Search

Background: Colorectal cancer currently presents the third-highest incidence of cancers worldwide, making secondary prevention through screening programs for colorectal cancer, usually by Fecal Occult Blood Testing, an essential preventive medicine intervention. First-degree relatives of colorectal cancer patients are a particularly at-risk group, with indications to consider direct screening by full colonoscopy. Colonoscopy is considered the gold standard for diagnosing colorectal cancer, as it has high sensitivity and specificity, and is both a diagnostic and therapeutic tool. However, it requires significant organizational and financial resources, and has a small but relatively higher risk of complications as opposed to fecal occult blood testing. The present study aimed to assess the appropriateness of a screening program without age restrictions of CRC by full colonoscopy in asymptomatic, first-degree adult relatives of patients with colorectal cancer, aiming both to actively increase screening coverage and to determine the detection rate of precancerous lesions and colorectal cancer in this population., Study Design: Uncontrolled interventional study - colorectal cancer screening by full colonoscopy for at-risk population., Methods: The Italian League for the Fight against Cancer started a colorectal cancer screening program by full colonoscopy for first-degree relatives of colorectal cancer patients in 1998 in the province of Latina, Lazio Region, Italy. The program was expanded to the provinces of Rieti, Lazio Region, and Sassari, Sardinia Region, in 2014 and 2016 respectively, and was concluded in 2018. Subjects were actively and voluntarily recruited by the study's working group. Subjects that had already been subjected to a full colonoscopy in the preceding 5 years were excluded from this study. Identified neoplastic lesions were treated either directly or referred to the Day Hospital setting, and histologically diagnosed following World Health Organization guidelines., Results: In total, 2,288 subjects (age range 15-88, mean 52.3 yrs, M/F = 946/1,204) were screened by colonoscopy, of which 103 (4.5%) were incomplete and 2,173 (95.0%) complete, with data on colonoscopy performance missing for 12 participants. Out of 468 positive outcomes on colonoscopy, diagnosis for 422 (204M/173F), 19.4% of total subjects, was adenomatous polyps and 46 (20M/20F), 2.1% of total subjects, was colorectal cancer. Female sex was a protective factor against a positive test outcome, with a 35% reduction compared to male sex, with OR=0.64 95%CI (0.52-0.80). On the other hand, being over 50 years of age was found to be a risk factor, making a positive outcome more than twice as likely, with OR=2.3 95%CI (1.8-2.9). Subjects over 50 also had significantly more instances of multiple adenomas being found, however the size distribution of found adenomas was not significantly different between subjects under and over 50, despite size being a predictor of risk of neoplastic progression., Conclusions: Given the high detection rate of precancerous lesions and colorectal cancer in the studied population, it is our opinion that guidelines should continue to recommend earlier and more frequent screening in first-degree relatives of patients with colorectal cancer, and, barring the introduction of more cost-effective and/or lower risk procedures with a similar efficacy profile, maintain the use of colonoscopy as the main screening option.

The first annual meeting of the Italian Society for Artificial Intelligence in Medicine (Società Italiana Intelligenza Artificiale in Medicina, SIIAM) on December 7, 2023, marked a significant milestone in integrating artificial intelligence (AI) into Italy's healthcare framework. This paper reports on the collaborative workshop conducted during this event, highlighting the collective efforts of 51 professionals from diverse fields including medicine, engineering, data science, and law. The interdisciplinary background of the participants played a crucial role in generating ideas for innovative AI solutions tailored to healthcare challenges. Central to the discussions were several AI applications aimed at improving patient care and streamlining healthcare processes. Notably, the use of Large Language Models (LLMs) in remote monitoring of chronic patients emerged as an area of focus. These models promise enhanced patient monitoring through detailed symptom checking and anomaly detection, thereby facilitating timely medical interventions. Another significant proposal involved employing LLMs to improve empathy in medical communication, addressing the challenges posed by cultural diversity and high-stress levels among healthcare professionals. Additionally, the development of Machine Learning algorithms for standardizing treatment in pediatric emergency departments was discussed, along with the need for educational initiatives to enhance AI adoption in rural healthcare settings. The workshop also explored using LLMs for efficient data extraction and analysis in scientific literature, interpreting healthcare norms, and streamlining hospital discharge records. This paper provides a comprehensive overview of the ideas and solutions proposed at the workshop, reflecting the participants' forward-thinking vision and the potential of AI to revolutionize healthcare., Competing Interests: The authors declare no conflict of interest., (©2024 Pacini Editore SRL, Pisa, Italy.)