[Accuracy of the cytopathological telediagnosis of palpable limphoadenopathies. A study design simulating the diagnostic support from remote access to health workers in the developing countries]

Recenti Prog Med. 2005 May;96(5):226-30.
[Article in Italian]

Abstract

It has been estimated the accuracy of the cytologic diagnosis, carried out by a cytopathologist, on digital images (DID) of fine needle aspiration lymphonode biopsies (FNAB), in order to validate the telematic support to health workers practising in specialist-lacking remote areas of the developing countries. A blind-cross check simulation design was drawn-up. A non-pathologist practictioner selected the microscopic fields of 39 cases of FNAB, took the pictures with a digital camera, stored them and submitted to a cytopathologist for DID. The DID were compared against the original diagnosis in direct view (DVC). The histologic diagnosis on the surgical biopsies were the golden standard to refer to. The results demonstrate that: 1) DVC and DID have similar and satisfactory sensitivity rates (88.2 e 85.3%) and positive predictive values (96.8 e 90.1%); 2) DID have lower specificity and efficiency rates against DVC (40.0 vs 80.0 and 79.5 vs 79.2%); 3) the negative predictive values are low both in DVC and DID (50.0 e 28.6%), but the data ought to be confirmed in a well-balanced ratio malignancy-benignancy; d) the accuracy of DID could be increased by improving the selection process of the microscopic fields; e) the remote access-assisted cytologic diagnosis of FNAB can be a procedure to carry out easily and cheaply in areas where the histopathology facilities are unavailable. The accuracy is the highest in metastasis, in high-grade non-Hodgkin lymphoma and in lymphocite depletion and mixed cell sub-groups of Hodgkin lymphoma.

Publication types

  • English Abstract

MeSH terms

  • Developing Countries
  • Health Personnel
  • Humans
  • Lymphatic Diseases / pathology*
  • Palpation
  • Reproducibility of Results
  • Surveys and Questionnaires
  • Telepathology*