Hierarchical convolutional models for automatic pneu-monia diagnosis based on X-ray images: new strategies in public health.

Conclusions: Despite some limits, our findings support the notion that deep learning methods can be used to simplify the diagnostic process and improve disease management. Background: In order to help physicians and radiologists in diagnosing pneumonia, deep learning and other artificial intelligence methods have been described in several researches to solve this task. The main objective of the present study is to build a stacked hierarchical model by combining several models in order to increase the procedure accuracy. Methods: Firstly, the best convolutional network in terms of accuracy were evaluated and described. Later, a stacked hierarchical model was built by using the most relevant features extracted by the selected two models. Finally, over the stacked model with the best accuracy, a hierarchically dependent second stage model for inner-classification was built in order to detect both inflammation of the pulmonary alveolar space (lobar pneumonia) and interstitial tissue involvement (interstitial pneumonia). Results: The study shows how the adopted staked model lead to a higher accuracy. Having a high accuracy on pneumonia detection and classification can be a paramount asset to treat patients in real health-care environments.

Severe Acute Respiratory Infections (SARI) surveillance in over-65-years-old patients: the experience of a University hospital (seasons 2017-2018 and 2018-2019).

Background: Influenza is a relevant public health problem, also due to the risk of complications. The most effective measure to prevent influenza is vaccination; therefore, at present, there is consensus among European countries, regarding the need for routine seasonal influenza vaccination of elderly and individuals at increased risk of severe influenza. At the same time, influenza surveillance is necessary to understand the viruses circulating and effectiveness of vaccination strategies. The present study reports the results of two seasons influenza surveillance (2017/2018 and 2018/2019) conduced in an University Hospital in Rome among hospitalized patients aged ≥65 years. Study design: A prospective cohort study. Methods: The study consisted of systematic daily screening of all admissions among patients aged ≥65 years meeting a syndromic SARI case definition during two consecutive influenza seasons: 2017/2018 and 2018/2019. Characteristics of patients and their risk factors were collected by a standardized questionnaire and nose-pharyngeal swabs were performed to each patient. Influenza vaccine effectiveness (IVE), rates of vaccinated subjects and case fatality rate were also evaluated. Results: Influenza was laboratory confirmed in 11 (9.9%) of the 111 and 11 (9.6%) of the 115 enrolled patients in seasons 2017/18 and 2018/19, respectively. Adjusted IVE against all influenza type, calculated for each season, was 88.5% (95% CI: 38.9 to 97.8) and 61.7% (95% CI: -59.9 to 90.9) for 2017/2018 and 2018/2019 seasons, respectively. Our analysis shows a Case Fatality Rate of 2.7% and 4.3% for the 2017/18 and 2018/19 seasons, respectively. Conclusions: The surveillance of SARI conduced in one hospital in Rome confirmed that influenza is an important cause of hospital admissions. Routine monitoring of infectious diseases and related aetiology associated with SARI, also at the local-level, is useful for targeting the right preventive measures.

Surveillance of healthcare acquired infections by "alert microorganisms": preliminary results.

BACKGROUND: We evaluated the trend of four years (2015 - 2018) of "alert organisms" surveillance carried out at the 450 bed teaching hospital S. Andrea in Rome. METHODS: All patients with an "Alert organism" isolation were screened. In accordance with definitions used by the Centers for Disease Control patients with an "alert organism" isolation were evaluated for infection or colonization, by an infection control team (ICT). RESULTS: Between April 2015 and December 2018 a total 4,762 specimens with "Alert organism" isolation were screened and 1,601 patients were surveyed and included in the study. Overall 780 (48.8%) patients developed an healthcare acquired infection (HAI) at our institution, whereas 311 (19.4%) entered with a community acquired infection, 254 (15.8%) with an infection acquired in another healthcare setting and 256 (16.0%) resulted simply colonized. The 780 patients who developed an HAI at our institution presented 878 infectious episodes and the isolation of 931 microorganisms. C. difficile infections were the most common (27.2%), followed by 21.3% respiratory tract infections, 16.9% urinary tract infections, 15.5% surgical site infections, 12.5% bloodstream infections, 3.6% ulcers and 3.0% others. Among HAI group Gram negatives (54.1%) were more frequent than Gram positives (45.9%), whereas in patients entering in the hospital already with a community infection Gram positives overpassed Gram negatives (58.7% vs. 41.3%; p<0.001). Most common pathogens responsible for HAI were C. difficile (25.6%), Klebsiella spp. (25.5%), MRSA (19.6%) and Acinetobacter spp. (15.3%). Notably 30.0% HAI at other institutions were represented by C. difficile. Impressively, >40% of community acquired infections were related to MRSA. CONCLUSIONS: The present study provided some useful insight into the major multi-resistant pathogens epidemiology at our institution. The Authors succeeded in organizing a multidisciplinary ICT that created a partnership feeling with the hospital personnel.

Legionellosis in health care facilities: state of the art in control and prevention in Italy.

BACKGROUND: Nosocomial water systems may be contaminated by Legionella spp; therefore, health care facilities represent a potential health risk for patients and health care staff. Active, well-planned clinical and environmental surveillance in hospitals is the most important instrument of prevention. AIMS AND METHODS: The aim of the present article was to outline the state of the art in legionellosis control and prevention among Italian health care facilities by reporting some experiences in the field. RESULTS: Our results showed that Legionella spp. are largely reported as both hospital water system contaminants and etiological agents in water-related health care-associated infections (HCAI) in Italy. Among the numerous sources of HCAI, water is the most investigated, although it has been demonstrated that air sampling may provide additional information for risk assessment. CONCLUSIONS: More appropriate risk assessment is needed, especially in large facilities. In addition, more sensitive diagnostic tests should be used and dedicated training courses should be implemented in health care facilities.

Official controls regarding artisanal ice cream shops: public health policies and consumer protection in the Italian and European legislative frameworks.

Ice cream is a widely enjoyed food that is especially popular during summer. To ensure it is safe and ready-to-eat for consumers, legislation imposes a series of obligations for food business operators, and for competent authorities that have to carry out official controls, including official sampling. This article reviews the general and specific requirements applicable to the premises where ice cream is produced, concerning aspects related to health notification obligations and to implementing and maintaining procedures based on the principles of the Hazard Analysis and Critical Control Point system. The review extends to results stemming from the most recent official control activities conducted in Italy, and to future perspectives on control methods that will have to be compliant with the provisions of the new EU Regulation 625/2017 applicable from 2019.