This was a retrospective observational study carried out in the Department of Critical Care Medicine of Sir Run Run Shaw Hospital (a university-affiliated hospital in Hangzhou, China). The study included consecutive patients with sHTGP (admission TG > 11.3mmol/L) who were admitted within 72h after AP onset. From January 2019 to February 2020, there were 494 patients admitted to the Sir Run Run Shaw Hospital during the period, and the main etiologies were cholelithiasis (47.4 %), hypertriglyceridemia (HTG, 39.9 %) and alcohol (4.9 %) respectively. AP was diagnosed according to the 2012 Atlanta classification . Exclusion criteria included: age < 18 years old, pregnancy, complicated with malignant tumor and incomplete information. The study was approved by the Ethics Committee of Sir Run Run Shaw Hospital (20190215-3). Because of the retrospective characteristic of the study, Informed consents from individuals were waived.
According to the AP treatment procedures of the center, all the sHTGP patients were suggested to receive DFPP within 24h after admission. DFPP was conducted via femoral double lumen using Plasauto EZ machine machine (Asahi-Kasei, Tokyo, Japan), which was loaded with a blood cell separator column (Plasmaflo, OP-08) and a plasma separator (Cascadeflo, EC-40W). Heparin was applied for the anticoagulation of the system. Large weight molecules such as VLDL were discarded and clean plasma was returned to circulation system. The frequencies of DFPP sessions were decided by the clinicians in each case.
Baseline characteristics included demographic data, body mass index (BMI), co-morbidities, disease severity, the Acute Physiology and Chronic Health Evaluation II score (APACHE II) and the Sequential Organ Failure Assessment score (SOFA). The lipid profile was collected in each case on admission, consecutive 4 days after DFPP and at discharge. Lipid profile included serum TG, cholesterol (TC), very low density lipoprotein (VLDL), low density lipoprotein (LDL), HDL and Apoprotein A1 (Apo A1). TG, TC, VLDL, LDL, HDL and Apo A1 were measured by homogeneous assays carried out by the Abbott ARCHITECT c16000 Clinical Chemistry Analyser (Abbott Diagnostics, USA). The primary outcome in this study was the hospitalization duration.
For continuous variable with normal distribution, it was presented as mean ± SD and analyzed using t test. For continuous variable without normal distribution, it was presented as median (25%, 75% interquartile ranges) and analyzed using Mann-Whitney test. Categorical variables were presented as percentages and analyzed with the χ2 test. Propensity score matching was used to minimize the effect of confounding factors which may lead to outcome bias. A one-to-one nearest neighbor matching algorithm was applied using a caliper width of 0.05. The following variables were selected to generate the propensity score: age, sex, BMI, APCHE II, SOFA and initial TG level. Kernel density plots of the p score were applied to examine the propensity score matching degree. Finally, 10 matched pairs were generated and applied to further analyses. After coinciding with all assumptions, multiple linear regression analyses were applied to explore the relationship between parameters of lipid profile and hospitalization duration. Four models were used in the linear regression analyses: model 1 was performed as an adjusted model; model 2 was adjusted for age, sex and BMI; model 3 was adjusted for age, sex BMI and APACHE II score; model 4 was adjusted for age, sex BMI, APACHE II score and co-morbidities. The statistical analysis was performed by SPSS 23.0 and Graphpad Prism 6.0. P values < 0.05 were considered statistically significant (two-tailed). In this study, the continuous variables with missing values were less than 20% and consequently the missing values were replaced by the mean or median values.