Table 2.

Variables, measures, and methods of analysis

Outcome	Hypothesis	Outcome measure	Covariates	Methods of analysis
Feasibility outcomes
Adherence to treatment	>85% adherence	Proportion receiving first dose of study treatment within 2 hours of furosemide	None	95% confidence interval of a proportion
Completion of treatment	>80% completion	Proportion receiving all six doses of study treatment	None	95% confidence interval of a proportion
Absence of hyperoncotic albumin in control arm	>85% in control group without 25% albumin	Proportion of patients assigned to control receiving no 25% albumin	None	95% confidence interval of a proportion
Randomization rate of eligible patients	>50% of eligible patients recruited	Proportion of eligible patients randomized	None	95% confidence interval of a proportion
Randomization rate of patients by clinical site	Average recruitment of one patient per site per week	Rate of recruitment per week	None	Average rate of recruitment
Clinical outcomes
Primary outcome
Duration of mechanical ventilation	Increased number of ventilator-free days in treatment group	Number of ventilator-free days	Duration of ventilation at time of randomization	t test, linear regression
Secondary outcomes
Duration of mechanical ventilation	Decreased duration of mechanical ventilation in treatment group	Duration of mechanical ventilation (days)	Duration of ventilation at time of randomization	t test, linear regression
Episodes interrupting treatment with furosemide	Fewer episodes of interruption in treatment group	Number of episodes	None	t test
Need for dialysis	Smaller proportion of patients requiring dialysis in treatment group during ICU stay	Proportion of patients requiring dialysis during 30 days	None	Chi-squared or Fisher’s exact statistic
Length of ICU stay	Shorter length of stay in treatment group	Duration of ICU stay	Duration of ICU stay at time of randomization	t test, linear regression
ICU mortality	Decreased ICU mortality in treatment group during ICU stay	All-cause mortality (binary)	APACHE-2 and SOFA scores at randomization	Kaplan-Meyer survival curve (time to death)
30-day mortality	Decreased 30-day mortality in treatment group	All-cause mortality (binary)	APACHE-2 and SOFA scores at randomization	Kaplan-Meyer survival curve (time to death)
Physiologic outcomes
Change in oxygenation	Greater increase PaO2/FiO2 ratio in treatment group at day 3 and day 5	Change in PaO2/FiO2 ratio	None	t test
	Greater decrease in oxygenation index in treatment group at day 3 and day 5	Change in oxygenation index	None	t test
Change in lung compliance	Greater increase in dynamic compliance in treatment group at day 3 and day 5	Change in dynamic compliance during study treatment (ml/cmH2O)	None	t test
Change in fluid balance	Greater net decrease in fluid balance in treatment group at day 3 and day 5	Change in net fluid balance (ml)	Net fluid balance at time of randomization	t test, linear regression
Change in body weight	Greater decrease in body weight in treatment group at day 3 and day 5	Change in body weight (kg)	Baseline weight	t test, linear regression
Urine output	Greater net urine output in treatment group at day 3 and day 5	Urine output during 3 days of study treatment (ml)	Net fluid balance at time of randomization	t test, linear regression
Dose of furosemide	Lower total amount of furosemide used in treatment group at day 3	Dose of furosemide during 3 days of study treatment (mg)	None	t test
Changes in serum albumin	Greater increase in serum albumin level in treatment group at day 3 and day 5	Change in serum albumin (g/l)	None	t test
Changes in colloid osmotic pressure	Greater increase in colloid osmotic pressure in treatment group at day 3 and day 5	Change in colloid osmotic pressure (mmHg)	None	t test
Changes in serum total protein	Greater increase in total protein in treatment group at day 3 and day 5	Change in total protein (g/l)	None	t test
Change in electrolytes	No difference between study groups for major electrolytes (sodium, potassium) at day 3 and day 5	Change in electrolytes between beginning and end of study treatment (mEq/l)	None	t test
Subgroup analyses
ARDS	Improved oxygenation in patients with ARDS		None	Regression methods with appropriate interaction term
Severity of disease	Improved hemodynamic stability in patients with severe disease		None	Regression methods with appropriate interaction term
Time since recovery of hemodynamic stability	Improved hemodynamic stability in patients with recent hemodynamic instability	>48 hours versus <48 hours of hemodynamic stability (as defined by study eligibility criteria)	None	Regression methods with appropriate interaction term
Sensitivity analyses
Per-protocol analysis		All outcomes	None
Adjusting for baseline covariates		All outcomes	As described	Multivariable regression
Adjusted for dose of furosemide	Treatment with albumin results in improvement beyond that explained by dose of furosemide	All outcomes	Dose of furosemide	Multivariable regression

APACHE-2, Acute Physiology and Chronic Health Evaluation-2; ARDS, acute respiratory distress syndrome; FiO_2,, fraction of inspired oxygen; ICU, intensive care unit; PaO₂, partial pressure of oxygen in arterial blood; SOFA, Sepsis-related Organ Failure Assessment.