Multilevel models using logistic regression (MLogRM) and random forest models (RFM) are increasingly deployed in industry for the purpose of binary classification. The European Commission’s proposed Artificial Intelligence Act (AIA) necessitates, under certain conditions, that application of such models is fair, transparent, and ethical, which consequently implies technical assessment of these models. This paper proposes and demonstrates an audit framework for technical assessment of RFMs and MLogRMs by focussing on model-, discrimination-, and transparency & explainability-related aspects. To measure these aspects 20 KPIs are proposed, which are paired to a traffic light risk assessment method. An open-source dataset is used to train a RFM and a MLogRM model and these KPIs are computed and compared with the traffic lights. The performance of popular explainability methods such as kernel- and tree-SHAP are assessed. The framework is expected to assist regulatory bodies in performing conformity assessments of binary classifiers and also benefits providers and users deploying such AI-systems to comply with the AIA.
DOCUMENT
BACKGROUND: Recent evidence suggests that an increase in baccalaureate-educated registered nurses (BRNs) leads to better quality of care in hospitals. For geriatric long-term care facilities such as nursing homes, this relationship is less clear. Most studies assessing the relationship between nurse staffing and quality of care in long-term care facilities are US-based, and only a few have focused on the unique contribution of registered nurses. In this study, we focus on BRNs, as they are expected to serve as role models and change agents, while little is known about their unique contribution to quality of care in long-term care facilities. METHODS: We conducted a cross-sectional study among 282 wards and 6,145 residents from 95 Dutch long-term care facilities. The relationship between the presence of BRNs in wards and quality of care was assessed, controlling for background characteristics, i.e. ward size, and residents' age, gender, length of stay, comorbidities, and care dependency status. Multilevel logistic regression analyses, using a generalized estimating equation approach, were performed. RESULTS: 57% of the wards employed BRNs. In these wards, the BRNs delivered on average 4.8 min of care per resident per day. Among residents living in somatic wards that employed BRNs, the probability of experiencing a fall (odds ratio 1.44; 95% CI 1.06-1.96) and receiving antipsychotic drugs (odds ratio 2.15; 95% CI 1.66-2.78) was higher, whereas the probability of having an indwelling urinary catheter was lower (odds ratio 0.70; 95% CI 0.53-0.91). Among residents living in psychogeriatric wards that employed BRNs, the probability of experiencing a medication incident was lower (odds ratio 0.68; 95% CI 0.49-0.95). For residents from both ward types, the probability of suffering from nosocomial pressure ulcers did not significantly differ for residents in wards employing BRNs. CONCLUSIONS: In wards that employed BRNs, their mean amount of time spent per resident was low, while quality of care on most wards was acceptable. No consistent evidence was found for a relationship between the presence of BRNs in wards and quality of care outcomes, controlling for background characteristics. Future studies should consider the mediating and moderating role of staffing-related work processes and ward environment characteristics on quality of care.
DOCUMENT
Objective: A key aspect of psychiatric rehabilitation is supporting individuals with serious mental illness in reaching personal goals. This study aimed to investigate whether various aspects of the working alliance predict successful goal attainment and whether goal attainment improves subjective quality of life, independent of the ehabilitation approach used. Methods: Secondary analyses were conducted of data from a Dutch randomized clinical trial on goal attainment by individuals supported with the Boston University approach to psychiatric rehabilitation (N=80) or a generic approach (N=76). Working alliance was measured with the Working Alliance Inventory (WAI) from the practitioner’s perspective. Rehabilitation practitioners had backgrounds in social work, nursing, or vocational rehabilitation. Multiple logistic regression and multiple regression analyses explored effects of working alliance on goal attainment and of goal attainment on subjective quality of life at 24 months. Analyses were controlled for client- and process-related predictors, baseline quality of life, and rehabilitation approach. Results: The WAI goal subscale predicted goal attainment at 24 months. No effect was found for the bond or task subscale. Goal attainment significantly predicted quality of life at 24 months. These effects were independent of the rehabilitation approach used. Conclusions: A good bond between client and practitioner is not enough to attain successful rehabilitation outcomes. Findings suggest that it is important to discuss clients’ wishes and ambitions and form an agreement on goals. Attaining rehabilitation goals directly influenced the subjective quality of life of individuals with serious mental illness, which underscores the importance of investing in these forms of client support.
MULTIFILE
Among other things, learning to write entails learning how to use complex sentences effectively in discourse. Some research has therefore focused on relating measures of syntactic complexity to text quality. Apart from the fact that the existing research on this topic appears inconclusive, most of it has been conducted in English L1 contexts. This is potentially problematic, since relevant syntactic indices may not be the same across languages. The current study is the first to explore which syntactic features predict text quality in Dutch secondary school students’ argumentative writing. In order to do so, the quality of 125 argumentative essays written by students was rated and the syntactic features of the texts were analyzed. A multilevel regression analysis was then used to investigate which features contribute to text quality. The resulting model (explaining 14.5% of the variance in text quality) shows that the relative number of finite clauses and the ratio between the number of relative clauses and the number of finite clauses positively predict text quality. Discrepancies between our findings and those of previous studies indicate that the relations between syntactic features and text quality may vary based on factors such as language and genre. Additional (cross-linguistic) research is needed to gain a more complete understanding of the relationships between syntactic constructions and text quality and the potential moderating role of language and genre.
DOCUMENT
Background Variations in childbirth interventions may indicate inappropriate use. Most variation studies are limited by the lack of adjustments for maternal characteristics and do not investigate variations in adverse outcomes. This study aims to explore regional variations in the Netherlands and their correlations with referral rates, birthplace, interventions, and adverse outcomes, adjusted for maternal characteristics. Methods In this nationwide retrospective cohort study, using a national data register, intervention rates were analysed between twelve regions among single childbirths after 37 weeks’ gestation in 2010–2013 (n = 614,730). These were adjusted for maternal characteristics using multivariable logistic regression. Primary outcomes were intrapartum referral, birthplace, and interventions used in midwife- and obstetrician-led care. Correlations both between primary outcomes and between adverse outcomes were calculated with Spearman’s rank correlations. Findings Intrapartum referral rates varied between 55–68% (nulliparous) and 20–32% (multiparous women), with a negative correlation with receiving midwife-led care at the onset of labour in two-thirds of the regions. Regions with higher referral rates had higher rates of severe postpartum haemorrhages. Rates of home birth varied between 6–16% (nulliparous) and 16–31% (multiparous), and was negatively correlated with episiotomy and postpartum oxytocin rates. Among midwife-led births, episiotomy rates varied between 14–42% (nulliparous) and 3–13% (multiparous) and in obstetrician-led births from 46–67% and 14–28% respectively. Rates of postpartum oxytocin varied between 59–88% (nulliparous) and 50–85% (multiparous) and artificial rupture of membranes between 43–52% and 54–61% respectively. A north-south gradient was visible with regard to birthplace, episiotomy, and oxytocin. Conclusions Our study suggests that attitudes towards interventions vary, independent of maternal characteristics. Care providers and policy makers need to be aware of reducing unwarranted variation in birthplace, episiotomy and the postpartum use of oxytocin. Further research is needed to identify explanations and explore ways to reduce unwarranted intervention rates.
DOCUMENT
Objective: This study evaluated the effect of an after-school group-based medium-intensity multicomponent behavioural intervention programme for children aged 8–12 years classified as overweight, obese or at risk for overweight on body mass index standard deviation score (BMI SDS). In accordance with standardized protocols body weight and height were measured in 195 participants (88 boys, 107 girls) at baseline and at the end of the programme. A total of 166 children derived from a school-based monitoring system served as control group. Multivariate regression analyses examined the effect of the intervention and the independent factors associated with better outcomes in the intervention group. Results: Analysis of covariance showed a significant intervention effect on BMO SDS in favour of the intervention group (b-coefficient - 0.13 ± 0.03; p < 0.01) compared with the control group. Change in BMI SDS between baseline and follow-up in the intervention group was associated with baseline age (b-coefficient 0.03 ± 0.02; p=0.04) but was independent from gender, ethnicity, baseline BMI SDS, time between baseline and follow-up, school year and attendance rate.
DOCUMENT
Aggressive incidents occur frequently in health care facilities, such as psychiatric care and forensic psychiatric hospitals. Previous research suggests that civil psychiatric inpatients may display more aggression than forensic inpatients. However, there is a lack of research comparing these groups on the incident severity, even though both frequency and severity of aggression influence the impact on staff members. The purpose of this study is to compare the frequency and severity of inpatient aggression caused by forensic and civil psychiatric inpatients in the same Dutch forensic psychiatric hospital. Data on aggressive incidents occurring between January 1, 2014, and December 31, 2017, were gathered from hospital files and analyzed using the Modified Overt Aggression Scale, including sexual aggression (MOAS+). Multilevel random intercept models were used to analyze differences between forensic and civil psychiatric patients in severity of aggressive incidents. In all, 3,603 aggressive incidents were recorded, caused by 344 different patients. Civil psychiatric patients caused more aggressive incidents than forensic patients and female patients caused more inpatient aggression compared with male patients. Female forensic patients were found to cause the most severe incidents, followed by female civil psychiatric patients. Male forensic patients caused the least severe incidents. The findings have important clinical implications, such as corroborating the need for an intensive treatment program for aggressive and disruptive civil psychiatric patients, as well as emphasizing the importance of gender-responsive treatment
MULTIFILE
Introduction: In March 2014, the New South Wales (NSW) Government (Australia) announced the NSW Integrated Care Strategy. In response, a family-centred, population-based, integrated care initiative for vulnerable families and their children in Sydney, Australia was developed. The initiative was called Healthy Homes and Neighbourhoods. A realist translational social epidemiology programme of research and collaborative design is at the foundation of its evaluation. Theory and Method: The UK Medical Research Council (MRC) Framework for evaluating complex health interventions was adapted. This has four components, namely 1) development, 2) feasibility/piloting, 3) evaluation and 4) implementation. We adapted the Framework to include: critical realist, theory driven, and continuous improvement approaches. The modified Framework underpins this research and evaluation protocol for Healthy Homes and Neighbourhoods. Discussion: The NSW Health Monitoring and Evaluation Framework did not make provisions for assessment of the programme layers of context, or the effect of programme mechanism at each level. We therefore developed a multilevel approach that uses mixed-method research to examine not only outcomes, but also what is working for whom and why.
LINK
Background Running-related injuries (RRIs) can be considered the primary enemy of runners. Most literature on injury prediction and prevention overlooks the mental aspects of overtraining and under-recovery, despite their potential role in injury prediction and prevention. Consequently, knowledge on the role of mental aspects in RRIs is lacking. Objective To investigate mental aspects of overtraining and under-recovery by means of an online injury prevention programme. Methods and analysis The ‘Take a Mental Break!’ study is a randomised controlled trial with a 12 month follow-up. After completing a web-based baseline survey, half and full marathon runners were randomly assigned to the intervention group or the control group. Participants of the intervention group obtained access to an online injury prevention programme, consisting of a running-related smartphone application. This app provided the participants of the intervention group with information on how to prevent overtraining and RRIs with special attention to mental aspects. The primary outcome measure is any self-reported RRI over the past 12 months. Secondary outcome measures include vigour, fatigue, sleep and perceived running performance. Regression analysis will be conducted to investigate whether the injury prevention programme has led to a lower prevalence of RRIs, better health and improved perceived running performance. Ethics and dissemination The Medical Ethics Committee of the University Medical Center Utrecht, the Netherlands, has exempted the current study from ethical approval (reference number: NL64342.041.17). Results of the study will be communicated through scientific articles in peer-reviewed journals, scientific reports and presentations on scientific conferences.
DOCUMENT
Background: The environment affects children’s energy balance-related behaviors to a considerable extent. A context-based physical activity and nutrition school- and family-based intervention, named KEIGAAF, is being implemented in low socio-economic neighborhoods in Eindhoven, The Netherlands. The aim of this study was to investigate: 1) the effectiveness of the KEIGAAF intervention on BMI z-score, waist circumference, physical activity, sedentary behavior, nutrition behavior, and physical fitness of primary school children, and 2) the process related to the implementation of the intervention. Methods: A quasi-experimental, controlled study with eight intervention schools and three control schools was conducted. The KEIGAAF intervention consists of a combined top-down and bottom-up school intervention: a steering committee developed the general KEIGAAF principles (top-down), and in accordance with these principles, KEIGAAF working groups subsequently develop and implement the intervention in their local context (bottom-up). Parents are also invited to participate in a family-based parenting program, i.e., Triple P Lifestyle. Children aged 7 to 10 years old (grades 4 to 6 in the Netherlands) are included in the study. Effect evaluation data is collected at baseline, after one year, and after two years by using a child questionnaire, accelerometers, anthropometry, a physical fitness test, and a parent questionnaire. A mixed methods approach is applied for the process evaluation: quantitative (checklists, questionnaires) and qualitative methods (observations, interviews) are used. To analyze intervention effectiveness, multilevel regression analyses will be conducted. Content analyses will be conducted on the qualitative process data. Discussion: Two important environmental settings, the school environment and the family environment, are simultaneously targeted in the KEIGAAF intervention. The combined top-down and bottom-up approach is expected to make the intervention an effective and sustainable version of the Health Promoting Schools framework. An elaborate process evaluation will be conducted alongside an effect evaluation in which multiple data collection sources (both qualitative and quantitative) are used.
DOCUMENT
