Indications For Use Research Articles

Applications of artificial intelligence in retinopathy of prematurity

Retinopathy of prematurity (ROP) is a leading cause of childhood blindness worldwide, despite the fact that treatment is available, and blindness can be prevented in most cases. There are two key gaps in care delivery. The first is access to care with many parts of the world underserved. The second is subjective diagnosis where the eye exam is challenging, and different clinicians may perceive different levels of severity. Multiple potential indications for use (IFUs) have been proposed for ROP AI algorithms that could impact both gaps in care. In this presentation I will review potential IFUs and the potential translational challenges in implementation.

An FDA Guide on Indications for Use and Device Reporting of Artificial Intelligence-Enabled Devices: Significance for Pediatric Use

Radiology has been a pioneer in adopting artificial intelligence (AI)-enabled devices into the clinic. However, initial clinical experience has identified concerns of inconsistent device performance across different patient populations. Medical devices, including those using AI, are cleared by the FDA for their specific indications for use (IFUs). IFU describes the disease or condition the device will diagnose or treat, including a description of the intended patient population. Performance data evaluated during the premarket submission support the IFU and include the intended patient population. Understanding the IFUs of a given device is thus critical to ensuring that the device is used properly and performs as expected. When devices do not perform as expected or malfunction, medical device reporting is an important way to provide feedback about the device to the manufacturer, the FDA, and other users. This article describes the ways to retrieve the IFU and performance data information as well as the FDA medical device reporting systems for unexpected performance discrepancy. It is crucial that imaging professionals, including radiologists, know how to access and use these tools to improve the informed use of medical devices for patients of all ages.

“Off-IFU” Results of the Zenith Fenestrated Platform Are Comparable to Clinical Trial Outcomes

The Zenith fenestrated device (ZFEN; Cook Medical, Bloomington, Ind) has been proved safe and effective in the treatment of short-neck abdominal aortic aneurysms. The indications for use (IFU) anatomic criteria include a neck length >4 mm, neck diameter <31 mm, aortic neck angle <45°, and nonaneurysmal common iliac arteries (<21 mm). The present study compared the safety of “off-IFU” use of the ZFEN compared with the original USFT (U.S. fenestrated trial) data and identified risk factors for failure.

Endurant stent graft demonstrates promising outcomes in challenging abdominal aortic aneurysm anatomy

ObjectiveWe aimed to assess the 5-year safety and effectiveness outcomes of patients enrolled in the Endurant Stent Graft Natural Selection Global Post Market Registry (ENGAGE) who were treated outside the approved indications for use (IFU) of the Endurant stent graft. MethodsOur primary outcome measure was 12-month treatment success, defined as successful endograft delivery and deployment and the absence of type I or III endoleak, stent migration or limb occlusion, late conversion, and abdominal aortic aneurysm diameter increase or rupture. Secondary outcome measures included 30-day all-cause mortality, major adverse events, secondary procedures, technical observations, aneurysm-related mortality, and all-cause mortality within 12 months. ResultsDemographic characteristics of ENGAGE patients treated outside (225 [17.8%]) and within (1038 [82.2%]) the IFUs were similar, except that female patients comprised a much higher percentage of the outside IFU group (19.1% vs 8.7%; P < .001). The outside IFU group presented with lower rates of coronary artery disease and cardiac revascularization and a greater number of symptomatic patients compared with the within IFU group (21.3% vs 15.0%; P = .020). Technical success was achieved in more than 99% of all patients. The outside and within IFU groups showed a comparable and low occurrence of uncorrected type I (0.9% vs 1.2%; P = 1.00) and type III endoleak (0.4% vs 0.3%; P = .54) immediately after device implantation. The 5-year freedom from type IA endoleaks was 89.4% vs 96.7% (P < .0001) for those patients outside and within the IFUs, respectively, although both groups had similar type III endoleaks through 5 years (P = .61). Stent graft limb occlusion estimated overall survival, and freedom from aneurysm-related mortality and endovascular interventions were comparable in both patient groups through the 5-year follow-up. The Kaplan-Meier estimates at 5 years showed a trend for low but increased need for type I or III endoleak correction procedures in the outside IFU group compared with the within IFU group (7.2% vs 5.2%; P = .099). ConclusionsDifferences were not observed in all-cause mortality, aneurysm-related mortality, and secondary procedures between within and outside IFU patients through a 5-year follow-up in the ENGAGE registry. Proximal necks with angulation or diameters outside the IFUs were the most common reasons for patients identified as being outside IFU, and the cohort had increased incidence of type IA endoleaks. Despite the challenges presented from the broad range of aortic and abdominal aortic aneurysm morphologies, the Endurant stent graft showed promising 5-year outcomes.

VALIDITY, RELIABILITY, AND PSYCHOMETRIC PROPERTIES OF COGNIVUE®, A QUANTITATIVE ASSESSMENT OF COGNITIVE IMPAIRMENT

Introduction Background: Many tools for assessing decline in cognitive function have limited utility due to issues of accuracy, testing bias, and uptake among clinicians. Cognivue®* is a brief, easy-to-use, FDA-cleared tool for the adjunctive assessment of cognitive impairment risk. Objective: To clinically validate Cognivue® via agreement analysis of impairment risk classifications, retest reliability assessment, and psychometric property comparison. Methods Adults age 55-95 y at risk for age-related cognitive decline or dementia completed Cognivue®, St. Louis University Mental Status (SLUMS), and other neuropsychological tests including Rey Auditory Verbal Learning Test (RAVLT) & Trail Making Test A/B (TMT-A, TMT-B). Analyses included: regression analyses for agreement and re-test reliability, and rank linear regression and factor analysis for psychometric comparisons. Results Data were available for 401 subjects who completed ≥1 testing session, and 358 who completed 2 sessions 1-2 weeks apart. Previously determined Cognivue® classification scores were validated, demonstrating good agreement with SLUMS scores (weighted κ 0.57; 95% CI 0.50-0.63). The study of test-retest reliability showed similar scores across repeated testing for Cognivue® (regression fit, R2, 0.81; r, 0.90), and SLUMS (regression fit, R2, 0.67; r, 0.82). The Cognivue® risk classifications of high, low to moderate, and no risk of impairment, did not differ significantly across repeat testing; however, for SLUMS, the relationship between scores and classifications across repeated testing was less robust. The psychometric validity of the Cognivue® cognitive test battery was demonstrated compared to traditional paper & pencil neuropsychological tests. Scores were most closely correlated with measures of verbal processing, manual dexterity/speed, visual contrast sensitivity, visuospatial/executive function, and speed/sequencing. Conclusions The Cognivue® validation study demonstrated good agreement between Cognivue® and the SLUMS test; good test-retest reliability of Cognivue® test results; and validated the psychometric properties of the Cognivue® test battery compared to traditional neuropsychological tests. This research was funded by This study was supported by Cognivue, Inc. *INDICATIONS FOR USE: Cognivue® testing is indicated as an adjunctive tool for evaluating perceptual and memory function in individuals aged 55-95 y. It is not intended to be used as a stand-alone device to identify the presence or absence of clinical diagnoses. Cognivue® is intended to be used by medical professionals qualified to interpret the results of a cognitive assessment examination.

COMPARISON OF THE COGNIVUE® QUANTITATIVE ASSESSMENT TOOL AND SLUMS TO CLASSIFY THE RISK OF COGNITIVE IMPAIRMENT

Introduction Background: Cognivue®* was developed based on clinical experience and NIH funded laboratory research into the neural mechanisms of functional impairment in aging and dementia. The computerized testing tool provides an automated brain functional assessment tool not tied to traditional question & answer testing. Cognivue® consists of 3 sub-batteries of 10 separately scored sub-tests presented in a 10 minute automated sequence. These include basic motor & visual ability (visuo-motor and visual salience), perceptual processing (letter, word, shape, and motion discrimination), and memory processing (letter, word, shape, and motion memory). Objective: To determine the Cognivue® cut-off scores corresponding to the St. Louis University Mental Status (SLUMS) 3-level classification for risk of cognitive impairment. Methods Adults (age 55-95 y) at-risk for age-related cognitive decline or dementia were invited via posters and email to complete the SLUMS and Cognivue® tests. Optimization analyses by positive percent agreement (PPA) and negative percent agreement (NPA), as well as by accuracy and error bias were conducted. Results 92 subjects, at 5 sites, completed SLUMS (reference standard) and Cognivue® tests. Based on SLUMS score, 50% were not impaired (>26), 38% were intermediate (26‒21), and 12% were impaired ( 26 (no impairment) corresponded to a Cognivue® score of 78.5 (NPA=0.5; PPA=0.79). Based on the 2 separate analysis techniques, results showed that Cognivue® scores between 55‒64 corresponded to SLUMS scores for impairment, and Cognivue® scores between 74‒79 corresponded to SLUMS scores for no impairment. Conclusions Cognivue® scores ≤50 provide a conservative standard for high risk of impairment that will avoid misclassification of an individual as impaired. Cognivue® scores ≥75 provide a conservative cut-off for no risk of impairment that will avoid misclassification of an individual as not unimpaired. This research was funded by This study was supported by Cognivue, Inc. *INDICATIONS FOR USE: Cognivue® testing is indicated as an adjunctive tool for evaluating perceptual and memory function in individuals aged 55-95 y. It is not intended to be used as a stand-alone device to identify the presence or absence of clinical diagnoses. Cognivue® is intended to be used by medical professionals qualified to interpret the results of a cognitive assessment examination.

МАЛОИНВАЗИВНЫЙ НАКОСТНЫЙ ОСТЕОСИНТЕЗ ДИСТАЛЬНОГО МЕТАЭПИФИЗА ЛУЧЕВОЙ КОСТИ: ПОКАЗАНИЯ К ПРИМЕНЕНИЮ И ОСОБЕННОСТИ МЕТОДИКИ

МАЛОИНВАЗИВНЫЙ НАКОСТНЫЙ ОСТЕОСИНТЕЗ ДИСТАЛЬНОГО МЕТАЭПИФИЗА ЛУЧЕВОЙ КОСТИ: ПОКАЗАНИЯ К ПРИМЕНЕНИЮ И ОСОБЕННОСТИ МЕТОДИКИ

Factors that affect cost and clinical outcome of endovascular aortic repair for abdominal aortic aneurysm

ObjectiveThis study evaluated the effect of indication for use (IFU), additional graft components, and percutaneous closure of endovascular aortic repair (PEVAR) on clinical outcomes and cost of endovascular aortic repair (EVAR). MethodsClinical and financial data were obtained for all elective EVARs completed at a university-affiliated medical center between January 2012 and June 2013. Data were analyzed by χ2, Student t-test for independent samples, and Kaplan-Meier survival. ResultsThere were 67 elective EVARs. Additional cuffs/extensions were used in 37%, increasing the baseline graft cost by 36% (P < .001), total costs by 20% (P < .001), and negatively affecting the contribution margin. Aortic neck IFU (P = .02), failure of the index graft to seal the neck (P = .02), and need for an additional cuff (P = .008) were related to the need for reintervention for type Ia endoleak for graft B (Excluder; W. L. Gore and Associates, Flagstaff, Ariz), whereas limb IFU was related to the need for additional limb extension for graft A (Powerlink; Endologix, Irvine, Calif; P < .001). Limb extension (P = .06) and failure of the index graft to provide an adequate seal (P < .001) were associated with reintervention for type Ib endoleak. Reintervention-free rates at 24 months were 96% for graft A and 94% for graft B (P =.54), but different patterns in reintervention emerged: graft A required reoperation early (<2 months) then stabilized; graft B did not require reintervention until 24 months, but rates increased substantially by 25 months. PEVAR was attempted in 61 (91%): 49 (73%) bilaterally, 7 (10%) unilaterally, and 5 (8%) failed. The mean number of closure devices was four (range, 1-9): $1000 (3.5% of total cost). Bilateral PEVAR was associated with shorter operating time than unilateral PEVAR/failed PEVAR (P < .001) and lower operating room use costs (P = .005) and total hospital costs (P = .003) than failed PEVAR. The contribution margin was higher for bilateral PEVAR than unilateral PEVAR/failed PEVAR (P = .005). Patients with bilateral PEVAR and unilateral PEVAR were more often discharged on postoperative day 1 than those with failed PEVAR (P = .002). Hospital length of stay (P = .49), operating room duration (P = .31), and total costs (P = .72) were similar for unsuccessful PEVAR and EVAR completed with cutdown. ConclusionsHigher rates of reintervention occurred when EVAR was performed outside of IFU guidelines or when additional components were needed. Additions raised graft costs significantly above baseline. Notable differences in graft performance in complex anatomy and varied patterns of reoperation could be useful in the graft selection process to improve outcome and contain costs. Bilateral PEVAR was associated with lower costs and postoperative day 1 discharge. Attempting PEVAR may be reasonable unless there is serious concern for failure.

When Not to Implant the Multilayer Flow Modulator: Lessons Learned From Application Outside the Indications for Use in Patients With Thoracoabdominal Pathologies

To scrutinize registry data on thoracoabdominal repairs performed using the Multilayer Flow Modulator (MFM) outside the indications for use (IFU) and analyze the adverse outcomes. Of 380 patients from Europe registered in the MFM Global Registry after treatment for thoracoabdominal aortic aneurysm (TAAA) or dissection, 38 (10%) patients (30 men; median age 71 years, range 30-91) treated on a compassionate basis outside the IFU were analyzed. Thirteen patients had chronic Stanford type B dissection with aneurysmal dilatation >6 cm. There were 6 mycotic and 4 saccular aneurysms in addition to 15 primary TAAAs. The mean aneurysm diameter was 7.1 cm. Ten patients presented with rupture, and 23 patients had previous open or thoracic endovascular aortic repair (TEVAR). Although no death, paraplegia, stroke, or renovisceral compromise was documented during the initial hospital stay, technical success was zero. There were 31 (81.6%) cases in which there was failure to land the device in normal aorta. Other violations of the IFU included 12 with inadequate stent overlap and 11 cases involving a small MFM being deployed inside a larger one. Five of the 9 cases in which an undersized device was used resulted in a type I endoleak (failure mode I). During a mean follow-up of 10.0±6.9 months, all-cause mortality was 89.5% (34/38), of which 27 (71.1%) were aneurysm-related deaths. Overall survival, freedom from aneurysm-related death, and rupture-free survival estimates were 17.5%, 25.0%, 31.5%, respectively, at 18 months. There were 8 visceral branch complications; in all, 14 secondary endovascular interventions were required in 11 patients for endoleak (failure modes I and II) or stent foreshortening. No false lumen was completely thrombosed in the dissecting aneurysms. All aneurysms showed a mean sac growth rate of 0.12±0.16 cm/month. Factors having a significance influence on risk of aneurysm-related death included maximum aneurysm diameter (p=0.025, HR 1.37, 95% CI 1.04 to1.82), previous TEVAR (p=0.03, HR 2.44, 95% CI 1.10 to 2.08), and inadequate overlap between MFM devices (p<0.002, HR 4.02, 95% CI 1.70 to 9.49). There are clinical scenarios in which the MFM does not perform well. The MFM is not a solution for patients living on borrowed time and should not be used indiscriminately in patients in whom other modalities of aortic repair are not feasible. Its use must adhere to the IFU, and robust clinical data are required before constructing a randomized controlled trial.

Commentary: Multilayer Stent for the Treatment of Complex Aortic Pathologies: A Long or Short Road Ahead?

‘‘Hic Sunt Leones’’ (here there are lions) was written on uncharted territories of old Roman maps; this notice alerted people leaving for a journey about the unknown and dangerous paths ahead. In this issue of the JEVT, two articles illustrated well that similar dangers awaited overenthusiastic interventionists who took the unexplored route to indiscriminate use of the Multilayer Flow Modulator (MFM) for compassionate or selected cases. The results were an assortment of data reporting negative and positive anecdotal experiences without any statistical power. Moreover, the early negative results alarmed the medical communities, which jeopardized their interest in this clinically untested but fascinating and innovative technology. The review of failed MFM cases by Sultan and colleagues and the sound clinical data from the structured STRATO trial contribute to our understanding of the indications for use (IFU) of the MFM. First off, let’s look at the question that is on everyone’s mind: do these endoprostheses live up to expectations in complex aortic pathologies? The answer: not completely. Treatment of thoracic and thoracoabdominal aortic aneurysms is attended by relatively high risk whether the procedure be surgical or endovascular, and the avoidance of rupture is at the core of these therapies. Chimney and fenestrated techniques (ChEVAR, F-EVAR) help to reduce mortality and morbidity, but they have precise IFUs and require well developed catheter skills. In ChEVAR, the maximum number of target vessels that can be safely stented has never been systematically studied, but two chimneys seems to be the number for a safe procedure. F-EVAR reports describe a 15% rate of secondary interventions within the first year. In this group, reintervention for endoleak accounted for 48% of cases; in the remaining 52%, a significant number of new surgical procedures were needed for mesenteric ischemia. Recently, Manning et al. reported their experience in F-EVAR using triple fenestration stent-grafting, underlining that, at present, these procedures are technically more demanding and associated with higher risks compared with double or single fenestrations. New technologies and increasing understanding of the pathophysiological mechanisms of vascular disease complement each other. A disease manifesting as an anatomical abnormality results from an interaction of molecular, biological, and biochemical processes. Novel, and somewhat unconventional, concepts of treating vascular pathologies rely on modulation of such processes through