This retrospective cohort at the Ocean Road Cancer Institute (ORCI) included 251 patients with Kaposi sarcoma (KS) treated between January 2021 and December 2023. The study estimated the cumulative incidence of loss to follow-up (LTFU) while explicitly treating death as a competing event, using the cumulative incidence function. When accounting for death, the cumulative incidence of LTFU was 27.6% (95% CI, 22.0–33.1) at 6 months, 36.4% (95% CI, 30.5–42.4) at 12 months, 43.3% (95% CI, 37.2–49.4) at 18 months, and 46.6% (95% CI, 40.4–52.7) at 24 months.
These findings indicate that nearly half of the cohort were no longer engaged in care within two years when death was accounted for as a competing risk, a magnitude that has important consequences for measuring cancer outcomes and mortality in this setting.
Among the 251 patients included, 214 (85.3%) had epidemic (HIV-associated) KS and 37 (14.7%) had endemic (non–HIV-associated) KS. Males constituted 62.2% of the cohort. The report provides these baseline distributions but does not present additional individual-level demographics or staging details in the abstract.
The authors used a competing-risks framework to recognize that death precludes later ascertainment of LTFU and that treating death as censoring could bias estimates of disengagement and downstream survival. They applied the Fine–Gray subdistribution hazards regression to evaluate predictors of LTFU while accounting for death as a competing event. Both univariable and multivariable Fine–Gray models were used to identify independent associations.
This analytic choice addresses the nonrandom nature of LTFU: patients who disengage may differ systematically from those retained with respect to disease severity, treatment response, and mortality risk, potentially underestimating cancer-related deaths if LTFU is ignored or misclassified.
In multivariable Fine–Gray regression analyses, three factors were independently associated with a lower subdistribution hazard of LTFU:
Absence of oral involvement (adjusted subdistribution hazard ratio [aSHR], 0.37). The presence of oral lesions was associated with higher risk of later disengagement compared with absence of oral involvement.
Reachable telephone contact (aSHR, 0.54). Having a functioning telephone contact was associated with better retention, suggesting contactability supports follow-up.
Initial treatment with chemotherapy rather than radiotherapy (aSHR, 0.44). Patients whose first treatment modality was chemotherapy had a lower risk of LTFU than those starting with radiotherapy.
The abstract does not report the full list of covariates included in the multivariable model or the precise variable definitions, which are not detailed in the provided source text.
The high cumulative incidence of LTFU—approaching 47% by 24 months when death is treated as a competing risk—has several implications:
Measurement bias: Substantial LTFU can bias survival estimates and lead to underestimation of cancer-related mortality if disengaged patients are assumed alive or are censored improperly. The competing-risks approach used here aims to mitigate that bias.
Programmatic focus: Interventions to improve retention in oncology clinics are critical. The associations identified point to feasible strategies such as ensuring accurate and functioning telephone contact information and using phone-based follow-up to re-engage patients.
Treatment pathways: The finding that initial chemotherapy was associated with lower LTFU than initial radiotherapy suggests that modality and care processes influence retention; reasons for this association (for example, scheduling, travel burden, side-effect profiles, or linkage to HIV care) are not described in the abstract and would require further exploration.
Surveillance systems: In resource-limited settings, strengthening routine follow-up mechanisms may improve continuity of care and the reliability of cancer surveillance data.
The study received ethical approval from the Muhimbili University of Health and Allied Sciences Institutional Research and Ethics Committee (MUHAS IRB). The authors declared no competing interests. The dataset used in the study is available from the corresponding author on reasonable request.
Limitations explicitly reported in the abstract are limited. The abstract notes that LTFU is unlikely to occur at random and may bias outcome estimates, which motivates the competing-risks analysis. The abstract does not provide detailed information on other potential limitations such as missing covariate data, reasons for LTFU, verification of vital status among those LTFU, or generalizability beyond the ORCI population.
The study concludes that nearly half of patients with KS at ORCI were lost to follow-up within two years when death was accounted for as a competing event. To address this gap, the authors recommend strengthening retention strategies in oncology care settings in resource-limited environments. Practical measures emphasized include maintaining reliable patient contact information and implementing routine phone-based follow-up to improve continuity of care, enhance survival monitoring, and strengthen cancer surveillance.
Note: The provided source text is the article abstract and metadata; detailed methods, full model covariates, additional subgroup analyses, and granular limitations were not reported in the abstract and therefore are not included here.