Diagnosis of idiopathic pulmonary fibrosis by virtual means using "IPFdatabase"- a new software.
Aged
Algorithms
Female
Humans
Idiopathic Pulmonary Fibrosis
/ diagnostic imaging
Lung Diseases, Interstitial
/ diagnostic imaging
Male
Middle Aged
Practice Guidelines as Topic
Predictive Value of Tests
Sensitivity and Specificity
Software
/ standards
Tomography, X-Ray Computed
/ instrumentation
User-Computer Interface
Database
Diagnosis
IPF
Software
Journal
Respiratory medicine
ISSN: 1532-3064
Titre abrégé: Respir Med
Pays: England
ID NLM: 8908438
Informations de publication
Date de publication:
02 2019
02 2019
Historique:
received:
16
08
2018
revised:
24
12
2018
accepted:
26
12
2018
entrez:
2
2
2019
pubmed:
2
2
2019
medline:
24
3
2020
Statut:
ppublish
Résumé
The diagnostic algorithm for idiopathic pulmonary fibrosis (IPF) guidelines has some shortcomings. The aim of the present study was to develop a novel software, "IPFdatabase", that could readily apply the diagnostic criteria per IPF guidelines and make a 'virtual' diagnosis of IPF. Software was developed as a step-by-step compilation of necessary information according to guidelines to enable a diagnosis of IPF. Software accuracy was validated primarily by comparing software diagnoses to those previously made at a Center for Interstitial Lung Diseases. Clinical validation on 98 patients (68 male, age 61.0 ± 8.5 years), revealed high software accuracy for IPF diagnosis when compared to historical diagnoses (sensitivity 95.5%, specificity 96.2%; positive predictive value 95.5%, negative predictive value 96.2%). A general radiologist and a general pathologist reviewed relevant data with and without the new software: interobserver agreement increased when they used the IPFdatabase (kappa 0.18 to 0.64 for radiology, 0.13 to 0.59 for pathology). IPFdatabase is a useful diagnostic tool for typical cases of IPF, and potentially restricts the need for MDDs to atypical and complex cases. We propose this web-designed software for instant accurate diagnosis of IPF by virtual means and for educational purposes; the software is readily accessed with mobile apps, allows incorporation of updated version of guidelines, can be utilized for gathering data useful for future studies and give physicians rapid feedback in daily practice.
Sections du résumé
BACKGROUND
The diagnostic algorithm for idiopathic pulmonary fibrosis (IPF) guidelines has some shortcomings. The aim of the present study was to develop a novel software, "IPFdatabase", that could readily apply the diagnostic criteria per IPF guidelines and make a 'virtual' diagnosis of IPF.
METHODS
Software was developed as a step-by-step compilation of necessary information according to guidelines to enable a diagnosis of IPF. Software accuracy was validated primarily by comparing software diagnoses to those previously made at a Center for Interstitial Lung Diseases.
RESULTS
Clinical validation on 98 patients (68 male, age 61.0 ± 8.5 years), revealed high software accuracy for IPF diagnosis when compared to historical diagnoses (sensitivity 95.5%, specificity 96.2%; positive predictive value 95.5%, negative predictive value 96.2%). A general radiologist and a general pathologist reviewed relevant data with and without the new software: interobserver agreement increased when they used the IPFdatabase (kappa 0.18 to 0.64 for radiology, 0.13 to 0.59 for pathology).
CONCLUSION
IPFdatabase is a useful diagnostic tool for typical cases of IPF, and potentially restricts the need for MDDs to atypical and complex cases. We propose this web-designed software for instant accurate diagnosis of IPF by virtual means and for educational purposes; the software is readily accessed with mobile apps, allows incorporation of updated version of guidelines, can be utilized for gathering data useful for future studies and give physicians rapid feedback in daily practice.
Identifiants
pubmed: 30704696
pii: S0954-6111(18)30400-1
doi: 10.1016/j.rmed.2018.12.011
pii:
doi:
Types de publication
Comparative Study
Journal Article
Langues
eng
Sous-ensembles de citation
IM
Pagination
31-36Informations de copyright
Copyright © 2019 Elsevier Ltd. All rights reserved.