The research in this dissertation aims to optimise blood donation processes in the framework of the Dutch national blood bank Sanquin. The primary health risk for blood donors is iron deficiency,... Show moreThe research in this dissertation aims to optimise blood donation processes in the framework of the Dutch national blood bank Sanquin. The primary health risk for blood donors is iron deficiency, which is evaluated based on donors' hemoglobin and ferritin levels. If either of these levels are inadequate, donors are deferred from donation. Deferral due to low hemoglobin levels occurs on-site, meaning that donors have already traveled to the blood bank and then have to return home without donating, which is demotivating for the donor and inefficient for the blood bank. A large part of this dissertation therefore has the objective to develop a prediction model for donors' hemoglobin levels, based on historical measurements and donor characteristics.The prediction model that was developed reduces the deferral rate by approximately 60\% (from 3\% to 1\% for women, and from 1\% to 0.4\% for men), showing the potential of using data to enhance blood bank policy efficiency. Additionally, the model predictions were made explainable, providing the blood bank with insights into why specific predictions are made. These insights increase our understanding of the relationships between donor characteristics and hemoglobin levels. If this prediction model would be implemented in practice, the explanations could also be shared with the donor to help them understand why they are (not) invited to donate, which could also contribute to donor satisfaction and retention.In a collaborative effort with blood banks in Australia, Belgium, Finland and South Africa, the same prediction model was applied on data from each blood bank. Despite differences in blood bank policies and donor demographics, the models found similar associations with the predictor variables in all countries. Differences in performance could mostly be attributed to differences in deferral rates, with blood banks with higher deferral rates obtaining higher model accuracy.Beyond hemoglobin prediction models, additional research questions are explored. One study aims to identify determinants of ferritin levels in donors through repeated measurements, and linking these to environmental variables. Another study involves modeling the pharmacokinetics of antibodies in COVID-19 recovered donors, and finding relationships between patient characteristics, symptoms, and antibody levels over time.In summary, the research in this dissertation shows the potential within the wealth of data collected by blood banks. The proposed data-driven donation strategies not only decrease deferral rates but also increase donor retention and understanding. This comprehensive approach allows Sanquin to provide more personalised feedback to donors regarding their iron status, ultimately optimising the blood donation process and contributing to the overall efficacy of blood banking systems. Show less
In the Innovative Medicine's Initiative Applied Public-Private Research enabling OsteoArthritis Clinical Headway (IMI-APPROACH) knee osteoarthritis (OA) study, machine learning models were trained... Show moreIn the Innovative Medicine's Initiative Applied Public-Private Research enabling OsteoArthritis Clinical Headway (IMI-APPROACH) knee osteoarthritis (OA) study, machine learning models were trained to predict the probability of structural progression (s-score), predefined as >0.3 mm/year joint space width (JSW) decrease and used as inclusion criterion. The current objective was to evaluate predicted and observed structural progression over 2 years according to different radiographic and magnetic resonance imaging (MRI)-based structural parameters. Radiographs and MRI scans were acquired at baseline and 2-year follow-up. Radiographic (JSW, subchondral bone density, osteophytes), MRI quantitative (cartilage thickness), and MRI semiquantitative [SQ; cartilage damage, bone marrow lesions (BMLs), osteophytes] measurements were obtained. The number of progressors was calculated based on a change exceeding the smallest detectable change (SDC) for quantitative measures or a full SQ-score increase in any feature. Prediction of structural progression based on baseline s-scores and Kellgren-Lawrence (KL) grades was analyzed using logistic regression. Among 237 participants, around 1 in 6 participants was a structural progressor based on the predefined JSW-threshold. The highest progression rate was seen for radiographic bone density (39%), MRI cartilage thickness (38%), and radiographic osteophyte size (35%). Baseline s-scores could only predict JSW progression parameters (most P>0.05), while KL grades could predict progression of most MRI-based and radiographic parameters (P<0.05). In conclusion, between 1/6 and 1/3 of participants showed structural progression during 2-year follow-up. KL scores were observed to outperform the machine-learning-based s-scores as progression predictor. The large amount of data collected, and the wide range of disease stage, can be used for further development of more sensitive and successful (whole joint) prediction models. Trial Registration: Clinicaltrials.gov number NCT03883568. Show less
Objective: The aim of the study was to evaluate the oxygen saturation index (OSI) as an early predictor of clinical deterioration in infants with congenital diaphragmatic hernia (CDH). Methods: A... Show moreObjective: The aim of the study was to evaluate the oxygen saturation index (OSI) as an early predictor of clinical deterioration in infants with congenital diaphragmatic hernia (CDH). Methods: A single-center retrospective cohort study was conducted in consecutive infants with isolated CDH with continuous OSI measurements collected in the first 24 h after birth between June 2017 and July 2021. Outcomes of interest were pulmonary hypertension, extracorporeal membrane oxygenation (ECMO)-therapy, and mortality. We evaluated the discriminative values of the maximum OSI value and of mean OSI values with receiver operator characteristic (ROC) analysis and the area under the ROC curve. Results: In 42 infants with 49,473 OSI measurements, the median OSI was 5.0 (interquartile range 3.1-10.6). Twenty-seven infants developed pulmonary hypertension on a median of day 1 (1-1), of which 15 infants had an indication for ECMO-therapy, and 6 infants died. Maximum OSI values were associated with pulmonary hypertension, ECMO-therapy, and mortality. Mean OSI values had an acceptable discriminative ability for pulmonary hypertension and an excellent discriminative ability for ECMO-therapy and mortality. Although OSI measurements were not always present in the first hours after birth, we determined discriminative cut-offs for mean OSI values already in these first hours for pulmonary hypertension, the need for ECMO-therapy, and mortality. Conclusions: Continuous OSI evaluation is a promising modality to identify those infants at highest risk for clinical deterioration already in the first hours after birth. This provides an opportunity to tailor postnatal management based on the individual patient's needs. Show less
Introduction: Lymph node ratio (LNR) is an important prognostic factor of survival in patients with pancreatic ductal adenocarcinoma (PDAC). This study aimed to validate three LNR-based nomograms... Show moreIntroduction: Lymph node ratio (LNR) is an important prognostic factor of survival in patients with pancreatic ductal adenocarcinoma (PDAC). This study aimed to validate three LNR-based nomograms using an international cohort. Materials and methods: Consecutive PDAC patients who underwent upfront pancreatoduodenectomy from six centers (Europe/USA) were collected (2000-2017). Patients with metastases, R2 resection, missing LNR data, and who died within 90 postoperative days were excluded. The updated Amsterdam nomogram, the nomogram by Pu et al., and the nomogram by Li et al. were selected. For the validation, calibration, discrimination capacity, and clinical utility were assessed. Results: After exclusion of 176 patients, 10113 patients were included. Median overall survival (OS) of the cohort was 23 months (95% CI: 21-25). For the three nomograms, Kaplan-Meier curves showed significant OS diminution with increasing scores (p < 0.01). All nomograms showed good calibration (non-significant Hosmer-Lemeshow tests). For the Amsterdam nomogram, area under the ROC curve (AUROC) for 3-year OS was 0.64 and 0.67 for 5-year OS. Sensitivity and specificity for 3-year OS prediction were 65% and 59%. Regarding the nomogram by Pu et al., AUROC for 3- and 5-year OS were 0.66 and 0.70. Sensitivity and specificity for 3-year OS prediction were 68% and 53%. For the Li nomogram, AUROC for 3- and 5-year OS were 0.67 and 0.71, while sensitivity and specificity for 3-year OS prediction were 63% and 60%. Conclusion: The three nomograms were validated using an international cohort. Those nomograms can be used in clinical practice to evaluate survival after pancreatoduodenectomy for PDAC. (c) 2022 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/). Show less
Hassan, S.; Ramspek, C.L.; Ferrari, B.; Diepen, M. van; Rossio, R.; Knevel, R.; ... ; COVID-19 Network Working Grp 2022
Background: The coronavirus disease 2019 (COVID-19) presents an urgent threat to global health. Prediction models that accurately estimate mortality risk in hospitalized patients could assist... Show moreBackground: The coronavirus disease 2019 (COVID-19) presents an urgent threat to global health. Prediction models that accurately estimate mortality risk in hospitalized patients could assist medical staff in treatment and allocating limited resources. Aims: To externally validate two promising previously published risk scores that predict in-hospital mortality among hospitalized COVID-19 patients. Methods: Two prospective cohorts were available; a cohort of 1028 patients admitted to one of nine hospitals in Lombardy, Italy (the Lombardy cohort) and a cohort of 432 patients admitted to a hospital in Leiden, the Netherlands (the Leiden cohort). The endpoint was in-hospital mortality. All patients were adult and testedCOVID-19 PCR-positive. Model discrimination and calibration were assessed. Results: The C-statistic of the 4C mortality score was good in the Lombardy cohort (0.85, 95CI: 0.82-0.89) and in the Leiden cohort (0.87, 95CI: 0.80-0.94). Model calibration was acceptable in the Lombardy cohort but poor in the Leiden cohort due to the model systematically overpredicting the mortality risk for all patients. The C -sta-tistic of the CURB-65 score was good in the Lombardy cohort (0.80, 95CI: 0.75-0.85) and in the Leiden cohort (0.82, 95CI: 0.76-0.88). The mortality rate in the CURB-65 development cohort was much lower than the mortality rate in the Lombardy cohort. A similar but less pronounced trend was found for patients in the Leiden cohort. Conclusion: Although performances did not differ greatly, the 4C mortality score showed the best performance. However, because of quickly changing circumstances, model recalibration may be necessary before using the 4C mortality score. Show less
Background Visual impairment frequently occurs amongst older people. Therefore, the aim of this study was to investigate the predictive value of visual impairment on functioning, quality of life... Show moreBackground Visual impairment frequently occurs amongst older people. Therefore, the aim of this study was to investigate the predictive value of visual impairment on functioning, quality of life and mortality in people aged 85 years. Methods From the Leiden 85-plus Study, 548 people aged 85 years were eligible for this study. Visual acuity was measured at baseline by Early Treatment Diabetic Retinopathy Study charts (ETDRS). According to the visual acuity (VA) three groups were made, defined as no (VA > 0.7), moderate (0.5 <= VA <= 0.7) or severe visual impairment (VA < 0.5). Quality of life, physical, cognitive, psychological and social functioning were measured annually for 5 years. For mortality, participants were followed until the age of 95. Results At baseline, participants with visual impairment scored lower on physical, cognitive, psychological and social functioning and quality of life (p < 0.001). Compared to participants with no visual impairment, participants with moderate and severe visual impairment had an accelerated deterioration in basic activities of daily living (respectively 0.27-point (p = 0.017) and 0.35 point (p = 0.018)). In addition, compared to participants with no visual impairment, the mortality risk was 1.83 (95% CI 1.43, 2.35) for participants with severe visual impairment. Discussion In very older adults, visual impairment predicts accelerated deterioration in physical functioning. In addition, severely visually impaired adults had an increased mortality risk. A pro-active attitude, focussing on preventing and treating visual impairment could possibly contribute to the improvement of physical independence, wellbeing and successful aging in very old age. Show less
Large and complex data sets are increasingly available for research in critical care. To analyze these data, researchers use techniques commonly referred to as statistical learning or machine... Show moreLarge and complex data sets are increasingly available for research in critical care. To analyze these data, researchers use techniques commonly referred to as statistical learning or machine learning (ML). The latter is known for large successes in the field of diagnostics, for example, by identification of radiological anomalies. In other research areas, such as clustering and prediction studies, there is more discussion regarding the benefit and efficiency of ML techniques compared with statistical learning. In this viewpoint, we aim to explain commonly used statistical learning and ML techniques and provide guidance for responsible use in the case of clustering and prediction questions in critical care. Clustering studies have been increasingly popular in critical care research, aiming to inform how patients can be characterized, classified, or treated differently. An important challenge for clustering studies is to ensure and assess generalizability. This limits the application of findings in these studies toward individual patients. In the case of predictive questions, there is much discussion as to what algorithm should be used to most accurately predict outcome. Aspects that determine usefulness of ML, compared with statistical techniques, include the volume of the data, the dimensionality of the preferred model, and the extent of missing data. There are areas in which modern ML methods may be preferred. However, efforts should be made to implement statistical frameworks (e.g., for dealing with missing data or measurement error, both omnipresent in clinical data) in ML methods. To conclude, there are important opportunities but also pitfalls to consider when performing clustering or predictive studies with ML techniques. We advocate careful valuation of new data-driven findings. More interaction is needed between the engineer mindset of experts in ML methods, the insight in bias of epidemiologists, and the probabilistic thinking of statisticians to extract as much information and knowledge from data as possible, while avoiding harm. Show less
This special issue of Reproductive Sciences is focusing on ethnic health disparity and its impact on (fe)male reproduction. Indeed, studies regarding underlying mechanisms, interventions and... Show moreThis special issue of Reproductive Sciences is focusing on ethnic health disparity and its impact on (fe)male reproduction. Indeed, studies regarding underlying mechanisms, interventions and prognosis in reproduction are underexposed for the non-White male and female. Here, we call for documentation of race and ethnicity in the analysis and management of couples with recurrent pregnancy loss. Show less
Background: The aim of this study was to develop a prediction model for 10-year overall survival (OS) after resection of colorectal liver metastasis (CRLM) based on patient, tumour and treatment... Show moreBackground: The aim of this study was to develop a prediction model for 10-year overall survival (OS) after resection of colorectal liver metastasis (CRLM) based on patient, tumour and treatment characteristics.Methods: Consecutive patients after complete resection of CRLM were included from two centres (1992-2019). A prediction model providing 10-year OS probabilities was developed using Cox regression analysis, including KRAS, BRAF and histopathological growth patterns. Discrimination and calibration were assessed using cross-validation. A web-based calculator was built to predict individual 10-year OS probabilities.Results: A total of 4112 patients were included. The estimated 10-year OS was 30% (95% CI 29 -32). Fifteen patient, tumour and treatment characteristics were independent prognostic factors for 10-year OS; age, gender, location and nodal status of the primary tumour, disease-free interval, number and diameter of CRLM, preoperative CEA, resection margin, extrahepatic disease, KRAS and BRAF mutation status, histopathological growth patterns, perioperative systemic chemotherapy and hepatic arterial infusion pump chemotherapy. The discrimination at 10-years was 0.73 for both centres. A simplified risk score identified four risk groups with a 10-year OS of 57%, 38%, 24%, and 12%.Conclusions: Ten-year OS after resection of CRLM is best predicted with a model including 15 patient, tumour, and treatment characteristics. The web-based calculator can be used to inform patients. This model serves as a benchmark to determine the prognostic value of novel biomarkers. (C) 2022 The Author(s). Published by Elsevier Ltd. Show less
Objective: The aim of the study is to assess the effect of perioperative chemo-therapy (CTX) in patients with grade II-III extremity soft tissue sarcoma (eSTS) on overall survival (OS) and evaluate... Show moreObjective: The aim of the study is to assess the effect of perioperative chemo-therapy (CTX) in patients with grade II-III extremity soft tissue sarcoma (eSTS) on overall survival (OS) and evaluate whether the PERSARC prediction tool could identify patients with eSTS more likely to benefit from CTX.Methods: Patients (18-70 years) with primary high-grade eSTS surgically treated with cura-tive intent were included in the retrospective cohort study. The effect of any perioperative CTX and anthracycline + ifosfamide (AI)-based CTX on OS was investigated in three PERSARC-risk groups (high/intermediate/low). The PERSARC-risk groups were defined by the 33% and 66% quantile of the predicted 5-year OS of the study population equal to a 5-year OS of 65.8% and 79.8%, respectively. The effect of CTX on OS was investigated with weighted Kaplan-Meier curves and multivariable Cox models with an interaction between risk group and CTX.Results: This study included 5683 patients. The weighted Kaplan-Meier curves did not demonstrate a beneficial effect of any CTX and AI-based CTX on OS in the overall population. However, in the high PERSARC-risk group the 5-year OS of AI-based CTX was significantly better than no CTX (69.8% vs 59.0%, respectively, p Z 0.004) (HR 0.66, 95% CI 0.53-0.83).Conclusions: This study demonstrated a beneficial effect of AI-based CTX on OS in a selected group of high-risk patients with an absolute survival benefit of 11% as stratified by the PERSARC prediction tool. However, no beneficial effect of CTX on OS was found in the overall population of patients with primary high-grade eSTS younger than 70 years.(c) 2022 The Author(s). Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/). Show less
Objective: Prediction models for cardiovascular disease (CVD) mortality come from high-income countries, comprising laboratory measurements, not suitable for resource-limited countries. This study... Show moreObjective: Prediction models for cardiovascular disease (CVD) mortality come from high-income countries, comprising laboratory measurements, not suitable for resource-limited countries. This study aims to develop and validate a non-laboratory model to predict CVD mortality in a middle-income setting. Study design and setting: We used data of population aged 40-80 years from three cohort studies: Tehran Lipid and Glucose Study (n = 5160), Isfahan Cohort Study (n = 4350), and Golestan Cohort Study (n = 45,500). Using Cox proportional hazard models, we developed prediction models for men and women, separately. Cross-validation and bootstrapping procedures were applied. The models' discrimination and calibration were assessed by concordance statistic (C-index) and calibration plot, respectively. We calculated the models' sensitivity, specificity and net benefit fraction in a threshold probability of 5%. Results: The 10-year CVD mortality risks were 5.1% (95%CI: 4.8-5.5) in men and 3.1% (95%CI: 2.9%-3.3%) in women. The optimism-corrected performance of the model was c = 0.774 in men and c = 0.798 in women. The models showed good calibration in both sexes, with a predicted-to-observed ratio of 1.07 in men and 1.09 in women. The sensitivity was 0.76 in men and 0.66 in women. The net benefit fraction was higher in men compared to women (0.46 vs. 0.35). Conclusion: A low-cost model can discriminate well between low-and high-risk individuals, and can be used for screening in low-middle income countries. (C)& nbsp;2021 Elsevier Inc. All rights reserved. Show less
Rossen, T.M. van; Ooijevaar, R.E.; Vandenbroucke-Grauls, C.M.J.E.; Dekkers, O.M.; Kuijper, E.J.; Keller, J.J.; Prehn, J. van 2022
Objectives: Clostridioides difficile infection (CDI), its subsequent recurrences (rCDIs), and severe CDI (sCDI) provide a significant burden for both patients and the healthcare system. Identifying... Show moreObjectives: Clostridioides difficile infection (CDI), its subsequent recurrences (rCDIs), and severe CDI (sCDI) provide a significant burden for both patients and the healthcare system. Identifying patients diagnosed with initial CDI who are at increased risk of developing sCDI/rCDI could lead to more cost-effective therapeutic choices. In this systematic review we aimed to identify clinical prognostic factors associated with an increased risk of developing sCDI or rCDI.Methods: PubMed, Embase, Emcare, Web of Science and COCHRANE Library databases were searched from database inception through March, 2021. The study eligibility criteria were cohort and caseecontrol studies. Participants were patients >= 18 years old diagnosed with CDI, in which clinical or laboratory factors were analysed to predict sCDI/rCDI. Risk of bias was assessed by using the Quality in Prognostic Research (QUIPS) tool and the Grading of Recommendations Assessment, Development and Evaluation (GRADE) tool modified for prognostic studies. Study selection was performed by two independent reviewers. Overview tables of prognostic factors were constructed to assess the number of studies and the respective effect direction and statistical significance of an association.Results: 136 studies were included for final analysis. Greater age and the presence of multiple comorbidities were prognostic factors for sCDI. Identified risk factors for rCDI were greater age, healthcareassociated CDI, prior hospitalization, proton pump inhibitors (PPIs) started during or after CDI diagnosis, and previous rCDI.Conclusions: Prognostic factors for sCDI and rCDI could aid clinicians to make treatment decisions based on risk stratification. We suggest that future studies use standardized definitions for sCDI/rCDI and systematically collect and report the risk factors assessed in this review, to allow for meaningful metaanalysis of risk factors using data of high-quality trials. (C) 2021 The Author(s). Published by Elsevier Ltd on behalf of European Society of Clinical Microbiology and Infectious Diseases. Show less
Noorduyn, J.C.A.; Graaf, V.A. van de; Willigenburg, N.W.; Scholten-Peeters, G.G.M.; Mol, B.W.; Heymans, M.W.; ... ; ESCAPE Res Grp 2022
Purpose Marker-by-treatment analyses are promising new methods in internal medicine, but have not yet been implemented in orthopaedics. With this analysis, specific cut-off points may be obtained,... Show morePurpose Marker-by-treatment analyses are promising new methods in internal medicine, but have not yet been implemented in orthopaedics. With this analysis, specific cut-off points may be obtained, that can potentially identify whether meniscal surgery or physical therapy is the superior intervention for an individual patient. This study aimed to introduce a novel approach in orthopaedic research to identify relevant treatment selection markers that affect treatment outcome following meniscal surgery or physical therapy in patients with degenerative meniscal tears. Methods Data were analysed from the ESCAPE trial, which assessed the treatment of patients over 45 years old with a degenerative meniscal tear. The treatment outcome of interest was a clinically relevant improvement on the International Knee Documentation Committee Subjective Knee Form at 3, 12, and 24 months follow-up. Logistic regression models were developed to predict the outcome using baseline characteristics (markers), the treatment (meniscal surgery or physical therapy), and a marker-by-treatment interaction term. Interactions with p < 0.10 were considered as potential treatment selection markers and used these to develop predictiveness curves which provide thresholds to identify marker-based differences in clinical outcomes between the two treatments. Results Potential treatment selection markers included general physical health, pain during activities, knee function, BMI, and age. While some marker-based thresholds could be identified at 3, 12, and 24 months follow-up, none of the baseline characteristics were consistent markers at all three follow-up times. Conclusion This novel in-depth analysis did not result in clear clinical subgroups of patients who are substantially more likely to benefit from either surgery or physical therapy. However, this study may serve as an exemplar for other orthopaedic trials to investigate the heterogeneity in treatment effect. It will help clinicians to quantify the additional benefit of one treatment over another at an individual level, based on the patient's baseline characteristics. Show less
Fleuren, L.M.; Dam, T.A.; Tonutti, M.; Bruin, D.P. de; Lalisang, R.C.A.; Gommers, D.; ... ; Dutch ICU Data Sharing Covid-19 Co 2021
Introduction Determining the optimal timing for extubation can be challenging in the intensive care. In this study, we aim to identify predictors for extubation failure in critically ill patients... Show moreIntroduction Determining the optimal timing for extubation can be challenging in the intensive care. In this study, we aim to identify predictors for extubation failure in critically ill patients with COVID-19. Methods We used highly granular data from 3464 adult critically ill COVID patients in the multicenter Dutch Data Warehouse, including demographics, clinical observations, medications, fluid balance, laboratory values, vital signs, and data from life support devices. All intubated patients with at least one extubation attempt were eligible for analysis. Transferred patients, patients admitted for less than 24 h, and patients still admitted at the time of data extraction were excluded. Potential predictors were selected by a team of intensive care physicians. The primary and secondary outcomes were extubation without reintubation or death within the next 7 days and within 48 h, respectively. We trained and validated multiple machine learning algorithms using fivefold nested cross-validation. Predictor importance was estimated using Shapley additive explanations, while cutoff values for the relative probability of failed extubation were estimated through partial dependence plots. Results A total of 883 patients were included in the model derivation. The reintubation rate was 13.4% within 48 h and 18.9% at day 7, with a mortality rate of 0.6% and 1.0% respectively. The grandient-boost model performed best (area under the curve of 0.70) and was used to calculate predictor importance. Ventilatory characteristics and settings were the most important predictors. More specifically, a controlled mode duration longer than 4 days, a last fraction of inspired oxygen higher than 35%, a mean tidal volume per kg ideal body weight above 8 ml/kg in the day before extubation, and a shorter duration in assisted mode (< 2 days) compared to their median values. Additionally, a higher C-reactive protein and leukocyte count, a lower thrombocyte count, a lower Glasgow coma scale and a lower body mass index compared to their medians were associated with extubation failure. Conclusion The most important predictors for extubation failure in critically ill COVID-19 patients include ventilatory settings, inflammatory parameters, neurological status, and body mass index. These predictors should therefore be routinely captured in electronic health records. Show less
Ramspek, C.L.; Teece, L.; Snell, K.I.E.; Evans, M.; Riley, R.D.; Smeden, M. van; ... ; Diepen, M. van 2021
Background: External validation of prognostic models is necessary to assess the accuracy and generalizability of the model to new patients. If models are validated in a setting in which competing... Show moreBackground: External validation of prognostic models is necessary to assess the accuracy and generalizability of the model to new patients. If models are validated in a setting in which competing events occur, these competing risks should be accounted for when comparing predicted risks to observed outcomes. Methods: We discuss existing measures of calibration and discrimination that incorporate competing events for time-to-event models. These methods are illustrated using a clinical-data example concerning the prediction of kidney failure in a population with advanced chronic kidney disease (CKD), using the guideline-recommended Kidney Failure Risk Equation (KFRE). The KFRE was developed using Cox regression in a diverse population of CKD patients and has been proposed for use in patients with advanced CKD in whom death is a frequent competing event. Results: When validating the 5-year KFRE with methods that account for competing events, it becomes apparent that the 5-year KFRE considerably overestimates the real-world risk of kidney failure. The absolute overestimation was 10%age points on average and 29%age points in older high-risk patients. Conclusions: It is crucial that competing events are accounted for during external validation to provide a more reliable assessment the performance of a model in clinical settings in which competing risks occur. Show less
Maas, S.C.E.; Vidaki, A.; Teumer, A.; Costeira, R.; Wilson, R.; Dongen, J. van; ... ; Kayser, M. 2021
Background Information on long-term alcohol consumption is relevant for medical and public health research, disease therapy, and other areas. Recently, DNA methylation-based inference of alcohol... Show moreBackground Information on long-term alcohol consumption is relevant for medical and public health research, disease therapy, and other areas. Recently, DNA methylation-based inference of alcohol consumption from blood was reported with high accuracy, but these results were based on employing the same dataset for model training and testing, which can lead to accuracy overestimation. Moreover, only subsets of alcohol consumption categories were used, which makes it impossible to extrapolate such models to the general population. By using data from eight population-based European cohorts (N = 4677), we internally and externally validated the previously reported biomarkers and models for epigenetic inference of alcohol consumption from blood and developed new models comprising all data from all categories. Results By employing data from six European cohorts (N = 2883), we empirically tested the reproducibility of the previously suggested biomarkers and prediction models via ten-fold internal cross-validation. In contrast to previous findings, all seven models based on 144-CpGs yielded lower mean AUCs compared to the models with less CpGs. For instance, the 144-CpG heavy versus non-drinkers model gave an AUC of 0.78 +/- 0.06, while the 5 and 23 CpG models achieved 0.83 +/- 0.05, respectively. The transportability of the models was empirically tested via external validation in three independent European cohorts (N = 1794), revealing high AUC variance between datasets within models. For instance, the 144-CpG heavy versus non-drinkers model yielded AUCs ranging from 0.60 to 0.84 between datasets. The newly developed models that considered data from all categories showed low AUCs but gave low AUC variation in the external validation. For instance, the 144-CpG heavy and at-risk versus light and non-drinkers model achieved AUCs of 0.67 +/- 0.02 in the internal cross-validation and 0.61-0.66 in the external validation datasets. Conclusions The outcomes of our internal and external validation demonstrate that the previously reported prediction models suffer from both overfitting and accuracy overestimation. Our results show that the previously proposed biomarkers are not yet sufficient for accurate and robust inference of alcohol consumption from blood. Overall, our findings imply that DNA methylation prediction biomarkers and models need to be improved considerably before epigenetic inference of alcohol consumption from blood can be considered for practical applications. Show less
Background: Given the strong relationship between depression and anxiety, there is an urge to investigate their shared and specific long-term course determinants. The current study aimed to... Show moreBackground: Given the strong relationship between depression and anxiety, there is an urge to investigate their shared and specific long-term course determinants. The current study aimed to identify and compare the main determinants of the 9-year trajectories of combined and pure depression and anxiety symptom severity. Methods: Respondents with a 6-month depression and/or anxiety diagnosis (n=1,701) provided baseline data on 152 sociodemographic, clinical and biological variables. Depression and anxiety symptom severity assessed at baseline, 2-, 4-, 6- and 9-year follow-up, were used to identify data-driven course-trajectory subgroups for general psychological distress, pure depression, and pure anxiety severity scores. For each outcome (classprobability), a Superlearner (SL) algorithm identified an optimally weighted (minimum mean squared error) combination of machine-learning prediction algorithms. For each outcome, the top determinants in the SL were identified by determining variable-importance and correlations between each SL-predicted and observed outcome (rho pred) were calculated. Results: Low to high prediction correlations (rho pred: 0.41-0.91, median=0.73) were found. In the SL, important determinants of psychological distress were age, young age of onset, respiratory rate, participation disability, somatic disease, low income, minor depressive disorder and mastery score. For course of pure depression and anxiety symptom severity, similar determinants were found. Specific determinants of pure depression included several types of healthcare-use, and of pure-anxiety course included somatic arousal and psychological distress. Limitations: Limited sample size for machine learning. Conclusions: The determinants of depression- and anxiety-severity course are mostly shared. Domain-specific exceptions are healthcare use for depression and somatic arousal and distress for anxiety-severity course. Show less
Ramspek, C.L.; Steyerberg, E.W.; Riley, R.D.; Rosendaal, F.R.; Dekkers, O.M.; Dekker, F.W.; Diepen, M. van 2021
Etiological research aims to uncover causal effects, whilst prediction research aims to forecast an outcome with the best accuracy. Causal and prediction research usually require different methods,... Show moreEtiological research aims to uncover causal effects, whilst prediction research aims to forecast an outcome with the best accuracy. Causal and prediction research usually require different methods, and yet their findings may get conflated when reported and interpreted. The aim of the current study is to quantify the frequency of conflation between etiological and prediction research, to discuss common underlying mistakes and provide recommendations on how to avoid these. Observational cohort studies published in January 2018 in the top-ranked journals of six distinct medical fields (Cardiology, Clinical Epidemiology, Clinical Neurology, General and Internal Medicine, Nephrology and Surgery) were included for the current scoping review. Data on conflation was extracted through signaling questions. In total, 180 studies were included. Overall, 26% (n = 46) contained conflation between etiology and prediction. The frequency of conflation varied across medical field and journal impact factor. From the causal studies 22% was conflated, mainly due to the selection of covariates based on their ability to predict without taking the causal structure into account. Within prediction studies 38% was conflated, the most frequent reason was a causal interpretation of covariates included in a prediction model. Conflation of etiology and prediction is a common methodological error in observational medical research and more frequent in prediction studies. As this may lead to biased estimations and erroneous conclusions, researchers must be careful when designing, interpreting and disseminating their research to ensure this conflation is avoided. Show less
Hond, A. de; Raven, W.; Schinkelshoek, L.; Gaakeer, M.; Avest, E. ter; Sir, O.; ... ; Groot, B. de 2021
Objective: Early identification of emergency department (ED) patients who need hospitalization is essential for quality of care and patient safety. We aimed to compare machine learning (ML) models... Show moreObjective: Early identification of emergency department (ED) patients who need hospitalization is essential for quality of care and patient safety. We aimed to compare machine learning (ML) models predicting the hospitalization of ED patients and conventional regression techniques at three points in time after ED registration.Methods: We analyzed consecutive ED patients of three hospitals using the Netherlands Emergency Department Evaluation Database (NEED). We developed prediction models for hospitalization using an increasing number of data available at triage, similar to 30 min (including vital signs) and similar to 2 h (including laboratory tests) after ED registration, using ML (random forest, gradient boosted decision trees, deep neural networks) and multivariable logistic regression analysis (including spline transformations for continuous predictors). Demographics, urgency, presenting complaints, disease severity and proxies for comorbidity, and complexity were used as covariates. We compared the performance using the area under the ROC curve in independent validation sets from each hospital.Results: We included 172,104 ED patients of whom 66,782 (39 %) were hospitalized. The AUC of the multi-variable logistic regression model was 0.82 (0.78-0.86) at triage, 0.84 (0.81-0.86) at similar to 30 min and 0.83 (0.75-0.92) after similar to 2 h. The best performing ML model over time was the gradient boosted decision trees model with an AUC of 0.84 (0.77-0.88) at triage, 0.86 (0.82-0.89) at similar to 30 min and 0.86 (0.74-0.93) after similar to 2 h.Conclusions: Our study showed that machine learning models had an excellent but similar predictive performance as the logistic regression model for predicting hospital admission. In comparison to the 30-min model, the 2-h model did not show a performance improvement. After further validation, these prediction models could support management decisions by real-time feedback to medical personal. Show less
Background Notwithstanding the firmly established cross-sectional association of happiness with psychiatric disorders and their symptom severity, little is known about their temporal relationships.... Show moreBackground Notwithstanding the firmly established cross-sectional association of happiness with psychiatric disorders and their symptom severity, little is known about their temporal relationships. The goal of the present study was to investigate whether happiness is predictive of subsequent psychiatric disorders and symptom severity (and vice versa). Moreover, it was examined whether changes in happiness co-occur with changes in psychiatric disorder status and symptom severity. Methods In the Netherlands Study of Depression and Anxiety (NESDA), happiness (SRH: Self-Rated Happiness scale), depressive and social anxiety disorder (CIDI: Composite Interview Diagnostic Instrument) and depressive and anxiety symptom severity (IDS: Inventory of Depressive Symptomatology; BAI: Beck Anxiety Inventory; and FQ: Fear Questionnaire) were measured in 1816 adults over a three-year period. Moreover, we focused on occurrence and remittance of 6-month recency Major Depressive Disorder (MDD) and Social Anxiety Disorders (SAD) as the two disorders most intertwined with subjective happiness. Results Interindividual differences in happiness were quite stable (ICC of .64). Higher levels of happiness predicted recovery from depression (OR = 1.41; 95% CI = 1.10-1.80), but not social anxiety disorder (OR = 1.31; 95%CI = .94-1.81), as well as non-occurrence of depression (OR = 2.41; 95%CI = 1.98-2.94) and SAD (OR = 2.93; 95%CI = 2.29-3.77) in participants without MDD, respectively SAD at baseline. Higher levels of happiness also predicted a reduction of IDS depression (sr = - 0.08; 95%CI = -0.10 - -0.04), and BAI (sr = - 0.09; 95%CI = -0.12 - -0.05) and FQ (sr = - 0.06; 95%CI = -0.09 - -0.04) anxiety symptom scores. Conversely, presence of affective disorders, as well as higher depression and anxiety symptom severity at baseline predicted a subsequent reduction of self-reported happiness (with marginal to small sr values varying between -.04 (presence of SAD) to -.17 (depression severity on the IDS)). Moreover, changes in happiness were associated with changes in psychiatric disorders and their symptom severity, in particular with depression severity on the IDS (sr = - 0.46; 95%CI = -.50 - -.42). Conclusions Results support the view of rather stable interindividual differences in subjective happiness, although level of happiness is inversely associated with changes in psychiatric disorders and their symptom severity, in particular depressive disorder and depression severity. Show less