Remko Soer

Background: Occupational Health Check-ups (OHCs) have evolved from identifying work-related health risks to supporting sustainable employability through individualized lifestyle advice, but translating multidimensional demographic, occupational, and health characteristics into effective, personalized recommendations remains complex and lacks validated decision-support methods. Objective: Aims of this study are to: 1) develop a Machine Learning Decision Support Tool (ML-DST) based on workers’ demographic, occupational, and health characteristics to support personalized advice about lifestyle changes to improve sustainable employability, and 2) evaluate performance of the ML-DST. Methods: Real-world OHC data from a population-based cohort (2010–2020) were used to develop ML-DST models providing lifestyle advice (physical activity, smoking, alcohol, nutrition, relaxation, or no change) to improve workability and vitality. Paired baseline and 1–4-year follow-up data were used to train and evaluate 12 models. Per model, three rule-based algorithms were applied. Performance was considered useful when both True Positive Rate (TPR) and macroF1 score were ≥0.7. The best performing algorithm is proposed per model for inclusion in the final ML-DST. Results: Of 20,440 individual OHCs, 3412 paired sets were included. Best TPRs ranged from 0.51 to 0.71. One model achieved TPR and macroF1 scores ≥0.7, supporting nutritional behavior change to improve vitality; however, its profiles were not conceptually meaningful. The other 11 models did not achieve useful quantitative performance. Conclusion: Training algorithms for an ML-DST to support advice about lifestyle changes to improve long-term sustainable employability using data from regular Occupational Health Check-ups did not lead to useful model performance for improving workability and vitality.

Introduction: Chronic pain is a public health priority that affects 18% of Dutch and 16% of German adults, representing a major societal burden. The Ems Dollart Region (EDR), which forms the northern part of the Dutch–German border area, is particularly characterised by lower incomes, more physically demanding jobs and lower health literacy, which are associated with higher pain prevalence. Individuals with lower socioeconomic status face greater pain-related disability and reduced quality of life. Public health campaigns underpinned by pain science education provide a strategy to improve understanding of pain among the wider public by addressing common misconceptions about its causes and management. However, large-scale population-level interventions and evaluations of such campaigns remain limited. Methods and analysis: This protocol describes the design and implementation of a cross-border public health campaign in the EDR, aimed at improving public understanding of pain, changing attitudes and beliefs and promoting healthy behaviours. We will conduct a quasi-experimental before–after study in the EDR with a control group in Flanders (Belgium). The intervention is guided by behaviour change and biopsychosocial models. The campaign will deliver tailored content via social media, broadcast, print and online advertisements over a 24-month period. Two independent cross-sectional samples will be surveyed at baseline and at 2-year follow-up, comprising 3200 participants per measurement wave from three regions: the Netherlands and Germany as intervention regions, and Flanders, Belgium, as the control region. The primary outcome is the prespecified Pain Concepts Questionnaire ratio score, assessing pain-related knowledge, attitudes and beliefs. The primary analysis will use an adjusted difference-in-differences framework to estimate whether change in the primary outcome differs between the combined intervention regions and the control region. Secondary analyses will examine the Netherlands and Germany separately and will evaluate healthcare utilisation, medication use, health-related quality of life and work absenteeism. Ethics and dissemination: This study protocol was reviewed by the Medical Ethics Review Board of the University Medical Center Groningen and the University of Oldenburg. The study was considered not subject to the Medical Research Involving Human Subjects Act (non-WMO declaration; reference number M24.345555, issued on 31 December 2024) and was approved by the University of Oldenburg Ethics Committee (reference number 2025–056, issued on 07 March 2025). Informed consent was obtained electronically by the research agency at the time of participants’ registration with the research panel. Study findings, regardless of outcome, will be disseminated through peer-reviewed publications and presentations at national and international conferences.

Purpose: Chronic low back pain (CLBP) is a leading cause of disability worldwide, requiring coordinated referrals across different healthcare services. Despite clinical guidelines emphasising a biopsychosocial approach, referral decisions are often unstructured and poorly understood. Existing research has mainly focused on clinical management, with limited attention to the factors influencing referral decisions and their impact on referral pathways. The objective of this scoping review is to map the existing literature on factors influencing referral decisions for patients with non-specific CLBP and identify gaps for future research. Methods: Following PRISMA-ScR guidelines and Joanna Briggs Institute guidance for scoping reviews, a comprehensive literature search was conducted across the PubMed, EMBASE, and Web of Science databases. Studies focusing on referral decisions for patients with CLBP were included. Data extraction and descriptive synthesis were performed. Results: The search yielded 4,816 records, with 13 studies meeting inclusion criteria. Included studies, conducted between 1996 and 2025, spanned Europe, North America, and Israel, and comprised diverse designs including qualitative interviews, surveys, experimental vignettes, cross-sectional and cohort studies. Referral decisions in CLBP care were influenced by an interplay of clinical, psychosocial, patient-related, healthcare professional-related, systemic and contextual factors. Conclusion: This scoping review highlights the multifactorial nature of referral decisions in CLBP care and identifies critical gaps in current research. Future prospective studies are needed to better understand these factors and referral pathways, aiming to improve guideline adherence, patient outcomes, and equity in CLBP management.

Background: Posterior cervical foraminotomy (posterior surgery) is a valid alternative to anterior discectomy with fusion (anterior surgery) as a surgical treatment of cervical radiculopathy, but the quality of evidence has been limited. The purpose of this study was to compare the clinical outcome of these treatments after 2 years of follow-up. We hypothesized that posterior surgery would be noninferior to anterior surgery. Methods: This multicenter, randomized, noninferiority trial assessed patients with single-level cervical radiculopathy in 9 Dutch hospitals with a follow-up duration of 2 years. The primary outcomes measured reduction of cervical radicular pain and were the success ratio based on the Odom criteria, and arm pain and decrease in arm pain, evaluated with the visual analog scale, with a 10% noninferiority margin, which represents the maximum acceptable difference between the new treatment (posterior surgery) and the standard treatment (anterior surgery), beyond which the new treatment would be considered clinically unacceptable. The secondary outcomes were neck pain, Neck Disability Index, Work Ability Index, quality of life, complications (including reoperations), and treatment satisfaction. Generalized linear mixed effects modeling was used for analyses. The study was registered at the Overview of Medical Research in the Netherlands (OMON), formerly the Netherlands Trial Register (NTR5536). Results: From January 2016 to May 2020, 265 patients were randomized (132 to the posterior surgery group and 133 to the anterior surgery group). Among these, 25 did not have the allocated intervention; 11 of these 25 patients had symptom improvement, and the rest of the patients did not have the intervention due to various reasons. At the 2-year follow-up, of 243 patients, primary outcome data were available for 236 patients (97%). Predicted proportions of a successful outcome were 0.81 after posterior surgery and 0.74 after anterior surgery (difference in rate, −0.06 [1-sided 95% confidence interval (CI), −0.02]), indicating the noninferiority of posterior surgery. The between-group difference in arm pain was −2.7 (1-sided 95% CI, 7.4) and the between-group difference in the decrease in arm pain was 1.5 (1-sided 95% CI, 8.2), both confirming the noninferiority of posterior surgery. The secondary outcomes demonstrated small between-group differences. Serious surgery-related adverse events occurred in 9 patients (8%) who underwent posterior surgery, including 9 reoperations, and 11 patients (9%) who underwent anterior surgery, including 7 reoperations (difference in reoperation rate, −0.02 [2-sided 95% CI, −0.09 to 0.05]). Conclusions: This trial demonstrated that, after a 2-year follow-up, posterior surgery was noninferior to anterior surgery with regard to the success rate and arm pain reduction in patients with cervical radiculopathy. Level of Evidence: Therapeutic Level I. See Instructions for Authors for a complete description of levels of evidence.

BACKGROUND: Central sensitization cannot be demonstrated directly in humans. Therefore, studies used different proxy markers (signs, symptoms and tools) to identify factors assumed to relate to central sensitization in humans, that is, Human Assumed Central Sensitization (HACS). The aims of this systematic review were to identify non-invasive objective markers of HACS and the instruments to assess these markers in patients with fibromyalgia (FM). METHODS: A systematic review was conducted with the following inclusion criteria: (1) adults, (2) diagnosed with FM, and (3) markers and instruments for HACS had to be non-invasive. Data were subsequently extracted, and studies were assessed for risk of bias using the quality assessment tools developed by the National Institute of Health. RESULTS: 78 studies ([ ] 5234 participants) were included and the findings were categorized in markers identified to assess peripheral and central manifestations of HACS. The identified markers for peripheral manifestations of HACS, with at least moderate evidence, were pain after-sensation decline rates, mechanical pain thresholds, pressure pain threshold, sound ‘pressure’ pain threshold, cutaneous silent period, slowly repeated evoked pain sensitization and nociceptive flexion reflex threshold. The identified markers for central manifestations of HACS were efficacy of conditioned pain modulation with pressure pain conditioning and brain perfusion analysis. Instruments to assess these markers are: pin-prick stimulators, cuff-algometry, repetitive pressure stimulation using a pressure algometer, sound, electrodes and neuroimaging techniques. CONCLUSIONS: This review provides an overview of non-invasive markers and instruments for the assessment of HACS in patients with FM. Implementing these findings into clinical settings may help to identify HACS in patients with FM.

Background: Decision support systems (DSSs) for suggesting optimal treatments for individual patients with low back pain (LBP) are currently insufficiently accurate for clinical application. Most of the input provided to train these systems is based on patient-reported outcome measures. However, with the appearance of electronic health records (EHRs), additional qualitative data on reasons for referrals and patients’ goals become available for DSSs. Currently, no decision support tools cover a wide range of biopsychosocial factors, including referral letter information to help clinicians triage patients to the optimal LBP treatment. Objective: The objective of this study was to investigate the added value of including qualitative data from EHRs and referral letters to the accuracy of a quantitative DSS for patients with LBP. Methods: A retrospective study was conducted in a clinical cohort of Dutch patients with LBP. Patients filled out a baseline questionnaire about demographics, pain, disability, work status, quality of life, medication, psychosocial functioning, comorbidity, history, and duration of pain. Referral reasons and patient requests for help (patient goals) were extracted via natural language processing (NLP) and enriched in the data set. For decision support, these data were considered independent factors for triage to neurosurgery, anesthesiology, rehabilitation, or minimal intervention. Support vector machine, k-nearest neighbor, and multilayer perceptron models were trained for 2 conditions: with and without consideration of the referral letter content. The models’ accuracies were evaluated via F1-scores, and confusion matrices were used to predict the treatment path (out of 4 paths) with and without additional referral parameters. Results: Data from 1608 patients were evaluated. The evaluation indicated that 2 referral reasons from the referral letters (for anesthesiology and rehabilitation intervention) increased the F1-score accuracy by up to 19.5% for triaging. The confusion matrices confirmed the results. Conclusions: This study indicates that data enriching by adding NLP-based extraction of the content of referral letters increases the model accuracy of DSSs in suggesting optimal treatments for individual patients with LBP. Overall model accuracies were considered low and insufficient for clinical application.

Background: Short-term neck pain after posterior cervical foraminotomy (posterior surgery) compared with anterior cervical discectomy with fusion (anterior surgery) treating cervical radiculopathy has only been assessed once, retrospectively, to our knowledge. The aim of this study was to prospectively evaluate the course of neck pain for 6 weeks after both treatments. Methods: This is a secondary analysis of the multicenter Foraminotomy ACDF Cost-Effectiveness Trial (FACET), conducted from January 2016 to May 2020. Of 389 patients who had single-level, 1-sided cervical radiculopathy and were screened for eligibility, 265 were randomly assigned to undergo posterior surgery (n = 132) or anterior surgery (n = 133). The primary outcome of the present analysis was neck pain, assessed weekly for 6 weeks using the visual analog scale (VAS), on a scale of 0 to 100. The secondary outcomes were arm pain, neck disability, work ability, quality of life, treatment satisfaction, motor and sensory changes, and hospital length of stay. Data were analyzed with mixed model analysis in intention-to-treat samples using 2-sided 95% confidence intervals (CIs). Results: In the first postoperative week, the mean VAS for neck pain was 56.2 mm (95% CI, 51.7 to 60.8 mm) after posterior surgery and 46.7 mm (95% CI, 42.2 to 51.2 mm) after anterior surgery. The mean between-group difference was 9.5 mm (95% CI, 3.3 to 15.7 mm), which gradually decreased to 2.3 mm (95% CI, −3.6 to 8.1 mm) at postoperative week 6. As of postoperative week 5, there was no significant difference between groups. Responder analyses confirmed this result. Secondary outcomes showed small differences between groups. Conclusions: Insight into the course of neck pain during the first 6 weeks after posterior compared with anterior surgery is provided. Despite initially more neck pain after posterior surgery, patients swiftly improved and, as of postoperative week 5, results similar to those after anterior surgery were observed. Our findings should enable improved patient counseling and enhanced shared decision-making between physicians and patients with cervical radiculopathy, where more neck pain in the first postoperative weeks should be balanced against the benefits of posterior surgery. Level of Evidence: Therapeutic Level I. See Instructions for Authors for a complete description of levels of evidence.

BACKGROUND Decision support systems (DSSs) for suggesting optimal treatments for individual patients with low back pain (LBP) are currently insufficiently accurate for clinical application. Most of the input provided to train these systems is based on patient-reported outcome measures. However, with the appearance of electronic health records (EHRs), additional qualitative data on reasons for referrals and patients’ goals become available for DSSs. Currently, no decision support tools cover a wide range of biopsychosocial factors, including referral letter information to help clinicians triage patients to the optimal LBP treatment. OBJECTIVE The objective of this study was to investigate the added value of including qualitative data from EHRs and referral letters to the accuracy of a quantitative DSS for patients with LBP. METHODS A retrospective study was conducted in a clinical cohort of Dutch patients with LBP. Patients filled out a baseline questionnaire about demographics, pain, disability, work status, quality of life, medication, psychosocial functioning, comorbidity, history, and duration of pain. Referral reasons and patient requests for help (patient goals) were extracted via natural language processing (NLP) and enriched in the data set. For decision support, these data were considered independent factors for triage to neurosurgery, anesthesiology, rehabilitation, or minimal intervention. Support vector machine, k-nearest neighbor, and multilayer perceptron models were trained for 2 conditions: with and without consideration of the referral letter content. The models’ accuracies were evaluated via F1-scores, and confusion matrices were used to predict the treatment path (out of 4 paths) with and without additional referral parameters. RESULTS Data from 1608 patients were evaluated. The evaluation indicated that 2 referral reasons from the referral letters (for anesthesiology and rehabilitation intervention) increased the F1-score accuracy by up to 19.5% for triaging. The confusion matrices confirmed the results. CONCLUSIONS This study indicates that data enriching by adding NLP-based extraction of the content of referral letters increases the model accuracy of DSSs in suggesting optimal treatments for individual patients with LBP. Overall model accuracies were considered low and insufficient for clinical application.

Background: Inconsistent descriptions of Lumbar multifidus (LM) morphology were previously identified, especially in research applying ultrasonography (US), hampering its clinical applicability with regard to diagnosis and therapy. The aim of this study is to determine the LM-sonoanatomy by comparing high-resolution reconstructions from a 3-D digital spine compared to standard LM-ultrasonography. Methods: An observational study was carried out. From three deeply frozen human tissue blocks of the lumbosacral spine, a large series of consecutive photographs at 78 μm interval were acquired and reformatted into 3-D blocks. This enabled the reconstruction of (semi-)oblique cross-sections that could match US-images obtained from a healthy volunteer. Transverse and oblique short-axis views were compared from the most caudal insertion of LM to L1. Results: Based on the anatomical reconstructions, we could distinguish the LM from the adjacent erector spinae (ES) in the standard US imaging of the lower spine. At the lumbosacral junction, LM is the only dorsal muscle facing the surface. From L5 upwards, the ES progresses from lateral to medial. A clear distinction between deep and superficial LM could not be discerned. We were only able to identify five separate bands between every lumbar spinous processes and the dorsal part of the sacrum in the caudal anatomical cross-sections, but not in the standard US images. Conclusion: The detailed cross-sectional LM-sonoanatomy and reconstructions facilitate the interpretations of standard LM US-imaging, the position of the separate LM-bands, the details of deep interspinal muscles, and demarcation of the LM versus the ES. Guidelines for electrode positioning in EMG studies should be refined to establish reliable and verifiable findings. For clinical practice, this study can serve as a guide for a better characterisation of LM compared to ES and for a more reliable placement of US-probe in biofeedback.