage at the pin site); these are expected to subside after 72 hours. Colonisation was defined as warmth and red skin colour around the pin site, increased drainage, possible pain and the presence of microbes on culture. Infection includes all the changes seen with reaction and colonisation, together with possible pus, pin loosening and increased microbial growth. Two classification systems are used to grade the level of pin site infection (Checketts 1999; Sims 1996), but neither system has published validity, or reliability, results. In addition, pin site infections have been categorised as either major or minor (Ward 1998). Minor infections are considered to be benign, easily treatable with antibiotics and characterised by prolonged drainage, crusting, swelling and erythema (redness). Major infections require removal of one or more of the pins before the infection can be resolved (Lee-Smith 2001). The threat of infection remains a constraint to the use of external fixators.
Description of the intervention
It has been proposed that effective insertion techniques and subsequent nursing care of pin sites will reduce the frequency of pin site infections, loosening of fixation, and osteomyelitis (Green 1984; Kroll 1973; Sisk 1983). Many different regimens for pin site care have been described. These include regular cleansing with solutions such as hydrogen peroxide (Jones-Walton 1991), 0.9% normal saline or cooled boiled water (Sims 1996). At the Kurgan Ilizarov Institute in Russia, a very specific dressing regimen of 70% alcohol and 0.2% chlorhexidine has been prescribed immediately postoperatively. In this regimen, pins are never left uncovered, and care is carried out by hospital personnel on a weekly basis (Grant 1992). There is no consensus, however, on the optimal frequency of pin site care, which ranges from daily (Tolo 1983), to weekly (Ahlborg 1999), or even fortnightly (Grant 1992). Other uncertainties include whether or not scabs around the pin sites should be removed, and whether massage to promote drainage of exudate should be practised (Gordon 2000; McKenzie 1999; Sims 1996).
Mahan 1991 and Wissing 1988 have emphasised that a number of factors (including the surgical technique used to apply the fixator) might affect infection rates and might be more important than any local care. While experimental studies have shown the benefit in terms of either reduced pin loosening or pin site infection from silver-coated (Collinge 1994), tobramycin-impregnated (Voos 1999), or hydroxyapetite-coated pins (Moroni 1998), this aspect falls outside the scope of this review.
Why it is important to do this review
Most of the studies investigating this issue has come from research on animals or concern patient outcomes obtained from case series or descriptive studies. As it is unclear which care regimen is the most effective, a systematic review of the evidence has been undertaken.
To assess the evidence for the effects of cleansing, massage and dressing techniques for pin sites on postoperative infection.
Criteria for considering studies for this review
Types of studies
Randomised controlled trials which compare different methods of managing pin sites were included in this review.
For clarity, this review used the terms "pin (s)" rather than orthopaedic pins, percutaneous pins or fixator wires; "pin-site" rather than pin-tract; "external fixator" rather than external apparatus or framework or halo-frame. The review also considered material about skeletal traction.
Types of participants
Adults and children with pins inserted for either external fixators or skeletal traction. Studies of treatment regimens that set out to manage established infections were not included. Studies of people of any age and any care setting were included.
Types of interventions
A. Cleansing solutions
- Cleansing (any technique) compared with no cleansing.
- Comparisons between cleansing solutions (including sterile normal saline, alcohol solutions, iodine solutions, cooled boiled water or ordinary tap water).
B. Methods of cleansing
- Sterile technique compared with a non-sterile technique (e.g. a simple shower or washing with a soft toothbrush).
C. Primary dressing
- No dressing compared with any dressing.
- Comparisons between dressings (e.g. simple gauze, cut foam dressings and those made specifically by fixator manufacturers).
- Massage compared with no massage.
- Comparisons between different massage regimens.
- The management of 'skin-tenting' (slow return of the skin to its normal position) around pins.
Comparisons between specific comprehensive regimens of treatment such as the Kurgan Ilizarov method were to be included.
Types of outcome measures
Studies reported their outcomes in terms of:
- Incidence of infection (as classified by Sims 1996) or other documented method
The definition of infection given by Sims 1996 was used in the review, where possible. Although it does not quantify bacterial counts, it does provide a qualitative measure of infection.
- Grade 1: responds to local treatment, increased cleaning and massage (local manipulation of the skin to prevent adherence to the pin and allow drainage).
- Grade 2: responds to oral antibiotics.
- Grade 3: responds to intravenous antibiotics.
- Grade 4: responds to removal of pin.
- Grade 5: removal of pin and surgery required to control infection.
- Grade 6: chronic osteomyelitis (unresponsive to treatment).
We took the definitions of infection used in the trial reports at face value.
- Frequency of pin re-siting due to infection.
- Frequency of external fixator apparatus removal due to infection.
- Patient comfort, as defined by the patient's expression of discomfort at the pin site.
- Patient acceptability, as being defined by the patient's refusal to accept the treatment any longer.
- Duration of treatment (i.e. time to pin removal).
- Duration of overall treatment time (i.e. any prolongation of treatment beyond the norm for this type of fixation).
- Cost of treatment regimen, in particular as linked to the treatment of pin site infections.
- Limb amputation.
- All cause mortality.
Search methods for identification of studies
In September 2013, for the third update of this review, we searched the following electronic databases:
- The Cochrane Wounds Group Specialised Register (searched 5 September 2013);
- The Cochrane Central Register of Controlled Trials (CENTRAL) (The Cochrane Library 2013, Issue 8);
- Ovid MEDLINE (2011 to August Week 4 2013);
- Ovid MEDLINE (In-Process & Other Non-Indexed Citations, September 04, 2013);
- Ovid EMBASE (2011 to 2013 Week 35);
- EBSCO CINAHL (2011 to 29 August 2013).
The search strategies for Central, Ovid MEDLINE, Ovid EMBASE and EBSCO CINAHL can be found in Appendix 1,Appendix 2, Appendix 3 and Appendix 4 respectively. The Ovid MEDLINE search was combined with the Cochrane Highly Sensitive Search Strategy for identifying randomised trials in MEDLINE: sensitivity- and precision-maximizing version (2008 revision); Ovid format (Lefebvre 2011). The EMBASE and CINAHL searches were combined with the trial filters developed by the Scottish Intercollegiate Guidelines Network (SIGN 2008). There was no restriction by language or publication status.
Searching other resources
For the original review the following journals were handsearched: Orthopaedic Clinics of North America July 2000, Oct 2000, April 2001; Current Orthopaedics 1999-2000; Orthopaedic Nursing January 2000 to June 2001. The handsearching produced no additional studies and was not continued.
For all versions of this review, citations within obtained reviews and papers were scrutinised to identify additional studies.
Data collection and analysis
Selection of studies
Titles and abstracts of studies identified by the search strategy were assessed initially in terms of their relevance and design, according to the selection criteria, by the review author in charge of updating the review (AL). Copies of all relevant and potentially relevant papers were obtained, if they satisfied the inclusion criteria of this initial assessment. This process was repeated independently by the primary original author (JT), and selection results compared. Where there was disagreement over the selection of studies that could not be resolved by discussion, a third review author (JS) made the final decision. While papers published in languages other than English will be translated in sufficient detail to determine whether they are randomised trials, no non-English papers were identified for this update.
Data extraction and management
In addition to the assessment of risk of bias, other data were extracted from each study as follows.
- Study design.
- Number of patients randomised, excluded or lost to follow-up.
- Whether an intention-to-treat analysis was done.
- Whether a power calculation was done.
- Duration, timing and location of the study.
- Number of centres.
- Source of funding.
Characteristics of the study participants
- Age and any other recorded characteristics of participants in the study.
- Other inclusion criteria.
- Exclusion criteria.
- Type of surgery.
- Source of participants.
- Proportion of those eligible.
- Number of pin sites per participant.
- Types of cleansing used (if any).
- Methods of cleansing used (e.g. type of technique).
- Types of dressings used (if any).
- Methods used to measure infection.
- Methods used to measure patient comfort.
- Methods used to determine the norm in treatment time for pin fixation.
Assessment of risk of bias in included studies
All assessments of the risk of bias and data extraction were performed independently by two review authors (AL and JT), using forms designed according to Cochrane guidelines. Two of the review authors (JT and JS) are experts on clinical issues. The other review author (AL) has statistical and methodological expertise. Where necessary, additional information on trial methodology or original trial data were sought from the principal or corresponding author of any trials that appeared to meet the eligibility criteria (see Acknowledgements for details of the authors who provided additional clarification of data beyond what was reported in the publications).
Included trials were assessed for risk of bias using the risk of bias tool developed by the Cochrane Collaboration (Higgins 2011). The tool assesses:
- Sequence generation (whether the allocation sequence was adequately generated, for example using random number tables, computer random number generator, coin tossing, or shuffling).
- Allocation concealment (whether the allocation was adequately concealed, for example using central randomisation, or serially numbered, opaque, sealed envelopes).
- Blinding of participants, personnel and outcome assessors (whether knowledge of the allocated intervention was adequately prevented during the study by reporting the method of blinding, or by reporting that the participants, personnel and/or outcome assessors were blinded or, if the study was unblinded, whether knowledge of the intervention was unlikely to cause bias).
- Incomplete outcome data (whether incomplete outcome data were adequately addressed, for example whether the reasons for missing data that are unlikely to be related to the true outcome, or missing outcome data, were balanced in numbers across intervention groups).
- Selective outcome reporting (whether the reports of the study were free of suggestion of selective outcome reporting, for example, the publishing of a prior protocol for the study listing the prespecified outcomes for the study, or any other evidence that outcomes have not been selectively reported).
- Other sources of bias (whether the study was apparently free of other problems that could put it at a high risk of bias, e.g. baseline imbalance, bias related to study design, early termination of study, unit of analysis errors).
The criteria for assessment of each of these parameters for each trial was as follows:
- Low risk of bias - for example, the allocation sequence was adequately generated using random number tables.
- Unclear risk of bias - for example, it was unclear whether blinded outcome assessment had been undertaken in the study.
- High risk of bias - for example, allocation was not concealed, as report stated that an open list was used.
Details of the risk of bias for each included study are in Characteristics of included studies.
Assessment of heterogeneity
A priori, it was decided that results from the included studies would be combined in meta-analyses only if there were similarities in the included studies. Clinical heterogeneity was explored by close examination of the characteristics of the participants, interventions, outcomes and duration of the included studies. It was intended that, if there was no evidence of clinical heterogeneity, studies would be pooled and the forest plots inspected for statistical heterogeneity by reference to the chi-squared test (with a cut off level of P value 0.10) and the I
Data were collected on the unit of randomisation; that is, the participants in the trials or the pin sites, where this information was provided. Statistical analysis was performed in accordance with the guidelines for statistical analysis in the Cochrane Handbook for Systematic Reviews of Interventions (Deeks 2011).
If meta-analysis was appropriate, data were entered and analysed using Cochrane Revman software and results presented with 95% confidence intervals (CI). Estimates for dichotomous outcomes (e.g. proportion of participants who had infected sites) were reported as risk ratio (RR) with 95% CIs. Where outcomes were measured by different scales, it was intended that continuous data (for example, reduction in pain as measured by a visual analogue (continuous) scale (VAS) would be converted to the weighted mean difference (WMD) or standardised mean difference (SMD), and overall effect sizes with 95% CIs would be calculated. It was also intended that outcomes such as time to wound healing would be calculated as survival (time to event) data, using the appropriate analytical method, as suggested by the Cochrane Handbook (Deeks 2011).
Where meta-analysis was not appropriate, the results of the included studies were reported in narrative format.
A priori, it was planned to undertake sensitivity analyses and subgroup analyses to compare results from the pooled data according to differences in the quality of the trials, and differences in participants, interventions and outcomes.
Description of studies
Results of the search
In the first published version of this review in 2004, we identified a total of 21 potentially relevant randomised controlled trials (RCTs) and controlled clinical trials (CCTs) that considered any aspect of pin management. Eleven of these made reference to some aspects of pin site care, but only one trial met the criteria for inclusion and was included in the review (Henry 1996). For the updates of this review prior to 2013, a further seven RCTs met the inclusion criteria for the review (Camilo 2005; Cavusoglu 2009; Chan 2009; Egol 2006; Grant 2005; Patterson 2005; W-Dahl 2003). During the 2013 update of the review, a further 64 potential studies were identified, of which 3 appeared to meet the inclusion criteria. After independent selection processes, two of these studies were included (Lee 2012; Camathias 2012) and one study that was waiting assessment in previous versions of the review was also included (Yuenyongviwat 2011). One study (Lagerquist 2012) was excluded because it was a review.
A total of eleven RCTs with 572 participants were included in the review, although not all the participants contributed to the assessment of all comparative outcomes. Full descriptions of each of the included studies are provided in Characteristics of included studies.
Although in some trials data were collected using randomised participants as the unit of analysis, seven trials conducted analyses of the severity of pin site infections and their frequency per pin or frequency per observation rather than per participant (as participants had multiple pin sites or observations), without making adjustments for within patient clustering (Cavusoglu 2009; Chan 2009; Grant 2005; Henry 1996; W-Dahl 2003; Lee 2012; Camathias 2012). The reported analyses from these trials must be viewed with caution because of unit of analysis errors. For these trials, no information was provided by the authors on the infection rate per participant.
There was significant heterogeneity in the types of participants (in particular, type and location of surgery and protocol for pin placement), types of regimens used for pin site care and outcomes (different methods for classifying infection). Meta-analysis was undertaken only for the comparison of cleansing regimens with no cleansing regimens (3 studies), sterile antiseptic cleansing with sterile non antiseptic cleansing (3 studies), and sterile cleansing with non sterile cleansing (2 studies). For all other comparisons, meta-analysis was not considered appropriate because of extreme heterogeneity. The findings from each trial are discussed below and summarised in Table 1.
Sensitivity analyses were not undertaken, either because the data were not pooled, or all the studies were at high risk of bias.
This double blinded RCT recruited 56 participants from a hospital in the Soloman Islands. Consecutive patients who were treated with a standard external fixator (ExFix) in the lower or upper extremity over 2 years were included. They remained inpatients for the duration of their treatment. Antibiotics were given intravenously as soon as possible and continued for a total of 48 hours in all cases. The ExFix constructs normally had 4 pins, 2 proximal and 2 distal and randomisation was applied to the pins according to location rather than the patients, resulting in the patients being their own control. The pins were randomised to daily care regimen (removal of crusts, sterile saline irrigation, drying with sterile swabs and sterile dressing pre-soaked in povidone-iodine with dry dressing wrap) or no pin site care. The outcomes included condition of the soft tissue interface, stability of the pins, torsional stability of the pins as determined with a torque metre during their removal, osteolysis on pre-removal radiographs and pin site pain. The authors noted that observed inflammatory changes may not necessarily represent an infection but they stated that microbiological swabs were not reliable in the diagnosis of pin site infection as they may represent normal skin colonisation.
Camilo 2005 recruited 30 participants (average age 30.5 years) with Ilizarov external fixation apparatus from two hospitals in Brazil. The same team of surgeons inserted the apparatus in both settings. The entry point to the study was the first postoperative follow-up, although the researchers did not state how many days after surgery this was, or whether this was consistent across the sample groups. Patients with signs of active infection in the same limb as the external fixator at the entry point were excluded. Patients in both treatment arms of the study were instructed to apply the wound care protocol after a shower, after washing their hands with clean water and soap. In the control group, the skin around each pin site was cleaned with sterile gauze soaked in 0.9% saline solution to remove all 'dirt'. The sites were then dried with sterile gauze and each site was covered with folded gauze. The experimental group followed the same protocol except that, in addition to all other aspects, gauze soaked in polyvinylpyrrolidone-iodine (PVPI) was applied to each site.
Observation of each of the pin sites for the presence, or absence, of signs and symptoms of infection took place at each outpatient visit. The presence of purulent exudate (pus) was the criterion adopted for defining the presence of infection. Any drainage fluid or purulent exudate was collected for culture according to a described protocol. Follow-up continued until the fixator was removed (mean of 273 days; range 95 to 276 days).
The RCT by Cavusoglu 2009 compared two different pin site care protocols in 39 patients in Turkey: (1) pin care with daily showering and brushing the pin sites with soap and a toothbrush, and (2) pin site care with daily showering and cleaning crusts with sterile gauze impregnated with iodine solution. All patients had tibial or fibula fractures with Ilizarov external fixation, and used sterile gauze with iodine solution for the first 15 days before continuing with the two separate protocols. No dressing was applied throughout the period. Pin sites were inspected and graded on a scale of 0 to 5 according to a modification of the system of Dahl, described by Gordon 2000, during multiple follow-up periods from 5 to 150 days after surgery, until removal. Grade 1 and 2 infections were categorized as minor and grade 3 or more as major. Systemic antibiotics were prescribed for major infections.
The RCT by Chan 2009 compared infection rates in 62 patients in Malaysia (according to a grading system developed in-house) associated with two dressing solutions, diluted povidone-iodine and saline. Participants (46 males and 16 females; children and adults) had undergone distraction osteogenesis (bone lengthening) with external fixators and follow-up was every two weeks for six months. The fixations used were either rigid stainless steel 5 mm diameter half pins (with one metal-skin interface), or smooth stainless steel 1.8 mm diameter wires (with two interfaces).
Egol 2006 recruited 118 adults (mean age 54 years), with 120 distal radial fractures, from two hospitals in the USA, and allocated them to one of three treatment groups one week after surgery: (1) daily pin site care with a solution of 1/2 normal saline and 1/2 hydrogen peroxide, (2) weekly application of chlorhexidine-impregnated dressings (Biopatch) by the treating surgeon, or (3) weekly dry dressing changes without pin site care. One of the objectives of the study was to ascertain if the chlorhexidine-impregnated foam disc provided a significant reduction in pin site infection. Surgery and care after surgery was performed according to a specific protocol. All patients received three doses of first-generation cephalosporin antibiotics, one dose before and two after surgery. Patients undertook pin site care after discharge from hospital. The groups were equivalent at baseline in terms of age or mechanism of injury, but not equivalent in terms of gender. Patients returned to the clinic for weekly evaluation and X-rays were performed every two weeks. It is not clear how infection was identified, although all infections were treated with antibiotics, so antibiotic use can be used as a surrogate outcome for rate of infection. The external fixator was removed at six weeks after surgery (unless infection necessitated earlier removal), and patients were followed up for a minimum of six months. In addition to antibiotic use, patients were assessed weekly for cellulitis, erythema, drainage, pin loosening and whether pin removal was necessary before fracture healing due to infection.
Twenty patients (at least 18 years old), who required skeletal pins for acute injury, were recruited from a hospital in Australia for the small, randomised trial by Grant 2005. At 72 hours after surgery, participants were randomised to (1) cleansing with normal saline flush and application of soft white paraffin ointment, or (2) twice daily cleansing with normal saline and application of 10% povidone-iodine solution. Both groups underwent cleansing for 14 days, or until discharge, followed by daily care at home until either the pins were removed or infection occurred. The length of time the pins remained in situ ranged from four to 120 days. The number of pin sites per patient ranged from two to 17. Patients managed their own care, and follow-up occurred once a week after discharge from hospital. Prophylactic antibiotics were prescribed either during surgery or immediately after for 83% of patients. Clinical signs of infection (defined as redness, induration (skin hardening), haemo-serous ooze and pain), and requirement for antibiotic use were measured.
Henry 1996 was a randomised trial of pin site care in the UK. Thirty women aged 11 to 18 years with 120 pin sites, who were undergoing leg-lengthening surgery, had daily pin site care according to a specified protocol. They were randomised to cleansing with either (1) 0.9% saline; (2) 70% alcohol; or (3) no cleansing. Crust removal, gentle massage, spraying with dry povidone iodine and dressing with dry gauze was undertaken at all sites. The majority of sites were in the femur or tibia. Infection was defined by pain, redness and swelling and the culturing of a significant number of pathogenic bacteria. The length of time the pins were in situ ranged from 56 to 244 days (mean 50 days).
This randomised trial conducted in Malaysia (38 participants) compared pin site dressings of gauze impregnated with 0.2% polyhexamethylene biguanide with plain gauze (control) in 38 patients scheduled for limb lengthening or deformity correction. Rates of pin site infection were assessed blindly at multiple follow up times (2, 4, 8 and 12 weeks) according to the grading system suggested by Saw 2006. The patients and their caretakers were taught how to perform pin site dressing according to a standard protocol used in the hospital. The dressings appeared identical and both patients and assessors were blinded. The study was designed to look at infection rates per observations rather than per number of pin sites or per patient. The researchers also compared infection rates between the two groups according to infection grades (1, 2 or 3) and between the two groups according to types of bone fixation (wire-skin and pin-skin interfaces).
Patterson 2005 was a pilot study, exploring the differences in infection rate and reaction rate among seven different pin care protocols that varied both the cleansing agent (half-strength peroxide cleansing, saline cleansing, or antibacterial soap and water cleansing) and dressing type used (stable gauze or sponge, or Xeroform/Xeroflo) in 92 participants. The control group had no cleansing, and a dry dressing which was only changed if it became wet or soiled. Participants were recruited from two centres in the USA. Two-thirds of participants were male, 31% were children or adolescents; the most common diagnosis was closed fracture with lower extremity involvement, and each participant had an average of six pins. Patients undertook pin site care after discharge from hospital.
Pin sites were evaluated for redness, swelling, discomfort, tenting, loosening of pins, crusting and drainage by a rating scale designed to evaluate pin site reaction. This scale provided the criteria for determining a stage II or stage III pin site infection. Data were collected over six weeks.
The randomised controlled trial by W-Dahl 2003 compared (1) daily cleansing of pins sites with 0.9% saline solution and a dry dressing and bandage, to (2) the same procedure conducted weekly. A sample of 50 adult patients (mean age 54 years), who had undergone elective surgery for gonarthrosis by the hemicallatasis technique and insertion of external fixation apparatus, were recruited from a hospital in Sweden. Participants had an average of four hydroxyapatite-coated pins each. The treatment regimen was carried out by the hospital-based staff and any one of several district nurses after discharge. In hospital, a technique with sterile dressings and sterile gloves was used, and at home or in outpatients the sterile gloves were replaced with clean gloves. Patients could shower, provided they protected the whole device in a plastic bag. Outcomes were evaluated weekly by district nurses. As well as bacterial cultures of each pin, a published assessment tool was used for the identification of pin site infection (Checketts 1999), though the reliability and validity of this tool has not been clearly established. The criteria rate minor infections as grades I to lll and major infections as grades lV to VI, which are equivalent to Sims & Saleh 1996 grades (Sims 1996). The researchers also considered the pain levels experienced by the patients (using a Visual Analogue Scale (VAS)), antibiotic use, whether the pins were still tightly fixed (or could be removed by hand), and all pin tips were cultured on removal.
This randomised trial compared the prevalence of pin site infection in 30 patients, in Thailand, with open tibial fracture requiring external fixation and randomised to either daily pin site dressing with normal saline and 0.5ml of 1% silver sulfadiazine or daily dry dressing with optional removal of dry scale by patients. Assessments were made monthly at an outpatient clinic by an independent but unblinded observer according to the Checketts 1999 classification. After discharge from hospital, patients carried out their own pin site care. Mean duration of follow up was 106 and 109 days.
In the first published version of this review, we excluded 10 trials (Bednar 1996; Botte 1989; Hutchinson 2000; Kapoor 2000; Masse 2000; Moroni 2001a; Moroni 2001b; Sproles 1985; Turcic 1998; Vossinakis 2002) as they considered some aspects of pin form, pin function or infection but did not use pin site care (n=9) or were not properly randomised (n=1). A further three studies retrieved during the updating of the review prior to 2013 indicated that allocation to groups was not randomised, and these trials were also excluded (Bhattacharya 2006; Davies 2005; W-Dahl 2004). During the 2013 update of the review, one further study was excluded because it was a systematic review rather than a randomised controlled trial (Lagerquist 2012). All reasons for exclusion are detailed in the Characteristics of excluded studies table.
Risk of bias in included studies
|Figure 1. Methodological quality graph: review authors' judgements about each methodological quality item presented as percentages across all included studies.|
|Figure 2. Methodological quality summary: review authors' judgements about each methodological quality item for each included study.|
Six of the eleven included studies provided sufficient detail of the adequacy of the randomisation method (Cavusoglu 2009; Egol 2006; Henry 1996; W-Dahl 2003; Lee 2012; Camathias 2012). Participants in these studies were randomised by random number tables or randomised schedules generated by computer. Two studies provided evidence that allocation had been adequately concealed (Camilo 2005; Egol 2006). These studies used sealed, consecutively numbered envelopes. The remaining studies did not report clearly on allocation and were considered at unclear risk of bias for allocation (either randomisation method or allocation concealment).
Two studies had blinding of both participants and assessors, with the use of identical dressings (Camathias 2012; Lee 2012). None of the other studies appeared to be blinded (investigators and participants) and there was also no indication whether the assessment of outcomes was also blinded.
Incomplete outcome data
Eight of the eleven included studies did not report any loss to follow-up or exclusions after randomisation (one study confirmed this after personal communication). The other three studies reported withdrawals of 10% (Grant 2005), 9% (Patterson 2005) and 3% (Chan 2009). In the Grant study, all withdrawals were from the soft white paraffin group.
All studies except two reported all prespecified outcomes and were considered at low risk of bias. The outcomes in one study (Henry 1996) were not reported clearly and this study was considered at high risk of bias. The other study (Camathias 2012) reported on all prespecified outcomes but did not provide quantitative estimates for some outcomes and was considered at unclear risk of bias.
Other potential sources of bias
Baseline imbalance was reported in three trials (Camilo 2005; Cavusoglu 2009; Egol 2006); for five other trials, comparability of groups at baseline could not be determined (Chan 2009; Henry 1996; Patterson 2005; Yuenyongviwat 2011; Camathias 2012); and for the three remaining trials, baseline demographic data indicated no differences between randomised groups (Grant 2005; W-Dahl 2003; Lee 2012). Eight studies, where randomised participants had multiple pin sites, presented analyses per pin site or pin site location without adjustment for lack of independence (Cavusoglu 2009; Chan 2009; Egol 2006; Grant 2005; Henry 1996; Patterson 2005; W-Dahl 2003; Camathias 2012). Patterson, however, did provide analyses per randomised participant, and these data have been used in this review, while Egol included 118 participants with 120 pin sites, so it was judged that there was minimal effect of unit of analysis errors in this study. Such unit of analysis errors have the potential to overestimate effects.
Three studies were considered at high risk of other potential sources of bias. In Grant 2005, a proportion of participants (83%) received prophylactic antibiotics in the intraoperative or postoperative period; the likelihood of confounding from differential antibiotic use in this trial cannot be excluded. In the Patterson 2005 trial, the possibility of confounding could not be excluded, as there was insufficient information provided about types of fixators, application of pins, insertion techniques and location of sites, all of which can affect infection rates differentially. One study, where participants were observed at multiple time points, presented analyses per observation without accounting for dependent data (Lee 2012).
In summary, the majority of the included studies had major flaws in methodological quality, which weakens the confidence we can place on the individual and pooled results.
Effects of interventions
It was possible to pool the results for cleansing compared with no cleansing (three studies); sterile antiseptic cleansing compared with sterile non-antiseptic cleansing (three studies); and sterile cleansing compared with non-sterile cleansing (two studies). Forest plots displayed the results of individual studies for the comparisons between dressings because of the heterogeneity of the interventions. Results are also reported individually in narrative form below, according to the outcome comparisons, and a summary is provided in Additional Table 1.
Cleansing compared with no cleansing
Three trials addressed the issue of whether cleansing solution of any type reduced pin site infection when compared with no cleansing (application of dry dressing without cleansing) (Egol 2006; Henry 1996; Patterson 2005).
Henry 1996 compared cleansing with either 0.9% saline (sodium chloride), or 70% alcohol, with a control group that had no cleansing procedure, but had daily pin site care with dry povidone-iodine spray, gentle massage and change of dressing. Patterson 2005 compared cleansing with half-strength hydrogen peroxide, saline or antibacterial soap and water with a control group that had no cleansing procedure (dressing was changed only if it became wet or soiled). In Egol 2006, one of the experimental groups had daily cleansing with a solution of one-half normal saline solution and one-half hydrogen peroxide, and the control group had the weekly application of a dry dressing with no cleansing. The other experimental group in this study used a specific type of dressing which is not relevant to this comparison.
Henry 1996 defined infection as pain, redness and swelling and a significant number of pathogenic bacteria on culture at the pin site with, or without, systemic symptoms. Infections occurred in 25% (10/40) and 18% (7/40) of those participants who had cleansing with either solution, and in 8% (3/40) of those in the control group who had no cleansing. There was no statistically significant difference in infection rates between cleansing or no-cleansing regimens. The infection rate was 35% higher in femoral than tibial pins, particularly in those situated near the groin, or on swollen thighs; this was reported incidentally and location of infection was not an objective of the study. All pins were massaged to remove crusts and none had to be surgically released because of tension. All pin sites were sprayed with povidone-iodine dry spray. Ninety-two per cent of the infections were due to Staphylococcus aureus and all resolved with oral antibiotics. There were 120 pin sites in the 30 participants in this study, and these were in situ for 56 to 244 days. The infection rate did not relate to the length of time pins were in situ.
Patterson 2005 assessed the effects of cleansing solution and type of dressing on pin site infections when compared with a control group with no cleansing. A ratings scale was piloted in this trial (from the Mercy Hospital in Kansas City, Missouri) which required scores on a number of different parameters: redness, swelling, discomfort, tenting, loosening of pins, crusting and drainage. The criteria for a stage II pin site infection were a rating score greater than three and requiring treatment with antibiotics. The criteria for a stage III infection were receiving a score greater than seven, treatment with IV antibiotics and/or removal of the pin. Data were collected over a follow-up time of six weeks to control for increasing infection rate associated with the length of time pins were in place. There was no significant difference in infection rates depending on treatment. Pin site infections - mostly stage II - developed in 27%, 30% and 45% of participants in the groups having cleansing and 36% of participants in the control group with no cleansing. Infection rate was also analysed per pin, but is not reported here because of unit of analysis errors with this method of analysis.
Egol 2006 evaluated pin site complications weekly (erythema, cellulitis, drainage, pin loosening, need for antibiotics and need for pin removal before fracture healing due to infection), and participants carried out their own care. Oral antibiotics were administered when participants developed erythema, cellulitis or pin-track infection (which is equivalent to the Sims and Saleh grade 2 categorisation) (Sims 1996), and pins were removed when the infection did not respond to antibiotics (equivalent to Sims and Saleh grade 4 categorisation). There was no significant difference in the rate of pin site complications, antibiotic use or requirement for early removal between the group that had cleansing and the control group, however, a lack of statistical power cannot be ruled out as a contributory factor in this result. Of those having cleansing with saline/hydrogen peroxide, 22.5% required treatment with antibiotics compared with 2.5% of those that were not cleansed (RR 9.00; 95% CI 1.20 to 67.77) ( Analysis 1.1); 7.5% of the former group required early fixator removal compared with 0% in the control group. Fixators were left in place for an average of six weeks, and all participants were followed up at six months.
In spite of the high heterogeneity between studies (I
In summary, there is insufficient evidence of an effect of pin site cleansing on pin site infection from three small studies at high risk of bias.
Comparisons of different cleansing solutions
Three trials compared the incidence of infection or pin site complications between cleansing solutions (Chan 2009; Henry 1996; Patterson 2005). Cleansing with saline 0.9% was compared with cleansing with 70% alcohol in Henry 1996, cleansing with half-strength peroxide, saline, or soap and water were compared by Patterson 2005, and cleansing with a diluted povidone-iodine solution was compared with saline in Chan 2009
In Henry 1996, infections occurred in 25% of those participants whose pins were cleansed with 0.9% saline (10 participants), and 18% of those whose pins were cleansed with 70% alcohol (seven participants). Patterson 2005 assessed the effects of cleansing solution, methods of cleansing (sterile versus non sterile) and type of dressing on infection rates. When the groups were combined according to the type of cleansing that was used, there was no evidence that infection rates differed (infection rate was 27% in participants having half-strength peroxide cleansing, and 30% in those having saline cleansing). Infection rates in Chan 2009 were almost identical in the iodine group and the saline group, at 19% and 17% respectively, and were 45% in those having soap and water cleansing. Studies were pooled in a meta-analysis comparing cleansing solutions with sterile antiseptic properties to those with sterile non-antiseptic properties.
The three trials that compared cleansing solutions had major methodological shortcomings, with small sample sizes, and there was no evidence of differences in infection rates in meta-analysis (RR 1.07; 95% CI 0.82 to 1.39) ( Analysis 2.1). Thus, there is insufficient evidence to determine whether one particular cleansing solution is more effective than any other.
Comparisons between different methods of cleansing
Two studies were identified that compared cleansing methods, such as sterile compared with non-sterile techniques (Cavusoglu 2009; Patterson 2005). Patterson 2005 compared multiple protocols; by combining some of these protocols, infection rates from cleansing with two sterile cleansing solutions, either half-strength peroxide or saline, could be compared with non-sterile soap and water. Patterson 2005 reported a 29% infection rate among all sterile protocols compared with a 45% infection rate for the two non-sterile protocols (these rates were not tested statistically in the publication). Cavusoglu 2009 compared daily showering and cleansing with sterile gauze containing a povidone iodine solution, with daily showering and brushing the pin sites with soap and a soft toothbrush; both minor and major infection rates were measured, but per site rather than per participant. There was no statistically significant difference in minor infection rates (RR 0.86; 95% CI 0.73 to 1.02; one study) ( Analysis 3.1). When the two studies (Cavusoglu 2009 and Patterson 2005) were pooled for the outcome of major infection rates (I
One study compared the effects of daily cleansing with weekly cleansing with 0.9% sodium chloride solution using a sterile technique in 50 patients having surgery for gonarthrosis by the hemicallotasis technique (W-Dahl 2003). There was no evidence of a difference in the frequency and severity of pin site infections (data were reported per pin, rather than per randomised participant), pain (mean VAS score 1.5 (SD 0.5) for daily care compared with mean VAS score 1.6 (SD 0.4) for weekly care) or the use of antibiotics (mean days 41 (SD 30) for daily care versus mean days 53 (SD 22) for weekly care) or analgesics. Unit of analysis errors resulting from the analysis of infection rates by pin sites limits the confidence that we can have in the results from this study.
Dressing compared with no dressing
One study compared daily pin site care with no pin site care (Camathias 2012). The pin site care regimen included removal of crusts, sterile saline irrigation, drying with sterile swabs, sterile dressing pre-soaked in povidone-iodine and dry dressing wrap. Most participants had 4 pins, 2 proximal and 2 distal and these groups of pins, proximal and distal, were randomised rather than the participants. The outcome of interest in this RCT was the condition of the soft tissue interface, categorised as either dry, or with secretions or granulation. The authors considered that dry pin sites were considered a success of treatment and secretions and granulation were considered a failure of treatment. They reported that inflammatory changes (secretions or formation of granulation tissue) did not necessarily represent an infection as studies had shown that microbiological swabs were not reliable in the diagnosis of pin tract infection, often representing normal skin colonisation. In pin sites not receiving treatment, 35% were considered not dry and in pin sites with daily care 36% were not dry (RR 1.02, 95% CI 0.70 to 1.48) ( Analysis 4.1). Analyses taking into account possible confounding factors such as age, sex, duration of fixation, location of pins and presence of infection in the primary wound did not find evidence of a significant difference in the paired pin sites according to treatments.There was also no evidence of a difference in pin site pain. No cases of osteomyelitis were identified in 2 years follow up. The authors concluded that routine pin site care is unnecessary in external fixation treatment of injuries.
Comparisons between different types of dressings
Six studies compared the effects of different types of dressings on pin site complications and infections (Camilo 2005; Egol 2006; Grant 2005; Patterson 2005; Lee 2012; Yuenyongviwat 2011). It was not possible to pool the studies as the dressing types were heterogeneous, but results are shown for each individual study in forest plots.
Egol 2006 compared the weekly application of a dry dressing to the weekly application (by the treating surgeon) of a chlorhexidine-impregnated patch (Biopatch). Grant 2005 compared twice daily application of 10% povidone-iodine solution (for two weeks or until hospital discharge followed by daily application) with daily application of sterile, soft white paraffin ointment. Camilo 2005 compared daily topical application of 10% polyvinylpyrrolidone-iodine (PVPI) solution on gauze dressing with dry gauze dressing. Patients were evaluated at outpatient return visits for identification of signs and symptoms of superficial infection at wire and pin insertion sites. Patterson 2005 compared the effects of the application of stable gauze or sponge twice a day, the application of 3% bismuth tribromophenate and white petroleum emulsion impregnated gauze (Xeroform/Xeroflo) twice a day and application of stable gauze or sponges which remained in situ (only changed if wet or soiled). Lee 2012 compared plain gauze dressing with gauze impregnated with polyhexamethylene (PHMB) in patients, with final assessments of infection rates made at 12 weeks after surgery. Yuenyongviwat 2011 compared dressings of saline and silver sulfadiazine with dry dressings where assessments were made monthly for up to 3 months.
In Egol 2006 5% of those having the Biopatch dressing and 2.5% of those having a dry dressing required antibiotics, but these rates were not compared statistically in the trial report (RR 2.00; 95% CI 0.19 to 21.18)( Analysis 5.1). There was also no evidence of a difference in the other outcomes: rates of erythema, cellulitis, drainage, or pin loosening between dressing types. No participants in the groups where dressings were compared required early fixator removal. Grant 2005 analysed effects on pin sites rather than the randomised unit of analysis, i.e. participants in the study. The authors reported that a greater proportion of pin sites in the soft white paraffin ointment group (34.1%) developed clinical signs of infection when compared with the povidone-iodine group (18.1%), but this difference did not reach the prespecified level of significance (P value less than 0.05) (RR 0.53; 95% CI 0.28 to 1.01) ( Analysis 5.2). There was no evidence of a difference in the incidence of superficial infection between groups (PVPI-impregnated gauze compared with dry gauze) in Camilo 2005 (66.7% versus 46.7%) (RR 1.43; 95% CI 0.75 to 2.73) ( Analysis 5.3). There was no significant difference in the rates of rehospitalisation due to infection (26.7% in the PVPI group and 6.7% in the dry gauze group), however, the study was underpowered. When groups were combined in the Patterson 2005 trial, those with gauze/sponge dressings had an infection rate of 38.5%, those with Xeroform had an infection rate of 31%, and those in the control group in which the gauze/sponge remained in situ had an infection rate of 36.5%. There was no evidence of a difference between groups (RR 0.8; 95% CI 0.44 to 1.47)( Analysis 5.4). The Lee 2012 study analysed the infection rate risk according to number of observations and these data could not be displayed in a forest plot because of unit of analysis errors. The study found that risk of pin site infection was significantly reduced with PHMB gauze when compared to plain gauze (RR 0.23, 95% CI 0.12 to 0.44). The infection rate was 1% in the PHMB group and 4.5% in the control group. In Yuenyongviwat 2011, there was no evidence of a significant difference in the infection rate between sliver sulfadiazine dressing and dry dressing (RR 1.17; 95% CI 0.51 to 2.66) ( Analysis 5.5).
Overall, it was not possible to determine whether any particular dressing was more effective than any other, given the heterogeneity and methodological shortcomings of the included studies.
There were no trials identified that compared the effects of massage regimens on pin site infections or complications.
Summary of main results
The studies included in this review were not able to determine an optimal strategy for pin site care.
It is not surprising that few differences were reported between pin site care regimens in the included studies, as few studies ensured that the sample size was adequately powered to find differences. Although the combination of dressings and cleansing agents was not a defined objective of this review, one study (Patterson 2005) found that patients having cleansing with hydrogen peroxide followed by the application of Xeroform dressing had a lower rate of infection (9%) than the other combinations of cleansing and dressing and the control group with no cleansing; however, no adjustments were made for multiple comparisons, controlling for age (which was found to significantly influence infection rate) and we cannot exclude the possibility that this is a chance finding. Another blinded study (Lee 2012) found that the risk of pin site infection was significantly reduced with polyhexamethylene biguanide (PHMB) gauze when compared to plain gauze (RR 0.23, 95% CI 0.12 to 0.44) (infection rate of 1% in the PHMB group and 4.5% in the control group). It is necessary that these findings be confirmed by further research before any recommendations can be made.
Quality of the evidence
Overall, study quality was poor. Although randomised, most studies were small and underpowered with methodological flaws, and only two were blinded. In most studies, minimal efforts were made to control for other factors that might have influenced the results, such as patient compliance with pin care protocols after discharge from hospital, antibiotic use, length of time pins were in place, location of pins, pin insertion procedures, types of hardware used, heterogeneity (variation) in the age of participants and condition being treated by external fixators. Some of the studies included in this review indicated that age of the participants and location of the pins were independently associated with the likelihood of infection. One blinded study of reasonable quality attempted to control for pin site location, age, sex, and duration of fixation, but did not directly measure infection rates. Other factors that may affect these rates are: active correction phase, pins compared with wires, fixator type, pin stability and type of pin coating. Moreover, the primary outcome, pin site infection, was not consistently defined in the included studies, and was often reported using pin sites or observations as the denominators without adjustment for unit of analysis errors.
Potential biases in the review process
Efforts were made to retrieve all eligible studies but it is not possible to exclude the likelihood that unpublished studies may have been missed.
Agreements and disagreements with other studies or reviews
The findings of this review are broadly in agreement with a systematic review of the literature resulting in a National Association of Orthopaedic Nurses Guidelines in pin-site care (Holmes 2005). This group assessed the evidence, and made four recommendations. Firstly, it was suggested that pins located in areas with considerable soft tissue should be considered at greater risk of infection. In this review, Henry 1996 reported higher rates of infection in femoral pins compared to tibial pins, and proximal femoral when compared with lower and middle pins. A recommendation was also made that after the first 48 to 72 hours, pin site care should be done daily, or weekly for sites with mechanically-stable bone-pin interfaces. W-Dahl 2003 reported no differences between daily or weekly pin site care. Thirdly, it was suggested that chlorhexidine 2 mg/ml solution may be the most effective cleansing solution for pin site care. This finding was based on the results of a non-randomised observational study (W-Dahl 2004). One study in this review found no differences in the rates of infection with chlorhexidine dressings when compared with the use of hydrogen peroxide care. This review has not been able to determine whether one particular cleansing agent is more effective at reducing infection than any another. Finally, the Guidelines recommend that patients and their families should be taught pin site care before discharge from the hospital. Since most patients with external bone fixators and pins are discharged from hospital before healing at the pin sites has occurred, compliance with a prescribed pin site regimen is an important factor that is likely to influence the rate of pin site complications and infections. Most of the studies included in this review explicitly described instructions for pin site care at home after discharge from hospital (using clean rather than sterile procedures), but there were no attempts to measure patient compliance with these instructions.
Another review (Lagerquist 2012) was also unable to reach definitive solutions on pin site care. This review included both randomised and non randomised studies. It also recommended that patients and their families should be taught pin site care, including signs and symptoms of infection, and noted that further research was required to reach more evidence based recommendations for care.
In conclusion, this review has not found sufficient evidence to recommend a particular strategy of pin site care. Adequately-powered, randomised trials are required to examine the effects of different pin care regimens and their co-interventions such as antibiotic use; furthermore, other extraneous factors must be controlled in the study designs. Any new trials should be reported in line with CONSORT guidance, and data regarding baseline distribution of prognostic factors, such as pre-drilling holes for pins, incising the pin sites prior to pin insertion and use of antimicrobial coating on pins should be clearly displayed, and adjustments made for imbalances (Begg 1996). If multiple pin sites are located on each randomised participant or multiple observations are made at different follow up times, adjustments should be made to prevent unit of analysis errors.
Implications for practice
There is insufficient evidence of adequate quality to inform a strategy for the best management of pin sites. In the absence of good evidence, the implementation of general strategies for minimising the risk of cross infection are advised.
Implications for research
Adequately-powered, well designed randomised controlled trials are required to determine optimal pin site management. Studies should focus on using a consistent method for identification of infection rates and randomisation should be per participant, rather than by pin site to avoid unit of analysis errors. Subgroup analyses should be used to control for participant age and location of pin sites, to avoid confounding.
The review authors thank the editorial base of the Cochrane Wounds Group for their support and advice on the review. The review authors would also like to thank the referees of the Cochrane Wounds Group (Bill Gillespie, Eileen Scott, Zena Moore, Michael Dean) and the editors (Andrea Nelson, David Margolis) for their comments on drafts of the review. The review authors would like to acknowledge the contribution of Elizabeth Royle who copy edited the updated review.
In addition, the authors would like to thank the authors of the following publications who supplied additional data which were either missing or not clear in the publications: Egol 2006; Grant 2005; W-Dahl 2003.
Data and analyses
- Top of page
- Authors' conclusions
- Data and analyses
- What's new
- Contributions of authors
- Declarations of interest
- Sources of support
- Differences between protocol and review
- Index terms
Appendix 1. Central search strategy
We used the following search strategy in the Cochrane Central Register of Controlled Trials (CENTRAL):
#1 MeSH descriptor External Fixators explode all trees
#2 MeSH descriptor Bone Nails explode all trees
#3 MeSH descriptor Bone Wires explode all trees
#4 MeSH descriptor Bone Screws explode all trees
#5 external NEAR/2 (fixat* or wire* or pin* or rod* or nail*)
#6 bone NEAR/2 (fixat* or wire* or pin* or rod* or nail*)
#7 fracture* NEAR/3 (fixat* or wire* or pin* or rod* or nail*)
#8 (orthopaedic or orthopedic) NEAR/2 (wire* or pin* or rod* or nail*)
#9 percutaneous NEAR/2 pin*
#10 steinmann NEXT pin*
#11 skeletal traction
#12 (#1 OR #2 OR #3 OR #4 OR #5 OR #6 OR #7 OR #8 OR #9 OR #10 OR #11)
#13 MeSH descriptor Detergents explode all trees
#14 MeSH descriptor Soaps explode all trees
#15 MeSH descriptor Alcohols explode all trees
#16 MeSH descriptor Iodine explode all trees
#17 MeSH descriptor Iodophors explode all trees
#18 MeSH descriptor Povidone-Iodine explode all trees
#19 MeSH descriptor Chlorhexidine explode all trees
#20 MeSH descriptor Disinfectants explode all trees
#21 MeSH descriptor Anti-Infective Agents, Local explode all trees
#22 MeSH descriptor Sodium Chloride explode all trees
#23 MeSH descriptor Peroxides explode all trees
#24 detergent* or soap* or alcohol or alcohols or iodine or iodophor* or
chlorhexidine or betadine or disinfectant* or saline or peroxide or
(boiled NEXT water) or (tap NEXT water)
#25 cleans* or shower* or wash* or bath*
#26 MeSH descriptor Bandages explode all trees
#27 MeSH descriptor Alginates explode all trees
#28 MeSH descriptor Hydrogels explode all trees
#29 dressing* or gauze or tulle or foam* or hydrocolloid* or alginat* or
#30 MeSH descriptor Massage explode all trees
#32 MeSH descriptor Postoperative Care explode all trees
#33 (post-operative or postoperative) NEXT care
#34 MeSH descriptor Surgical Wound Infection explode all trees with
#35 MeSH descriptor Infection Control explode all trees
#36 pin NEXT site
#37 (#13 OR #14 OR #15 OR #16 OR #17 OR #18 OR #19 OR #20 OR #21 OR #22
OR #23 OR #24 OR #25 OR #26 OR #27 OR #28 OR #29 OR #30 OR #31 OR #32 OR
#33 OR #34 OR #35 OR #36)
#38 (#12 AND #37)
The search strategies for Ovid MEDLINE, Ovid EMBASE and EBSCO CINAHL can be found in Appendix 2, Appendix 3 and Appendix 4 respectively. The Ovid MEDLINE search was combined with the Cochrane Highly Sensitive Search Strategy for identifying randomised trials in MEDLINE: sensitivity- and precision-maximizing version (2008 revision); Ovid format. The EMBASE and CINAHL searches were combined with the trial filters developed by the Scottish Intercollegiate Guidelines Network. There was no restriction by language or publication status. For the second update of the review, the electronic searches retrieved 31 records in total.
Appendix 2. Ovid MEDLINE search strategy
1 exp External Fixators/
2 exp Bone Nails/
3 exp Bone Wires/
4 exp Bone Screws/
5 (external adj (fixat$ or wire$ or pin$ or rod$ or nail$)).mp.
6 (bone adj2 (fixat$ or wire$ or pin$ or rod$ or nail$)).mp.
7 (fracture$ adj2 (fixat$ or wire$ or pin$ or rod$ or nail$)).mp.
8 ((orthopaedic or orthopedic) adj2 (wire$ or pin$ or rod$ or nail$)).mp.
9 (percutaneous adj2 pin$).mp.
10 steinmann pin$.mp.
11 skeletal traction.mp.
13 exp Detergents/
14 exp Soaps/
15 exp Alcohols/
16 exp Iodine/
17 exp Iodophors/
18 exp Povidone-Iodine/
19 exp Chlorhexidine/
20 exp Disinfectants/
21 exp Anti-Infective Agents, Local/
22 exp Sodium Chloride/
23 exp Peroxides/
24 (detergent$ or soap$ or alcohol or alcohols or iodine or iodophor$ or chlorhexidine or betadine or disinfectant$ or saline or peroxide or (boiled adj water) or (tap adj water)).mp.
25 (cleans$ or shower$ or wash$ or bath$).mp.
26 exp Bandages/
27 exp Alginates/
28 exp Hydrogels/
29 (dressing$ or gauze or tulle or foam$ or hydrocolloid$ or alginat$ or hydrogel$).mp.
30 exp Massage/
32 exp Postoperative Care/
33 ((post-operative or postoperative) adj care).mp.
34 exp Surgical Wound Infection/pc [Prevention & Control]
35 exp Infection Control/
36 pin site.mp.
38 12 and 37
Appendix 3. Ovid EMBASE search strategy
1 exp External Fixator/
2 exp Bone Nail/
3 exp Kirschner Wire/
4 exp Bone Screw/
5 (external adj (fixat$ or wire$ or pin$ or rod$ or nail$)).ti,ab.
6 (bone adj2 (fixat$ or wire$ or pin$ or rod$ or nail$)).ti,ab.
7 (fracture$ adj2 (fixat$ or wire$ or pin$ or rod$ or nail$)).ti,ab.
8 ((orthopaedic or orthopedic) adj2 (wire$ or pin$ or rod$ or nail$)).ti,ab.
9 (percutaneous adj2 pin$).ti,ab.
10 steinmann pin$.ti,ab.
11 skeletal traction.ti,ab.
13 exp Detergent/
15 exp Alcohol Derivative/
16 exp IODINE/
17 exp Iodophor/
18 exp Povidone Iodine/
19 exp CHLORHEXIDINE/
20 exp Disinfectant Agent/
21 exp Topical Antiinfective Agent/
22 exp Sodium Chloride/
23 exp Peroxide/
24 (detergent$ or soap$ or alcohol or alcohols or iodine or iodophor$ or chlorhexidine or betadine or disinfectant$ or saline or peroxide or (boiled adj water) or (tap adj water)).ti,ab.
25 (cleans$ or shower$ or wash$ or bath$).ti,ab.
26 exp Bandage/
27 exp Alginic Acid/
28 exp Hydrogel/
29 exp Wound Dressing/
30 (dressing$ or gauze or tulle or foam$ or hydrocolloid$ or alginat$ or hydrogel$).ti,ab.
31 exp Massage/
33 exp Postoperative Care/
34 ((post-operative or postoperative) adj care).ti,ab.
35 Surgical Infection/pc [Prevention]
36 exp Infection Control/
37 pin site.ti,ab.
39 12 and 38
Appendix 4. EBSCO CINAHL search strategy
S32 S10 and S31
S31 S11 or S12 or S13 or S14 or S15 or S16 or S17 or S18 or S19 or S20 or S21 or S22 or S23 or S24 or S25 or S26 or S27 or S28 or S29 or S30
S30 TI pin site* or AB pin site*
S29 (MH "Infection Control+")
S28 (MH "Surgical Wound Infection/PC")
S27 TI ( post-operative care or postoperative care ) or AB ( post-operative care or postoperative care )
S26 TI massag* or AB massag*
S25 (MH "Massage+")
S24 TI ( dressing* or gauze or tulle or foam* or hydrocolloid* or alginat* or hydrogel* ) or AB ( dressing* or gauze or tulle or foam* or hydrocolloid* or alginat* or hydrogel* )
S23 (MH "Alginates")
S22 (MH "Bandages and Dressings+")
S21 TI ( cleans* or shower* or wash* or bath* ) or AB ( cleans* or shower* or wash* or bath* )
S20 TI ( detergent* or soap* or alcohol or alcohols or iodine or iodophor* or chlorhexidine or betadine or disinfectant* or saline or peroxide or boiled water or tap water ) or AB ( detergent* or soap* or alcohol or alcohols or iodine or iodophor* or chlorhexidine or betadine or disinfectant* or saline or peroxide or boiled water or tap water )
S19 (MH "Peroxides+")
S18 (MH "Sodium Chloride+")
S17 (MH "Antiinfective Agents, Local+")
S16 (MH "Chlorhexidine")
S15 (MH "Povidone-Iodine")
S14 (MH "Iodine")
S13 (MH "Alcohols+")
S12 (MH "Soaps")
S11 (MH "Detergents+")
S10 S1 or S2 or S3 or S4 or S5 or S6 or S7 or S8 or S9
S9 TI skeletal traction or AB skeletal traction
S8 TI steinmann pin* or AU steinmann pin*
S7 TI percutaneous N2 pin* or AB percutaneous N2 pin*
S6 TI ( orthopedic N2 fixat* or orthopedic N2 wire* or orthopedic N2 orthopedic N2 pin* or orthopedic N2 rod* or orthopedic N2 nail* ) or AB ( orthopedic N2 fixat* or orthopedic N2 wire* or orthopedic N2 orthopedic N2 pin* or orthopedic N2 rod* or orthopedic N2 nail* )
S5 TI ( orthopaedic N2 fixat* or orthopaedic N2 wire* or orthopaedic N2 orthopaedic N2 pin* or orthopaedic N2 rod* or orthopaedic N2 nail* ) or AB ( orthopaedic N2 fixat* or orthopaedic N2 wire* or orthopaedic N2 orthopaedic N2 pin* or orthopaedic N2 rod* or orthopaedic N2 nail* )
S4 TI ( fracture N2 fixat* or fracture N2 wire* or fracture N2 fracture N2 pin* or fracture N2 rod* or fracture N2 nail* ) or AB ( fracture N2 fixat* or fracture N2 wire* or fracture N2 fracture N2 pin* or fracture N2 rod* or fracture N2 nail* )
S3 TI ( bone N2 fixat* or bone N2 wire* or bone N2 bone N2 pin* or bone N2 rod* or bone N2 nail* ) or AB ( bone N2 fixat* or bone N2 wire* or bone N2 bone N2 pin* or bone N2 rod* orbone N2 nail* )
S2 TI ( external N2 fixat* or external N2 wire* or external N2 external N2 pin* or external N2 rod* or external N2 nail* ) or AB ( external N2 fixat* or external N2 wire* or external N2 external N2 pin* or external N2 rod* or external N2 nail* )
S1 (MH "Orthopedic Fixation Devices+")
Last assessed as up-to-date: 5 September 2013.
Protocol first published: Issue 1, 2002
Review first published: Issue 1, 2004
Contributions of authors
Jenny Temple (formerly Lee-Smith) and Julie Santy separately assessed trials for inclusion and trial methodological quality, and Jenny Temple drafted the original version of the review.
Jenny Temple: assessed the results of all updated searches. Extracted data and undertook quality assessment. Approved a final draft of the review update prior to submission.
Anne Lethaby: assessed the results of all updated searches. Extracted data, checked the quality of data extraction and entered data. Wrote to study authors, experts and companies for information. Undertook quality assessment, checked the quality assessment and completed the risk of bias tables. Analysed, interpreted data, performed and checked the statistical analysis. Completed first draft of the review update and made an intellectual contribution to the review. Approved final review update prior to submission. Independently selected relevant trials, extracted data and assessed the included trials for risk of bias. Is guarantor for the review.
Contributions of editorial base
Nicky Cullum: edited the review, advised on methodology, interpretation and review content. Approved the final review and review update prior to submission.
Sally Bell-Syer: coordinated the editorial process. Advised on methodology, interpretation and content. Edited and copy edited the review and the updated review.
Ruth Foxlee: designed the search strategy and edited the search methods section for the update.
Rachel Richardson: checked and edited the most recent update.
Declarations of interest
Sources of support
- University of Plymouth, UK.
- University of Bradford, UK.
- University of Hull, UK.
- The Cochrane Collaboration, Updating Project, UK.
- NIHR/Department of Health (England), (Cochrane Wounds Group), UK.
Differences between protocol and review
The title of the published protocol was 'Interventions to prevent infection associated with external bone fixators and pins'; this has been amended to 'Pin site care for preventing infections associated with external bone fixators and pins' at the review stage.
Medical Subject Headings (MeSH)
Bandages; Bone Nails [*adverse effects]; Bone Wires [*adverse effects]; Disinfectants [administration & dosage]; External Fixators [*adverse effects]; Fracture Fixation [*instrumentation]; Randomized Controlled Trials as Topic; Surgical Wound Infection [*prevention & control]; Therapeutic Irrigation [methods]; Traction [instrumentation]
MeSH check words
* Indicates the major publication for the study