Overview

Introduction Throughout human history, surgical interventions have been frequently used in human treatment. However, despite their therapeutic properties, the pain experienced by patients, especially in the acute postoperative period, can be quite challenging for clinicians. Acute postoperative pain is an important public health issue. While 80% of patients report experiencing pain in the postoperative period, 88% of them experience moderate or higher levels of pain. According to another study, more than 60% of surgical patients suffer from moderate to severe acute postoperative pain, and this pain has been associated with the development of chronic postoperative pain. Poorly managed postoperative pain can lead to negative outcomes such as lower patient satisfaction, delayed patient recovery, increased length of hospital stay, increased care costs, chronic pain, unnecessary opioid prescription, opioid abuse, overdose, and death. In addition, in order to provide effective pain management, the method of providing preventive analgesic treatment before the pain begins is frequently used. However, this situation may lead to unnecessary medication administration in many patients and consequently, many adverse events such as bleeding, respiratory depression, cardiac events or gastrointestinal system side effects of opioids, nonsteroidal anti-inflammatory drugs and other analgesics. As a result, the difficulty in predicting acute postoperative pain leads to suboptimal pain management. Therefore, being able to predict which patients will suffer from moderate to severe acute postoperative pain will optimize the risk-benefit ratio of perioperative analgesic treatments and ensure that appropriate treatment is given. Although different studies on this subject have tried to predict postoperative pain with logistic regression analysis, the desired result has not yet been achieved. This situation becomes even more important in surgeries with a high risk of severe pain in the postoperative period, such as lung resection. In order to reduce or prevent postoperative pulmonary complications in patients undergoing lung resection, it is very important for patients to be able to cough without feeling pain and thus to remove secretions from the respiratory tract. If sufficient analgesia is not provided, these patients cannot perform this effectively. This increases complications, hospital stay, and patient care costs. In order to prevent these negative situations and provide optimal analgesia, new methods are needed to predict postoperative pain levels. Numerous models have been proposed in studies to understand the risk factors that will exacerbate severe acute postoperative pain. Most of the research in this area has focused on determining risk factors for postoperative pain using statistical methodology. Previous studies suggest that machine learning models can outperform linear statistical models in classifying postoperative pain-related outcomes when similar features are considered. Therefore, artificial intelligence (AI) algorithms are algorithms that can combine and analyze complex data with hundreds of variables and provide new outputs, and can guide an effective solution in predicting and managing the postoperative process. Previous studies have shown promising results in predicting acute postoperative pain with an area under the curve (AUC) of 0.70 using artificial intelligence algorithms to predict pain with perioperative data. However, studies on this topic are needed in a specific surgery such as lung resection, which has the potential for severe pain.

This study aimed to predict postoperative pain by analyzing perioperative data using AI algorithms in lung resections and to determine the effectiveness of AI algorithms in this regard. Thus, it aimed to reduce unnecessary analgesic use in patients, eliminate possible side effects of these drugs, and start effective analgesic treatment in a timely manner in patients with high pain risk.

Purpose/Hypothesis:

This study aimed to predict postoperative pain by analyzing perioperative data using AI algorithms in lung resections and to determine the effectiveness of AI algorithms in this regard.

H0: Artificial intelligence algorithms are not effective in predicting postoperative pain in lung resections.

H1: Artificial intelligence algorithms are effective in predicting postoperative pain in lung resections.

Material-Method:

This study will be conducted in accordance with the Declaration of Helsinki and will be carried out at the SBÜ Ankara Atatürk Sanatorium Training and Research Hospital after receiving ethics committee approval. Our study is a retro-prospective study. Retrospectively collected patient data will be evaluated prospectively with artificial intelligence algorithms.

Eligibility

Inclusion Criteria:

• Age > 18,
• Patients who underwent lung resection between January 2023 and October 2024 will be included.

Exclusion Criteria:

• Patients under 18 years of age
• Patients with missing postoperative pain form
• Patients who did not undergo lung resection