Materials and methods

Study design and ethical approval

This was a single-center, prospective, cross-sectional, observational correlative study conducted from June 2024 to May 2025 at Era’s Lucknow Medical College and Hospital, Era University, Lucknow, India. This study was approved by the Institutional Ethics Committee, Era’s Lucknow Medical College and Hospital, Era University, Lucknow (IEC Registration No. ECR/717/Inst./UP/2015/RR-21; approval reference ELMC&H/R.Cell/2024-A/208; approved on May 18, 2024), and was conducted in accordance with the Declaration of Helsinki. Written informed consent was obtained from every participant, and confidentiality was maintained throughout.

Study population and sampling

Consecutive patients with clinically palpable or incidentally imaging-detected thyroid swellings were screened for eligibility. Of the 192 patients screened, 178 met the inclusion criteria and were enrolled: age >16 years, palpable or imaging-detected swelling, willingness to provide consent, and complete TSH, USG, and FNAC data. Patients with prior radioactive iodine therapy or neck irradiation, previous thyroid surgery, a non-diagnostic aspirate (Bethesda category I), refusal of consent, or incomplete data were excluded. Of the 192 screened patients, 14 were excluded according to the prespecified eligibility criteria. A detailed flow diagram is shown in Figure 1.

Figure 1: Flow diagram of patient enrollment, index testing, and histopathological verification.The image was created by Farheen Khan and Nishi Tandon using Microsoft Paint (Microsoft Corporation, Redmond, WA).FNAC, fine-needle aspiration cytology; HPE, histopathological examination; TSH, thyroid-stimulating hormone; USG, ultrasonography; NIFTP, non-invasive follicular thyroid neoplasm with papillary-like nuclear features

Unit of analysis and index nodule selection

The unit of analysis was the patient. In patients with a solitary thyroid nodule, the same nodule was assessed by ultrasonography, FNAC, and histopathology. In patients with multinodular thyroid disease, the index nodule was defined as the nodule with the highest ACR TI-RADS category or, when categories were similar, the largest nodule that underwent FNAC and guided the surgical decision. Cytological, ultrasonographic, and histopathological findings were matched using lesion location, size, and operative/pathology records.

Thyroid function assessment

Serum TSH was measured in all patients using an Abbott Alinity i chemiluminescent immunoassay analyzer (Abbott Diagnostics, Abbott Park, IL; reference range 0.4-4.5 mIU/L) according to the manufacturer’s instructions. Patients were classified as euthyroid, hypothyroid (TSH >4.5 mIU/L), or hyperthyroid (TSH <0.4 mIU/L).

Ultrasonographic evaluation

A single experienced radiologist performed high-resolution B-mode ultrasonography using a GE LOGIQ P9 ultrasound system (GE Healthcare, Chicago, IL, USA) equipped with a 7.5-12 MHz linear-array transducer to limit interobserver variability. Nodules were scored according to ACR TI-RADS across composition, echogenicity, shape, margin, and echogenic foci, yielding a category from TR1 through TR5 [6].

Cytological evaluation

FNAC was performed by an experienced cytopathologist using a 23-25 gauge needle, with ultrasound guidance for complex or cystic lesions. Smears were stained using Papanicolaou and May-Grünwald-Giemsa methods and classified according to the Bethesda System, categories II through VI; category I specimens were excluded per protocol [3]. The radiologist was blinded to cytological findings at the time of ultrasonographic interpretation. The cytopathologist was blinded to the ACR TI-RADS category. Histopathological assessment was performed as part of routine clinical care; complete blinding to FNAC and TI-RADS findings was not ensured.

Histopathological evaluation

Of the 178 patients, 68 (38.20%) underwent surgery. Specimens were fixed in 10% neutral buffered formalin, routinely processed, paraffin-embedded, sectioned at 3-5 µm, and stained with hematoxylin and eosin. Immunohistochemistry was performed where required to establish the final diagnosis. Diagnoses followed the 2022 WHO Classification [8]. For the primary diagnostic performance analysis, histopathological outcomes were classified as benign versus malignant/borderline, with NIFTP included in the malignant/borderline group because it is a diagnostically challenging neoplasm that commonly requires surgical excision for definitive classification.

Statistical analysis

Analyses were performed in Python (pandas, NumPy, SciPy, scikit-learn). Bethesda V/VI and TI-RADS 4/5 were defined as positive for malignancy in the 68 surgically confirmed cases. The reference outcome was defined as malignant/borderline histopathology, with NIFTP included in this group for the primary analysis. Sensitivity, specificity, PPV, NPV, and accuracy were calculated against HPE, with 95% CIs calculated using the Wilson score method [9]. Categorical associations were tested using the chi-square test or Fisher’s exact test, as appropriate. Intermodality agreement was quantified using Cohen’s kappa (κ): 0.01-0.20, slight; 0.21-0.40, fair; 0.41-0.60, moderate; 0.61-0.80, substantial; and 0.81-1.00, almost perfect [10]. Significance was set at p < 0.05. Diagnostic performance was evaluated in the surgically verified subgroup only. For the primary FNAC analysis, Bethesda categories II, III, and IV were classified as negative, and Bethesda categories V and VI were classified as positive. As surgery was performed according to routine clinical indications rather than the study protocol, the estimates represent test performance among operated patients and should not be interpreted as population-level estimates for all patients presenting with thyroid nodules.