Referral criteria trumps screening technology for amblyopia

July 1, 2015

Physicians should focus on referral criteria in screening technology rather than the instrument itself.

 

Take-Home Message: Physicians should focus on referral criteria in screening technology rather than the instrument itself.

 

 

By Lynda Charters; Reviewed by Sean P. Donahue, MD, PhD

Nashville-New screening technologies have revolutionized the early detection of amblyopia, by allowing children to be screened long before they are able to cooperate with reading an eye chart.

CTR doesn't avert spontaneous IOL dislocation

These instruments-photoscreeners and autorefractors-work by detecting risk factors for amblyopia (anisometropia, high bilateral hyperopia and astigmatism, and strabismus) rather than amblyopia itself.

Another advantage is that the instrument software can be customized based on the age of the patient or the desired sensitivity or specificity, by altering the referral criteria the package uses.

The most important factor, according to Sean P. Donahue, MD, PhD, is not what instrument should be purchased but the referral criteria put into the instrument in order to appropriately refer children. Dr. Donahue is professor of ophthalmology, neurology, and pediatrics, Vanderbilt University Medical Center, and chief, Pediatric Ophthalmology, Vanderbilt Children’s Hospital, Nashville.

Four types of refractive screening instruments are available:

  • Off-axis photoscreeners,

  • Automated refractors,

  • Retinal birefringence, and

  • Visual-evoked potentials.

Of these, the first two devices are the most commercially available.

Off-axis photoscreening

MTI (Medical Technologies & Innovations, Inc.), the oldest and the primary example of these instruments, was based on an analogue picture. The fundamental concept was a linear flash away from the visual axis. By evaluating how the flash returned from the retina, clinicians were able to interpret the possible refractive error and strabismus because the flash was out of focus in a nonemmetropic eye, Dr. Donahue explained.

Next: Autorefractors

 

Having used this instrument for 20 years on millions of patients in Tennessee and worldwide, he pointed out that most children can be screened successfully.

4 reasons you need to be reading your online reviews-and why you should care

“Depending on the referral criteria, very high predictive values can be obtained,” he said. “The Iowa Group has predictive values over 85% in detecting AAPOS [American Association for Pediatric Ophthalmology and Strabismus] vision screening criteria for amblyopia risk factors.”

Disadvantages of the technology are its age, analogue versus digital images, and delayed interpretation because of the absence of automated reading.

The next off-axis photoscreener to become available was the iScreen (iScreen Vision) with advances, such as a digital camera in a single-flash unit. The image is transmitted by computer to a reading center, and the results are returned the following day. The data from this instrument is good, according to Dr. Donahue.

Autorefractors

These instruments were the next step in screening technology. Advantages include a target for the child to look at and ultrasound measurement of the refractive error. The Nikon Retinomax and the Welch Allyn SureSight became available in the late 1990s, and the latter received the most attention with its hard-copy printout and two sets of referral criteria, one from the manufacturer and the other criteria proposed by the Vision in Preschoolers (VIP) Program, which has a higher specificity and slightly less sensitivity, Dr. Donahue noted.

Despite these advantages, the SureSight instrument had a high referral rate of 7.3%, but a predictive value positive of only 48% in his studies. Adjusting the referral criteria to decrease the referral rate jeopardized the sensitivity, Dr. Donahue noted.

When ordering this instrument, physicians must choose between the manufacturer or the VIP criteria.

Top-paying states for ophthalmic techs

“With the manufacturer’s criteria, about 15% of patients will be referred with a predictive value positive of 10%, meaning that only one of 10 children referred is going to have a visual problem,” he said. “The VIP criteria have 94% specificity and a predictive value of about 45%, which is better for the vast majority of preschool vision screening applications.”

PlusOptix Mobile Vision Screener (PlusOptix GmbH), the next autorefractor, has the advantage of being able to alter the referral criteria. The S12 is the most currently available iteration of the technology. The characteristic that Dr. Donahue likes about this device is the availability of five sets of referral criteria, which the user can choose.

Next: Screening considerations

 

The instrument also has high sensitivity (85%) and specificity (90%), and a lower referral rate with a high predictive value and a higher referral rate with a lower predictive value.

Pushing the envelope for the sake of ophthalmology

“All of these have published data to support them,” Dr. Donahue said. “The choice is based on preference. Do you want to refer a lot of children and not miss anything, or do are you in situation where the children are in a remote location and must come to a center and a high number of false positives is not desired?”

A third autorefractor (PediaVision Spot, Welch Allyn) is a handheld wireless device with adjustable referral criteria that has higher sensitivity than the PlusOptix instrument but poorer specificity of 75% when the manufacturer’s criteria are used. Dr. Donahue has been working with the manufacturer to collect data in an attempt to improve the specificity and decrease the referral rate.

Gobiquity has introduced a device that uses a smartphone with an application called GoCheck Kids. There currently are no published data on this instrument.

However, as Dr. Donahue pointed out, “If this is proven to work, it is not an $8,000 capital investment as the other instruments are.”

Several other instruments are under development but are not yet commercially available.

Screening considerations

Patient age has always been a factor in screening.

“Age, however, has less of an impact on results than previously thought based on the Amblyopia Treatment Studies,” Dr. Donahue said.

For anisometropic amblyopia, the prevalence of amblyopia seems to be fairly constant based on the magnitude of the difference between the eyes. However, with time, the depth of the amblyopia increases, he noted.

Dr. Donahue advised identifying the amblyopia early and at the latest, by 5 years of age. However, because some amblyopia resolves with spectacle treatment alone, it may be important to treat the 1- and 2-year-old patients, which underscores the importance of early screening.

Next: USPSTF statement

 

He identified three elements that ophthalmologists should know:

  • The U.S. Preventive Services Task Force (USPSTF);

  • CPT code 99173 for traditional screening, and

  • CPT code 99174 for photoscreening and photorefraction.

The USPSTF statement, published in 2011, advises that photoscreening is useful in children aged 3 to 5 years.

“This is the first time that a major organization has stated this,” Dr. Donahue said.

More clinical diagnosis news

The USPSTF also said there is insufficient evidence to recommend photoscreening for patients less than 3 years of age. Despite that level of recommendation, Donahue believes the data that support the effectiveness of photoscreening in that age range “is overwhelming.”

Another important consideration for ophthalmologists is from The American Academy of Pediatrics Instrument-Based Pediatric Vision Screening Policy Statement, which states that photoscreening is no longer experimental and should be reimbursed by payers.

“Photoscreening is an alternative to traditional screening for younger children,” Dr. Donahue said. “It is not yet mandated because of the cost.”

Considering which instrument is the best to purchase is not the best route.

“The consideration should be: When I buy this instrument, what criteria should be put in to maximize what I want to detect-i.e., sensitivity or specificity? The choice does not depend on the instrument, but on the referral criteria,” Dr. Donahue summarized.

 

 

Sean P. Donahue, MD, PhD

E: sean.donahue@vanderbilt.edu

This article is adapted from Dr. Donahue’s presentation during Pediatric Subspecialty Day at the 2014 meeting of the American Academy of Ophthalmology. Dr. Donahue has been a consultant for all of the above mentioned vision screening manufacturers, as well as for competing products; is the lead author of the AAPOS Vision Screening Committee Recommendations regarding the detection of Amblyopia Risk Factors using automated screening; and is the lead author of the upcoming AAPOS-AAO-AAP-AACO combined Policy Statement about Vision Screening in Children.