A review of lung cancer care at a public hospital in Costa Rica found that 190 of 223 patients in the study had died by the close of follow-up in 2025. That was 85.20% of the cohort. The estimated five-year survival was 1.9%, although the researchers measured survival from the first recorded care episode linked to suspected lung cancer, not from the later pathological diagnosis.
The pattern raises questions about how patients moved from suspicion to diagnosis and treatment, but the study does not establish that delays or access routes caused the deaths. It was a retrospective observational review with descriptive and exploratory bivariate analyses, and no multivariable model was performed.
What the records showed
Researchers examined adults older than 18 with pathological primary lung cancer treated at Hospital México, a public tertiary-level hospital in Costa Rica, from 2015 to 2020. The pathology database contained 278 records. Eighteen physical records could not be accessed and 37 were excluded after review, leaving 223 patients for analysis.
Among the symptoms recorded before diagnosis, cough was most common, appearing in 126 patients, or 56.50%. Dyspnea was recorded in 110 patients, or 49.33%, and weight loss in 98, or 43.95%. These figures reflect what was documented in the charts, so they may not capture every symptom or when it first began.
Adenocarcinoma accounted for 194 cases, or 91.51%, while poorly differentiated tumors were the most frequent grade, recorded in 80 patients, or 45.45%. The records also listed the tumor-extent categories T4, N3 and M1a in 70 patients, 41 patients and 45 patients, respectively. The researchers did not reconstruct formal overall stage groups because complete information was not available.
The path from suspicion to treatment
For the 213 patients with the relevant dates recorded, the median time from documented suspicion to pathological diagnosis was 34 days. The middle half of recorded intervals lay between 21 and 73 days, but some exceeded 1,000 days. Because the starting point was the documented suspicion date, the measure may miss time spent before referral or the first hospital consultation.
The median interval from diagnosis to first treatment was 46 days, with the middle half of recorded values ranging from 29 to 81 days. The median time from the first care episode related to suspected lung cancer to death was 188 days, with an interquartile range of 71 to 476 days. Both measures were calculated from the available retrospective records.
The study estimated five-year survival at 1.9%, with a 95% confidence interval of 0.5% to 5.1%. That time origin matters: the estimate begins with the first suspected-cancer care episode, so it is not the conventional overall-survival figure measured from pathological diagnosis.
Testing and newer treatments were not routine
Molecular biomarker testing was recorded for 106 patients, or 47.53% of the cohort. The most common markers were EGFR, ALK and PD-L1. Among reports that included information about where testing was performed, 30, or 60.00%, came from hospitals in the national CCSS system. The records did not consistently identify testing platforms or centers.
Treatment was recorded for 150 patients, or 67.26%. Chemotherapy appeared in the records of 83 patients, radiotherapy in 58, and targeted therapy or immunotherapy in 45. Nine patients, or 4.04%, were recorded as not being candidates for treatment and received palliative management. The percentages for individual treatments used the documented-treatment denominator for each variable, so missing information matters.
Among the patients who received targeted therapy or immunotherapy, 29, or 64.44%, accessed those treatments through administrative requests to higher institutional authorities outside the standard care protocol. The finding describes the recorded route to treatment. It does not establish why those requests were needed or whether the route changed clinical outcomes.
What the study can and cannot say
Male sex was associated with a higher probability of mortality than female sex in the study’s bivariate analysis, with an odds ratio of 6.71 and a 95% confidence interval from 1.48 to 30.34. The result was exploratory and unadjusted. Without a multivariable analysis, it cannot be treated as an independent effect of sex.
The researchers used descriptive summaries, distribution-appropriate statistical tests and Kaplan-Meier survival analysis, with a 5% significance threshold. Patients without a documented death were censored at the end of follow-up or at their last available record. The retrospective, single-center design included incomplete and heterogeneous documentation.
The results describe patients treated at one public tertiary hospital, not all people with lung cancer in Costa Rica. They also cannot determine whether diagnostic delays, incomplete biomarker testing or administrative treatment routes contributed to the observed survival. More complete information on the time before referral and formal stage would be needed to answer those questions.
Paper data and sources
Original title: Clinical Characteristics and Delays in the Care Pathway of Patients With Lung Cancer Treated at a Public Hospital in Costa Rica, 2015-2020
Authors: Alvarado-Prado R, Ortiz-Barboza A, Blanco-Saborío A et al.
Journal/Repository: Cureus
Status: Peer-reviewed
First online: 2026-10-19
DOI: Not available
Original paper · Full text