Unlocking the Value: Quantifying the Return on Investment of Hospital Artificial Intelligence.

PURPOSE: A comprehensive return on investment (ROI) calculator was developed to evaluate the monetary and nonmonetary benefits of an artificial intelligence (AI)-powered radiology diagnostic imaging platform to inform decision makers interested in adopting AI. METHODS: A calculator was constructed to calculate comparative costs, estimated revenues, and quantify the clinical value of using an AI platform compared with no use of AI in radiology workflows of a US hospital over a 5-year time horizon. Parameters were determined on the basis of expert interviews and a literature review. Scenario and deterministic sensitivity analyses were conducted to evaluate calculator drivers. RESULTS: In the calculator, the introduction of an AI platform into the hospital radiology workflow resulted in labor time reductions and delivery of an ROI of 451% over a 5-year period. The ROI was increased to 791% when radiologist time savings were considered. Time savings for radiologists included more than 15 8-hour working days of waiting time, 78 days in triage time, 10 days in reading time, and 41 days in reporting time. Using the platform also provided revenue benefits for the hospital in bringing in patients for clinically beneficial follow-up scans, hospitalizations, and treatment procedures. Results were sensitive to the time horizon, health center setting, and number of scans performed. Among those, the most influential outcome was the number of additional necessary treatments performed because of AI identification of patients. CONCLUSIONS: The authors demonstrate a substantial 5-year ROI of implementing an AI platform in a stroke management-accredited hospital. The ROI calculator may be useful for decision makers evaluating AI-powered radiology platforms.

Economic evaluation of supplemental breast cancer screening modalities to mammography or digital breast tomosynthesis in women with heterogeneously and extremely dense breasts and average or intermediate breast cancer risk in US healthcare.

OBJECTIVE: To evaluate the cost-effectiveness of supplemental breast imaging modalities for women with heterogeneously and extremely dense breasts and average or intermediate risk of breast cancer (BC) in the USA, and analyze capacity requirements for supplemental magnetic resonance imaging (MRI) and contrast-enhanced mammography (CEM). METHODS: Clinical and economic outcomes for supplemental imaging modalities including full- and abbreviated-protocol MRI (Fp-MRI, Ab-MRI), CEM, and ultrasound (U/S) as add-on to x-ray mammography (XM) or digital breast tomosynthesis (DBT), were compared to XM or DBT alone, in a decision tree linked to a Markov chain validated by comparison with a microsimulation analysis. A Delphi panel supplemented model input parameters from the literature. A capacity model evaluated the number of additional daily scans and scanners required for Fp-MRI and CEM. RESULTS: Compared to XM or DBT alone, all supplemental imaging protocols were cost-effective. Both Fp- and Ab-MRI, and to a lesser extent CEM and U/S, yielded superior clinical outcomes to XM or DBT. Compared to XM alone, U/S and Ab-MRI had the lowest incremental cost-effectiveness ratios (ICER). For U/S, the ICER was $23,394 for the average-risk population and $13,241 for the intermediate-risk population. For CEM, the ICER was $38,423 and $23,772, respectively. For the extremely dense subpopulation with intermediate risk, supplemental screening requirements could be accommodated by conducting one Fp-MRI scan per day per existing general scanner. CONCLUSIONS: While ultrasound had the lowest ICER, MRI and CEM demonstrated the best clinical outcomes, compared to XM or DBT alone for women with dense breasts and intermediate and high risk. Existing MRI scanner capacity has the potential to meet most of the supplemental screening needs of this population.

Performance of Supplemental Imaging Modalities for Breast Cancer in Women With Dense Breasts: Findings From an Umbrella Review and Primary Studies Analysis.

Breast cancer screening performance of supplemental imaging modalities by breast density and breast cancer risk has not been widely studied, and the optimal choice of modality for women with dense breasts remains unclear in clinical practice and guidelines. This systematic review aimed to assess breast cancer screening performance of supplemental imaging modalities for women with dense breasts, by breast cancer risk. Systematic reviews (SRs) in 2000 to 2021, and primary studies in 2019 to 2021, on outcomes of supplemental screening modalities (digital breast tomography [DBT], MRI (full/abbreviated protocol), contrast enhanced mammography (CEM), ultrasound (hand-held [HHUS]/automated [ABUS]) in women with dense breasts (BI-RADS C&D) were identified. None of the SRs analyzed outcomes by cancer risk. Meta-analysis of the primary studies was not feasible due to lack of studies (MRI, CEM, DBT) or methodological heterogeneity (ultrasound); therefore, findings were summarized narratively. For average risk, a single MRI trial reported a superior screening performance (higher cancer detection rate [CDR] and lower interval cancer rate [ICR]) compared to HHUS, ABUS and DBT. For intermediate risk, ultrasound was the only modality assessed, but accuracy estimates ranged widely. For mixed risk, a single CEM study reported the highest CDR, but included a high proportion of women with intermediate risk. This systematic review does not allow a complete comparison of supplemental screening modalities for dense breast populations by breast cancer risk. However, the data suggest that MRI and CEM might generally offer superior screening performance versus other modalities. Further studies of screening modalities are urgently required.

Evaluation of health economic impact of initial diagnostic modality selection for colorectal cancer liver metastases in suspected patients in China, Japan and the USA.

AIMS: To compare cost offsets and contributing factors (false-negative rates and confirmatory imaging requirements, potentially leading to longer waiting times for diagnosis) as well as long-term cost effectiveness associated with the diagnostic and treatment pathways for colorectal cancer liver metastases (CRCLM) in the US, Japan, and China according to initial imaging modality used. Gadoxetate disodium (ethoxylbenzyl-diethylenetriaminepentaacetic acid)-enhanced magnetic resonance imaging (EOB-MRI) was compared to multidetector computed tomography (MDCT), extracellular contrast media enhanced-MRI (ECCM-MRI) (the US and China only) and contrast-enhanced ultrasound (CEUS). MATERIALS AND METHODS: Decision tree models were developed to simulate the clinical pathway, from first diagnostic test to initial treatment decision, based on local clinical guidelines and validated by experts. Input data were derived from the literature (up to 31st December 2020) as well as from interviews with local experts. A Markov model extension was built to evaluate the number of false-negative patients and associated costs, over a lifetime horizon. RESULTS: The decision-tree models showed that, increasing proportionate use of initial EOB-MRI resulted in a cost-offset per patient (excluding false-negative patients) in all countries (USD 201 for the US, JPY 6,284 for Japan and CNY 446 for China) driven by reductions in follow-on diagnostic procedures and unnecessary treatment. The use of EOB-MRI was also associated with a shorter average waiting time to a final diagnosis and treatment decision compared to MDCT, ECCM-MRI and CEUS. The Markov model showed that with an increase in EOB-MRI use, there are fewer false-negative diagnoses over a lifetime horizon. In all three countries, the incremental cost-effectivenes ratio (ICER) was below standard willingness-to-pay thresholds. CONCLUSION: The findings of these models demonstrate that use of EOB-MRI early in the diagnostic pathway for CRCLM results in short-term cost savings, as well as being cost effective in the long term.

Evaluation of the health economic impact of initial diagnostic modality selection in patients suspected of having HCC in China and the USA.

AIMS: To compare relative costs associated with the diagnostic pathways for hepatocellular carcinoma (HCC) in the US and China according to the initial imaging modality used. Gadoxetate disodium (ethoxylbenzyl-diethylenetriaminepentaacetic acid)-enhanced magnetic resonance imaging (EOB-MRI) was compared to contrast-enhanced multidetector computed tomography (MDCT), extracellular contrast media enhanced-MRI (ECCM-MRI) and contrast-enhanced ultrasound (CEUS). MATERIALS AND METHODS: Decision tree models were developed to simulate the clinical pathway, based on local clinical guidelines, and validated by experts. Input data were derived from the literature (up to 31 December 2020) as well as from interviews with local experts. RESULTS: The models showed that compared to alternative initial imaging modalities, EOB-MRI was associated with higher diagnostic accuracy (fewer false-positive and fewer false-negative results). Increasing proportionate use of EOB-MRI resulted in a cost offset per patient (excluding false-negative patients) in both the US (USD 337) and China (CNY 1,443), driven by reductions in scan costs and unnecessary treatment costs. The use of EOB-MRI was also associated with a shorter average waiting time for a final diagnosis and treatment decision for patients compared to MDCT, ECCM-MRI, and CEUS. CONCLUSION: The findings of these models demonstrate that EOB-MRI is the most accurate and rapid imaging modality for the diagnosis of HCC in the US and China, resulting in cost offsets that may benefit the healthcare system.

Chronic kidney disease progression among patients with type 2 diabetes identified in US administrative claims: a population cohort study.

BACKGROUND: Chronic kidney disease (CKD), one of the most common complications of type 2 diabetes (T2D), is associated with poor health outcomes and high healthcare expenditures. As the CKD population increases, a better understanding of the prevalence and progression of CKD is critical. However, few contemporary studies have explored the progression of CKD relative to its onset in T2D patients using established markers derived from real-world care settings. METHODS: This retrospective, population-based cohort study assessed CKD progression among adults with T2D and with newly recognized CKD identified from US administrative claims data between 1 January 2008 and 30 September 2018. Included were patients with T2D and laboratory evidence of CKD as indicated by the established estimated glomerular filtration rate (eGFR) and urine albumin:creatinine ratio (UACR) criteria. Disease progression was described as transitions across the eGFR- and UACR-based stages. RESULTS: A total of 65 731 and 23 035 patients with T2D contributed to the analysis of eGFR- and UACR-based CKD stage progression, respectively. CKD worsening was observed in approximately 10-17% of patients over a median follow-up of 2 years. Approximately one-third of patients experienced an increase in eGFR values or a decrease in UACR values during follow-up. CONCLUSIONS: A relatively high proportion of patients were observed with disease progression over a short period of time, highlighting the need for better identification of patients at risk of rapidly progressive CKD. Future studies are needed to determine the clinical characteristics of these patients to inform earlier diagnostic and therapeutic interventions aimed at slowing disease progression.

Adherence to Chronic Kidney Disease Screening Guidelines Among Patients With Type 2 Diabetes in a US Administrative Claims Database.

OBJECTIVE: To examine the screening rates for kidney damage and function among patients with type 2 diabetes (T2D) and chronic kidney disease stage at diabetes diagnosis using a US administrative claims database. PATIENTS AND METHODS: This cohort study used a claims database enriched with laboratory results data. Patients with T2D (defined as 1 inpatient or 2 outpatient claims for diabetes), aged 18 years or older, and with at least 1 year of follow-up enrollment were identified. Patients with type 1 diabetes, kidney disease, or other related conditions at baseline were excluded. We estimated screening rates using laboratory orders for serum creatinine and estimated glomerular filtration rate (eGFR) measurement and urine albumin to creatinine ratio (UACR). Chronic kidney disease severity was reported using the Kidney Disease: Improving Global Outcomes classification based on laboratory results. RESULTS: A total of 1,881,447 patients with T2D were eligible for analysis. Mean ± SD age was 63.1±13.1 years; 947,150 patients (50.3%) were male. Serum creatinine tests were ordered within 14 days of the index date among 290,722 patients of 622,915 (46.7%) patients with newly-recognized T2D. Overall, 1,595,964 patients (84.8%) had at least one serum creatinine test ordered during the 1-year follow-up period. Fewer patients received a UACR test during follow-up (814,897 [43.3%]). Less than half of all patients with T2D received a laboratory test order for both serum creatinine and urine albumin measurements during the follow-up period. CONCLUSION: Physicians treating patients with diabetes are selectively adhering to chronic kidney disease screening guidelines, as indicated by high rates of eGFR testing, but less frequent UACR testing. Despite recommendations to monitor both eGFR and UACR, less than half of patients were screened for albuminuria during the 1-year follow-up.

Cardiovascular and Renal Outcomes in Patients with Type-2 Diabetes and Chronic Kidney Disease Identified in a United States Administrative Claims Database: A Population Cohort Study.

INTRODUCTION: CKD, a common complication of type-2 diabetes (T2D), causes considerable disease burden. Patients with T2D and CKD are considered high-risk for complications; however, studies describing patients with T2D and incident CKD identified from real-world data using the diagnostic gold-standard criteria - estimated glomerular filtration rate and urine albumin-to-creatinine ratio (UACR) - are scarce. METHODS: In this population-based cohort study, we sought to estimate the rates of cardiovascular and renal outcomes among patients with T2D and CKD by comorbidity subgroups and CKD severity. Patients were sampled between 2008 and 2017 from de-identified US administrative claims enriched with laboratory data. Analyses were stratified by prevalent heart failure (HF), anemia, and resistant hypertension and the KDIGO categories at index. RESULTS: We identified 106,369 patients with T2D and incident CKD. The rate of all-cause hospitalization was 189 [95% CI: 187, 191] per 1,000 person-years with cardiovascular-related hospitalizations being more frequent than kidney-related outcomes. The rate of acute kidney failure was 77.3 [95% CI: 76.2, 78.5] per 1,000 person-years. Patients with HF experienced a 4-times higher rate for cardiovascular events compared to those without. Rates of hospitalization increased from 5- to 6-fold with increasing KDIGO severity. CONCLUSIONS: Multimorbidity and advance stages of CKD increase the risk of cardiovascular and renal complications among patients with T2D diabetes. Earlier CKD diagnosis as well as interventions and coordinated care addressing other comorbid conditions present at diagnosis may reduce the overall disease burden in this population.

Annual health care resource utilization and cost among type 2 diabetes patients with newly recognized chronic kidney disease within a large U.S. administrative claims database.

BACKGROUND: Chronic kidney disease (CKD) is one of the most common complications of type 2 diabetes mellitus (T2D) and results in considerable economic burden. Current studies describing cost and health care resource utilization (HCRU) in T2D patients with CKD in real-world data are few. Even more scarce is evidence that takes into account disease severity and other comorbidities. OBJECTIVES: To (a) describe T2D patients with CKD identified in U.S. administrative claims data using laboratory test results for kidney function that are considered the gold standard criteria for kidney disease diagnosis and (b) estimate the annual HCRU and costs among these patients, overall and by disease severity and comorbidity subgroup. METHODS: Optum CDM data between the years 2008 and 2017 were used to identify T2D patients with newly recognized CKD, using laboratory test results for estimated glomerular filtration rate (eGFR) or urine albumin-to-creatinine ratio (UACR). The study estimated annualized total, inpatient, outpatient, and pharmacy costs and the number of outpatient, inpatient, and emergency room visits in the first year after CKD identification. Analyses were stratified by prevalent anemia, heart failure (HF), resistant hypertension, and by CKD stages. RESULTS: T2D patients with newly recognized CKD (n = 106,369) had a high prevalence of cardiovascular comorbidities and incurred on average $24,029 of total cost per person per year in the first year after CKD identification. Patients with HF and anemia incurred on average $41,951 and $31,127 of total annual cost, respectively. Patients identified at stage 5 CKD incurred on average $110,210 of total annual cost and had roughly a 7-fold higher annual inpatient hospitalization rate compared with patients identified at stage 1 CKD. CONCLUSIONS: Administrative claims data linked to laboratory results provide an opportunity to identify CKD patients using the gold standard criteria from clinical practice, minimizing potential misclassification of patients. Identified CKD patients, particularly those with HF, anemia, and more advanced CKD stage, incur high HCRU and cost. Better monitoring, earlier CKD diagnosis, and interventions that are effective in halting or slowing the progression of CKD, as well as at managing comorbid conditions, could be effective means to reduce the economic burden of CKD in T2D. DISCLOSURES: This study was funded by Bayer. Kelly is an employee of, and owns stock options in, Aetion, which was contracted by Bayer to conduct the study. Petruski-Ivleva was an employee of Aetion during the planning, analysis, and interpretation stages of the study. Kovesdy received honoraria from Amgen, Astra Zeneca, Bayer, Cara Therapeutics, Reata, Takeda, and Tricida. Fried received consultant fees from Bayer, Novo Nordisk, and Bristol-Meyers Squibb. Folkerts, Blankenburg, and Gay are Bayer employees. This work was presented as a poster at the annual European Association for the Study of Diabetes (EASD) conference held in Barcelona, Spain, on September 16-20, 2019.

Disease characteristics and outcomes in patients with chronic kidney disease and type 2 diabetes: a matched cohort study of spironolactone users and non-users.

BACKGROUND: Limited evidence has indicated that addition of a steroidal mineralocorticoid receptor antagonist (MRA) to the standard of care reduces proteinuria in patients with diabetic kidney disease (DKD); however, there are limited data regarding real-world MRA use in these patients. This study aimed to describe the characteristics of spironolactone users and non-users with DKD, and to explore their clinical outcomes. METHODS: This was a non-interventional, retrospective cohort study using demographic and clinical data from a US claims database (PharMetrics Plus) and the Experian consumer data asset during 2006-2015. Baseline characteristics (e.g. comorbidities) and post-inclusion clinical outcomes were described in matched cohorts of spironolactone users and non-users (n = 5465 per group). RESULTS: Although matching aligned key demographic and clinical characteristics of the cohorts, a significantly greater proportion of spironolactone users than non-users had oedema, proteinuria, and cardiovascular disease at baseline (P < 0.0001). During the post-inclusion period, disease progression and clinical events of interest such as acute kidney injury were more commonly observed in spironolactone users than non-users. Users also had higher healthcare resource utilization and costs than non-users; however, these differences diminished at later stages of disease. CONCLUSIONS: In this study, spironolactone users had a greater comorbidity burden at baseline than matched non-users, suggesting that the presence of certain comorbidities may be contributing factors in the decision to prescribe spironolactone. High healthcare resource utilization and costs for patients at later stages of disease, irrespective of spironolactone use, highlight the need for new therapies for DKD.

Patient characteristics and initiation of mineralocorticoid receptor antagonists in patients with chronic kidney disease in routine clinical practice in the US: a retrospective cohort study.

BACKGROUND: Steroidal mineralocorticoid receptor antagonists (MRAs) are recommended for the treatment of heart failure (HF) and resistant hypertension, both common comorbidities in patients with diabetes and chronic kidney disease (CKD). This study explored the clinical characteristics of, and steroidal MRA use in, patients with CKD with and without type 2 diabetes mellitus (T2D) and/or HF. METHODS: This retrospective cohort study used PharMetrics Plus US claims database data (October 2009-September 2014) to identify two patient populations aged ≥18 years with a first diagnosis of CKD or a first prescription for steroidal MRAs. Demographic characteristics, comorbidities, clinical events, medication use and healthcare costs were reported by population and stratified by diagnosis: CKD, CKD + T2D (DKD), CKD + HF and DKD + HF. The CKD population cohorts were further stratified by steroidal MRA treatment duration (no MRAs, < 6 and ≥ 6 months' treatment). RESULTS: The CKD and MRA populations comprised 229,004 patients and 5899 patients, respectively. Median age and the proportion of men were similar in the CKD and MRA populations across disease cohorts. Disease burden increased across the cohorts as comorbidity and clinical event incidences increased. Hypertension was reported in 70-92% of patients, irrespective of disease cohort or population. In the CKD population, MRA use was low but increased with disease burden: CKD, 1.2%; DKD, 1.8%; CKD + HF, 6.5%; and DKD + HF, 6.6%. Moreover, MRA users presented with higher rates of comorbidities and medication use, and higher healthcare costs than MRA non-users. Longer MRA treatment duration was associated with reduced polypharmacy, lower event rates and lower healthcare costs. In the MRA population, patients almost exclusively received spironolactone (≥ 96%; median dose across all groups 25 mg; one-year persistence, ≤ 43%); up to 16% of patients had end-stage renal disease at baseline despite steroidal MRAs being contraindicated. CONCLUSIONS: Steroidal MRA use was low across all cohorts, but increased with disease severity, driven particularly by HF. Steroidal MRAs were used in patients with advanced CKD, despite being contraindicated. The persistent morbidity and clinical event rates in CKD and DKD patients highlight the disease burden and the need for treatments that effectively target both cardio-vascular and kidney-related events.

Central nervous system drug consumption depending on the time between symptom onset and the diagnosis of Alzheimer's disease: an analysis by the Registry of Dementias of Girona.

AIMS: To describe central nervous system (CNS) drug consumption patterns depending on the time to diagnosis of Alzheimer's disease (AD), and to check whether the cases diagnosed later are associated with greater severity and consuming more CNS drugs. METHODS: Cross-sectional study using 952 cases of the Registry of Dementias of Girona. A binary logistic regression was used to detect variables associated with the use of CNS drugs depending on the time to diagnosis. RESULTS: CNS drugs were consumed by 95.8% of the AD patients. Only antipsychotics presented a statistically significant increase in the frequency of prescription to patients with longer time elapsed from symptom onset to AD diagnosis. CONCLUSION: Longer time elapsed from the onset of symptoms to the diagnosis resulted in increased probability of antipsychotic consumption.

Florbetaben PET in the Early Diagnosis of Alzheimer's Disease: A Discrete Event Simulation to Explore Its Potential Value and Key Data Gaps.

The growing understanding of the use of biomarkers in Alzheimer's disease (AD) may enable physicians to make more accurate and timely diagnoses. Florbetaben, a beta-amyloid tracer used with positron emission tomography (PET), is one of these diagnostic biomarkers. This analysis was undertaken to explore the potential value of florbetaben PET in the diagnosis of AD among patients with suspected dementia and to identify key data that are needed to further substantiate its value. A discrete event simulation was developed to conduct exploratory analyses from both US payer and societal perspectives. The model simulates the lifetime course of disease progression for individuals, evaluating the impact of their patient management from initial diagnostic work-up to final diagnosis. Model inputs were obtained from specific analyses of a large longitudinal dataset from the New England Veterans Healthcare System and supplemented with data from public data sources and assumptions. The analyses indicate that florbetaben PET has the potential to improve patient outcomes and reduce costs under certain scenarios. Key data on the use of florbetaben PET, such as its influence on time to confirmation of final diagnosis, treatment uptake, and treatment persistency, are unavailable and would be required to confirm its value.

Review and meta-analysis of biomarkers and diagnostic imaging in Alzheimer's disease.

Mild Alzheimer's disease (AD) is often difficult to differentiate from mild cognitive impairment (MCI) or non-AD dementias. A multitude of diagnostic biomarkers and advanced imaging strategies have been developed to aid in the diagnosis and management of AD. We sought to review and analyze the published evidence on key test characteristics of major diagnostic strategies to formulate best estimates of sensitivity (SN) and specificity (SP). A systematic review was undertaken to locate and abstract all studies of biomarkers or diagnostic imaging for AD published in English from January 1990 to March 2010 that provided estimates of SN and SP. Meta-analysis was performed using a bivariate mixed-effects binary regression model. We calculated -SN, SP, and area under the receiver operating curves (AUROC), with confidence and prediction contours. Of 1,840 unique studies identified, 119 presented primary data sufficient for analysis. SN and SP were calculated against non-demented controls, non-AD dementias with and without MCI, if available. Compared to non-demented controls, FDG-PET demonstrated the highest AUROC (0.96), with 90% SN (95%CI 84% to 94%), and 89% SP (95% CI 81% to 94%). FDG-PET also was most accurate in discriminating AD from demented controls (including MCI) with AUROC 0.91, and 92% SN (95%CI 84% to 96%) and 78% SP (95% CI 69% to 85%). For discrimination of AD from non-AD dementias (excluding MCI), CSF Ptau, and SPECT produced identical AUROC (0.86). Diagnostic strategies for AD show wide variation in test characteristics and some show promise for use in clinical practice.