Utility Measures Research Articles

Interpretable Machine Learning Model to Predict Bone Cement Leakage in Percutaneous Vertebral Augmentation for Osteoporotic Vertebral Compression Fracture Based on SHapley Additive exPlanations.

Retrospective study. Our objective is to create comprehensible machine learning (ML) models that can forecast bone cement leakage in percutaneous vertebral augmentation (PVA) for individuals with osteoporotic vertebral compression fracture (OVCF) while also identifying the associated risk factors. We incorporated data from patients (n = 425) which underwent PVA. To predict cement leakage, we devised six models based on a variety of parameters. Evaluate and juxtapose the predictive performances relied on measures of discrimination, calibration, and clinical utility. SHapley Additive exPlanations (SHAP) methodology was used to interpret model and evaluate the risk factors associated with cement leakage. The occurrence rate of cement leakage was established at 50.4%. A binary logistic regression analysis identified cortical disruption (OR 6.880, 95% CI 4.209-11.246), the basivertebral foramen sign (OR 2.142, 95% CI 1.303-3.521), the fracture type (OR 1.683, 95% CI 1.083-2.617), and the volume of bone cement (OR 1.198, 95% CI 1.070-1.341) as independent predictors of cement leakage. The XGBoost model outperformed all others in predicting cement leakage in the testing set, with AUC of .8819, accuracy of .8025, recall score of .7872, F1 score of .8315, and a precision score of .881. Several important factors related to cement leakage were drawn based on the analysis of SHAP values and their clinical significance. The ML based predictive model demonstrated significant accuracy in forecasting bone cement leakage for patients with OVCF undergoing PVA. When combined with SHAP, ML facilitated a personalized prediction and offered a visual interpretation of feature importance.

Assessing privacy leakage in synthetic 3-D PET imaging using transversal GAN

Background and objectiveTraining computer-vision related algorithms on medical images for disease diagnosis or image segmentation is difficult in large part due to privacy concerns. For this reason, generative image models are highly sought after to facilitate data sharing. However, 3-D generative models are understudied, and investigation of their privacy leakage is needed. MethodsWe introduce our 3-D generative model, Transversal GAN (TrGAN), using head & neck PET images which are conditioned on tumor masks as a case study. We define quantitative measures of image fidelity and utility, and propose a novel framework for evaluating privacy-utility trade-off through membership inference attack. These metrics are evaluated in the course of training to identify ideal fidelity, utility and privacy trade-offs and establish the relationships between these parameters. ResultsWe show that the discriminator of the TrGAN is vulnerable to attack, and that an attacker can identify which samples were used in training with almost perfect accuracy (AUC = 0.99). We also show that an attacker with access to only the generator cannot reliably classify whether a sample had been used for training (AUC = 0.51). We also propose and demonstrate a general decision procedure for any deep learning based generative model, which allows the user to quantify and evaluate the decision trade-off between downstream utility and privacy protection. ConclusionsTrGAN can generate 3-D medical images that retain important image features and statistical properties of the training data set, with minimal privacy loss as determined by a membership inference attack. Our utility-privacy decision procedure may be beneficial to researchers who wish to share data or lack a sufficient number of large labeled image datasets.

Cost-effectiveness analysis of child eye health interventions in India

Abstract Background India has 4.8 million blind persons including 250,000 blind children, and 35 million visually impaired individuals. Prevalence of blindness is 0.36%, it was 1% two decades ago due to significant public health programs towards prevention and treatment by various stakeholders. Time is opportune for cost-effectiveness analyses of these programs. There is a lack of research evidence related to quality-of-life utility values for eye health conditions in India. This study aims to develop utility values for vision impairment and conduct cost-effectiveness analysis of eye health interventions for childhood vision impairment. Methods Cross-sectional questionnaire-based study. 306 Children aged 4-16 years of age with unaddressed vision impairment included. Random sampling was done at a tertiary eye hospital in central India. EQ-5D-Y-3L questionnaire along with VAS score was used to measure utility measures for each of the vision impairment levels. Vision impairment was defined based on WHO classification. Mixed methods were used to collect cost data related to interventions, sensitivity analyses were performed to account for uncertainties. The protocol was approved by institutional ethics committee of university of Hyderabad (UH/IEC/2020/222). Results Median utility value of mild and moderate vision impairment is 0.8 (95% CI 0.7-0.9), severe vision impairment is 0.7 (95% CI 0.65-0.75), and for blindness is 0.55 (95% CI 0.44-0.72). Willingness to pay (WTP) is USD 2,935. ICERs for mild, moderate, severe vision impairment and blindness are USD 81, USD 367, USD 44, USD 954 respectively. Conclusions The utility value of 0.55 for blindness correlates with existing literature. ICERs for various levels of vision impairment show the interventions to be cost-effective. There is a significant improvement in utility values post intervention. The utility measures are recommended to be used to evaluate eye health programs, and as advocacy tools for preventive eye care. Key messages • There is a significant improvement in utility values post-intervention and the interventions are highly cost-effective. • Important decision-making tool for allocating scarce resources to preventive eye care services.

Impact of interventions on the quality of life of cancer patients: a systematic review and meta-analysis of longitudinal research

The impact of cancer interventions has been conducted in several research due to the significant burden of this non-communicable disease. The interventions that played an important role in the improvement of the patient’s quality of life (QoL) and health-related quality of life (HRQL) can be classified into two main groups: pharmaceutical and non-pharmacological methods. However, studies so far often analyze a specific group of interventions for specific types of cancer. Thus, in this systematic review and meta-analysis, we synthesized the overall impact of cancer interventions on patients’ quality of life in several cancers.In this research, we followed the Preferred Reporting Items for Systematic Review and Meta-Analysis Protocols (PRISMA-P) to search the longitudinal original research on the Web of Science (WOS) database. After that, the Newcastle-Ottawa Scale (NOS) and Jadad Scale were used to assess the quality of non-randomized control trials and randomized control trials, respectively. Then, the characteristics of the included studies were described in the six main fields table and the random effect model with robust estimation was applied to analyze the impact of interventions on the health utility of patients.From the database, 122 longitudinal original research were included in the meta-regression, with most of them having high or fair quality. The European Organization for the research and treatment of cancer scale for quality of life (EORTC-QLQ) was the most used health utility measurement at 65.15%. In the adjusted effect models, the Visual Analogue Scale (VAS) had significant statistics in all models when we compared it with the EQ-5D Scale (p < 0.05) and several types of cancer such as breast, lung, and prostate cancer had significant statistics when comparing with hematological cancer in the model types of cancer (p < 0.01). Moreover, radiotherapy, screening, and a combination of chemotherapy and best supportive care also had significant statistics (p < 0.01) in the model of interventions when compared with radiotherapy applied only. Our research can suggest a vital combination of both pharmaceutical and non-pharmacological interventions to improve the quality of life of some common types of cancer patients.

Basics of health economics for clinical trials in orthopaedic trauma

With the rise in global healthcare spending, medical decision-making is increasingly based on health economics outcomes. This narrative review aims to provide an overview of cost-effectiveness analysis (CEA) and cost-utility analysis (CUA), including their advantages and limitations, and outline practical aspects for implementing health economics in clinical trials for orthopaedic trauma.Both CEA and CUA offer several advantages. Firstly, they consider the costs as well as benefits of an intervention, providing a more comprehensive picture of its economic impact. Secondly, they provide a clear and straightforward metric for comparing interventions, which can help decision-makers make informed choices. However, there are methodological shortcomings that must be acknowledged, such as the lack of standardized instruments for assessing health utility measures, which can result in a wide range of cost-benefit ratios. In addition, a consensus on the willingness-to-pay threshold still needs to be reached to develop decision rules for cost-effectiveness similar to clinical effectiveness.Methods such as CEA and CUA should be incorporated into clinical trials in orthopaedic trauma research. Practical aspects for this include planning in advance, preferably in cooperation with a health economist. Selecting appropriate outcome measures is crucial, and both the medical effects of interventions and quality of life instruments should be carefully chosen to ensure comparability with previous studies. Additionally, the potential impact on clinical practice and healthcare policies should be considered. Direct as well as indirect costs should be assessed, and quality assurance with well-established checklists should be confirmed.

Financial Toxicity and Patient Experience Outcomes on a Multi-Institutional Phase I Single Fraction Stereotactic Partial Breast Irradiation Protocol for Early-Stage Breast Cancer

Given the demonstrated financial toxicity (FT) of radiation treatment on breast cancer patients shown in both conventional and our recent 5 fraction stereotactic APBI (S-PBI) study, we assessed the FT, as well as patient-reported utility, quality-of-life and patient experience measures, on patients treated in our phase I single fraction S-PBI trial. A phase I single fraction dose escalation trial of S-PBI for early-stage breast cancer was conducted. Women with in-situ or stage I-II (AJCC 6) invasive breast cancer following breast conserving surgery were treated with S-PBI in 1 fraction to a total dose of 22.5, 26.5 or 30 Gy (Clinical trials.gov ID NCT02685332). At one month follow-up, patients were asked to complete our novel "Patient Perspective Cost and Convenience of Care Questionnaire". Patients also completed the EQ-5D-5L, including the visual analogue scale of overall health (VAS), at enrollment, 6, 12-, 24-, 36-, and 48-month follow-up. Of 29 patients enrolled and treated, questionnaire data was available for all patients. Our trial encompassed a wide range of annual household incomes, education, and employment status. Overall, 44.8% (n = 13/29) of patients reported that radiation treatment presented a financial burden. Interestingly, no demographic information, such as patient race, marital status, education, household income, or employment during treatment predicted perceived FT. Patients reporting FT trended towards younger age (median 64 vs 70.5) and having a cancer related co-pay similar to our 5 fraction S-PBI FT trial; however, due to the small size of this study, this did not reach significance (p = 0.24 and 0.10, respectively). VAS and utility scores were calculated per the EQ-5D-5L and remained unchanged from baseline through 4-year follow-up. Likewise, there was no difference in the utility or VAS between patients who reported FT and those who did not. Interestingly, while patient reported cosmesis was similar for all patients at enrollment, patients who reported FT noted significantly worse cosmesis scores (fair/poor vs good/excellent) at 6 month and 2-year follow-ups (p = 0.01 and 0.04, respectively). Finally, patients were surveyed on treatment related disruption to their daily activities and enjoyment of life. The median values were 0 (scale 0-10, with 0 being no disruption) regardless of perceived FT. Patients were also uniformly satisfied with treatment time with a median score of 10 (scale 0-10, 10 being most satisfied). Here, we show that despite using SPBI in a single fraction, nearly half of the patients treated still reported FT of treatment. Importantly, single fraction S-PBI has no negative impact on patient VAS or utility scores, and all patients were uniformly satisfied with treatment time without significant disruption to their life.

Generic Health Utility Measures in Exercise Oncology: A Scoping Review and Future Directions.

Despite the evidence that exercise is effective at mitigating common side effects in adults with cancer, it is rarely part of usual cancer care. One reason for this is the lack of economic evidence supporting the benefit of exercise. Economic evaluations often rely on the use of generic utility measures to assess cost effectiveness. This review identifies and synthesizes the literature on the use of generic utility measures used to evaluate exercise interventions for adults with cancer. A systematic search of the literature from January 2000 to February 2023 was conducted using four databases (Medline, EMBASE, CINAHL, Academic Search Complete). Exercise studies involving adults with any type of cancer that used a generic utility measure were eligible for inclusion. Of the 2780 citations retrieved, 10 articles were included in this review. Seven articles included economic evaluations, with varying results. Four studies reported on cost-effectiveness; however, detailed effectiveness data derived from the generic utility measure were often not reported. Generic utility measures help to compare baseline values of and changes in health utility weights across studies and to general population norms; however, to date, they are underutilized in exercise oncology studies. Consideration should be given to the identified research evidence, population, and methodological gaps.

Cost of Illness and Health-Related Quality of Life for Stuttering: Two Systematic Reviews.

For those who stutter, verbal communication is typically compromised in social situations. This may attract negative responses from listeners and stigmatization by society. These have the potential to impair health-related quality of life across a range of domains, including qualitative and quantitative impacts on speech output, mental health issues, and failure to attain educational and occupational potential. These systematic reviews were designed to explore this matter using traditional health economics perspectives of utility measures and cost of illness. Studies were included if they involved children, adolescents, or adults with stuttering as a primary diagnosis. The quality of life search strategy identified 2,607 reports, of which three were included in the quality of life analysis. The cost of illness search strategy identified 3,778 reports, of which 39 were included in the cost of illness analysis. Two of the three studies included in the quality of life analysis had a high risk of bias. When measured using utility scores, quality of life for people who stutter was in the range of those reported for chronic health conditions such as diabetes mellitus, cardiovascular disease, and cancer. However, there is little such evidence of quality of life impairment during the preschool years. Studies included in the cost of illness analysis carried considerable risk of bias overall. For people who stutter, there are substantive direct and indirect costs of illness. These include impairment, challenges, and distress across many domains throughout life, including income, education, employment, and social functioning. Evidence of quality of life impairment using utility measures is extremely limited. If this situation is not remedied, the lifetime impairment, challenges, and distress experienced by those who stutter cannot be documented in a form that can be used to influence health policy and health care spending. https://doi.org/10.23641/asha.24168201.

Implementation of Vikor Algorithm in Web-Based Recommendation System for Obstetrician Selection

The process of childbirth is a natural event for women; with the increase in birth rates, selecting a suitable obstetrician becomes essential. However, many couples find it challenging to choose the right obstetrician, from consultations to hospital births. This study aims to develop an obstetrician recommendation system to overcome decision-making challenges that involve many criteria. The method used is VIseKriterijumska Optimizacija Kompromisno Resenje (VIKOR), which is one of the best methods in Multiple Criteria Decision Making (MCDM) for ranking and selecting alternatives. This method involves several steps, including determining the weight of the criteria, normalizing the matrix, calculating the Utility Measure and Regret Measure, and calculating the VIKOR index. Testing was carried out through scenario tests and user survey tests using the USE questionnaire. The scenario test results show the successful implementation of VIKOR in this system. The results of the user survey test indicated a high level of satisfaction with a score of 83.34%, which fulfilled the four variables in the USE questionnaire: usability (79.76%), ease of use (84.81%), ease of learning (84.81), and satisfaction (83.31%). The results of this study have important implications for increasing the effectiveness of decision-making in selecting the right obstetrician. By incorporating the VIKOR method into an application, this research provides a valuable contribution to developing a VIKOR-based recommendation system. It emphasizes the importance of applying this method in complex decision-making environments.

In Pursuit of Beauty: Aesthetic-Aware and Context-Adaptive Photo Selection in Crowdsensing

The pervasive view of the mobile crowd bridges various real-world scenes and people's perceptions with the gathering of distributed crowdsensing photos. To elaborate informative visuals for viewers, existing techniques introduce photo selection as an essential step in crowdsensing. Yet, the aesthetic preference of viewers, at the very heart of their experiences under various crowdsensing contexts (e.g., travel planning), is seldom considered and hardly guaranteed. We propose CrowdPicker, a novel photo selection framework with adaptive aesthetic awareness for crowdsensing. With the observations on aesthetic uncertainty and bias in different crowdsensing contexts, we exploit a joint effort of mobile crowdsourcing and domain adaptation to actively learn contextual knowledge for dynamically tailoring the aesthetic predictor. Concretely, an aesthetic utility measure is invented based on the probabilistic balance formalization to quantify the benefit of photos in improving the adaptation performance. We prove the NP-hardness of sampling the best-utility photos for crowdsourcing annotation and present a (1-1/e) approximate solution. Furthermore, a two-stage distillation-based adaptation architecture is designed based on fusing contextual and common aesthetic preferences. Extensive experiments on three datasets and four raw models demonstrate the performance superiority of CrowdPicker over four photo selection baselines and four typical sampling strategies. Cross-dataset evaluation illustrates the impacts of aesthetic bias on selection.

A Cost-Effectiveness Analysis of Pegcetacoplan for the Treatment of Geographic Atrophy

To evaluate the cost-effectiveness of the treatment of geography atrophy (GA) with intravitreal pegcetacoplan and to identify utility measurement surrogates. cost analysis based on data from a published study SUBJECTS: none; based on data from published sham control compared to 2 treatment groups in the index study METHODS: Costs were based on 2022 Medicare reimbursement data. Specific outcomes were extrapolated from the DERBY and OAKS trials. Assumptions were made for the lifetime analysis based on a theoretical logistic growth model of the atrophy. cost, cost-utility, cost per QALY, cost per area of GA (in US$) RESULTS: The cost to treat GA in every month (EM) and every other month (EOM) treatment groups over the 2 years as reported was $70,000 and $34,600, respectively. The cost per area of delaying GA for 2 years in all patients was $87,300/mm2 (EM) and $49,200/mm2 (EOM), and in initially extrafoveal patients, $53,900/mm2 (EM) and $32,100/mm2 (EOM). The cost per day of delaying GA for 2 years was $295 (EM) and $170 (EOM); the marginal cost (EM vs EOM) per RPE cell saved was $30. The modeled lifetime costs were $350,000 (EM) and $172,000 (EOM), or $309,000/mm2 (EM) and $180,000 (EOM) /mm2 . The modeled time to 95% atrophy at 13 years was delayed by 2.5 years (EM) and 2.1 years (EOM). The costs/QALY gained based on modeled visual loss with 95% atrophy were $706,000 (EM) and $397,000 (EOM). Treatment of GA with intravitreal pegcetacoplan EOM was more cost effective than EM. Treatment of extrafoveal lesions yielded greater utility compared to the treatment of the entire group. As atrophy progression approaches an upper limit, the marginal cost/benefit ratios increase.

A cross-sectional study to evaluate utility measure and health-related quality of life (HRQoL) among patients with severe uncontrolled asthma in Spain

Background and purpose of the study The utility measure is a method to quantify health-related quality of life according to the preference values that patients attach to their health status. This study aimed to estimate the utility measure of patients with controlled and uncontrolled severe asthma (SA) in Spain, separately. Additionally, other characteristics (sociodemographic, clinical, and healthcare resource use [HCRU]) were also assessed for both SA populations. Methods This cross-sectional study included 159 patients with SA in Spain. Data were collected from medical records and directly from the patients during the study visit. Asthma Control Questionnaire (ACQ)-5 was used to classify patients with controlled and uncontrolled SA. Results Most of the patients were female (72.0% uncontrolled SA and 63.6% controlled SA). The mean (SD) EuroQol-5D (EQ-5D-5L) score was 0.88 (0.14) and 0.70 (0.25) in controlled and uncontrolled SA, respectively. The mean (SD) Asthma Quality-of-Life-5D (AQL-5D) score was 0.93 (0.09) and 0.85 (0.09) in controlled and uncontrolled SA, respectively. Emergency visits (19.2 vs. 2.7%) and hospitalizations (7.7% vs. no hospitalization) were more common among uncontrolled SA than controlled SA. Mean (SD) number of visits to primary care and pneumologists in uncontrolled SA vs. controlled SA was 4.1 (2.8) vs. 2.5 (3.0) and 3.7 (3.5) vs. 2.8 (2.2), respectively. Conclusion The study provides data on utility measures among patients with SA in Spain for the first time. Patients with uncontrolled SA had lower HRQoL and higher HCRU than patients with controlled SA. Therefore, the implementation of measures that improve HRQoL among patients with uncontrolled SA is highly recommended.

Cost-effectiveness of cardiovascular magnetic resonance imaging compared to common strategies in the diagnosis of coronary artery disease: a systematic review.

Cardiovascular magnetic resonance imaging (CMR) has established exceptional diagnostic utility and prognostic value in coronary artery disease (CAD). An assessment of the current evidence on the cost-effectiveness of CMR in patients referred for the investigation of CAD is essential for developing an economic model to evaluate the cost-effectiveness of CMR in CAD. We conducted a comprehensive search of multiple electronic databases, including PubMed, Scopus, Web of Science core collection, Embase, National Health Service Economic Evaluation Database (NHS EED), and health technology assessment, to identify relevant literature. After removing duplicates and screening the title/abstract, a total of 13 articles were deemed eligible for full-text assessment. We included studies that reported one or more of the following outcomes: incremental cost-effectiveness ratio (ICER), cost per quality-adjusted life year (QALYs), cost per life year gained, sensitivity and specificity rate as the primary outcome, and health utility measures or health-related quality of life as the secondary outcome. The quality of the included studies was assessed using the CHEERS 2022 guidelines. The findings of this study demonstrate that in patients undergoing urgent percutaneous coronary intervention, CMR over a one-year and lifetime horizon leads to higher quality-adjusted life years (QALYs) compared to current strategies in cases of multivessel disease. The systematic review indicates that the CMR-based strategy is more cost-effective when compared to standard methods such as single-photon emission computed tomography (SPECT), coronary computed tomography angiography (CCTA), and coronary angiography (CA) (CMR = $19,273, SPECT = $19,578, CCTA = $19,886, and immediate CA = $20,929). The results also suggest that the CMR strategy can serve as a cost-effective gatekeeping tool for patients at risk of obstructive CAD. A CMR-based strategy for managing patients with suspected CAD is more cost-effective compared to both invasive and non-invasive strategies, particularly in real-world patient populations with a low to intermediate prevalence of the disease.

Translation and Cultural Adaptation of Multi-Attribute Utility Instrument (MAUI) Indonesian Version of the 15D Questionnaire

Background: Diabetes mellitus is a chronic metabolic disorder that cannot be cured and progressively develop into a complication affecting the patient's health-related quality of life (HRQoL). The utility is a measure of preference-based HRQoL. Indirect utility measurements can be measured using the Multi-Attribute Utility Instrument (MAUI) instrument, one of which is the 15D questionnaire that has never been translated and adapted culturally in Indonesian. Objective: Translating and culturally adapting the 15D questionnaire to Indonesian diabetes mellitus patients. Methods: The translation process was carried out through the stages of forward translation, reconciliation, and backward translation involving two qualified translators and three experts in their field. The pilot test stage involved eight respondents consisting of 6 diabetes mellitus patients and two healthy individuals. Results: Problems in the linguistic validation process led to more conformity of word equivalents from the original to the target language. The agreement found was in the area of semantic equivalence, idiomatic equivalence, and experiential equivalence. The problem was resolved by reconciliation during the Focus Group Discussion, which translators and experts in their field attended and discussed with the original author to get equality of meaning in terms of language and culture. The Indonesian version of the 15D questionnaire tested on eight respondents showed results that were easy to understand and straightforward. Conclusion: The Indonesian version of the 15D questionnaire is valid from the linguistic and cultural adaptation stage. Further research is needed relating to the validation and reliability of the questionnaire.

On the Quality of Synthetic Generated Tabular Data

Class imbalance is a common issue while developing classification models. In order to tackle this problem, synthetic data have recently been developed to enhance the minority class. These artificially generated samples aim to bolster the representation of the minority class. However, evaluating the suitability of such generated data is crucial to ensure their alignment with the original data distribution. Utility measures come into play here to quantify how similar the distribution of the generated data is to the original one. For tabular data, there are various evaluation methods that assess different characteristics of the generated data. In this study, we collected utility measures and categorized them based on the type of analysis they performed. We then applied these measures to synthetic data generated from two well-known datasets, Adults Income, and Liar+. We also used five well-known generative models, Borderline SMOTE, DataSynthesizer, CTGAN, CopulaGAN, and REaLTabFormer, to generate the synthetic data and evaluated its quality using the utility measures. The measurements have proven to be informative, indicating that if one synthetic dataset is superior to another in terms of utility measures, it will be more effective as an augmentation for the minority class when performing classification tasks.

Protocol for a Prospective, Observational Cost-effectiveness Analysis of Returning Secondary Findings of Genome Sequencing for Unexplained Suspected Genetic Conditions

PurposeAlthough costly, genome-wide sequencing (GWS) detects an extensive range of variants, enhancing our ability to diagnose and assess risk for an increasing number of diseases. In addition to detecting variants related to the indication for testing, GWS can detect secondary variants in BRCA1, BRCA2, and other genes for which early intervention may improve health. As the list of secondary findings grows, there is increased demand for surveillance and management by multiple specialists, adding pressure to constrained health care budgets. Secondary finding testing is actively debated because some consider it opportunistic screening for future health risks that may not manifest. Given the economic implications of secondary finding testing and follow-up and its unproven clinical utility, the objective is to assess the incremental cost-effectiveness of secondary finding ascertainment per case detected and per unit of improved clinical utility in families of children with unexplained suspected genetic conditions undergoing clinical GWS. MethodsThose undergoing trio genome or exome sequencing are eligible for the study. Positive secondary finding index cases will be matched to negative controls (1:2) based on age group, primary result(s) type, and clinical indication. During the 2-year study, 71 cases and 142 matched controls are expected. Health service use will be collected in patients and 1 adult family member every 6 months. The per-child and per-dyad total cost will be determined by multiplying use of each resource by a corresponding unit price and summing all cost items. Costs will be estimated from the public and societal payer perspectives. The mean cost per child and per dyad for secondary finding–positive and secondary finding–negative groups will be compared statistically. If important demographic differences are observed between groups, ordinary least-squares regression, log transformation, or other nonparametric technique will be used to compare adjusted mean costs. The ratio of the difference in mean cost to the secondary finding yield will be used to estimate incremental cost-effectiveness. In secondary analyses, effectiveness will be estimated using the number of clinical management changes due to secondary findings or the Clinician-Reported Genetic Testing Utility Index (C-GUIDE) score, a validated measure of clinical utility. Sensitivity analysis will be undertaken to assess the robustness of the findings to variation in key parameters. ImplicationsThis study generates key evidence to inform clinical practice and funding allocation related to secondary finding testing. The inclusion of family members and a new measure of clinical utility represent important advancements in economic evaluation in genomics.

Saibot: A Differentially Private Data Search Platform

Recent data search platforms use ML task-based utility measures rather than metadata-based keywords, to search large dataset corpora. Requesters submit a training dataset, and these platforms search for augmentations ---join or union-compatible datasets---that, when used to augment the requester's dataset, most improve model (e.g., linear regression) performance. Although effective, providers that manage personally identifiable data demand differential privacy (DP) guarantees before granting these platforms data access. Unfortunately, making data search differentially private is nontrivial, as a single search can involve training and evaluating datasets hundreds or thousands of times, quickly depleting privacy budgets. We present Saibot , a differentially private data search platform that employs Factorized Privacy Mechanism (FPM), a novel DP mechanism, to calculate sufficient semi-ring statistics for ML over different combinations of datasets. These statistics are privatized once, and can be freely reused for the search. This allows Saibot to scale to arbitrary numbers of datasets and requests, while minimizing the amount that DP noise affects search results. We optimize the sensitivity of FPM for common augmentation operations, and analyze its properties with respect to linear regression. Specifically, we develop an unbiased estimator for many-to-many joins, prove its bounds, and develop an optimization to redistribute DP noise to minimize the impact on the model. Our evaluation on a real-world dataset corpus of 329 datasets demonstrates that Saibot can return augmentations that achieve model accuracy within 50--90% of non-private search, while the leading alternative DP mechanisms (TPM, APM, shuffling) are several orders of magnitude worse.

Exploring the association between time series features and forecasting by temporal aggregation using machine learning

When a forecast of the total value over several time periods ahead is required, forecasters are presented with two temporal aggregation (TA) approaches to produce required forecasts: i) aggregated forecast (AF) or ii) aggregate data using non-overlapping temporal aggregation (AD). Often, the recommendation is to aggregate data to a frequency relevant to the decision the eventual forecast will support and then produce the forecast. However, this might not be always the best choice and we argue that both AF and AD approaches may outperform each other in different situations. Moreover, there is a lack of evidence on what indicators may determine the superiority of each approach. We design and execute an empirical experiment framework to first explore the performance of these approaches using monthly time series of M4 competition dataset. We further turn the problem into a classification supervised learning by constructing a database consisting of features of each time series as predictor and model class labelled as AF/AD as response/outcome. We then build machine learning algorithms to investigate the association between time series features and the performance of AF and AD. Our findings suggest that both AF and AD approaches may not consistently generate accurate results for every individual series. AF is shown to be significantly better than AD for the monthly M4 time series, especially for longer horizons. We build several machine learning approaches using a set of extracted time series features as input to predict accurately whether AD or AF should be used. We find out that Random Forest (RF) is the most accurate approach in correctly classifying the outcome assessed both by statistical measures such as misclassification error, F-statistics, area under the curve, and a utility measure. The RF approach reveals that curvature, nonlinearity, seas_pacf, unitroot_pp, mean, ARCHM.LM, Coefficient of Variation, stability, linearity, and max_level_shif are among the most important features in driving the predictions of the model. Our findings indicate that the strength of trend, ARCH.LM, hurst, autocorrelation lag 1, unitroot_pp, and seas_pacf may favour AF approach, while lumpiness, entropy, nonlinearity, curvature, and strength of seasonality may increase the chance of AD performing better. We conclude the study by summarising the findings and present an agenda for further research.

A comparison of synthetic data approaches using utility and disclosure risk measures

Seongbin An, Trang Doan, Juhee Lee, Jiwoo Kim, Yong Jae Kim, Yunji Kim, Changwon Yoon, Sungkyu Jung, Dongha Kim, Sunghoon Kwon, Hang J Kim, Jeongyoun Ahn, Cheolwoo Park. Korean J Appl Stat 2023;36:141-66. https://doi.org/10.5351/KJAS.2023.36.2.141

Understanding the Clinical Utility of Genome Sequencing in Critically Ill Newborns

Diagnostic genome sequencing (GS) in newborns may have many benefits. More accurate diagnosis could spur the development of innovative genomic therapies. A precise diagnosis could help doctors and parents anticipate clinical problems and inform a family's future reproductive choices. However, the integration of GS into neonatal care remains associated with a variety of ethical controversies, including concerns about informed consent, about interpreting uncertain results, about resource allocation and whether access to genomic services could exacerbate health disparities, and about the effect of genome diagnostics on people with disabilities. There also remains significant uncertainty about which babies should be tested and when and how the potential benefits of GS ought to be measured. Probably related to these challenges, some payors have been reluctant to cover the cost of GS for critically ill newborns. Much of the reluctance appears to turn on questions about the clinical benefit associated with GS and whether and for whom GS will be cost-effective. These situations point to the urgent need for careful assessments of the clinical utility of GS in critically ill infants. In this paper, we critically examine the ways in which the clinical utility of GS has been evaluated in this patient population. We focus on "change of management" (COM), a widely used measure of clinical utility for diagnostic GS. We suggest that this measure is often ambiguous because not all COMs can be attributed to genomic results and because not all COMs lead to patient benefit. Finally, we suggest ways that measurement of clinical utility could be improved.