Patients included in the study were 1252, 907 female and 345 male, with a female to male ratio of 2.6:1 and an average age of 53.428. Considering the site of the intervention, 474 procedure were executed at Parma University Hospital, 504 at Cagliari and 274 at Ferrara University Hospital. Patients with missing data were previously excluded from the study.
Medium preoperative TSH was 1.562 uU/ml and calcium was 9.425 mg/dl. Demographic and preoperative laboratory tests are reported in Table 1.
Preoperative cytological examination by FNA was performed in 715 patients and the result was reported according to the Bethesda System for Reporting Thyroid Cytopathology. (6) Results are reported in Table 2. Patients without nodularity at preoperative thyroid ultrasound examination did not undergo cytological examination.
Total thyroidectomy was performed in 1022 cases, lobectomy in 230; in 130 patients, a lymph node dissection was also performed.
NIM was used in 958 procedures, according to the habits of the different operating units. In fact, in two centers it was routinely used while in one it was used only in complex selected cases, such as neoplasms suspected for perithyroid infiltration, voluminous goiters, reoperations or laterocervical lymphadenomies. In 24 cases a loss of signal was recorded. Overall, 8.9% of patients underwent postoperative fibroscopy, which in one center was performed routinely while in the others only in case of persistent dysphonia 30 days after surgery. Speech therapy was necessary in 3.4% of cases.
Complications arising on day 1, day 7, day 30 and at 6 months were then evaluated. On day 1, bleeding was found in 3.2% of patients, paresthesia in 6.4%, dysphonia in 4.8%. Calcemic in the first postoperative day showed an average value of 8.47 mg / dl, ranging from a minimum of 5 mg / dl to a maximum of 12 mg / dl. On the 7th day, paresthesia were found in 1.7% of cases, dysphonia in 4.4%, dysphagia in 0.9%. On day 30, paresthesia were found in 1.3% of patients, dysphonia in 3.8% and dysphagia in 0.3%. After 6 months we recorded paresthesia in 0.5%, dysphonia in 1.8% and dysphagia in 0.5%.
The relationship between the type of surgery (lobectomy or total thyroidectomy) and the onset of complications at the various established follow-up steps (1st, 7th, 30th day and 6 months) was analyzed. (Table 3)
Bleeding on day 1 occurred in 1.9% of patients undergoing lobectomy and in 3.7% of patients undergoing total thyroidectomy, with no statistically significant difference. Paresthesia occurred on day 1 in 0.9% of patients undergoing lobectomy, while in patients undergoing total thyroidectomy they occurred in 8%, thus showing a statistically significant correlation (p = 0.000). The correlation between dysphonia in day 1 and type of intervention also proved to be statistically significant, occurring in 5.9% of patients who underwent total thyroidectomy (5.9%), while only in 1.9% of lobectomy cases. Regarding the complications arising on day 7, 30 and after 6 months (paresthesia, dysphonia, dysphagia) no significant correlations emerged; however, there is a higher percentage of dysphonia among lobectomies (6.3%) compared to total thyroidectomies (3.7%), in all follow-up phases. Nevertheless, among patients undergoing lobectomy who presented dysphonia on day 7, a loss of NIM signal was recorded in 76.9%, there was no loss of IONM signal in 7.7% and the IONM has not been used in 15.4%
As this is a retrospective study, it was not possible to trace which of these patients were originally enrolled for a total thyroidectomy and therefore how many procedures became lobectomies due to intraoperative NIM signal loss, according to the indications of the two stage thyroidectomy. (7) In the pre-IONM era, these cases would have relapsed into the group of patients undergoing total thyroidectomy.
Considering the loss of NIM signal and the onset of dysphonia on the entire sample of patients, only one patient presented this association on the 1st postoperative day, while we found a statistically significant association on the 7th (p = 0.000) and 30th day (p = 0.000) and after 6 months (p < 0.001).
It was then assessed whether the onset of paresthesia could correlate with the calcemic values detected on the 1st postoperative day. Paresthesia were recorded in 6.48% of cases in the first postoperative day.
In patients without paresthesia on day 1, the calcium has an average value of 8.53 mg / dl, with a minimum of 7 mg / dl and a maximum 12 mg / dl, while in patients who presented paresthesia the average calcium is attests to a significantly lower value of 7.59 mg / dl on average, with a minimum value of 5 mg / dl and a maximum of 9 mg / dl. A correlation between the calcemic value in the first day and the development of paresthesia was highlighted in all the stages considered (p = 0.000 in the 1st, 7th and 30th day, p = 0.004 at 6 months).
A correlation was sought between the onset of paresthesia and the presence of parathyroid glands in the surgical specimen. However, no statistically significant relationships emerged. In fact, only 9.2% of patients with presence of parathyroid glands at the definitive histological examination, presented paresthesia on the first day, on the 7th day 2.5% and on the 30th day 1.7%; on the other hand, among patients negative for the presence of parathyroid glands, paresthesia were recorded in 1.6% in day 7 and in 1.3% in 30th.
After 6 months, among patients in whom the presence of parathyroid glands was found, there was not even one case of paresthesia, while 0.6% of those who did not have parathyroid glands accidentally removed at the final histological examination were recorded.
The mean length of hospitalization was 2.459 days, showing a minimum of 1 day and a maximum of 14 days. Patients who experienced bleeding on day 1 had a hospital stay of 4.174 days while those who did not have bleeding reported a mean hospital stay of 2.076 days. Even the onset of paresthesia in the first day minimally prolonged hospitalization, with an average duration of 2.625 days compared to 2.265 days for patients without paresthesia. Similarly, the development of dysphonia in day 1 led to an average hospital stay of 3.730 days compared to 2.062 days for patients without dysphonia. Analyzing data with the Kruskall-Wallis test, a statistically significant difference was highlighted between the hospitalization of patients who presented at least one of the complications considered in the 1first day and the duration of hospitalization (p < 0.000 in all cases).
Finally, we evaluated the possible association between thyroid disease, divided into 5 diagnostic categories (differentiated carcinoma, medullary carcinoma, non-hyperfunctioning benign disease, hyperfunctioning benign disease, NIFTP) and the onset of complications.
The analysis found only a statistically significant correlation between the hyperfunctioning benign disease and the onset of bleeding and paresthesia in first postoperative day. Full results are shown in Table 4.
At multivariate analysis, intraoperative corticosteroid administration (OR = 5.682; CI: 1.2329–26.1859; p = 0.025) and the use of haemostatic agent during surgery (OR = 2,928; CI: 1.1383–7.5345; p = 0.025) were found as independent risk factors for postoperative dysphonia (Fig. 1).
Male sex (OR = 4.606; CI: 1.8132–11,7053; p = 0.001) was found as independent predictive factor for postoperative bleeding (Fig. 2); and a postoperative calcemia < 8.0 mg/dl (OR = 7.994; CI: 3.555–17.9763; p < 0.001) was identified as independent predictive factor for paraesthesia (Fig. 3).