The research team, led by Randa Al-Kharabsheh and Muayyad Ahmad, analyzed the electronic health records of 2,795 adult neurology admissions treated between 2022 and 2024. Each patient’s diagnosis was coded and grouped according to the International Classification of Diseases, Tenth Revision, allowing the investigators to compare readmission patterns across disease categories rather than treating every neurological admission as interchangeable. From the same records, they extracted the clinical and demographic variables that hospitals routinely collect, and then put nine different prediction models head to head to see which could best identify patients at risk of an unplanned return.

The headline numbers are sobering. Across the full cohort, 8.9 percent of neurological patients were readmitted unexpectedly after discharge. But that average conceals dramatic variation: among patients whose primary diagnosis was a neoplasm, roughly one in four, 24 percent, came back to the hospital unplanned, a rate nearly three times the overall figure. That single statistic illustrates why the researchers argue that population-level risk stratification, rather than one-size-fits-all discharge planning, could meaningfully change how neurology services allocate follow-up resources.

After comparing the nine candidate models, the team selected the Chi-square Automatic Interaction Detection algorithm, better known as CHAID. The method works by repeatedly splitting the patient population on the variable that shows the statistically strongest association with the outcome, in this case unplanned readmission, and then splitting each branch again, producing a tree whose paths can be read like a flowchart. A clinician can trace a branch and see, in plain terms, that a patient with a particular combination of comorbidity burden, age, sex, and region of residence lands in a high-risk leaf. That interpretability is precisely why the authors chose CHAID over more accurate but opaque competitors: in a hospital setting, a model that clinicians can understand and challenge is often more useful than one that performs marginally better but cannot explain itself.

The model’s raw discriminative power was modest. On a held-out test set, it achieved an area under the receiver operating characteristic curve, or AUC, of 0.671, a value the authors themselves characterize as poor to fair. In practical terms, the model distinguishes between patients who will and will not be readmitted only somewhat better than chance would allow at the individual level. For risk stratification across the full cohort, it reached a balanced accuracy of 0.63, with sensitivity of 0.52, meaning it caught about half of eventual readmissions, and specificity of 0.74, meaning it correctly cleared most patients who would not return. The researchers are candid that these numbers do not support using the tool to predict any single patient’s fate, and they frame the model instead as a screening instrument for identifying subgroups that deserve intensified attention.

What the model lacks in individual precision, it partly compensates for in the clarity of its risk profiles. The strongest single predictor was the Charlson Comorbidity Index, a weighted score that tallies a patient’s burden of chronic illnesses; patients scoring five or higher sat firmly in the high-risk branches of the tree. Age, sex, length of hospital stay, and geographic region also shaped the splits. The subgroup analyses are where the study becomes genuinely revealing. Vascular patients, such as those admitted after stroke, who also carried high comorbidity scores formed a distinctly high-risk cluster, an association so strong that it reached a p-value below .001 with a chi-square statistic of 66.368. Among younger female patients, 37.3 percent had neoplasms, a statistically significant enrichment that flags a group in whom cancer-related readmissions drive much of the risk. And patients diagnosed with functional disorders who came from northern regions emerged as a third distinct profile, a finding the authors report with a p-value of .009.

Each of these profiles tells a different operational story. High-comorbidity stroke patients may need tighter medication reconciliation, earlier outpatient follow-up, and coordinated management of diabetes, kidney disease, and cardiac conditions that neurologists alone cannot address. Young women with brain tumors may require oncology-linked discharge planning, symptom management protocols, and clearer pathways back to specialized care when complications arise. The geographic signal among functional disorder patients hints at structural inequities, possibly in access to follow-up services or in the burden of travel, that no amount of bedside clinical care can fix on its own. A decision tree cannot say which intervention will work, but by making the risk architecture visible, it tells hospital administrators where to look first.

The study carries additional weight because of where it was conducted. The authors note that this is one of the first investigations from the Middle East to demonstrate the potential of CHAID-based risk stratification for neurology readmissions. Much of the readmission-prediction literature comes from North American and European health systems with different payment structures, discharge practices, and patient populations, and models trained in those settings often transfer poorly. Building local evidence from Jordanian electronic health records is a step toward risk tools that reflect the actual epidemiology and health-service geography of the region, including the role of regional referral patterns that emerged as a predictor in this analysis.

The researchers are equally clear about the limits of their work. The analysis was retrospective and internally validated only; the risk-stratification metrics were calculated descriptively from the complete cohort, while independent validation rested on the held-out test-set AUC. Before any clinical deployment, the model must undergo external validation in different hospitals, different regions, and ideally prospective data collection. Readmission is also a notoriously messy outcome, influenced by social factors, access to primary care, and hospital discharge practices that electronic records capture imperfectly. An AUC of 0.671 is a floor, not a ceiling, and the authors position their model as a foundation for refinement rather than a finished clinical instrument.

Even so, the study lands at a moment when health systems everywhere are under pressure to reduce avoidable returns, and it makes a case that is easy to underestimate: sometimes the most valuable algorithm is not the most accurate one but the most legible. A gradient-boosted ensemble might have squeezed a few extra points of AUC out of the same data, but no clinician could read it at a glance. A CHAID tree, by contrast, hands neurology departments a structured picture of their readmission problem, anchored in variables they already measure, and points them toward the specific combinations of diagnosis, comorbidity, age, sex, and geography where intervention is most likely to matter. For the one in eleven neurological patients in this cohort who returned to the hospital unplanned, and for the one in four cancer patients among them, that kind of clarity is a practical first step toward care that anticipates, rather than reacts to, the revolving door.

Subject of Research: AI-based prediction of unplanned hospital readmissions among neurological patients

Article Title: Predicting unplanned readmissions in neurological patients: A large-scale AI-driven analysis using CHAID decision trees

Article References: Al-Kharabsheh, R., & Ahmad, M. (2026). Predicting unplanned readmissions in neurological patients: A large-scale AI-driven analysis using CHAID decision trees. PLOS One, 21(10), e0360077. https://doi.org/10.1371/journal.pone.0360077

Image Credits: AI Generated

DOI: 10.1371/journal.pone.0360077

Keywords: unplanned readmissions, neurology, CHAID decision tree, machine learning, electronic health records, risk stratification, Charlson Comorbidity Index, Jordan, PLOS One, predictive modeling, hospital discharge, health informatics

The research, led by Julio Navío-Marco, Diego Ardura, and Arturo Galán of the Universidad Nacional de Educación a Distancia in Madrid, was conducted within an Applied Statistics in Education course at a major hybrid university. Students were offered the option of tackling one of their assignments individually or in teams of two or three. After the term ended, the entire cohort was surveyed with an instrument designed specifically for the study, measuring perceived utility, anticipated difficulties, skills acquired, and—for those who chose teams—the demands of the experience and its contribution to building learning communities. The design was ex-post-facto, meaning the researchers analyzed choices and perceptions after the fact rather than manipulating conditions, and the survey was resent weekly until every student had responded.

The statistical machinery behind the study was substantial. Exploratory factor analyses confirmed that the survey’s scales were robust: the 13-item perceived utility scale explained nearly 65 percent of item variance with a reliability of 0.97 on McDonald’s omega, while the difficulty, skills, effort, and learning community scales all showed similarly strong psychometric properties. To compare perceptions across students, the team used repeated measures ANOVA with Greenhouse–Geisser corrections for violated sphericity, and effect sizes were estimated with eta-squared and Cohen’s d. But the most revealing analysis came from a decision tree algorithm called CHAID—Chi-squared Automatic Interaction Detector—which segmented students into groups based on the variables that best predicted whether they would choose teamwork again.

The first CHAID model, predicting which modality students actually chose for the assignment, classified 72.1 percent of the sample correctly. Its verdict was unambiguous: the single best predictor of choosing teamwork was whether students believed collaboration was necessary for their education. Students scoring above three on that item were far more likely to opt for a team. Among those who saw teamwork as less essential, the next fork was performance—those hoping to improve their results leaned toward group work—while students who neither valued teamwork for learning nor sought performance gains chose teams only when their schedules allowed it. In short, students who preferred working alone generally did not see collaborative tasks as essential, did not expect performance benefits, and had limited availability.

The perceived benefits that separated team players from solo workers were strikingly consistent. Students who chose groups believed teamwork would make them more efficient (d = 0.63), more engaged (d = 0.58), higher performing (d = 0.51), and more responsible for their own learning (d = 0.51), with motivation and the perceived indispensability of teamwork for their education showing similarly moderate effects. Across the whole sample, the highest-rated contributions were improvements in communication skills, the generation of new knowledge, meeting educational needs, getting to know peers, reducing the isolation of distance learning, and creating a learning community. On the skills side, only employability-oriented competencies showed a meaningful difference between the groups, with a moderate effect size of 0.48—a signal that students drawn to teamwork are, at least in part, investing in their professional futures.

The negative side of the ledger was dominated by practical friction. Scheduling and availability difficulties and uneven member involvement topped the list of anticipated problems, while communication challenges ranked lowest. Students who worked individually rated nearly every anticipated difficulty higher than their team-working peers, with the largest gaps involving remote communication (d = 0.34), lack of commitment (d = 0.35), and participation in meetings (d = 0.36). The researchers point to the persistent specter of the free rider—the teammate who coasts on others’ effort—as a key driver of distrust, echoing a long line of studies identifying unequal participation as the most frequent drawback of online group work. Notably, not knowing teammates in advance was not a significant barrier in either group, suggesting that the real obstacle is not unfamiliarity but confidence that everyone will pull their weight.

The second decision tree, modeling whether team-working students would choose collaboration again, was even more accurate, classifying 91.4 percent of cases correctly. Its master variable was the value students placed on helping peers and benefiting from the resulting feedback. Students who rated mutual assistance above three points were highly likely to return to teams, and that likelihood rose further among those interested in building learning communities. For students with intermediate motivation to help and be helped, the prospect of improving grades tipped the balance, while those with low mutual-help motivation rarely persisted—unless they believed teamwork would lighten their workload. When the researchers compared students eager to repeat teamwork with those who preferred to retreat to individual work, the effect sizes on learning community perceptions were enormous, reaching d = 2.46 for the belief that teamwork enables mutual assistance among peers.

These results land at a particularly consequential moment for teacher education. The students in the study are future teachers, and their attitudes toward collaboration today will shape the classroom practices of tomorrow. Previous research has shown that many educators receive little formal training in teamwork pedagogy and are often unsure how to teach collaborative skills, which imposes additional transaction costs on already stretched instructors. The Spanish team argues that if teachers themselves are not convinced of the intrinsic value of teamwork, they will struggle to sell it to their students. Their data suggest that awareness and training efforts should be concentrated most intensely on future educators who gravitate toward individual work to avoid the difficulties of coordination and trust.

The theoretical grounding of the study connects these findings to a deeper tradition. Teamwork, the authors note, functions as a Vygotskian Zone of Proximal Development in which students learn from peers as well as instructors, and it aligns with Bandura’s social cognitive theory, in which learning emerges through dynamic, mutual interaction with others. In distance environments, collaboration also demands self-regulation, and the study taps into the emerging concept of socially shared regulation of learning, where teams collectively manage their own processes. The Community of Inquiry framework—teaching, cognitive, and social presence—provides the scaffolding for the instructor’s role: defining group sizes, assigning warm-up team-building activities in scarce face-to-face sessions, supplementing with online interactive games, and assessing individual contributions within team projects.

The practical prescriptions that follow are concrete. Instructors in hybrid settings should reserve precious in-person time for tasks best done face to face—team building, strategic discussion, defining objectives—while providing detailed written instructions for the online component, using synchronous meetings and interactive software deliberately, and evaluating both the process and the product through meeting minutes, peer and self-evaluations, and progress reports. Preparatory measures that build interpersonal familiarity before teams begin could directly address the trust deficit the data exposed. The authors also flag promising lines for future research: identifying which assignments suit blended environments best, using teamwork to reduce the isolation that remains a chronic weakness of distance education, and training future educators to champion collaboration with their own students. What the study makes clear is that teamwork in online universities is not a default that happens when students are placed in groups—it is a perception-driven choice, won or lost in the minds of learners long before the first meeting is scheduled.

Subject of Research: Perceptions of teamwork skill development among future teachers in hybrid and online university education

Article Title: Teamwork skills development in hybrid and online universities: the perspective of the future teachers

Article References: Teamwork skills development in hybrid and online universities: the perspective of the future teachers. (n.d.). https://doi.org/10.1007/s44322-025-00032-1

Image Credits: AI Generated

DOI: 10.1007/s44322-025-00032-1

Keywords: teamwork, blended learning, online education, collaborative learning, teacher training, employability skills, learning communities, CHAID decision tree, free rider problem, self-regulated learning, higher education, distance learning