Improving Keyword-Based Topic Classification in Cancer Patient Forums with Multilingual Transformers.

Online forums play an important role in connecting people who have crossed paths with cancer. These communities create networks of mutual support that cover different cancer-related topics, containing an extensive amount of heterogeneous information that can be mined to get useful insights. This work presents a case study where users' posts from an Italian cancer patient community have been classified combining both count-based and prediction-based representations to identify discussion topics, with the aim of improving message reviewing and filtering. We demonstrate that pairing simple bag-of-words representations based on keywords matching with pre-trained contextual embeddings significantly improves the overall quality of the predictions and allows the model to handle ambiguities and misspellings. By using non-English real-world data, we also investigated the reusability of pretrained multilingual models like BERT in lower data regimes like many local medical institutions.

A continuous-time Markov model approach for modeling myelodysplastic syndromes progression from cross-sectional data.

The integration of both genomics and clinical data to model disease progression is now possible, thanks to the increasing availability of molecular patients' profiles. This may lead to the definition of novel decision support tools, able to tailor therapeutic interventions on the basis of a "precise" patients' risk stratification, given their health status evolution. However, longitudinal analysis requires long-term data collection and curation, which can be time demanding, expensive and sometimes unfeasible. Here we present a clinical decision support framework that combines the simulation of disease progression from cross-sectional data with a Markov model that exploits continuous-time transition probabilities derived from Cox regression. Trajectories between patients at different disease stages are stochastically built according to a measure of patient similarity, computed with a matrix tri-factorization technique. Such trajectories are seen as realizations drawn from the stochastic process driving the transitions between the disease stages. Eventually, Markov models applied to the resulting longitudinal dataset highlight potentially relevant clinical information. We applied our method to cross-sectional genomic and clinical data from a cohort of Myelodysplastic syndromes (MDS) patients. MDS are heterogeneous clonal hematopoietic disorders whose patients are characterized by different risks of Acute Myeloid Leukemia (AML) development, defined by an international score. We computed patients' trajectories across increasing and subsequent levels of risk of developing AML, and we applied a Cox model to the simulated longitudinal dataset to assess whether genomic characteristics could be associated with a higher or lower probability of disease progression. We then used the learned parameters of such Cox model to calculate the transition probabilities of a continuous-time Markov model that describes the patients' evolution across stages. Our results are in most cases confirmed by previous studies, thus demonstrating that simulated longitudinal data represent a valuable resource to investigate disease progression of MDS patients.

Trusting telemedicine: A discussion on risks, safety, legal implications and liability of involved stakeholders.

OBJECTIVES: The main purpose of the article is to raise awareness among all the involved stakeholders about the risks and legal implications connected to the development and use of modern telemedicine systems. Particular focus is given to the class of "active" telemedicine systems, that imply a real-world, non-mediated, interaction with the final user. A secondary objective is to give an overview of the European legal framework that applies to these systems, in the effort to avoid defensive medicine practices and fears, which might be a barrier to their broader adoption. METHODS: We leverage on the experience gained during two international telemedicine projects, namely MobiGuide (pilot studies conducted in Spain and Italy) and AP@home (clinical trials enrolled patients in Italy, France, the Netherlands, United Kingdom, Austria and Germany), whose development our group has significantly contributed to in the last 4 years, to create a map of the potential criticalities of active telemedicine systems and comment upon the legal framework that applies to them. Two workshops have been organized in December 2015 and March 2016 where the topic has been discussed in round tables with system developers, researchers, physicians, nurses, legal experts, healthcare economists and administrators. RESULTS: We identified 8 features that generate relevant risks from our example use cases. These features generalize to a broad set of telemedicine applications, and suggest insights on possible risk mitigation strategies. We also discuss the relevant European legal framework that regulate this class of systems, providing pointers to specific norms and highlighting possible liability profiles for involved stakeholders. CONCLUSIONS: Patients are more and more willing to adopt telemedicine systems to improve home care and day-by-day self-management. An essential step towards a broader adoption of these systems consists in increasing their compliance with existing regulations and better defining responsibilities for all the involved stakeholders.

Standardization of immunotherapy adverse events in patient information leaflets and development of an interface terminology for outpatients' monitoring.

INTRODUCTION: Immunotherapy is effective for treating cancer, but it is also associated with a wide spectrum of adverse events. In order to detect them early, the patients need to be monitored at home, between the therapy administrations, e.g., by asking them to report outcomes, usually including symptoms and quality of life measures. For the collected data to be reusable, the symptoms need to be in a standardized form. The aim of this study is to explore the standardization of the information contained in the patient information leaflets (PILs) of immunotherapy drugs, by creating an interface terminology of immunotherapy-related adverse events, which should support a consistent collection of symptoms from the patients. METHODS: PILs contain a significant amount of information in free text, but they mix patient-reportable and clinically assessable events. We extracted a list of patient-reportable adverse events, mapped them to reference terminologies and compared the mapping results to choose the best-performing reference terminology. RESULTS: The PILs standardization led to the extraction of 151 symptoms and 424 terms, including both preferred terms and synonyms in English and Italian. Among the reference terminologies we considered, SNOMED CT allowed us to map all concepts and became, hence, the main reference terminology for the resulting interface terminology. A preliminary validation on the PIL of a new immunotherapy drug showed that our interface terminology already contained all the mentioned symptoms. CONCLUSION: PILs provide a valuable source for determining adverse events. The resulting interface terminology includes Italian and English terms for patient-reportable adverse events for five immunotherapy drugs representative of their category. Further work will be undertaken to evaluate the usability of the interface terminology and the patients' experience and satisfaction with the proposed terms, made available for example through an app, as well as its effectiveness on data quality and quality of care.

Parental evaluation of a telemonitoring service for children with Type 1 Diabetes.

Introduction In the past years, we developed a telemonitoring service for young patients affected by Type 1 Diabetes. The service provides data to the clinical staff and offers an important tool to the parents, that are able to oversee in real time their children. The aim of this work was to analyze the parents' perceived usefulness of the service. Methods The service was tested by the parents of 31 children enrolled in a seven-day clinical trial during a summer camp. To study the parents' perception we proposed and analyzed two questionnaires. A baseline questionnaire focused on the daily management and implications of their children's diabetes, while a post-study one measured the perceived benefits of telemonitoring. Questionnaires also included free text comment spaces. Results Analysis of the baseline questionnaires underlined the parents' suffering and fatigue: 51% of total responses showed a negative tendency and the mean value of the perceived quality of life was 64.13 in a 0-100 scale. In the post-study questionnaires about half of the parents believed in a possible improvement adopting telemonitoring. Moreover, the foreseen improvement in quality of life was significant, increasing from 64.13 to 78.39 ( p-value = 0.0001). The analysis of free text comments highlighted an improvement in mood, and parents' commitment was also proved by their willingness to pay for the service (median = 200 euro/year). Discussion A high number of parents appreciated the telemonitoring service and were confident that it could improve communication with physicians as well as the family's own peace of mind.

Cord and blood levels of newborn IgE: Correlation, role and influence of maternal IgE.

The role of cord blood immunoglobulin E (IgE) levels in predicting the development of atopy has been widely investigated. The aim of the study was to evaluate the correlation between serum and cord blood total IgE in newborns and the possible influence of the atopic status of the mother on them. It was also investigated the possible role of gestational age on neonatal total IgE levels. We considered 763 deliveries, 724≥37 weeks of gestation and 39<37 weeks of gestation. 14% of mothers (13.7% at term, 15.4% preterm) showed high total IgE levels. The results showed a significant correlation between serum and cord IgE levels both in preterm and term newborns. The data revealed also that mother's total IgE levels affect both neonatal serum and cord total IgE levels. For the latters we also found child gender as an additional independent predictor. On the contrary total IgE levels are not affected by gestational age. Clinical limitations of total IgE is known but their determination can be useful to define atopy and to suggest follow-up of the children.

Profiles of Birch Sensitization (Bet v 1, Bet v 2, and Bet v 4) and Oral Allergy Syndrome Across Italy.

BACKGROUND: Birch allergy (BA) is a common pollinosis caused by the allergens Bet v 1, Bet v 2, and Bet v 4. Oral allergy syndrome (OAS) is frequently associated with BA. A gradient of sensitization to birch allergen across Europe has been reported. Therefore, this study aimed to investigate the birch sensitization profile, including OAS, across Italy. METHODS: We performed a retrospective study of 854 patients (391 males, mean age 35.9 years, range 18-93 years): 196 patients were recruited in Genoa, 188 in northern Italy, 359 in central Italy, and 111 in southern Italy. Serum IgE to Bet v 1, Bet v 2, and Bet v 4 was assessed, and OAS was analyzed. RESULTS: With respect to the geographical path Genoa-North-Center-South, the frequency of sensitization to Bet v 1 decreased significantly (P<.0001) from Genoa (95.41%) to southern Italy (58.56%). The frequency of sensitization to Bet v 2 increased significantly (P<.0001) from Genoa (6.12%) to southern Italy (52.25%). The frequency of Bet v 4 also increased significantly (P=.0002) from Genoa (6.12%) to southern Italy (14.41%). The distribution of patients with OAS differed significantly across the areas (P<.0001), the most marked difference ranging between 33.5% in Genoa and 76.9% in northern Italy. The frequency of birch allergens correlated with OAS in central Italy only. CONCLUSIONS: The present study demonstrated a significant difference between sensitization to birch and its clinical expression across Italy.

Profiles of birch Sensitization (Bet v 1, Bet v 2, and Bet v 4) and oral allergy syndrome acrsss Italy

Background: Birch allergy (BA) is a common pollinosis caused by the allergens Bet v 1, Bet v 2, and Bet v 4. Oral allergy syndrome (OAS) is frequently associated with BA. A gradient of sensitization to birch allergen across Europe has been reported. Therefore, this study aimed to investigate the birch sensitization profile, including OAS, across Italy. Methods: We performed a retrospective study of 854 patients (391 males, mean age 35.9 years, range 18-93 years): 196 patients were recruited in Genoa, 188 in northern Italy, 359 in central Italy, and 111 in southern Italy. Serum IgE to Bet v 1, Bet v 2, and Bet v 4 was assessed, and OAS was analyzed. Results: With respect to the geographical path Genoa-North-Center-South, the frequency of sensitization to Bet v 1 decreased significantly (P<.0001) from Genoa (95.41%) to southern Italy (58.56%). The frequency of sensitization to Bet v 2 increased significantly (P<.0001) from Genoa (6.12%) to southern Italy (52.25%). The frequency of Bet v 4 also increased significantly (P=.0002) from Genoa (6.12%) to southern Italy (14.41%). The distribution of patients with OAS differed significantly across the areas (P<.0001), the most marked difference ranging between 33.5% in Genoa and 76.9% in northern Italy. The frequency of birch allergens correlated with OAS in central Italy only. Conclusions: The present study demonstrated a significant difference between sensitization to birch and its clinical expression across Italy (AU)

Introducción: El síndrome de alergia oral (SAO) se encuentra frecuentemente asociado a la alergia al polen de abedul. Se han descrito diferentes gradientes de sensibilización a polen de abedul en Europa. Este estudio pretende determinar el perfil de sensibilización a polen de abedul, incluyendo la presencia de SAO, en Italia. Métodos: Estudio retrospectivo con 854 pacientes (391 hombres, edad media 35,9 años, rango 18-93 años): 196 pacientes procedían de Génova, 188 del Norte de Italia, 359 de Italia Central y 111 del Sur de Italia. Se determinó la IgE específica a Bet v 1, Bet v 2 y Bet v 4, así como la presencia de SAO. Resultados: De acuerdo a la procedencia geográfica Génova-Norte-Centro-Sur de Italia, la sensibilización a Bet v 1 disminuye significativamente (P<0,0001) desde Génova (95,41%) hasta el Sur de Italia (58,56%). La sensibilización a Bet v 2 aumenta significativamente (P<0,0001) desde Génova (6,12%) hasta el Sur de Italia (52,25%). También la sensibilización a Bet v 4 aumenta significativamente (P<0,0002) desde Génova (6,12%) hasta el Sur de Italia (14,41%). Existe una distribución del SAO significativamente diferente entre las diferentes áreas geográficas consideradas, siendo la máxima diferencia la presentada entre Génova (33,5%) y el Norte de Italia (76,9%). Las frecuencias de sensibilización a las diferentes moléculas del polen de abedul se correlacionan con el SAO solo en la región Central de Italia. Conclusiones: El presente estudio demuestra la existencia de diferencias significativas entre la sensibilización a las diferentes moléculas del polen de abedul y su expresión clínica en diferentes regiones italianas (AU)

TRYPTASE AND HISTAMINE MAY SUPPORT ORAL FOOD CHALLENGE IN THE DIAGNOSIS OF ALLERGY.

Allergen-specific immunoglobulin E (IgE) reactions lead to acute degranulation of mast cells and basophils and release of stored mediators, particularly tryptase and histamine, which can be measured in vitro after reactions. The aim of this study was to investigate the utility of serum tryptase and plasma histamine during oral food challenge (OFC) in 103 children with suspected food allergy, in order to support the diagnosis of a IgE-mediated reaction. Blood samples for serum tryptase and plasma histamine were collected before the OFC and after the onset of allergic symptoms or after 60 minutes from test completion. Serum tryptase and plasma histamine were measured by a fluoroenzyme immunoassay (ImmunoCAP; ThermoFisher, Uppsala, Sweden) according to the manufacturer’s instructions. A correlation between serum tryptase and plasma histamine distributions was observed after OFC (p=0.0035). A correlation was also observed for both serum tryptase and plasma histamine before and after OFC (p less than0.0001). Subjects with positive response to OFC had significantly higher values (p = 0.0375) of serum tryptase compared to subjects with negative response. The plasma histamine distribution showed a significant difference between measurements before and after OFC, both in the complete population (p less than 0.0001), and considering the response (negative OFC: p less than 0.0001; positive OFC: p=0.0181). The diagnostic work-up of IgE- mediated food allergy may include determination of serum tryptase and plasma histamine, in order to support the results of OFC. These markers are strongly related to the same IgE-mediated mechanism and, as they can be both easily measured, can confirm the allergic nature of a reaction in the real-life setting of food allergy.

WALNUT SENSITIZATION IN PEDIATRIC AGE: A PRELIMINARY STUDY.

Walnut consumption has recently become a healthy dietary habit worldwide, due to its positive benefits in reducing cholesterol levels and oxidative stress; this has resulted in an increase in individual consumption, global production and risk of developing sensitization and allergy. In general, clinical manifestations of walnut allergy are frequently severe and systemic potentially life-threatening, leading to anaphylaxis both in the pediatric and adult populations. In light of these findings, we performed a preliminary study considering the walnut native allergen and the recombinant Jug r1 in order to evaluate their role in atopic diseases.

Personalization and Patient Involvement in Decision Support Systems: Current Trends.

OBJECTIVES: This survey aims at highlighting the latest trends (2012-2014) on the development, use, and evaluation of Information and Communication Technologies (ICT) based decision support systems (DSSs) in medicine, with a particular focus on patient-centered and personalized care. METHODS: We considered papers published on scientific journals, by querying PubMed and Web of ScienceTM. Included studies focused on the implementation or evaluation of ICT-based tools used in clinical practice. A separate search was performed on computerized physician order entry systems (CPOEs), since they are increasingly embedding patient-tailored decision support. RESULTS: We found 73 papers on DSSs (53 on specific ICT tools) and 72 papers on CPOEs. Although decision support through the delivery of recommendations is frequent (28/53 papers), our review highlighted also DSSs only based on efficient information presentation (25/53). Patient participation in making decisions is still limited (9/53), and mostly focused on risk communication. The most represented medical area is cancer (12%). Policy makers are beginning to be included among stakeholders (6/73), but integration with hospital information systems is still low. Concerning knowledge representation/management issues, we identified a trend towards building inference engines on top of standard data models. Most of the tools (57%) underwent a formal assessment study, even if half of them aimed at evaluating usability and not effectiveness. CONCLUSIONS: Overall, we have noticed interesting evolutions of medical DSSs to improve communication with the patient, consider the economic and organizational impact, and use standard models for knowledge representation. However, systems focusing on patient-centered care still do not seem to be available at large.

Predictors of renal survival in ANCA-associated vasculitis. Validation of a histopatological classification schema and review of the literature.

OBJECTIVES: In 2010 a histopathological classification of ANCA-associated glomerulonephritis was proposed to predict the outcomes at diagnosis. Our aim was to validate the proposed classification in our cohort of patients and to compare the studies already published. METHODS: The data of 93 patients who underwent kidney biopsy in a single Italian centre within 15 years were retrospectively collected. RESULTS: The 10-year renal and patients' survival were 60% and 81%, respectively. Biopsies were classified as 21% focal, 30% crescentic, 39% mixed and 10% sclerotic. Survival without ESRD at 5 years was 82% in focal, 37% in crescentic, 81% in mixed and 51% in sclerotic group. The Kaplan-Meier analysis highlights that renal survival was not different between sclerotic and crescentic groups (p=0.9) but both had a significantly worse prognosis than focal (p=0.04 and 0.015 respectively) and mixed groups (p=0.05 and 0.03 respectively). Focal and mixed groups had the same renal survival (p=0.7). At multivariate analysis the independent predictors of end-stage renal disease were less than 20% of normal glomeruli at kidney biopsy (p=0.022), high serum creatinine (p=0.009) and arterial hypertension at presentation (p= 0.006). CONCLUSIONS: In our cohort, the proposed histological classification was not predictive of renal prognosis. The focal and the mixed classes had the same prognosis and a significantly better renal outcome than both the crescentic and the sclerotic classes. At multivariate analysis among the histological features only less than 20% of normal glomeruli defines the renal prognosis together with renal function and arterial hypertension at baseline.

UceWeb: a web-based collaborative tool for collecting and sharing quality of life data.

OBJECTIVES: This work aims at building a platform where quality-of-life data, namely utility coefficients, can be elicited not only for immediate use, but also systematically stored together with patient profiles to build a public repository to be further exploited in studies on specific target populations (e.g. cost/utility analyses). METHODS: We capitalized on utility theory and previous experience to define a set of desirable features such a tool should show to facilitate sound elicitation of quality of life. A set of visualization tools and algorithms has been developed to this purpose. To make it easily accessible for potential users, the software has been designed as a web application. A pilot validation study has been performed on 20 atrial fibrillation patients. RESULTS: A collaborative platform, UceWeb, has been developed and tested. It implements the standard gamble, time trade-off and rating-scale utility elicitation methods. It allows doctors and patients to choose the mode of interaction to maximize patients' comfort in answering difficult questions. Every utility elicitation may contribute to the growth of the repository. CONCLUSION: UceWeb can become a unique source of data allowing researchers both to perform more reliable comparisons among healthcare interventions and build statistical models to gain deeper insight into quality of life data.

Gquest: modeling patient questionnaires and administering them through a mobile platform application.

BACKGROUND: The use of surveys is becoming popular in the health care industry for acquiring information useful to the accomplishment of several studies. Besides their exploitation on a large scale for conducting epidemiological studies, surveys are being increasingly carried out on a narrower perspective through the administration of questionnaires aimed at assessing the quality of life perceived by patients or their clinical status during mid- or long-term treatments. This is useful for managing resources or optimizing and individualizing treatments. OBJECTIVES: This paper describes Gquest, a platform for modeling, generating and administering questionnaires through mobile devices such as smartphones or tablets. Gquest was motivated by the need of administering questionnaires during home treatments, albeit its applicability is rather general. The main requirement was to have a very simple, clean and easy to use platform able to support (a) physicians in the design and delivery of questionnaires and (b) outpatients in sending self-recorded outcomes to the clinical staff. METHODS: Gquest has two basic components. The first one is a model devised for representing questionnaires which is extremely flexible. It allows the generation of questions and answers of different types, supports adaptivity in the dialog with the user and enforces simple consistency rules for checking his input. The second component is an application able to run instances of those questionnaires. It downloads questionnaires over the air in terms of XML files from a server and stores them locally into the mobile repository. Questionnaires become then available to the user, who in our case is a patient or one of his relatives. The user can select which one to fill-in, according to his needs and/or the specific treatment protocol. The selected questionnaire may be filled-in all at once or be completed in subsequent steps over time since any input provided is persisted on a local database. Finally, when a questionnaire is closed all the answers are transparently synchronized to a server for further evaluation or statistical purposes. RESULTS: The platform has been exploited and tested in two medical domains where questionnaires have very different purposes; a quality of life pilot survey involving patients with atrial fibrillation within the EU-funded project MobiGuide, and an investigation of the users' needs during the requirements elicitation phase of a telemedicine project for the safe discharge of fragile patients. CONCLUSIONS: Gquest proved to be a suitable tool for dealing with a wide variety of question types and has been positively accepted by the patients enrolled in the pilot study.

Progressive improvement of patient and renal survival and reduction of morbidity over time in patients with lupus nephritis (LN) followed for 20 years.

Whether the long-term patient and renal survival of those diagnosed with lupus nephritis (LN) has improved over the decades is still debated. Eighty-nine patients diagnosed between 1968 and 1990 entered this study and their outcome was evaluated after 20 years. At presentation 54% of patients had class IV LN, 39.3% had renal insufficiency and 59.5% had nephrotic syndrome. Patients were divided into two groups: Group 1 consisted of 30 patients diagnosed between 1968 and 1980; Group 2 consisted of 59 patients diagnosed between 1981 and 1990. In Group 1 patient survival at 20 years was 84% versus 95% in Group 2 (p=0.05). Survivals without end-stage renal failure were respectively 75% and 84% at 20 years (p=0.05). Survivals without severe infection at 20 years were 44% in Group 1 and 66.5% in Group 2 (p=0.02). Survivals without cardiovascular events at 20 years were: 53% in Group 1 and 90% in Group 2 (p=0.005). At presentation, patients in Group 1 had higher serum creatinine (1.96 vs 1.15 mg/dl, p=0.01), higher activity index (8 vs 5.5, p=0.01), lower hematocrit (31% v s6%, p=0.008) and lower serum C4 levels (p=0.04) than Group 2 patients. Patients in Group 1 also received less frequent methylprednisolone pulses (43% vs 81%, p=0.0006). In Italian patients with LN, long-term life expectancy and renal survival progressively improved over the decades, while morbidity progressively declined. An earlier referral and refinement of therapy achieved this goal.

Improving thrombolysis for acute ischemic stroke in Lombardia stroke centers.

The purpose of this study is to identify which factors are able to limit or hamper the access to systemic thrombolysis (evTPA) in Lombardia to define corrective interventions. We analyzed 1,015 patients with ischemic stroke admitted to emergency departments (ED) participating to the Lombardia Stroke Unit Registry and eligible for evTPA; 303 (29.9%) patients were treated with evTPA (evTPA+ group) and 712 (70.1%) were not (evTPA- group). We collected case-mix and stroke care process variables.The evTPA+ group was characterized by a shorter ED arrival time, a greater neurological impairment, a more chance to be admitted to ED linked to comprehensive stroke center (CSC) and a shorter waiting time to access to diagnostic procedures. The chance to be treated with evTPA was greater if neurological evaluation anticipated neuroimaging (p = 0.0003). The multivariate analysis confirmed that the admission to ED linked to CSC (OR: 2.50, 95% CI: 1.39-4.48, p < 0.0001) and neurological evaluation performed before neuroimaging (OR: 2.34, 95% CI: 1.35-4.04, p = 0.002) increased the probability to receive rtPA. The evTPA treatment is strictly dependent on pre-hospital and ED care process phases and strongly influenced by the degree of stroke severity. Door-to-needle time is shorter in patients with a greater stroke severity and a shorter ED arrival time. A 24-h/week availability of the neurologist in ED can increase the percentage of thrombolysis optimizing the selection of patients and the timing of the diagnostic procedures.

Tetanus toxoid IgE may be useful in predicting allergy during childhood.

Hypersensitivity reactions after immunization with tetanus toxoid are occasionally observed in atopic and non-atopic individuals. High IgE levels in infancy may predict subsequent allergy. The aims of this study were: i) to evaluate the role of specific IgE to tetanus toxoid in children in response to tetanus immunization and the possible factors associated with specific IgE levels, and ii) to investigate the correlation between specific IgE levels to tetanus toxoid and the late development of allergy (up to 12 years). Initially, 278 healthy infants (152 males and 126 females, aged 12 months) living in an urban city were screened for serum total IgE and specific IgE to tetanus toxoid, after having obtained informed consent from parents. After 12 years, 151 children could be evaluated. Total IgE summed with tetanus specific IgE were significantly associated with allergy at 12 years. In conclusion, this study demonstrates that serum total IgE and tetanus specific IgE may be predictive of subsequent allergy onset.

Adding docetaxel to cisplatin and fluorouracil in patients with unresectable head and neck cancer: a cost-utility analysis.

BACKGROUND: Adding docetaxel (Taxotere, T) to induction chemotherapy with platinum/infusional 5-FU (PF) has been shown to improve overall survival of patients with head and neck cancer. The aim of the study was to analyze the cost-utility of TPF in patients with unresectable disease. DESIGN: We developed a Markov model to represent patient's weekly transitions among different health states, related to treatment or disease status. Transition probabilities were obtained from the TAX 324 clinical trial report and from the European Organization for Research and Treatment of Cancer (EORTC) 24971/TAX 323 raw data. Costs were estimated in Italy from a Regional Healthcare System perspective. A 5-year temporal horizon was adopted and a 3.5% yearly discount rate was applied. RESULTS: When compared with PF, TPF treatment increases life expectancy by 0.33 quality-adjusted life-years (QALYs) in TAX 323 and 0.41 QALYs in TAX 324. The benefit was achieved at a cost of €11,822/QALY for TAX 323 and €6757/QALY for TAX 324. Monte Carlo sensitivity analysis showed that 69% (TAX 323) and 99% (TAX 324) of the results lie below the threshold of €50,000/QALY saved. CONCLUSIONS: In our analysis, TPF induction chemotherapy proved to be cost-effective when compared with PF, having a cost-utility ratio comparable to other widely accepted healthcare interventions.

Summary of Product Characteristics content extraction for a safe drugs usage.

The use of medications has a central role in health care provision, yet on occasion, it may injure the person taking them as result of adverse drug events. A correct drug choice must be modulated to acknowledge both patients' status and drug-specific information. However, this information is locked in free-text and, as such, cannot be actively accessed and elaborated by computerized applications. The goal of this work lies in extracting content (active ingredient, interaction effects, etc.) from the Summary of Product Characteristics, focusing mainly on drug-related interactions, following a machine learning based approach. We compare two state of the art classifiers: conditional random fields with support vector machines. To this end, we introduce a corpus of 100 interaction sections, hand annotated with 13 labels that have been derived from a previously developed conceptual model. The results of our empirical analysis demonstrate that the two models perform well. They exhibit similar overall performance, with an overall accuracy of about 91%.