Developing and validation of a smartphone app for post-discharge early follow-up after colorectal cancer surgeries

Digit Health. 2024 Oct 24;10:20552076241292389. doi: 10.1177/20552076241292389. eCollection 2024 Jan-Dec.

ABSTRACT

BACKGROUND: Colorectal surgeries are complex procedures associated with high rates of complications and hospital readmission.

OBJECTIVE: This study aimed to develop an electronic post-discharge follow-up plan to remotely monitor patients' symptoms in the postoperative period of colorectal surgeries and evaluate the outcomes of emergency department visits and the rate of severe complications within 15 days after hospital discharge.

DESIGN: We developed a digital tool capable of remotely assessing symptoms that could indicate complications related to colorectal surgical procedures and directing early management. This project was divided into two stages. The first was platform development with an algorithm for identifying symptoms and directing conduct, and the second was clinical validation of the program and evaluation of patient's experience. Patients who underwent elective oncological colorectal surgery were invited to participate in this study. We used commercial software (CleverCare) that was adjusted according to the clinical algorithm developed in this study, predicting complications and directing conduct with minimal human intervention using a Chatbot with Natural Language Processing (NPL) and artificial intelligence.

RESULTS: We planned three Interim Analyses to evaluate the outcomes of complications, referrals to the Emergency Department (ED), ED visits, adherence, and patient satisfaction. After each analysis, specialists validated the changes before implementation. A total of 92 eligible participants agreed to participate in the study. The ability to detect complications increased with each adjustment phase, and after the third and last phase, the digital solution identified 3(4.8%) real complications, with a sensitivity of 75%, specificity of 83%, accuracy of 82%, positive predictive value of 27%, and negative predictive value of 97%. Complete adherence to the monitoring program was 83.7% with an NPS score of 94 in the last evaluation phase.

CONCLUSION: The digital platform is safe with high adherence rates and good patient acceptance.

PMID:39465225 | PMC:PMC11512466 | DOI:10.1177/20552076241292389