Brief Report

Clinical characteristics and management of wound foreign bodies in the ED

Matthew R. Levine MD?, Stephen M. Gorman BS, Christian F. Young MD, D. Mark Courtney MD

Department of Emergency Medicine, Feinberg School of Medicine, Northwestern University, Chicago, IL 60611, USA

Received 19 July 2007; revised 15 November 2007; accepted 17 November 2007

Abstract

Objectives: Wound foreign body (FB) management is challenging. Little data exist describing ED management. We sought to describe case characteristics and ED management of wound FBs.

Methods: retrospective case series of all patients with ICD-9 code diagnosis of wound FB over 4 years in 2 EDs.

Results: Four hundred ninety cases met inclusion criteria. A total of 75.6% presented within 48 hours of injury. Wounds from foreign material or stepping on something accounted for two thirds of the mechanisms. Most were wood, metal, or glass. Diagnosis was primarily by physical exam with exploration (77.6%; 95% confidence interval, 73.4%-81.5%), but 22.4% (95% confidence interval, 18.5-26.6%) were only diagnosed radiographically. X-rays missed 25% of glass and 93% of wood FBs. Eighty-nine percent were removed in the ED. Consultation was uncommon (9.6%).

Conclusions: Most, but not all, patients with wound FBs presented within 48 hours of injury and had suspicious complaints or mechanisms. Diagnosis was usually clinical but sometimes only by radiography. Radiographs were not sufficiently sensitive to reliably detect wood and glass. Most FBs were successfully removed without specialty consultation.

(C) 2008

Wound care is one of the most common presentations for Emergent care–approximately 6.5 million ED visits occur in the USA annually for open wounds [1]. The possibility of a retained soft tissue foreign body (FB) is common. Diagnosis and removal can be challenging and time consuming. Soft tissue FBs are frequently missed on initial evaluation. One study of hand FBs from a hand practice reported that 75 of 200 FBs were missed by the initial physician [2]. Missed diagnosis results in potential morbidity such as infection, chronic pain, structural injury, and impairment of function. Missed FBs also are high medico-legal risk for the physician with one report citing this to be among the top 3 causes of litigation after wound care [3].

* Corresponding author. Tel.: +1 312 694 7000; fax: +1 312 926 6274.

E-mail address: mrlevine@comcast.net (M.R. Levine).

0735-6757/$ – see front matter (C) 2008 doi:10.1016/j.ajem.2007.11.026

The importance of wound FB diagnosis is commonly taught to medical students and residents. However, there are little quantitative data to describe the types of foreign bodies and the clinical course in the ED for these patients. We performed this descriptive study of a large sample of patients with wound FBs to describe the following: (1) patient characteristics, (2) wound and FB characteristics, and (3) patient management and disposition.

We performed an institutional review board-approved retrospective case series of all subjects with a final ED diagnosis of wound FB. There were 2 study sites–an urban university ED with an EM residency with an annual census of more than 70000 patients, and an affiliated community ED with an annual census of more than 40000. Data were gathered from January 2000 to November 2004 at the university ED and from June 2001 to July 2005 at the community ED. For inclusion, there had

Fig. 1 Time since injury on day of presentation (n = 450).

to be a definite FB documented either by imaging, visualization on physical exam, or actual FB removal. Any cases of possible, probable, or “rule-out” FB without a documented definite FB by imaging, direct visualization, or removal were excluded. Patients with FBs of the eye, airway, gastrointestinal tract, or any other nonwound or soft tissue FBs were excluded. Acute gunshot wounds were also excluded.

Foreign-body cases were identified by ICD-9 code search of all ED visits during this time that related to a soft tissue FB. ICD-9 codes found to be relevant to soft tissue FBs included

7094, 9106, 9107, 9116, 9117, 9126, 9136, 9137, 9146, 9147,

9156, 9157, 9166, and 9176. Electronic medical records were reviewed at both sites with data abstracted into a Microsoft Access database (Microsoft Corp, Redmond, Wash) using a predefined data collection instrument. There were 2 data abstractors, both of whom were trained by the principal investigator (PI). The first 20 charts were reabstracted by the PI and no disagreements were found.

Data included institution, sex, age, date of presentation, date of injury, mechanism of injury, FB material, number of FBs, FB size measured from positive x rays, how the FB was initially diagnosed (physical exam/exploration vs imaging), studies ordered, study results, removal attempts and results, use of specialty consultation in the ED, antibiotic use, and wound closure. To minimize misclassification bias, we used a standardized predefined data collection instrument with specific definitions of categories of chief complaint types and mechanism of injury classifications, and adjudicated any discrepancies by consensus between 3 study authors. Data were abstracted by one of the physician authors and a trained research associate who had participated in many other projects involving data abstraction. Hundreds of charts were reviewed by a second physician author to ensure accuracy of abstraction. Every chart abstracted was reviewed by at least one physician author. Missing data were neither assumed to be positive nor negative, but simply unavailable and excluded from analysis.

Statistical analysis was done with Stata (version 9.2, StataCorp, College Station, Tex, 2006). Proportions were reported with 95% confidence intervals. Continuous data were

Fig. 2 Mechanisms of injury. Data available for 424 patients. *Cut on material–Patient sustained wound on a foreign material as opposed to an implement such as a knife. Materials included: splinters or slivers (78), glass or ceramic (35), wooden objects (34), pieces of metal (7), metal door or gate (4), miscellaneous material (cactus, paper, toothpick, clothespin, paint chip, white chips, shell) (7). ^#Cleaning surface– Patient sustained wound cleaning a surface by rubbing one’s hand over it.

n		%	95% CI
Material (total = 490)
Wood	168	34.3	30.1-38.7
Metal	134	27.3	23.4-31.5
Glass/ceramic	134	27.3	23.4-31.5
Other/unknown	32	6.5	4.5-9.1
Pencil tip	7	1.4	0.6-2.9
Gravel/rocks/shells	7	1.4	0.6-2.9
Other vegetative	5	1.0	0.3-2.4
Paint chips	3	0.6	0.1-1.8
Location (total = 489)
Upper extremity	285	58.2	53.6-62.6
Lower extremity	178	36.3	32.1-40.8
Head and neck	21	4.3	2.7-6.5
Trunk	5	1.0	0.3-2.4

reported with means, medians, SDs, and ranges. Preplanned subgroup analysis was performed on cases with multiple FBs, those who had radiographic imaging, cases involving specialist consultation, and those with Delayed presentations. A total of 490 subjects met the inclusion criteria. Males accounted for 58.2% of the cohort (285/490; 95% confidence interval [CI], 53.7%-62.6%). Mean age was 35.6 years (SD,

Table 1 Wound FB materials and locations

+-16.6 years). The range of ages was from 1 to 88 years (interquartile range, 24-46 years).

Most patients presented on the day of injury or the next day (75.6%; 95% CI, 71.3%-79.5%) (Fig. 1). However, 8.7%

(95% CI 6.2-11.6 %) presented more than 1 week from time of injury, which we arbitrarily defined as late presentation. All except 2 of these late presenters came within one year of injury, but one presented 26 years later.

Lacerations on foreign material (sustaining a wound on foreign material as opposed to an implement such as a knife) and stepping on something were the most common mechan- isms, accounting for 66.1% (95% CI, 61.4%-70.6%) of all cases with documented mechanisms. Most other mechanisms had minor representations (Fig. 2). Cleaning a surface with one’s hand emerged as a mechanism not traditionally associated with retained FBs to our knowledge.

Three materials accounted for the majority of FBs: wood (168, or 34.3%; 95% CI, 30.0%-38.7%), metal (134, or

27.3%; 95% CI, 23.4%-31.5%), and glass or ceramic (134, or 27.3%; 95% CI, 23.4%-31.5%). Most were located in the

extremities; 58.2% (95% CI, 53.6%-62.6%) were in the

upper extremity and 36.3% (95% CI, 32.0%-40.8%) in the lower extremity (Table 1).

Clear data specifying the number of FBs present were available for 474 cases. Eighty-eight percent had a single FB (95% CI, 84%-90%). Patients with glass FBs were more likely to have multiple FBs than those with wood or metal FBs (odds ratio, 2.84; 95% CI, 1.6-5.1). Size of 146 FBs was measured precisely from all available positive x-rays (n = 121). Foreign-body sizes varied significantly (median

size, 4 mm; range, 1-172 mm; 25%-75% interquartile range, 2-10 mm). There were 15 FBs measuring only 1 mm, 7 of which were removed in the ED.

Clear information regarding how the FB was diagnosed was available for 425 patients. Of these, most (77.6%; 95% CI, 73.4%-81.5%) were initially diagnosed by physical exam. The remaining 22% were diagnosed by plain film radio- graphy. For 65 cases, it was unclear whether the Initial diagnosis was made by physical exam or radiography. Plain film radiography was ordered in 246 patients, or 50% (95% CI, 46%-55%). Three patients had CTscans, all for scalp FBs. One quarter of all radiographs were negative despite FBs being ultimately recovered from these patients (n = 59; 25 wood, 24 glass, 1 metal, 9 other). Although this study was not designed to determine the sensitivity of plain film radiography for FB detection, as every case in the cohort was a true positive, we were able to calculate radiograph sensitivity for plain films by FB type. This was 75.5% for glass, 98.6% for metal, and 7.4% for wood (Table 2). Also,

gravel was found in 5 of 5 radiographs ordered.

X-rays that were done after FB removal attempts were low yield. Of the 42 postprocedure x-rays with available results, 3 were positive and 39 were negative for FBs. None of the 3 positive postprocedure x-rays diagnosed new FBs; they just confirmed the persistence of unremoved material.

Foreign bodies were removed in the ED in 89% of cases (95% CI, 86.0%-91.8%). It was usually unclear whether the FBs that were not removed were due to failed attempts or that removal was considered too difficult or not indicated. Consultation of surgical specialists in the ED was not common, occurring in 47 cases (9.6%; 95% CI, 7.1%- 12.6%). When consultants were involved, 36.4% of the FBs were still not removed in the ED.

Systemic antibiotics were prescribed to 44.4% (95% CI, 39.9%-49.0%) of subjects. Cases involving consultants had high rates of systemic antibiotic usage (84.4%; 95% CI, 70.5%-93.5%). It is unclear whether this was due primarily to higher antibiotic use by consultants or that consultation cases may have been more complicated and at risk for infection. During the initial ED visit, 53 patients (10.8%; 95% CI, 8.2%-13.9%) had their wounds closed.

We arbitrarily defined late presenters as those presenting greater than one week after the injury. Of the 40 late presenters, glass was the most common FB (n = 18) and wood was uncommon (n = 5). More than two thirds (26/40) of these had successful FB removal in the ED.

Table 2 Sensitivity of plain film radiography by FB type

FB X-rays X-rays X-rays Sensitivity 95% CI Type ordered positive negative (%)

Glass 98 74 24 75.5 65.8-83.6

Metal 72 71 1 98.6 92.5-100.0

Wood 27 2 25 7.4 0.91-24.3

Number of positive x-rays (true positives) divided by the sum of the positive x-rays (true positives) and the negative x-rays (false negatives).

We describe here the clinical characteristics of the largest sample of ED wound FB cases to date. The largest prior wound FB cohort described is a report of 200 cases from a hand clinic [2]. This sample may not be generalizable to the ED as referral clinic patients would be more likely to have the FBs in place for longer at the time of presentation and may have already had complications requiring the referral. Other ED-based studies are limited by small samples [4-6] or focus exclusively on specific injury types. Morgan et al [7] described 164 consecutive hand FB cases in an ED in an industrial city in England, most resulting from machinery injuries from metal. Our series, from both urban and suburban communities with foreign bodies not limited to one anatomic location or material, is the largest and most representative report to date of typical ED patients.

Although most wound FBs presented within 48 hours of injury, one quarter presented weeks, months, or rarely, years later. These presentation patterns and percentages are consistent with prior smaller reports [7,8]. More than two thirds of the late presenters had their FBs successfully removed in the ED, so an extended period since injury should not be an exclusionary factor when deciding whether to seek or remove an FB. Notably, in one case a glass FB was removed in the ED from the forehead of a patient 26 years after an MVC because it migrated near the surface and became bothersome.

Data presented here support prior work indicating wood, metal, and glass are the most frequent FB types in the ED with wood being slightly more common than the others [2,7]. Almost 95% of the FBs we describe were located in the extremities, a finding also consistent with prior smaller reports [8-10].

One in 12 of our subjects had multiple FBs. The most common material in these cases was glass. Approximately 1 in 5 patients with glass FB had multiple FBs. This finding has not been previously reported, but is consistent with typical wound care teaching–that patients with one FB are at risk for multiple FBs in the same wound. Clinicians should maintain a high index of suspicion that a second FB may be present with glass wounds.

Approximately 90% of FBs detected in the ED were also successfully removed in the ED. This is consistent with prior reports: Steele et al [11] reported ED removal of 24 of 28 FBs, whereas Avner and Baker [9] reported ED removal of 31 of 33. Our report also indicates that even many FBs as tiny as 1 mm were successfully removed in the ED (7 of 10 instances with available data on removal). It is true that some small inert FBs may be safely left in situ, but factors such as location, accessibility, pain, functional impairment, Infection risk, and psychological distress are additional considerations when deciding to attempt FB removal.

The most valuable element of the FB assessment was the physical examination and exploration, which was the means of diagnosis for 78% of the FBs. Many wood and glass FBs were recovered after negative plain films, emphasizing the

importance of a careful physical exam and wound exploration. It is commonly stated that radiographs are needed for medico-legal reasons [12]. Other data have suggested that radiography can identify FBs that were not visualized clinically [2,13]. In our series, plain radiographs were ordered in half of the cases. These films were nearly perfect for detecting metal. However, wood FBs were missed in 93% of cases and glass was missed 25% of the time. Our data suggest that radiography has limitations when glass is suspected and little value at all for wood. One patient in our series with a negative x-ray subsequently had a 2-cm piece of wood removed. If an Imaging study is indicated for wood, ultrasound is more sensitive [14].

Neither physical exam with wound exploration nor radiography alone can sufficiently rule out all FBs, so the two should be considered complementary in the evaluation of possible wound FBs. Because of limitations of both plain films and ED wound exploration, some FBs will invariably be missed. Patients should always be told of the possibility of retained FBs and given detailed instructions regarding subsequent symptoms that could occur in the setting of retained FBs and, if these develop, to follow up to optimize recovery.

With respect to ED management, we report that antibiotics were sometimes prescribed (44%), consultants were rarely used (10%), and successful removal usually occurred (89%).

We acknowledge that because this work does not include comparison Control subjects without FBs, we cannot estimate specific predictor variables to indicate who does and does not have an FB. Similarly, we cannot comment on the overall diagnostic accuracy, particularly specificity, of radiographic studies. The sensitivity calculations for plain film radio- graphy are limited by the inability to control for different sizes and locations of the FBs.

Interrater agreement was not formally measured. Most of the data elements gathered (sex, type of FB, FB location) were straightforward and easy to abstract. However, the PI (MRL) reviewed all plain films and more than 100 charts personally to examine for concordance with the primary data abstractor. There were few discrepancies, and these were handled by consensus among the 3 authors.

It is unlikely that every FB case was captured by our ICD-9 code search. For example, wounds coded as lacerations without mention of FB in the diagnosis line may not have been captured. Also, FBs from an MVC in which other more severe injuries may be written in the final diagnosis, omitting the FB diagnosis, would not have been captured. The goal of this study was not to provide a complete estimate of FB incidence, but to describe the clinical characteristics and management of this common yet underreported problem.

Most, but not all, patients with a wound FB presented within 48 hours of injury and had suspicious complaints or mechanisms suggestive of a possible FB. Diagnosis was usually by wound exploration but some FBs were identified

only by radiography. Depending on radiographs exclusively to rule out the possibility of wood or glass foreign bodies may be problematic for almost all wood and some cases of glass. Emergency physicians successfully removed most FBs, uncommonly requesting specialty consultation.

References

1. McCaig LF, Nawar EW. National Hospital Ambulatory Medical Care Survey: 2004 emergency department summary. Adv Data 2006; 372:1-29.
2. Anderson MA, Newmeyer WL, Kilgore ES. Diagnosis and treatment of retained foreign bodies in the hand. Am J Surg 1982;144:63-7.

[3]. Karcz A, Korn R, Burke MC, et al. malpractice claims against emergency physicians in Massachusetts: 1975-1993. Am J Emerg Med 1996;14:341-5.

1. Montano JB, Steele MT, Watson WA. Foreign body retention in glass- caused wounds. Ann Emerg Med 1992;21:1360-3.
2. Ousby J, Wilson DH. 1086 consecutive injuries caused by glass. Injury

1982;12:427-30.

1. Banerjee B, Das RK. Sonographic detection of foreign bodies of the extremities. Br J Rad 1991;64:107-12.
2. Morgan WJ, Leopold T, Evans R. Foreign bodies in the hand. J Hand Surg 1984;9:194-6.
3. Crawford R, Matheson AB. clinical value of ultrasonography in the detection and removal of radiolucent foreign bodies. Injury 1989;20: 341-3.
4. Avner JR, Baker MD. Lacerations involving glass: the role of routine roentgenograms. AJDC 1992;146:600-2.
5. Friedman DI, Forti RJ, Wall SP, et al. The utility of bedside ultrasound and patient perception in detecting soft tissue foreign bodies in children. Ped Emerg Care 2005;21:487-92.
6. Steele MT, Tran LV, Watson WA. Retained glass foreign bodies in wounds: predictive value of wound characteristics, patient perception, and wound exploration. Am J Emerg Med 1998;16:627-30.
7. Kaiser CW, Slowick T, Spurling KP, et al. Retained foreign bodies. J Trauma 1997;43(1):107-11.
8. Gron P, Andersen K, Vraa A. Detection of glass foreign bodies by radiography. Injury 1986;17(6):404-6.
9. Turkcuer I, Atilla R, Topacoglu H, et al. Do we really need plain and soft-tissue radiographies to detect radiolucent foreign bodies in the ED? Am J Emerg Med 2006;24(7):763-8.