Forecasting Inequalities in Survival to Retirement Age by Socioeconomic Status in Denmark and Sweden.

In Denmark and Sweden, statutory retirement age is indexed to life expectancy to account for mortality improvements in their populations. However, mortality improvements have not been uniform across different sub-populations. Notably, in both countries, individuals of lower socioeconomic status (SES) have experienced slower mortality improvements. As a result, a uniform rise in the statutory retirement age could disproportionally affect these low-SES groups and may unintentionally lead to a reverse redistribution effect, shifting benefits from short-lived low-SES individuals to long-lived high-SES individuals. The aim of this study is twofold: to quantify and contextualise mortality inequalities by SES in Denmark and Sweden, and to assess how indexing retirement age will affect future survival to retirement age by SES in these countries. We used Danish and Swedish registry data (1988-2019), to aggregate individuals aged 50 + based on their demographic characteristics and SES. We computed period life tables by year, sex, and SES to estimate the difference in survival across different SES groups. We then forecast mortality across SES groups to assess how indexing retirement age will affect survival inequalities to retirement age, using two forecasting models-the Mode model and the Li-Lee model. Mortality inequalities are comparable in Denmark and Sweden, even though the latter generally has higher survival. We also find that indexing retirement age to life expectancy will have two main consequences: it will reduce the probability of reaching retirement for all SES groups, particularly those of low SES, and time spent in retirement will be reduced, particularly for those of high SES.

Comparing code-free deep learning models to expert-designed models for detecting retinal diseases from optical coherence tomography.

BACKGROUND: Code-free deep learning (CFDL) is a novel tool in artificial intelligence (AI). This study directly compared the discriminative performance of CFDL models designed by ophthalmologists without coding experience against bespoke models designed by AI experts in detecting retinal pathologies from optical coherence tomography (OCT) videos and fovea-centered images. METHODS: Using the same internal dataset of 1,173 OCT macular videos and fovea-centered images, model development was performed simultaneously but independently by an ophthalmology resident (CFDL models) and a postdoctoral researcher with expertise in AI (bespoke models). We designed a multi-class model to categorize video and fovea-centered images into five labels: normal retina, macular hole, epiretinal membrane, wet age-related macular degeneration and diabetic macular edema. We qualitatively compared point estimates of the performance metrics of the CFDL and bespoke models. RESULTS: For videos, the CFDL model demonstrated excellent discriminative performance, even outperforming the bespoke models for some metrics: area under the precision-recall curve was 0.984 (vs. 0.901), precision and sensitivity were both 94.1% (vs. 94.2%) and accuracy was 94.1% (vs. 96.7%). The fovea-centered CFDL model overall performed better than video-based model and was as accurate as the best bespoke model. CONCLUSION: This comparative study demonstrated that code-free models created by clinicians without coding expertise perform as accurately as expert-designed bespoke models at classifying various retinal pathologies from OCT videos and images. CFDL represents a step forward towards the democratization of AI in medicine, although its numerous limitations must be carefully addressed to ensure its effective application in healthcare.

A modal age at death approach to forecasting adult mortality.

Recent studies have shown that there are some advantages to forecasting mortality with indicators other than age-specific death rates. The mean, median, and modal ages at death can be directly estimated from the age-at-death distribution, as can information on lifespan variation. The modal age at death has been increasing linearly since the second half of the twentieth century, providing a strong basis from which to extrapolate past trends. The aim of this paper is to develop a forecasting model that is based on the regularity of the modal age at death and that can also account for changes in lifespan variation. We forecast mortality at ages 40 and above in 10 West European countries. The model we introduce increases forecast accuracy compared with other forecasting models and provides consistent trends in life expectancy and lifespan variation at age 40 over time.

MBT: Model-Based Transformer for retinal optical coherence tomography image and video multi-classification.

BACKGROUND AND OBJECTIVE: The detection of retinal diseases using optical coherence tomography (OCT) images and videos is a concrete example of a data classification problem. In recent years, Transformer architectures have been successfully applied to solve a variety of real-world classification problems. Although they have shown impressive discriminative abilities compared to other state-of-the-art models, improving their performance is essential, especially in healthcare-related problems. METHODS: This paper presents an effective technique named model-based transformer (MBT). It is based on popular pre-trained transformer models, particularly, vision transformer, swin transformer for OCT image classification, and multiscale vision transformer for OCT video classification. The proposed approach is designed to represent OCT data by taking advantage of an approximate sparse representation technique. Then, it estimates the optimal features, and performs data classification. RESULTS: The experiments are carried out using three real-world retinal datasets. The experimental results on OCT image and OCT video datasets show that the proposed method outperforms existing state-of-the-art deep learning approaches in terms of classification accuracy, precision, recall, and f1-score, kappa, AUC-ROC, and AUC-PR. It can also boost the performance of existing transformer models, including Vision transformer and Swin transformer for OCT image classification, and Multiscale Vision Transformers for OCT video classification. CONCLUSIONS: This work presents an approach for the automated detection of retinal diseases. Although deep neural networks have proven great potential in ophthalmology applications, our findings demonstrate for the first time a new way to identify retinal pathologies using OCT videos instead of images. Moreover, our proposal can help researchers enhance the discriminative capacity of a variety of powerful deep learning models presented in published papers. This can be valuable for future directions in medical research and clinical practice.

Assessing the feasibility, acceptability, and fidelity of a tele-retinopathy-based intervention to encourage greater attendance to diabetic retinopathy screening in immigrants living with diabetes from China and African-Caribbean countries in Ottawa, Canada: a protocol.

BACKGROUND: Diabetic retinopathy is a leading cause of preventable blindness in Canada. Clinical guidelines recommend annual diabetic retinopathy screening for people living with diabetes to reduce the risk and progression of vision loss. However, many Canadians with diabetes do not attend screening. Screening rates are even lower in immigrants to Canada including people from China, Africa, and the Caribbean, and these groups are also at higher risk of developing diabetes complications. We aim to assess the feasibility, acceptability, and fidelity of a co-developed, linguistically and culturally tailored tele-retinopathy screening intervention for Mandarin-speaking immigrants from China and French-speaking immigrants from African-Caribbean countries living with diabetes in Ottawa, Canada, and identify how many from each population group attend screening during the pilot period. METHODS: We will work with our health system and patient partners to conduct a 6-month feasibility pilot of a tele-retinopathy screening intervention in a Community Health Centre in Ottawa. We anticipate recruiting 50-150 patients and 5-10 health care providers involved in delivering the intervention for the pilot. Acceptability will be assessed via a Theoretical Framework of Acceptability-informed survey with patients and health care providers. To assess feasibility, we will use a Theoretical Domains Framework-informed interview guide and to assess fidelity, and we will use a survey informed by the National Institutes of Health framework from the perspective of health care providers. We will also collect patient demographics (i.e., age, gender, ethnicity, health insurance status, and immigration information), screening outcomes (i.e., patients with retinopathy identified, patients requiring specialist care), patient costs, and other intervention-related variables such as preferred language. Survey data will be descriptively analyzed and qualitative data will undergo content analysis. DISCUSSION: This feasibility pilot study will capture how many people living with diabetes from each group attend the diabetic retinopathy screening, costs, and implementation processes for the tele-retinopathy screening intervention. The study will indicate the practicability and suitability of the intervention in increasing screening attendance in the target population groups. The study results will inform a patient-randomized trial, provide evidence to conduct an economic evaluation of the intervention, and optimize the community-based intervention.

The co-development of a linguistic and culturally tailored tele-retinopathy screening intervention for immigrants living with diabetes from China and African-Caribbean countries in Ottawa, Canada.

BACKGROUND: Diabetic retinopathy is a sight-threatening ocular complication of diabetes. Screening is an effective way to reduce severe complications, but screening attendance rates are often low, particularly for newcomers and immigrants to Canada and people from cultural and linguistic minority groups. Building on previous work, in partnership with patient and health system stakeholders, we co-developed a linguistically and culturally tailored tele-retinopathy screening intervention for people living with diabetes who recently immigrated to Canada from either China or African-Caribbean countries. METHODS: Following an environmental scan of diabetes eye care pathways in Ottawa, we conducted co-development workshops using a nominal group technique to create and prioritize personas of individuals requiring screening and identify barriers to screening that each persona may face. Next, we used the Theoretical Domains Framework to categorize the barriers/enablers and then mapped these categories to potential evidence-informed behaviour change techniques. Finally with these techniques in mind, participants prioritized strategies and channels of delivery, developed intervention content, and clarified actions required by different actors to overcome anticipated intervention delivery barriers. RESULTS: We carried out iterative co-development workshops with Mandarin and French-speaking individuals living with diabetes (i.e., patients in the community) who immigrated to Canada from China and African-Caribbean countries (n = 13), patient partners (n = 7), and health system partners (n = 6) recruited from community health centres in Ottawa. Patients in the community co-development workshops were conducted in Mandarin or French. Together, we prioritized five barriers to attending diabetic retinopathy screening: language (TDF Domains: skills, social influences), retinopathy familiarity (knowledge, beliefs about consequences), physician barriers regarding communication for screening (social influences), lack of publicity about screening (knowledge, environmental context and resources), and fitting screening around other activities (environmental context and resources). The resulting intervention included the following behaviour change techniques to address prioritized local barriers: information about health consequence, providing instructions on how to attend screening, prompts/cues, adding objects to the environment, social support, and restructuring the social environment. Operationalized delivery channels incorporated language support, pre-booking screening and sending reminders, social support via social media and community champions, and providing using flyers and videos as delivery channels. CONCLUSION: Working with intervention users and stakeholders, we co-developed a culturally and linguistically relevant tele-retinopathy intervention to address barriers to attending diabetic retinopathy screening and increase uptake among two under-served groups.

Gemcitabine promotes autophagy and lysosomal function through ERK- and TFEB-dependent mechanisms.

Gemcitabine is a first-line treatment agent for pancreatic ductal adenocarcinoma (PDAC). Contributing to its cytotoxicity, this chemotherapeutic agent is primarily a DNA replication inhibitor that also induces DNA damage. However, its therapeutic effects are limited owing to chemoresistance. Evidence in the literature points to a role for autophagy in restricting the efficacy of gemcitabine. Autophagy is a catabolic process in which intracellular components are delivered to degradative organelles lysosomes. Interfering with this process sensitizes PDAC cells to gemcitabine. It is consequently inferred that autophagy and lysosomal function need to be tightly regulated to maintain homeostasis and provide resistance to environmental stress, such as those imposed by chemotherapeutic drugs. However, the mechanism(s) through which gemcitabine promotes autophagy remains elusive, and the impact of gemcitabine on lysosomal function remains largely unexplored. Therefore, we applied complementary approaches to define the mechanisms triggered by gemcitabine that support autophagy and lysosome function. We found that gemcitabine elicited ERK-dependent autophagy in PDAC cells, but did not stimulate ERK activity or autophagy in non-tumoral human pancreatic epithelial cells. Gemcitabine also promoted transcription factor EB (TFEB)-dependent lysosomal function in PDAC cells. Indeed, treating PDAC cells with gemcitabine caused expansion of the lysosomal network, as revealed by Lysosome associated membrane protein-1 (LAMP1) and LysoTracker staining. More specific approaches have shown that gemcitabine promotes the activity of cathepsin B (CTSB), a cysteine protease playing an active role in lysosomal degradation. We showed that lysosomal function induced by gemcitabine depends on TFEB, the master regulator of autophagy and lysosomal biogenesis. Interfering with TFEB function considerably limited the clonogenic growth of PDAC cells and hindered the capacity of TFEB-depleted PDAC cells to develop orthotopic tumors.

IL-15 Prevents the Development of T-ALL from Aberrant Thymocytes with Impaired DNA Repair Functions and Increased NOTCH1 Activation.

We previously reported that NOD.Scid mice lacking interleukin-15 (IL-15), or IL-15 receptor alpha-chain, develop T-acute lymphoblastic leukemia (T-ALL). To understand the mechanisms by which IL-15 signaling controls T-ALL development, we studied the thymocyte developmental events in IL-15-deficient Scid mice from NOD and C57BL/6 genetic backgrounds. Both kinds of mice develop T-ALL characterized by circulating TCR-negative cells expressing CD4, CD8 or both. Analyses of thymocytes in NOD.Scid.Il15-/- mice prior to T-ALL development revealed discernible changes within the CD4-CD8- double-negative (DN) thymocyte developmental stages and increased frequencies of CD4+CD8+ double-positive cells with a high proportion of TCR-negative CD4+ and CD8+ cells. The DN cells also showed elevated expressions of CXCR4 and CD117, molecules implicated in the expansion of DN thymocytes. T-ALL cell lines and primary leukemic cells from IL-15-deficient NOD.Scid and C57BL/6.Scid mice displayed increased NOTCH1 activation that was inhibited by NOTCH1 inhibitors and blockers of the PI3K/AKT pathway. Primary leukemic cells from NOD.Scid.Il15-/- mice survived and expanded when cultured with MS5 thymic stromal cells expressing Delta-like ligand 4 and supplemented with IL-7 and FLT3 ligand. These findings suggest that IL-15 signaling in the thymus controls T-ALL development from aberrant thymocytes with an impaired DNA repair capacity and increased NOTCH1 activation.

Focused Attention in Transformers for interpretable classification of retinal images.

Vision Transformers have recently emerged as a competitive architecture in image classification. The tremendous popularity of this model and its variants comes from its high performance and its ability to produce interpretable predictions. However, both of these characteristics remain to be assessed in depth on retinal images. This study proposes a thorough performance evaluation of several Transformers compared to traditional Convolutional Neural Network (CNN) models for retinal disease classification. Special attention is given to multi-modality imaging (fundus and OCT) and generalization to external data. In addition, we propose a novel mechanism to generate interpretable predictions via attribution maps. Existing attribution methods from Transformer models have the disadvantage of producing low-resolution heatmaps. Our contribution, called Focused Attention, uses iterative conditional patch resampling to tackle this issue. By means of a survey involving four retinal specialists, we validated both the superior interpretability of Vision Transformers compared to the attribution maps produced from CNNs and the relevance of Focused Attention as a lesion detector.

Probability of males to outlive females: an international comparison from 1751 to 2020.

OBJECTIVE: To measure sex differences in lifespan based on the probability of males to outlive females. DESIGN: International comparison of national and regional sex-specific life tables from the Human Mortality Database and the World Population Prospects. SETTING: 199 populations spanning all continents, between 1751 and 2020. PRIMARY OUTCOME MEASURE: We used the outsurvival statistic ( φ ) to measure inequality in lifespan between sexes, which is interpreted here as the probability of males to outlive females. RESULTS: In random pairs of one male and one female at age 0, the probability of the male outliving the female varies between 25% and 50% for life tables in almost all years since 1751 and across almost all populations. We show that φ is negatively correlated with sex differences in life expectancy and positively correlated with the level of lifespan variation. The important reduction of lifespan inequality observed in recent years has made it less likely for a male to outlive a female. CONCLUSIONS: Although male life expectancy is generally lower than female life expectancy, and male death rates are usually higher at all ages, males have a substantial chance of outliving females. These findings challenge the general impression that 'men do not live as long as women' and reveal a more nuanced inequality in lifespans between females and males.

[Teaching in a pandemic: the transformation of teaching and clinical supervision]. / Enseigner en situation de pandémie : La transformation de l'enseignement et de la supervision clinique.

Introduction: The COVID-19 pandemic is an event that deeply impacts our personal, professional, and collective lives. How do we teach in these times of great upheaval? What are the main changes that have occurred? Method: Using the Cartel logic, four professors and a qualitative researcher carried out an autoethnographic research aimed at documenting the main changes that have occurred in the teaching of family medicine in their respective practices located in four different academic family medicine groups at the University of Sherbrooke. Results: Five key moments in teaching that occurred during a pandemic were identified: a) the declaration of a pandemic, b) the approach with the graduating/advanced cohort of residents, c) the anticipation and preparation for the arrival of new residents, d) arrival of first year residents and e) adaptation to the second wave. For each moment, we present the issues encountered in our care and teaching practices under three transversal relational axes: the relationship of humans to their cultural context, the patient-doctor relationship, and the teacher-resident relationship. Conclusion: Our analysis shows that the transmission of medical knowledge and the art of medicine cannot take place without specific attention to the overall cultural context, the contextual relationship of clinical care, and the teaching relationship. Our study also makes it possible to recommend the opening of spaces for reflection and dialogue in our teaching environments.

Central Nervous System Metastases in Thymic Epithelial Tumors: A Brief Report of Real-World Insight From RYTHMIC.

Thymic epithelial tumors (TETs) are rare malignancies ranging from indolent thymoma A to aggressive thymic carcinomas (TCs). Brain metastases are extremely infrequent for TETs and have only been described in case reports or small single-center series. RYTHMIC (Réseau tumeurs THYMiques et Cancer) is a French nationwide network mandated to systematically review every TET case and prospectively includes all consecutive patients discussed by national or regional tumor boards. We analyzed patients with TETs and central nervous system (CNS) metastasis during their cancer history from this large French registry. In an 8-year period, 2909 patients were included in the database, including 248 TCs (8.5%). A total of 14 patients had CNS metastases, five (36%) at diagnosis and nine (64%) at relapse. Among them, 12 patients (86%) had a diagnosis of TC and two (14%) had thymoma A and B3. Surgical biopsies were performed, and the histologic subtype for non-TC tumors was centrally confirmed. Median overall survival was 22 months (95% confidence interval [CI]: 9.8-34.2), with longer, albeit not significant, overall survival when CNS metastases were present at diagnosis versus relapse (not reached versus 17 mo; p = 0.29); median progression-free survival was 13 versus 8 months (p = 0.06), respectively. A higher risk of death (hazard ratio = 5.34, 95% CI: 1.3-21.9, p = 0.02) and relapse (hazard ratio = 1.89, 95% CI: 0.9-3.7, p = 0.06) was observed for patients suffering from TC with brain metastases compared with those without CNS extension. CNS disease was extremely rare in our TET cohort (0.48%), reported at both diagnosis and progression, present primarily in TC, with prevalence rising to 4.9%.

Impact of continuous positive airway pressure mode on adherence to treatment in obstructive sleep apnea patients awaiting bariatric surgery.

Obstructive sleep apnea is prevalent in the bariatric population, and is associated with various complications. Despite increasing popularity, automatic positive airway pressure has not yet been studied in this population. The objective was to compare treatment adherence between automatic positive airway pressure and fixed positive airway pressure (continuous positive airway pressure) in obstructive sleep apnea patients awaiting bariatric surgery. This randomized controlled trial involved obese patients newly diagnosed with severe obstructive sleep apnea and awaiting bariatric surgery. The primary outcome was the difference in adherence between automatic positive airway pressure and continuous positive airway pressure pre-operatively. Secondary outcomes included positive airway pressure efficacy, adherence at 1 month, adverse effects, quality of life and peri-operative complications. Analyses were conducted using a modified intention-to-treat methodology. Fifty patients were randomized. Baseline characteristics and duration of positive airway pressure therapy were comparable between groups. At the time of surgery, the percentage of overall nights positive airway pressure used was 96.9% [95% confidence interval: 93.5-100] and 86.0% [95% confidence interval: 66.9-100] in the automatic positive airway pressure and continuous positive airway pressure groups, respectively (p = .047). Average use was 6.3 hr per night [95% confidence interval: 5.1-7.2] and 5.9 hr per night [95% confidence interval: 3.0-8.8], with a difference of 0.4 hr favouring automatic positive airway pressure (p = .75). Nightly use ≥ 4 hr per night was 86.4% and 74.0% in the automatic positive airway pressure and fixed continuous positive airway pressure groups, respectively (p = .22). There were no statistically significant differences regarding adherence at 1 month, efficacy parameters, adverse effects, quality of life and peri-operative complications. With no difference on the safety profile and efficiency parameters, treatment adherence is not improved with automatic positive airway pressure compared with fixed continuous positive airway pressure in obstructive sleep apnea patients awaiting bariatric surgery.

Demographic perspectives on the rise of longevity.

This article reviews some key strands of demographic research on past trends in human longevity and explores possible future trends in life expectancy at birth. Demographic data on age-specific mortality are used to estimate life expectancy, and validated data on exceptional life spans are used to study the maximum length of life. In the countries doing best each year, life expectancy started to increase around 1840 at a pace of almost 2.5 y per decade. This trend has continued until the present. Contrary to classical evolutionary theories of senescence and contrary to the predictions of many experts, the frontier of survival is advancing to higher ages. Furthermore, individual life spans are becoming more equal, reducing inequalities, with octogenarians and nonagenarians accounting for most deaths in countries with the highest life expectancy. If the current pace of progress in life expectancy continues, most children born this millennium will celebrate their 100th birthday. Considerable uncertainty, however, clouds forecasts: Life expectancy and maximum life span might increase very little if at all, or longevity might rise much faster than in the past. Substantial progress has been made over the past three decades in deepening understanding of how long humans have lived and how long they might live. The social, economic, health, cultural, and political consequences of further increases in longevity are so significant that the development of more powerful methods of forecasting is a priority.

Diversification in causes of death in low-mortality countries: emerging patterns and implications.

INTRODUCTION: An important role of public health organisations is to monitor indicators of variation, so as to disclose underlying inequality in health improvement. In industrialised societies, more individuals than ever are reaching older ages and have become more homogeneous in their age at death. This has led to a decrease in lifespan variation, with substantial implications for the reduction of health inequalities. We focus on a new form of variation to shed further light on our understanding of population health and ageing: variation in causes of death. METHODS: Data from the WHO Mortality Database and the Human Mortality Database are used to estimate cause-of-death distributions and life tables in 15 low-mortality countries. Cause-of-death variation, using 19 groups of causes, is quantified using entropy measures and analysed from 1994 to 2017. RESULTS: The last two decades have seen increasing diversity in causes of death in low-mortality countries. There have been important reductions in the share of deaths from diseases of the circulatory system, while the share of a range of other causes, such as diseases of the genitourinary system, mental and behavioural disorders, and diseases of the nervous system, has been increasing, leading to a more complex cause-of-death distribution. CONCLUSIONS: The diversification in causes of death witnessed in recent decades is most likely a result of the increase in life expectancy, together with better diagnoses and awareness of certain diseases. Such emerging patterns bring additional challenges to healthcare systems, such as the need to research, monitor and treat a wider range of diseases. It also raises new questions concerning the distribution of health resources.

Magnetic resonance sounding dataset of a hard-rock headwater catchment for assessing the vertical distribution of water contents in the subsurface.

Magnetic Resonance Sounding (MRS) measurements are acquired at 16 stations in the Strengbach headwater catchment (Vosges Mountains - France). These data, rendering the vertical distribution of water contents in the subsurface, are used to show their potential in conditioning a hydrological model of the catchment, as described in the article "Magnetic resonance sounding measurements as posterior information to condition hydrological model parameters: Application to a hard-rock headwater catchment" - Journal of Hydrology (2020). Acquisition protocols follow a free induction decay scheme. Data are filtered by applying a band-pass filter at the Larmor frequency. A filter removing the 50 Hz noise is also applied with the exception of data at a Larmor frequency close to the 50 Hz harmonic. The signal envelopes are then fitted by a decaying exponential function over time to estimate the median characteristic relaxation time of each MRS sounding.

Assessment of Training Outcomes of Nurse Readers for Diabetic Retinopathy Telescreening: Validation Study.

BACKGROUND: With the high prevalence of diabetic retinopathy and its significant visual consequences if untreated, timely identification and management of diabetic retinopathy is essential. Teleophthalmology programs have assisted in screening a large number of individuals at risk for vision loss from diabetic retinopathy. Training nonophthalmological readers to assess remote fundus images for diabetic retinopathy may further improve the efficiency of such programs. OBJECTIVE: This study aimed to evaluate the performance, safety implications, and progress of 2 ophthalmology nurses trained to read and assess diabetic retinopathy fundus images within a hospital diabetic retinopathy telescreening program. METHODS: In this retrospective interobserver study, 2 ophthalmology nurses followed a specific training program within a hospital diabetic retinopathy telescreening program and were trained to assess diabetic retinopathy images at 2 levels of intervention: detection of diabetic retinopathy (level 1) and identification of referable disease (level 2). The reliability of the assessment by level 1-trained readers in 266 patients and of the identification of patients at risk of vision loss from diabetic retinopathy by level 2-trained readers in 559 more patients were measured. The learning curve, sensitivity, and specificity of the readings were evaluated using a group consensus gold standard. RESULTS: An almost perfect agreement was measured in identifying the presence of diabetic retinopathy in both level 1 readers (κ=0.86 and 0.80) and in identifying referable diabetic retinopathy by level 2 readers (κ=0.80 and 0.83). At least substantial agreement was measured in the level 2 readers for macular edema (κ=0.79 and 0.88) for all eyes. Good screening threshold sensitivities and specificities were obtained for all level readers, with sensitivities of 90.6% and 96.9% and specificities of 95.1% and 85.1% for level 1 readers (readers A and B) and with sensitivities of 86.8% and 91.2% and specificities of 91.7% and 97.0% for level 2 readers (readers A and B). This performance was achieved immediately after training and remained stable throughout the study. CONCLUSIONS: Notwithstanding the small number of trained readers, this study validates the screening performance of level 1 and level 2 diabetic retinopathy readers within this training program, emphasizing practical experience, and allows the establishment of an ongoing assessment clinic. This highlights the importance of supervised, hands-on experience and may help set parameters to further calibrate the training of diabetic retinopathy readers for safe screening programs.

Comparison of Fast-Progression, Hyperprogressive Disease, and Early Deaths in Advanced Non-Small-Cell Lung Cancer Treated With PD-1/PD-L1 Inhibitors or Chemotherapy.

PURPOSE: Hyperprogressive disease (HPD), fast progression (FP), and early death (ED) have been described in 13.8%, 4.7%, and 5.6% and in 5.1%, 2.8%, and 6.8%, respectively, of patients with non-small-cell lung cancer (NSCLC) treated with single-agent programmed cell death ligand 1 inhibitors (ICI) or chemotherapy, respectively. Whether FP/ED and HPD represent overlapping patterns is unknown. PATIENTS AND METHODS: FP, ED, and HPD were retrospectively assessed in patients with NSCLC treated with single-agent ICI or chemotherapy. Eligibility required 2 computed tomography (CT) scans before and 1 CT scan during treatment. (1) HPD, (2) FP, (3) ED were defined as (1) RECIST version 1.1 progression at first CT scan and tumor growth rate variation per month > 50%, (2) ≥ 50% increase in the sum of the longest diameters of target lesions within 6 weeks from baseline, and (3) death as a result of radiologic progression within 12 weeks from baseline CT scan, respectively. RESULTS: Of 406 ICI-treated NSCLC, 56 patients (13.8%), 9 patients (2.2%), and 36 patients (8.8%) were HPD, FP, and ED, respectively. Eight (14.2%) and 20 (35.7%) of 56 patients with HPD were also FP and ED. ED significantly correlated with baseline Eastern Cooperative Oncology Group performance status ≥ 2 compared with HPD (33% v 13%, P = .02). Overall survival was significantly longer for HPD (3.4 months [95% CI, 2.7 to 4.0 months]) compared with FP (0.7 months [95% CI, 0.6 to 0.8 months]); HR, 0.18 [95% CI, 0.08 to 0.42]; P < .0001) and ED (1.4 months [95% CI, 1.3 to 1.6 months]); HR, 0.19 [95% CI, 0.11 to 0.34]); P < .0001), whereas it did not differ between FP and ED (HR, 1.3 [95% CI, 0.56 to 3.0]; P = .55). Of 59 patients with NSCLC treated with single-agent chemotherapy, the HPD, FP, and ED rates were 5.1%, 1.7%, and 6.7%, respectively. CONCLUSION: FP, ED, and HPD represent distinct progression patterns with limited overlap and different survival outcomes.