This study assessed the feasibility of using MSF to determine the presence of a miscounted needle and identify misplaced needle in the surgical field. MSF significantly improved the time to determine the presence of a needle and the time to find a lost needle in the both the ex-vivo laparoscopic model and the in-vivo porcine model of laparoscopy and laparotomy.

In Clinical Scenario 1, the instrument allowed the users to quickly process through the hay and locate the needle without any visual cues based on auditory and visual signal. As each user sifted through the hay, the device was specific enough to guide the user to the needle within millimeters using the changes in the signals. As a result, the success rate of identification of the needle was 96% and the average time to localize was less than two minutes. The difficulty of directly visualizing the needle in the haystack is demonstrated by the 22.0% rate of needle identification without the device. This rate likely would not have improved significantly even if the users were allowed more than five minutes.

Clinical Scenario 2 tested the user’s ability to accurately determine whether a needle was present in the haystack. The MSF group had 100% accuracy while taking less than 5 min to decide. It is worth noting that in most trials, the user made this determination based solely on the device’s auditory feedback without visualizing the needle. This is akin to the detection of RFID embedded sponges with a wand external to the patient [18]. Without the device, the user was entirely dependent on visualization of the needle, which proved to be inefficient.

This laparoscopic model simulated the difficulties encountered in finding a needle in a real surgical field: poor visualization, narrow optics, a large heterogenous space, and the potential for the needle to shift positions. Recognizing the inherent limitations of this ex vivo model, the MSF was further evaluated in live porcine models.

The in-vivo trials were designed to simulate an intraoperative needle miscount in a human patient. Our porcine model aimed to simulate clinical scenarios of RSS with anatomic similarity, dynamic location of the needle in midst of the search, interference of standard operating table with its metallic components, respiratory and cardiovascular variations and limitations of surgical approach (laparotomy and laparoscopy).

The in-vivo results suggest that MSF is non-inferior to C-arm in its accuracy of determining the presence of a needle. Groups that used MSF had a very high accuracy in determining the presence of a needle (overall 92%). When compared to manual search, MSF demonstrated a significant improvement in user accuracy (97% vs. 46.7%). MSF had a positive predictive value (needle present) of 95.4% and negative predictive value (needle absent) of 100%. The time to determine the presence of a needle was also significantly shorter when using the MSF by an average of almost two minutes. In addition, multivariable analysis demonstrated that MSF use is independently correlated with a significant increase in the accuracy of identifying a lost needle in the abdomen, and MSF use had a stronger correlation to accurate determination of needle presence than X-ray.

The time to determine the presence of a needle was similar between C-arm and MSF. However, this does not factor in the additional time needed to obtain and interpret an X-ray, though in our trials both the C-arm and radiologist interpretation were immediately available. Our trials also imaged the abdomen by quadrants with the option to obtain additional images as needed to better examine certain quadrants, which may increase its sensitivity by focusing the radiologist’s attention. This arrangement is likely unrealistic in a clinical setting. The delay in obtaining and interpreting an X-ray has significant impact as it lengthens anesthesia time and increases perioperative risk for the patient while also taking up valuable OR time for the staff and the health system. The use of MSF may help to eliminate this time by effectively ruling out presence of needle without waiting for the X-ray or its interpretation.

Though there have been significant technological advancements to prevent soft retained items such as the radiofrequency identification of sponges, there has not been a widespread adoption of devices to prevent or aid in finding an RSS. One study explored the use of UV fluoroscopy to retrieve fluorescent coated needles and found that using the device improved surgeon’s time to retrieve the needles [19]. However, this would require hospital wide implementation of exclusive use of fluorescent coated needles. Another study found that using a magnetic retriever significantly reduced the search time for a lost needle [20]. This device inherently carries a risk of causing additional injury to organs during the search, which may be preventing its widespread adoption.

Currently, in the case of a miscount, the patient is kept under anesthesia while the surgical team performs a manual search and/or awaits X-ray images to be obtained and interpreted by a radiologist. In recent survey data, this is estimated to take 31–40 min of additional time per miscount event [21]. The combination of undue patient risk from the exposure to additional anesthesia and radiation coupled with the significant OR delays can create a frustrating environment for the surgical team. This study demonstrates the feasibility of a FDA-approved device which may be valuable in both the prevention and identification of RSS in the surgical field.

This study has several limitations and shortcomings. included a limited sample size. It is possible that the results may have been different with more trials. However, our outcomes were consistent throughout both ex-vivo and in-vivo experiments and the statistical analysis was robust to demonstrate that MSF is effective in determining a presence of a needle and finding it, if needed.

Due to the nature of the device indiscriminately finding metallic objects, there was an unusual situation that required the research coordinator to intervene during one of the laparoscopic trials. Persistent false positive signals were found while searching through the pig’s stomach and small bowel, assumed to be due to the pig’s ingestion of bars of the cage and metal filings. As a result, this required adjustment of sensitivity of the device, which allowed the user to facilitate effective search for the needle. During the laparotomy trials, this was not an issue.

Because MSF is a user-operated device, the results from this study may be affected by human error. As previously stated, the users were not instructed in how to conduct an intra-abdominal cavity search with MSF. As a result, each user had a different approach to their search, and this may have been a confounder in this study. While beyond the scope of this study, the authors encourage the use of a systematic search when using MSF to maximize accuracy and efficiency. We recommend the following methodology:

1.

Sweep the right upper quadrant, making sure to include the large area obscured by the liver, especially the posterior side of the liver and the hepatoduodenal ligament.

2.

Search above the body of the stomach and the gastrocolic ligament.

3.

Retract the stomach caudad and search the left upper quadrant including perisplenic region.

4.

Sweep the right and left paracolic gutters.

5.

Perform a pass over the bowel prior to manipulation. When running each portion of the bowel, sweep ahead of the area you are about to manipulate, beginning with the jejunum at the ligament of Treitz and running the bowel to the terminal ileum.

6.

Perform a sweep of the lower mid-abdomen and pelvis making sure to arrive on either side of bladder.

In addition, when the users were surveyed regarding their experience with the device after the trials, they collectively felt that the use of a sweeping motion or making concentric enlarging circles while working in different quadrants was the most effective way to rule out a sharp in a specific area.