[Systematic review about effectiveness of digital pathology as a diagnostic tool for pathology and cytopathology]

Carmona M, García-Carpintero E, Palma Ruiz, M
Record ID 32018005209
Spanish
Original Title: Revisión sistemática sobre la efectividad de la patología digital como herramienta de diagnóstico asistencial en anatomía patológica y citopatología
Authors' objectives: The aim of this report is to know the effectiveness of diagnostic tools based on digital systems of microscopic images as a routine practice in anatomic pathology departments. Secondary objectives are as follows: • To analyse the degrees of concordance and discordance of diagnoses made using DP compared to those made using conventional microscopy. • To find out the main characteristics of cases of major discordance in digital pathology diagnosis. • To analyse the workload in a department with digital pathology compared to a department without it. The research question is whether diagnosis in histopathology and cytopathology using DP systems is not inferior to diagnosis using conventional microscopy.
Authors' results and conclusions: RESULTS: The SR carried out produced a result of 1,544 references to studies, not counting duplicates. From these, a SR from 2020 was selected as the starting source, which allowed us to restrict the search to those studies sub-sequent to it (Azam 2020). When all screening criteria were applied, three more SRs and five individual studies were selected. The heterogeneous characteristics of the included studies and the type of outcome variable found in most of them, diagnostic concordance, did not allow us to perform a meta-analysis. However, as performed by Azam et al. in their SR, a weighted mean calculation has been carried out, taking into account the diagnostic concordance data and the size of the sample from which it was obtained. This calculation results in a mean concordance between DP and microscopy of 96.17% in a total of 53,535 samples (histology and cytology slides). The safety of these DP diagnoses is adequate as no major diagnostic discordances (implying differences in treatment) of clinical relevance were found in the total number of studies included. CONCLUSIONS: Digital pathology is an effective and safe diagnostic technique in clinical practice. • Concordance with conventional microscopic diagnosis exceeds 95%. • Major discordances observed are minimal. • The implementation of a DP system in an anatomical pathology department requires training and education of professionals, pathologists and cytotechnicians. • Its implementation requires the acquisition of quality equipment and software that complies with current EU regulations. • Any DP system must be integrated with health service information systems, including the electronic medical record. • An image storage system that is integrated with information systems and complies with data protection regulations is needed.
Authors' methods: A systematic review of the available scientific literature on the efficacy and safety of DP by complete digitisation of histological and cytological slides was carried out. The scientific literature on efficacy, safety and efficiency was searched in the following electronic databases: Medline, Embase, PROSPERO, DARE (Database of Abstracts of Reviews of Effects), Health Technology Assessment (HTA) Database and NHS-EED (National Health System Economic Evaluation Database) Centre for Reviews and Dissemination (CRD), INAHTA (The International Network of Agencies of Health Technology Assessment), CADTH (Canadian Agency for Drugs and Technologies in Health), AHRQ (Agency for Healthcare research and Quality), NICE (The National Institute for Health and Care excellence) and FDA (US Food and Drug Administration). We selected those studies with randomised clinical trial (RCT) or systematic reviews (SR) methodology that had results on the efficacy and safety of DP in pathology diagnosis in clinical practice. Summary tables were made of the results of all the included studies, containing concordance values, minor discordances and major discordances. Data on the efficiency of the diagnostic tool were added for those studies that included them. Where possible, an estimate of the weighted mean with its 95%CI was made for the diagnostic concordances with both interventions (conventional microscopic diagnosis and digital diagnosis).
Details
Project Status: Completed
Year Published: 2021
Requestor: Spanish Ministry of Health
English language abstract: An English language summary is available
Publication Type: Not Assigned
Country: Spain
MeSH Terms
  • Telepathology
  • Diagnosis
  • Digital Technology
  • Pathology, Clinical
  • Pathology
  • Workload
  • Artificial Intelligence
  • Image Processing, Computer-Assisted
  • Image Interpretation, Computer-Assisted
Keywords
  • Digital pathology
  • Digital cytopathology
  • Systematic review
  • Effectiveness
  • Security
Contact
Organisation Name: Agencia de Evaluacion de Tecnologias Sanitarias
Contact Address: Instituto de Salud "Carlos III", Calle Sinesio Delgado 6, Pabellon 4, 28029 Madrid, Spain. Tel: +34 9 1 822 2005; Fax: +34 9 1 387 7841;
Contact Name: Luis M. Sánchez Gómez
Contact Email: luism.sanchez@isciii.es
Copyright: <p>Agencia de Evaluacion</p>
This is a bibliographic record of a published health technology assessment from a member of INAHTA or other HTA producer. No evaluation of the quality of this assessment has been made for the HTA database.