lence of comorbid conditions. There is substantial 1-year mean cost associated with pediatric ADHD, even for stable responders on their pharmacotherapy. Notably, only one quarter of the cost of ADHD is due to medication.

The purpose of this investigation was to evaluate the budget impact and cost-effectiveness of direct-acting antivirals (DAAs) for the treatment of hepatitis C virus (HCV) infection in Hong Kong. A decision analytic model was developed to compare short-term costs and health outcomes of patients with chronic HCV genotype 1 infection in Hong Kong who were treated with an interferon (INF)-based treatment (dual therapy of pegylated interferon and ribavirin) or DAA-based treatments (sofosbuvir or ledipasvir/sofosbuvir or ombitasvir/paritaprevir/ritonavir plus dasabuvir). Compared to INF-based treatment, DAA-based treatments yielded an incremental cost of $24,677-$31,171 per course while improving the rate of sustained virologic response (SVR) from 59-66% to 82.3-99.8%. The incremental cost-effective ratios of DAA-based treatments ranged from $9724 to $29,189 per treatment success, which were all below the cost-effectiveness threshold of local GDP per capita ($42,423 in 2015). Introducing DAAs resulted in a 126.1% ($383.7 million) budget increase on HCV infection management over 5 years. A 50% change in DAA medication costs reflected a change in the incremental budget from $55.2 to $712.3 million. DAA-based treatments are cost-effective alternatives to INF-based treatment in Hong Kong. Introducing DAAs to the public hospital formulary yields a considerable budget increase but is still economically favorable to the local government.

No abstract

about contemporary cost of illness (COI) studies and understand the full burden of patients with recent-onset and established RA. METHODS: We performed a systematic review of all COI studies of RA, published after the year 2000. Multiple databases using relevant keywords and search filters were searched. We synthesised the data from all relevant studies through a narrative synthesis, with specific focus on the impact of the various settings and methodological approaches used to estimate costs in the studies. RESULTS: There were substantial variations in patient casemix, healthcare settings, and methodological approaches to estimating COI of RA. The majority of the studies adopted the prevalence-based approach from a societal perspective to estimate an annual cost per patient. Direct costs tended to increase over the past decade mainly because of the introduction of biologics. However, this rising drug costs reached a plateau around 2009 and was subsequently offset by a general reduction in hospitalisation and in productivity loss. Productivity loss was mostly estimated by the human capital approach, accounted for the largest part of total annual cost, (45%-70% of total costs in the Western society). CONCLUSIONS: In the management of RA, early and intensive treatment may increase the short-term direct costs, but in the longer-term, this was associated with reduced hospitalisation and productivity loss in the society.

Objective Frailty may be found in heart failure patients especially in the elderly and is associated with a poor prognosis. However, assessment of frailty status is time-consuming and the electronic frailty indices developed using health records have served as useful surrogates. We hypothesized that an electronic frailty index developed using machine learning can improve short-term mortality prediction in patients with heart failure. Methods This was a retrospective observational study included patients admitted to nine public hospitals for heart failure from Hong Kong between 2013 and 2017. Age, sex, variables in the modified frailty index, Deyo's Charlson comorbidity index (≥2), neutrophil-to-lymphocyte ratio (NLR) and prognostic nutritional index (PNI) were analyzed. Gradient boosting, which is a supervised sequential ensemble learning algorithm with weak prediction submodels (typically decision trees), was applied to predict mortality. Variables were ranked in the order of importance with a total score of 100 and used to build the predictive models. Comparisons were made with decision tree and multivariate logistic regression. Results A total of 8893 patients (median: age 81, Q1-Q3: 71–87 years old) were included, in whom 9% had 30-day mortality and 17% had 90-day mortality. PNI, age and NLR were the most important variables predicting 30-day mortality (importance score: 37.4, 32.1, 20.5, respectively) and 90-day mortality (importance score: 35.3, 36.3, 14.6, respectively). Gradient boosting significantly outperformed decision tree and multivariate logistic regression (area under the curve: 0.90, 0.86 and 0.86 for 30-day mortality; 0.92, 0.89 and 0.86 for 90-day mortality). Conclusions The electronic frailty index based on comorbidities, inflammation and nutrition information can readily predict mortality outcomes. Their predictive performances were significantly improved by gradient boosting techniques. Funding Acknowledgement Type of funding sources: None.