OBJECTIVE: To investigate the immune and inflammation confusion state in severe sepsis and the effects of two way immunomodulation therapy with continuous blood purification (CBP), thymosin alpha1, and combined therapy of CBP and thymosin alpha(1). METHODS: 91 Patients with severe sepsis aged > 18, with Marshall score>5. were randomly divided into 4 groups: CBP Group (n = 22) undergoing continuous renal replacement therapy (CRRT) or molecular adsorbents recirculating system (MARS) therapy once a day for 3 days in addition to classical Surviving Sepsis Campaign (SSC) therapy, Thymosin alpha(1) Group (n = 23) undergoing subcutaneous injection of thymosin alpha(1) 1.6 mg once a day for 7 days in addition to SSC therapy, Combined Therapy Group (n = 22) undergoing CBP combined with thymosin alpha(1) treatment in addition to SSC therapy, and SSC Group (treatment control group, n = 24) undergoing SSC therapy only. Peripheral blood samples were collected before treatment, and 3 and 7 days after the beginning of treatment (days 4 and 8) to detect the serum interleukin (IL)-6, IL-10, and tumor necrosis factor (TNF)-alpha. The levels of CD(14)(+) monocyte human leucocyte antigen (HLA)-DR and T lymphocytes were monitored. The mechanical ventilation time, ICU stay length, and mortality within 28 d and mortality within 90 d were observed. Ten healthy persons were used as healthy control group. RESULTS: Thirty-four of the 91 patients died within 28 d with a mortality of 77.4% (Death Group) and other 57 patients were put in Survival Group. The levels of serum IL-6, IL-10, and TNFalpha, and IL-6/IL-10 at different time points of both Death and Survival Groups were all significantly higher, and the HLA-DR level, and CD(3)(+), CD(4)(+), and CD(8)(+) T lymphocyte numbers at different time points of both Death and Survival Groups were all significantly lower than those of the healthy controls (P < 0.05 or < 0.01). The levels of serum IL-6, IL-6/IL-10, TNFalpha, HLA-DR, and CD(3)(+), CD(4)(+), and CD(8)(+) T lymphocyte at different time points of Death Group were all significantly higher than those of Survival Group (P < 0.05 or < 0.01). The CD(3)(+) T lymphocyte number on day 8 of Thymosin Group was significantly higher than that of SSC Group (all P < 0.05). The serum IL-6 and TNFalpha and IL-6/IL-10 were decreased, and HLA-DR, and CD(3)(+), CD(4)(+), and CD(8)(+) were increased significantly on day 8 in CBP and Combined Therapy Groups. The level of TNFalpha decreased, and the numbers of CD(3)(+) and CD(4)(+) T lymphocytes increased significantly on day 4 in Combined Therapy Group (P < 0.05 or P < 0.01). Compared with Thymosin Group, almost all the indexes of CBP and Combined Therapy Groups were improved, only the CD(3)(+) T lymphocyte level on day 4 increased and the IL-6/IL-10 ratio on day 8 was decreased significantly in Combined Therapy Group (both P < 0.05). Compared with those of SSC Group, the mechanical ventilation time, length of ICU stay within 28 days, and 28 days mortality and 90 days mortality of the 3 treatment groups were all decreased, and there were statistical differences in the length of ICU stay of CBP Group and in the mechanical ventilation time and length of ICU stay within 28 days of Combined Therapy Group (both P < 0.05). CONCLUSION: Systemic inflammatory response and immunodepression exist simultaneously in severe sepsis. Thymosin alpha(1) increases the cellular immunity, and CBP bi-modulates the immune turbulence, reduces the inflammatory mediators, and ameliorates the immune homeostasis. These 2 therapies also improve the clinical prognosis and the combination of both would be more effective.