Infection Patterns and Predictors of 28-Day Mortality Among Cancer Patients in Najran, Saudi Arabia: A 10-Year...

Background and aim Infections are a major cause of morbidity and mortality in cancer patients; however, regional data from southern Saudi Arabia remain scarce. The epidemiology of infectious complications can vary significantly due to differences in local pathogens, resistance patterns, and healthcare practices. Understanding these patterns is crucial for guiding empirical therapy and improving clinical outcomes. Therefore, this study aimed to characterize infection patterns and identify independent predictors of 28-day mortality among adult cancer patients in Najran, Saudi Arabia. Methods This retrospective cohort study at King Khalid Hospital Oncology Center (Najran, Saudi Arabia; January 2014 to December 2024) included 199 adult cancer patients (≥18 years) with 577 documented infection episodes. Primary outcomes were infection site distribution and microbiological profiles. Secondary outcomes included 28-day and overall mortality. The Andersen-Gill extension of the Cox model analyzed recurrent episodes, accounting for within-patient clustering. Multivariable Cox proportional hazards regression identified mortality predictors, reported as HRs with 95% CIs. ICU admission served as a proxy for clinical severity. Results Patients had a median age of 62 years; 133 (66.8%) had metastatic disease, and 100 (50.3%) were female. The most frequent malignancies were breast cancer in 35 (17.6%), colorectal cancer in 33 (16.6%), upper gastrointestinal cancers in 31 (15.6%), and hematologic malignancies (lymphoma/myeloma) in 30 (15.1%). Of 577 infection episodes, 351 (60.8%) were culture positive. Among positive cultures, Gram-negative organisms predominated, with Klebsiella (69; 12.0% of all episodes) and Escherichia coli (64; 11.1%) being the most common. Multidrug-resistant organisms accounted for 49 (8.5%) episodes. Among 199 patients, 179 (89.9%) survived, and 20 (10.1%) were non-survivors (28-day mortality). The 28-day mortality rate was 12.3% per episode. Independent predictors of mortality included ICU admission (HR 8.56, 95% CI 3.17-23.13, p < 0.001), neutropenia (HR 2.52, 95% CI 1.07-5.91, p = 0.034), noninvasive ventilation (HR 3.78, 95% CI 1.06-13.47, p = 0.041), and polymicrobial infection (HR 4.16, 95% CI 1.18-14.69, p = 0.027). Additionally, 440 (76.3%) episodes lacked a localized infection site, 106 (55%) patients had recurrent infections, and 137 (23.7%) required ICU admission. Culture-positive episodes trended toward higher ICU use (94 (26.8%) vs 45 (19.9%); p = 0.0765), but this was not statistically significant. Cause of death analysis showed underlying malignancy (51.0%) was more frequent than organ failure (38.5%; p = 0.068), with treatment-related complications (6.7%) and unknown causes (3.8%) contributing less. Conclusions In southern Saudi Arabia, Gram-negative pathogens and polymicrobial infections contribute significantly to mortality. The low multidrug resistance rate (8.5%) compared to other Saudi regions presents a critical opportunity for intervention. Local healthcare administration should urgently implement standardized diagnostic bundles and antimicrobial stewardship to maintain this favorable profile. Markers of acute deterioration (ICU admission, neutropenia, and noninvasive ventilation) strongly predict short-term mortality. These findings highlight the need for comprehensive clinical assessment and prospective studies with standardized definitions.