Issarti 2020.

Study characteristics
Patient Sampling	Retrospective, multicentre, case‐control study. Scheimpflug Pentacam measurements (Oculus GmbH, Wetzlar, Germany) of 812 eyes were retrospectively collected from 2 centres in Belgium.
Patient characteristics and setting	Normal eyes: eyes without slit‐lamp findings suggestive of corneal ectasia and normal tomography Keratoconic eyes were divided into the following 4 groups. Suspected keratoconus Early keratoconus Mild keratoconus Moderate to advanced keratoconus
Index tests	A score‐based machine learning system based on feedforward neural network (Logistic Index for Keratoconus, Logik), capable of classifying keratoconus according to its severity, to objectively discriminate suspect keratoconus from healthy eyes, and to provide a consistent, time‐continuous scoring system for keratoconus progression
Target condition and reference standard(s)	Normal: eyes without slit‐lamp findings suggestive of corneal ectasia and normal tomography Suspect keratoconus (forme fruste keratoconus): the contralateral, asymptomatic eye of a person with clinical keratoconus in 1 eye without clinical or tomographic signs of ectasia Early keratoconus: eyes with a subtle sign of keratoconus, such as localized steeping in anterior or posterior surface, without significant changes of the cornea in the slit lamp Mild keratoconus: tomographic changes consistent with keratoconus (anterior or posterior corneal steepening, corneal thinning, stromal thinning), Fleischer's rings at the cone base, partial or circular Fleischer's rings, but no visible Vogt's striae Moderate to advanced keratoconus: clear cornea, corneal thinning at the apex, visible Vogt's striae, clearly visible circular Fleischer's ring, and corneal tomography findings An ophthalmologist and an optometrist performed the classification. Cases were classified before inclusion.
Flow and timing	All cases were included in the reference standard and index test. All data were included in a 2 × 2 table.
Comparative	Not applicable
Notes	This work was supported in part by a research grant from the Flemish government agency for Innovation by Science and Technology (grant no. TBM‐T000416N).
Methodological quality
Item	Authors' judgement	Risk of bias	Applicability concerns
DOMAIN 1: Patient selection
Was a consecutive or random sample of patients enrolled?	No
Was a case‐control design avoided?	No
Did the study avoid inappropriate exclusions?	No
Could the selection of patients have introduced bias?		High risk
Are there concerns that the included patients and setting do not match the review question?			High
DOMAIN 2: Index test (All tests)
Were the index test results interpreted without knowledge of the results of the reference standard?	Yes
If a threshold was used, was it pre‐specified?	No
Was the model designed in an appropriate manner?	Yes
Could the conduct or interpretation of the index test have introduced bias?		Low risk
Are there concerns that the index test, its conduct, or interpretation differ from the review question?			Low concern
DOMAIN 3: Reference standard
Is the reference standard likely to correctly classify the target condition?	Yes
Were the reference standard results interpreted without knowledge of the results of the index tests?	Yes
Could the reference standard, its conduct, or its interpretation have introduced bias?		Low risk
Are there concerns that the target condition as defined by the reference standard does not match the question?			Low concern
DOMAIN 4: Flow and timing
Did all patients receive the same reference standard?	Yes
Were all patients included in the analysis?	Yes
Could the patient flow have introduced bias?		Low risk
DOMAIN 5: Comparative