Register-based Data Research Articles

Associations between the Abilitator, a self-reported measurement tool of work ability and functioning, and national register-based indicators of health and employment

ObjectivesThe Abilitator is a recently developed self-reported measurement tool for work ability and functioning of people in a weak labour market position. The aim of this study was to describe...

Association of nurse understaffing and limited nursing work experience with in-hospital mortality among patients: A longitudinal register-based study

BackgroundAlthough nurse understaffing and limited nursing work experience may affect hospital patients' risk of mortality, relatively little longitudinal patient-level evidence on these associations is available. Hospital administrative data could provide important information about the level of staffing, nurses' work experience and patient mortality over time. ObjectiveTo examine whether daily exposure to nurse understaffing and limited nursing work experience is associated with patient mortality, using patient-level data with different exposure time windows and accounting for several patient-related characteristics. MethodsThis longitudinal register-based study combined administrative data on patients (clinical database Auria) and employees (Titania® shift-scheduling) from one hospital district in Finland in 2013–2019, covering a total of 254,446 hospital stays in 40 units. We quantified nurse understaffing as the number of days with low nursing hours in relation to target hours (<90 % of the annual unit median), and limited work experience as the number of days with a low proportion of nurses with >3 years of in-hospital experience, and those aged over 25 (<90 % of the annual unit median). We used two survival model designs to analyze the associations between nurse understaffing and limited nursing work experience and the in-hospital mortality of the patients: we considered these exposures during the first days in hospital and as a cumulative proportion of days with suboptimal staffing during the first 30 days. ResultsIn total, 1.5 % (N = 3937) of the hospital stays ended in death. A 20 % increase in the proportion of days with nurse understaffing was associated with an increased, 1.05-fold mortality risk at the patient level (95 % confidence interval, 1.01–1.10). The cumulative proportion of days with limited nursing work experience, or the combination of nurse understaffing and limited work experience were not associated with increased risk of death among all patients. However, both indicators of limited nursing work experience were associated with an increased mortality risk among patients with comorbidities (HR 1.05, 95 % CI 1.02–1.08 and HR 1.05, 95 % CI 1.00–1.10, respectively). ConclusionsNurse understaffing was associated with a slight, but a potentially critical increase in patient in-hospital mortality. Limited nursing work experience was associated with increased in-hospital mortality in a subgroup of patients with comorbidities. Increased use of administrative data on planned and realized working hours could be a routine tool for reducing avoidable in-hospital mortality.

Risk factors for severe respiratory syncytial virus infection during the first year of life: development and validation of a clinical prediction model

Novel immunisation methods against respiratory syncytial virus (RSV) are emerging, but knowledge of risk factors for severe RSV disease is insufficient for optimal targeting of interventions against them. Our aims were to identify predictors for RSV hospital admission from registry-based data and to develop and validate a clinical prediction model to guide RSV immunoprophylaxis for infants younger than 1 year. In this model development and validation study, we studied all infants born in Finland between June 1, 1997, and May 31, 2020, and in Sweden between June 1, 2006, and May 31, 2020, along with the data for their parents and siblings. Infants were excluded if they died or were admitted to hospital for RSV within the first 7 days of life. The outcome was hospital admission due to RSV bronchiolitis during the first year of life. The Finnish study population was divided into a development dataset (born between June 1, 1997, and May 31, 2017) and a temporal hold-out validation dataset (born between June 1, 2017, and May 31, 2020). The development dataset was used for predictor discovery and selection in which we screened 1511 candidate predictors from the infants', parents', and siblings' data, and developed a logistic regression model with the 16 most important predictors. This model was then validated using the Finnish hold-out validation dataset and the Swedish dataset. In total, there were 1 124 561 infants in the Finnish development dataset, 130 352 infants in the Finnish hold-out validation dataset, and 1 459 472 infants in the Swedish dataset. In addition to known predictors such as severe congenital heart defects (adjusted odds ratio 2·89, 95% CI 2·28-3·65), we confirmed some less established predictors for RSV hospital admission, most notably oesophageal malformations (3·11, 1·86-5·19) and lower complexity congenital heart defects (1·43, 1·25-1·63). The prediction model's C-statistic was 0·766 (95% CI 0·742-0·789) in Finnish data and 0·737 (0·710-0·762) in Swedish validation data. The infants in the highest decile of predicted RSV hospital admission probability had 4·5 times higher observed risk compared with others. Calibration varied according to epidemic intensity. The model's performance was similar to a machine learning (XGboost) model using all 1511 candidate predictors (C-statistic in Finland 0·771, 95% CI 0·754-0·788). The prediction model showed clinical utility in decision curve analysis and in hypothetical number needed to treat calculations for immunisation, and its C-statistic was similar across different strata of parental income. The identified predictors and the prediction model can be used in guiding RSV immunoprophylaxis in infants, or as a basis for further immunoprophylaxis targeting tools. Sigrid Jusélius Foundation, European Research Council, Pediatric Research Foundation, and Academy of Finland.

Precarious employment at a young age and alcohol-related health problems

Abstract Background Precarious employment, a known determinant of poor health, is increasing in prevalence, particularly among young adults, where the long-term health consequences are not fully understood. The aim is to examine the effects of precarious employment at a young age on later alcohol-related health problems. Methods We used register-based data from the Swedish Work, Illness, and Labor-market Participation (SWIP) cohort to follow a cohort of 367 655 young adults aged 27 years between 2000-2003. Register data on labour market establishment (precarious employment, sub-standard employment, unemployment, and standard employment) was collected three years after graduation from school. The Swedish index of alcohol-related diagnoses was used to define the outcome, which was collected during a 26-year follow-up. Alcohol-related morbidity and mortality was combined into the outcome of alcohol-related health problems and collected from the National Patient Registers and Cause of Death Register. Cox Regression models were used to obtain hazard ratios (HR) with 95% confidence intervals (CI). Results The preliminary results suggest, after adjusting for important covariates (e.g. sex, parents highest level of socioeconomic status and prior mental and alcohol-related health problems), that young adults who were precariously employed three years after graduation from school were at an increased risk of alcohol-related health problems compared to young adults with standard employment (HR 1.49, 95% CI: 1.37, 1.61). Young adults in long-term unemployment had an even higher risk (HR 2.05, 95%CI: 1.89, 2.22). Conclusions In Sweden, a nationwide register-based study with a long-term follow-up suggests that young adults with lower labour market attachment when entering the labour market are at an increased risk of later alcohol-related health problems. Key messages • Exposure to precarious employment at labour market establishment can have long-term determinantal consequences on later well-being. • Policies targeting the transition from school to the workforce are needed to ensure decent employment conditions for the vulnerable group of young adults.

Mortality around the anniversary of a sibling’s death. Findings from Swedish register-data

Abstract Background Studies have found that the death of a sibling is associated with adverse health in bereaved siblings. Anniversaries may trigger grief reactions and contribute to an acute deterioration of health and are exogenous to shared intergenerational characteristics and may provide an indication of a causal effect of bereavement. Previous studies have found anniversary effects following the loss of a child and parent. We examined whether mortality increases among adult siblings around the anniversary following sibling death. Methods This case-crossover study used Swedish register-based longitudinal data from 1990 to 2016, based on the entire population. Participants included all adults aged 18 to 65 years who experienced sibling death and subsequently died. Conditional logistic regression was used to quantify the association between the anniversary (or preanniversary and postanniversary periods) and mortality, controlling for time-invariant confounding. All analyses were stratified by sex of the bereaved. Further analyses will also be stratified by socio-demographic characteristics such as the sex of the deceased sibling, time since sibling death, age, marital status and cause of sibling death. Results The results show that the anniversary of a siblings’ death is not associated with an increased mortality risk in bereaved siblings. We rather found somewhat reduced mortality risks around the anniversary dates in men (OR 0.78 95% CI 0.55-1.10 for the period ranging from 2 days before up to the anniversary) and women (OR 0.44 95% CI 0.20-0.97 for the anniversary). Further analysis will scrutinize whether anniversary effects are found when the analysis are stratified by socio-demographic characteristics and cause of sibling death. Conclusions Anniversaries following sibling death might not trigger grief to the same extent as anniversaries following the death of a parent or child. Sibling bereavement may still have health consequences although not during anniversaries. Key messages • Anniversaries following sibling death might not trigger grief to the same extent as child and parental bereavement. • Sibling bereavement may still have health consequences although not around anniversaries.

The use of employment and social services and rehabilitation after the onset of unemployment

Abstract Background Even though employment and social services and rehabilitation are available to the unemployed, there is little information about their service paths after the onset of unemployment. The aim was to follow different service spells using sequence analysis among those who became unemployed. Methods The register-based data included all 18-59-year-old individuals living in the City of Oulu, Finland, who became unemployed in 2017 (N = 8,935). For the 12 months after the onset of unemployment, eight states were constructed: rehabilitative work activities, mental health and substance abuse services, social work, rehabilitation, work trial, labour market training (LMT), several services, and no services. We used sequence analysis and clustering to identify groups with different service paths. Sociodemographic and health-related factors and later employment were examined using multinomial regression analysis. Results We identified eight clusters: 1) no services (57%), 2) short LMT (11%), 3) several and simultaneous services (6%), 4) mental health or substance abuse services (2%), 5) work trial (10%), 6) rapid LMT (6%), 7) intermediate LMT (5%), and 8) long LMT (4%). Persons in cluster 1 had a more advantaged background. Compared to them, clusters 3 and 5 were associated with being female and cluster 7 was associated with being male. All service clusters were associated with being under 50 years old. Clusters 3-8 were associated with primary education and clusters 3, 5, and 8 also with secondary education. All service clusters were associated with no employment days before the onset of unemployment. Clusters 3-5 were associated with chronic diseases. Later employment was highest among clusters 1, 6, and 8. Conclusions The unemployed can have very different service paths. Lower SES predicts the use of employment and social services and rehabilitation, but the paths also differ by sex and health. Duration and timing of LMT vary by sociodemographic characteristics. Key messages • Most of the unemployed do not use any employment or social services or rehabilitation. Those who do can have very different service paths. • Lower education predicts more frequent use of services. Paths vary by sex, age, and health.

Parental education, ADHD in family and ADHD diagnosis incidence in Finnish children and adolescents

Abstract Background Lower parental education is linked to higher incidence of attention-deficit hyperactivity disorder (ADHD) diagnosis in offspring, as is having ADHD in family, but these associations are rarely studied together. We examined whether parental education predicts offspring ADHD diagnosis differently for children and adolescents with or without ADHD family members. Methods We used Finnish total population register-based data on individuals born between 1994-2003 with non-migrant background. We followed 228,247 boys and 217,866 girls for first ADHD diagnosis, measured from health care or medication registers, in childhood (4-12) and adolescence (13-17), and identified ADHD for biological parents or full siblings. We used Poisson regression to examine differences in diagnoses separately by offspring sex and age group. Results Preliminary results show that boys are more often diagnosed for ADHD in childhood and girls in adolescence, but incidence is higher for boys in both age groups. Among boys and girls without ADHD family members, lower parental education is associated with higher incidence in both childhood and adolescence. Among those with ADHD family members, lower parental education predicts higher incidence in childhood, but in adolescence the differences by parental education even out for boys, and for girls, higher parental education predicts higher incidence. Conclusions While lower parental education predicts higher ADHD diagnosis incidence in general, for adolescents in ADHD families, higher parental education seems to increase the likelihood of getting diagnosed -particularly for girls. This suggests that adolescents in ADHD families with lower parental education may remain without ADHD diagnosis if they are not diagnosed as children. Support measures should be developed to ensure equal access to diagnosis and care irrespective of age, gender and family background. Key messages • Lower parental education predicts higher ADHD diagnosis incidence in general. Yet, adolescents in ADHD families with lower educated parents may remain underdiagnosed. • Support measures should be developed to guarantee equal access to ADHD diagnosis and care regardless of family background.

Use of outpatient healthcare services before and after the onset of unemployment

Abstract Background The onset of unemployment may entail adverse health effects and, thus, increase the need for healthcare services. On the other hand, the onset of unemployment may weaken access to the healthcare services, which may as such decrease the overall use of healthcare services. The aim was to examine the use of outpatient healthcare services in different sectors of healthcare before and after the onset of unemployment and to study whether job loss affected the use of these services. Methods Longitudinal individual-level register-based data was utilized on all individuals living in the City of Oulu, Finland, who became unemployed in 2017 (N = 1,999), their propensity matched controls (N = 1,999), and unmatched controls (N = 58,459) in a quasi-experimental design. Use of outpatient healthcare services was examined in one-month periods from 12 months before to 12 months after the onset of unemployment. Several socio-demographic factors, along with sickness and employment histories, were used for propensity score matching. Difference-in-differences analysis was used to measure the differences in the use of outpatient healthcare services between the unemployed and their matched controls. Results The use of healthcare services decreased significantly after the onset of unemployment. This was due to a decrease in the use of occupational healthcare services. No change related to job loss was observed in the use of public or private healthcare services. The number of healthcare visits increased again after the unemployment ended. Difference-in-differences analyses showed that compared to propensity score matched controls, becoming unemployed reduced the use of healthcare services. Conclusions When access to occupational healthcare services ceases, other healthcare services do not appear to fill the gap among those who become unemployed. Adequate healthcare services should be guaranteed to all population groups equally based on need, irrespective of the labour market status. Key messages • The overall use of outpatient healthcare services decreases significantly after job loss in Finland because of a decrease in the use of occupational healthcare services. • Adequate and easily accessible healthcare services should be guaranteed to all population groups equally based on need, irrespective of the labour market status or other socio-demographic factors.

Part-Time Workers’ Employment Trajectories by Length of Hours and Reason for Working Part-Time: An 8-Year Follow-Up Study

Using the Finnish Labour Force Surveys merged with register-based follow-up data, we analyzed how different characteristics of part-time work predict employees’ and entrepreneurs’ employment trajectories in an 8-year follow-up. We analyzed careers by the length of weekly working hours and the reason for part-time work, that is, childcare, studies, health, part-time pension, other voluntary choice, or if full-time work was not available (involuntary part-time). We applied sequence analysis to define work career clusters based on the continuum of spells spent in different labor market statuses, that is, in upper and lower white-collar, manual, or entrepreneurial employment, unemployment, studying, pensioned, or inactivity. According to the results, involuntary part-timers have a significantly higher probability of entering the unemployment trajectory than full-time workers. Those working part-time hours due to care responsibilities were also more likely to face frequent periods of unemployment, whereas part-time work combined with studies was associated with stable white-collar trajectories. Our results also show that weakened labor outcomes following marginal part-time jobs associate with disability retirement instead of unemployment later in time, most probably determined by ill health. Therefore, we suggest further studies to consider marginal part-time workers’ health as the determinant of weakening career outcomes. Overall, our results highlight the need to improve part-time working conditions, a concern that organizations like the OECD have also raised. This improvement could reduce the risk of unemployment, promote health, extend work careers, and consequently increase the employment rate.

Effect of chronic antiplatelet therapy on clinical outcomes of endovascular thrombectomy for treatment of acute ischemic stroke.

The objective of this study was to investigate the prognostic significance of chronic antiplatelet therapy (APT) usage in acute ischemic stroke (AIS) treated with endovascular thrombectomy (EVT). Long-term APT may enhance recanalization but may also predispose patients to an increased risk of hemorrhagic transformation. Weighted hospitalizations for anterior-circulation AIS treated with EVT were identified in a large United States claims-based registry. Baseline clinical characteristics and outcomes were compared between patients with and without chronic APT usage prior to admission. Multivariable logistic regression analysis was performed to assess adjusted associations between APT and study endpoints. This analysis identified 36,560 patients, of whom 8170 (22.3%) were on a chronic APT regimen prior to admission. These patients were older and demonstrated a higher burden of comorbid disease, but had similar stroke severity on presentation in comparison with those not on APT. On unadjusted analysis, patients with prior APT demonstrated higher rates of favorable outcomes (24.3% vs 21.5%, p < 0.001), lower rates of mortality (7.0% vs 10.1%, p < 0.001), and lower rates of any intracranial hemorrhage (ICH; 20.3% vs 24.2%, p < 0.001), but no difference in rates of symptomatic ICH (sICH). Following multivariable adjustment for baseline clinical characteristics including age, acute stroke severity, and comorbidity burden, prior APT was associated with favorable outcome (adjusted odds ratio [aOR] 1.21, 95% CI 1.17-1.24, p < 0.001) and a lower likelihood of mortality (aOR 0.73, 95% CI 0.70-0.77, p < 0.001), without an increased likelihood of ICH (any ICH aOR 0.84, 95% CI 0.81-0.87, p < 0.001; sICH aOR 0.92, 95% CI 0.82-1.03, p = 0.131). Retrospective evaluation of patients with AIS treated with EVT using registry-based data demonstrated an association of prior APT usage with favorable outcomes, without an increased risk of hemorrhagic transformation.

Contribution of compositional changes in the workforce to sickness absence trends in Finland

In this study, we assessed whether the long-term decrease in sickness absences in Finland is explained by observed and unobserved compositional changes in the workforce. Utilizing register-based panel data on Finnish wage earners aged 30–62, we examined the annual onset of compensated sickness absence (granted after 10 weekdays) in the period 2005–2016. We applied random effects models adjusting for changes in the observed sociodemographic and occupational characteristics of the study population. We also applied fixed effects models, with corrections of the estimates for cohort ageing, to additionally account for the unobserved time-invariant characteristics of the study population over the years. Of the observed characteristics, increasing educational level partly explained the decreasing trend in sickness absences, and the further contribution of the occupational class was weak. Additionally, accounting for unobserved individual characteristics further explained the decreasing trend in sickness absences among those aged 30–47 years and led to a reverse increasing trend among those aged 48–62 years irrespective of sex and employment sector. Particularly for those over 47 years old, the decrease in sickness absences appeared to be more strongly influenced by compositional changes in characteristics that are established before fully entering the labour market — such as educational level as well as unmeasured individual characteristics that remain unchanged after childhood and early adulthood — than in the work environment or other factors contributing at working age. Sickness absence trends fluctuated during economic cycles, which did not appear to be explained by immediate changes in the observed or unobserved characteristics. Different mechanisms are likely to explain long-term sickness absence trends and trends around economic cycles. Attempts to improve work ability and labour market inclusion in long-term should rely more on increasing educational levels among the workforce and on interventions carried out early during the life course.

Impact of developmental coordination disorder in childhood on educational outcomes in adulthood among neonatal intensive care recipients: a register-based longitudinal cohort study

ObjectivesDevelopmental coordination disorder (DCD) is related to poorer educational outcomes among children and adolescents. Evidence on this association into adulthood is lacking. Therefore, we aimed to investigate whether probable DCD...

Elevated risk of developing type 2 diabetes in people with a psychiatric disorder: What is the role of health behaviors and psychotropic medication?

AimsSeveral psychiatric disorders are linked with an increased risk of developing type 2 diabetes (T2D), but the mediating mechanisms are unclear. We aimed to investigate health behaviors, obesity, psychotropic medication use, and comorbidity as potential mediating mechanisms explaining these associations. MethodsWe combined data from a large population-based survey with register-based data and followed a sample of 250,013 Danes (≥16 years) for up to 8.9 years. We conducted mediation analyses investigating 10 potential mediators of the associations between psychiatric disorders and incident T2D. ResultsIndividuals with a substance use disorder, schizophrenia, mood disorder, neurotic disorder, eating disorder, or a personality disorder had a significantly higher risk of developing T2D. Organic disorders, intellectual disabilities, developmental and behavioral disorders were not associated with T2D-risk. For all psychiatric disorders significantly associated with T2D, the use of antidepressant medication had the largest proportional mediating effect on the association (13–32 %). ConclusionsUse of antidepressant medication had the largest contribution to the associations between psychiatric disorders and incident T2D. Future epidemiological studies and prevention studies should focus on optimizing the use of antidepressant medication with minimal side effects, and the promotion of health behaviors in individuals with a psychiatric disorder to prevent T2D.

Frailty and heart failure: State-of-the-art review.

At least half of all patients with heart failure (HF) are affected by frailty, a syndrome that limits an individual ability to recover from acute stressors. While frailty affects up to 90% of patients with HF with preserved ejection fraction, it is also seen in ~30-60% of patients with HF with reduced ejection fraction, with ~26% higher prevalence in women compared with men. The relationship between frailty and HF is bidirectional, with both conditions exacerbating the other. Frailty is further complicated by a higher prevalence of sarcopenia (by ~20%) in HF patients compared with patients without HF, which negatively affects outcomes. Several frailty assessment methods have been employed historically including the Fried frailty phenotype and Rockwood Clinical Frailty Scale to classify HF patients based on the severity of frailty; however, a validated HF-specific frailty assessment tool does not currently exist. Frailty in HF is associated with a poor prognosis with a 1.5-fold to 2-fold higher risk of all-cause death and hospitalizations compared to non-frail patients. Frailty is also highly prevalent in patients with worsening HF, affecting >50% of patients hospitalized for HF. Such patients with multiple readmissions for decompensated HF have markedly poor outcomes compared to younger, non-frail cohorts, and it is hypothesized that it may be due to major physical and functional limitations that limit recovery from an acute episode of worsening HF, a care aspect that has not been addressed in HF guidelines. Frail patients are thought to confer less benefit from therapeutic interventions due to an increased risk of perceived harm, resulting in lower adherence to HF interventions, which may worsen outcomes. Multiple studies report that <40% of frail patients are on guideline-directed medical therapy for HF, of which most are on suboptimal doses of these medications. There is a lack of evidence generated from randomized trials in this incredibly vulnerable population, and most current practice is governed by post hoc analyses of trials, observational registry-based data and providers' clinical judgement. The current body of evidence suggests that the treatment effect of most guideline-based interventions, including medications, cardiac rehabilitation and device therapy, is consistent across all age groups and frailty subgroups and, in some cases, may be amplified in the older, more frail population. In this review, we discuss the characteristics, assessment tools, impact on prognosis and impact on therapeutic interventions of frailty in patients with HF.

Prevalence and incidence of personality disorders among children and adolescents in Danish mental health services: a nationwide register study

A few epidemiological studies have examined personality disorders (PDs) among children and adolescents in secondary mental health services. This study aims to describe the prevalence and incidence of PDs among children and adolescents who have attended Danish child and adolescent psychiatric services (CAPS). Using register-based data, we studied all patients under the age of 18 years who were admitted to in- and outpatient CAPS (N = 115,121) in Denmark from 2007 to 2017. A total of 4952 patients were diagnosed with a PD during the study period. The mean prevalence was 859 patients per year, and the mean incidence was 274 patients per year, including an increased incidence and prevalence of borderline, anxious, and unspecified PDs over the decade. The number of patients diagnosed with PDs increased from 700 to 851 per year, but the proportion of patients with PDs compared to all psychiatric diagnoses decreased from 4.2% to 2.8% over the study period. The PD population had an older age (14.8 years vs. 11.3 years; p < 0.001), a higher likelihood of being female (74% vs. 44%; p < 0.001), and four times more contacts with the psychiatric emergency departments than other patients with a psychiatric diagnosis. Future studies should focus on (a) implementing further epidemiological studies in different countries; (b) tracking diagnostic practices to facilitate comparisons and provide feedback for training clinicians and raising awareness; and (c) estimating trajectories of PDs, including costs within the CAPS, to facilitate informed decision-making regarding the future organization and provision of services to these children, adolescents, and their families.

Use of outpatient healthcare services before and after the onset of unemployment: A register-based propensity score matched study from Finland.

The aim was to examine the use of outpatient healthcare services in different sectors of healthcare before and after the onset of unemployment and to study whether job loss affected the use of these services. Longitudinal individual-level register-based data was utilized on all individuals living in the City of Oulu, Finland, who became unemployed in 2017 (N = 1,999), their propensity matched controls (N = 1,999), and unmatched controls (N = 58,459) in a quasi-experimental design. Use of outpatient healthcare services was examined in one-month periods from 12 months before to 12 months after the onset of unemployment. Several socio-demographic factors, along with sickness and employment histories, were used for matching. Difference-in-differences analysis was used to measure the differences in the use of outpatient healthcare services between the unemployed and their matched controls. The use of health services decreased significantly after the onset of unemployment. This was due to a decrease in the use of occupational health services. No change related to job loss was observed in the use of public or private healthcare services. The number of healthcare visits increased again after the unemployment ended. Difference-in-differences analyses showed that compared to propensity score matched controls, becoming unemployed reduced the use of health services. When access to occupational healthcare services ceases, other health services do not appear to fill the gap among those who become unemployed. Adequate healthcare services should be guaranteed to all population groups equally based on need, irrespective of the labour market status.

Early parental death and its association with children’s mental and economic well-being in adulthood: a nationwide population-based register study

BackgroundThis study examined the association between early parental death and children’s subsequent mental health, years of schooling, and labour-market outcomes (ie, employment and earnings) in adulthood.MethodsWe used nationwide register-based data...

Machine learning using multimodal clinical, electroencephalographic, and magnetic resonance imaging data can predict incident depression in adults with epilepsy: A pilot study.

This study was undertaken to develop a multimodal machine learning (ML) approach for predicting incident depression in adults with epilepsy. We randomly selected 200 patients from the Calgary Comprehensive Epilepsy Program registry and linked their registry-based clinical data to their first-available clinical electroencephalogram (EEG) and magnetic resonance imaging (MRI) study. We excluded patients with a clinical or Neurological Disorders Depression Inventory for Epilepsy (NDDI-E)-based diagnosis of major depression at baseline. The NDDI-E was used to detect incident depression over a median of 2.4 years of follow-up (interquartile range [IQR] = 1.5-3.3 years). A ReliefF algorithm was applied to clinical as well as quantitative EEG and MRI parameters for feature selection. Six ML algorithms were trained and tested using stratified threefold cross-validation. Multiple metrics were used to assess model performances. Of 200 patients, 150 had EEG and MRI data of sufficient quality for ML, of whom 59 were excluded due to prevalent depression. Therefore, 91 patients (41 women) were included, with a median age of 29 (IQR = 22-44) years. A total of 42 features were selected by ReliefF, none of which was a quantitative MRI or EEG variable. All models had a sensitivity > 80%, and five of six had an F1 score ≥ .72. A multilayer perceptron model had the highest F1 score (median = .74, IQR = .71-.78) and sensitivity (84.3%). Median area under the receiver operating characteristic curve and normalized Matthews correlation coefficient were .70 (IQR = .64-.78) and .57 (IQR = .50-.65), respectively. Multimodal ML using baseline features can predict incident depression in this population. Our pilot models demonstrated high accuracy for depression prediction. However, overall performance and calibration can be improved. This model has promise for identifying those at risk for incident depression during follow-up, although efforts to refine it in larger populations along with external validation are required.

Long term outcomes and causal modelling of compulsory inpatient and outpatient mental health care using Norwegian registry data: Protocol for a controversies in psychiatry research project.

Compulsory mental health care includes compulsory hospitalisation and outpatient commitment with medication treatment without consent. Uncertain evidence of the effects of compulsory care contributes to large geographical variations and a controversy on its use. Some argue that compulsion can rarely be justified and should be reduced to an absolute minimum, while others claim compulsion can more frequently be justified. The limited evidence base has contributed to variations in care that raise issues about the quality/appropriateness of care as well as ethical concerns. To address the question whether compulsory mental health care results in superior, worse or equivalent outcomes for patients, this project will utilise registry-based longitudinal data to examine the effect of compulsory inpatient and outpatient care on multiple outcomes, including suicide and overall mortality; emergency care/injuries; crime and victimisation; and participation in the labour force and welfare dependency. By using the natural variation in health providers' preference for compulsory care as a source of quasi-randomisation we will estimate causal effects of compulsory care on short- and long-term trajectories. This project will provide valuable insights for service providers and policy makers in facilitating high quality clinical care pathways for a high risk population group.

O-012 Risk of epithelial ovarian tumors among women with polycystic ovary syndrome (PCOS): A nationwide population-based cohort study

Abstract Study question Is polycystic ovary syndrome (PCOS) associated with epithelial ovarian cancer and borderline ovarian tumors? Summary answer Overall, no association between PCOS and epithelial ovarian tumors was seen, however, we observed an increased risk of ovarian cancer among postmenopausal women with PCOS. What is known already Few studies have investigated the association between PCOS and epithelial ovarian tumors. In general, studies have reported conflicting results, possibly due to methodological differences and varying confounder adjustments. For ovarian cancer, most studies have reported no associations, while two studies reported an increased risk for women with PCOS compared to women without PCOS. For borderline ovarian tumors, most studies found no associations while one study observed an increased risk for serous borderline ovarian tumors among women with PCOS. Only a few studies have investigated the risk associated with PCOS according to menopausal status and histological subtypes of these tumors. Study design, size, duration This is a nationwide cohort study including all women born in Denmark between January 1, 1940 and December 31, 2000. The study cohort consisted of 1,7 million women after exclusion of women who emigrated, died, were diagnosed with cancer, or had bilateral oophorectomy before study start. Participants/materials, setting, methods The study cohorts for ovarian cancer analyses and borderline ovarian tumor analyses were linked to national registers to obtain information on PCOS diagnosis, cancer diagnosis, covariates, migration, and vital status. Cox proportional hazard regression models were used to calculate hazard ratios (HRs) and 95% confidence intervals (CI) for ovarian cancer and for borderline ovarian tumors, as well as separately for histological subtypes and according to menopausal status (landmark analysis at attained age of 51 years). Main results and the role of chance During a median 26.0 years of follow-up, 6,490 women were diagnosed with epithelial ovarian cancer and 2,990 women were diagnosed with borderline ovarian tumors. Overall, we found no marked associations for ovarian cancer and borderline ovarian tumors among women with PCOS compared to women without PCOS after adjustments for highest obtained level of education and obesity (HR 1.32 95% CI 0.80-2.20 and HR 1.59 95% CI 0.92-2.75, respectively). However, we found an increased risk of ovarian cancer among postmenopausal women (HR 2.18 95% CI 1.04-4.57). Lastly, we observed an increased risk for serous borderline ovarian tumors (HR 2.98 95% CI 1.64-5.42) for women with PCOS. Limitations, reasons for caution The low number of exposed cases, despite the large study population, is a main limitation. Although the register-based exposure data are highly valid, diagnosis of PCOS is challenging, and we are not able to account for changes in diagnostic practices over time. Wider implications of the findings The results from our study and previous studies should be taken into consideration when revising clinical guidelines for the management of long-term health outcomes for women with PCOS. Low statistical precision is an important limitation in our study and in previous studies, and larger studies with pooled analyses are warranted. Trial registration number not applicable