Large-scale Implementation Research Articles

Hybrid 1 randomized controlled trial of an integrated stepped-care mental health intervention for traumatic injury patients

BackgroundAnnually, nearly 3 million individuals in the US are hospitalized after experiencing a traumatic injury (e.g., serious automobile crash, gunshot wound, stab injury). Many traumatically injured patients experience a trajectory of resilience. However, 20–40 % develop mental health problems such as posttraumatic stress disorder and depression; population estimates exceed 600,000 patients annually. Most trauma centers do not provide direct services to address mental health recovery, but the 2022 American College of Surgeons guidelines have established this as a priority. Cost-effective interventions are needed that meet the needs of patients at each stage of the recovery process while achieving sustainability at the level of implementation. This protocol paper describes a study that rigorously tests the Trauma Resilience and Recovery Program (TRRP), a scalable, sustainable technology-enhanced intervention to support the mental health recovery of patients who have experienced a traumatic injury. MethodsWe describe a randomized controlled trial with 1-year follow up of TRRP vs. enhanced usual care with 350 traumatically injured patients, including recruitment and retention procedures, assessment, implementation and fidelity monitoring, and statistical plans. ConclusionNovel components of our design include integration of technology-based elements, use of a stepped-care model, and implementation in a trauma center that did not previously have a mental health program. Data collected address the impact of TRRP and inform improvements to the model and its implementation in preparation for large-scale testing and implementation initiatives. This body of work is critical to informing the field as it continues to move toward national standards and recommendations.Trial registration:NCT05497115Clinicaltrials.gov

External validation of a deep learning model for automatic segmentation of skeletal muscle and adipose tissue on abdominal computed tomography images.

Body composition assessment using computed tomography (CT) images at the L3-level is increasingly applied in cancer research. Robust high-throughput automated segmentation is key to assess large patient cohorts and to support implementation of body composition analysis into routine clinical practice. We trained and externally validated a deep learning neural network (DLNN) to automatically segment L3-CT images. Expert-drawn segmentations of visceral and subcutaneous adipose tissue (VAT/SAT) and skeletal muscle (SM) of L3-CT-images of 3,187 patients undergoing abdominal surgery were used to train a DLNN. The external validation cohort was comprised of 2,535 patients with abdominal cancer. DLNN performance was evaluated with (geometric) Dice Similarity (DS) and Lin's Concordance Correlation Coefficient. There was a strong concordance between automatic and manual segmentations with median DS for SM, VAT, and SAT of 0.97 (interquartile range, IQR: 0.95-0.98), 0.98 (IQR: 0.95-0.98), and 0.95 (IQR: 0.92-0.97), respectively. Concordance correlations were excellent: SM 0.964 (0.959-0.968), VAT 0.998 (0.998-0.998), and SAT 0.992 (0.991-0.993). Bland-Altman metrics indicated only small and clinically insignificant systematic offsets; SM radiodensity: 0.23 hounsfield units (0.5%), SM: 1.26 cm2.m-2 (2.8%), VAT: -1.02 cm2.m-2 (1.7%), and SAT: 3.24 cm2.m-2 (4.6%). A robustly-performing and independently externally validated DLNN for automated body composition analysis was developed. CT-based body composition analysis is highly prognostic for long-term overall survival in oncology. This DLNN was succesfully trained and externally validated on several large patient cohorts and will therefore enable large scale population studies and implementation of body composition analysis into clinical practice.

Nanomaterials for Modified Asphalt and Their Effects on Viscosity Characteristics: A Comprehensive Review.

The application of nanomaterials as modifiers in the field of asphalt is increasingly widespread, and this paper aims to systematically review research on the impact of nanomaterials on asphalt viscosity. The results find that nanomaterials tend to increase asphalt's viscosity, enhancing its resistance to high-temperature rutting and low-temperature cracking. Zero-dimension nanomaterials firmly adhere to the asphalt surface, augmenting non-bonding interactions through van der Waals forces and engaging in chemical reactions to form a spatial network structure. One-dimensional nanomaterials interact with non-polar asphalt molecules, forming bonds between tube walls, thereby enhancing adhesion, stability, and resistance to cyclic loading. Meanwhile, these bundled materials act as reinforcement to transmit stress, preventing or delaying crack propagation. Two-dimensional nanomaterials, such as graphene and graphene oxide, participate in chemical interactions, forming hydrogen bonds and aromatic deposits with asphalt molecules, affecting asphalt's surface roughness and aggregate movement, which exhibit strong adsorption capacity and increase the viscosity of asphalt. Polymers reduce thermal movement and compact asphalt structures, absorbing light components and promoting the formation of a cross-linked network, thus enhancing high-temperature deformation resistance. However, challenges such as poor compatibility and dispersion, high production costs, and environmental and health concerns currently hinder the widespread application of nanomaterial-modified asphalt. Consequently, addressing these issues through comprehensive economic and ecological evaluations is crucial before large-scale practical implementation.

The Use of Triaxial Accelerometers and Machine Learning Algorithms for Behavioural Identification in Domestic Dogs (Canis familiaris): A Validation Study.

Assessing the behaviour and physical attributes of domesticated dogs is critical for predicting the suitability of animals for companionship or specific roles such as hunting, military or service. Common methods of behavioural assessment can be time consuming, labour-intensive, and subject to bias, making large-scale and rapid implementation challenging. Objective, practical and time effective behaviour measures may be facilitated by remote and automated devices such as accelerometers. This study, therefore, aimed to validate the ActiGraph® accelerometer as a tool for behavioural classification. This study used a machine learning method that identified nine dog behaviours with an overall accuracy of 74% (range for each behaviour was 54 to 93%). In addition, overall body dynamic acceleration was found to be correlated with the amount of time spent exhibiting active behaviours (barking, locomotion, scratching, sniffing, and standing; R2 = 0.91, p < 0.001). Machine learning was an effective method to build a model to classify behaviours such as barking, defecating, drinking, eating, locomotion, resting-asleep, resting-alert, sniffing, and standing with high overall accuracy whilst maintaining a large behavioural repertoire.

Mapping forest-based natural climate solutions

Natural climate solutions are critical actions of ecosystem stewardship to mitigate climate change. However, prioritizing locations and possible actions is challenging. We demonstrate a generalizable approach for identifying potential opportunities for natural climate solutions by creating a spatial hierarchy of land management restrictions. Global forest carbon stocks and flux models were then used to explore forest-based natural climate solutions in the high-carbon density coastal temperate rainforests of western North America. Our results show 13 million hectares are available for action, an area that holds 4,900 ± 640 megatonnes of carbon dioxide equivalent and represents 45% of regional and 0.5% of global aboveground forest carbon stocks. Based on historical trends, a 10% reduction in average annual forest carbon loss through improved forest management and conservation could reduce forest carbon emissions by 9.1 megatonnes of carbon dioxide equivalent per year, corresponding to 5.2% of the 2030 land-based climate commitments made by the United States and Canada. Large-scale implementation of natural climate solutions will require collaborative planning with forest-dependent communities, industry, governments, and Indigenous peoples.

User Experiences of Transitioning From a Homegrown Electronic Health Record to a Vendor-Based Product in the Department of Veterans Affairs: Qualitative Findings From a Mixed Methods Evaluation.

The Department of Veterans Affairs (VA), the largest nationally integrated health system in the United States, is transitioning from its homegrown electronic health record (EHR) to a new vendor-based EHR, Oracle Cerner. Experiences of the first VA site to transition have been widely discussed in the media, but in-depth accounts based on rigorous research are lacking. We sought to explore employee perspectives on the rationale for, and value of, transitioning from a VA-tailored EHR to a vendor-based product. As part of a larger mixed methods, multisite, formative evaluation of VA clinician and staff experiences with the EHR transition, we conducted semistructured interviews at the Mann-Grandstaff VA Medical Center before, during, and after going live in October 2020. In total, we completed 122 interviews with 26 participants across multiple departments. Before the new vendor-based EHR went live, participants initially expressed cautious optimism about the transition. However, in subsequent interviews following the go-live, participants increasingly critiqued the vendor's understanding of VA's needs, values, and workflows, as well as what they perceived as an inadequate fit between the functionalities of the new vendor-based EHR system and VA's characteristic approach to care. As much as a year after going live, participants reiterated these concerns while also expressing a desire for substantive changes to the transition process, with some questioning the value of continuing with the transition. VA's transition from a homegrown EHR to a vendor-based EHR system has presented substantial challenges, both practical and cultural in nature. Consequently, it is a valuable case study for understanding the sociotechnical dimension of EHR-to-EHR transitions. These findings have implications for both VA leadership and the broader community of policy makers, vendors, informaticists, and others involved in large-scale health information technology implementations.

A heat-measurement-free strategy for Economic Model Predictive Control of hydronic radiators

Cost-efficient measurement of room-level heat output from hydronic radiators is a major barrier to large-scale implementation of Economic Model Predictive Control (EMPC) in residential space heating for demand-side management. This paper therefore presents a novel EMPC strategy for hydronic radiators that relies on measurements of radiator pipe temperatures as a proxy for the radiator heat output, thus eliminating the need for costly flow-based meters at each radiator in a building. Simulation-based experiments indicate that the proposed proxy-based EMPC matches the performance of its heat-based counterpart. The proxy-based EMPC achieved a 16.6 % cost reduction compared to the heat-based EMPC's 16.8 %, with no comfort violations in both cases. Furthermore, the strategy shows resilience towards uncertainties in the user-estimated radiator exponent and maximum heating capacity. The proposed EMPC scheme also allows system operators to fine-tune the balance between cost savings and return temperatures using the proxy's upper limit. The findings presented in this paper suggest that the proposed proxy-based EMPC scheme provides a practical pathway for broader applications of EMPC in hydronic-based space heating with the prospect of unlocking significant load shifting potential, cost savings for end-users, and enhanced efficiency in individual and collective energy systems.

Hydrogen adsorption on fcc metal surfaces towards the rational design of electrode materials

The successful large-scale implementation of hydrogen as an energy vector requires high performance electrodes and catalysts made of abundant materials. Rational materials design strategies are the most efficient means of reaching this goal. Here we present a study on the adsorption of H-atoms onto fcc transition metal surfaces and propose descriptors for the rational design of electrodes and catalysts by means of correlations between fundamental properties of the materials and among other properties, their experimentally measured performance as hydrogen evolution electrodes (HEE). A large set of quantum mechanical modelling data at the DFT level was produced, covering the adsorption of H-atoms onto the most stable surfaces (100), (110) and (111) of: Ag, Au, Co, Cu, Ir, Ni, Pd, Pt and Rh. For each material and surface, a coverage dependent set of minimum energy structures was produced and chemical potentials for adsorption of H-atoms were obtained. Averaging procedures are here proposed to approach modelling to the experiments. Several correlations between the computed data and experimentally measured quantities are done to validate our methodology: surface plane dependent adsorption energies, chemical potentials and experimentally determined surface energies and work functions. We search for descriptors of catalytic activity by testing correlations between the DFT data obtained from our averaging procedures and experimental data on HEE performance. Our methodology allows us to obtain linear correlations between the adsorption energy of H-atoms and the exchange current density (i0) in a HEE, avoiding the volcano-like plots. We show that the chemical potential has limitations as a descriptor of i0 because it reaches an early plateau in terms of i0. Simple quantities obtained from database data such as the first stage electronegativity (χ) as devised by Mulliken has a strong linear correlation i0. With a quantity we denominate modified second-stage electronegativity (χ2m) we can reproduce the typical volcano plot in a correlation with i0. A theoretical and conceptual framework is presented. It shows that both χ and χ2m, that depend on the first ionization potential, second ionization potential and electron affinity of the elements can be used as descriptors in rational design of electrodes or of catalysts for hydrogen systems.

Assessment of a solar-powered trigeneration plant integrated with thermal energy storage using phase change materials

This study presents a comprehensive thermodynamic assessment of a trigeneration plant producing electricity, fresh water through multi-effect desalination (MED), and cooling through an absorption refrigeration cycle. The MED and absorption refrigeration systems utilize the rejected heat from the power cycle, driven by concentrated solar power (CSP). Situated in Qatar, the present system leverages the abundant solar irradiance to optimize the efficiency of electricity generation, water desalination, and cooling. The design features parabolic trough collectors with synthetic oil as the heat transfer fluid, direct thermal storage, and a Rankine steam turbine cycle with three turbine stages. The system also incorporates phase change materials (PCMs) based thermal energy storage (TES) to improve the system performance and offset the mismatch between demand and supply. The present system evaluation is based on energy and exergy analyses, while the Aspen Plus is used to simulate the power production and desalination operations, providing detailed insights into its efficiency and potential for large-scale implementation. The proposed system operates with an energy efficiency of 56.72 % and an exergy efficiency of 31.24 %, respectively.

California's Recovery Incentives Program: Implementation strategies

IntroductionDue to the increasing role of psychostimulants in the US drug poisoning crisis, there is an increasing need to effectively implement evidence-based treatment for individuals with stimulant use disorder (StimUD). Contingency management is a behavioral strategy with robust evidence of support for the treatment of StimUD. In 2023, California initiated a large-scale effort to implement CM as a treatment for individuals with a stimulant use disorder (cocaine, methamphetamine, amphetamine) called the Recovery Incentives Program: California's Contingency Management Benefit. MethodsThe Recovery Incentives Program is being systematically implemented using the Becker et al. Science to Service Lab (SSL) implementation approach with several augmentations for this project. The SSL features three core components: didactic training, performance feedback, and external facilitation. We have augmented this approach with a readiness assessment process for sites prior to CM service launch, and an ongoing fidelity monitoring and feedback component post-launch. ResultsThe present paper is a preliminary report describing the use of this augmented SSL strategy for CM implementation in a large-scale implementation effort. Data are presented to describe the implementation activities during the first ten months of the Recovery Incentives Program. ConclusionThe California Recovery Incentives Program has been systematically implemented and appears to be receiving a positive response from treatment program staff and enrolled members. Future papers and evaluation reports will continue to document member response to the Program and report on the ongoing training and implementation process.

Symptom vs context: lessons learned from a large-scale implementation of the Cultural Formulation Interview.

Mental health services in multicultural societies require culturally sensitive approaches to reduce health disparities. The Cultural Formulation Interview (CFI) is thought to enhance shared decision making and to facilitate culturally and contextually informed treatment. There is, however, little known regarding its implementability in large-scale psychiatric services. The current paper reports on (a) efforts to implement the CFI in a large organization for mental health services in the Netherlands, and (b) two studies that evaluated this implementation process and identified barriers to CFI adoption in clinical practice. Implementation of the CFI was facilitated by developing an online course, an advanced training of "team coaches", (digital) resources, and integrating the CFI questions into the intake format. A preliminary evaluation revealed that the CFI was administered in only 13.2% of all intakes across the entire organization, with minimal utilization of training resources. Study 1 aimed to investigate how the CFI was perceived by clinicians and stakeholders. A survey of 150 clinicians found a great lack of familiarity with the CFI and its purpose. While 67% reported partial CFI use, 50% saw no added value, and 61% deemed it relevant only for ethnic minorities. Study 2 examined which patient and clinician variables were associated with adequate CFI use (i.e., correct documentation of the CFI information in initial intake reports). The sample consisted of 112 intakes of patients conducted by ten clinicians. Regression analysis showed a significant association between clinicians' cultural competences and adequate CFI use, meaning that more culturally competent clinicians tended to generate better cultural assessments using the CFI. In addition, the CFI information was documented more adequately among patients who were unemployed at the time of assessment. In conclusion, implementation of the CFI requires a fundamental rethinking of the entire intake assessment, shifting it from a symptom-oriented approach towards a context- and person-centered one. Future trainings may benefit from embedding the CFI within a broader cultural competency training, rather than solely focusing on the CFI, which is currently the common practice.

Modeling and optimization for arrays of water turbine OWC devices

The large-scale implementation of wave energy conversion requires the installation of parks of devices. We study the problem of optimizing park layout and control for wave energy converters of the oscillating water column type. As a test case, we consider a device with a Wells turbine working in water. First, a novel model based on a nonlinear ordinary differential equation is derived to describe the behavior of the water column and used to estimate the power matrix. Then, its linearization is derived in order to enable the fast simulation of large parks of devices. The choice of the hydrodynamic model allows obtaining the gradient of the power with respect to the positions through an adjoint approach, making it especially convenient for optimization. We consider in particular the case of interaction with the piles of a floating wind energy plant. The results from the developed computational framework allow us to draw interesting conclusions about park layout design. In particular, we observe that mutual interaction effects can be significant even in parks made up of devices of small size, and that wave reflection from the piles of an offshore structure can improve energy production.

Availability of Evidence for Predictive Machine Learning Algorithms in Primary Care

The aging and multimorbid population and health personnel shortages pose a substantial burden on primary health care. While predictive machine learning (ML) algorithms have the potential to address these challenges, concerns include transparency and insufficient reporting of model validation and effectiveness of the implementation in the clinical workflow. To systematically identify predictive ML algorithms implemented in primary care from peer-reviewed literature and US Food and Drug Administration (FDA) and Conformité Européene (CE) registration databases and to ascertain the public availability of evidence, including peer-reviewed literature, gray literature, and technical reports across the artificial intelligence (AI) life cycle. PubMed, Embase, Web of Science, Cochrane Library, Emcare, Academic Search Premier, IEEE Xplore, ACM Digital Library, MathSciNet, AAAI.org (Association for the Advancement of Artificial Intelligence), arXiv, Epistemonikos, PsycINFO, and Google Scholar were searched for studies published between January 2000 and July 2023, with search terms that were related to AI, primary care, and implementation. The search extended to CE-marked or FDA-approved predictive ML algorithms obtained from relevant registration databases. Three reviewers gathered subsequent evidence involving strategies such as product searches, exploration of references, manufacturer website visits, and direct inquiries to authors and product owners. The extent to which the evidence for each predictive ML algorithm aligned with the Dutch AI predictive algorithm (AIPA) guideline requirements was assessed per AI life cycle phase, producing evidence availability scores. The systematic search identified 43 predictive ML algorithms, of which 25 were commercially available and CE-marked or FDA-approved. The predictive ML algorithms spanned multiple clinical domains, but most (27 [63%]) focused on cardiovascular diseases and diabetes. Most (35 [81%]) were published within the past 5 years. The availability of evidence varied across different phases of the predictive ML algorithm life cycle, with evidence being reported the least for phase 1 (preparation) and phase 5 (impact assessment) (19% and 30%, respectively). Twelve (28%) predictive ML algorithms achieved approximately half of their maximum individual evidence availability score. Overall, predictive ML algorithms from peer-reviewed literature showed higher evidence availability compared with those from FDA-approved or CE-marked databases (45% vs 29%). The findings indicate an urgent need to improve the availability of evidence regarding the predictive ML algorithms' quality criteria. Adopting the Dutch AIPA guideline could facilitate transparent and consistent reporting of the quality criteria that could foster trust among end users and facilitating large-scale implementation.

Environmental savings from concrete reuse: examining the limitations and optimal practices for cutting thresholds of concrete building components for reuse

Abstract Concrete, a widely used construction material, presents both opportunities and challenges for promoting environmentally responsible practices. This research focuses on the concept of concrete reuse as a sustainable strategy, examining the limitations of cutting dimensions. To investigate the greenhouse gas (GHG) emissions associated with preparing concrete for reuse, a mathematical model was developed considering several factors, such as transportation emissions, energy used in cutting, and the effects of varying cutting performance and energy sources. By incorporating real-life data and connecting with industrial practices, this study establishes a screening standard to determine the minimum cutting size for concrete reuse. The results indicate that the ratio of transportation distances for the reuse case versus the new production of concrete case has an influence on the minimum cutting size for concrete reuse. Moreover, the cutting size itself significantly impacts the overall GHG emissions of the reuse strategy. We offer practical insights for industry stakeholders to optimize concrete reuse practices and effectively reduce GHG emissions. As further research explores cost-effectiveness and on-site case studies, the potential for large-scale implementation of concrete reuse has become increasingly promising. Thus, concrete reuse has emerged as a viable and effective approach to sustainable construction, provided that the minimum cutting dimensions are carefully considered.

Simulation Evaluation of a Large-Scale Implementation of Virtual-Phase Link–Based Model Predictive Control

Simulation Evaluation of a Large-Scale Implementation of Virtual-Phase Link–Based Model Predictive Control

Mitigating risk of credit reversal in nature-based climate solutions by optimally anticipating carbon release

Nature-based climate solutions supply carbon credits generated from net carbon drawdown in exchange for project funding, but their credibility is challenged by the inherent variability and impermanence of drawdown. By evaluating drawdown benefits from a social cost of carbon perspective, project developers can enhance credibility and estimate impermanence by conservatively anticipating drawdowns to be eventually released following a release schedule, issuing additional credits when actual release is less severe than anticipated. We demonstrate how we can use ex post observations of drawdowns to construct optimal release schedules that limit the risk of credit reversals (when net drawdown is negative). We simulate both theoretical and real-life projects to examine how this approach balances the trade-off between generating credits evaluated as more permanent and limiting reversal risk. We discuss how this approach incentivizes project performance and provides a pragmatic solution to challenges facing larger-scale implementation of nature-based climate solutions.

Enhancing Preoperative Outcome Prediction: A Comparative Retrospective Case-Control Study on Machine Learning versus the International Esodata Study Group Risk Model for Predicting 90-Day Mortality in Oncologic Esophagectomy.

Risk prediction prior to oncologic esophagectomy is crucial for assisting surgeons and patients in their joint informed decision making. Recently, a new risk prediction model for 90-day mortality after esophagectomy using the International Esodata Study Group (IESG) database was proposed, allowing for the preoperative assignment of patients into different risk categories. However, given the non-linear dependencies between patient- and tumor-related risk factors contributing to cumulative surgical risk, machine learning (ML) may evolve as a novel and more integrated approach for mortality prediction. We evaluated the IESG risk model and compared its performance to ML models. Multiple classifiers were trained and validated on 552 patients from two independent centers undergoing oncologic esophagectomies. The discrimination performance of each model was assessed utilizing the area under the receiver operating characteristics curve (AUROC), the area under the precision-recall curve (AUPRC), and the Matthews correlation coefficient (MCC). The 90-day mortality rate was 5.8%. We found that IESG categorization allowed for adequate group-based risk prediction. However, ML models provided better discrimination performance, reaching superior AUROCs (0.64 [0.63-0.65] vs. 0.44 [0.32-0.56]), AUPRCs (0.25 [0.24-0.27] vs. 0.11 [0.05-0.21]), and MCCs (0.27 ([0.25-0.28] vs. 0.15 [0.03-0.27]). Conclusively, ML shows promising potential to identify patients at risk prior to surgery, surpassing conventional statistics. Still, larger datasets are needed to achieve higher discrimination performances for large-scale clinical implementation in the future.

Microbial-Derived Carotenoids and Their Health Benefits

Natural carotenoids (CARs) such as β-carotene, astaxanthin, lutein, norbixin, bixin, capsanthin, lycopene, β-Apo-8-carotenal, canthaxanthin, β-apo-8-carotenal-ester, and zeaxanthin are being explored for possible applications in feed, food, cosmeceuticals, and nutraceuticals. Three primary areas of carotenoid research are emerging: (1) encapsulations for improved chemical and physical properties; (2) natural source carotenoid manufacturing; and (3) preclinical, epidemiological, and clinical studies of carotenoids’ potential health benefits. The recent advancements in research on the chemistry and antioxidant activity, marketing strategies, dietary sources, bioavailability, and bioaccessibility, extraction, dietary consumption, encapsulating techniques, and health advantages of carotenoids are all extensively discussed in this review. Carotenoids are pigments found naturally in most fruits and vegetables, algae, plants, and photosynthetic bacteria. Carotenoids cannot be synthesized by humans and must be consumed in the form of food or supplements. There are several roles for carotenoids in human health. Although individual carotenoids may function in different ways, their main action is to act as antioxidants. There are validated techniques for separating and purifying carotenoids, yet, industrial production requires the development of economically viable techniques for larger-scale implementation. Carotenoids have been shown to boost cognitive performance and cardiovascular health, as well as help prevent some types of cancer. Despite evidence for carotenoids’ health benefits, major population-based supplementation trials have yielded conflicting outcomes for several carotenoids. This review includes recent developments in carotenoid metabolism and nutritional and health advantages. It also offers an outlook on future directions in these areas.

Enhancing maternal health service utilization among highly vulnerable pregnant women through a human-centered design process: Study protocol for a quasi-experimental study in Oromia, Ethiopia

Background Disproportionate uptake of and access to maternal and child health services remain significant challenges across and within countries. Differing geographic, economic, environmental, and social factors contribute to varying degrees of vulnerabilities among individuals, which manifest as disparities in maternal and newborn health outcomes. Designing solutions according to need is vital to improve maternal and child health outcomes. In this paper, we describe our study protocol on evaluating the effectiveness of an intervention package developed through a human-centered design (HCD) process in improving maternal health service uptake among vulnerable pregnant women in rural areas of Ethiopia. Methods The study has two distinct phases. For both phases, solutions will be developed leveraging HCD approaches to improve institutional delivery and antenatal care uptake among vulnerable pregnant women. Final solutions will be implemented in collaboration with community health workers and health officers. A community-based, quasi-experimental, mixed-method study design is used to assess differences between intervention and control arms. A panel sample will be enrolled after screening for pregnancy and vulnerability level and surveyed at baseline and midline in Phase 1. In Phase 2, a post-only study will be conducted in the same intervention and control sites. Pregnant women will receive refined solutions for six months, which will be evaluated using end-line assessment. Qualitative interviews will be conducted among vulnerable pregnant women and healthcare providers. Conclusions and Implications Our sequential approach to evaluating initial solutions, which in turn will be used to inform the enhancement of solutions, will provide practical insights into how solutions are accepted among vulnerable women and how they can be better integrated into women’s lives and health systems. Our findings on program effectiveness and implementation research will be disseminated to the Ethiopian Ministry of Health and its partners to inform large-scale implementation at the national level.

Evaluating the efficacy of machine learning in calibrating low-cost sensors

Ambient air quality monitoring requires low-cost environmental sensor devices that are affordable and feasible for large-scale implementation. However, issues such as sensor drift, environmental sensitivity, and inter-sensor variability affect data accuracy and cannot be adequately addressed by traditional calibration methods. This paper summarizes the use of machine learning techniques for calibrating low-cost sensors. The literature review shows that machine learning models like Random Forest, Support Vector Regression, and Neural Networks significantly improve sensor accuracy and reliability. For instance, Random Forest models reduced the root mean squared error by 30% for PM2.5 measurements, while Neural Networks achieved an R value of 0.997 for methane sensors. Integrating machine learning with IoT and mobile technologies enhances real-time monitoring and spatial resolution. Identified gaps include the quality of training datasets, managing environmental variability, and improving model transferability across different contexts. Addressing these gaps through advanced models and real-time calibration methodologies will further enhance sensor performance, ensuring more precise and reliable environmental data.