Effects of simulated sample sizes on mortality estimates-Protocol for a study in 3 randomised ICU trials

BACKGROUND

An increasing number of trials are stopped earlier than originally planned. It has been suggested that trials stopped pre-maturely overestimate the treatment effect. With the outlined observational study, we aim to simulate the results of stopping trials before they reach their planned sample size to assess the effects on mortality estimates.

METHODS AND STATISTICS

Based on 3 international, randomised clinical trials (RCTs) in critical care: Scandinavian Starch for Severe Sepsis and Septic Shock (6S) trial, the Transfusion Requirements in Septic Shock (TRISS) trial and the Stress Ulcer Prophylaxis in the Intensive Care Unit (SUP-ICU) trial, we will estimate relative risks with 95% confidence intervals for the primary outcome 90-day mortality after the inclusion of each individual patient in each RCT. This will be presented graphically with the primary outcome as a function of the number of included patients.

DISCUSSION

The outlined study will provide important knowledge about the effects of stopping critical care trials early. This may have important implications for patients, relatives, clinicians, researchers, guideline committee members and policy makers.

ETHICS AND DISSEMINATION

We will use data from consenting patients enrolled in RCTs approved by the relevant ethical committees; this study requires no further permissions. We will report the results in accordance with the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement and submit the final approved manuscript to a peer-reviewed journal.

Intensive care doctors' preferences for arterial oxygen tension levels in mechanically ventilated patients

BACKGROUND

Oxygen is liberally administered in intensive care units (ICUs). Nevertheless, ICU doctors' preferences for supplementing oxygen are inadequately described. The aim was to identify ICU doctors' preferences for arterial oxygenation levels in mechanically ventilated adult ICU patients.

METHODS

In April to August 2016, an online multiple-choice 17-part-questionnaire was distributed to 1080 ICU doctors in seven Northern European countries. Repeated reminder e-mails were sent. The study ended in October 2016.

RESULTS

The response rate was 63%. When evaluating oxygenation 52% of respondents rated arterial oxygen tension (PaO₂ ) the most important parameter; 24% a combination of PaO₂ and arterial oxygen saturation (SaO₂ ); and 23% preferred SaO₂ . Increasing, decreasing or not changing a default fraction of inspired oxygen of 0.50 showed preferences for a PaO₂ around 8 kPa in patients with chronic obstructive pulmonary disease, a PaO₂ around 10 kPa in patients with healthy lungs, acute respiratory distress syndrome or sepsis, and a PaO₂ around 12 kPa in patients with cardiac or cerebral ischaemia. Eighty per cent would accept a PaO₂ of 8 kPa or lower and 77% would accept a PaO₂ of 12 kPa or higher in a clinical trial of oxygenation targets.

CONCLUSION

Intensive care unit doctors preferred PaO₂ to SaO₂ in monitoring oxygen treatment when peripheral oxygen saturation was not included in the question. The identification of PaO₂ as the preferred target and the thorough clarification of preferences are important when ascertaining optimal oxygenation targets. In particular when designing future clinical trials of higher vs lower oxygenation targets in ICU patients.

Corticosteroids and risk of gastrointestinal bleeding in critically ill adults: Protocol for a systematic review

BACKGROUND

Corticosteroids are frequently prescribed to critically ill patients. However, their use may increase the risk of gastrointestinal (GI) bleeding, which is associated with morbidity and mortality. Accordingly, we aim to assess whether continued administration of corticosteroids for >24 hours increases the rate of GI bleeding in adult critically ill patients compared to placebo or no treatment.

METHODS/DESIGN

We will conduct a systematic review of randomized clinical trials with meta-analysis and trial sequential analysis. The participants will be adult (as defined in the included trials) critically ill patients. The intervention will be any corticosteroid administered systematically for >24 hours and the comparator will be placebo or no treatment. The primary outcome will be rate of clinically important GI bleeding. We will systematically search EMBASE, MEDLINE, Medline In-Process, Cochrane Library, Epistemonikos and trial registries for relevant literature, as well as perform a hand search. We will follow the recommendations by the Cochrane Collaboration and the Preferred Reporting Items for Systematic Review and Meta-Analysis (PRISMA) statement. The risk of systematic errors (bias) and random errors will be assessed and the overall quality of evidence will be evaluated using the Grading of Recommendations Assessment, Development, and Evaluation (GRADE) approach.

DISCUSSION

The risk of GI bleeding in adult critically ill patients treated with corticosteroids is unknown. Hence, there is need for a robust systematic review to assess this risk and provide clinicians with a clearer understanding of the strength and limitations of existing data.

Oxygen supplementation for critically ill patients-A protocol for a systematic review

BACKGROUND

In critically ill patients, hypoxaemia is a common clinical manifestation of inadequate gas exchange in the lungs. Supplemental oxygen is therefore given to all critically ill patients. This can result in hyperoxaemia, and some observational studies have identified harms with hyperoxia. The objective of this systematic review is to critically assess the evidence of randomised clinical trials on the effects of higher versus lower inspiratory oxygen fractions or targets of arterial oxygenation in critically ill adult patients.

METHODS

We will search for randomised clinical trials in major international databases. Two authors will independently screen and select references for inclusion using Covidence, extract data and assess the methodological quality of the included randomised clinical trials using the Cochrane risk of bias tool. Any disagreement will be resolved by consensus. We will analyse the extracted data using Review Manager and Trial Sequential Analysis. To assess the quality of the evidence, we will create a 'Summary of Findings' table containing our primary and secondary outcomes using the GRADE assessment.

DISCUSSION

Supplemental oxygen administration is widely recommended in international guidelines despite lack of robust evidence of its effectiveness. To our knowledge, no systematic review of randomised clinical trials has investigated the effects of oxygen supplementation in critically ill patients. This systematic review will provide reliable evidence to better inform future trialists and decision-makers on clinical practice on supplemental oxygen administration in critically ill patients.

Forced fluid removal in intensive care patients with acute kidney injury: The randomised FFAKI feasibility trial

BACKGROUND

Accumulation of fluids is frequent in intensive care unit (ICU) patients with acute kidney injury and may be associated with increased mortality and decreased renal recovery. We present the results of a pilot trial assessing the feasibility of forced fluid removal in ICU patients with acute kidney injury and fluid accumulation of more than 10% ideal bodyweight.

METHODS

The FFAKI-trial was a pilot trial of forced fluid removal vs standard care in adult ICU patients with moderate to high risk acute kidney injury and 10% fluid accumulation. Fluid removal was done with furosemide and/or continuous renal replacement therapy aiming at net negative fluid balance > 1 mL/kg ideal body weight/hour until cumulative fluid balance calculated from ICU admission reached less than 1000 mL.

RESULTS

After 20 months, we stopped the trial prematurely due to a low inclusion rate with 23 (2%) included patients out of the 1144 screened. Despite the reduced sample size, we observed a marked reduction in cumulative fluid balance 5 days after randomisation (mean difference -5814 mL, 95% CI -2063 to -9565, P = .003) with forced fluid removal compared to standard care. While the trial was underpowered for clinical endpoints, no point estimates suggested harm from forced fluid removal.

CONCLUSIONS

Forced fluid removal aiming at 1 mL/kg ideal body weight/hour may be an effective treatment of fluid accumulation in ICU patients with acute kidney injury. A definitive trial using our inclusion criteria seems less feasible based on our inclusion rate of only 2%.

Stress ulcer prophylaxis in adult intensive care unit patients - a protocol for a systematic review

BACKGROUND

In the intensive care unit (ICU), stress ulcer prophylaxis with proton pump inhibitors or histamine-2-receptor antagonists is standard of care although gastrointestinal bleeding remains uncommon. It remains unknown whether its use is associated with benefits or harms and the quality of evidence supporting the use of stress ulcer prophylaxis has been questioned. Accordingly, the objective of this systematic review was to critically assess the evidence from randomized clinical trials on the benefits and harms of stress ulcer prophylaxis vs. placebo or no prophylaxis in adult ICU patients.

METHODS

We will systematically search for randomized clinical trials in major international databases. Two authors will independently screen and select trials for inclusion, extract data and assess the methodological quality using the Cochrane risk of bias tool. Any disagreement will be resolved by consensus. We will perform conventional meta-analyses using Review Manager, and STATA 15, and we will assess the risk of random errors using Trial Sequential Analysis. Also, we will assess and report the overall quality of evidence for all outcomes according to GRADE.

DISCUSSION

The evidence on the benefits and harms of stress ulcer prophylaxis in adult ICU patients is unclear and an updated systematic review is warranted as new trials have been published. To control risks of systematic and random errors, we will use Cochrane and GRADE methodology and Trial Sequential Analysis. Our ambition with this systematic review is to provide updated, reliable and precise data to better inform decision makers on the use of stress ulcer prophylaxis in adult ICU patients.

Fluid accumulation during acute kidney injury in the intensive care unit

INTRODUCTION

Fluid therapy is a ubiquitous intervention in patients admitted to the intensive care unit, but positive fluid balance may be associated with poor outcomes and particular in patients with acute kidney injury. Studies describing this have defined fluid overload either at specific time points or considered patients with a positive mean daily fluid balance as fluid overloaded. We wished to detail this further and performed joint model analyses of the association between daily fluid balance and outcome represented by mortality and renal recovery in patients admitted with acute kidney injury.

METHOD

We did a retrospective cohort study of patients admitted to the intensive care unit with acute kidney injury during a 2-year observation period. We used serum creatinine measurements to identify patients with acute kidney injury and collected sequential daily fluid balance during the first 5 days of admission to the intensive care unit. We used joint modelling techniques to correlate the development of fluid overload with survival and renal recovery adjusted for age, gender and disease severity.

RESULTS

The cohort contained 863 patients with acute kidney injury of whom 460 (53%) and 254 (29%) developed 5% and 10% fluid overload, respectively. We found that both 5% and 10% fluid overload was correlated with reduced survival and renal recovery.

CONCLUSION

Joint model analyses of fluid accumulation in patients admitted to the intensive care unit with acute kidney injury confirm that even a modest degree of fluid overload (5%) may be negatively associated with both survival and renal recovery.

Timing of onset of gastrointestinal bleeding in the ICU: Protocol for a preplanned observational study

BACKGROUND

Critically ill patients are at risk of gastrointestinal bleeding, but clinically important gastrointestinal bleeding is rare. The majority of intensive care unit (ICU) patients receive stress ulcer prophylaxis (SUP), despite uncertainty concerning the balance between benefit and harm. For approximately half of ICU patients with gastrointestinal bleeding, onset is early, ie within the first two days of the ICU stay. The aetiology of gastrointestinal bleeding and consequently the balance between benefit and harm of SUP may differ between patients with early vs late gastrointestinal bleeding.

METHODS

This is a protocol and statistical analysis plan for a preplanned exploratory substudy of the Stress Ulcer Prophylaxis in the Intensive Care Unit (SUP-ICU) randomized clinical trial, comparing intravenous pantoprazole (40 mg once daily) with placebo in 3350 acutely ill adult ICU patients. We will describe baseline characteristics and assess the time to onset of the first clinically important episode of GI bleeding accounting for survival status and allocation to SUP or placebo. In addition, we will describe differences in therapeutic and diagnostic procedures used in patients with clinically important gastrointestinal bleeding according to early vs late bleeding and 90-day vital status.

CONCLUSIONS

The study outlined in this protocol will provide detailed information on patient characteristics and the timing of onset of gastrointestinal bleeding in the patients enrolled in the SUP-ICU trial. This may provide additional knowledge and incentives for future studies on which patients benefit from SUP.

Haloperidol for delirium in critically ill patients - protocol for a systematic review

BACKGROUND

In the intensive care unit, the prevalence of delirium is high. Delirium has been associated with morbidity and mortality including more ventilator days, longer intensive care unit stay, increased long-term mortality, and cognitive impairment. Thus, the burden of delirium for patients, relatives, and societies is considerable. The objective of this systematic review was to critically access the evidence of randomised clinical trials on the effects of haloperidol vs. placebo or any other agents for delirium in critically ill patients.

METHODS

We will search for randomised clinical trials in the following databases: Cochrane Library, MEDLINE, EMBASE, Science Citation Index, BIOSIS, Cumulative Index to Nursing and Allied Health Literature, Latin American and Caribbean Health Sciences Literature, and Allied and Complementary Medicine Database. Two authors will independently screen and select references for inclusion using Covidence, extract data and assess the methodological quality of the included randomised clinical trials using the Cochrane risk of bias tool. Any disagreement will be resolved by consensus. We will analyse the extracted data using Review Manager, STATA 15, and Trial Sequential.

ANALYSIS

The aim of this study was to assess the quality of the evidence, we will create a 'Summary of Findings' table containing our primary and secondary outcomes using the GRADE assessment.

DISCUSSION

Our ambition with this systematic review is to provide reliable and powered evidence to better inform decision makers on the use of or future trials with haloperidol for the management of delirium in critically ill patients.

Empirical metronidazole for patients with severe bacterial infection: protocol for a systematic review

INTRODUCTION

Anaerobic bacteria are believed to be common pathogens in severe infections. Yet, they are difficult to culture and consequently often unrecognised in clinical infections. Metronidazole is often used empirically for potential anaerobic infections, as the resistance to metronidazole is low. However, disadvantages of metronidazole use exist, including drug interactions, side effects and economical expenses. Currently, the balance between the benefits and harms of empirical metronidazole for severe bacterial infections is unknown. We aim to assess patient-important benefits and harms of empirical metronidazole vs. placebo or no treatment in adult patients with severe bacterial infection of any origin in a systematic review of randomised clinical trials with meta-analysis and trial sequential analysis.

METHODS AND ANALYSIS

This protocol provides details on the planned systematic review, which will be prepared according to the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) statement, the Cochrane Handbook, and the Grading of Recommendations, Assessment, Development and Evaluation (GRADE) criteria. The primary outcome is all-cause mortality. Secondary outcomes include adverse events, secondary infections, use of life support, antibiotic resistance and hospital length of stay. We will conduct conventional meta-analyses, including predefined subgroup- and sensitivity analyses. Additionally, we will assess the risk of random errors by trial sequential analysis.

ETHICS AND DISSEMINATION

Ethical approval is not needed, as the outlined review exclusively will include previously published data. We aim to publish in an international, peer-reviewed journal.

Changes in colloid solution sales in Nordic countries

BACKGROUND

Administration of resuscitation fluid is a common intervention in the treatment of critically ill patients, but the right choice of fluid is still a matter of debate. Changes in medical practice are based on new evidence and guidelines as well as traditions and personal preferences. Official warnings against the use of hydroxyl-ethyl-starch (HES) solutions have been issued. Nordic guidelines have issued several strong recommendations favouring crystalloids over colloids in all patient groups. Our objective was to describe the patterns of colloid use in Nordic countries from 2012 to 2016.

METHODS

The data were obtained from companies that provide pharmaceutical statistics in different countries. The data are sales figures from pharmaceutical companies to pharmacies and health institutions.

RESULTS

We found a 56% reduction in the total sales of all colloids in Nordic countries over a 5-year period. These findings were mainly related to a 92% reduction in the sales of HES solutions. A reduction in sales of other synthetic colloids has also occurred. During the same period, we found a 46% increase in albumin sales, but these numbers varied between Nordic countries.

CONCLUSION

The general reduction in colloid sales likely reflects the recommendation that colloids should be used only in special circumstances. The dramatic reduction in the sales of HES solutions was expected given evidence of harm and the official warnings. The steady increase in albumin sales and the notable differences between the five Nordic countries cannot be explained.

Scandinavian SSAI clinical practice guideline on choice of inotropic agent for patients with acute circulatory failure

Background

Adult critically ill patients often suffer from acute circulatory failure and those with low cardiac output may be treated with inotropic agents. The aim of this Scandinavian Society of Anaesthesiology and Intensive Care Medicine guideline was to present patient‐important treatment recommendations on this topic.

Methods

This guideline was developed according to GRADE. We assessed the following subpopulations of patients with shock: (1) shock in general, (2) septic shock, (3) cardiogenic shock, (4) hypovolemic shock, (5) shock after cardiac surgery, and (6) other types of shock, including vasodilatory shock. We assessed patient‐important outcome measures, including mortality and serious adverse reactions.

Results

For all patients, we suggest against the routine use of any inotropic agent, including dobutamine, as compared to placebo/no treatment (very low quality of evidence). For patients with shock in general, and in those with septic and other types of shock, we suggest using dobutamine rather than levosimendan or epinephrine (very low quality of evidence). For patients with cardiogenic shock and in those with shock after cardiac surgery, we suggest using dobutamine rather than milrinone (very low quality of evidence). For the other clinical questions, we refrained from giving any recommendations or suggestions.

Conclusions

We suggest against the routine use of any inotropic agent in adult patients with shock. If used, we suggest using dobutamine rather than other inotropic agents for the majority of patients, however, the quality of evidence was very low, implying high uncertainty on the balance between the benefits and harms of inotropic agents.

Factors associated with non-response at health-related quality of life follow-up in a septic shock trial

BACKGROUND

Follow-up of intensive care unit (ICU) patients often includes health-related quality of life (HRQoL) surveying, but non-responders hamper the interpretation. Our aim was to assess factors for non-response to HRQoL survey in ICU patients with septic shock at follow-up in a clinical trial.

METHODS

In a post hoc follow-up registry study, we assessed all the Danish survivors in the Transfusion-Requirements in Septic Shock trial patients, who were mailed the Short Form 36-item Survey (SF-36) 1-year after randomization. We used covariates from the trial database merged with covariates from nation-wide registries using the unique national identification number to explore possible factors for not responding. Five covariates were pre-specified to be included in the primary multivariate analysis: age, number of days in hospital from randomization to follow-up, level of education, cohabitation and employment status at follow-up. We compared the mortality from 1-year survival (2012-2014) till end of final follow-up (January 2016) between non-responders and responders.

RESULTS

We assessed 308 survivors of whom 108 (35%) were non-responders. In the primary analysis lower age (odds ratio 1.03, 95% CI [1.01-1.05]), more admission days in hospital (1.006 [1.001-1.011]) and living alone (4.33 [2.46-7.63]) were associated with non-responding, whereas the level of education and employment status were not. Non-responders had a hazard ratio of 1.63 [0.97-2.72] for mortality from 1-year follow-up to final follow-up as compared to the responders.

CONCLUSION

Being younger, spending more days in hospital and living alone were all associated with non-response at 1-year HRQoL follow-up among ICU patients with septic shock.

Development and internal validation of the Simplified Mortality Score for the Intensive Care Unit (SMS-ICU)

BACKGROUND

Intensive care unit (ICU) mortality prediction scores deteriorate over time, and their complexity decreases clinical applicability and commonly causes problems with missing data. We aimed to develop and internally validate a new and simple score that predicts 90-day mortality in adults upon acute admission to the ICU: the Simplified Mortality Score for the Intensive Care Unit (SMS-ICU).

METHODS

We used data from an international cohort of 2139 patients acutely admitted to the ICU and 1947 ICU patients with severe sepsis/septic shock from 2009 to 2016. We performed multiple imputations for missing data and used binary logistic regression analysis with variable selection by backward elimination, followed by conversion to a simple point-based score. We assessed the apparent performance and validated the score internally using bootstrapping to present optimism-corrected performance estimates.

RESULTS

The SMS-ICU comprises seven variables available in 99.5% of the patients: two numeric variables: age and lowest systolic blood pressure, and five dichotomous variables: haematologic malignancy/metastatic cancer, acute surgical admission and use of vasopressors/inotropes, respiratory support and renal replacement therapy. Discrimination (area under the receiver operating characteristic curve) was 0.72 (95% CI: 0.71-0.74), overall performance (Nagelkerke's R² ) was 0.19 and calibration (intercept and slope) was 0.00 and 0.99, respectively. Optimism-corrected performance was similar to apparent performance.

CONCLUSIONS

The SMS-ICU predicted 90-day mortality with reasonable and stable performance. If performance remains adequate after external validation, the SMS-ICU could prove a valuable tool for ICU clinicians and researchers because of its simplicity and expected very low number of missing values.

Prediction of bleeding and thrombosis by standard biochemical coagulation variables in haematological intensive care patients

PURPOSE

We assessed the value of standard biochemical coagulation parameters in predicting bleeding, thrombosis and mortality in adult Intensive Care Unit (ICU) patients with haematological malignancies.

METHODS

We screened all patients with acute leukaemia and myelodysplastic syndrome admitted to a university hospital ICU during 2008-2012. Data were obtained from the clinical chemistry laboratory database and patient files. We graded bleeding according to the World Health Organisation (WHO)-system within 24-h, within 5-days and during the whole ICU stay. We analysed the predictive values of laboratory parameters using multiple logistic regression and receiver operator characteristics (ROC) curves. As we previously have established that platelet count at admission was associated with bleeding, we focused on International Normalised Ratio (INR), activated pro-thrombin time (APTT), anti-thrombin, D-dimer and fibrinogen, and markers of infection (C-reactive protein, pro-calcitonin), kidney function (creatinine) and tissue damage (lactate dehydrogenase (LDH)).

RESULTS

We included 116 patients; 66 (57%) had at least one bleeding episode and 11 (9%) patients had at least one thrombotic event. The differences in coagulation values when bleeding compared to baseline values were minor. INR was the only variable we found associated with subsequent bleeding within 24 h from admission to ICU (odds ratio 2.91, 95% CI: 1.01-8.43, P = 0.048). ROC analyses did not show predictive value of any of the other variables with regards to bleeding and none of the variables were associated with thrombosis in adjusted analyses. Increased levels of LDH at admission were associated with increased 7-day and 30-day mortality.

CONCLUSIONS

Increased INR at admission was associated with a higher rate of bleeding in ICU patients with haematological malignancies. No other biochemical coagulation or other parameter had any association with bleeding, thrombosis or mortality except increased LDH, which at ICU admission was associated with increased 30-day mortality.

Necrotizing soft tissue infections - a multicentre, prospective observational study (INFECT): protocol and statistical analysis plan

BACKGROUND

The INFECT project aims to advance our understanding of the pathophysiological mechanisms in necrotizing soft tissue infections (NSTIs). The INFECT observational study is part of the INFECT project with the aim of studying the clinical profile of patients with NSTIs and correlating these to patient-important outcomes. With this protocol and statistical analysis plan we describe the methods used to obtain data and the details of the planned analyses.

METHODS

The INFECT study is a multicentre, prospective observational cohort study. Patients with NSTIs are enrolled in five Scandinavian hospitals, which are all referral centres for NSTIs. The primary outcomes are the descriptive variables of the patients. Secondary outcomes include identification of factors associated with 90-day mortality and amputation; associations between affected body part, maximum skin defect and Laboratory Risk Indicator for Necrotizing Fasciitis (LRINEC) score and 90-day mortality; 90-day mortality in patients with and without acute kidney injury (AKI) and LRINEC score of six and above or below six; and association between affected body part at arrival and microbiological findings. Exploratory outcomes include univariate analyses of baseline characteristics associations with 90-day mortality. The statistical analyses will be conducted in accordance with the predefined statistical analysis plan.

CONCLUSION

Necrotizing soft tissue infections result in severe morbidity and mortality. The INFECT study will be the largest prospective study in patients with NSTIs to date and will provide important data for clinicians, researchers and policy makers on the characteristics and outcomes of these patients.

Untargeted antifungal therapy in adult patients with complicated intra-abdominal infection: a systematic review

BACKGROUND

Systematic reviews comparing untargeted antifungal treatment with placebo or no treatment in critically ill patients have provided conflicting results. We aimed to assess patient-important benefits and harms of untargeted antifungal therapy vs. placebo or no treatment in adult patients with complicated intra-abdominal infection.

METHODS

We conducted a systematic review with meta-analysis and trial sequential analysis of randomised clinical trials assessing untargeted antifungal therapy compared to placebo or no treatment in adults with complicated intra-abdominal infection. We used the Cochrane and GRADE methodologies and exclusively assessed patient-important outcomes. Two independent authors screened trials for eligibility, extracted data and assessed risk of bias. We performed conventional meta-analyses, including sensitivity and subgroup analyses, and trial sequential analysis to assess the risk of random errors and to estimate trial sequential analysis adjusted confidence intervals.

RESULTS

We included six trials (1,067 patients) in the review, and four trials reported data on the predefined outcome measures and were included in the meta-analysis. Three of the four trials had high risk of bias. We observed no statistically significant difference in mortality (relative risk 0.58, 95% confidence interval 0.24-1.39) or in any of the other patient-important outcomes between untargeted antifungal treatment and placebo or no treatment (low/very low quality of evidence). Trial sequential analysis demonstrated lack of data and high risk of random errors.

CONCLUSIONS

The quantity and quality of evidence supporting untargeted antifungal treatment in adult patients with complicated intra-abdominal infection are low to very low with no firm evidence for benefit or harm.

Effects of red blood cell storage time on transfused patients in the ICU-protocol for a systematic review

BACKGROUND

Patients in the intensive care unit (ICU) are often anaemic due to blood loss, impaired red blood cell (RBC) production and increased RBC destruction. In some studies, more than half of the patients were treated with RBC transfusion. During storage, the RBC and the storage medium undergo changes, which lead to impaired transportation and delivery of oxygen and may also promote an inflammatory response. Divergent results on the clinical consequences of storage have been reported in both observational studies and randomised trials. Therefore, we aim to gather and review the present evidence to assess the effects of shorter vs. longer storage time of transfused RBCs for ICU patients.

METHODS

We will conduct a systematic review with meta-analyses and trial sequential analyses of randomised clinical trials, and also include results of severe adverse events from large observational studies. Participants will be adult patients admitted to an ICU and treated with shorter vs. longer stored RBC units. We will systematically search the Cochrane Library, MEDLINE, Embase, BIOSIS, CINAHL and Science Citation Index for relevant literature, and we will follow the recommendation by the Cochrane Collaboration and the Preferred Reporting Items for Systemtic Review and Meta-Analysis (PRISMA)-statement. We will assess the risk of bias and random errors, and we will use the Grading of Recommendations, Assessment, Development and Evaluation (GRADE) approach to evaluate the overall quality of evidence.

CONCLUSION

We need a high-quality systematic review to summarise the clinical consequences of RBC storage time among ICU patients.

Effects of hydroxyethyl starch 130/0.42 vs. Ringer's acetate on cytokine levels in severe sepsis

BACKGROUND

The Scandinavian Starch for Severe Sepsis/Septic Shock (6S) trial showed increased 90-day mortality with hydroxyethyl starch (HES) 130/0.42 vs. Ringer's acetate. To explore the underlying pathophysiology, we compared early changes in plasma cytokine concentrations between patients resuscitated with HES vs. Ringer's acetate.

METHODS

In a subgroup of 226 patients from the 6S trial, we calculated delta plasma concentrations of tumour necrosis factor alpha (TNF-α), interleukin (IL)-6 and IL-10 from randomization to day 2. We used multiple linear and logistic regression analyses to assess differences between the groups and associations between delta cytokine concentrations and 90-day mortality, respectively.

RESULTS

Baseline characteristics and day 2 mortality were comparable between the groups. We observed similar delta cytokine concentrations in the HES vs. Ringer's group (mean difference in delta TNF-α: -1.5 pg/ml (95% CI, -4.9 to 1.9), P = 0.39; IL-6: 36.0 pg/ml (-24.1 to 96.1), P = 0.24; IL-10: -3.9 pg/ml (-21.1 to 28.9), P = 0.76). In all included patients, we observed a linear relationship between increases in TNF-α and 90-day mortality (P = 0.005).

CONCLUSION

Resuscitation with HES 130/0.42 vs. Ringer's acetate did not appear to affect plasma concentrations of TNF-α, IL-6 or IL-10 differently during the first days after randomization into the 6S trial. In the overall cohort, increases in TNF-α were associated with increased 90-day mortality. Although interpretation should be done with caution, it seems unlikely that the increased mortality observed with the use HES in the 6S trial is signalled by early changes in three biomarkers of systemic inflammation.

Lower vs. higher fluid volumes in sepsis-protocol for a systematic review with meta-analysis

BACKGROUND

Intravenous fluid administration with crystalloids is recommended in the initial management of sepsis. However, the quality of evidence supporting the recommendation on fluid volumes is low, and clinical equipoise exists. Potential benefits of restricting fluid volumes has been suggested, but the overall benefit or harm in patients with sepsis is unknown. Accordingly, we aim to assess patient-important benefits and harms of lower vs. higher fluid volumes in resuscitation of adult patients with sepsis.

METHODS/DESIGN

We will conduct a systematic review with meta-analysis and trial sequential analysis of randomised clinical trials comparing different strategies to obtain separation in fluid volumes or balances during resuscitation of adult patients with sepsis. We will systematically search the Cochrane Library, MEDLINE, EMBASE, Science Citation Index, BIOSIS and Epistemonikos for relevant literature. We will follow the recommendations by the Cochrane Collaboration and the Preferred Reporting Items for Systematic Review and Meta-Analysis (PRISMA) statement. The risk of systematic errors (bias) and random errors will be assessed, and the overall quality of evidence will be evaluated using the Grading of Recommendations Assessment, Development, and Evaluation (GRADE) approach.

DISCUSSION

The outlined systematic review will provide important data on how patient-important outcomes are affected by higher vs. lower resuscitation fluid volumes in adults with sepsis. Using trial sequential analysis to assess the risk of random errors will increase the validity of the summary estimates calculated and help estimate the required information size for future trials.

Stress ulcer prophylaxis in the intensive care unit trial: detailed statistical analysis plan

BACKGROUND

In this statistical analysis plan, we aim to provide details of the pre-defined statistical analyses of the Stress Ulcer Prophylaxis in the Intensive Care Unit (SUP-ICU) trial. The aim of the SUP-ICU trial is to assess benefits and harms of stress ulcer prophylaxis with a proton pump inhibitor in adult patients in the intensive care unit (ICU).

METHODS

The SUP-ICU trial is an investigator-initiated, international, multicentre, randomised, blinded, parallel-group trial of intravenously pantoprazole 40 mg once daily vs. placebo in 3350 acutely ill adult ICU patients at risk of gastrointestinal bleeding. The primary outcome measure is 90-day mortality. Secondary outcomes include the proportion of patients with clinically important gastrointestinal bleeding, pneumonia, Clostridium difficile infection or myocardial ischaemia, days alive without life support, serious adverse reactions, 1-year mortality, and a health economic analysis. Two formal interim analyses will be performed. The statistical analyses will be conducted according to the outlined pre-defined statistical analysis plan. The primary analysis will be a logistic regression analysis adjusted for stratification variables comparing the two intervention groups in the intention-to-treat population. In a secondary analysis, we will additionally adjust the primary outcome for potential random differences in baseline characteristics. The conclusion will be based on the intention-to-treat population.

CONCLUSION

Stress ulcer prophylaxis is standard of care in ICUs worldwide, but has never been tested in large high-quality randomised placebo-controlled trials. The SUP-ICU trial will provide important high-quality data on the balance between the benefits and harms of stress ulcer prophylaxis in adult critically ill patients.

Use of life support in acutely admitted ICU patients. An international cohort study

BACKGROUND

Use of life support in intensive care unit (ICU) patients has been associated with increased risk of poor outcome. The prognostic importance of the duration of support is less studied. We assessed the use of life support and the association between its duration and 90-day mortality in a contemporary cohort of acutely admitted adult ICU patients.

METHODS

We performed a post-hoc analysis of the SUP-ICU 7-day inception cohort study (n = 1034), which was conducted in 97 ICUs in 11 countries. We included patients with an ICU stay of 3 days or more. We assessed the use of life support during the first 3 days in ICU and the crude and adjusted association between its duration and 90-day mortality using logistic regression analyses.

RESULTS

We included 690 patients; their 90-day mortality was 23%. During the first 3 days in ICU mechanical ventilation was used in 65%, vasopressors/inotropes in 57% and renal replacement therapy in 13%. Renal replacement therapy for 3 days or more was associated with a higher 90-day mortality as compared with 1 day of renal replacement therapy [odds ratio 6.5 (95% confidence interval 1.3 to 32.8)]. For mechanical ventilation and vasopressors/inotropes the odds ratios were 2.2 [0.9 to 5.3] and 1.2 [0.5 to 2.6], respectively.

CONCLUSIONS

Among acutely admitted adult ICU patients, a higher number of days of renal replacement therapy in the initial ICU stay were associated with increased risk of death within 90 days. We did not observe such an association for mechanical ventilation or vasopressor/inotropic therapy.

Changes from 2012 to 2015 in intravenous fluid solutions issued to hospital departments

BACKGROUND

In recent years, large trials have increased the level of evidence for intravenous (IV) fluid therapy, at least in the intensive care setting. It is less clear whether this change in the evidence base has been associated with changes in IV fluid use in different hospital departments.

METHODS

We obtained details from the regional pharmacy regarding IV fluids issued to hospital departments in the Danish Capitol Region from January 2012 to May 2015. We used paired Wilcoxon's signed-rank test to analyse changes in the issuing in different departments.

RESULTS

Total regional issuing of IV fluids showed increase in crystalloid solutions (9%; P = 0.001) and decrease in colloid solutions (59%; P = 0.005). Subtype analysis showed increased issuing of buffered crystalloids (36%; P = 0.001), human albumin (30%; P < 0.0001) and decreased issuing in synthetic colloid solutions (82%; P < 0.0001) from Q1 2012 to Q2 2015. At the departmental level, the issuing of synthetic colloid solutions decreased markedly to all settings. The issuing of buffered crystalloids increased to orthopaedic (226%; P = 0.03) and to general surgery departments (686%; P = 0.002). Albumin solutions were increasingly issued to anaesthesia departments (63%; P = 0.005) and was rarely issued to general surgery and orthopaedic departments.

CONCLUSIONS

The issuing of IV fluid solutions to hospital departments has changed markedly over the last years to less colloid, in particular the synthetic solutions, and relatively more issuing of crystalloids, in particular the buffered solutions.

Effects of fluid restriction on measures of circulatory efficacy in adults with septic shock

BACKGROUND

The haemodynamic consequences of fluid resuscitation in septic shock have not been fully elucidated. Therefore, we assessed circulatory effects in the first 24 h of restriction of resuscitation fluid as compared to standard care in intensive care unit (ICU) patients with septic shock.

METHODS

This was a post-hoc analysis of the multicentre CLASSIC randomised trial in which patients with septic shock, who had received the initial fluid resuscitation, were randomised to a protocol restricting resuscitation fluid or a standard care protocol in nine ICUs. The highest plasma lactate, highest dose of noradrenaline, and the urinary output were recorded in five time frames in the first 24 h after randomisation. We used multiple linear mixed effects models to compare the two groups.

RESULTS

We included all 151 randomised patients; the cumulated fluid resuscitation volume in the first 24 h after randomisation was median 500 ml (Interquartile range (IQR) 0-1500) and 1250 ml (500-2500) in the fluid restriction group and standard care group, respectively. The estimated differences in the fluid restriction group vs. the standard care group were 0.1 mM (95% confidence interval -0.7 to 0.9; P = 0.86) for lactate, 0.01 μg/kg/min (-0.02 to 0.05; P = 0.48) for dose of noradrenaline, and -0.1 ml/kg/h (-0.3 to 0.2; P = 0.70) for urinary output during the first 24 h after randomisation.

CONCLUSIONS

We observed no indications of worsening of measures of circulatory efficacy in the first 24 h of restriction of resuscitation fluid as compared with standard care in adults with septic shock who had received initial resuscitation.

Three-year follow-up of implementation of evidence-based transfusion practice in a tertiary hospital

BACKGROUND AND OBJECTIVES

Traditionally, Denmark has had a high rate of allogeneic red blood cell transfusion caused by a liberal transfusion practice despite the existence of restrictive guidelines. We established a Patient Blood Management programme in a tertiary hospital and report the results of the implementation of evidence-based transfusion practice.

MATERIALS AND METHODS

Red blood cell transfusion quality indicators were compared with the evidence-based guideline at hospital and department level. Based on this evaluation, wards were selected for interventions targeting doctors and nurses. The implementation process was monitored by transfusion quality and utilization data over a 3-year period with totally 166 341 admissions in 98 960 mixed, adult medical and surgical patients.

RESULTS

At the hospital level, transfusion above the upper guideline limit decreased from 23 to 10% (P < 0·001), and transfusion at or below the restrictive haemoglobin trigger of 7·3 g/dl increased from 7 to 19% (P < 0·001). The percentage of single-unit transfusions increased from 72 to 78% (P < 0·001), and the majority of transfusion rates and volumes decreased significantly. Red cell use decreased with 41% in surgical procedures and 28% in admissions (P < 0·001).

CONCLUSION

The intervention was associated with a significant and sustained overall increase in compliance with national guidelines for red blood cell transfusion for non-bleeding patients, and led to significantly fewer patients being exposed to transfusion.

Higher vs. lower haemoglobin threshold for transfusion in septic shock: subgroup analyses of the TRISS trial

BACKGROUND

Using a restrictive transfusion strategy appears to be safe in sepsis, but there may be subgroups of patients who benefit from transfusion at a higher haemoglobin level. We explored if subgroups of patients with septic shock and anaemia had better outcome when transfused at a higher vs. a lower haemoglobin threshold.

METHODS

In post-hoc analyses of the full trial population of 998 patients from the Transfusion Requirements in Septic Shock (TRISS) trial, we investigated the intervention effect on 90-day mortality in patients with severe comorbidity (chronic lung disease, haematological malignancy or metastatic cancer), in patients who had undergone surgery (elective or acute) and in patients with septic shock as defined by the new consensus definition: lactate above 2 mmol/l and the need for vasopressors to maintain a mean arterial pressure above 65 mmHg.

RESULTS

The baseline characteristics were mostly similar between the two intervention groups in the different subgroups. There were no differences in the intervention effect on 90-day mortality in patients with chronic lung disease (test of interaction P = 0.31), haematological malignancy (P = 0.47), metastatic cancer (P = 0.51), in those who had undergone surgery (P = 0.99) or in patients with septic shock by the new definition (P = 0.20).

CONCLUSION

In exploratory analyses of a randomized trial in patients with septic shock and anaemia, we observed no survival benefit in any subgroups of transfusion at a haemoglobin threshold of 90 g/l vs. 70 g/l.

Scandinavian SSAI clinical practice guideline on choice of first-line vasopressor for patients with acute circulatory failure

Background

Adult critically ill patients often suffer from acute circulatory failure, necessitating use of vasopressor therapy. The aim of the Scandinavian Society of Anaesthesiology and Intensive Care Medicine (SSAI) task force for Acute Circulatory Failure was to present clinically relevant, evidence‐based treatment recommendations on this topic.

Methods

This guideline was developed according to standards for trustworthy guidelines, including a systematic review of the literature and use of the GRADE methodology for assessment of the quality of evidence and for moving from evidence to recommendations. We assessed the following subpopulations of patients with acute circulatory failure: 1) shock in general, 2) septic shock, 3) cardiogenic shock, 4) hypovolemic shock and 5) other types of shock, including vasodilatory shock. We assessed patient‐important outcome measures, including mortality, serious adverse reactions and quality‐of‐life.

Results

For patients with shock in general and those with septic shock, we recommend using norepinephrine rather than dopamine, and we suggest using norepinephrine rather than epinephrine, vasopressin analogues, and phenylephrine. For patients with cardiogenic shock and those with hypovolemic shock, we suggest using norepinephrine rather than dopamine, and we provide no recommendations/suggestions of norepinephrine vs. epinephrine, vasopressin analogues, and phenylephrine. For patients with other types of shock, including vasodilatory shock, we suggest using norepinephrine rather than dopamine, epinephrine, vasopressin analogues, and phenylephrine.

Conclusions

We recommend using norepinephrine rather than other vasopressors as first‐line treatment for the majority of adult critically ill patients with acute circulatory failure.

Systematic review of the association of venous oxygenation and outcome in adult hospitalized patients

BACKGROUND

Adequate tissue oxygenation is necessary to maintain organ function. Low venous oxygen saturation may reflect impaired tissue oxygenation, and may be used as a predictive tool and a therapeutic target to improve the care of critically ill patients. We therefore conducted a systematic review of the existing literature reflecting these aspects.

METHODS

We searched electronic databases in January 2016 for relevant studies on venous oxygen saturation for treatment guidance and patient outcome. We sub-grouped results based on patient groups and setting.

RESULTS

The search resulted in 5590 papers of which 42 studies were deemed relevant. The majority of the studies in cardiac and abdominal surgery patients showed associations between low venous oxygen saturation and increased mortality and morbidity, in particular increased length of intensive care. However, the cut-off level for low venous oxygen saturation varied between < 55 and 70% and all studies had high risk of bias. In patients with septic shock, recent randomized trials showed no benefit of early resuscitation guided by venous oxygen saturation.

CONCLUSION

Low venous oxygen saturation may be associated with increased mortality, morbidity and length of intensive care in patients following cardiac or abdominal surgery. However, the wide range of cut-off levels and low quality of evidence hampers the clinical application. In patients with septic shock, the present evidence does not support goal-directed therapy using venous oxygen saturation during early resuscitation.

Long term end-stage renal disease and death following acute renal replacement therapy in the ICU

INTRODUCTION

In ICU the need for acute renal replacement therapy (RRT) associates with high mortality and risk of end-stage renal disease (ESRD), but there are limited long-term data. We investigated these outcomes and their risk factors.

METHODS

Retrospective analysis of all adult patients admitted to a general, university hospital ICU 2005-2012, excluding chronic dialysis patients. ESRD was defined as need of RRT > 90 days or kidney transplant.

RESULTS

Of 5766 patients included, 1004 (16%) received acute RRT; their 30-day mortality was 42% vs. 16% for those not requiring acute RRT (adjusted hazard ratio (HR) 1.13 (0.96-1.32)). The 90-day mortality was 55% for patients receiving acute RRT vs. 22% for those who did not (adjusted HR 1.32 (1.15-1.51)) and 1-year mortality was 63% vs. 30%, respectively, (adjusted HR 1.31 (1.16-1.48)). The 7-year risk of ESRD for ICU patients surviving 90 days was 10% for patients who received acute RRT vs. 0.5% among those who did not (adjusted HR 5.9 (2.9-12.4)). Independent risk factors for ESRD included pre-existing kidney disease, pre-existing peripheral vascular disease and use of acute RRT in ICU.

CONCLUSIONS

The need of acute RRT was associated with markedly increased long term risk of death and ESRD; in contrast its use was not associated with 30-day mortality. In addition to acute RRT, decreased kidney function and peripheral vascular disease before ICU admission were risk factors for ESRD. It seems warranted offering medical follow-up to patients after acute RRT in ICU.

Shock treatment in a cohort of Scandinavian intensive care units in 2014

BACKGROUND

Shock is common in intensive care units, and treatment includes fluids, vasopressor and/or inotropic drugs, guided by hemodynamic monitoring. The aim of this study was to identify current practice for treatment of shock in Scandinavian intensive care units.

METHODS

Seven-day inception cohort study in 43 intensive care units in Scandinavia. Patients ≥ 15 years old receiving more than 4 h of cardiovascular acting drug infusion were included. The use of fluids, vasopressor and inotropic drugs, type of monitoring, and target values were recorded.

RESULTS

One hundred and seventy-one patients were included. At inclusion, 136/168 (81%) had received vasopressor and/or inotropic drug therapy for less than 24 h, and 143/171 (84%) had received volume loading before the onset of vasoactive drug treatment. Ringer's solution was given to 129/143 (90%) of patients and starches in 3/143 (2%) patients. Noradrenaline was the most commonly used cardiovascular acting drug, given in 168/171 (98%) of cases while dopamine was rarely used. Mean arterial pressure was considered the most important variable for hemodynamic monitoring. Invasive arterial blood pressure was monitored in 166/171 (97%) of patients, arterial pulse wave analysis in 11/171 (7%), and echocardiography in 50/171 (29%).

CONCLUSION

In this survey, Ringer's solution and noradrenaline were the most common first-line treatments in shock. The use of starches and dopamine were rare. Almost all patients were monitored with invasive arterial blood pressure, but comprehensive hemodynamic monitoring was used only in a minority of patients.

Association between fluid balance and mortality in patients with septic shock: a post hoc analysis of the TRISS trial

BACKGROUND

Several studies have shown an association between a positive fluid balance and increased mortality in patients with septic shock. This may have led to a more restrictive use of intravenous fluids. The association between fluid accumulation and mortality in the setting of a more restrictive use of intravenous fluids, however, is uncertain. We therefore aimed to investigate the association between a cumulative fluid balance 3 days after randomization and 90-day mortality in a recent Nordic multicentre cohort of patients with septic shock.

METHODS

A post hoc analysis of patients from the Transfusion Requirements in Septic Shock (TRISS) trial treated for 3 days or more in the ICU after randomization. The patients were categorized into four groups depending on their weight-adjusted cumulative fluid balance after 3 days. We performed multivariable Cox regression analysis, adjusting for important prognostics (study site, age, chronic cardiovascular and chronic lung disease, haematologic malignancy, chronic dialysis, source of infection, baseline SOFA score and plasma lactate).

RESULTS

The median cumulative fluid balance of the 841 included patients was 2480 ml (IQR 47-5045). The median time from ICU admission to inclusion in the trial was 22 h. The overall 90-day mortality was 52%. There was no statistically significant association between fluid balance 3 days from inclusion and 90-day mortality after the adjustment for the prognostics (P = 0.37).

CONCLUSION

In our cohort of patients with septic shock and a comparably low cumulative fluid balance, there was no association between fluid balance and mortality. However, the study design and the limited power preclude strong conclusions. There is an urgent need for high-quality trials assessing the benefit and harm of different fluid volume strategies in patients with septic shock.

Variability in targeted arterial oxygenation levels in patients with severe sepsis or septic shock

BACKGROUND

Supplemental oxygen therapy is used for intensive care (ICU) patients with severe sepsis, but with no general guidelines and few safety data. The aim of this observational study was to describe the variability in oxygen administration as well as the association between partial pressure of arterial oxygen (PaO2 ) and mortality.

METHODS

We extracted data from two Scandinavian clinical trials of ICU patients with severe sepsis or septic shock. We calculated average PaO2 and fraction of inspired oxygen (FiO2 ) from trial inclusion and the following 5 days, and assessed the association between PaO2 and 90-day mortality.

RESULTS

The median PaO2 was 9.8 kPa [5-95% range 6.4-19.9] and FiO2 was 0.51 [5-95% range 0.27-1.00], respectively. Eight hundred and five of 1,770 patients (45%) died. The relative risk of mortality was 1.43 [95% CI: 1.19-1.65] in patients with average PaO2 < 8 kPa and 1.29 [95% CI: 0.84-1.68] in patients with average PaO2 ≥ 16 kPa, as compared to patients with average PaO2 10-12 kPa. The relative risk of mortality was 1.38 [95% CI: 1.17-1.58] in patients with an average FiO2 0.60-0.80 and 2.10 [95% CI: 1.88-2.23] in patients with an average FiO2  ≥ 0.80 as compared to patients with an average FiO2  ≤ 0.40.

CONCLUSION

Administration of oxygen in patients with severe sepsis resulted in a wide range of PaO2 . Significantly higher mortality was observed in patients with an average PaO2 < 8 kPa and FiO2 ≥ 0.60. The results do not imply causation and the associations between average PaO2 and adverse outcomes have to be assessed further.

Use of selective digestive tract decontamination in European intensive cares: the ifs and whys

BACKGROUND

Several studies have shown that the use of selective digestive tract decontamination (SDD) reduces mortality. However, fear for increasing multidrug resistance might prevent wide acceptance. A survey was performed among the units registered in the European Registry for Intensive Care (ERIC), in order to investigate the number of ICUs using SDD and the factors that prevented the use of SDD.

METHODS

One invitation to the electronic survey was sent to each ERIC unit. The survey focused on department characteristics (intensive care type, local resistance levels), local treatment modalities (antibiotic stewardship) and doctors' opinions (collaborative issues concerning SDD). All ICU's in countries participating in the European Centre for Disease Prevention and Control resistance surveillance program were analysed.

RESULTS

Seventeen percent of the ICUs registered in the ERIC database used SDD prophylaxis. Most of these ICUs were located in the Netherlands or Germany. ICUs using SDD were four times more likely to use antibiotic stewardship. Also larger ICUs were more likely to use SDD. On the contrary, resistance to antibiotics was not related to the use of SDD. Also the doctor's opinion that SDD is proven in cluster-randomized trials was not a determinant for not using SDD.

CONCLUSION

SDD is used in a minority of the European ICUs registered in the ERIC database. Larger ICUs and ICUs with a prudent antibiotic policy were more likely to use SDD. Neither antibiotic resistance nor the cluster randomized study design were determinants of the non-use of SDD.