Effect of midodrine versus placebo on time to vasopressor discontinuation in patients with persistent hypotension in the intensive care unit (MIDAS): an international randomised clinical trial

Design

Randomized, multicenter, international, double-blind, placebo-controlled trial

N=132

Objective

To determine whether the addition of midodrine to standard care reduces the time to discontinuation of IV vasopressor use in otherwise resuscitated patients

Study Groups

Midodrine (n=66)

Placebo (n=66)

Inclusion Criteria

Adults who were hypotensive (systolic blood pressure <90 mm Hg); required IV vasopressor for >24 hours; otherwise resuscitated; have stable blood pressure on single-agent infusion (noradrenaline <8 mcg/min, or phenylephrine <100 mcg/min, or metaraminol <60 mcg/min) or required multiple vasopressors

Exclusion Criteria

Clinical evidence of inadequate tissue oxygenation; hypovolaemic shock or hypotension due to adrenal insufficiency; liver failure; chronic renal failure (SCr >2 mg/dL); Severe organic heart disease (ejection fraction <30%); acute urinary retention; pregnancy; pheochromocytoma; thyrotoxicosis; midodrine used before admission; allergy to midodrine; bradycardia (heart rate <50 bpm)

Methods

Participants were randomized 1:1 to receive midodrine 20 mg or matching placebo orally q8h while intravenous vasopressor therapy was continued as needed. Both groups received other medications following standard of care guidelines.

The study drugs were continued until ICU discharge or until: worsening hypotension requiring high-dose vasopressors (>100 mcg/min phenylephrine, >8 mcg/min norepinephrine, or >60 mcg/min metaraminol), epinephrine requirement, signs or symptoms of organ failure or hypoperfusion, adverse events related to midodrine, or death.

Duration

October 2012 to June 2019

Outcome Measures

Primary: time to discontinuation of IV vasopressors

Secondary: time from ICU admission to discharge-ready; ICU and hospital length of stay; ICU readmission; adverse events

Baseline Characteristics

Midodrine (n=66)

Placebo (n=66)

Age, years

Female

White/Caucasian

APACHE II score

SOFA score on day 1 (IQR)

ICU admission indication

Postoperative

Sepsis

Medical/other reason

45 (68.2%)

13 (19.7%)

8 (12.1%)

42 (63.6%)

13 (19.7%)

11 (16.7%)

Baseline mean arterial pressure, mm Hg

75.9 ± 9.4

72.8 ± 8.2

Vasopressor dose at study enrollment, mcg/kg/min (IQR)

Norepinephrine (n=41)

Phenylephrine (n=28)

Metaraminol (n=60)

0.06 (0.04-0.08)

0.61 (0.37-0.84)

0.60 (0.38-0.72)

0.06 (0.02-0.09)

0.43 (0.20-1.1)

0.61 (0.46-0.83)

Total duration of vasopressor use, hours (IQR)

Before study drug administration

76.6 (50.5-107.4)

35.5 (28-55)

60.6 (44.3-92)

35.4 (24.7-43.8)

Epidural analgesia

18 (27.3%)

13 (19.7%)

Results

Midodrine (n=66)

Placebo (n=66)

P-value

Time to vasopressor discontinuation, hours (IQR)

Time to ICU discharge readiness, days (IQR)

Length of stay, days

ICU

Hospital

6 (5-8)

11 (9-21)

6 (4-8)

14 (9-22)

0.46

0.45

0.62

The only significant difference in post-hoc subgroup analyses was seen in the 31 patients who received epidural analgesia. The time to vasopressor discontinuation was significantly shorter with midodrine compared to placebo (difference, -18.4 h; 95% confidence interval -33.5 to -3.3 h; incidence rate ratio 0.53; 95% confidence interval 0.28 to 0.99; P=0.045).

There was no significant difference in the 101 patients who did not receive epidural analgesia (P=0103).

Post-hoc sensitivity analyses showed no significant differences in the primary outcome of time to vasopressor discontinuation across different study sites or enrollment years.

Another post-hoc analysis addressed the possibility of delayed or impaired gastrointestinal absorption of the study drug due to opioid analgesic use. The total opioid doses administered during the first 24 hours of study drug administration were not different between the groups (P=0.65). Similarly, there was as no difference in time to vasopressor discontinuation in patients who received opioids (P=0.55) or patients without opioids during the first 24 hours of enrollment (P=0.77).

Adverse Events

Any adverse events: (30.3% vs 25.8%)

Hypertension (10.6% vs 4.6%), bradycardia (7.6% vs 0%; P=0.02), atrial fibrillation (4.6% vs 1.5%)

Study Author Conclusions

Midodrine did not reduce the time to discontinuation of intravenous vasopressors in critically ill patients with persistent hypotension. The lack of effectiveness combined with a higher rate of bradycardia does not support the routine use of midodrine as off-label medication to accelerate liberation from intravenous vasopressors in the ICU.

InpharmD Researcher Critique

This study did not specify specific weight-based doses for the vasopressors used, and there may be a discrepancy based on the location of the hospital (USA or Australia). There no information regarding fluid status in these patients. The baseline MAP seemed higher than normal when the patients remained on a small vasopressor dose. Discontinuation may have occurred by altering the MAP target (e.g., to 65mmHg) without the need for midodrine.

The broad eligibility criteria may have resulted in a heterogeneous cohort with different underlying etiologies of vasopressor-dependent hypotension; however, a majority of patients were postoperative or surgical. Excluding patients already receiving midodrine in the ICU prior to enrolment may have increased the possibility of selection bias.

The post-hoc and subgroup analyses were not pre-specified, suggesting bias in these results.

Midodrine for the weaning of vasopressor infusions

Design

Retrospective, single-center, observational, cohort study

N=188

Objective

Study Groups

Midodrine (n=94)

Control (n=94)

Inclusion Criteria

Aged ≥18 years; admitted to an adult ICU; received ≥1 vasopressor (dopamine, epinephrine, norepinephrine, phenylephrine or vasopressin); had a diagnosis code related to a cardiovascular issue,  trauma, or sepsis

For the midodrine group: received ≥3 midodrine doses

Exclusion Criteria

Died within 24 hours of ICU admission; received IV vasopressors for <2 hours; used midodrine fro an indication other than IV vasopressor weaning

Methods

This was a retrospective study from an institution in North Carolina that had three ICUs (cardiovascular, neuro-trauma. and medical-surgical). Patients who received at least three doses of oral midodrine for IV vasopressor weaning were compared to similar patients who received no midodrine. 

The decision to use midodrine, parameters for midodrine, and IV vasopressor titration and weaning were at the discretion of individual prescribers. Intravenous vasopressor doses were expressed as phenylephrine equivalent rates to adjust for potency differences between agents: 10 mcg/min of dopamine, 0.1 mcg/min of norepinephrine/epinephrine, and 0.0002 units/min of vasopressin to be equivalent to 1 mcg/min of phenylephrine. 

Duration

January 2007 to March 2012

Outcome Measures

Primary: time to IV vasopressor discontinuation after midodrine initiation

Secondary: the time from IV vasopressor discontinuation to ICU discharge, ICU length of stay, the number of ICU readmissions, hospital length of stay, mean arterial pressure (MAP) at the time of IV vasopressor initiation and discontinuation

Baseline Characteristics

Midodrine (n=94)

Control (n=94)

Age, years

Male

APACHE IV score (IQR)

Systemic corticosteroid use

Most common diagnoses

Myocardial infarction

Coronary atherosclerosis

Spinal column injury

Sepsis

16 (17%)

16 (17%)

16 (17%)

23 (24%)

14 (15%)

9 (10%)

4 (4%)

23 (24%)

Vasopressor used

Phenylephrine

Norepinephrine

Dopamine

Vasopressin

Epinephrine

65 (69%)

47 (50%)

42 (45%)

11 (12%)

8 (9%)

56 (60%)

34 (36%)

33 (35%)

10 (11%)

9 (10%)

0.17

0.06

0.18

0.82

0.80

Required multiple vasopressors (≥2)

56 (60%)

35 (37%)

<0.01

Phenylephrine-requivialant IV vasopressor rate, mcg/kg/min (IQR)

Patients in the midodrine group were on IV vasopressors for a median of 1.6 days (IQR, 0.8-0.42 days) before midodrine initiation. Midodrine was mainly dosed 10 mg TID (range, 2.5-10 mg; 2-6 times daily) with a median duration of 4.4 days (IQR, 3.2-7.8 days).

Results

Midodrine (n=94)

Control (n=94)

P-value

Time to IV vasopressor discontinuation after midodrine initiation, days

Cardiovascular

Trauma

Sepsis

1.2

1.0

1.3

2.2

Time to ICU discharge after midodrine initiation, days (IQR)

Time to ICU discharge after vasopressor discontinuation, days (IQR)

Length of stay, days (IQR)

ICU

Hospital

5.5 (3.0-14.8)

12.0 (8.0-21.8)

5.0 (3.0-10.0)

9.5 (5.0-16.0

0.29

<0.01

ICU readmission

11 (12%)

10 (11%)

0.82

Required addition or restart of a vasopressor*

75 (80%)

44 (47%)

<0.01

In-hospital mortality

Mean arterial pressure (MAP), mm Hg

At vasopressor start

At vasopressor discontinuation

67.7 ± 11.9

74.3 ± 9.5

65.9 ± 13.3

75.0 ± 12.5

0.35

0.69

*In the midodrine group, a majority of the vasopressor additions or restarts were before the midodrine was started (60%).

Adverse Events

During the 24 hours period after starting midodrine, six patients (6.4%) experienced hypertension and 12 patients (12.8%) experienced bradycardia not present during the 24-h period before starting midodrine.

Study Author Conclusions

These results suggest that midodrine has potential as a useful adjunctive treatment in the weaning of IV vasopressor infusions in difficult to wean patients who are otherwise stable. At this institution, midodrine has been used successfully in a variety of medical and surgical ICU patient populations.

InpharmD Researcher Critique

This study is limited by its single-center and retrospective nature, which relies on accurate documentation. While the study and control group had an equal number of patients, they were not matched. This allows for potential confounding and selection bias. Additionally, the control group did not have anything to compare the primary endpoint to, making it inadequate.

Although only some secondary endpoints could be compared, the control group ad a significantly higher APACHE IV score, which may have impacted mortality and length of stay. The main utility of the control group in this study is to narrow down the population in which midodrine was used, as any conclusions drawn on midodrine efficacy in comparison would be invalid. Midodrine was particularly used in patients who required systemic corticosteroids and multiple IV vasopressors, which supports that these were patients who had more difficulty maintaining hemodynamic stability compared to the control patients. Patients prescribed midodrine also had more failed attempts at weaning IV vasopressors (before the decision to use midodrine).

Oral midodrine treatment accelerates the liberation of intensive care unit patients from intravenous vasopressor infusions

Design

Prospective, observational, open-label, cohort study

N=20

Objective

To evaluate the  magnitude in the decline of IV vasopressor rate is greater during midodrine administration than prior to midodrine treatment

Study Groups

Inclusion Criteria

Exclusion Criteria

Administered midodrine as a pre-admission medicine; received <3 doses of midodrine; hypotension secondary to hypovolemia or adrenal insufficiency; history of orthostatic hypotension

Methods

Midodrine was given at a modal dose of 20 mg (range, 5-20 mg) TID. Patients were followed throughout their hospital stay and any ICU readmissions were recorded.

Duration

The median duration of midodrine for the entire hospitalization was 4 days (IQR, 3-7). Patients were followed for at least 4 midodrine doses.

Outcome Measures

Primary: time to IV vasopressor discontinuation

Secondary: hemodynamic variables, length of stay

Baseline Characteristics

Midodrine (n=20)

Age, years

Female

APACHE II score

Vasopressor days pre-midodrine (IQR)

Baseline vasopressor rate, mcg/min of phenylephrine equivalents

Results

Midodrine (n=20)

P-value

Time to IV vasopressor discontinuation after midodrine, hours (IQR)

Patients off IV vasopressors after 24 hours of midodrine

Mean arterial pressure (MAP), mm Hg

Before midodrine

After four midodrine doses

76 ± 8

77 ± 8

Heart rate (HR), beats per minute

Before midodrine

After four midodrine doses

82 ± 13

81 ± 15

Time to discharge, days (IQR)

ICU

Hospital

4 (3-6)

8.5 (5-16)

The decline in IV vasopressor rate changed from -0.62 ± 1.40 mcg/min/h of phenylephrine equivalents before midodrine to -2.20 ± 2.45 mcg/min/h after midodrine (P=0.012).

There was no significant correlation between the change in slope of phenylephrine equivalent rates and either MAP, HR, total body fluid balance (TBFB), and white blood cell count (WBC) pre-midodrine to after four midodrine doses.

Adverse Events

Two patients (10%) discontinued midodrine prematurely due to a bleeding episode requiring re-operation (5%) and recurrence of sepsis (5%).

Study Author Conclusions

This study shows that administration of midodrine significantly increased the magnitude of the decline of the phenylephrine equivalent rate. Midodrine may help liberate ICU patients from an IV vasopressor infusion once their clinical condition otherwise no longer necessitates critical care level of services.

InpharmD Researcher Critique

This study is limited by the single-group, observational nature which allows for confounding variables. Changes in vasopressor dosing appeared to be nonparametric, as they were presented as a mean ± standard deviation where the deviation sometimes exceeded the mean. The authors should have represented this data as a median, as using a mean to calculate significance is misleading and likely not clinically extrapolatable.

Feasibility, Utility, and Safety of Midodrine During Recovery Phase From Septic Shock

Design

Retrospective, single-center, observational study

N=275

Objective

To assess the feasibility, utility, and safety of oral midodrine to replace intravenous (IV) vasopressors during recovery from septic shock

Study Groups

IV vasopressors only (n=140)

Midodrine (n=135)

Inclusion Criteria

Exclusion Criteria

None mentioned

Methods

This was a retrospective study of a single center in New York. The decision to administer midodrine (and its dosing) was made on an individual basis by the clinical team caring for the patient. The typical midodrine starting dose was 10 mg q8h, which were incrementally increased until IV vasopressor cessation. No standardized protocol was used to administer, dose, or taper midodrine.

Duration

November 2013 to November 2014

Midodrine was given for an average of 6.15 days.

Outcome Measures

Primary: duration of IV vasopressor administration

Secondary: length of stay, in-hospital mortality

Baseline Characteristics

IV vasopressors only (n=140)

Midodrine (n=135)

Age, years

Female

APACHE IV score

Mechanical ventilation

0.72

There were no significant differences in the source of sepsis between the two groups.

The maximum midodrine dose used was 40 mg q8h.

Results

IV vasopressors only (n=140)

Midodrine (n=135)

P-value

IV vasopressor days

Vasopressor reinitiation

Length of stay, days

ICU

Hospital

9.4 ± 6.7

24.2 ± 14.3

7.5 ± 5.9

21.9 ± 14.4

0.017

0.30

Change in creatinine, mg/dL

Mortality

ICU

Hospital

18.6%

25.7%

11.1%

23%

0.08

0.60

Adverse Events

Midodrine was discontinued prior to IV vasopressor discontinuation in one patient secondary to bradycardia which resolved without further treatment.

Study Author Conclusions

Our results indicate that the use of oral midodrine during recovery from septic shock is a safe and effective method to reduce the duration of IV vasopressors and ICU length of stay.

InpharmD Researcher Critique

This study is limited by the retrospective, single-center design; these studies may suffer from bias and confounding variables such as the possibility that the study groups had unrecognized differences. The decision to administer midodrine and its dosing were made by the clinical team as was the decision to taper the IV vasopressor, and these decisions were not protocol-driven. Additionally, the decision to use midodrine inherently introduces selection bias, limiting a meaningful comparison between the two groups.