The present issue of Przegląd Lekarski – Jagiellonian Medical Review brings together contributions that reflect several of the most important directions in contemporary medicine: the long-term consequences of COVID-19, determinants of preventive behavior, promises and limits of artificial intelligence (AI), complexity of musculoskeletal and hematologic diagnostics, and growing role of advanced imaging in individualized care. Taken together, these publications demonstrate the evolution of the discipline toward a paradigm that is concurrently more data-driven and increasingly reliant on nuanced, expert clinical judgment.

The issue opens with the original study by Sternalski et al,1 which addresses one of the still unresolved questions of the postpandemic period: the biological persistence of long COVID. The authors focus on patients with elevated postinfection D-dimer concentrations and show that, after a median follow-up of more than 16 months, persistent elevation remains common and is associated with fatigue, higher fibrinogen and N-terminal pro–B-type natriuretic peptide levels, higher neutrophil and monocyte indices, and lower lymphocyte percentage and estimated glomerular filtration rate. Their findings indicate that, in at least a subset of patients, long COVID is associated with persistent laboratory abnormalities that are consistent with ongoing inflammatory, hemostatic, and potentially cardiovascular dysregulation. This paper thus contributes to the ongoing paradigm shift from conceptualizing long COVID as a purely symptom-based clinical syndrome to recognizing it as a condition characterized by quantifiable biological correlates.

A complementary public-health perspective is offered by the cross-sectional study by Kieruzal et al,2 focusing on willingness to use personal protective equipment and adopt personal protective behaviors among unvaccinated individuals. In a cohort of 1276 participants, most respondents reported using at least 1 preventive measure, with mask use in closed spaces being the most common. At the same time, the study demonstrates that adherence is not socially uniform: women and younger participants, especially those aged 18–29 years, were more likely to report protective behaviors, and sex and age emerged as the key determinants of willingness to adopt them. This is more than a retrospective reflection on COVID-19. It is also a reminder that preparedness for future epidemics depends not only on recommendations issued from above, but on understanding which groups are more likely to comply, which are less likely, and how public-health messaging should be tailored accordingly.

The review by Michalska and Błaż3 widens the horizon still further by surveying AI applications across neurology. Their umbrella review includes 58 systematic reviews and meta-analyses spanning stroke, dementia, movement disorders, neuro-oncology, epilepsy, multiple sclerosis, neuromuscular disorders, headache, and neurocritical care. The overall picture is strikingly ambivalent. On the one hand, reported model accuracies often range from 70% to above 90%, and the principal applications—imaging analysis, diagnosis and classification, and prediction of prognosis and outcomes—indicate how deeply AI has entered neurologic research. On the other hand, the review makes it equally clear that the evidence base remains methodologically fragile: many included studies were of low or critically low quality, and the lack of external validation and sufficiently diverse datasets remains a major barrier to implementation. This type of rigorous, integrative assessment is exactly what the field requires. The trajectory of AI in medicine will not be determined solely by impressive performance metrics; rather, it will depend critically on reproducibility, external validity and generalizability, as well as demonstrable clinical utility.

Diagnostic ambiguity is the central theme of the review by Dyngosz and Krzyżewski,4 which discusses hip–spine syndrome. As degenerative changes in the hip and lumbar spine frequently coexist and may produce overlapping symptoms, identifying the primary source of pain can be difficult. The authors underscore the enduring centrality of precise clinical history-taking and comprehensive physical examination, supplemented, in appropriately selected cases, by diagnostic intra-articular anesthetic injections. Equally important is their discussion of treatment sequencing. Although uncertainty persists, the reviewed literature increasingly favors performing total hip arthroplasty before spinal surgery, as this strategy appears to reduce postoperative complications, including hip dislocation and revision procedures, while also improving functional outcomes. In an era enamored with ever more sophisticated tests, this review is a useful reminder that tackling many of the hardest problems in medicine still begins with a differential diagnosis.

The review by Holcman et al5 turns to a different diagnostic frontier: molecular imaging in cardiac amyloidosis. As clinical recognition of transthyretin- and light-chain–related amyloidosis increases, there is a correspondingly greater demand for diagnostic algorithms that are both highly accurate and minimally invasive. The authors show how nuclear imaging has transformed this landscape. Bone-avid tracer scintigraphy now enables highly sensitive and specific diagnosis of transthyretin amyloidosis, reducing reliance on endomyocardial biopsy, while single-photon emission computed tomography with computed tomography enhances diagnostic precision, allows quantification of amyloid burden, and adds prognostic value. Meanwhile, positron emission tomography is presented as an emerging modality holding promise for therapy monitoring, although its role remains under investigation. Importantly, the article frames imaging not as an isolated tool, but as part of an integrated algorithm that includes magnetic resonance, echocardiography, and biomarkers. This multimodal perspective is exactly what precision cardiology now requires.

The issue then moves to hematology, with the review by Iwaniec et al6 discussing the challenges of diagnosing von Willebrand disease. Despite being the most common inherited bleeding disorder, von Willebrand disease remains underdiagnosed and diagnostically challenging. The review clearly outlines the current classification into types 1, 2, and 3, with the clinically relevant subtype distinctions within type 2, and it maps the diagnostic pathway from initial assays (factor VIII coagulant activity, von Willebrand factor antigen, and ristocetin cofactor activity), through newer tests assessing glycoprotein Ib binding, to specialized analyses, such as multimer assessment, platelet aggregation testing, factor VIII binding assays, and genetic testing. The central message is both clinically relevant and of substantial practical importance: accurate typing is not merely an academic endeavor but constitutes the essential foundation for selecting appropriate therapeutic strategies and implementing effective bleeding prophylaxis. In that sense, this paper exemplifies how diagnostic precision directly shapes patient safety.

The final contribution, a clinical image by Stażyk et al,7 provides a concise but memorable infectious-disease vignette. It presents a 19-year-old man with unilateral axillary lymphadenopathy, fever, and a history of repeated cat scratches, ultimately diagnosed with cat scratch disease due to Bartonella henselae infection. The clinical image highlights a classic lesson: common presentations, such as lymphadenitis, require attention not only to laboratory findings and empiric therapy, but also to epidemiologic clues and exposure history. It additionally underscores that the clinical course may be complicated by the development of suppuration, potentially necessitating surgical drainage.

What unites these diverse contributions is a shared commitment to precision. In some papers, precision is technological: biomarker follow-up, molecular imaging, or AI. In others, it is behavioral, epidemiologic, or phenomenological: identifying who adopts preventive measures, distinguishing hip pain from spinal pain, or recognizing the clinical signature of a zoonotic infection. A single principle underlies all these developments: progress in medicine arises not solely from the accumulation of novel instruments and technologies, but from the refinement of the knowledge and judgment required for their appropriate application. This issue offers a useful cross-section of that effort and, in doing so, captures both the opportunities and the discipline required in modern clinical science.