Continuous veno‐venous hemofiltration for severe acute pancreatitis

Abstract

Background

Severe acute pancreatitis is associated with high rates of mortality and life‐threatening complications. Continuous veno‐venous hemofiltration (CVVH) has been used in some centers to reduce mortality and avoid local or systemic complications; however, its efficacy and safety are uncertain.

Objectives

To assess the benefits and harms of CVVH in patients suffering from severe acute pancreatitis; to compare the effects of different CVVH techniques; and to evaluate the optimal time for delivery of CVVH.

Search methods

We searched the Cochrane Library (2019, Issue 8), MEDLINE (1946 to 13 September 2019), Embase (1974 to 13 September 2019), and Science Citation Index Expanded (1982 to 13 September 2019).

Selection criteria

We included all randomized controlled trials (RCTs) that compared CVVH versus no CVVH in participants with severe acute pancreatitis. We also included RCTs that compared different types of CVVH and different schedules for CVVH in participants with severe acute pancreatitis.

Data collection and analysis

Two review authors independently identified the trials for inclusion, collected the data, and assessed the risk of bias. We performed the meta‐analyses using Review Manager 5. We calculated the risk ratio (RR) for dichotomous outcomes, and the mean difference (MD) for continuous outcomes, with 95% confidence intervals (CIs). We used GRADE to assess the certainty of the evidence for all outcomes.

Main results

We included three RCTs, involving a total of 219 adults with severe acute pancreatitis with various etiologies, including alcoholic abuse, biliary disease, high‐fat diet, hyperlipidemia, etc. All of the participants had single or multiple organ failure. All of the RCTs were single‐center studies conducted in China. The mean age of participants was 47.4 years. The mean proportion of females was 42.0%. All RCTs were at high risk of bias.

CVVH versus no CVVH

We included two RCTs in which 189 participants with severe acute pancreatitis were randomized to undergo CVVH (97 participants) or no CVVH (92 participants). The evidence is very uncertain about the effect of CVVH on in‐hospital mortality (RR 0.62, 95% CI 0.27 to 1.40; 2 studies, 189 participants; very low‐certainty evidence) compared with no CVVH. The evidence suggests that CVVH may reduce length of stay in the intensive care unit (ICU) (MD ‐8.80 days, 95% CI ‐10.24 to ‐7.36 days; 1 study, 125 participants; very low‐certainty evidence), length of hospital stay (MD ‐26.40 days, 95% CI ‐30.17 to ‐22.63 days; 1 study, 125 participants; very low‐certainty evidence), and total hospital cost (MD ‐2800.00 dollars, 95% CI ‐3881.74 to ‐1718.26 dollars; 1 study, 125 participants; very low‐certainty evidence) compared with no CVVH but the evidence is very uncertain. Adverse events and quality of life were not reported in the studies.

One type ofCVVH versus a different type of CVVH

We included one RCT in which 30 participants with severe acute pancreatitis were randomized to undergo high‐volume CVVH (15 participants) or standard CVVH (15 participants). High‐volume CVVH may result in little to no difference in in‐hospital mortality compared with standard CVVH (RR 0.60, 95% CI 0.17 to 2.07; 1 study, 30 participants; low‐certainty evidence). The evidence is very uncertain about the effect of high‐volume CVVH on adverse events compared with standard CVVH (RR 1.00, 95% CI 0.16 to 6.20; 1 study, 30 participants; very low‐certainty evidence). Length of ICU stay, length of hospital stay, total hospital cost, and quality of life were not reported in the study.

Authors' conclusions

The certainty of the current evidence is very low or low. For both comparisons addressed in this review, data are sparse. The evidence is very uncertain about the effect of CVVH on mortality in patients with severe acute pancreatitis. Very low‐certainty evidence suggests CVVH may reduce length of ICU stay, length of hospital stay, and total hospital cost but the evidence is very uncertain. The evidence is also very uncertain whether high‐volume CVVH is superior, equivalent or inferior to standard CVVH in patients with severe acute pancreatitis.

Plain language summary

Continuous venous blood filtration for severe acute pancreatitis

Review question

Can continuous venous blood filtration reduce the number of deaths in people with severe acute pancreatitis?

Background

Severe acute pancreatitis is associated with a high death rate and life‐threatening complications. Continuous veno‐venous hemofiltration (CVVH; continuous filtration of blood from a vein) has been used in some centers to reduce complications and death, but it is unknown if CVVH is beneficial for patients.

Search date

The evidence is current to September 2019.

Study characteristics

We searched for all relevant, well‐conducted studies conducted up to September 2019. We included three randomized controlled trials (experiments in which participants are randomly allocated to two or more interventions, possibly including a control intervention or no intervention, and the results are compared). We included three studies, involving a total of 219 adults with severe acute pancreatitis due to various reasons, including alcoholic abuse, biliary disease, high‐fat diet, excess lipids in the blood, etc. All people had single or multiple organ failure. All of the studies were single‐center studies conducted in China. The mean age of participants was 47.4 years. The mean proportion of females was 42.0%. Two studies randomized 189 people with severe acute pancreatitis to receive either CVVH (97 people) or no CVVH (92 people). One study randomized 30 people with severe acute pancreatitis to receive either high‐volume (high‐speed) CVVH (15 people) or standard CVVH (15 people).

Study funding sources

One study was sponsored by a non‐commercial grant. The other two studies did not report on funding sources.

Key results: CVVH versus no CVVH

We cannot tell from our results whether CVVH has an important effect on in‐hospital deaths for people with severe acute pancreatitis because the sample size was small. CVVH may reduce length of stay in the intensive care unit, length of hospital stay, and total hospital cost. However, the evidence is very uncertain because both studies had some limitations and the results were imprecise.

Key results: One type ofCVVH versus a different type of CVVH

High‐volume CVVH may result in little to no difference in numbers of in‐hospital deaths. We are uncertain whether high‐volume CVVH reduces adverse events. We cannot tell from our results whether high‐volume CVVH is superior, equivalent or inferior to standard CVVH for people with severe acute pancreatitis because the sample size was small and the results were imprecise.

Certainty of the evidence

Most of the included studies had some limitations in terms of how they were conducted or reported. Overall, the certainty of the evidence ranged from very low to low.

Summary of findings

Summary of findings 1. Continuous veno‐venous hemofiltration (CVVH) compared to no CVVH for severe acute pancreatitis.

CVVH compared with no CVVH for severe acute pancreatitis
Patient or population: adults with severe acute pancreatitis Settings: in hospital Intervention: CVVH Comparison: no CVVH
Outcomes	Anticipated absolute effects* (95% CI)		Relative effect (95% CI)	№ of participants (studies)	Certainty of the evidence (GRADE)	Comments
	Risk with no CVVH	Risk with CVVH
All‐cause mortality Follow‐up: in hospital	130 per 1,000	81 per 1,000 (35 to 183)	RR0.62 (0.27 to 1.40)	189 (2 studies)	⊕⊝⊝⊝ Verylow1,2,3	The evidence is very uncertain about the effect of CVVH on in‐hospital mortality compared with no CVVH.
Adverse events	Not reported
Length of ICU stay Follow‐up: in hospital	The mean length of ICU stay in the no CVVH group was 21.3 days	The mean length of ICU stay in the CVVH group 8.8 days lower (10.24 days lower to 7.36 days lower)	MD‐8.80 (‐10.24, ‐7.36)	125 (1 study)	⊕⊝⊝⊝ Verylow2.4,5	The evidence suggests that CVVH may reduce length of ICU stay compared with no CVVH but the evidence is very uncertain.
Length of hospital stay Follow‐up: in hospital	The mean length of hospital stay in the CVVH group was 61.4days	The mean length of hospital stay in the CVVH group 26.4 days lower (30.17 days lower to 22.63 lower days)	MD‐26.40 (‐30.17, ‐22.63)	125 (1 study)	⊕⊝⊝⊝ Verylow2,4,5	The evidence suggests that CVVH may reduce length of hospital stay compared with no CVVH but the evidence is very uncertain.
Total hospital cost Follow‐up: in hospital	The mean total hospital cost in the no CVVH group was 8300.0 dollars	The mean total hospital cost in the CVVH group was MD 2800 dollars lower (3881.74 dollars lower to 1718.26 lower dollars)	MD ‐2800.00 (‐3881.74, ‐1718.26)	125 (1 study)	⊕⊝⊝⊝ Verylow2,4,5	The evidence suggests that CVVH may reduce total hospital cost compared with no CVVH but the evidence is very uncertain.
Quality of life	Not reported
*The basis for the assumed risk was the control group proportion in the study. The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the control group and the relative effect of the intervention (and its 95% CI). CVVH: continuous veno‐venous hemofiltration; CI: confidence interval; ICU: intensive care unit; RR: risk ratio; MD: mean difference.
GRADE Working Group grades of evidence High certainty: We are very confident that the true effect lies close to that of the estimate of the effect Moderate certainty: We are moderately confident in the effect estimate: The true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different Low certainty: Our confidence in the effect estimate is limited: The true effect may be substantially different from the estimate of the effect Very low certainty: We have very little confidence in the effect estimate: The true effect is likely to be substantially different from the estimate of effect

¹ Not downgraded for risk of bias because this is considered an objective outcome which is unlikely to be affected by performance bias

² Downgraded one level for serious indirectness (the study population had zero in‐hospital mortality even among the control group in one study; it is unlikely that these patients were truly suffering from severe acute pancreatitis, which is normally associated with substantial mortality rates).

³ Downgraded two levels for very serious imprecision (small sample sizes, very few events, and wide confidence intervals that include both potential benefit and potential harm from the intervention)

⁴ Downgraded two levels for very serious risk of bias: high risk of selective reporting and high risk of performance bias and this outcome determined largely by the doctors

⁵ Downgraded one level due to serious imprecision: total population size was less than 400

Summary of findings 2. High‐volume continuous veno‐venous hemofiltration (CVVH) compared to standard CVVH for severe acute pancreatitis.

High‐volume CVVH compared to standard CVVH for severe acute pancreatitis
Patient or population: adults with severe acute pancreatitis Settings: in hospital Intervention: high‐volume CVVH Comparison: standard CVVH
Outcomes	Anticipated absolute effects* (95% CI)		Relative effect (95% CI)	№ of participants (studies)	Certainty of the evidence (GRADE)	Comments
	Risk with standard CVVH	Risk with high‐volume CVVH
All‐cause mortality Follow‐up: in hospital	333 per 1,000	200 per 1,000 (57 to 690)	RR 0.60 (0.17 to 2.07)	30 (1 study)	⊕⊕⊝⊝ Low1,2	The evidence suggests that high‐volume CVVH may result in little to no difference in in‐hospital mortality compared with standard CVVH.
Adverse events Follow‐up: in hospital	133 per 1,000	133 per 1,000 (21 to 827)	RR 1.00 (0.16 to 6.20)	30 (1 study)	⊕⊝⊝⊝ Very low2,3	The evidence is very uncertain about the effect of high‐volume CVVH on adverse events compared with standard CVVH.
Length of ICU stay	Not reported
Length of hospital stay	Not reported
Total hospital cost	Not reported
Quality of life	Not reported
* The basis for the assumed risk was the control group proportion in the study. The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the control group and the relative effect of the intervention (and its 95% CI). CVVH: continuous veno‐venous hemofiltration; CI: confidence interval; ICU: intensive care unit; RR: risk ratio.
GRADE Working Group grades of evidence High certainty: We are very confident that the true effect lies close to that of the estimate of the effect Moderate certainty: We are moderately confident in the effect estimate: The true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different Low certainty: Our confidence in the effect estimate is limited: The true effect may be substantially different from the estimate of the effect Very low certainty: We have very little confidence in the effect estimate: The true effect is likely to be substantially different from the estimate of effect

¹ Not downgraded for risk of bias because this is considered an objective outcome which is unlikely to be affected by selection bias and performance bias
² Downgraded two levels for very serious imprecision (small sample sizes, very few events, and wide confidence intervals that included both potential benefit and potential harm from the intervention)
³ Downgraded one level for serious risk of bias: unclear risk of selection bias and performance bias

Background

See Appendix 1 for a glossary of terms.

Description of the condition

Acute pancreatitis is a sudden inflammatory process of the pancreas and a common acute abdominal emergency (Banks 2013; Whitcomb 2006). Although the exact incidence of acute pancreatitis worldwide is unknown, the estimated incidence ranges from five to 30 cases per 100,000 people per year and is increasing (Omdal 2011; Roberts 2013; Spanier 2008; Yadav 2006). The two most common causes of acute pancreatitis are gallstone migration and alcohol abuse, which account for more than three‐quarters of acute pancreatitis cases (Roberts 2013; Villatoro 2010; Yadav 2006).

It is international consensus that a diagnosis of acute pancreatitis requires two of the following three features (Banks 2013): 1) acute onset of a persistent, severe, epigastric pain often radiating to the back; 2) serum lipase activity (or amylase activity) at least three times greater than the upper limit of normal; and 3) characteristic findings of acute pancreatitis on contrast‐enhanced computed tomography (CECT) and, less commonly, magnetic resonance imaging (MRI) or transabdominal ultrasonography.

Acute pancreatitis can be classified as either mild, moderately severe or severe (Banks 2013). Most people who develop pancreatitis develop mild acute pancreatitis. The proportion of people with severe acute pancreatitis has been found to vary between 5% and 20% (Banks 2013). The prognosis of acute pancreatitis depends on the severity of acute pancreatitis. The general mortality rate of acute pancreatitis is estimated to be between 5% and 10% (Bradley 1993; Roberts 2013). In people with mild acute pancreatitis, there is no organ failure and no local or systemic complications (Banks 2013). Mild acute pancreatitis is often a self‐limiting illness and mortality is very rare (Banks 2013). In patients with moderately severe acute pancreatitis, there may be transient organ failure (within 48 hours), as well as local or systemic complications (Banks 2013). Mortality due to moderately severe acute pancreatitis is about 2% (Vege 2009). In people with severe acute pancreatitis, there is persistent single or multiple organ failure (more than 48 hours) and the mortality rate is high, ranging from 25% to 46% (Johnson 2004; Mofidi 2006; Vege 2009). Local complications include acute peripancreatic fluid collection, pancreatic pseudocyst, acute necrotic collection, and walled‐off necrosis; while systemic complications include exacerbations of pre‐existing comorbidities related to acute pancreatitis (Banks 2013).

Although the pathogenesis of acute pancreatitis is not completely understood, it is believed to be associated with intra‐pancreatic activation of proteolytic enzymes (e.g. trypsin, elastase) (Boxhoorn 2020; Sah 2013; Whitcomb 2006). The uncontrolled activation of proteolytic enzymes can cause activation of the inflammatory pathways and release of inflammatory cytokines and oxidative stress, all of which contribute to the development of systemic inflammatory response syndrome (SIRS) (Banks 2013; Rada 2011; Sah 2013). Uncontrolled SIRS can affect various organs (e.g. lung, kidney, heart), which may result in single or multiple organ failure and death (Boxhoorn 2020).

Description of the intervention

The main objective when treating severe acute pancreatitis is to decrease mortality and avoid local or systemic complications (Bakker 2014; Gurusamy 2016a). Nowadays, the treatment for severe acute pancreatitis consists of intensive supportive management (e.g. respiration support and renal replacement therapy) (Bakker 2014; Borthwick 2017; Rada 2011); endoscopic or surgical intervention, in some situations (Gurusamy 2016a; Tenner 2013; Tse 2012); prevention of infection surrounding the pancreas (Villatoro 2010); resumption of enteral nutrition (Al‐Omran 2010; Poropat 2015); and management of local or systemic complications (Bakker 2014; Gurusamy 2016b; Vallance 2014).

People with both severe acute pancreatitis and acute renal failure (ARF) often require renal replacement therapy in the intensive care unit (ICU) (Bakker 2014; Fayad 2016; Tenner 2013; Yokoe 2015). In recent years, continuous veno‐venous hemofiltration (CVVH) has been introduced as an alternative to hemodialysis for the treatment of critically ill people with ARF (Borthwick 2017; Klouche 2002; Ronco 2000). CVVH is a blood purification technique, in which the person's blood is passed through a membrane filter that removes waste products and excess water by convection (Nistor 2015; Tian 2015). It is reported that CVVH may be beneficial for critically ill people with SIRS, as well as single or multiple organ failure (Bouman 2007; Honore 2009).

How the intervention might work

Inflammatory cytokines can initiate the inflammatory response and regulate the host defence against pathogens (Wikipedia 2017). Excessive inflammatory cytokines have been shown to cause symptoms correlated to a disease, e.g. fever, tissue damage, and, in some cases, shock and death (Dinarello 2000). Dysregulation of inflammatory cytokines has also been linked to inflammatory diseases (Scarpioni 2016). A balance between inflammatory cytokines and anti‐inflammatory cytokines is necessary to maintain health (Sallam 2016; Wikipedia 2017).

The rationale of CVVH for the treatment of severe acute pancreatitis is based on the mechanisms of injury in acute pancreatitis. Some clinical studies of acute pancreatitis have shown that inflammatory cytokines play a key role in local or systemic complications, and that cytokine‐induced SIRS causes organ failure and death in severe acute pancreatitis (Nieminen 2014; Norman 1998). CVVH can remove excessive inflammatory cytokines (e.g. interleukin 1 (IL‐1), interleukin 6 (IL‐6), interleukin 8 (IL‐8), tumor necrosis factor alpha (TNFα)) from the blood circulation and help maintain balance between inflammatory cytokines and anti‐inflammatory cytokines (De Vriese 1999; Tian 2015). Theoretically, CVVH may reduce the chain reaction initiated by inflammatory cytokines and limit damage resulting from SIRS in severe acute pancreatitis. As a result, CVVH may improve the prognosis of people with severe acute pancreatitis.

Why it is important to do this review

The role of CVVH for people with severe acute pancreatitis is controversial. CVVH may potentially reduce mortality and avoid local or systemic complications, but its efficacy and safety are uncertain. Recently, we identified some studies, which were not included in the previous version of this review (Lin 2019), through scanning reference lists of the systematic reviews on this topic (Guo 2020; Hu 2019). This is an amendment of a previous Cochrane Review assessing the role of CVVH for severe acute pancreatitis.

Objectives

To assess the benefits and harms of continuous veno‐venous hemofiltration (CVVH) in patients suffering from severe acute pancreatitis; to compare the effects of different CVVH techniques; and to evaluate the optimal time for delivery of CVVH.

Methods

Criteria for considering studies for this review

Types of studies

We included all randomized controlled trials (RCTs), including cluster RCTs — regardless of sample size, language, or publication status — which compared: 1) CVVH versus no CVVH, 2) different types of CVVH, or 3) different schedules for CVVH in participants with severe acute pancreatitis. We excluded quasi‐randomized studies, in which the allocation was performed on the basis of a pseudo‐random sequence (e.g. odd or even hospital number or date of birth, alternation), and non‐randomized studies, because of the potential for bias (Reeves 2021).

Types of participants

We included adults (irrespective of sex or race) who were diagnosed with severe acute pancreatitis according to the revised Atlanta criteria (persistent organ failure for more than 48 hours; Banks 2013). We included people with severe acute pancreatitis and ARF, as well as those with severe acute pancreatitis without ARF. We accepted the study authors' definitions of ARF.

Types of interventions

We assessed the following comparisons for people with severe acute pancreatitis:

• CVVH versus no CVVH;

• one type of CVVH versus a different type of CVVH; and

• early versus delayed CVVH.

There is no universally accepted definition of 'early' CVVH. In patients with severe acute pancreatitis, we considered any CVVH performed within three days of the onset of pancreatitis as early CVVH. The reason for choosing three days was that this allowed time for the clinicians to make the diagnosis of severe acute pancreatitis and organize the CVVH. We considered any CVVH performed after three days as 'delayed' CVVH.

We included any additional interventions provided they were not part of the randomized treatment.

Types of outcome measures

Primary outcomes

• All‐cause mortality (in‐hospital mortality or mortality within three months).

• Adverse events (number of people with at least one adverse event within three months). We accepted all adverse events reported by the study authors, irrespective of the severity of the adverse event.

Secondary outcomes

• Length of ICU stay.

• Length of hospital stay.

• Total hospital cost.

• Quality of life (using any validated scale).

The main reason to justify CVVH is the assumption that it will reduce mortality, therefore we chose mortality as our primary outcome. Other clinical outcomes were chosen to assess whether CVVH resulted in a shorter ICU stay, earlier discharge from the hospital, improvement in health‐related quality of life and cost‐effectiveness.

Length of ICU stay or length of hospital stay might not follow a normal distribution. Therefore, we performed analysis using mean differences or standardized mean differences with caution. Reporting of the outcomes listed here was not an inclusion criterion for the review.

Search methods for identification of studies

We designed the search strategies with the help of the Cochrane Gut Group Information Specialist before performing literature searches. No restrictions were placed on the language of publications when searching the electronic databases or reviewing reference lists in identified studies.

Electronic searches

We conducted a literature search to identify all published and unpublished RCTs. The literature search identified potential studies in all languages. We translated the non‐English language papers and fully assessed them for potential inclusion in the review as necessary.

We searched the following electronic databases for identifying potential studies:

• Cochrane Central Register of Controlled Trials (CENTRAL) (September 2019) (Appendix 2);

• MEDLINE (1966 to September 2019) (Appendix 3);

• Embase (1988 to September 2019) (Appendix 4); and

• Science Citation Index Expanded (1982 to September 2019) (Appendix 5).

Searching other resources

We checked the reference lists of all primary studies and review articles for additional references. We contacted authors of identified studies and asked them to identify other published and unpublished studies. We also contacted manufacturers and experts in the field.

We searched PubMed (www.ncbi.nlm.nih.gov/pubmed) for errata or retractions from eligible studies and reported the date this was done within the review.

Grey literature databases

We searched the following databases to identify unpublished studies (accessed 13 September 2019):

• OpenGrey opengrey.eu

• PsycEXTRA www.apa.org/pubs/databases/psycextra/index

Clinical trials registers/trial result registers

We searched the following databases to identify ongoing studies (accessed 13 September 2019):

• ClinicalTrials.gov;

• World Health Organization International Clinical Trials Registry Platform Search Portal;

• Chinese Clinical Trial Register.

Data collection and analysis

Selection of studies

Two review authors (Junhua Gong and XD) independently screened titles and abstracts for inclusion. All the potential studies we identified as a result of the search were coded as either 'retrieve' (eligible, potentially eligible, or unclear) or 'do not retrieve'. We retrieved the full text of potentially eligible studies and two review authors (Junhua Gong and XD) independently screened the full text, identified studies for inclusion, and recorded reasons for exclusion of the ineligible studies. We resolved any disagreement through discussion or, if required, consulted a third review author (YC). We identified and excluded duplicates and collated multiple reports of the same study so that each study rather than each report was the unit of interest in the review. We recorded the selection process in sufficient detail to complete a PRISMA flow diagram and a 'Characteristics of excluded studies' table.

Data extraction and management

We used a standardized data collection form for study characteristics and outcome data, which was piloted on at least one study included in the review. Two review authors (ZL and ZZ) extracted the following study characteristics from the included studies.

• Methods: study design, total study duration and run‐in (the time period before entering a clinical trial, in which participants may discontinue or begin treatment), number of study centers and location, study setting, withdrawals, date of study.

• Participants: number (N), mean age, age range, gender, severity of condition, diagnostic criteria, inclusion criteria, exclusion criteria.

• Interventions: intervention, comparison.

• Outcomes: primary and secondary outcomes specified and collected, time points reported.

• Notes: funding for study, notable conflicts of interest of study authors.

Two review authors (ZL and ZZ) independently extracted outcome data from included studies. We noted in the 'Characteristics of included studies' table if the study authors reported outcome data in an unusable way. We resolved disagreements by consensus or by involving a third review author (Jianping Gong). One review author (ZL) copied the data from the data collection form into the Review Manager 5 file (Review Manager 2020). We double‐checked that the data were entered correctly by comparing the study reports with how the data were presented in the systematic review. A second review author (ZZ) spot‐checked study characteristics for accuracy against the study report.

Assessment of risk of bias in included studies

Two review authors (Junhua Gong and ZZ) independently assessed risk of bias for each included study using the criteria outlined in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2017). We resolved any disagreement by discussion or by involving a third review author (YC). We assessed the risk of bias according to the following domains:

• random sequence generation;

• allocation concealment;

• blinding of participants and personnel;

• blinding of outcome assessment;

• incomplete outcome data;

• selective outcome reporting; and

• other bias.

We graded each potential source of bias as high, low, or unclear and provided a quote from the study report along with a justification for our judgment in the ‘Risk of bias’ table. We summarized the ‘Risk of bias’ judgments across different studies for each of the domains listed. We considered blinding separately for different key outcomes where necessary, e.g. for unblinded outcome assessment, risk of bias for all‐cause mortality may be very different than for a person‐reported pain scale). Where information on risk of bias related to unpublished data or correspondence with a trial author, we noted this in the ‘Risk of bias' table.

We considered a trial to be at low risk of bias overall if we assessed it as being at low risk of bias across all domains. Otherwise, we considered trials to be at high risk of bias overall if they were at unclear or high risk of bias for one or more domains. We resolved any difference in opinion by discussion. When considering treatment effects, we took into account the risk of bias for the studies that contributed to that outcome.

Assessment of bias in conducting the systematic review

We conducted the review according to the published protocol (Cheng 2018) and reported any deviations from it in Differences between protocol and review.

Measures of treatment effect

We analyzed dichotomous data as a risk ratio (RR) and continuous data as a mean difference (MD) or a standardized mean difference (SMD). We ensured that higher scores for continuous outcomes had the same meaning for the particular outcome, explained the direction to the reader, and reported where the directions were reversed if this was necessary.

We undertook meta‐analyses only where this was meaningful, i.e. if the treatments, participants, and the underlying clinical question were similar enough for pooling to make sense.

A common way that trialists indicated when they had skewed data was by reporting medians and interquartile ranges. When we encountered this we noted that the data were skewed and considered the implication of this.

Where multiple study arms were reported in a single study, we included only the relevant arms. If two comparisons (e.g. drug A versus placebo and drug B versus placebo) had to be entered into the same meta‐analysis, we halved the control group to avoid double counting.

Unit of analysis issues

The unit of analysis was individual participants with severe acute pancreatitis. We did not find any cross‐over or cluster‐randomized trials for this comparison.

Dealing with missing data

We contacted investigators or study sponsors in order to verify key study characteristics, and obtained missing numerical outcome data where possible (e.g. when a study was identified as abstract only).

Assessment of heterogeneity

We used the I² statistic to measure heterogeneity among the studies in each analysis (Higgins 2003). If we identified substantial heterogeneity (greater than 50%), we explored it by pre‐specified subgroup analysis, and we interpreted summary effect measures with caution.

Assessment of reporting biases

We attempted to contact study authors, asking them to provide missing outcome data. When this was not possible, and the missing data were thought to introduce serious bias, we explored the impact of including such studies in the overall assessment of results by a sensitivity analysis.

If we were able to pool more than 10 studies, we intended to create and examine a funnel plot to explore possible publication biases. Visual asymmetry in the funnel plot was used to determine the reporting biases (Higgins 2021).

Data synthesis

We performed the meta‐analyses using Review Manager 5 (Review Manager 2020). For all analyses, we employed a random‐effects model.

‘Summary of findings' table

We created a 'Summary of findings' table using the following outcomes: all‐cause mortality, adverse events, length of ICU stay, length of hospital stay, total hospital cost, and quality of life. We used the five GRADE considerations (study limitations, consistency of effect, imprecision, indirectness and publication bias) to assess the certainty of a body of evidence as it related to the studies which contributed data to the meta‐analyses for the pre‐specified outcomes. We used methods and recommendations described in Chapter 14 of the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2021), and GRADEpro GDT software (GRADEpro 2015). Two review authors (ZL, ZZ) independently justified all decisions to downgrade their assessments of the certainty of studies. We justified all decisions to downgrade our assessments using footnotes and we made comments to aid the reader's understanding of the review where necessary. We considered whether there was any additional outcome information that could not be incorporated into meta‐analyses, noted this in the comments, and stated if it supports or contradicts the information from the meta‐analyses.

We defined the certainty of levels of evidence as 'high', 'moderate', 'low', or 'very low' as follows:

• High certainty: we are very confident that the true effect lies close to that of the estimate of the effect.

• Moderate certainty: we are moderately confident in the effect estimate; the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different.

• Low certainty: our confidence in the effect estimate is limited; the true effect may be substantially different from the estimate of the effect.

• Very low certainty: we have very little confidence in the effect estimate; the true effect is likely to be substantially different from the estimate of effect.

Subgroup analysis and investigation of heterogeneity

We planned to carry out the following subgroup analyses:

• severe acute pancreatitis with ARF versus severe acute pancreatitis without ARF; and

• different etiologies (e.g. alcohol abuse, biliary disease, familial chylomicronemia condition, and others).

The following outcomes were used in subgroup analyses:

• all‐cause mortality; and

• adverse events.

We used the formal Chi² test for subgroup differences to test for subgroup interactions.

Sensitivity analysis

We planned to perform a sensitivity analysis, defined a priori, to assess the robustness of our conclusions. This involved:

• changing between worst‐case scenario analysis and best‐case scenario analysis for missing data; and

• excluding studies in which either the mean or standard deviation, or both, were imputed.

Reaching conclusions

We based our conclusions only on findings from the quantitative or narrative synthesis of included studies for this review. We avoided making recommendations for practice and our implications for research give the reader a clear sense of where the focus of any future research in the area should be and what the remaining uncertainties are.

Results

Description of studies

See: Characteristics of included studies; Characteristics of excluded studies.

Results of the search

We identified 562 records through the electronic searches of the Cochrane Library (82 records), MEDLINE (Ovid) (162 records), Embase (Ovid) (284 records), and Science Citation Index Expanded (Web of Science) (34 records). We identified seven records through scanning reference lists of the identified non‐Cochrane reviews (Guo 2020; Hu 2019). We excluded 96 duplicates and 451 clearly irrelevant records after reading the titles and abstracts. We retrieved the remaining 22 records for further assessment. We excluded 18 studies for the reasons listed in the Characteristics of excluded studies table and one ongoing study (Zhao 2012). In total, three RCTs fulfilled the inclusion criteria for this review ( Chu 2013; Guo 2016; Wang 2017). The study flow diagram is shown in Figure 1.

1.

Study flow diagram.

Included studies

Two RCTs that were included by a non‐Cochrane review (Guo 2020) were found to be missed in the last published version. One RCT was added in this amendment (Wang 2017). However, the other RCT was excluded because it did not report any relevant outcome data for this review (Liu 2017). We included three RCTs, involving a total of 219 participants, in the review (Chu 2013; Guo 2016; Wang 2017). Details of the RCTs are shown in the Characteristics of included studies table.

CVVH versus no CVVH

Guo 2016 included 64 participants with severe acute pancreatitis complicated with acute lung injury, who were randomized to either undergo CVVH (N = 32) or receive no CVVH (N = 32). This study did not mention the etiologies of severe acute pancreatitis. This RCT was a single‐center study conducted in China. The mean age of participants was 51.9 years. This study measured mortality, laboratory and physiological variables. The study authors declared no conflict of interest and did not report any funding sources.

Wang 2017 included 245 participants with severe acute pancreatitis complicated with persistent organ failure, who were randomized to CVVH (N = 65), no CVVH (N = 60), laparoscopic peritoneal lavage and drainage (LDLP) (N = 62) and CVVH+LDLP (N = 58). We did not include participants in either LDLP group or CVVH+LDLP group in this review. The etiologies of severe acute pancreatitis included alcoholic abuse (20.8%), biliary disease (39.2%), hyperlipidemia (20.8%), and other diseases (19.2%). This RCT was a single‐center study conducted in China. The mean age of participants was 46.1 years. This study measured mortality, length of ICU stay, length of hospital stay, total hospital cost, laboratory and physiological variables. This study was sponsored by non‐commercial grants (National Natural Science Foundation of China and the Science Foundation of Science and Technology Hall of Jiangxi Province). The study authors declared no conflict of interest.

One type of CVVH versus a different type of CVVH

Chu 2013 included 30 participants with severe acute pancreatitis complicated with multiple organ dysfunction syndrome, who were randomized to either undergo high‐volume CVVH (N = 15; 85 ml/kg per hour) or receive standard CVVH (N = 15; 35 ml/kg per hour). The etiologies of severe acute pancreatitis included alcoholic abuse (10.0%), biliary disease (66.7%), high‐fat diet (16.7%), and hyperlipidemia (6.7%). This RCT was a single‐center study conducted in China. The mean age was of participants 43.1 years. This study measured mortality, adverse events, laboratory and physiological variables. Neither funding source nor conflict of interest was reported in this study.

Excluded studies

We excluded 18 studies. Details are listed in the Characteristics of excluded studies table. We excluded two RCTs because they assessed the role of CVVH for mild acute pancreatitis, but not for severe acute pancreatitis (He 2016; Jiang 2005). We excluded another six RCTs because the severe acute pancreatitis was not diagnosed with the revised Atlanta criteria (Gao 2018; Mao 1999; Xia 2012; Yang 2004; Yang 2010; Zhang 2010). We excluded one RCT because this study did not report any relevant outcome data for this review (Liu 2017). We excluded a quasi‐randomized study because the allocation was performed on the basis of a pseudo‐random sequence (odd or even hospital number) (Abulimiti 2018). The remaining studies were identified as non‐randomized studies (Aleksandrova 2012; Cui 2014; Gong 2010; Guo 2014; He 2013; Sun 2015; Wang 2015; Zhang 2014).

Ongoing studies

We found one ongoing study, in which fifty‐six participants with severe acute pancreatitis will be randomized to high‐volume CVVH or no CVVH (Zhao 2012). This trial is currently recruiting participants. It is being performed in China, and was initiated in April 2012. The primary outcomes are persistent organ failure and death. The secondary outcomes are infectious complications, input fluid volume, output fluid volume, physiological parameters, number of surgical interventions, length of ICU stay, total hospital costs, persistent multiple organ dysfunction syndrome, new‐onset organ failure, and new‐onset multiple organ dysfunction syndrome.

Risk of bias in included studies

All trials were at high risk of bias (Chu 2013; Guo 2016; Wang 2017). For graphical representations of the risk of bias, see Figure 2 and Figure 3.

2.

Risk of bias graph: review authors' judgements about each risk of bias item presented as percentages across all included studies.

3.

Risk of bias summary: review authors' judgements about each risk of bias item for each included study.

Allocation

Random sequence generation was at low risk of bias in one study where participants were randomized using a random number table (Wang 2017), and unclear risk of bias in two studies as no specific information was provided regarding the randomization process with no response from the authors (Chu 2013; Guo 2016). Allocation concealment was at unclear risk of bias in three studies with no information provided (Chu 2013; Guo 2016; Wang 2017).

Blinding

Due to the nature of the interventions, participants and performing doctors could not be blinded to group allocation (CVVH versus no CVVH) in two studies (Guo 2016; Wang 2017). We, therefore, judged both studies to be at high risk of blinding of participants and personnel (Guo 2016; Wang 2017). Blinding of participants and personnel was at unclear risk of bias in the other study (high‐volume CVVH versus standard CVVH) with no information provided (Chu 2013). The outcome parameters relating to time and cost (length of ICU stay, length of hospital stay, total hospital cost) were judged to be at risk of performance bias as these outcomes can be largely determined by the doctors.

All studies were at unclear risk of bias for blinding of outcome assessment with no information provided (Chu 2013; Guo 2016; Wang 2017).

Incomplete outcome data

There were no post‐randomization dropouts in all studies (Chu 2013; Guo 2016; Wang 2017). Thus we judged all studies to be at low risk of bias regarding incomplete outcome data (Chu 2013; Guo 2016; Wang 2017).

Selective reporting

The trial protocol was not available for all studies (Chu 2013; Guo 2016; Wang 2017). One study included results for key outcomes that would be expected to have been reported (Chu 2013). The review authors judged this study to be free of selective reporting (Chu 2013). The other two studies failed to include results for key outcomes that would be expected to have been reported for such studies (Guo 2016; Wang 2017). Thus, we judged both studies to be at high risk of bias regarding selective reporting (Guo 2016; Wang 2017).

Other potential sources of bias

All studies were at low risk of bias for baseline imbalance (Chu 2013; Guo 2016; Wang 2017).

Effects of interventions

See: Table 1; Table 2

See: Table 1; Table 2.

CVVH versus no CVVH

Two studies (189 participants) compared CVVH with no CVVH (Guo 2016; Wang 2017). Ninety‐seven participants were randomized to the CVVH group and 92 participants to the no‐intervention group.

All‐cause mortality (in hospital)

Both studies including 189 participants (97 CVVH, 92 no CVVH) reported on the in‐hospital mortality (Guo 2016; Wang 2017). Overall, there were twenty deaths in 720 participants: eight in the CVVH group and 12 in the no CVVH group. The evidence is very uncertain about the effect of CVVH on in‐hospital mortality compared with no CVVH (RR 0.62, 95% CI 0.27 to 1.40; 2 studies, 189 participants; very low‐certainty evidence; Analysis 1.1).

1.1. Analysis.

Comparison 1: Continuous veno‐venous hemofiltration (CVVH) vs no CVVH, Outcome 1: All‐cause mortality

Adverse events

Neither studies report this outcome.

Length of ICU stay

One study including 125 participants (65 CVVH, 60 no CVVH) reported on the length of ICU stay (Wang 2017). The other study did not report this outcome (Guo 2016). The evidence suggests that CVVH may reduce length of ICU stay compared with no CVVH but the evidence is very uncertain (MD ‐8.80 days, 95% CI ‐10.24 to ‐7.36 days; 1 study, 125 participants; very low‐certainty evidence; Analysis 1.2).

1.2. Analysis.

Comparison 1: Continuous veno‐venous hemofiltration (CVVH) vs no CVVH, Outcome 2: Length of ICU stay

Length of hospital stay

One study including 125 participants (65 CVVH, 60 no CVVH) reported on the length of hospital stay (Wang 2017). The other study did not report this outcome (Guo 2016). The evidence suggests that CVVH may reduce length of hospital stay compared with no CVVH but the evidence is very uncertain (MD ‐26.40 days, 95% CI ‐30.17 to ‐22.63 days; 1 study, 125 participants; very low‐certainty evidence; Analysis 1.3).

1.3. Analysis.

Comparison 1: Continuous veno‐venous hemofiltration (CVVH) vs no CVVH, Outcome 3: Length of hospital stay

Total hospital cost

One study including 125 participants (65 CVVH, 60 no CVVH) reported on the total hospital cost (Wang 2017). The other study did not report this outcome (Guo 2016). The evidence suggests that CVVH may reduce total hospital cost compared with no CVVH but the evidence is very uncertain (MD ‐2800.00 dollars, 95% CI ‐3881.74 to ‐1718.26 dollars; 1 study, 125 participants; very low‐certainty evidence; Analysis 1.4).

1.4. Analysis.

Comparison 1: Continuous veno‐venous hemofiltration (CVVH) vs no CVVH, Outcome 4: Total hospital cost (dollars)

Quality of life

Neither studies report this outcome.

One type of continuous veno‐venous hemofiltration versus a different type of CVVH

One study (30 participants) compared high‐volume CVVH with standard CVVH (Chu 2013). Fifteen participants were randomized to the high‐volume CVVH group and 15 participants to the standard CVVH group.

All‐cause mortality (in hospital)

One study including 30 participants (15 high‐volume CVVH, 15 standard CVVH) reported on the in‐hospital mortality (Chu 2013). Overall, there were eight deaths in 30 participants: three in the high‐volume CVVH group and five in the standard CVVH group. The evidence suggests that high‐volume CVVH may result in little to no difference in in‐hospital mortality compared with standard CVVH (RR 0.60, 95% CI 0.17 to 2.07; 1 study, 30 participants; low‐certainty evidence; Analysis 2.1).

2.1. Analysis.

Comparison 2: High‐volume continuous veno‐venous hemofiltration (CVVH) versus standard CVVH, Outcome 1: All‐cause mortality

Adverse events (in hospital)

One study including 30 participants (15 high‐volume CVVH, 15 standard CVVH) reported on the adverse events (Chu 2013). Overall, there were four adverse events in 30 participants: two in the high‐volume CVVH group and two in the standard CVVH group. The evidence is very uncertain about the effect of high‐volume CVVH on adverse events compared with standard CVVH (RR 1.00, 95% CI 0.16 to 6.20; 1 study, 30 participants; very low‐certainty evidence; Analysis 2.2).

2.2. Analysis.

Comparison 2: High‐volume continuous veno‐venous hemofiltration (CVVH) versus standard CVVH, Outcome 2: Adverse events

Length of ICU stay

The study did not report this outcome.

Length of hospital stay

The study did not report this outcome.

Total hospital cost

The study did not report this outcome.

Quality of life

The study did not report this outcome.

Subgroup analysis

We did not perform any of the planned subgroup analyses because the studies included in each comparison were too few.

Sensitivity analysis

We did not perform any of the planned sensitivity analyses because there was no post‐randomization dropout or imputed data.

Discussion

Summary of main results

This review included evidence from three studies, consisting of 219 participants with severe acute pancreatitis. For the comparison of CVVH versus no CVVH, the evidence is very uncertain about the effect of CVVH on in‐hospital mortality. we found that CVVH may reduce length of ICU stay, length of hospital stay, and total hospital cost compared with no CVVH but the evidence is very uncertain. For the comparison of high‐volume CVVH versus standard CVVH, we found that high‐volume CVVH may result in little to no difference in in‐hospital mortality. The evidence is very uncertain about the effect of high‐volume CVVH on adverse events compared with standard CVVH.

A previous version of this review included two studies with 94 participants with severe acute pancreatitis (Lin 2019). A small study (30 participants) compared high‐volume CVVH versus standard CVVH group (30 participants). Another small study compared CVVH versus no CVVH (64 participants) and suggested it was unclear whether CVVH had any effect on mortality or complications in patients with severe acute pancreatitis. With one recent study added to this review (Wang 2017), this review showed that CVVH was associated with better outcomes in terms of length of ICU stay, length of hospital stay, and total hospital cost when compared CVVH versus no CVVH but the evidence is very uncertain.

Overall completeness and applicability of evidence

All trials included adults with severe acute pancreatitis with various etiologies, including alcoholic abuse, biliary disease, high‐fat diet, hyperlipidemia, etc (Chu 2013; Guo 2016; Wang 2017). All of the participants had single or multiple organ failure. Thus, the results of this review are applicable to adults with severe acute pancreatitis for various reasons, especially for people with single or multiple organ failure.

Quality of the evidence

For the comparison of CVVH versus no CVVH, we rated the certainty of evidence for four outcomes. We found very low‐certainty evidence for in‐hospital mortality, length of ICU stay, length of hospital stay, and total hospital cost. The certainty of evidence for in‐hospital mortality was not downgraded for risk of bias because it was unlikely to be affected by performance bias. We downgraded the certainty of evidence for length of ICU stay, length of hospital stay, and total hospital cost by two levels for very serious risk of bias because of selective reporting and unblinding of participants and personnel, as they were determined largely by the doctors. The certainty of evidence for in‐hospital mortality was downgraded two levels for very serious imprecision because of small sample sizes, very few events, and wide confidence intervals that included both potential benefit and potential harm from the intervention. We downgraded the certainty of evidence for length of ICU stay, length of hospital stay, and total hospital cost by one level for serious imprecision because of small sample sizes. We downgraded the certainty of the evidence for all outcomes assessed by one level for indirectness (the study population had zero in‐hospital mortality even among the control group in one study (Guo 2016); it is unlikely that these patients were truly suffering from severe acute pancreatitis, which is normally associated with substantial mortality rates). There were too few studies included in this comparison to assess publication bias (Table 1).

For the comparison of high‐volume CVVH versus standard CVVH, we rated the certainty of evidence for two outcomes. We found low‐certainty evidence for in‐hospital mortality and very low‐certainty evidence for adverse events. The certainty of evidence for in‐hospital mortality was not downgraded for risk of bias because it was unlikely to be affected by selection bias and performance bias. We downgraded the certainty of evidence for adverse event by one level for very serious risk of bias because of unclear risk of selection bias and performance bias. The certainty of evidence for in‐hospital mortality and adverse events was downgraded two levels for very serious imprecision because of small sample sizes, very few events, and wide confidence intervals that included both potential benefit and potential harm from the intervention. There was no indirectness of evidence, as we included only one RCT for this comparison. There were too few studies included in this comparison to assess publication bias (Table 2).

Potential biases in the review process

There were several potential biases of note in the review process. First, whilst we performed a thorough search of the literature and identified three studies (Chu 2013; Guo 2016; Wang 2017), all studies were of small sample sizes and the mortality rates in participants with severe acute pancreatitis were relatively low. Our review may not have adequate power to rule out clinically important differences in mortality between the two management strategies. Second, some potentially relevant data were not reported in included studies. We were able to obtain some of these data by contacting the authors of the studies, but despite repeated contact, some data remained missing. There was incomplete correspondence with study investigators. Additionally, we were unable to explore publication bias because we did not have access to protocols for the included studies, and we did not use a funnel plot to test for publication bias because the included studies were too few. Based on the available data, the question of whether CVVH has any effect on the reduction of mortality for people with severe acute pancreatitis remains unanswered. This highlights the need for more randomized trials, with sufficient sample sizes, on this topic.

Agreements and disagreements with other studies or reviews

The systematic review by Wang and colleagues included ten trials, including two RCTs and eight non‐randomized studies (Wang 2013). Because of the heterogeneity and lack of high‐level evidence, Wang 2013 did not draw a definitive conclusion on the role of CVVH for severe acute pancreatitis. The systematic review by Hu and colleagues included twelve trials, including four RCTs and eight non‐randomized studies (Hu 2019). Hu 2019 found that CVVH improved clinical outcomes for patients with severe acute pancreatitis when compared with conventional treatment. The systematic review by Guo and colleagues included nine trials, including seven RCTs and two non‐randomized studies (Guo 2020). Five of these seven RCTs included participants with severe acute pancreatitis without using the revised Atlanta criteria, so was not included in this review (Gao 2018; Mao 1999; Xia 2012; Yang 2004; Zhang 2010). Guo 2020 found that CVVH reduced mortality and improved clinical outcomes for patients with severe acute pancreatitis when compared with conventional treatment. We included three studies in the review (Chu 2013; Guo 2016; Wang 2017), which reported severe acute pancreatitis according to the updated Atlanta classification and definitions (Banks 2013). We found that the impact of CVVH on mortality was very uncertain for people with severe acute pancreatitis because the certainty of the evidence is very low.

Authors' conclusions

Implications for practice.

The certainty of the current evidence is very low or low. For both comparisons addressed in this review, data are sparse. The evidence is very uncertain about the effect of CVVH on mortality in patients with severe acute pancreatitis. Very low‐certainty evidence suggests CVVH may reduce length of ICU stay, length of hospital stay, and total hospital cost but the evidence is very uncertain. The evidence is also very uncertain whether high‐volume CVVH is superior, equivalent or inferior to standard CVVH in patients with severe acute pancreatitis.

Implications for research.

More trials with sufficient sample sizes are necessary to assess the benefits and harms of CVVH for people with severe acute pancreatitis. A sample size of 258 (129 in each group) would be required to detect an absolute reduction in the mortality of 10% (from 35% to 25%) at 80% power and an alpha‐error set at 0.05. Future trials should report severe acute pancreatitis according to the updated Atlanta classification and definitions (Banks 2013). Future trials should analyze the data on an intention‐to‐treat basis in case of postrandomization dropouts.

What's new

Date	Event	Description
10 April 2021	Amended	Two additional trials published within the search timeframe that were included by a non‐Cochrane review (Guo 2020) were not included in the last published version. One trial was added in this amendment (Wang 2017). However, the other trial was excluded because it did not report any relevant outcome data for this review (Liu 2017). Conclusions changed after including the trial.

History

Protocol first published: Issue 2, 2018
Review first published: Issue 10, 2019

Notes

One additional trial published within the search timeframe that was missed in the initial publication was incorporated in the latest amendment of this review. Any new studies published since the original publication date will be incorporated in a future update. 

Appendices

Appendix 1. Glossary of terms

Acute: sudden.

Amylase: an enzyme that break down starches into sugars.

Computed tomography: makes use of computer‐processed combinations of many X‐ray images taken from different angles to produce cross‐sectional images of specific areas.

Convection: the movement of groups of molecules within fluids.

Correlated: related by a correlation, especially having corresponding characteristics.

Cytokines: chemicals made by the cells that act on other cells to stimulate or inhibit their function.

Elastase: an enzyme of the pancreatic juice that digests elastin.

Endoscopic: with the help of an endoscope, a tube inserted into body.

Enteral: within, or by way of, the intestine or gastrointestinal tract.

Enzyme: proteins that speed up chemical reactions in the body without themselves undergoing change.

Epigastric: the upper central region of the abdomen.

Hemodialysis: a process of purifying the blood of a person whose kidneys are not working normally.

Hemofiltration: a renal replacement therapy to treat acute kidney injury, in which a person's blood is passed through a set of tubing via a machine to a semipermeable membrane where waste products and water are removed by convection.

Homeostasis: the property of a system in which a variable is actively regulated to remain very nearly constant.

Incidence: the number of new people diagnosed per year.

Lipase: an enzyme that break down fats into glycerol and fatty acids.

Magnetic resonance imaging: a medical imaging technique using strong magnetic fields, radio waves and field gradients to generate images of the inside of the body.

Morbidity: the proportion of people with any complications.

Mortality: the proportion of deaths.

Necrosis: the death or decay of body tissue.

Necrotic: relating to or affected by necrosis.

Organ failure: a condition where an organ does not perform its expected function and normal homeostasis cannot be maintained without external clinical intervention.

Oxidative stress: a condition of increased oxidant production in animal cells characterized by the release of free radicals and resulting in cellular degeneration.

Pancreatitis: inflammation of the pancreas.

Pathogens: any agent that can cause disease.

Pathogenesis: the origin and development of the disease.

Peripancreatic: surrounding the pancreas.

Prognosis: a medical term for predicting the likely outcome of one's current standing.

Proteolytic: relating to the breakdown of proteins into peptides and amino acids, as occurs during digestion.

Pseudocyst: a cystic lesion that may appear as a cyst, but lacks epithelial cells.

Renal: relating to the kidney.

Systemic inflammatory response syndrome: an inflammatory state affecting the whole body, frequently a response of the immune system to infection.

Transient: temporary.

Trypsin: an enzyme of the pancreatic juice, capable of converting proteins into peptides and amino acids.

Ultrasonography: the technique of using high‐frequency sound waves to produce pictures of structures within the body.

Venous: of or relating to the blood circulating in the veins.

Appendix 2. CENTRAL search strategy (September 2019)

1. exp Pancreatitis/

2. pancreatitis.mp.

3. or/1‐2

4. exp Renal Dialysis/

5. exp Hemofiltration/

6. exp ultrafiltration/

7. exp hemoperfusion/

8. (dialysis or dialyses or hemodialysis or hemodialyses or haemodialysis or haemodialyses).tw,kw.

9. (hemodiafiltration* or haemodiafiltration* or biofiltration* or hemofiltration* or haemofiltration*).tw,kw.

10. (ultrafiltration* or ultra filtration* or haemoperfusion or hemoperfusion* or hemosorption* or haemosorption*).tw,kw.

11. (purificat* adj5 therap*).tw,kw.

12. (blood adj5 purificat*).tw,kw.

13. CVVH.tw,kw.

14. or/4‐13

15. 3 and 14

16. remove duplicates from 15

Appendix 3. MEDLINE search strategy (1946 to September 2019)

1. exp Pancreatitis/

2. pancreatitis.mp.

3. or/1‐2

4. exp Renal Dialysis/

5. exp Hemofiltration/

6. exp ultrafiltration/

7. exp hemoperfusion/

8. (dialysis or dialyses or hemodialysis or hemodialyses or haemodialysis or haemodialyses).tw,kw.

9. (hemodiafiltration* or haemodiafiltration* or biofiltration* or hemofiltration* or haemofiltration*).tw,kw.

10. (ultrafiltration* or ultra filtration* or haemoperfusion or hemoperfusion* or hemosorption* or haemosorption*).tw,kw.

11. (purificat* adj5 therap*).tw,kw.

12. (blood adj5 purificat*).tw,kw.

13. CVVH.tw,kw.

14. or/4‐13

15. 3 and 14

16. randomized controlled trial.pt.

17. controlled clinical trial.pt.

18. random*.mp.

19. placebo.ab.

20. drug therapy.fs.

21. trial.ab.

22. groups.ab.

23. or/16‐22

24. exp animals/ not humans.sh.

25. 23 not 24

26. 15 and 25

Appendix 4. Embase search strategy (1974 to September 2019)

1. exp pancreatitis/

2. pancreatitis.mp.

3. or/1‐2

4. exp renal replacement therapy/

5. exp ultrafiltration/

6. exp hemoperfusion/

7. (dialysis or dialyses or hemodialysis or hemodialyses or haemodialysis or haemodialyses).tw,kw.

8. (hemodiafiltration* or haemodiafiltration* or biofiltration* or hemofiltration* or haemofiltration*).tw,kw.

9. (ultrafiltration* or ultra filtration* or haemoperfusion or hemoperfusion* or hemosorption* or haemosorption*).tw,kw.

10. (purificat* adj5 therap*).tw,kw.

11. (blood adj5 purificat*).tw,kw.

12. CVVH.tw,kw.

13. or/4‐12

14. 3 and 13

15. Randomized controlled trial/

16. Random$.ti,ab.

17. randomization/

18. intermethod comparison/

19. placebo.ti,ab.

20. (compare or compared or comparison).ti.

21. ((evaluated or evaluate or evaluating or assessed or assess) and (compare or compared or comparing or comparison)).ab.

22. (open adj label).ti,ab.

23. ((double or single or doubly or singly) adj (blind or blinded or blindly)).ti,ab.

24. double blind procedure/

25. parallel group$1.ti,ab.

26. (crossover or cross over).ti,ab.

27. ((assign$ or match or matched or allocation) adj5 (alternate or group$1 or intervention$1 or patient$1 or subject$1 or participant$1)).ti,ab.

28. (assigned or allocated).ti,ab.

29. (controlled adj7 (study or design or trial)).ti,ab.

30. (volunteer or volunteers).ti,ab.

31. human experiment/

32. trial.ti.

33. or/15‐32

34. (random$ adj sampl$ adj7 ("cross section$" or questionnaire$1 or survey$ or database$1)).ti,ab. not (comparative study/ or controlled study/ or randomi?ed controlled.ti,ab. or randomly assigned.ti,ab.)

35. Cross‐sectional study/ not (randomized controlled trial/ or controlled clinical study/ or controlled study/ or randomi?ed controlled.ti,ab. or control group$1.ti,ab.)

36. (((case adj control$) and random$) not randomi?ed controlled).ti,ab.

37. (nonrandom$ not random$).ti,ab.

38. "Random field$".ti,ab.

39. (random cluster adj3 sampl$).ti,ab.

40. (rat or rats or mouse or mice or swine or porcine or murine or sheep or lambs or pigs or piglets or rabbit or rabbits or cat or cats or dog or dogs or cattle or bovine or monkey or monkeys or trout or marmoset$1).ti. and animal experiment/

41. Animal experiment/ not (human experiment/ or human/)

42. or/34‐41

43. 33 not 42

44. 14 and 43

Appendix 5. Science Citation Index Expanded search strategy (1982 to September 2019)

1. TS=((acute or necro* or inflam* or interstitial or edema* or oedema*) near/3 pancrea*)

2. TS=(dialysis or dialyses or hemofiltrat* or hemodiafiltrat* or ultrafiltrat* or biofiltrat*)

3. TS=(random* OR rct* OR crossover OR masked OR blind* OR placebo* OR meta‐analysis OR systematic review* OR metaanalys*)

4. #3 AND #2 AND #1

Data and analyses

Comparison 1. Continuous veno‐venous hemofiltration (CVVH) vs no CVVH.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1.1 All‐cause mortality	2	189	Risk Ratio (M‐H, Random, 95% CI)	0.62 [0.27, 1.40]
1.2 Length of ICU stay	1	125	Mean Difference (IV, Random, 95% CI)	‐8.80 [‐10.24, ‐7.36]
1.3 Length of hospital stay	1	125	Mean Difference (IV, Random, 95% CI)	‐26.40 [‐30.17, ‐22.63]
1.4 Total hospital cost (dollars)	1	125	Mean Difference (IV, Random, 95% CI)	‐2800.00 [‐3881.74, ‐1718.26]

Comparison 2. High‐volume continuous veno‐venous hemofiltration (CVVH) versus standard CVVH.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
2.1 All‐cause mortality	1	30	Risk Ratio (M‐H, Random, 95% CI)	0.60 [0.17, 2.07]
2.2 Adverse events	1	30	Risk Ratio (M‐H, Random, 95% CI)	1.00 [0.16, 6.20]

Characteristics of studies

Characteristics of included studies [ordered by study ID]

Chu 2013.

Study characteristics
Methods	Randomized controlled trial Country: China, single‐center Study dates: January 2005 to December 2010 Intention‐to‐treat‐analysis: yes Description of sample size calculation: no
Participants	Number randomized: 30 Postrandomization dropout: 0 (0%) Mean age: 43.1 years Females: 15 (50.0%) Biliary disease: 20 (66.7%) Alcohol: 3 (10.0%) High‐fat diet: 5 (16.7%) Hyperlipidemia: 2 (6.7%) Inclusion criteria: participants with severe acute pancreatitis complicated with multiple organ dysfunction syndrome Exclusion criteria: not mentioned
Interventions	Participants (N = 30) were randomly assigned to 1 of 2 groups Group 1: high‐volume CVVH (N = 15) Group 2: standard CVVH (N = 15)
Outcomes	Mortality, adverse events, laboratory and physiological variables
Notes	Funding source: not reported Declarations of interest: not reported
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Comment: no information provided
Allocation concealment (selection bias)	Unclear risk	Comment: no information provided
Blinding of participants and personnel (performance bias) All outcomes	Unclear risk	Comment: no information provided
Blinding of outcome assessment (detection bias) All outcomes	Unclear risk	Comment: no information provided
Incomplete outcome data (attrition bias) All outcomes	Low risk	Comment: there were no postrandomization dropouts
Selective reporting (reporting bias)	Low risk	Comment: the study protocol was not available, the study included results for key outcomes that would be expected to have been reported for such a study
Other bias	Low risk	Comment: the study appeared to be free of other sources of bias

Guo 2016.

Study characteristics
Methods	Randomized controlled trial Country: China, single‐center Study dates: May 2008 to May 2015 Intention‐to‐treat‐analysis: yes Description of sample size calculation: no
Participants	Number randomized: 64 Postrandomization dropout: 0 (0%) Mean age: 51.9 years Females: 20 (31.3%) Biliary disease: not mentioned Alcohol: not mentioned High‐fat diet: not mentioned Hyperlipidemia: not mentioned Inclusion criteria: diagnosis of severe acute pancreatitis according to the diagnostic criteria developed by the Chinese Medical Association; diagnosis of acute lung injury according to the diagnostic criteria developed by the Chinese Medical Society of Respiratory Diseases; time from the onset of admission ≤ 72 hours; age ranging between 18 and 70 years old. Exclusion criteria: treatment time ≤ 24 hours; administration of mechanical ventilation at admission; pre‐existing chronic lung disease or a history of left ventricular dysfunction; multiple organ dysfunction syndrome or kidney failure or other severe condition.
Interventions	Participants (N = 64) were randomly assigned to 1 of 2 groups Group 1: CVVH (N = 32) Group 2: no CVVH (N = 32)
Outcomes	Mortality, laboratory and physiological variables
Notes	Funding source: not reported Declarations of interest: none declared
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Quote: "These patients were randomly divided into study and control groups" Comment: no information provided on the method of sequence generation
Allocation concealment (selection bias)	Unclear risk	Comment: no information provided
Blinding of participants and personnel (performance bias) All outcomes	High risk	Comment: no information provided regarding measures to achieve and protect blinding. It is very unlikely that such laborious and complex measures were implemented and the authors subsequently forgot to mention them in their report; therefore, the study was regarded as unblinded
Blinding of outcome assessment (detection bias) All outcomes	Unclear risk	Comment: no information provided
Incomplete outcome data (attrition bias) All outcomes	Low risk	Comment: there were no postrandomization dropouts
Selective reporting (reporting bias)	High risk	Comment: The study protocol was not available, the study report failed to include results for key outcomes that would be expected to have been reported for such a study
Other bias	Low risk	Comment: the study appeared to be free of other sources of bias

Wang 2017.

Study characteristics
Methods	Randomized controlled trial Country: China, single‐center Study dates: January 2004 to September 2015 Intention‐to‐treat‐analysis: yes Description of sample size calculation: no
Participants	Number randomized: 125 (to groups 1 and 2) Postrandomization dropout: 0 (0%) Mean age: 46.1 years Females: 57 (45.6%) Biliary disease: 49 (39.2%) Alcohol: 26 (20.8%) High‐fat diet: not mentioned Hyperlipidemia: 26 (20.8%) Other: 24 (19.2%) Inclusion criteria: systemic inflammatory response syndrome (SIRS); persistent organ failure (defined by the Modified Marshall Scoring System) (> 48 hours); CT scores of pancreas ≥ 6; Acute Physiology and Chronic Health Evaluation II (APACHE II) scores ≥8. Exclusion criteria: evidence or a known history of renal dysfunction (creatinine >1.5 mg/dL); pregnancy, malignancy, or immunodeficiency; preexisting chronic kidney diseases requiring regular hemodialysis.
Interventions	Participants (N = 235) were randomly assigned to 1 of 4 groups Group 1: CVVH (N = 65) Group 2: no CVVH (N = 60) Group 3: laparoscopic peritoneal lavage and drainage (N = 62) This group was not included in the review Group 4: continuous veno‐venous hemofiltration and laparoscopic peritoneal lavage and drainage (N = 58) This group was also not included in the review
Outcomes	Mortality, length of ICU stay, length of hospital stay, total hospital cost, laboratory and physiological variables
Notes	Funding source: National Natural Science Foundation of China and the Science Foundation of Science and Technology Hall of Jiangxi Province Declarations of interest: none declared
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Quote: "According to a random number table, the patients were divided into four groups"
Allocation concealment (selection bias)	Unclear risk	Comment: no information provided
Blinding of participants and personnel (performance bias) All outcomes	High risk	Comment: no information provided regarding measures to achieve and protect blinding. It is very unlikely that such laborious and complex measures were implemented and the authors subsequently forgot to mention them in their report; therefore, the study was regarded as unblinded
Blinding of outcome assessment (detection bias) All outcomes	Unclear risk	Comment: no information provided
Incomplete outcome data (attrition bias) All outcomes	Low risk	Comment: there were no postrandomization dropouts
Selective reporting (reporting bias)	High risk	The study protocol was not available, the study report failed to include results for key outcomes that would be expected to have been reported for such a study
Other bias	Low risk	Comment: the study appeared to be free of other sources of bias

CVVH: continuous veno‐venous hemofiltration
N: number

Characteristics of excluded studies [ordered by study ID]

Study	Reason for exclusion
Abulimiti 2018	Quasi‐randomized study
Aleksandrova 2012	A non‐randomized study
Cui 2014	A non‐randomized study
Gao 2018	Randomized controlled trial about continuous veno‐venous hemofiltration for severe acute pancreatitis; severe acute pancreatitis was not diagnosed with the revised Atlanta criteria
Gong 2010	A non‐randomized study
Guo 2014	A non‐randomized study
He 2013	A non‐randomized study
He 2016	Randomized controlled trial about acute hypertriglyceridemic pancreatitis, not about severe acute pancreatitis
Jiang 2005	Randomized controlled trial about acute pancreatitis, not about severe acute pancreatitis
Liu 2017	Randomized controlled trial without reporting any relevant outcome data for this review
Mao 1999	Randomized controlled trial about continuous veno‐venous hemofiltration for severe acute pancreatitis; severe acute pancreatitis was not diagnosed with the revised Atlanta criteria
Sun 2015	A non‐randomized study
Wang 2015	A non‐randomized study
Xia 2012	Randomized controlled trial about continuous veno‐venous hemofiltration for severe acute pancreatitis; severe acute pancreatitis was not diagnosed with the revised Atlanta criteria
Yang 2004	Randomized controlled trial about continuous veno‐venous hemofiltration for severe acute pancreatitis; severe acute pancreatitis was not diagnosed with the revised Atlanta criteria
Yang 2010	Randomized controlled trial about peritoneal dialysis and continuous veno‐venous hemofiltration for severe acute pancreatitis; severe acute pancreatitis was not diagnosed with the revised Atlanta criteria
Zhang 2010	Randomized controlled trial about continuous veno‐venous hemofiltration for severe acute pancreatitis; severe acute pancreatitis was not diagnosed with the revised Atlanta criteria
Zhang 2014	A non‐randomized study

Characteristics of ongoing studies [ordered by study ID]

Zhao 2012.

Study name	Early high‐volume continuous veno‐venous hemofiltration for patients with severe acute pancreatitis: single‐center, randomized, controlled trial
Methods	Randomized controlled trial
Participants	Country: China Number of enrolment: 56 Inclusion criteria: 18 years to 80 years old; clinical diagnosis of severe acute pancreatitis; written informed consent; time from onset of abdominal pain to admission ≤ 72 hours; SIRS score ≥ 2. Exclusion criteria: confirmed infection; pregnancy; patients needing emergency operation for abdominal compartment syndrome; chronic renal diseases needing blood purification; previous exploratory laparotomy for acute abdomen and diagnosis of pancreatitis during laparotomy; acute flare‐up of chronic pancreatitis; malignancy.
Interventions	Participants are randomly assigned to 1 or 2 groups. Group 1: high‐volume CVVH Group 2: no CVVH
Outcomes	Primary outcome: persistent organ failure and death Secondary outcomes: infectious complications, input fluid volume, output fluid volume, physiological parameters, number of surgical interventions, length of ICU stay, total hospital costs, persistent multiple organ dysfunction syndrome, new‐onset organ failure, and new‐onset multiple organ dysfunction syndrome
Starting date	April 2012
Contact information	Principal investigator: Qingchuan Zhao, the First Affiliated Hospital of Fourth Military Medical University, Xi'an, Shanxi, China, 710032 Tel: 86‐29‐84771503 Email: zhaoqc@fmmu.edu.cn
Notes

ICU: intensive care unit
SIRS: systemic inflammatory response syndrome

Differences between protocol and review

We did not perform any subgroup analysis because of insufficient number of studies included in the meta‐analyses. In addition, we did not perform funnel plots to assess reporting biases because there were fewer than 10 included studies.

Contributions of authors

Conceiving the review: YC.

Designing the review: Junhua Gong.

Co‐ordinating the review: YL.

Designing search strategies: YC.

Study selection: Junhua Gong, XD

Data extraction: ZL, ZZ.

Writing the review: YL, SH.

Providing general advice on the review: ZL, ZZ.

Securing funding for the review: Jianping Gong.

Performing previous work that was the foundation of the current study: YC.

Sources of support

Internal sources

• Chongqing Medical University, China

  This review was supported by National Natural Science Foundation of China (Grant No. 81701950), Medical Research Projects of Chongqing (Grant No. 2018MSXM132), and the Kuanren Talents Program of the second affiliated hospital of Chongqing Medical University (Grant No. KY2019Y002).

• Kunming Medical University, China

External sources

• No sources of support provided

Declarations of interest

YL: none known.

SH: none known.

Junhua Gong: none known.

XD: none known.

ZL: none known.

Jianping Gong: none known.

ZZ: none known.

YC: none known.

Edited (no change to conclusions)