Common mistakes in sampling: Optimal cytology smears, correct sample handling and contextualisation of the findings with the clinical history are all essential steps for achieving an accurate cytological diagnosis. Good quality cytology smears provide excellent morphologic details of cells and infectious agents, often allowing to differentiate between inflammatory and neoplastic processes, identify the tumour type and its behaviour (benign or malignant). When performing a fine-needle aspirate (FNA), the aim is to produce a monolayer of cells with minimal cell rupture. An incorrect technique can produce unsuitable samples precluding adequate evaluation and identification of the cells. Another common pre-analytical mistake in cytology is to collect a single aspirate, especially from larger masses. A single mass may contain areas of necrosis, inflammation, neoplasia or normal tissue cells and a single slide is unlikely to be representative of the entire lesion. If a mass is fluid-filled, collection of fluid and adjacent solid areas would be recommended, as fluid cytology alone rarely reveals the nature of the surrounding mass. Labelling of the slides with patient name and origin of the FNAs is another crucial step. The importance of sample handling before processing and staining should not be underestimated. Fluid samples should be collected in the correct tubes and adequately stored; unstained cytology slides should not be exposed to formalin fumes. Finally, adequate staining procedures are essential to guarantee and highlight the cellular details that are required for the diagnosis. Taking care of all these simple steps will prevent the most common sampling mistakes, increasing the diagnostic power of cytology.

Common mistakes in interpreting: When interpreting cytology, it is vital to consider the clinical history and appearance of the lesion as well as the cytological appearance and to have likely differential diagnoses in mind. Organisms may not be visible in infected lesions if antibiotics are given before sampling. Bacteria are rarely seen in septic arthritis. Fungi and mycobacteria can be difficult to see with routine stains. The lesion may have mixed pathology such as focal areas of necrosis or inflammation within a tumour and sometimes the fine needle aspirate may harvest only some of these components and not be wholly representative. Hence if neoplasia is suspected but only inflammation is seen, resampling different areas would be recommended. We are familiar with looking for criteria of malignancy to make a diagnosis of neoplasia. However, hyperplastic or dysplastic cells can sometimes be impossible to distinguish from neoplastic cells, since all three can show criteria of malignancy. This is a particular problem of mesenchymal cells because the fibroblasts in granulation tissue or in inflammatory lesions can resemble the neoplastic cells seen in soft tissue sarcomas. The history and appearance may help distinguish these although biopsy will often be required. Just as non-neoplastic cells can look malignant, the converse is also true. Some malignant tumours consist of cells which do not display marked criteria of malignancy. Examples include haemophagocytic histiocytic sarcoma, some malignant melanomas and thyroid carcinoma. Knowledge of the clinical presentation and expected pathology will help minimise errors in interpretation. Cytology should never be performed in isolation.