Robot-assisted surgery and artificial intelligence-based tumour diagnostics: social preferences with a representative cross-sectional survey.

BACKGROUND: The aim of this study was to assess social preferences for two different advanced digital health technologies and investigate the contextual dependency of the preferences. METHODS: A cross-sectional online survey was performed among the general population of Hungary aged 40 years and over. Participants were asked to imagine that they needed a total hip replacement surgery and to indicate whether they would prefer a traditional or a robot-assisted (RA) hip surgery. To better understand preferences for the chosen method, the willingness to pay (WTP) method was used. The same assessment was conducted for preferences between a radiologist's and AI-based image analysis in establishing the radiological diagnosis of a suspected tumour. Respondents' electronic health literacy was assessed with the eHEALS questionnaire. Descriptive methods were used to assess sample characteristics and differences between subgroups. Associations were investigated with correlation analysis and multiple linear regressions. RESULTS: Altogether, 1400 individuals (53.7% female) with a mean age of 58.3 (SD = 11.1) years filled in the survey. RA hip surgery was chosen by 762 (54.4%) respondents, but only 470 (33.6%) chose AI-based medical image evaluation. Those who opted for the digital technology had significantly higher educational levels and electronic health literacy (eHEALS). The majority of respondents were willing to pay to secure their preferred surgical (surgeon 67.2%, robot-assisted: 68.8%) and image assessment (radiologist: 70.9%; AI: 77.4%) methods, reporting similar average amounts in the first (p = 0.677), and a significantly higher average amount for radiologist vs. AI in the second task (p = 0.001). The regression showed a significant association between WTP and income, and in the hip surgery task, it also revealed an association with the type of intervention chosen. CONCLUSIONS: Individuals with higher education levels seem to accept the advanced digital medical technologies more. However, the greater openness for RA surgery than for AI image assessment highlights that social preferences may depend considerably on the medical situation and the type of advanced digital technology. WTP results suggest rather firm preferences in the great majority of the cases. Determinants of preferences and real-world choices of affected patients should be further investigated in future studies.

The Impact of Parental Electronic Health Literacy on Disease Management and Outcomes in Pediatric Type 1 Diabetes Mellitus: Cross-Sectional Clinical Study.

BACKGROUND: Despite the growing uptake of smart technologies in pediatric type 1 diabetes mellitus (T1DM) care, little is known about caregiving parents' skills to deal with electronic health information sources. OBJECTIVE: We aimed to assess the electronic health literacy of parents caring for children with T1DM and investigate its associations with disease management and children's outcomes. METHODS: A cross-sectional survey was performed involving 150 parent-child (8-14 years old with T1DM) dyads in a university pediatric diabetology center. Parents' electronic health literacy (eHealth Literacy Scale [eHEALS]), general health literacy (Chew questionnaire and Newest Vital Sign [NVS]), and attitudes toward T1DM care (Parental Self-Efficacy Scale for Diabetes Management [PSESDM] and Hypoglycemia Fear Survey [HFS]) were investigated. Children's treatment, HbA1c level, and quality of life (Pediatric Quality of Life Inventory Diabetes Module [PedsQL Diab] and EQ-5D-Y-3L) were assessed. Multiple linear regression analysis was performed to investigate the determining factors of 6-month average HbA1c. RESULTS: Of the 150 children, 38 (25.3%) used a pen, 55 (36.7%) used a pen plus a sensor, 6 (4.0%) used an insulin pump, and 51 (34.0%) used an insulin pump plus a sensor. Parents' average eHEALS score (mean 31.2, SD 4.9) differed significantly by educational level (P=.04) and the children's treatment (P=.005), being the highest in the pump + sensor subgroup. The eHEALS score showed significant Pearson correlations with the Chew score (r=-0.45; P<.001), NVS score (r=0.25; P=.002), and PSESDM score (r=0.35; P<.001) but not with the children's HbA1c (r=-0.143; P=.08), PedsQL Diab (r=-0.0002; P>.99), and EQ-5D-Y-3L outcomes (r=-0.13; P=.12). Regression analysis revealed significant associations of the child's HbA1c level with sex (ß=0.58; P=.008), treatment modality (pen + sensor: ß=-0.66; P=.03; pump + sensor: ß=-0.93; P=.007), and parents' self-efficacy (PSESDM; ß=-0.08; P=.001). CONCLUSIONS: Significantly higher parental electronic health literacy was found in T1DM children using a glucose sensor. The electronic health literacy level was associated with parents' diabetes management attitude but not with the child's glycemic control. Studies further investigating the role of parental electronic health literacy in T1DM children managed at different levels of care and the local context are encouraged.

The Reporting Quality of Machine Learning Studies on Pediatric Diabetes Mellitus: Systematic Review.

BACKGROUND: Diabetes mellitus (DM) is a major health concern among children with the widespread adoption of advanced technologies. However, concerns are growing about the transparency, replicability, biasedness, and overall validity of artificial intelligence studies in medicine. OBJECTIVE: We aimed to systematically review the reporting quality of machine learning (ML) studies of pediatric DM using the Minimum Information About Clinical Artificial Intelligence Modelling (MI-CLAIM) checklist, a general reporting guideline for medical artificial intelligence studies. METHODS: We searched the PubMed and Web of Science databases from 2016 to 2020. Studies were included if the use of ML was reported in children with DM aged 2 to 18 years, including studies on complications, screening studies, and in silico samples. In studies following the ML workflow of training, validation, and testing of results, reporting quality was assessed via MI-CLAIM by consensus judgments of independent reviewer pairs. Positive answers to the 17 binary items regarding sufficient reporting were qualitatively summarized and counted as a proxy measure of reporting quality. The synthesis of results included testing the association of reporting quality with publication and data type, participants (human or in silico), research goals, level of code sharing, and the scientific field of publication (medical or engineering), as well as with expert judgments of clinical impact and reproducibility. RESULTS: After screening 1043 records, 28 studies were included. The sample size of the training cohort ranged from 5 to 561. Six studies featured only in silico patients. The reporting quality was low, with great variation among the 21 studies assessed using MI-CLAIM. The number of items with sufficient reporting ranged from 4 to 12 (mean 7.43, SD 2.62). The items on research questions and data characterization were reported adequately most often, whereas items on patient characteristics and model examination were reported adequately least often. The representativeness of the training and test cohorts to real-world settings and the adequacy of model performance evaluation were the most difficult to judge. Reporting quality improved over time (r=0.50; P=.02); it was higher than average in prognostic biomarker and risk factor studies (P=.04) and lower in noninvasive hypoglycemia detection studies (P=.006), higher in studies published in medical versus engineering journals (P=.004), and higher in studies sharing any code of the ML pipeline versus not sharing (P=.003). The association between expert judgments and MI-CLAIM ratings was not significant. CONCLUSIONS: The reporting quality of ML studies in the pediatric population with DM was generally low. Important details for clinicians, such as patient characteristics; comparison with the state-of-the-art solution; and model examination for valid, unbiased, and robust results, were often the weak points of reporting. To assess their clinical utility, the reporting standards of ML studies must evolve, and algorithms for this challenging population must become more transparent and replicable.

Augmented or Mixed Reality Enhanced Head-Mounted Display Navigation for In Vivo Spine Surgery: A Systematic Review of Clinical Outcomes.

BACKGROUND: This research paper provides a systematic literature review (SLR) on the current status of augmented-reality head-mounted devices (AR-HMDs) that guide and navigate spine surgeries and pedicle screw placement. METHODS: Embase, Scopus, PubMed, Cochrane Library and IEEE Xplore databases were screened for the systematic literature search to collect and statistically analyze live patient clinical, procedural and user experience data. Multi-level Poisson and binominal models were used for analysis. RESULTS: In vivo patient data, only the clinically widely used Gertzbein-Robbins Scale, were published as an outcome in the recent heterogeneous literature. The statistical analysis supports the hypothesis that using AR-HMDs has the same clinical outcomes as using more expensive robot-assisted surgical (RAS) systems. CONCLUSIONS: AR-HMD-guided pedicle screw insertion is reaching its technology readiness, providing similar benefits to RAS. Further meta-analysis is expected in the future from higher case-numbered and standardized randomized clinical trials.

Epidemiology and patients' self-reported knowledge of implantable medical devices: Results of a cross-sectional survey in Hungary.

BACKGROUND: Implantable medical devices (IMDs) are medical instruments embedded inside the body. Well-informed and empowered patients living with IMDs are key players of improving IMD-related patient safety and health outcomes. However, little is known about IMD patients' epidemiology, characteristics, and current awareness levels. Our primary aim was to investigate the point and lifetime prevalence of patients living with IMDs. Patients' IMD-related knowledge and determinants of IMDs' impact on their life were also explored. METHODS: An online cross-sectional survey was conducted. Respondents' IMD history, whether they received instructions for use and IMD's overall impact on life were recorded by self-reports. Patients' knowledge about living with IMDs was assessed on visual analogue scales (VAS, 0-10). Shared decision-making was analyzed by the 9-item Shared Decision Making Questionnaire (SDM-Q-9). Descriptive statistics and subgroup comparisons between IMD wearers were performed for statistical differences. Significant determinants of IMD's overall impact on life were examined in linear regression analysis. RESULTS: In the total sample (N = 1400, mean age 58.1 ±11.1; female 53.7%), nearly one third of respondents were living with IMD (30.9%; 433/1400). Among them, the most frequent IMDs were tooth implants (30.9%) and intraocular lens (26.8%). Mean knowledge VAS scores were similar (range: 5.5 ±3.8-6.5 ±3.2) but differences by IMD types were observed. Patients who received instructions for use or reported better impact on life indicated higher self-reported knowledge. Regression confirmed that patients' knowledge was significant predictor of IMD's impact on life, but this effect was overwritten by the SDM-Q-9. CONCLUSIONS: This first comprehensive epidemiological study on IMDs provides basic data for public health strategy planning alongside the implementation of MDR. Improved self-perceived outcomes were associated with higher knowledge hence education of patients receiving IMD deserves consideration. We suggest to investigate further the role of shared decision-making on IMD's overall impact on patients' life in future prospective studies.

Validation of the Musculoskeletal Health Questionnaire in a general population sample: a cross-sectional online survey in Hungary.

BACKGROUND: The Versus Arthritis Musculoskeletal Health Questionnaire (MSK-HQ) measures symptom severity and health-related quality of life (HRQoL) of people with musculoskeletal (MSK) conditions. We aimed to test the psychometric properties of the MSK-HQ among the general adult population and identify the determinants of MSK-HQ states. In addition, we aimed to explore the relationship between MSK-HQ and standard well-being measurement tools. METHODS: The translation proccess of the MSK-HQ into Hungarian followed the standard methods provided by the developer. A cross-sectional online survey was performed in Hungary involving a population normative sample (N = 2004, women: 53.1%; mean age: 48.3, SD = 16.6 years). Socio-demographic characteristics and self-reported MSK disorders were recorded. Alongside the MSK-HQ, standard measures of HRQoL (EQ-5D-5L), physical functioning (HAQ-DI) and well-being (ICECAP-A/O, WHO-5, Happiness VAS) were applied. Clinical and convergent validity were assessed by subgroup comparisons (Mann-Whitney-U and Kruskal-Wallis tests) and Spearman's rank correlations. Internal consistency was assessed by Cronbach's alpha. Test-retest reliability (N = 50) was evaluated by intraclass correlation coefficient (ICC). Predictors of MSK-HQ were analysed by ordinary least square multiple regressions. RESULTS: The mean MSK-HQ index score was 44.1 (SD = 9.9). MSK-HQ scores were significantly lower in subgroups with self-reported MSK disorders. Correlations were strong between MSK-HQ and EQ-5D-5L (0.788), EQ VAS (0.644) and HAQ-DI (-0.698) and moderate with the well-being measures (p < 0.05). Cronbach's alpha was 0.924 and ICC was 0.936 (p < 0.05). Being a man, living in the capital, having higher income and education were positively associated with MSK-HQ scores. CONCLUSIONS: This is the first study to prove the validity and reliability of the MSK-HQ among the general public. The impact of socio-demographic characteristics on MSK-HQ scores deserves consideration in clinical studies.

Development, testing, and implementation of a new procedure to assess the clinical added benefit of pharmaceuticals.

OBJECTIVES: The reimbursement process for innovative health technologies in Hungary lacks any formalized assessment of clinical added benefit (CAB). The aim of this research is to present the development, retrospective testing, and implementation of a local assessment framework for determining the CAB of cancer treatments at the Department of Health Technology Assessment of the National Institute of Pharmacy and Nutrition in Hungary. METHODS: The assessment framework was drafted after screening existing methods and a retrospective comparison of local reimbursement dossiers to that of German and French methods. The Magnitude of Clinical Benefit Scale of the European Society for Medical Oncology was chosen to rate the extent of CAB in oncology, as part of a conclusion complemented by the assessment of endpoint relevance and the quality of evidence. Several rounds of retrospective assessments have been conducted involving all clinical assessors, iterated with semistructured discussions to consolidate divergence between assessors. External stakeholders were consulted to provide feedback on the framework. RESULTS: Retrospective assessments resulted in average more than 75 percent concordance between assessors on each element of the conclusion. Input from ten stakeholders was also incorporated; stakeholders were generally supportive, and they mostly commented on the concept, the elements of the framework, and its implementation. CONCLUSIONS: The procedure is suitable for routine use in the decision-making process to describe the CAB of antineoplastic technologies in Hungary. Further extension of the framework is required to cover more disease areas for structured and comparable conclusions on CAB of innovative health technologies.

A critical assessment framework to identify, quantify and interpret the sources of uncertainty in cost-effectiveness analyses.

BACKGROUND: Using a standardized approach to describe the sources of uncertainty in cost-effectiveness analyses might bring added value to the local critical assessment procedure of reimbursement submissions in Hungary. The aim of this research is to present a procedural framework to identify, quantify and interpret sources of uncertainty, using the reimbursement dossier of darolutamide as an illustrative example. METHODS: In the procedural framework designed for the critical assessment of cost-effectiveness analyses, the quantifiability of an identified source of uncertainty is assessed through the input parameters of the originally submitted model, which is followed by the interpretation of its impact on estimates of costs and outcomes compared to the base case cost-effectiveness conclusion. RESULTS: Based on our experiences with the recent reimbursement dossier of darolutamide, the significant and quantifiable sources of uncertainty were the time horizon of the economic analysis; the restriction of the efficacy analysis population; long-term relative effectiveness of darolutamide; price discount on subsequent therapies. We identified resource use patterns for comparator and subsequent therapies as a quantifiable, yet non-significant source of uncertainty. The EQ-5D value set used to estimate utility values was identified as a non-quantifiable and potentially not significant source of uncertainty. CONCLUSIONS: The procedural framework, demonstrated with an example, was sufficiently flexible and coherent to document and structure the sources of uncertainty in cost-effectiveness analyses. The full-scale use of this framework is desirable during the decision-making process for reimbursement in Hungary. The further formalization of identifying sources of uncertainty is a possible subject of methodological development.

Patient-centred care in Hungary: Contributions to foster a policy agenda.

In Hungary, the National Health Insurance Fund provides health care coverage for nearly all residents, but healthcare spending is below the EU's average (6.4% versus 9.9% of the GDP in 2019, respectively). In 1997, patients' rights were established by laws of the healthcare system. The patients' voice, however, has remained weakly embedded in decision-making processes both on the system and individual patient levels. Policy progress achieved in the past years may foster patient-centeredness in health policy decision-making. However, people-reported data are not yet embedded in the Hungarian health information system and national population or household surveys, thus undermining the monitoring of the performance of the health system regarding patient-centred aspects. From the academic research side, several advances have occurred regarding the availability of validated instruments for the measurement of patient-centred aspects. These recent studies have placed Hungary in a uniquely advanced position compared with other countries in the Central and Eastern European (CEE) region. The use of those instruments in clinical guidelines and practices, to the education curricula of future health workers, is still in an early stage.

[Relative frequency of urinary tract infections in patients affected by diabetes mellitus type 2 treated with metformin and SGLT2 inhibitor. Network meta-analysis]. / A húgyúti fertozések relatív gyakorisága metforminnal és SGLT2-gátlóval kezelt 2-es típusú diabetes mellitusban szenvedo betegekben. Hálózati metaanalízis.

Introduction and aim: The of this research was to conduct a network meta-analysis based on a systematic literature search to compare the relative frequency of urinary tract infections using sodium-glucose cotransporter-2 (SGLT2) inhibitors combined with metformin in the therapy of type 2 diabetes. Method: MEDLINE and EMBASE databases were searched to identify publications of randomized, controlled trials investigating SGLT2 inhibitors combined with metformin in the therapy of type 2 diabetes and providing information on the frequency of urinary tract infections. Results: 10 165 unique citations were screened to identify 10 publications to be included in the network meta-analysis. The network meta-analysis showed reduced risk of urinary tract infections for low-dose ertugliflozin compared to other SGLT2 inhibitors (ertugliflozin 5 mg vs. empagliflozin 10 mg: RR: 0.606, 95% CrI: 0.264-1.415; ertugliflozin 5 mg vs. dapagliflozin 10 mg: RR = 0.853, 95% CrI: 0.301-2.285). For high-dose comparisons, empagliflozin 25 mg showed reduced risk of urinary tract infections compared to both ertugliflozin 15 mg (RR = 0.745, 95% CrI 0.330-1.610) and dapagliflozin 10 mg (RR = 0.680, 95% CrI: 0.337-1.289). The difference between active substances and their doses was not statistically significant for the relative frequency of urinary tract infections. The meta-regression revealed a statistically significant association between baseline fasting plasma glucose level and relative frequency of urinary tract infections (ß = 0.785, 95% CrI: 0.062-1.587). Conclusion: There was no statistically significant difference between SGLT2 inhibitors investigated in this study in terms of the relative frequency of urinary tract infections. This research demonstrates the applicability of network meta-analyses when assessing the relative effectiveness and safety of interventions. Orv Hetil. 2020; 161(13): 491-501.