Introduction

Retained surgical needles after ophthalmic surgery pose a significant risk to patients' ocular health. Foremost is the concern for ocular morbidity, including globe perforation, inflammation, infection, or toxicity that may exist with an orbital foreign body. Beyond the ocular sequelae is the enduring contraindication for performing magnetic resonance imaging in such a patient.

Numerous healthcare and surgical facilities impose blanket protocols mandating plain film radiographs of the surgical site be taken and interpreted by the surgeon and/or a radiologist should there be a miscount of suture needles used during any surgical procedure. Completion of the surgical case may be delayed until radiologic confirmation that a suture needle is not present in the surgical field. However, the application of such policies when applied to ophthalmic surgical cases is questionable in light of the current literature.

According to the Patient Care Memorandum-112-002-SAC, June 2006, of the Department of Veterans Affairs, Boston Healthcare System, VA, USA, ‘surgical instruments must be counted, except for procedures that are routinely concluded with a radiograph (for example, an orthopaedic case to assure proper alignment of a bone or implant). In these cases, a radiograph is mandatory if an instrument count is not performed, and reading the radiograph to determine whether any instruments have been retained must be performed before the patient is transferred from the OR.’

This study sought to evaluate the efficacy of plain film radiographs in reliably identifying retained suture needles used in ophthalmologic surgical cases.

Materials and methods

Ophthalmologic 10–0 nylon suture needles (Ethicon) were placed randomly into the conjunctival fornices, sclera, or cornea of 10 out of 20 porcine eyes. All eyes were placed within the orbit of a human cadaver skull, which was positioned on a soft tissue phantom. This allowed for the reproduction of the soft tissue and bony artefact normally present on radiographs taken from human patients.

Portable, plain film radiographs were taken of all eyes with a C-arm fluoroscopy machine (Figures 1 and 2). All radiographs were performed using radio-opaque markers to maintain a key, which could be used to identify films with a needle and those without. Control films (Figures 3, and 4) of an eye with and without a suture needle were also taken to ensure the ability to accurately identify the presence of the suture needle in the absence of the skull and soft tissue artefact.

Figure 1
figure 1

Portable plain film without a retained suture needle.

Full size image
Figure 2
figure 2

Portable plain film with a retained suture needle.

Full size image
Figure 3
figure 3

Control films identifying the presence of a needle (arrow).

Full size image
Figure 4
figure 4

Control films identifying the absence of a needle.

Full size image

All films were then reviewed independently by six radiologists and six ophthalmologists who were asked to identify the presence or absence of a surgical needle. The readers were blinded to the experimental design, and no time limit was imposed. The sensitivity and specificity of identifying orbital foreign bodies by portable plain films was analysed for all subjects and further analysed on the basis of the subgroups of radiologists and ophthalmologists.

Results

Evaluation of the responses of all 12 reviewers showed a significant number of false-negative and false-positive responses. The overall sensitivity and specificity for all physicians reviewing radiographs with suspected intraoperative suture needles was 54 and 77%, respectively.

Subgroup analysis showed that among radiologists reviewing the films, false-negative responses remained approximately the same with a sensitivity of 57%. However, they did tend to have fewer false-positive responses, as demonstrated by a specificity of 87%. On the other hand, ophthalmologists reviewing the films identified retained needles less reliably with a sensitivity of 51%, and a specificity of 67%.

Discussion

The value of plain film radiography in identifying intraocular foreign bodies has been a subject of significant controversy. During the advent of X-ray technology in the late 19th century, a landmark article found the technique to be a reliable test in indicating the mere presence of an intraocular foreign body; however, it was not consistent in accurately localizing the foreign body.1 Otto et al2concluded that plain film radiographs can be used as an effective screening test for metallic foreign bodies in the orbit before MRI. However, their study did demonstrate a threshold size at which identification could only be achieved by CT scan and not by plain film.

Further complicating the matter, numerous case reports have been published in which plain film radiographs could not be relied upon to accurately identify intraocular foreign bodies.3, 4, 5 Bray and Griffiths6 found a low sensitivity and specificity, when screening for intraocular foreign bodies was performed by plain film radiographs. The presence of such clinical examples and the low sensitivities and specificities of plain film X-ray in identifying intraocular foreign bodies highlights the fact that such an imaging modality will not exclude any degree of certainty of the presence of an intraocular foreign body. As such, many have even suggested that clinical evaluation by an ophthalmologist may be a more effective means of evaluation.3

The general consensus throughout the literature is that the most effective means of evaluating the presence of soft tissue foreign bodies is through the use of CT scanning.4, 7, 8 Moreover, the use of ultrasound biomicroscopy has also been found by several investigators to be significantly better in identifying intraocular foreign bodies compared with X-ray.9, 10, 11 In fact, ultrasound has even been reported to detect intraocular foreign bodies in several instances when CT scanning failed to do so.10, 11

To the best of our knowledge, this is the first study that we are aware of in which the efficacy of plain film radiographs as a diagnostic test for retained surgical needles was evaluated experimentally. We have shown a low sensitivity and specificity for both ophthalmologists and radiologists reviewing X-rays used as a screening method for suspected retained suture needles. As we only used 10–0 nylon suture needles, extrapolation of these results to surgical cases in which larger bore needles are used should be performed cautiously in light of the aforementioned data by Otto et al.2 However, given the results in that study, which found that intraorbital steel wires of less than 0.12 mm3 were undetectable by plain film radiographs, it might be inferred that suture needles greater than 8–0 would likewise be poorly identifiable.

The institution of generalized protocols mandating portable, plain film radiographs to be taken on every patient in the operating room when suture needle counts are inaccurate may not be appropriate in ophthalmologic surgical cases. In most ophthalmologic cases at this institution, inaccurate needle counts are on account of dropping of the needle during transfer from nurse to surgeon or surgeon to nurse and difficulty in locating the needle because of the small size. In those instances, the needle is not dropped near the surgical field. In most ophthalmic cases, with a small surgical field and microscopic wounds, it would be very difficult for a surgeon to lose a needle in the eye without being aware of the occurrence. The cost of such a mandate requiring the use of plain films, technician time, operating room personnel time, and radiological interpretation may not be the best way to handle inaccurate needle counts in ophthalmic surgical cases. It may be in the best interest of the patients and the institution to reform the X-ray policy in ophthalmic cases in which the needle count is inaccurate at the end of the case. It may be more effective to have a policy in eye cases, which mandates exploration of the surgical site by the ophthalmologist. If there is suspicion on the part of the surgeon that a needle or needle remnant is misplaced in the surgical site, a policy requiring CT scanning or ultrasound evaluation depending on the clinical cirmcumstance, may be more appropriate.