The age group commonly affected with thyroid lesions in this study is 20-59 which accounts for 88.9% of the cases with average age of 34.8. This is in agreement with a study done in Tikur Anbessa Hospital, Pakistan, in which 85% of the patients with thyroid lesion fall under the age group 20-59 [7]. Diseases of the thyroid are more common in women with female to male ratio of 4.1:1 with 468 (80.6%) cases seen in females. This is consistent with studies done in Kenya, Nigeria and Saudi Arabia with variable ratios reflecting different study populations [8–10]. High ratio of thyroid lesions in female sex was suggested to be from level of hormonal variation after puberty and this study does seem to validate this scientific explanation [11].

Non-neoplastic thyroid lesions constitute 432 (74.3%) of overall cases. MNG by itself accounts for 417 (71.8%) of overall thyroid lesions. This is consistent with a study done in SPHMMC (Addis Ababa) in which 152 (68.4%) of the cases were MNG [12]. This picture is also found to be a consistent finding when compared to similar studies done in Uganda that showed MNG to be the commonest consisting of 90 (65.7%), Yemen 164 (63.1%) and Bangladesh 81 (75%) of the cases [8,13,14].

Thyroiditis is seen in twelve (2.1%) of the cases with female to male ratio of 5:1 and a mean age at presentation of 34.8 + 11.4 with 8/581 (1.3%) of the overall thyroid lesions being chronic lymphocytic thyroiditis/Hashimotos thyroiditis. Two studies done in India show frequency of Lymphocytic thyroiditis at 30 (2%) and 37 (1%) with mean age of 40+ 10 which is in agreement with this study.15,16 A study done in Italy also shows the frequency to be 54 (1.4%) of the cases [17].

Neoplastic diseases of the thyroid were found to be 149 (25.6%) constituting the second most common thyroid lesion in the study with female to male ratio of 5:1. Benign neoplasms were seen in 54 (9.3%) of the cases with follicular adenoma being the most common 42 (77.8%) followed by 7 cases of hurthle cell adenoma (12.9%), NIFTP 5 cases (9.3%) and 1 case (1.8%) of well differentiated tumour of uncertain malignant potential (WDTUMP). This figure is comparable with similar study conducted by Tsegaye B and Ergete W. in Addis Ababa in which 164 (21%) were thyroid neoplasms in which 100 (60.9%) were adenomas.7 Study done in Western Kenya indicates occurrence of benign neoplasms to be 51 (13.1%) which is comparable with this study [18]. A retrospective study done in Southern Nigeria shows frequency of 36 (12.1%) benign neoplasms with similar female to male ratio of 5:1, with follicular adenoma constituting thirty-three (91%) of the cases [19]. A slight variation in the frequencies can be due to different sample sizes and duration of the study. Similar study done in Pakistan shows a total of 255 (26%) neoplastic lesions which is consistent with the current study, however, among the benign neoplasm’s follicular adenomas account for 67 (65.9%) cases which is lower than the figure seen in the current study [20]. This can be explained by the strict histopathologic criteria employed in the study in classifying benign thyroid neoplasms in the aforementioned study which was technically difficult to employ in the current study.

Among the 581 biopsies Ninety-five (16.4% of total cases and 63.8% of all thyroid neoplasms) cases were malignant with female to male ratio of 2.65: 1 and mostly are seen in the age group 20- 29 accounting for 26 (27.4%) of cases. There is overall predominance of malignant thyroid neoplasms over the benign counterparts in this study. As the study hospital is the largest tertiary referral hospital in the southwestern part of the country, most cases are referral cases with suspected malignancy. This can contribute to the higher incidence of cancers in the hospital and this is also a reflection of increasing prevalence of thyroid cancer in the region and mandates larger scale studies to determine the underlying causes and associated factors. However, such figures are also seen in other studies across the continent [10,20,21].

Among the malignant thyroid neoplasms Papillary thyroid carcinoma was the most common with frequency of Sixty-nine (72.6%) cases with female to male ratio of 2.8:1, mean age at presentation of 35.6 and 30 (78.9%) cases have a 1–5-year history of symptoms. This figure is similar with the study conducted in tikur anbessa hospital which shows 49 (76.9%) cases to be papillary thyroid carcinoma [7]. Similar predominance of papillary thyroid carcinoma is reported in other literatures with variable proportions [22–25]. Papillary thyroid carcinoma was the most common malignancy in study done in Nigeria which accounts for 98 (56.3%) cases with peak incidence in the 3rd decade.9 Another study done in Saudi Arabia shows papillary thyroid carcinoma to be the most common malignancy accounting for 105 (74%) of cases with mean age of 37.5 years [26]. Study conducted in Ireland from consecutive 1003 thyroidectomy specimens 261 were malignant in which papillary carcinoma accounted for 75.1% of cases.8 Large scale study in Sao Paulo, Brazil which shows 9607 (72%) cases of papillary carcinoma and similar study done in USA shows 28,414 (86%) cases [24].

Only Fourteen (14.7%) cases of Follicular carcinoma are seen in this study with mean age of 36.9 and presenting within 1 year of development of symptoms in 8 (57.4%) cases. This is in agreement with study conducted in Gonder in which eighteen (29%) of the cases turn out to be follicular [27]. However, this is contrary to the finding of a study done in Harare, Zimbabwe, in which follicular carcinoma was the most common malignancy seen in 60 (70%) patients [25]. This shows that follicular carcinomas may be a reflection of a more severe iodine lack in certain populations or indeed a reflection of some etiological differences for this tumour in the different populations.

Medullary carcinoma is still relatively uncommon with only three (0.5%) cases identified in this study which is in agreement with findings from two studies in Addis Ababa in which only one and two cases were found, Port-Harcourt, Nigeria (one case), Yemen (two cases) as well as reports from Pakistan (four cases) [7,14,20,28]. Only one case of Anaplastic thyroid carcinoma was identified in this study (A 70-year-old male patient) which is in harmony with other reports of rare incidence of this disease entity [21,22,24,29].

Diagnostic sensitivity and specificity of FNAC in this study are 71% and 92.7% respectively. These figures are slightly higher than similar study done by W. Ergete and D. Abebe in which the sensitivity and specificity were 67% and 84.7% respectively [30]. Similar study done in Tanzania shows 66.7% sensitivity and 92.5% specificity [31]. Even lower sensitivity of FNAC (61.5%) in diagnosing malignancy is reported in Uganda [32]. This is also in harmony with other studies in which in areas with a high prevalence of multinodular goitre, the sensitivity for detecting malignancy is relatively low whereas specificity is high [33].

Among 286 FNAC’s cyto-histologic correlation was achieved in 233 (86.6%) cases. Such concordance rate is in agreement with study done in Tikur Anbesa hospital in which 274 cases (82%) were correlated.30 Similarly correlation study conducted in USA shows 88% correlation [33].

False negative and false positive results account for 22 (8.2%) & 14 (5.2%) respectively. This is in agreement with an Egyptian study in which the false negative and false positive rates were 7.2% and 5.8 % respectively [4]. Similar figures are also seen in a study conducted to determine discordance rate between thyroid cytology and histopathology in USA in which 8% false negative and 4% false positive results were found [33]. The overall reported false-negative rates range from 1–11%. According to various literatures it is difficult to establish the true frequency of false- negative results, since only a small percentage (approximately 10%) of patients with benign cytologic findings undergo surgery. However, all reports agree that if all patients with thyroid FNA would also have histologic examinations, the true false-negative rate would be below 5% [6,8,34,35].

Diagnostic accuracy of FNAC in this study was 86.6% with Negative predictive value of 89.1% and Positive predictive value of 80%. Accuracy is slightly higher than a study done in Tikur Anbesa Hospital which shows 82% diagnostic accuracy [30]. A Study done in Tanzania to assess the cytologic correlation of thyroid lesions shows accuracy of 81.4% with Negative predictive value (NPV) of 86% and Positive Predictive Value (PPV) of 75%.31 Similar study conducted in Uganda shows diagnostic accuracy, NPV & PPV of 85.9%,93.9% & 47% respectively [32]. Study done in USA shows diagnostic accuracy of FNAC to be 86.8% with NPV & PPV of 90% & 85.4% respectively [33].

With regard to the risk of malignancy in suspicious or indeterminate groups seven out of fifteen suspicious cases (46.7%) correlated with histopathologic diagnosis of thyroid carcinoma. Thus, there is a significant probability that a cytologically indeterminate lesion could be neoplastic, and there is a good chance that this lesion will be malignant. This is in agreement with study done in Tikur Anbessa hospital showing four out of seven suspicious cases (57%) turning out to be malignant and a study done in USA in which 19 out of 40 suspicious cases (47%) were indeed malignant [30,33].