Little information is available regarding the frequency of thyroidectomy-related malpractice claims. Few studies have assessed claims that were settled or dropped before trial, providing only a limited view of the medico-legal environment. From the database of the Physician Insurers Association of America (PIAA), claims from 1985 to 2008 with thyroid-related procedure codes were evaluated for claimant information, insured's specialty, loss description, causation, and claim outcomes (1). During the 24-year period reviewed, 380 claims related to thyroid surgery were reported. One hundred twenty-eight claims (33.7% of total claims) resulted in an indemnity payment either due to settlement or a finding against the defendant. The average indemnity payment was $185,366 (range, $363 to $2,000,000). Among the cases in which a specific outcome was reported, 55 were related to laryngeal nerve injury or voice disturbance. No substantial change occurred in the incidence of claims across the study period. During this time, approximately $2,585,000 thyroidectomies were performed. Extrapolating from the PIAA data, this represents an estimated 5.9 claims per 10,000 cases (1).

Using the LexisNexis Academic legal database, a retrospective review of all relevant federal and state cases from 1989 to 2009 was performed using the search terms “thyroid,” “surgery,” and “medical malpractice” (2). A total of 143 medical malpractice cases involving thyroid surgery was retrieved from our search from 1989 to 2009. After reviewing all cases, 33 cases in which the alleged negligence occurred after thyroid surgery were used for analysis. Of these cases, 15 involved recurrent laryngeal nerve (RLN) injury. Interestingly, no mention of RLN monitoring was noted in any of the cases (2).

For the evaluation of a vocal cord paralysis (VCP) that has occurred after a thyroid surgery using intraoperative neuromonitoring (IONM), it is of crucial importance to take into account whether or not the published standards for proper use of IONM were applied correctly, and whether or not changes of neuromonitoring signal and electromyogram (EMG) were correctly interpreted intraoperatively and accordingly responded in a risk-oriented manner with regard to the further operational strategy of the second side thyroid lobe (34).

On the basis of results from the current literature, it can be concluded that the frequency of VCP has decreased over the years since the introduction of IONM in thyroid operations (5). However, in individual cases, bilateral paresis can be reliably avoided, but unilateral paresis cannot be safely avoided with the use of the IONM (67).

The current report derives from a review of the literature and personal experience on malpractice claims. Since no systematic analysis has been performed so far for the assessment of postoperative paresis of the superior laryngeal nerve (SLN) and continuous intraoperative neuromonitoring (CIONM), the present study is limited to the assessment of postoperative VCP as a result of injury to the RLN and of intermitted IONM.

Briefly, the IONM legal issues that will be presented here can be summarized into: 1) bilateral RLN palsy prevention by IONM; 2) EMG documentation from monitoring; 3) the use of IONM in difficult cases; 4) standard of care of monitoring in thyroid surgery; 5) appropriate and standardized use of IONM; and 6) costs of IONM and RLN palsy (Table 1).

RLN injury is a well-known and serious complication of thyroid surgery. A recent study estimated the frequency of post-thyroidectomy RLN palsy and identified possible risk factors (8). Based on the Danish national thyroid surgery database, 6,859 patients treated with thyroid surgery from January 1, 2001 to December 31, 2008, at the Danish Departments of Otolaryngology-Head and Neck Surgery (ENT-HNS) were analyzed (8). Unilateral RLN palsy was found in 2.1%, while 0.1% had bilateral RLN palsy. In benign histology, RLN palsies were registered in 1.3%. Malignant histology and accordingly neck dissection were the most predominant risk factors with a relative risk (RR) of 5.4 and 5.8, respectively (8). In benign cases, previously performed thyroid surgery had an RR of 10.4. Malignant histology, neck dissection, and previous thyroid surgery are the strongest predictors for RLN palsy (8).

The surgical experience pertaining to the identification, exposure, and protection of the RLN has a demonstrable influence on the nerve paresis rates (89). A study showed that the rate of postoperative permanent laryngeal nerve paresis correlated directly with the number of nerves at risk (NAR). A palsy rate of less than 1% was achieved when the operating frequency was 60 and more NAR per surgeon and year (9).

An increased risk of RLN palsy is preoperatively detectable when a locally advanced cancer is diagnosed or a retrosternal huge struma is detected on imaging. Graves' disease, Hashimoto, de Quervain, or Riedel's thyroiditis also indicate a potentially increased risk of RLN palsy (89). Particularly high is the risk of bilateral paresis with pre-existing unilateral paresis and the need for bilateral resection. In this situation, it is critical to check whether there is a need for a resection to be performed contralaterally to the pre-existing RLN paretic side.

Anatomic criteria determining patients with high-risk thyroid surgery have been described and defined based mostly on expert opinion. Preoperatively unpredictable risk situations, which can only be detected intraoperatively, are one of the main reasons for using neuromonitoring not only selectively but also routinely. Examples of intraoperative risk situations include: 1) atypical RLN pattern anterior to the thyroid gland; 2) RLN anterior to the Zuckerkandl's tuberculum nodule; 3) fixed, splayed, or entrapped RLN with capsular association through fascial bands; 4) vessels or goiters change; 5) invaded RLN; 6) posterior ligament of Berry nerve; 7) thin RLN (<1 mm); 8) ramificated or antevascular RLN; or 9) non-RLN occurring in about 1.5% of the thyroid gland operations (Table 2). Atypical courses of the RLN are observed in approximately a quarter of the thyroid gland operations. Therefore, IONM-based treatments are recommended and the centralization of thyroid surgery in larger units using IONM might improve quality (8).

The correct reading and interpretation of neuromonitoring EMG signals are essential to avoid any medico-legal errors.

Patient thyroid history, preoperative radiologic imaging, pre-operative symptoms, pre- and postoperative laryngeal examination, exhaustive surgical report with intraoperative findings, identification of RLN, pictures, and videos are important to document any pre-existing airway impairment or correct surgical deliberations.

The documentation of the EMG muscle action potential is possible with all modern IONM devices. IONM converts muscle activity into recorded electromyographical signals, which are often possible to print out. It not only serves the patient's legitimate interest in understanding the use of neuromonitoring but also primarily communicates to the surgeon the important information during the operation (i.e., artifact signal, amplitude decline, and latency increase). Documentation of the final normal neurophysiologic signals of the RLNs at the end of the surgical procedure may have a forensic function, with early differentiation between RLN related and unrelated voice changes. From a medico-legal point of view, it is recommended to record and document the muscle action potential at the beginning (V1; R1) and at the end (V2; R2) of the resection per each side of resection. When IONM is used, its documentation is required and includes time-traceable measures of amplitude, latency, waveform morphology, and magnitude of stimulating current. EMG curves are a proof of intact nerve function. Monitoring nerves during surgical procedures may potentially reduce the medico-legal liability for surgeons, with a reduction of economic loss for the patient, health system, and insurance companies. Finally, the recorded nerve signal is important for early differentiation between RLN related and unrelated voice changes as well (24).

Reduction of major injury to patients such as bilateral RLN palsy is a great advantage in the modern era of thyroid surgery. A possible consequence of LOS of RLN at the first side is an accurate consideration of optimal contralateral surgery timing. With LOS, the surgeon must consider that the ipsilateral nerve is injured at least temporarily and can subsequently consider whether it is important and to the patient's best interest to perform surgery on the contralateral side on this day given the new intraoperative information of ipsilateral paralysis (3). According to Goretzki et al. (25), a failed IONM stimulation of RLN after resection of the first thyroid lobe is specific enough to reconsider the surgical strategy in patients with bilateral thyroid disease to prevent bilateral RLN palsy (25). Indeed, in 85% of cases with known nerve injury and in 56% with negative IONM stimulation at the first side of dissection, the surgical strategy was changed with no postoperative bilateral RLN palsy. This was in contrast to the 17% bilateral RLN palsy (P<0.05) when surgeons were not aware of a preexisting or highly likely nerve injury at the first side of thyroid dissection (25).

After LOS on the first side of resection in bilateral goiter, more than 93% of surgeons declared their willingness to change the resection plan for the contralateral side to avoid the risk of bilateral RLN palsy. This resulted in a discontinuation of surgery (84%) or in undertaking a less extensive resection for completion of the other side than originally planned (9%) (7). Hospitals with the heaviest institutional caseload reportedly changed their surgical plans more often than those with a lower institutional caseload (7). Likewise, hospitals that routinely employed vagal stimulation and recording of EMG responses refrained from continuing the resection on the unaffected side more frequently than those that did not (7).

IONM is mentioned in the preoperative informed consent for primary and secondary interventions. If the patients are informed about the use of IONM, they should also be informed about the intraoperative consequences such as LOS. The possibility to stage the procedure in case of LOS on the first side must be specified in the preoperative informed consent. Patients are informed of the possibility of staging the contralateral side in case of LOS, in respect of the patient autonomy and with emphasis on shared decision making. In fact, most patients perfectly agree on this concept and comprehend the tremendous benefit in deciding whether to go to the other side in case of LOS at the first dominant side.

In a malpractice analysis, IONM provides important prognostic information regarding the ipsilateral VC function at the completion of the initial side of the surgery. IONM allows the surgeon to stage contralateral surgery if RLN damage is diagnosed, thereby avoiding the potential for bilateral VCP. If IONM is used, the surgeon has to consider the result and eventually stop the operation at the first side. If not otherwise stated, the operation has to be stopped in case of benign diseases. If the surgeon proceeds, he or she should outline the reasons (e.g., malignant bilateral disease, severe patient co-morbidity).

If an injury occurs, the surgeon must carefully examine arguments and justify the fact that monitoring was not used selectively in a difficult case. In 2009, we tested the hypothesis that the RLN identification during thyroid surgery reduces injury, and that intraoperative nerve monitoring may be of additional benefit (5). This represented the only randomized prospective trial for IONM vs. no-IONM. One thousand consenting patients scheduled to have bilateral thyroid surgery were randomized to standard protection or additional nerve monitoring. The primary outcome measure was the prevalence of RLN injury. Of 1,000 NAR in each group, transient and permanent RLN injuries were found respectively in 38 and 12 nerves without RLN monitoring (P=0.011), and 19 and 8 nerves with RLN monitoring (P=0.368). The prevalence of transient RLN paresis was lowered by 2.9% in high-risk patients who had RLN monitoring (P=0.011) and 0.9% in low-risk patients (P=0.249). The negative and positive predictive values of RLN monitoring in predicting postoperative VC function were 98.9% and 37.8%, respectively. We concluded that nerve monitoring decreased the incidence of transient, but not permanent, RLN paresis compared with visualization alone, particularly in high-risk patients (5).

The national thyroid surgeons' guidelines for IONM use propose IONM in risk operations (26). Thus, according to the above prospective trial and societies guidelines, thyroid surgeons should carefully consider justification if IONM was not used in high risk procedures (526).

A recent analysis of claims of the years 1995 to 2014 (n=102) revealed a different pattern of published operational statistics due to the subjective and objective impairments of the quality of life (27). While the rate of persistent hypocalcemia and vocal palsy was low (<3%) in surgical statistics, these complications dominated the causes of complaints in courts (unilateral vocal palsy, 30%; bilateral vocal palsy, 20%; hypoparathyroidism, 15%). Unilateral and bilateral RLN paresis occurred during the first assessment period (1995–2000) predominantly after thyroid gland partial resections, and in the third observation period (2008–2014) after total thyroidectomy (27).

IONM has been used for the RLN functional control since 2001, and in 2008, in cases of vocal palsy, the dispute between plaintiffs and defendants at conciliation centers and courts has been the subject of almost all cases (27). In 80% (2001–2007) and 70% (2008–2014) of the RLN damage cases using the IONM, the specific use of neuromonitoring did not follow the recommendations published in the literature at the time of the damage (327). The necessary standardized stimulation of the VN (V1 and V2) was dispensed, and only the signal tone after nerve stimulation was considered. Type 1 signal failures could not be detected in the absence of vagal stimulation, or artifact signals were confounded with regular muscle action potentials if the EMG were not considered (27).

Considering the fact that most vocal palsy is the result of nonstructural and exclusively functional RLN lesions, which are equally effective in vocal folds, it is understandable that bilateral vocal palsy occurred predominantly after thyroid operations without neuromonitoring. However, there is cause for concern that almost all bilateral vocal palsy have occurred during the last reported period (2008–2014) despite the use of neuromonitoring, mostly as a result of non-compliance with the standards (27).

We stated that, for the evaluation of a VCP after a thyroid surgery using IONM, it is of crucial importance to determine whether or not the standards for proper IONM use were applied correctly, and whether or not changes of neuromonitoring signal and EMG were correctly interpreted intraoperatively (3427). This results in the need for continuous surgical training in the correct application of standardized procedures of neuromonitoring, especially with regard to the indispensable consideration of EMG and vagal stimulation before and after resection of the first page in all planned bilateral resections.

The basis of a malpractice litigation (especially in the US) is the breach of the standard of care, i.e., if a surgeon provided a patient with care that is less than the standard and the patient was harmed, the surgeon will likely lose the lawsuit (28). The minimum standard of care is what other doctors in the area would provide. The questions remain as to: 1) what constitutes the standard of RLN management in the modern era of thyroid surgery; 2) what is the prevalence of neuromonitoring in thyroid surgery; and 3) how many surgeons use RLN monitoring. Horne et al. (29) reported in 2007 that more than 45% of the otolaryngologists use RLN monitoring. Sturgeon et al. (30) reported in 2009 that 37% of the general surgeons attending the American Association of Endocrine Surgeon meeting in 2006 used RLN monitoring; young surgeons (age, 35–55 years) and academic surgeons with high volume cases (probably dealing with more difficult cases) were most likely to use IONM always (30). This percentage increased from 7% to 37% between 2001 to 2007 (31). In Germany, more than 50% of thyroidectomies are performed with IONM (32). In Northern Europe, about 77% of thyroidectomies are monitored (32). Thus, IONM will be the standard of care for thyroidectomy in a very near future.

In Italy, hospitalization costs for a thyroidectomy with IONM range from €3,713 to €3,770, that is 5%–7% higher than those for traditional thyroidectomy without IONM (33). Direct IONM cost ranged between €250 and €270 per single procedure (33). From personal communications, IONM costs are extremely different between countries, with even higher costs of €800 per procedure. There are patients who cannot afford the costs of such technology. Moreover, there are countries and health care systems that cannot afford such expenses. For this reason, the abovementioned issues should be carefully considered in these contexts.

In conclusion, we believe that the clinical benefits of IONM are definitely more important than the medico-legal issues presented here. Clinical benefits of IONM are both for the patients and the thyroid surgeons and include: 1) reduction of RLN injury in high risk procedures; 2) reduction of bilateral VC injuries; 3) reduction of permanent nerve injuries (34); 4) clear benefits for residents in training and less-experienced surgeons (6); 5) intraoperative, but not postoperative, evidence of RLN injury; 6) identification and dissection of the RLN; 7) facilitating the completeness of a total thyroidectomy; 8) serving the endoscopic thyroidectomy; and 9) providing an important instrument for research (35).

In case of a RLN palsy, the correct evaluation of all issues of the single and unique thyroid operation should be considered as: 1) the individual case (i.e., thyroid disease); 2) the procedure (i.e., central neck dissection); 3) the position of the surgeon (i.e., experience); 4) the experience of the surgeon with IONM; and 5) the documentation (printed\traceable\digital) of the VN stimulation. It is important in a monitored procedure, in the operative notes a statement that demonstrates that the RLN was visualized and was intact (IONM) at the end of the procedure. All these issues, together, are useful to draw conclusions from the standpoint of surgery, both for the surgical deliberation and for forensic.

The patient should be informed preoperatively not only of whether neuromonitoring is to be used during the imminent operation but also about the limitations of this technique and the surgical consequences resulting from a signal failure. This mainly includes the question of the strategy change of a planned bilateral operation if there has been a signal failure on the first operated side. Pre-operative RLN paresis should be clarified preoperatively and the patient should be informed of whether the resection of the opposite side can be spared, even if the benign struma is the larger, but not malignant, side. If there is an indispensable indication for the thyroid resection of the vocal fold palsy side, high-threshold prerequisites are attached to the surgeon's expertise.

IONM, comparably to the intraoperative parathormone determination established for decades, has shifted the previously postoperatively possible diagnosis into the operating room. Therefore, the use of neuromonitoring makes sense only if operative consequences are observed in the case of a recognizable loss of function. There are also marginal cases, which the surgeon “in dubio” should choose for the patient's safer procedure. For neuromonitoring, this means that, in the event of a signal failure of the VN on the first side, the resection of the second side should be avoided in order to prevent bilateral paresis.

Further studies should be carried out to determine whether a postoperatively non-injured vocal function can always be interpreted in cases where continuous neuromonitoring after signal failure indicates a recovery of the stimulation amplitude above 50% of the initial amplitude. For intraoperative management, it would be important to determine, without forcing any nerve-compromising actions (e.g., tension or compression of the nerve), how long after signal failure the amplitude should be maintained and whether, for example, intraoperative corticosteroids should be given.