Benefits and Pitfalls of Using Administrative Data to Study Hospitalization Patterns in Patients With Cancer Treated With Chemotherapy

The study by O’Neill et al adds to the growing number of population studies that have reportedhigh rates of emergency room visits and hospitalizations in patients with cancer treated with chemotherapy in both the curative and metastatic setting. In the study by O’Neill et al, 92% of patients receiving chemotherapy for advanced cancers had an unplanned hospitalization, a rate that was up to 1.9 times higher than the matched nonchemotherapy control group and much higher than would be expected from clinical trial reports. This not only raises a significant concern regarding quality of care of patients with advanced cancers but also has important implications for health care use at the end of life. The use of administrative databases to study chemotherapy toxicities is both powerful and limited. On the one side, administrative data reflect the real-world experience of patients being treated in nontrial settings. Patients on clinical trials represent a minority of all patients with cancer; they are usually younger, healthier, and have higher socioeconomic status than thegeneral cancerpopulationand therefore are less likely to experience serious toxicity of chemotherapy. The outcomes reported in large administrative database studies such as O’Neill et al are more informative of the expected outcomes of the majority of patients seen in daily practice and are important for patients, providers, and health systems to understand some of the risks and health-system implications of therapy. Such information is important for both individual decision making regarding therapy and to assist with health-system planning and quality improvement. However, administrative data contain limited clinical information, which can make it challenging to attribute negative outcomes (such as hospitalizations) to the chemotherapy as opposed to other unmeasured clinical confounders, such as performance status or symptom burden, or to understand how to prevent future events. In the study by O’Neill and colleagues, the matched patients receiving chemotherapy hadworse survival thanunmatchedpatients receiving chemotherapy and thus may reflect the sickest patients treated with chemotherapy. In an attempt to determine the role of chemotherapy in driving hospitalization rates in patients with cancer, studies have used various approaches and algorithms to define hospitalizations that are likely chemotherapy associated. In general, these algorithms have been generated to reflect the common toxicities of chemotherapy using clinical experience and consensus. Although the algorithms used in the literature in general reflect the same scope of toxicities, they each vary in the complement of diagnoses considered, and,

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.