Dear Editor, Extracorporeal membrane oxygenation (ECMO) has a significant impact on intensive care unit (ICU) patient costs. Human resources are a major cost driver for ECMO in severe acute respiratory distress syndrome (ARDS) [1]. However, recent developments in ECMO technology has allowed safe patient management by a single caregiver [2–4]. In the present report, we analyze the direct ICU costs of patients with influenza A (H1N1)-related ARDS supported with miniaturized ECMO systems and managed by a single caregiver. We performed an observational study of all patients (n = 10) with H1N1-associated ARDS treated with ECMO in Hospital S. Joao (Porto, Portugal), between November 2009 and April 2011. The clinical data and ECMO treatment details of this ICU patient cohort have been described in a previous report [5]. Patients were admitted in medical ICUs and managed by a single caregiver using miniaturized ECMO circuits (Rotaflow Centrifugal Pump plus Quadrox PLS and CARDIOHELP; Maquet Cardiopulmonary AG). In our single caregiver model, the ICU nurse primarily managed the ECMO patient under the coordination of 24/7 inhouse ICU physicians, with the backup of an on-call ECMO specialist. Our multidisciplinary ECMO specialist team (ICU physicians, ICU nurses, and cardiovascular perfusionists) is responsible for managing equipment and supplies, circuit preparation, patient transportation, troubleshooting, daily rounds, education, and service administration. Per patient ICU direct costs were analyzed retrospectively from clinical records and hospital databases using both top-down and bottom-up approaches. Bottom-up analysis was performed for all drugs, transfusion requirements, laboratory and imagiology exams. Regarding disposable items, bottom-up analysis was possible for ECMO-specific disposables and for renal replacement therapies. Costs with human resources were assessed using the top-down approach. Specific ECMO costs were also individualized for each ICU episode: (1) disposables devices; (2) laboratory tests; (3) transfusion requirements; and (4) drugs. The Ethics Committee of the Hospital S. Joao approved this study and waived the need for patients’ informed consent. Variables are reported as median and interquartile ranges, if not otherwise indicated. The cost drivers of H1N1-related ARDS patients treated with miniaturized ECMO are presented in Table 1. ECMO support corresponded to 63.6 (62.5–88.5) % of the ICU length of stay (LOS). The ICU LOS positively correlated (r = 0.97; p \ 0.001) with ICU patient costs [42,721 (33,358–56,801) €], but not with specific ECMO costs [r = 0.17; p = 0.63; 7,516 (5,599–7,736) €]. ECMO-specific costs represented 17.6 (13.1–18.1) % of the total ICU direct costs, mostly ECMO disposable devices [94.8 (87.1–96.1) %]. Daily ICU costs with antibiotics, sedation, analgesia and neuromuscular blockade, fluids and nutrition, vasoactive drugs, anticoagulation medications, and other unspecific drugs were higher during the period of ECMO support compared with the period without ECMO support. In conclusion, the use of miniaturized ECMO had a small impact on direct ICU costs in our cohort of H1N1-related severe ARDS patients managed by a single caregiver and did not correlate with the ICU LOS. Knowledge of the incremental costs associated with miniaturized ECMO will contribute to the clarification of its cost-effectiveness in severe ARDS.