Table 4.

Change in blood hemoglobin in first 6 mo and 2-yr mortality census (rate) among 26,668 MHD patients of the first quarter who had at least one additional averaged hemoglobin value during the subsequent quarter of the cohorta

Hemoglobin Change (g/dl in 6 mo)Sample Size (%)All-Cause Death (%)bCV Death (%)bBaseline HemoglobinBaseline Albumin
Dropped
    ≤−2.01,691 (6)557 (33)239 (15)13.2 ± 1.13.83 ± 0.36
    −1.99 to −1.51,562 (5)475 (30)208 (14)12.7 ± 1.03.84 ± 0.36
    −1.49 to −0.84,033 (13)1,087 (27)461 (12)12.5 ± 1.03.86 ± 0.34
Stable
    −0.79 to 0.7916,130 (53)4,225 (26)1,877 (12)12.0 ± 1.03.85 ± 0.33
Increased
    0.8 to 1.493,800 (12)1,102 (29)502 (14)11.4 ± 0.93.79 ± 0.35
    1.5 to 1.991,581 (5)443 (28)193 (13)11.0 ± 1.03.74 ± 0.38
    ≥2.01,904 (6)523 (27)235 (13)10.5 ± 1.13.68 ± 0.38
  • a Both ANOVA and trend P values for mortality rates are <0.001.

  • b The death rates in brackets pertain to the percentage of patients who died during a 2-yr interval in the given category. The denominator for all-cause mortality is the total sample of MHD patients with known hemoglobin change, whereas the denominator of CV death is slightly smaller as a result of missing documentation of causes of death. Death rates and baseline hemoglobin and albumin values are significantly different from each other across the categories (ANOVA P < 0.001).