Hospital care patterns among patients with psychotic-disorder coding and metastatic/secondary malignant neoplasm coding: A nationwide administrative cohort study in Spain

Support Care Cancer. 2026 Nov 4;34(11):1193. doi: 10.1007/s00520-026-11440-0.

ABSTRACT

PURPOSE: Patients with psychotic disorders may experience differences in cancer care, but evidence on hospital supportive and palliative care patterns in advanced cancer remains limited. We examined whether psychotic-disorder coding was associated with hospital palliative-care coding and selected hospital treatment-intensity markers among patients with metastatic/secondary malignant neoplasm coding in a Spanish hospital oncology extract.

METHODS: We conducted a nationwide retrospective cohort study using Spanish hospital administrative data from 2021 to 2024. The source extract was a requested oncology administrative hospital extract, composed overwhelmingly of episodes with malignant cancer recorded as the principal diagnosis and also containing a small subset of palliative-care principal-diagnosis episodes with malignant neoplasm/metastatic coding in secondary fields. Eligible patients had secondary malignant neoplasm/metastatic disease coding in secondary diagnosis fields (C77-C79). The first metastatic-coded episode observed during 2021-2024 was treated as an administrative index episode and not necessarily as the onset of metastatic disease. The unit of analysis was the patient. For each patient, a metastatic care window was defined from the first metastatic-coded hospital episode to the last available episode. Psychotic-disorder coding was defined primarily as F20-F29 within this window; a temporal audit examined F20-F29 coding before or at the index episode. Outcomes were considered present if coded at least once during the metastatic window. Multivariable logistic regression adjusted for age, sex, year, tumor group, autonomous community, hospital service, and approximate non-oncologic Charlson-based comorbidity measured at the first metastatic-coded episode.

RESULTS: The cohort included 272,809 patients, of whom 1989 had psychotic-disorder coding within the metastatic care window (0.73%). A temporal audit identified 1938 patients with F20-F29 coding before or at the first metastatic-coded episode in the observed extract; 1889 patients met both definitions, 100 had F20-F29 coding in the metastatic care window only after the index episode, and 49 had F20-F29 coding before or at the index episode but not within the metastatic care window. Psychotic-disorder coding was associated with higher odds of hospital palliative-care coding (OR 1.35, 95% CI 1.19-1.52) and mechanical ventilation (OR 1.44, 95% CI 1.15-1.82). In-hospital chemotherapy procedure coding showed lower adjusted odds among patients with psychotic-disorder coding (OR 0.73, 95% CI 0.59-0.91). Marginal adjusted probabilities for palliative-care coding were 20.6% with psychotic-disorder coding versus 17.0% without psychotic-disorder coding. Corresponding probabilities were 4.5% versus 5.8% for in-hospital chemotherapy procedure coding and 3.6% versus 2.6% for mechanical ventilation.

CONCLUSION: In this nationwide hospital administrative oncology extract, psychotic-disorder coding was associated with distinct hospital coding patterns after metastatic/secondary malignant neoplasm coding. The findings should be interpreted as hospital-coded associations and not as direct measures of appropriateness, treatment quality, outpatient systemic therapy, or specialist palliative-care access.

PMID:42982905 | DOI:10.1007/s00520-026-11440-0