Study:
(1) 32 patients treated with plasma exchange from 1980 to 1994.
(2) Data analyzed using Cox stepwise logistic regression.
Independent predictors of treatment failure:
(1) longer delay in initiating plasma exchanges
(2) presence of stupor or coma at the beginning of plasma exchange
(3) higher serum creatinine levels at the beginning of plasma exchange
|
Parameter
|
Odds Ratio
|
95% CI
|
p value
|
|
stupor or coma
|
148
|
99.4 – 8955
|
< 0.001
|
|
delay in plasma exchange
|
1.50
|
1.44 – 1.73
|
0.007
|
|
serum creatinine
|
6.07
|
4.61 – 81.45
|
0.009
|
from Table 2, page 322; delay is based on the number of days from presentation prior to starting exchange; serum creatinine was based on intervals of < 2, 2.0 – 4.9, 5 – 10, and > 10 mg/dL.
Predictors of mortality:
(1) stupor or coma present at the beginning of plasma exchanges
(2) failure to treat with plasma exchanges
Interestingly, the following parameters did not correlate with prognosis (see prognostic scores in previous sections) in this series:
(1) hemoglobin levels
(2) platelet count
(3) serum LDH level
Steps to improve prognosis:
(1) early diagnosis
(2) prompt initiation of intensive plasma exchange
