Association between pre-processing total protein and processing time in drop-type cell-free and concentrated ascites
Kazuha Sagiya1, Naoki Ogasawara2, Shinya Inahara3
1Central Medical Department Faculty of Clinical Engineering, Tohoku Medical and Pharmaceutical University Hospital, 1-12-1 Fukumuro, Miyagino-ku, Sendai, Miyagi, 983-8512, Japan. ce-sagiya@hosp.tohoku-mpu.ac.jp.
Abstract:
Predicting processing time in drop-type cell-free and concentrated ascites reinfusion therapy (CART) remains difficult. We examined whether pre-processing total protein (TP) was associated with processing time after accounting for ascites amount, processing height, ascites etiology, and repeated sessions. We retrospectively analyzed 97 CART sessions in 51 patients. Log-transformed processing time was analyzed using a linear mixed-effects model including pre-processing TP, pre-processing ascites amount, processing height (140 or 200 cm), and ascites etiology as fixed effects, with patient as a random intercept. Receiver operating characteristic (ROC) analysis of pre-processing TP for an above-preferred-range session-averaged processing rate (>2,000 mL/h) was secondary and exploratory. Higher pre-processing TP (β = 0.259, 95% confidence interval (CI) 0.194-0.325; P < 0.001), greater ascites amount per 1,000 g (β = 0.270, 95% CI 0.234-0.305; P < 0.001), and 140-cm versus 200-cm processing height (β = 0.161, 95% CI 0.038-0.284; P = 0.012) were associated with longer processing time, whereas ascites etiology was not significantly associated. The overall ROC cutoff was 1.2 g/dL (area under the curve (AUC), 0.931), and the first-session AUC was 0.947 (95% CI, 0.884-1.000). Higher pre-processing TP remained significantly associated with longer processing time after accounting for ascites amount, processing height, ascites etiology, and repeated sessions. Pre-processing TP may serve as a practical workflow-oriented indicator in drop-type CART, whereas ROC-derived cutoffs should be interpreted as exploratory.
