Introduction: In Serbia and other countries of Central and Eastern Europe, malignant disease incidence and mortality rates have risen continuously over recent decades. Despite progress in understanding cancer epidemiology, comprehensive analyses covering all tumor localizations across both sexes using modern statistical methods remain lacking for the full 2005-2024 period. Furthermore, the impact of socioeconomic factors, healthcare availability, and events such as the COVID-19 pandemic on cancer trends has not been sufficiently investigated. Aim: To determine the time trends of incidence and mortality rates of malignant diseases in central Serbia over the period 2005-2024, identify trend breakpoints and characteristic changes, and analyze risk factors and prevention strategies for each cancer localization. Material and Method: Incidence and mortality data for malignant diseases (both sexes, 2005-2024) were obtained from the official reports of the Institute for Public Health "Dr. Milan Jovanović Batut" Belgrade. Age-standardized rates per 100,000 inhabitants were calculated separately for women and men for each cancer type, including stomach (C16), colorectal (C18-C20), liver (C22), pancreas (C25), lung and bronchus (C34), bladder (C67), brain (C71), leukemias (C91-C95), all sites (C00-C97), and all sites except skin (C44), as well as sex-specific cancers (breast, cervix uteri, corpus uteri, ovary for women; mouth/pharynx, larynx, prostate, kidney for men). Trend analysis was performed using Joinpoint Regression Program version 5.4.0, calculating Annual Percent Change (APC), Average Annual Percent Change (AAPC), 95% confidence intervals, and p-values via t-test. The permutation test was used to determine the optimal number of breakpoints (0-3), with significance set at p < 0.05. Results: Overall cancer incidence increased significantly for both sexes: AAPC = 3.20% (95% CI 0.38-6.10; p = 0.028) for women and AAPC = 6.79% (95% CI 2.37-11.41; p = 0.004) for men. Mortality also rose significantly: AAPC = 3.93% (95% CI 2.87-5.00; p < 0.001) for women and AAPC = 5.34% (95% CI 1.01-9.84; p = 0.017) for men. In common cancers for both sexes, two trend breakpoints were identified around 2017 and 2020, with a sharp increase in incidence (APC for men = 19.05% in 2017-2020) and mortality (APC for women = 14.15% in 2017-2020), followed by a post-2020 decline. Conclusion: Cancer incidence and mortality in central Serbia rose significantly from 2005 to 2024, with men showing approximately twice the rate of increase compared to women. A sharp spike in trends during 2017-2020 likely reflects a combination of improved diagnostics, changes in healthcare organization, and indirect effects of the COVID-19 pandemic including delayed diagnoses and misattribution of mortality causes. Pancreatic, liver, bladder, and kidney cancers, and leukemias represent the most dynamically increasing malignancies. Conversely, the declining trends in stomach and cervical cancers reflect clear benefits from H. pylori eradication, improved hygiene, screening programs. Regarding HPV vaccination, its long-term efficacy and potential delayed adverse effects require multi-year monitoring; thus it should be applied exclusively on a voluntary basis by adults with signed informed consent. Prevention of the rising cancers requires targeted public health interventions addressing tobacco use, obesity, diabetes, alcohol consumption, and environmental exposures.
Introduction: Climate change has emerged as one of the greatest threats to global health, and the healthcare sector contributes substantially to greenhouse gas emissions. In the management of respiratory diseases, pressurized metered-dose inhalers (pMDIs) are an important source of greenhouse gas emissions because they contain hydrofluoroalkane propellants with a high global warming potential. Methodology: This review summarizes current evidence on the environmental impact of inhaler devices and evaluates the environmental, clinical, patient-related, and economic considerations that should guide sustainable inhaler prescribing for asthma and chronic obstructive pulmonary disease (COPD). A literature search was conducted in the PubMed database from inception to 1 June 2026 using the search terms "inhalation device" OR "environmental". Seventy-nine articles were identified and supplemented by relevant publications from reference lists. Topic: Current evidence demonstrates that dry powder inhalers (DPIs) and soft mist inhalers (SMIs) have substantially lower carbon footprints than pMDIs while providing comparable clinical outcomes when appropriately selected for patients. However, inhaler choice should not be based solely on environmental considerations, as clinical effectiveness, patient characteristics, inhaler technique, patient preferences, and device availability remain essential determinants of successful treatment. Economic evaluations suggest that switching clinically eligible patients from pMDIs to lower-carbon alternatives may reduce both greenhouse gas emissions and healthcare costs in selected settings, although the financial impact depends on the replacement strategy and local pricing. Conclusion: Overall, sustainable inhaler prescribing should integrate environmental sustainability with evidence-based clinical practice and shared decision-making to optimize patient outcomes while reducing the environmental impact of respiratory care.
Introduction: Quality healthcare is a fundamental right, yet it remains unaffordable for many, particularly in resource-limited settings such as the public healthcare sector. Majority of these limited resources are consumed by critically ill patients, that usually have low survival rates, leaving inadequate resources for the less critical patients. This results in inefficient resource utilization and suboptimal patient outcomes. Therefore, effective triage is essential to optimize resource allocation in such settings. Aim: This study aims to evaluate the association between APACHE II scores and outcomes in the surgical ICU of a tertiary care hospital. Material and Method: This observational longitudinal cohort study was conducted over a two-month period in 100 adult patients admitted to the Surgical/Trauma ICU of a tertiary care government hospital. All patients received standard ICU treatment as per institutional protocol. Data were collected using case record forms, to study the APACHE II scores, drug utilization patterns, and their association with patient outcomes. Results: Of the 100 patients, 53 died while 47 survived. Non-survivors had significantly higher APACHE II scores and received more drugs. We observed that patients with an APACHE II score below 16 had a 33.3% mortality rate, while those with a score of 16 or above had a 79.06% mortality rate, a statistically significant difference. However, the use of high-cost medications like broad-spectrum antibiotics and albumin did not significantly differ between survivors and non-survivors. The average treatment cost was slightly higher for non-survivors, but not statistically significant. Conclusion: Our study found that the APACHE II score is a reliable predictor of mortality in the ICU setting, with a score of 16 serving as an effective cut-off for categorizing patient risk. Higher APACHE II scores were associated with increased mortality, longer ICU stays, and relatively higher treatment costs. The use of expensive medications did not improve outcomes for critically ill patients. Nevertheless, future large-scale, multicentre studies utilizing extended tracking after discharge from ICU are warranted, to evaluate the intersection of clinical severity, pharmaceutical optimization, and healthcare economics, more comprehensively.
Introduction: Breast cancer remains the most frequently diagnosed malignancy and one of the leading causes of cancer mortality among women worldwide. Despite the implementation of organized mammography screening programmes, considerable disparities in mortality persist, particularly in lowand middle-income countries. Artificial intelligence (AI) has emerged as a promising tool for improving the accuracy, efficiency, and accessibility of breast cancer screening. To review current evidence regarding the role of artificial intelligence in breast cancer screening and early detection, with particular emphasis on its potential application within the healthcare system of the Republic of Serbia. Methodology: A comprehensive literature review was conducted using the MEDLINE, PubMed, and KOBSON databases. Literature published between 2012 and 2026 were identified using the keywords breast cancer, artificial intelligence, mammography, screening, and machine learning. After applying predefined inclusion criteria, 23 relevant publications were included in the final analysis. Topic: Current evidence demonstrates that AI-assisted mammography achieves diagnostic performance comparable to or exceeding that of experienced breast radiologists, while substantially reducing reading workload. Prospective clinical trials have shown that AI-supported screening increases early cancer detection rates without compromising patient safety. Beyond lesion detection, AI enables individualized risk assessment, prediction of interval cancers, and supports the transition toward risk-adapted screening strategies. For Serbia, AI-assisted centralized triage systems could optimize limited radiology resources, improve access to expert interpretation, and strengthen organized screening programmes. Successful implementation, however, requires local validation studies, interoperable digital infrastructure, appropriate regulatory frameworks, and continuous professional education. Conclusion: Artificial intelligence has the potential to substantially improve breast cancer screening by increasing diagnostic accuracy, reducing radiologists' workload, and supporting personalized screening approaches. Strategic integration aligned with European quality standards and national healthcare priorities could promote earlier diagnosis and reduced breast cancer mortality in Serbia.
Introduction: Epilepsy affects more than 50 million people worldwide. In approximately one-third of patients, seizures cannot be adequately controlled with medication. A major challenge in clinical practice is that the effectiveness of a drug can only be assessed after weeks or months of observation - during which the patient may be receiving suboptimal treatment. The electroencephalogram (EEG) provides a direct, continuous window into brain electrical activity, yet in routine practice it is used almost exclusively for diagnosis rather than for ongoing therapy monitoring. Aim: To develop and test a computational model that continuously quantifies how far the EEG signal is from a normal brain state - a measure we call latent distance - and to evaluate whether this measure could serve as a real-time, non-invasive biomarker of antiepileptic drug effect. Material and Method: A deep learning model was trained on a publicly available EEG dataset (Bonn, Germany) using a self-supervised approach that requires no manual labelling of the data. The model was then tested on an entirely independent dataset (CHB-MIT, Boston, USA) containing EEG recordings from paediatric patients with drug-resistant epilepsy - a strict test of whether the model generalises beyond its training data. Performance was assessed using standard diagnostic accuracy metrics. Results: On the external validation dataset, the model achieved an area under the ROC curve (AUC) of 0.80 (95% CI: 0.73-0.95), a precision-recall AUC of 0.97, and a balanced accuracy of 81%. Crucially, the latent distance signal followed a clinically meaningful trajectory: it rose before the documented seizure onset, peaked during the seizure, and gradually returned to baseline in the post-ictal period. Conclusion: This proof-of-concept study demonstrates that a self-supervised EEG model can detect seizures and continuously track deviations from normal brain activity in patients it has never encountered before. Latent distance may represent a novel, continuous EEG biomarker with potential utility in real-time monitoring of antiepileptic therapy - particularly in identifying drug effects before changes in seizure frequency become apparent. Prospective clinical studies are needed to confirm this hypothesis.
