Unburied Versus Buried Kirschner Wires in the Traumatic Osteosynthesis of the Upper Limb in Pediatric Patients: A Retrospective Cohort

Abstract

Introduction:

Kirschner wires are widely used for pediatric fracture fixation. They can be either buried under the skin or left unburied. Although buried Kirschner wires are thought to reduce complications by being protected by the skin, they require a second surgery for removal. Conversely, unburied Kirschner wires allow outpatient removal but may increase infection risk. There is still no consensus on which technique is preferable. This study aimed to compare complication rates between buried and unburied Kirschner wires in pediatric upper limb fractures.

Methods:

We conducted a retrospective cohort study of 92 children undergoing traumatic osteosynthesis with Kirschner wires at our institution between June 2022 and June 2024. Patients were classified into buried (n = 48) or unburied (n = 44) groups. Postoperative complications—including infection, Kirschner wire migration, neurovascular deficit, and cast-related complications—were assessed, along with time-to-complication. Logistic regression, Kaplan-Meier survival analysis, and sensitivity analyses were done.

Results:

Seventeen patients (18.5%) experienced complications, more frequently in the buried group (28%) than in the unburied group (7%). Buried Kirschner wires carried a markedly higher risk (OR 5.6, 95% CI, 1.5-21.2, P = 0.011). Complications included Kirschner wire migration (10% vs 2%), infection (4% vs 0%), skin erosion (8% vs 0%), cast-related pain requiring a replacement (4% vs 4%), and neurovascular palsy (2% vs 0%). Kaplan-Meier analysis showed no notable difference in overall complication-free survival (P = 0.161). Sensitivity analyses did not reveal any benefit favoring buried Kirschner wires in any of the scenarios.

Conclusion:

Burying Kirschner wires in pediatric fracture fixation was not associated with a lower complication risk and did not prove to offer a clear clinical advantage. Because unburied Kirschner wires can be removed in outpatient settings, sparing patients a second surgery and reducing financial costs, they may represent the preferable approach. Further research including larger multicenter studies is needed to validate these findings and establish consensus.

Kirschner wires are widely used in the field of orthopaedics. Although fractures in pediatric patients are often treated nonsurgically, sometimes, the displacement makes osteosynthesis by Kirschner wires necessary for the reduction and stabilization of the fracture.

Kirschner wires can be buried under the skin, or left unburied, exposed to the external environment. In the case of unburied Kirschner wires, the interface between the bent Kirschner wires and the skin, as well as the Kirschner wires themselves, are covered with a sterile dressing.

Each technique presents its advantages and drawbacks. The presupposition is that buried Kirschner wires are less likely to lead to complications due to the protection provided by the skin tissue. However, they require a secondary procedure in the operating room (OR) for removal, involving additional general anesthesia and notable logistical costs¹ for the hospital. Contrarily, the unburied Kirschner wires are exposed to the outside environment, leading in theory to a higher risk of complication specifically for infection, but they do not necessarily require a second surgery because they can easily be removed in the outpatient clinic under local anesthetic.

Complications associated with Kirschner wire osteosynthesis can markedly impact patient outcomes and healthcare costs. Despite their customary use, there remains controversy on whether which technique—buried or unburied Kirschner wires—leads to fewer rates of complications, including infection cases. Existing publications have yielded contradictory results, with some suggesting that one technique may be preferable to the other for medical or hospital logistical reasons,^2-6 whereas others have found no difference between the two surgical approaches.⁷

This study aims to compare postoperative complication rates between buried and unburied Kirschner wires in pediatric upper limb fracture fixation and to explore whether factors such as patient characteristics or fracture type.

Methods

Study Design

This retrospective cohort study included all patients from the Pediatric Surgery Department of our institution who underwent traumatic osteosynthesis by Kischner wires between June 2022 and June 2024. Patients were then classified into two groups based on the surgical procedure they had gone through: buried Kirschner wires (completely embedded under the skin) and unburied Kirschner wires (left protruding).

Objective

The primary goal of this retrospective cohort study is to compare the postoperative complication rates between buried and unburied Kirschner wires in children operated by osteosynthesis by Kirschner wires for fractures of the upper limb. Secondarily, it aims to evaluate time-to-complication and whether specific factors, such as age, sex, comorbidities, and fracture type, influenced outcomes in both techniques.

Data Collection

Demographic and clinical data were collected from medical records. We identified a total of 98 patients who underwent traumatic osteosynthesis surgery using Kirschner wires between June 2022 and June 2024. Two patients were excluded because the Kirschner wire fixation was performed at another institution, and four other patients were excluded because they were operated in the lower limb. This resulted in a final cohort of 92 patients with 48 patients having buried Kirschner wires and 44 patients having unburied Kirschner wires (Figure 1). Age, sex, prior medical history, and type of fixation were extracted. The main outcome was the incidence of postoperative complications, including Kirschner wire migration, neurovascular deficit, cast removal due to pain, and infection, which manifested as either superficial infection, osteomyelitis, or osteoarthritis. The secondary outcome was time-to-complication, recorded in days from the osteosynthesis in the operating room to the first documented complication either immediately after surgery or during the follow-up period.

Figure 1.

Flow chart showing the cohort of the study.

Statistical Analysis

We evaluated the risk of postoperative complications according to surgical technique using logistic regression. Incidence rates were calculated with 95% Wald confidence intervals. Time-to-event data were analyzed with the Kaplan-Meier method; differences between groups were evaluated using the two-sided log-rank test. A sensitivity analysis was done to assess the assumption of noninformative censoring for patients lost to follow-up, considering both worst-case (events assigned at censoring date) and best-case (censoring date replaced by study end point) scenarios. For notable log-rank tests, the proportional hazards assumption was evaluated by assessing the time-dependent pattern of Schoenfeld residuals derived from the Cox proportional hazards model. Violations of this assumption prompted a piecewise hazard model reporting hazard ratios over time (before and after survival curve crossing). Non-normally distributed quantitative variables (assessed visually and with Shapiro-Wilk) are reported as median (interquartile range) and compared using Student t-test. Qualitative variables are summarized as counts (percentages) and compared using χ² or Fisher exact tests when expected counts were <5. Complication types and comorbidities were summarized descriptively. All analyses were two sided, with α = 0.05, and conducted using SAS software, version 9.4 (SAS Institute Inc., Cary, NC).

Results

Baseline Characteristics

The median age of patients in the buried group was 6 (4–8.5) years, markedly younger than the unburied group, which had a median age of 9 (5–12.5) years (P = 0.003). No notable difference was found in sex distribution between the two groups: the buried group comprised 58% males and 42% females, whereas the unburied group included 59% males and 41% females (P = 0.941). Comorbid conditions were uncommon in both groups. One patient in each group had a history of tobacco use disorder. Additional comorbidities in the buried group included one patient with Factor V Leiden and one with an atrial/ventricular septal defect (ASD/VSD). These conditions were not present in the unburied group (Table 1).

Table 1.

Demographic and Clinical Characteristics of the Patients at Baseline

	Buried (n = 48)	Unburied (n = 44)	P
Age (years)	6 [4; 8,5]	9 [5; 12,5]	0.003
Sex n, %
Male	28 (58)	26 (59)	0.897
Female	20 (42)	18 (41)
Comorbiditiesa n, %
Tobacco use disorder	2 (4)	0 (0)
Factor V Leiden	1 (2)	0 (0)
ASD/VSD	0 (0)	1 (2)

^aVariable described only by descriptive analysis.

Type of Fracture

The most common fracture in both groups was the supracondylar fracture, accounting for 25 patients (52%) in the buried group and 22 patients (50%) in the unburied group. Most of the supracondylar fractures were classified as type IV according to the Lagrange and Rigault classification.⁸ The second most frequent fracture was the lateral condyle fracture, followed by medial epicondyle, distal radius and ulna, metacarpal, olecranon, and Monteggia fracture (Table 2).

Table 2.

Type of Fracture

	Buried (n = 48)	Unburied (n = 44)	P
Fracture n, %
Supracondylar	25 (52)	22 (50)
Type II	3 (6)	2 (4)	<0.0001
Type III	6 (13)	9 (21)
Type IV	16 (33)	11 (25)
Lateral condyle	9 (20)	6 (14)
Medial epicondyle	5 (10)	4 (9)
Distal radius and ulna	3 (6)	4 (9)
Metacarpal	0 (0)	5 (11)
Olecranon	4 (8)	3 (7)
Monteggia	2 (4)	0 (0)

The distribution of the fracture type was markedly different between the two groups (P < 0.0001) because all the metacarpal fractures were in the unburied group and all the Monteggia fractures in the buried group.

Postoperative Complications

Out of the 92 patients, 17 patients experienced complications postoperatively. Of these, 14 patients (28%) were in the buried group, and three patients (7%) were in the unburied group. Five patients (10%) in the buried group and one patient (2%) in the unburied group experienced Kirschner wire migration, which was observed either during the follow-up appointments in the outpatient clinic or in the OR during the removal surgery in the patients who had buried Kirschner wires. In the buried group, migration occurred in three patients with supracondylar type IV fractures and in two patients with lateral condyle fractures. The single case in the unburied group occurred in a patient with a supracondylar type III fracture. Two cases (4%) of infection due to surgery were observed in the buried group. One patient with a supracondylar type IV fracture had a superficial infection, which was managed by local wound care. The other patient, with a medial epicondyle fracture treated with Kirschner wires, was diagnosed with a septic arthritis, which required joint irrigation, débridement and intravenous antibiotics. No cases of infection were observed in the unburied group. One case of ulnar neuropathy was observed in a patient with a type IV supracondylar fracture in the buried group. It fully recovered within a few months. Two patients in each group presented in the emergency department (ED) with severe pain under the cast, which had to be replaced. In the buried group, both patients had supracondylar type IV fractures, and in the buried group, one had a supracondylar type III fracture and the other one had a supracondylar type IV fracture. In addition, four patients (8%) in the buried group presented skin erosion in the site where the Kirschner wire was placed during surgery (Table 3). Two of them had a supracondylar type IV fracture, one had a supracondylar type III fracture and the other one had a distal radius fracture. Therefore, more absolute cases of complications were found in the buried group than in the unburied group. Most of the complications occurred in patients with supracondylar type IV fractures (58.82%).

Table 3.

Postoperative Complications in the 92 Patients of the Study

	Buried (n = 48)	Unburied (n = 44)
Type of complications n, %
Kirschner wire migration	5 (10)	1 (2)
Neurovascular palsy	1 (2)	0 (0)
Septic arthritis	1 (2)	0 (0)
Osteomyelitis	0 (0)	0 (0)
Pain under cast requiring replacement	2 (4)	2 (5)
Skin erosion	4 (8)	0 (0)
Superficial infection	1 (2)	0 (0)

The incidence and risk of complications was also assessed. The incidence in both groups resulted to be 18.5% (95% CI, 10.5-26.4). In the unburied group, complications occurred in 7% of patients compared with 28% in the unexposed group, corresponding to a markedly increased risk associated with the buried technique (OR = 5.6 [1.5-21.2], P = 0.011).

Time-to-Complication

Time-to-complication was analyzed using a Kaplan-Meier model and a log-rank test (Figure 2). The Kaplan-Meier survival analysis compared complication-free probabilities after surgery between the two groups. The curves show that both groups start complication free, but early complications (days 10-20) appear slightly more in the unburied group, whereas the buried group experiences later declines (days 30-50). By day 80, the complication-free probabilities are 0.89 in the unburied group and 0.57 in the buried group. Although trends suggest early advantage for the buried technique and later stabilization for the unburied group, the log-rank test (P = 0.161)” indicated no statistically significant difference, showing that overall complication-free survival is broadly similar between the two surgical approaches.

Figure 2.

Graph showing probability of freedom from postoperative complications over time in the population.

The sensitivity analysis done under the worst-case scenario showed no notable difference between the two groups (P = 0.34) (Figure 3). Conversely, the best-case scenario showed a notable benefit (P = 0.026) (Figure 4) for the unburied group, with 0.9 complication-free probabilities. These findings are supported by the piecewise Cox model, which showed comparable risk between the two techniques up to 37 days (HR = 1.13 (95% CI, 0.16-8.0), P = 0.903) and a markedly lower risk for the unburied group beyond this time point (HR = 0.11 (95% CI, 0.01-0.83), P = 0.032).

Figure 3.

Graph showing probability of freedom from postoperative complication over time in the worst-case scenario.

Figure 4.

Graph showing probability of freedom from postoperative complication over time in the best-case scenario.

Discussion

In this retrospective study, the complication rates of patients with buried and unburied Kirschner wires were compared and analyzed. In absolute numbers, more complications were observed when Kirschner wires were buried. The rate of infection, Kirschner wire migration, and neurovascular palsy were higher. Moreover, we could prove that there is an increased risk of developing complications with the buried technique (OR = 5.6 [1.5-21.2], P = 0.011).

When we analyzed the time-to-complication with the Kaplan-Meier analysis, although results were not statistically significant in the log-rank test (P = 0.161), when we performed a sensitivity analysis, either in our scenario and in the hypothetical worst-case scenario, there was no difference between the two techniques.

Our initial hypothesis was that buried Kirschner wires are less likely to lead to complications due to the protection provided by the skin tissue. After our analysis, this premise could not be validated because buried Kirschner wires did not prove to have less complications than unburied Kirschner wires in any of the scenarios of our analysis. Therefore, in our experience, there is not a notable advantage in burying the Kirschner wires. Hence, the decision to leave Kirschner wires unburied may be a wise one, as it is not inferior to the buried technique and, moreover, offers some important advantages.

Our results align with previous studies, such as McGonagle et al,² which found no justification for burying Kirschner wires, as they also did not observe any notable advantage and therefore recommended leaving them unburied. Abdullah et al³ did not find any notable difference in infection rate between buried and exposed Kirschner wires in patients with wrist and hand fractures, which also supports our findings. Similarly, Khaled et al⁴ reported in a prospective randomized study that leaving implants unburied constitutes a safe approach, yielding comparable complication and infection rates compared with the buried technique. All these results are also supported by a recent systematic review and meta-analysis by Pullan et al,⁵ in which after analyzing 15 studies with more than 1800 patients in total concluded that although buried Kirschner wires are associated with lower overall infection rates, the risk of deep infection is similar between the two techniques. Moreover, given the high rates of skin erosion associated with the buried technique, as well as the extra cost of undergoing a second surgical procedure for removal, suggest that unburied Kirschner wires are safer and more cost-effective than burying them.

Patient Comfort

The procedure for removing Kirschner wires differs depending on the technique used. When a Kirschner wire is buried, a second surgical procedure is required, often under general anesthesia, and in some cases, regional anesthesia. By contrast, the removal of an unburied Kirschner wire is done in the outpatient clinic using inhalation anesthesia, such as nitrous oxide, accompanied by oral analgesia. In addition, the buried Kirschner wire removal requires an incision for extraction, whereas the unburied Kirschner wire typically leaves only a small puncture hole after removal. Although formal evidence is lacking, the unburied technique may offer greater comfort for most patients, particularly children.

Healthcare Costs Implications

From an economic standpoint, leaving the Kirschner wires unburied offers a notable cost-saving advantage. The removal of an unburied Kirschner wire requires only a limited number of personnel, a limited time in a hospital facility, a simple analgesic, and a less-invasive intervention for the patient. Conversely, the removal of a buried Kirschner wire involves the use of an operating room and very often a general anesthesia procedure, which clearly has a higher risk associated. In our institution, the overall cost of each technique—including both the primary surgical procedure and subsequent Kirschner wire removal—was markedly higher for buried Kirschner wires. The mean cost for buried Kirschner wires was USD 4514.20, compared with USD 2653.67 for unburied Kirschner wires, resulting in an additional expense of USD 1860.53. This cost difference is economically notable and should be considered when selecting the appropriate fixation method. Opting for unburied Kirschner wires may therefore help allocate resources more efficiently to reduce healthcare spending.

Study Limitations

Despite the valuable insights provided by this study, several limitations should be acknowledged.

Sample Size

The retrospective design and the limited sample size may affect the strength and reliability of the conclusions. This limits the use of multivariate models (both logistic regression and the Cox model) for adjustment. Here, only univariate analysis results are presented. Results may be therefore interpreted with caution.

Lost To-Follow-up Patients

The number of patients lost to follow-up is high relative to the total sample (12 out of 92, i.e., 13%), particularly because the majority (10 out of 12) are from the exposed “unburied” group. The assumption of noninformative censoring had to be tested through a sensitivity analysis, which produced divergent results: the worst-case scenario (Figure 3) and overall cohort results were not notable (no difference between the curves), whereas the best-case scenario (Figure 4) showed a difference between the two methods favoring the exposed technique. These results therefore indicate information loss due to censoring in the exposed group, requiring cautious interpretation.

Cox Model Analysis for HR Estimation

For the best-case scenario, the risk was not proportional throughout the follow-up period, as the curves crossed at around 40 days. The model was therefore split to calculate risk before and after the crossing. During the first phase, the difference was not notable, whereas it became notable during the second period, with risk favoring the exposed group. This analysis aligns with the previous odds ratio calculation, but it should again be interpreted with caution, as it reflects the best-case scenario (patients who had not experienced an event by the index date).

Moreover, the study was conducted over a relatively short period of only two years and within a single institution, which inherently limits the generalizability of the findings. Multicenter studies involving a more diverse range of hospitals would enhance the external validity of the results and make them more applicable to a broader patient population.

Finally, our institution primarily serves a population from lower socioeconomic backgrounds, which may influence treatment outcomes due to factors such as limited access to healthcare services or reduced adherence to postoperative care recommendations.⁷

Conclusion

In conclusion, this study suggests that burying Kirschner wires for traumatic osteosynthesis in pediatric patients does not offer a notable advantage compared with leaving them unburied. Given that the unburied option most of the time eliminates the need for a second surgery and offers a more cost-effective approach, it may be considered in hospitals as the preferred method.

However, as the field evolves, additional studies are merited to help reach a broader consensus on the most effective and efficient technique to use.