Ruiz 2017.

Study characteristics
Patient Sampling	Retrospective, case‐control study including 131 eyes from 102 people during routine clinical practice at the Rothschild Foundation in the period between October 2015 and January 2016.
Patient characteristics and setting	Participants were grouped as follows. Normal Keratoconus Forme fruste and suspect keratoconus Postrefractive surgery for myopia All participants had a complete ophthalmic evaluation that consisted of manifest refraction, slit‐lamp examination, and corneal tomography evaluation by both Pentacam HR and Orbscan IIz. Exclusion criteria: a history of ocular surgery (apart from laser correction in the refractive surgery group), corneal cross‐linking, and very advanced keratoconus with corneal scarring. Rigid contact lens wearers were asked to remove their lenses at least 1 week before testing.
Index tests	The support vector machine uses 25 topography and tomography parameters of the anterior and posterior corneal surfaces from the Pentacam. The algorithm automatically classifies corneal patterns and shows the probability that the cornea has keratoconus, forme fruste or suspect keratoconus, photorefractive surgery, normal, or regular astigmatism.
Target condition and reference standard(s)	Participants were clinically diagnosed at the Rothschild Foundation based on slit‐lamp examination and corneal topography and classified into 4 groups. Topography and tomography were measured with the Pentacam HR and the Orbscan IIz. Participants were diagnosed before the corneal images were analysed with the algorithm.
Flow and timing	It was unclear whether all participants received the same reference standard. All data were included in a 2 × 2 table.
Comparative	Not applicable
Notes	Supported by a research grant of the Flemish government agency for Innovation by Science and Technology (grant nr. IWT/110684).
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?	Unclear
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?	Unclear
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?		Unclear risk
Are there concerns that the target condition as defined by the reference standard does not match the question?			Unclear
DOMAIN 4: Flow and timing
Did all patients receive the same reference standard?	Unclear
Were all patients included in the analysis?	Yes
Could the patient flow have introduced bias?		Unclear risk
DOMAIN 5: Comparative