Tag: statistics

Implement Sci Commun. 2025 Mar 3;6(1):24. doi: 10.1186/s43058-025-00707-y.

ABSTRACT

BACKGROUND: A framework to evaluate implementation of Major System Change (MSC) in healthcare has been developed and applied to implementation of longer-term system changes. This was the first study to apply the five domains of the MSC framework to rapid healthcare system change. We aimed to: i) evaluate implementation of rapid MSC, using England COVID-19 remote home monitoring services as a case study and ii) consider whether and how the MSC framework can be applied to rapid MSC.

METHODS: A mixed-methods rapid evaluation in England, across 28 primary and secondary healthcare sites (October 2020-November 2021; data collection: 4 months). We conducted 126 interviews (5 national leads, 59 staff, 62 patients/carers) and surveyed staff (n = 292) and patients/carers (n = 1069). Service providers completed cost surveys. Aggregated and patient-level national datasets were used to explore enrolment, service use and clinical outcomes. The MSC framework was applied retrospectively. Qualitative data were analysed thematically to explore key themes within each MSC framework domain. Descriptive statistics and multivariate analyses were used to analyse experience, costs, service use and clinical outcomes.

RESULTS: Decision to change/Decision on model: Service development happened concurrently: i) early local development motivated by urgent clinical need, ii) national rollout using standard operating procedures, and iii) local implementation and adaptation. Implementation approach: Services were tailored to local needs to consider patient, staff, organisational and resource factors. Implementation outcomes: Patient enrolment was low (59% services <10%). Service models and implementation approaches varied substantially. Intervention outcomes: No associations found between services and clinical outcomes. Patient and staff experiences were generally positive. However, barriers to delivery and engagement were found; with some groups finding it harder to engage.

CONCLUSIONS: Low enrolment rates and substantial variation due to tailoring services to local contexts meant it was not possible to conclusively determine service effectiveness. Process outcomes indicated areas of improvement. The MSC framework can be used to analyse rapid MSC. Implementation and factors influencing implementation may differ to non-rapid contexts (e.g. less uniformity, more tailoring). Our mixed-methods approach could inform future evaluations of large-scale rapid and non-rapid MSC in a range of conditions and services internationally.

PMID:40033389 | DOI:10.1186/s43058-025-00707-y

Implement Sci Commun. 2025 Mar 3;6(1):24. doi: 10.1186/s43058-025-00707-y.

ABSTRACT

BACKGROUND: A framework to evaluate implementation of Major System Change (MSC) in healthcare has been developed and applied to implementation of longer-term system changes. This was the first study to apply the five domains of the MSC framework to rapid healthcare system change. We aimed to: i) evaluate implementation of rapid MSC, using England COVID-19 remote home monitoring services as a case study and ii) consider whether and how the MSC framework can be applied to rapid MSC.

METHODS: A mixed-methods rapid evaluation in England, across 28 primary and secondary healthcare sites (October 2020-November 2021; data collection: 4 months). We conducted 126 interviews (5 national leads, 59 staff, 62 patients/carers) and surveyed staff (n = 292) and patients/carers (n = 1069). Service providers completed cost surveys. Aggregated and patient-level national datasets were used to explore enrolment, service use and clinical outcomes. The MSC framework was applied retrospectively. Qualitative data were analysed thematically to explore key themes within each MSC framework domain. Descriptive statistics and multivariate analyses were used to analyse experience, costs, service use and clinical outcomes.

RESULTS: Decision to change/Decision on model: Service development happened concurrently: i) early local development motivated by urgent clinical need, ii) national rollout using standard operating procedures, and iii) local implementation and adaptation. Implementation approach: Services were tailored to local needs to consider patient, staff, organisational and resource factors. Implementation outcomes: Patient enrolment was low (59% services <10%). Service models and implementation approaches varied substantially. Intervention outcomes: No associations found between services and clinical outcomes. Patient and staff experiences were generally positive. However, barriers to delivery and engagement were found; with some groups finding it harder to engage.

CONCLUSIONS: Low enrolment rates and substantial variation due to tailoring services to local contexts meant it was not possible to conclusively determine service effectiveness. Process outcomes indicated areas of improvement. The MSC framework can be used to analyse rapid MSC. Implementation and factors influencing implementation may differ to non-rapid contexts (e.g. less uniformity, more tailoring). Our mixed-methods approach could inform future evaluations of large-scale rapid and non-rapid MSC in a range of conditions and services internationally.

PMID:40033389 | DOI:10.1186/s43058-025-00707-y

Microbiome. 2025 Mar 3;13(1):64. doi: 10.1186/s40168-025-02041-w.

ABSTRACT

BACKGROUND: The microbiome is a complex ecosystem of interdependent taxa that has traditionally been studied through cross-sectional studies. However, longitudinal microbiome studies are becoming increasingly popular. These studies enable researchers to infer taxa associations towards the understanding of coexistence, competition, and collaboration between microbes across time. Traditional metrics for association analysis, such as correlation, are limited due to the data characteristics of microbiome data (sparse, compositional, multivariate). Several network inference methods have been proposed, but have been largely unexplored in a longitudinal setting.

RESULTS: We introduce LUPINE (LongitUdinal modelling with Partial least squares regression for NEtwork inference), a novel approach that leverages on conditional independence and low-dimensional data representation. This method is specifically designed to handle scenarios with small sample sizes and small number of time points. LUPINE is the first method of its kind to infer microbial networks across time, while considering information from all past time points and is thus able to capture dynamic microbial interactions that evolve over time. We validate LUPINE and its variant, LUPINE_single (for single time point analysis) in simulated data and four case studies, where we highlight LUPINE’s ability to identify relevant taxa in each study context, across different experimental designs (mouse and human studies, with or without interventions, and short or long time courses). To detect changes in the networks across time and groups or in response to external disturbances, we used different metrics to compare the inferred networks.

CONCLUSIONS: LUPINE is a simple yet innovative network inference methodology that is suitable for, but not limited to, analysing longitudinal microbiome data. The R code and data are publicly available for readers interested in applying these new methods to their studies. Video Abstract.

PMID:40033386 | DOI:10.1186/s40168-025-02041-w

Microbiome. 2025 Mar 3;13(1):64. doi: 10.1186/s40168-025-02041-w.

ABSTRACT

BACKGROUND: The microbiome is a complex ecosystem of interdependent taxa that has traditionally been studied through cross-sectional studies. However, longitudinal microbiome studies are becoming increasingly popular. These studies enable researchers to infer taxa associations towards the understanding of coexistence, competition, and collaboration between microbes across time. Traditional metrics for association analysis, such as correlation, are limited due to the data characteristics of microbiome data (sparse, compositional, multivariate). Several network inference methods have been proposed, but have been largely unexplored in a longitudinal setting.

RESULTS: We introduce LUPINE (LongitUdinal modelling with Partial least squares regression for NEtwork inference), a novel approach that leverages on conditional independence and low-dimensional data representation. This method is specifically designed to handle scenarios with small sample sizes and small number of time points. LUPINE is the first method of its kind to infer microbial networks across time, while considering information from all past time points and is thus able to capture dynamic microbial interactions that evolve over time. We validate LUPINE and its variant, LUPINE_single (for single time point analysis) in simulated data and four case studies, where we highlight LUPINE’s ability to identify relevant taxa in each study context, across different experimental designs (mouse and human studies, with or without interventions, and short or long time courses). To detect changes in the networks across time and groups or in response to external disturbances, we used different metrics to compare the inferred networks.

CONCLUSIONS: LUPINE is a simple yet innovative network inference methodology that is suitable for, but not limited to, analysing longitudinal microbiome data. The R code and data are publicly available for readers interested in applying these new methods to their studies. Video Abstract.

PMID:40033386 | DOI:10.1186/s40168-025-02041-w

Microbiome. 2025 Mar 3;13(1):64. doi: 10.1186/s40168-025-02041-w.

ABSTRACT

BACKGROUND: The microbiome is a complex ecosystem of interdependent taxa that has traditionally been studied through cross-sectional studies. However, longitudinal microbiome studies are becoming increasingly popular. These studies enable researchers to infer taxa associations towards the understanding of coexistence, competition, and collaboration between microbes across time. Traditional metrics for association analysis, such as correlation, are limited due to the data characteristics of microbiome data (sparse, compositional, multivariate). Several network inference methods have been proposed, but have been largely unexplored in a longitudinal setting.

RESULTS: We introduce LUPINE (LongitUdinal modelling with Partial least squares regression for NEtwork inference), a novel approach that leverages on conditional independence and low-dimensional data representation. This method is specifically designed to handle scenarios with small sample sizes and small number of time points. LUPINE is the first method of its kind to infer microbial networks across time, while considering information from all past time points and is thus able to capture dynamic microbial interactions that evolve over time. We validate LUPINE and its variant, LUPINE_single (for single time point analysis) in simulated data and four case studies, where we highlight LUPINE’s ability to identify relevant taxa in each study context, across different experimental designs (mouse and human studies, with or without interventions, and short or long time courses). To detect changes in the networks across time and groups or in response to external disturbances, we used different metrics to compare the inferred networks.

CONCLUSIONS: LUPINE is a simple yet innovative network inference methodology that is suitable for, but not limited to, analysing longitudinal microbiome data. The R code and data are publicly available for readers interested in applying these new methods to their studies. Video Abstract.

PMID:40033386 | DOI:10.1186/s40168-025-02041-w

BMC Med Inform Decis Mak. 2025 Mar 3;25(1):106. doi: 10.1186/s12911-025-02927-7.

ABSTRACT

BACKGROUND: Duration of surgery (DOS) varies substantially for patients with hip and knee arthroplasty (HA/KA) and is a major risk factor for adverse events. We therefore aimed (1) to identify whether machine learning can predict DOS in HA/KA patients using retrospective data available before surgery with reasonable performance, (2) to compare whether machine learning is able to outperform multivariable regression in predictive performance and (3) to identify the most important predictor variables for DOS both in a multi- and single-hospital context.

METHODS: eXtreme Gradient Boosting (XGBoost) and multivariable linear regression were used for predictions. Both models were applied to both the whole dataset which included multiple hospitals (3,704 patients), and a single-hospital dataset (1,815 patients) of the hospital with the highest case-volumes of our sample. Data was split into training (75%) and test data (25%) for both datasets. Models were trained using 5-fold cross-validation (CV) on the training datasets and applied to test data for performance comparison.

RESULTS: On test data in the multi-hospital setting, the mean absolute error (MAE) was 12.13 min (HA) / 13.61 min (KA) for XGBoost. In the single-hospital analysis, performance on test data was MAE 10.87 min (HA) / MAE 12.53 min (KA) for XGBoost. Predictive ability of XGBoost was tended to be better than of regression in all setting, however not statistically significantly. Important predictors for XGBoost were physician experience, age, body mass index, patient reported outcome measures and, for the multi-hospital analysis, the hospital.

CONCLUSION: Machine learning can predict DOS in both a multi-hospital and single-hospital setting with reasonable performance. Performance between regression and machine learning differed slightly, however insignificantly, while larger datasets may improve predictive performance. The study found that hospital indicators matter in the multi-hospital setting despite controlling for various variables, highlighting potential quality differences between hospitals.

TRIAL REGISTRATION: The study was registered at the German Clinical Trials Register (DRKS) under DRKS00019916.

PMID:40033378 | DOI:10.1186/s12911-025-02927-7

BMC Med Inform Decis Mak. 2025 Mar 3;25(1):106. doi: 10.1186/s12911-025-02927-7.

ABSTRACT

BACKGROUND: Duration of surgery (DOS) varies substantially for patients with hip and knee arthroplasty (HA/KA) and is a major risk factor for adverse events. We therefore aimed (1) to identify whether machine learning can predict DOS in HA/KA patients using retrospective data available before surgery with reasonable performance, (2) to compare whether machine learning is able to outperform multivariable regression in predictive performance and (3) to identify the most important predictor variables for DOS both in a multi- and single-hospital context.

METHODS: eXtreme Gradient Boosting (XGBoost) and multivariable linear regression were used for predictions. Both models were applied to both the whole dataset which included multiple hospitals (3,704 patients), and a single-hospital dataset (1,815 patients) of the hospital with the highest case-volumes of our sample. Data was split into training (75%) and test data (25%) for both datasets. Models were trained using 5-fold cross-validation (CV) on the training datasets and applied to test data for performance comparison.

RESULTS: On test data in the multi-hospital setting, the mean absolute error (MAE) was 12.13 min (HA) / 13.61 min (KA) for XGBoost. In the single-hospital analysis, performance on test data was MAE 10.87 min (HA) / MAE 12.53 min (KA) for XGBoost. Predictive ability of XGBoost was tended to be better than of regression in all setting, however not statistically significantly. Important predictors for XGBoost were physician experience, age, body mass index, patient reported outcome measures and, for the multi-hospital analysis, the hospital.

CONCLUSION: Machine learning can predict DOS in both a multi-hospital and single-hospital setting with reasonable performance. Performance between regression and machine learning differed slightly, however insignificantly, while larger datasets may improve predictive performance. The study found that hospital indicators matter in the multi-hospital setting despite controlling for various variables, highlighting potential quality differences between hospitals.

TRIAL REGISTRATION: The study was registered at the German Clinical Trials Register (DRKS) under DRKS00019916.

PMID:40033378 | DOI:10.1186/s12911-025-02927-7

Int J Behav Nutr Phys Act. 2025 Mar 3;22(1):25. doi: 10.1186/s12966-025-01716-9.

ABSTRACT

BACKGROUND: The time spent physically active outside of school (e.g., extracurricular physical activity) is an important contributor to children’s total daily physical activity for health and well-being. Little is known about the opportunities available to children to engage in extracurricular physical activity from low- to middle-income countries. This study aims to answer the question: What are the main perceived barriers and facilitators of extracurricular physical activity among school-age children in Mexico?

METHODS: A multi-method cross-sectional study was performed. Six focus groups with children (aged 9-12 years), six focus groups with parents, 10 one-on-one interviews with parents, 12 interviews with teachers, and six interviews with head teachers were conducted across Campeche, Morelos, and Mexico State, Mexico. A questionnaire was applied to explore children’s physical activity frequency and preferences for time inside and outside of school. Qualitative data analyses were performed with inductive thematic analysis supported with NVivo software. Quantitative data were analysed with descriptive statistics using IBM SPSS 26.

RESULTS: Three main themes summarise the study’s findings: (1) how children spend their time outside of school, (2) the places that children access, and (3) the social environment for physical activity outside of the school. The data suggest that children in Mexico dedicate their spare time to screen, work, do housework, or perform unstructured physical activity mostly at home instead of playing sports or actively outdoors. Family support, enjoyment of physical activity, access to programs and facilities, time, living in a housing complex with open common areas, and mild weather were important facilitators identified. 69.4% of children engage in extracurricular physical activity, none of which was provided by schools. More children commute by walking than riding a bike to and from school. Children living inland spent three times more time at home compared to those in seafront areas.

CONCLUSIONS: Children rely on their families to partake in extracurricular structured physical activity. Policies targeting children’s health and well-being should include school-based extracurricular physical activity programs.

PMID:40033367 | DOI:10.1186/s12966-025-01716-9

BMC Health Serv Res. 2025 Mar 4;25(1):330. doi: 10.1186/s12913-025-12456-8.

ABSTRACT

BACKGROUND: To enable people with stroke to achieve an active everyday life under altered conditions, the development of self-management programs is essential to facilitate the process of change that individuals must undergo. To improve access to self-management, internet-based solutions have been proposed. The aim of this study was to evaluate the feasibility of a novel internet-based intervention, “Strategies for Empowering activities in Everyday Life” (SEE, version 1.0), for clients with stroke.

METHODS: This feasibility study had a preposttest design without a control group and utilized a mixed-method approach. Data were collected through study-specific forms, outcome assessments, interviews, and field notes. Descriptive statistics and content analysis were subsequently applied.

RESULTS: The study involved fifteen clients and staff at clinics in a hospital-based open-care rehabilitation setting. The results indicate that SEE is feasible for clients with stroke. When adopted as expected, SEE has the potential to empower self-management and enhance engagement, balance, and values in everyday activities. The study also indicates that SEE is feasible in terms of adherent delivery of dosage, acceptability, and value, as perceived by clients, occupational therapists, and clinic managers. However, adjustments are needed in the study design, in terms of recruitment strategies, the selection of assessor-based outcome assessment, and the evaluation of adherence. Additionally, the educational program for professionals should be enhanced to better support the implementation of SEE.

CONCLUSION: After the study design, intervention, and educational program are refined, SEE can be prepared for a pilot randomized controlled trial.

TRIAL REGISTRATION: clinicaltrails.gov NCT04588116, date of registration: 8th October 2020.

PMID:40033363 | DOI:10.1186/s12913-025-12456-8

BMC Nurs. 2025 Mar 3;24(1):236. doi: 10.1186/s12912-025-02839-x.

ABSTRACT

BACKGROUND: Needle-stick injuries (NSIs) pose a substantial occupational hazard, exposing healthcare professionals to potentially infectious diseases. Nurses’ clinical competence plays a crucial role in preventing and mitigating the incidence of NSIs. This study aimed to investigate the relationship between clinical competence and the incidence of NSIs, as well as the factors contributing to these injuries, among nurses in Fars Province, southern Iran, from March 2023 to May 2023.

METHODS: This descriptive cross-sectional study included 264 nurses selected through convenience sampling. All participants were employed in various departments of teaching hospitals in Fasa city, southern Iran, during the study period. Data were collected using a demographic questionnaire and a clinical competence questionnaire specifically developed for nurses. The demographic questionnaire captured variables such as age, gender, marital status, educational background, departmental assignment, work experience, and weekly working hours. The clinical competence questionnaire consisted of 55 items assessing seven dimensions: clinical care, leadership, legal and ethical performance, professional development, interpersonal relationships, education and coaching, and critical thinking and research aptitude. Statistical analyses were performed using SPSS software (version 16), employing the Chi-square test, Kruskal-Wallis test, and multiple logistic regression analysis. A significance level of p < 0.05 was applied to all tests.

RESULTS: The findings revealed that 39.4% of the participating nurses exhibited high clinical competence, 51.5% demonstrated moderate competence, and 9.1% were classified as having low competence. Statistical analysis indicated a significant association between clinical competence levels and needle-stick status (P = 0.002). Moreover, a significant difference was identified between clinical competence levels and the frequency of NSIs (P = 0.001). A logistic regression model was employed to assess the likelihood of NSIs based on demographic variables. The results showed that 178 participants (67.42%) had experienced needle-stick or sharp injuries within the preceding year. Among these, 63 males (35.3%) and 115 females (64.6%) reported such incidents. The highest incidence of needle-stick and sharp injuries occurred in the Operating Room (91.7%), followed by Dialysis (88.9%), Pediatrics (80%), Surgical Intensive Care (76.5%), Emergency (74.3%), Women’s Surgery (70%), Post-Cardiac Intensive Care (69.2%), Oncology (63.6%), Internal Medicine (59.1%), Surgery and Infectious Diseases (54.5%), Laboratory and Cardiac Intensive Care (52.9%), Men’s Surgery (50%), and the Psychiatric Ward (41.2%).

CONCLUSIONS: Considering that the majority of nurses working in hospitals exhibited moderate to low levels of clinical competence, it is recommended that hospitals implement an annual clinical competence assessment for nurses. Regular evaluations and targeted training programs can enhance nurses’ competence levels, thereby improving patient care quality and reducing the incidence of NSIs among healthcare providers. Additionally, specific strategies should be developed and implemented in medical centers and hospitals to mitigate the risk of NSIs across all hospital departments, particularly in high-risk areas such as operating rooms and dialysis units, where the prevalence of NSIs is significantly higher.

CLINICAL TRIAL NUMBER: Not applicable.

PMID:40033355 | DOI:10.1186/s12912-025-02839-x