The thalassaemia syndromes (TS) show different phenotype severity. Developing a reliable, practical and global tool to determine disease severity and tailor treatment would be of great value. Overall, 7910 patients were analysed with the aim of constructing a complication risk score (CoRS) to evaluate the probability of developing one or more complications. Nine independent variables were included in the investigation as predictors. Logistic regression models were used for Group A [transfusion-dependent thalassaemia (TDT)], Group B [transfused non-TDT (NTDT)] and Group C (non-transfused NTDT). Statistically significant predictors included age (years), haemoglobin levels, hepatic transaminases [alanine aminotransferase (ALT) and aspartate aminotransferase] and left-ventricular ejection fraction (LVEF) for Group A; age (years), age at first chelation (months), ALT and LVEF for Group B; and age (years), mean serum ferritin (SF) levels and LVEF for Group C. The area under the receiver operating characteristic curve was 84·5%, 82·1% and 80·0% for Groups A, Group B and Group C respectively, suggesting the models had good discrimination. Finally, the CoRS for each group was categorised into four risk classes (low, intermediate, high, and very high) using the centiles of its distribution. In conclusion, we have developed a CoRS for TS that can assist physicians in prospectively tailoring patients' treatment.

A prognostic scoring system that can differentiate β-thalassemia patients based on mortality risk is lacking. We analysed data from 3145 β-thalassemia patients followed through a retrospective cohort design for the outcome of death. An a priori list of prognostic variables was collected. β Coefficients from a multivariate cox regression model were used from a development dataset (n = 2516) to construct a formula for a Thalassemia International Prognostic Scoring System (TIPSS) which was subsequently applied to a validation dataset (n = 629). The median duration of observation was 10.0 years. The TIPSS score formula was constructed as exp (1.4 × heart disease + 0.9 × liver disease + 0.9 × diabetes + 0.9 × sepsis + 0.6 × alanine aminotransferase ≥42 IU/L + 0.6 × hemoglobin ≤9 g/dL + 0.4 × serum ferritin ≥1850 ng/mL). TIPSS score thresholds of greatest differentiation were assigned as <2.0 (low-risk), 2.0 to <5.0 (intermediate-risk), and ≥5.0 (high-risk). The TIPSS score was a good predictor for the outcome of death in the validation dataset (AUC: 0.722, 95%CI: 0.641-0.804) and survival was significantly different between patients in the three risk categories (P < 0.001). Compared to low-risk patients, the hazard ratio for death was 2.778 (95%CI: 1.335-5.780) in patients with intermediate-risk and 6.431 (95%CI: 3.151-13.128) in patients with high-risk. This study provides a novel tool to support mortality risk categorization for patients with β-thalassemia that could help management and research decisions.

Health economic evaluations (HEEs) help healthcare decision makers understand the value of new technologies. Artificial intelligence (AI) is increasingly being used in healthcare interventions. We sought to review the conduct and reporting of published HEEs for AI-based health interventions. We conducted a systematic literature review with a 15-month search window (April 2021 to June 2022) on 17 June 2022 to identify HEEs of AI health interventions and update a previous review. Records were identified from 3 databases (Medline, Embase, and Cochrane Central). Two reviewers screened papers against predefined study selection criteria. Data were extracted from included studies using prespecified data extraction tables. Included studies were quality assessed using the National Institute for Health and Care Excellence (NICE) checklist. Results were synthesized narratively. A total of 21 studies were included. The most common type of AI intervention was automated image analysis (9/21, 43%) mainly used for screening or diagnosis in general medicine and oncology. Nearly all were cost-utility (10/21, 48%) or cost-effectiveness analyses (8/21, 38%) that took a healthcare system or payer perspective. Decision-analytic models were used in 16/21 (76%) studies, mostly Markov models and decision trees. Three (3/16, 19%) used a short-term decision tree followed by a longer-term Markov component. Thirteen studies (13/21, 62%) reported the AI intervention to be cost effective or dominant. Limitations tended to result from the input data, authorship conflicts of interest, and a lack of transparent reporting, especially regarding the AI nature of the intervention. Published HEEs of AI-based health interventions are rapidly increasing in number. Despite the potentially innovative nature of AI, most have used traditional methods like Markov models or decision trees. Most attempted to assess the impact on quality of life to present the cost per QALY gained. However, studies have not been comprehensively reported. Specific reporting standards for the economic evaluation of AI interventions would help improve transparency and promote their usefulness for decision making. This is fundamental for reimbursement decisions, which in turn will generate the necessary data to develop flexible models better suited to capturing the potentially dynamic nature of AI interventions.

OBJECTIVE: Inflammation has been associated with an increased risk for cancer development, while innate immune system activation could counteract the risk for malignancies. Familial Mediterranean fever (FMF) is a severe systemic inflammatory condition and also represents the archetype of innate immunity deregulation. Therefore, the aim of this study is to investigate the risk for cancer development in FMF.

METHODS: The risk ratio (RR) for malignancies was separately compared between FMF patients and fibromyalgia subjects, Still's disease patients and Behçet's disease patients. Clinical variables associated with cancer development in FMF patients were searched through binary logistic regression.

RESULTS: 580 FMF patients and 102 fibromyalgia subjects, 1012 Behçet's disease patients and 497 Still's disease patients were enrolled. The RR for the occurrence of malignant neoplasms was 0.26 (95% Confidence Interval [CI.] 0.10-0.73, p=0.006) in patients with FMF compared to fibromyalgia subjects; the RR for the occurrence of malignant cancer was 0.51 (95% CI. 0.23-1.16, =0.10) in FMF compared to Still's disease and 0.60 (95% CI. 0.29-1.28, =0.18) in FMF compared to Behçet's disease. At logistic regression, the risk of occurrence of malignant neoplasms in FMF patients was associated with the age at disease onset (β1 = 0.039, 95% CI. 0.001-0.071, =0.02), the age at the diagnosis (β1 = 0.048, 95% CI. 0.039-0.085, =0.006), the age at the enrolment (β1 = 0.05, 95% CI. 0.007-0.068, =0.01), the number of attacks per year (β1 = 0.011, 95% CI. 0.001- 0.019, =0.008), the use of biotechnological agents (β1 = 1.77, 95% CI. 0.43-3.19, =0.009), the use of anti-IL-1 agents (β1 = 2.089, 95% CI. 0.7-3.5, =0.002).

CONCLUSIONS: The risk for cancer is reduced in Caucasic FMF patients; however, when malignant neoplasms occur, this is more frequent in FMF cases suffering from a severe disease phenotype and presenting a colchicine-resistant disease.