Dexamethasone

What it is

Dexamethasone is a synthetic adrenocortical steroid — a laboratory-made glucocorticoid. The label describes it as a white to practically white, odourless, crystalline powder that is practically insoluble in water. It is given as tablets, an oral solution, injections, eye drops and ointments, an intravitreal implant and a tear-duct insert, so the same molecule reaches the body by very different routes.

What the research says

What the evidence actually shows depends entirely on which condition is being treated. There are large randomised trials with mortality outcomes in COVID-19 needing respiratory support, tuberculous meningitis, antenatal use before early preterm birth and prevention of postoperative sickness, and a Cochrane review in croup. For several of its long-standing approved indications there is no modern outcome trial at all, and the FDA-approved indication list is a regulatory position rather than an evidence summary. Against this sit well-documented harms: raised blood sugar, infection, bone loss, mood change and suppression of the body's own cortisol production after stopping.

Evidence grade: Well established.

How it works

Drug class: Synthetic glucocorticoid (adrenocortical steroid)

Dexamethasone is a man-made copy of the body's own adrenal steroid hormones. Taken by mouth it is absorbed from the gut, then acts on cells throughout the body to damp down inflammation and alter immune responses; it also changes how the body handles sugar, protein, calcium and salt. Unlike hydrocortisone it has almost no salt-retaining action at equivalent anti-inflammatory strength. In myeloma cells it triggers cell death, although the label states the precise mechanism there is not known. (Source 1)

What it is used for

• A placebo-controlled trial in 545 adolescents and adults found fewer deaths with dexamethasone (relative risk 0.69) but no reduction in severe disability. A later placebo-controlled trial restricted to 520 HIV-positive adults found no survival benefit. Evidence: established. (Source 2)
• The label records that corticosteroid trials speed the resolution of a relapse but do not change the eventual outcome or the natural history of the disease. We found no modern outcome trial of dexamethasone specifically for this use. Evidence: limited. (Source 3)
• A 20 mg dexamethasone tablet is approved only as part of combination myeloma regimens, and its dosing is taken from the partner drug's prescribing information. The label states the precise mechanism in myeloma is unknown. We did not review the combination-regimen trials in this slice. Evidence: established. (Source 4)
• This is a long-standing label indication. We found no outcome trial of dexamethasone alone for it within this slice's budget, so the evidence here rests on the regulatory position rather than on trials we read. Evidence: unknown. (Source 5)
• Off-label in the United States: COVID-19 does not appear among the approved indications on the dexamethasone tablet label. The RECOVERY trial found lower 28-day mortality in patients on oxygen or ventilation, with the largest absolute benefit in ventilated patients, and no benefit (results consistent with harm) in those needing no respiratory support. Evidence: established. (Source 6)
• Off-label against the US tablet label's indication list. A 2023 Cochrane review of 45 trials in 5,888 children found glucocorticoids reduce croup scores versus placebo, with low-certainty evidence and 98% of trials at high or unclear risk of bias, and found a lower dose may be as effective as the standard dose. Evidence: established. (Source 7)
• Off-label against the tablet label, which lists only tuberculous meningitis. The 2015 Cochrane review found no reduction in overall mortality (17.8% vs 19.9%, risk ratio 0.90, P=0.07), lower rates of hearing loss, no benefit in low-income countries, and no effect on severe hearing loss when only high-quality studies were analysed. Evidence: disputed. (Source 8)
• Off-label against the tablet label's indication list. The ACTION-I trial in five low-resource countries found neonatal death fell from 23.5% to 19.6% (relative risk 0.84), with a stated number needed to treat of 25 women, and no increase in possible maternal infection. Evidence: established. (Source 9)
• Off-label against the tablet label's indication list. In 8,725 adults having noncardiac surgery, a single 8 mg intravenous dose cut sickness in the first 24 hours from 53.9% to 42.2% and did not increase surgical-site infection, at the cost of more hyperglycaemic events (0.6% vs 0.2%). Evidence: established. (Source 10)

Interactions

• Calcium and vitamin D supplements (clinical trial): In people taking systemic corticosteroids, adding calcium with vitamin D slowed bone loss at the spine and forearm in a Cochrane meta-analysis of five trials; fracture rates and femoral neck bone mass were not significantly different. This is a protective combination rather than a harmful clash. The review's own recommendation went further than the figures: it concluded that because of low toxicity and cost, all patients being started on corticosteroids should receive prophylactic calcium and vitamin D. Limit: The two weighted mean differences carry no unit in the source; they are percent change in bone mineral density. The source also prints the sentence with unbalanced brackets. The review's searches ran only up to 1996, so nothing published since is included, and fracture rates and femoral neck bone mass were not significantly different. (Source 11)
• St John's wort and other strong CYP3A4 inducers (theoretical): Dexamethasone is broken down by the liver enzyme CYP3A4, and the label states that drugs which induce CYP3A4 speed up corticosteroid metabolism so that the corticosteroid dose has to be increased. St John's wort is a well-known strong CYP3A4 inducer, so the same effect is expected, but the label names only drugs (barbiturates, phenytoin, carbamazepine, rifampin) and we found no published pharmacokinetic study of St John's wort with dexamethasone. (Source 12)
• Grapefruit juice (pharmacokinetic study): Grapefruit juice inhibits CYP3A4, the enzyme that clears dexamethasone. In rats given grapefruit juice alongside several COVID-19-era medicines, dexamethasone plasma concentrations rose. This is an animal pharmacokinetic finding and has not been shown in people. Limit: This was measured in rats, not people, and the caution about patients is the authors' own extrapolation from that animal study. Of the four drugs the paper assayed, grapefruit juice raised plasma levels of three (azithromycin, apixaban and dexamethasone); favipiravir is not among them. (Source 13)
• Food (a high-fat, high-calorie meal) (pharmacokinetic study): A high-fat, high-calorie meal did not change total dexamethasone exposure after a 20 mg tablet but lowered the peak concentration by 23%. The myeloma label states the tablet can be taken with or without food. (Source 14)
• Alcohol (case reports): In a case-control study of 120 older people admitted with gastrointestinal bleeding while taking non-aspirin NSAIDs, heavy alcohol abuse carried an adjusted odds ratio of 4.44 and NSAIDs combined with corticosteroids an odds ratio of 10.70. Both were independent risk factors in the same model; the study did not test alcohol and corticosteroids together, so a specific alcohol-dexamethasone interaction has not been measured. Limit: Nothing here measures alcohol with dexamethasone: heavy alcohol use and NSAIDs-with-corticosteroids were separate risk factors in one model and were never tested together. The comparison is also narrower than it looks, because the people this study compared were all over 60 and all taking non-aspirin NSAIDs, cases and controls alike, so these odds ratios are not the risk of taking an NSAID. Two of the source's own figures are faulty: one interval is printed with no dash between its bounds ('2.10 17.43'), and the odds ratio of 9.94 for irregular food intake does not sit in the middle of the interval printed with it (3.29 to 24.28), so one of those three numbers is wrong in the paper. (Source 15)
• Cholestyramine (label): Cholestyramine increases how fast corticosteroids are cleared from the body, which can reduce the dexamethasone effect. (Source 12)
• Diabetes medicines (label): Corticosteroids raise blood glucose, so doses of diabetes medicines may need adjusting during treatment. (Source 12)
• Digitalis glycosides and potassium-losing drugs (label): Corticosteroids increase potassium loss; with digitalis glycosides this raises the risk of heart rhythm disturbances. (Source 12)
• Ketoconazole, erythromycin and other CYP3A4 inhibitors (label): The other half of the same CYP3A4 interaction. Drugs that block CYP3A4 slow the breakdown of dexamethasone and raise its blood levels: the label states ketoconazole has been reported to cut the metabolism of certain corticosteroids by up to 60%, and names itraconazole, clarithromycin, ritonavir and cobicistat-containing products as other strong inhibitors. The label asks for the benefit to be weighed against the risk of systemic corticosteroid side effects. (Source 12)
• Itraconazole and aprepitant (measured CYP3A4 inhibition) (pharmacokinetic study): Measured rather than predicted. In the studies recorded in the myeloma label, itraconazole (a strong CYP3A4 inhibitor) raised total dexamethasone exposure 3.7-fold after a single 4.5 mg oral dose, and aprepitant (a moderate inhibitor) raised it 2.2-fold. These are the numbers behind the label's general warning about CYP3A4 inhibitors. (Source 14)

Stopping it

• Stopping a glucocorticoid is not neutral: the body's own cortisol production can stay suppressed. A meta-analysis of 74 studies in 3,753 people found adrenal insufficiency after stopping is common and that no route, dose, duration or underlying disease rules it out, with the highest risk after higher doses and longer use. (Source 16)
• The label states that this drug-induced adrenal suppression can be reduced by tapering rather than stopping abruptly, and that the relative insufficiency may last for months after treatment ends, so stress during that period may need steroid cover again. (Source 17)
• The label also describes a distinct corticosteroid withdrawal syndrome after prolonged therapy, with muscle pain, joint pain and malaise, and warns against stopping abruptly or without medical supervision. (Source 18)
• The dosing section states the same thing as a general rule for long-term therapy: the drug should be withdrawn gradually rather than abruptly. (Source 19)

What goes wrong

Short courses of oral corticosteroids were followed by higher rates of sepsis, venous clots and fracture in a very large insurance-claims study. (Source 20)

• Cohort study, Low certainty.
• Size: 1,548,945 adults, of whom 327,452 (21.1%) had at least one short course.
• Who: US privately insured adults aged 18 to 64, 2012 to 2014; median 6 days of treatment, median prednisone-equivalent 20 mg/day.
• How long: 30-day and 31-90-day risk windows after starting the drug.
• Result: Within 30 days, incidence rate ratio 5.30 (95% CI 3.80 to 7.41) for sepsis, 3.33 (2.78 to 3.99) for venous thromboembolism and 1.87 (1.69 to 2.07) for fracture; the excess persisted below 20 mg/day prednisone equivalent (4.02, 3.61 and 1.83).
• Funding: not stated in the abstract; published in the BMJ.

Limit of this finding: This is an observational self-controlled and cohort analysis of all oral corticosteroids — most prescriptions were methylprednisolone dose-packs, not dexamethasone — so it shows an association for the class, not a measured effect of dexamethasone, and it cannot establish cause.

Within 30 days of drug initiation, there was an increase in rates of sepsis (incidence rate ratio 5.30, 95% confidence interval 3.80 to 7.41), venous thromboembolism (3.33, 2.78 to 3.99), and fracture (1.87, 1.69 to 2.07), which diminished over the subsequent 31-90 days. The increased risk persisted at prednisone equivalent doses of less than 20 mg/day (incidence rate ratio 4.02 for sepsis, 3.61 for venous thromboembolism, and 1.83 for fracture; all P<0.001).

The absolute excess risk from a short corticosteroid course was small even though the relative risks were large. (Source 21)

• Cohort study, Low certainty.
• Size: 327,452 users and 1,221,493 non-users.
• Who: US privately insured adults aged 18 to 64 who had a clinic visit.
• How long: 5 to 90 days after the visit.
• Result: Hospital admission for sepsis 0.05% in users vs 0.02% in non-users; venous thromboembolism 0.14% vs 0.09%; fracture 0.51% vs 0.39%. The absolute differences are about 3, 5 and 12 extra events per 10,000 people.
• Funding: not stated in the abstract.

Limit of this finding: Two cautions on these numbers. The paper's own paragraph gives two different windows for the same figures, opening with 'the five to 90 day period after a clinic visit' and closing with 'the 90 days after a clinic visit', while its abstract uses 0 to 30 days and 31 to 90 days; so the time window these absolute risks belong to is not pinned down. And this is an observational study of all oral corticosteroids, mostly methylprednisolone dose-packs, so it is a class-level association and not a measured dexamethasone effect.

the risk of hospital admission for sepsis was 0.05% (n=170 of 327 452) in steroid users compared with 0.02% (n=293 of 1 221 493) in non-users during this period.

Dexamethasone caused measurably more hyperglycaemic events than placebo in a large surgical trial, while not increasing surgical-site infection. (Source 10)

• Randomized trial, High certainty.
• Size: 8,725 adults; hyperglycaemia analysed in 7,563 without diabetes.
• Who: Adults having nonurgent, noncardiac surgery.
• How long: 30 days for infection; perioperative period for glucose.
• Result: Hyperglycaemic events in patients without diabetes 22 of 3,787 (0.6%) with dexamethasone vs 6 of 3,776 (0.2%) with placebo — an absolute increase of 0.4 percentage points, about one extra event per 250 people treated. Surgical-site infection 8.1% vs 9.1%, adjusted risk difference −0.9 percentage points (95.6% CI −2.1 to 0.3), meeting the prespecified 2.0-point noninferiority margin.
• Funding: Australian National Health and Medical Research Council and others.

Hyperglycemic events in patients without diabetes occurred in 22 of 3787 (0.6%) in the dexamethasone group and in 6 of 3776 (0.2%) in the placebo group.

Corticosteroid treatment for bacterial meningitis increased recurrent fever. (Source 22)

• Systematic review, Moderate certainty.
• Size: 25 studies, 4,121 participants.
• Who: Adults and children with acute bacterial meningitis.
• How long: as reported in each trial.
• Result: Recurrent fever risk ratio 1.27 (95% CI 1.09 to 1.47); no increase in other adverse events was found.
• Funding: Cochrane review.

Limit of this finding: Corticosteroids as a class: 22 of the 25 trials used dexamethasone and three used hydrocortisone or prednisone.

Corticosteroid treatment was associated with an increase in recurrent fever (RR 1.27, 95% CI 1.09 to 1.47), but not with other adverse events.

Suppression of the body's own cortisol after glucocorticoid treatment is common and no route, dose or duration is free of it. (Source 16)

• Meta-analysis, Moderate certainty.
• Size: 74 articles, 3,753 participants.
• Who: Adults who had taken corticosteroids for a range of conditions and by a range of routes.
• How long: from under 28 days to over a year.
• Result: Pooled proportion with adrenal insufficiency 4.2% for nasal use (95% CI 0.5 to 28.9) up to 52.2% for intra-articular use (40.5 to 63.6); by disease 6.8% in asthma on inhalers (3.8 to 12.0) up to 60.0% in haematological malignancy (38.0 to 78.6); by dose 2.4% (0.6 to 9.3) low dose to 21.5% (12.0 to 35.5) high dose; by duration 1.4% (0.3 to 7.4) under 28 days to 27.4% (17.7 to 39.8) over a year in asthma.
• Funding: not stated in the abstract.

Limit of this finding: The pooled percentages come from studies of many different corticosteroids and testing protocols, and some confidence intervals are extremely wide (for example 0.5% to 28.9% for nasal use), so individual figures should not be read as precise risks for dexamethasone.

Adrenal insufficiency after discontinuation of glucocorticoid occurs frequently; 2) there is no administration form, dosing, treatment duration, or underlying disease for which adrenal insufficiency can be excluded with certainty, although higher dose and longer use give the highest risk

The label records immune suppression and a dose-related rise in infection, including reactivation of tuberculosis, chickenpox, measles and hepatitis B. (Source 23)

• Official position, Certainty not rated.
• Size: not stated.
• Who: People taking systemic corticosteroids.
• How long: not stated.
• Result: No rates are given. The label states the rate of infectious complications rises with increasing dose and that infections can be severe and at times fatal.
• Funding: not applicable (regulatory label)

Limit of this finding: The closing sentence of this label subsection forbids giving dexamethasone into a joint, bursa, tendon or lesion where there is acute local infection. It is the label's own text, but this product is an oral tablet with no injectable form, so that sentence is about corticosteroid injections generally and not about the tablet.

Corticosteroid-associated infections can be mild but can be severe and at times fatal. The rate of infectious complications increases with increasing corticosteroid dosages. Monitor for the development of infection and consider DEXAMETHASONE TABLETS withdrawal or dosage reduction as needed.

The label records bone loss, osteoporosis and fracture as corticosteroid effects, with a stated mechanism. (Source 24)

• Official position, Certainty not rated.
• Size: not stated.
• Who: People taking systemic corticosteroids, at any age.
• How long: not stated.
• Result: No rates are given. The label attributes bone loss to reduced calcium absorption, increased calcium excretion, inhibition of osteoblast function, protein catabolism and reduced sex hormone production.
• Funding: not applicable (regulatory label)

Corticosteroids decrease bone formation and increase bone resorption both through their effect on calcium regulation (i.e., decreasing absorption and increasing excretion) and inhibition of osteoblast function. This, together with a decrease in the protein matrix of the bone secondary to an increase in protein catabolism and reduced sex hormone production, may lead to inhibition of bone growth in pediatric patients and the development of osteoporosis at any age.

The label records mood and psychiatric effects ranging from euphoria and insomnia to frank psychosis, and an acute myopathy at high doses. (Source 25)

• Official position, Certainty not rated.
• Size: not stated.
• Who: People taking systemic corticosteroids.
• How long: not stated; recovery from the myopathy may take weeks to years.
• Result: No rates are given. The acute myopathy is described as generalised, possibly involving ocular and respiratory muscles, and may result in quadriparesis.
• Funding: not applicable (regulatory label)

Psychic derangements may appear when corticosteroids are used, ranging from euphoria, insomnia, mood swings, personality changes and severe depression, to frank psychotic manifestations.

Listed endocrine and metabolic adverse reactions include raised blood sugar, a cushingoid state and suppression of the pituitary-adrenal axis. (Source 26)

• Official position, Certainty not rated.
• Size: not stated.
• Who: People taking dexamethasone or other corticosteroids.
• How long: not stated.
• Result: No rates are given; the label lists decreased glucose tolerance, hyperglycaemia, glycosuria, increased insulin requirements, unmasking of latent diabetes, osteoporosis, pathological fracture of long bones, vertebral compression fractures and peptic ulcer with possible perforation and haemorrhage.
• Funding: not applicable (regulatory label)

Endocrine:Decreased carbohydrate and glucose tolerance, development of cushingoid state, hyperglycemia, glycosuria, hirsutism, hypertrichosis, increased requirements for insulin or oral hypoglycemic agents in diabetes, manifestations of latent diabetes mellitus, menstrual irregularities, secondary adrenocortical and pituitary unresponsiveness (particularly in times of stress, as in trauma, surgery or illness), suppression of growth in pediatric patients.

The label instructs that people on corticosteroids be warned to avoid exposure to chickenpox or measles and to seek medical advice without delay if exposed. (Source 18)

• Official position, Certainty not rated.
• Size: not stated.
• Who: People taking systemic corticosteroids.
• How long: not stated.
• Result: No rates are given. This is the label’s patient-counselling instruction, and it sits alongside the WARNINGS statement that chickenpox and measles can have a more serious or even fatal course in people on corticosteroids.
• Funding: not applicable (regulatory label)

Persons who are on corticosteroids should be warned to avoid exposure to chickenpox or measles. Patients should also be advised that if they are exposed, medical advice should be sought without delay.

What the evidence supports

In hospitalised COVID-19 patients, dexamethasone lowered 28-day mortality overall and most in those on invasive mechanical ventilation. (Source 6)

• Randomized trial, High certainty.
• Size: 6,425 patients (2,104 dexamethasone, 4,321 usual care)
• Who: Patients hospitalised with COVID-19 in the UK; mean age 66.1 years.
• How long: up to 10 days of treatment, 28-day outcome.
• Result: 28-day mortality 22.9% (482/2104) vs 25.7% (1110/4321); age-adjusted rate ratio 0.83 (95% CI 0.75 to 0.93), P<0.001. Invasive mechanical ventilation subgroup 29.3% vs 41.4%, rate ratio 0.64 (0.51 to 0.81); oxygen without invasive ventilation 23.3% vs 26.2%, rate ratio 0.82 (0.72 to 0.94). Absolute difference overall 2.8 percentage points, giving a number needed to treat of about 36 in the whole trial population.
• Funding: Medical Research Council and National Institute for Health Research and others (independent of a manufacturer)

Limit of this finding: These figures are from the trial's preliminary report of July 2020; the final 2021 version of the paper prints the interval for the subgroup needing no respiratory support as 0.92 to 1.55 rather than 0.91 to 1.55, and is otherwise the same on this passage.

Overall, 482 patients (22.9%) in the dexamethasone group and 1110 patients (25.7%) in the usual care group died within 28 days after randomization (age-adjusted rate ratio, 0.83; 95% confidence interval [CI], 0.75 to 0.93; P<0.001). The proportional and absolute between-group differences in mortality varied considerably according to the level of respiratory support that the patients were receiving at the time of randomization. In the dexamethasone group, the incidence of death was lower than that in the usual care group among patients receiving invasive mechanical ventilation (29.3% vs. 41.4%; rate ratio, 0.64; 95% CI, 0.51 to 0.81) and among those receiving oxygen without invasive mechanical ventilation (23.3% vs. 26.2%; rate ratio, 0.82; 95% CI, 0.72 to 0.94)

The absolute mortality benefit in RECOVERY was concentrated in ventilated patients. (Source 27)

• Randomized trial, High certainty.
• Size: 6,425 patients.
• Who: Hospitalised COVID-19 patients by level of respiratory support.
• How long: 28 days.
• Result: Age-adjusted absolute reduction in 28-day mortality 12.3 percentage points (95% CI 6.3 to 17.6) with invasive mechanical ventilation (number needed to treat about 8) and 4.2 percentage points (95% CI 1.4 to 6.7) on oxygen only (number needed to treat about 24).
• Funding: Medical Research Council and National Institute for Health Research and others.

Limit of this finding: The paper is internally inconsistent about this number: the Results section gives the oxygen-only absolute reduction as 4.2 percentage points while the Discussion gives 4.1 age-adjusted percentage points. A reader should treat the oxygen-only absolute benefit as roughly four percentage points and not rely on either figure to one decimal place. These figures are from the trial's preliminary report of July 2020; the final 2021 version of the paper prints the interval for the subgroup needing no respiratory support as 0.92 to 1.55 rather than 0.91 to 1.55, and is otherwise the same on this passage.

The age-adjusted absolute reductions in 28-day mortality associated with the use of dexamethasone were 12.3 percentage points (95% CI, 6.3 to 17.6) among the patients who were receiving invasive mechanical ventilation and 4.2 percentage points (95% CI, 1.4 to 6.7) among those receiving oxygen only.

Adjunctive dexamethasone reduced death in adolescents and adults with tuberculous meningitis but did not reduce severe disability. (Source 2)

• Randomized trial, Moderate certainty.
• Size: 545 patients (274 dexamethasone, 271 placebo)
• Who: Patients over 14 years with tuberculous meningitis in Vietnam, with or without HIV.
• How long: nine months of follow-up.
• Result: Relative risk of death 0.69 (95% CI 0.52 to 0.92), P=0.01. Severe disability among survivors 18.2% (34/187) with dexamethasone vs 13.8% (22/159) with placebo, P=0.27; death or severe disability odds ratio 0.81 (0.58 to 1.13), P=0.22. Serious adverse events 26/274 vs 45/271, P=0.02.
• Funding: not stated in the abstract.

Limit of this finding: The trial reports fewer deaths with dexamethasone yet numerically more severe disability among survivors and no significant effect on the combined outcome of death or severe disability. It should not be read as showing that dexamethasone improves functional recovery.

Treatment with dexamethasone was associated with a reduced risk of death (relative risk, 0.69; 95 percent confidence interval, 0.52 to 0.92; P=0.01). It was not associated with a significant reduction in the proportion of severely disabled patients (34 of 187 patients [18.2 percent] among survivors in the dexamethasone group vs. 22 of 159 patients [13.8 percent] in the placebo group, P=0.27)

Antenatal dexamethasone reduced neonatal death in women at risk of early preterm birth in low-resource hospitals. (Source 9)

• Randomized trial, High certainty.
• Size: 2,852 women and 3,070 babies across 29 hospitals in Bangladesh, India, Kenya, Nigeria and Pakistan.
• Who: Pregnant women at risk of preterm birth between 26 weeks 0 days and 33 weeks 6 days.
• How long: 28 days of neonatal follow-up.
• Result: Neonatal death 19.6% (278/1417) vs 23.5% (331/1406), relative risk 0.84 (95% CI 0.72 to 0.97), P=0.03; absolute reduction 3.9 percentage points and a stated number needed to treat of 25 women (95% CI 14 to 110) to prevent one neonatal death. Any baby death 25.7% vs 29.2%, relative risk 0.88 (0.78 to 0.99).
• Funding: Bill and Melinda Gates Foundation (not a drug manufacturer)

There were 278 (19.6%) neonatal deaths among 1417 liveborn infants in the dexamethasone group and 331 (23.5%) neonatal deaths among 1406 liveborn infants in the placebo group (relative risk 0.84; 95% confidence interval [CI], 0.72 to 0.97; P=0.03) (Table 2). We determined that 25 women would need to be treated with dexamethasone to prevent one neonatal death (95% CI, 14 to 110).

A single intravenous dose of dexamethasone during anaesthesia reduced postoperative nausea and vomiting. (Source 10)

• Randomized trial, High certainty.
• Size: 8,725 adults in the modified intention-to-treat population.
• Who: Adults having nonurgent, noncardiac surgery lasting at least 2 hours with an overnight stay; 13.2% had diabetes.
• How long: first 24 hours after surgery for sickness; 30 days for infection.
• Result: Postoperative nausea and vomiting in the first 24 hours 42.2% with dexamethasone vs 53.9% with placebo, risk ratio 0.78 (95% CI 0.75 to 0.82); absolute reduction 11.7 percentage points, number needed to treat about 9.
• Funding: Australian National Health and Medical Research Council and others.

Postoperative nausea and vomiting in the first 24 hours after surgery occurred in 42.2% of patients in the dexamethasone group and in 53.9% in the placebo group (risk ratio, 0.78; 95% CI, 0.75 to 0.82)

Glucocorticoids, dexamethasone chief among them, reduce croup symptom scores compared with placebo, but the certainty of that evidence is low. (Source 7)

• Systematic review, Low certainty.
• Size: 45 randomised trials, 5,888 children (placebo comparisons: 7 to 11 trials, 426 to 959 children per timepoint)
• Who: Children aged 0 to 18 years with croup.
• How long: croup score measured at 2, 6, 12 and 24 hours.
• Result: Croup score vs placebo at two hours standardised mean difference −0.65 (95% CI −1.13 to −0.18; 7 RCTs, 426 children; low-certainty); six hours −0.76 (−1.12 to −0.40; 11 RCTs, 959 children; low-certainty); 12 hours −1.03 (−1.53 to −0.53; 8 RCTs, 571 children; low-certainty). The review graded 98% of the included studies as at high or unclear risk of bias.
• Funding: Review-level: government 11%, academic 7%, industry 18%, foundations 9%, with 55% of studies not reporting funding.

Limit of this finding: The same review says glucocorticoids 'probably reduced the return visits or (re)admissions for croup by almost half', but the interval it gives for that result (risk ratio 0.55, 95% CI 0.28 to 1.11) includes 1, so the review's own figures do not rule out no effect; and the 28% reduction in supplemental glucocorticoid use is the one estimate in the passage with no certainty rating attached. Read the croup-score reductions, which the review grades low certainty, as the finding here.

Compared to placebo, glucocorticoids may result in greater reductions in croup score after two hours (standardised mean difference (SMD) −0.65, 95% confidence interval (CI) −1.13 to −0.18; 7 RCTs, 426 children; low‐certainty evidence); six hours (SMD −0.76, 95% CI −1.12 to −0.40; 11 RCTs, 959 children; low‐certainty evidence); and 12 hours (SMD −1.03, 95% CI −1.53 to ‐0.53; 8 RCTs, 571 children; low‐certainty evidence)

What the evidence does not support

Dexamethasone was no better than placebo for 28-day mortality in COVID-19 patients who needed no respiratory support, and the result was consistent with harm. (Source 28)

• Randomized trial, High certainty.
• Size: 1,535 patients in the no-respiratory-support subgroup (of 6,425 randomised)
• Who: Hospitalised COVID-19 patients receiving no oxygen at randomisation.
• How long: 28 days.
• Result: 28-day mortality 17.8% with dexamethasone vs 14.0% with usual care; rate ratio 1.19 (95% CI 0.91 to 1.55) — a point estimate in the direction of harm with a confidence interval that includes no effect.
• Funding: Medical Research Council and National Institute for Health Research and others.

Limit of this finding: These figures are from the trial's preliminary report of July 2020; the final 2021 version of the paper prints the interval for the subgroup needing no respiratory support as 0.92 to 1.55 rather than 0.91 to 1.55, and is otherwise the same on this passage.

However, there was no evidence that dexamethasone provided any benefit among patients who were not receiving respiratory support at randomization, and the results were consistent with possible harm in this subgroup.

In HIV-positive adults with tuberculous meningitis, dexamethasone did not improve survival. (Source 29)

• Randomized trial, High certainty.
• Size: 520 adults (263 dexamethasone, 257 placebo)
• Who: HIV-positive adults with tuberculous meningitis in Vietnam and Indonesia; half had never had antiretroviral therapy.
• How long: 12 months.
• Result: Death at 12 months 44.1% (116/263) vs 49.0% (126/257); hazard ratio 0.85 (95% CI 0.66 to 1.10), P=0.22. Serious adverse events 73.0% vs 75.5% (P=0.52). No subgroup clearly benefited.
• Funding: Wellcome Trust (independent of a manufacturer)

Limit of this finding: The figure 49.0% appears twice in this passage for two different things: the share of participants who had never had antiretroviral therapy (255 of 520) and mortality in the placebo group (126 of 257). The trial was funded by the Wellcome Trust and registered as NCT03092817.

During the 12 months of follow-up, death occurred in 116 of 263 participants (44.1%) in the dexamethasone group and in 126 of 257 participants (49.0%) in the placebo group (hazard ratio, 0.85; 95% confidence interval, 0.66 to 1.10

Across all randomised trials in acute bacterial meningitis, corticosteroids did not reduce overall mortality, and there was no benefit in low-income countries. (Source 8)

• Systematic review, Moderate certainty.
• Size: 25 studies, 4,121 participants (2,511 children, 1,517 adults)
• Who: People of all ages with acute bacterial meningitis; 9 trials in low-income and 16 in high-income countries.
• How long: trial follow-up as reported in each study.
• Result: Mortality 17.8% vs 19.9%, risk ratio 0.90 (95% CI 0.80 to 1.01), P=0.07; adults risk ratio 0.74 (0.53 to 1.05), P=0.09. Severe hearing loss risk ratio 0.67 (0.51 to 0.88); any hearing loss 0.74 (0.63 to 0.87); neurological sequelae 0.83 (0.69 to 1.00).
• Funding: Cochrane review; funding of individual trials not stated in the abstract.

Limit of this finding: The review contradicts itself in places: neurological sequelae are described as reduced although the confidence interval reaches 1.00, and the quality-based subgroup analysis found no effect on severe hearing loss in the high-quality studies even though the conclusion supports use in high-income countries. A reader should not treat the hearing-loss benefit as settled.

There was insufficient evidence that corticosteroids caused a reduction in mortality overall (17.8% versus 19.9%; risk ratio (RR) 0.90, 95% confidence interval (CI) 0.80 to 1.01; P = 0.07), or for adults (RR 0.74, 95% CI 0.53 to 1.05; P = 0.09).

Restricting the meningitis analysis to high-quality studies removed the hearing-loss benefit, and no benefit was seen in low-income countries. (Source 22)

• Systematic review, Low certainty.
• Size: subgroups within 25 studies and 4,121 participants.
• Who: Adults and children with acute bacterial meningitis, split by country income and study quality.
• How long: as reported in each trial.
• Result: High-income countries: severe hearing loss risk ratio 0.51 (95% CI 0.35 to 0.73), any hearing loss 0.58 (0.45 to 0.73), short-term neurological sequelae 0.64 (0.48 to 0.85). No beneficial effect in low-income countries. Subgroup analysis by study quality showed no effect on severe hearing loss in the high-quality studies.
• Funding: Cochrane review.

Limit of this finding: These are subgroup results for corticosteroids as a class: 22 of the 25 trials used dexamethasone and three used hydrocortisone or prednisone, so none of these numbers is a dexamethasone-specific effect. The review's own conclusion supports corticosteroids in high-income countries even though its quality-based subgroup found no hearing-loss effect in the high-quality studies; the two statements pull in opposite directions.

There was no beneficial effect of corticosteroid therapy in low‐income countries. Subgroup analysis for study quality showed no effect of corticosteroids on severe hearing loss in high‐quality studies.

The label itself states that corticosteroid trials in multiple sclerosis relapses speed recovery but do not change the eventual outcome or the course of the disease. (Source 3)

• Official position, Certainty not rated.
• Size: not stated on the label.
• Who: People with acute exacerbations of multiple sclerosis.
• How long: not stated.
• Result: No numbers are given; the label states the trials do not show an effect on ultimate outcome or natural history and that relatively high doses are needed for any significant effect.
• Funding: not applicable (regulatory label, this version dated 2026 on DailyMed)

Although controlled clinical trials have shown corticosteroids to be effective in speeding the resolution of acute exacerbations of multiple sclerosis, they do not show that they affect the ultimate outcome or natural history of the disease.

Where the evidence is mixed

No serious adverse events from glucocorticoids were reported in the trials newly added to the 2023 croup review. (Source 30)

• Systematic review, Low certainty.
• Size: 2 new trials with 1,323 children, within 45 trials and 5,888 children overall.
• Who: Children with croup.
• How long: trial follow-up.
• Result: No quantified harm; the review states no serious adverse events were reported in the newly included studies, while also noting adverse events were observed for some comparisons in the review as a whole.
• Funding: government 11%, academic 7%, industry 18%, foundations 9%; 55% of studies did not report funding.

Limit of this finding: Absence of reported serious adverse events in two trials is not evidence of safety; the same review states that adverse events were observed for some of its comparisons, without pooling them. Three further things a reader should know. This passage is from the review's plain-language summary rather than its scientific abstract, so it carries no effect estimates or certainty ratings; its phrase 'reduced return visits or (re)admissions by almost half' restates a risk ratio of 0.55 whose interval (0.28 to 1.11) includes no effect. The same summary says the standard 0.60 mg/kg dose probably reduced croup severity at 24 hours compared with 0.15 mg/kg, while the review's own conclusions say 0.15 mg/kg may be as effective as 0.60 mg/kg. And 'the addition of supplemental glucocorticoid favoured dexamethasone' means less extra steroid was needed with dexamethasone, not that adding one is recommended.

No serious adverse events from the use of the glucocorticoids were reported in the newly included studies.

On the review's own dose comparison, the standard 0.60 mg/kg dexamethasone beat 0.15 mg/kg on croup score at 24 hours in a single 72-child trial, and the two doses differed on nothing else. (Source 31)

• Systematic review, Moderate certainty.
• Size: 1 to 3 randomised trials per outcome, 72 to 949 children, within 45 trials and 5,888 children overall.
• Who: Children aged 0 to 18 years with croup.
• How long: croup score at 2, 6, 12 and 24 hours; return visits, length of stay and additional treatments over trial follow-up.
• Result: 24-hour croup score standardised mean difference 0.63 (95% CI 0.16 to 1.10; 1 RCT, 72 children; moderate certainty) favouring 0.60 mg/kg. No difference at two hours (SMD -0.27, -0.76 to 0.22; 2 RCTs, 861 children; high certainty), six hours (SMD -0.45, -1.26 to 0.35; 3 RCTs, 178 children; moderate certainty) or 12 hours (SMD -0.60, -4.39 to 3.19; 2 RCTs, 113 children; very low certainty). Return visits or (re)admissions risk ratio 0.91 (0.71 to 1.17; 3 RCTs, 949 children; high certainty); length of stay mean difference 0.12 (-0.32 to 0.56; 2 RCTs, 892 children). Epinephrine, intubation and supplemental glucocorticoid use did not differ either.
• Funding: Review-level: government 11%, academic 7%, industry 18%, foundations 9%, with 55% of studies not reporting funding.

Limit of this finding: The 24-hour advantage for the standard dose rests on one trial of 72 children, and it is the only timepoint at which the doses differ; the 12-hour interval (-4.39 to 3.19) is so wide the review grades it very low certainty. The review's own Authors' conclusions put the emphasis the other way, saying a smaller dose of 0.15 mg/kg may be as effective as the standard 0.60 mg/kg. Read this as no dependable difference between the two doses on any clinical outcome, with one small trial favouring the higher dose on a symptom score.

Compared to 0.15 mg/kg, 0.60 mg/kg dexamethasone probably reduced the severity of croup as assessed by the croup scoring scale at 24‐hour postbaseline score (SMD 0.63, 95% CI 0.16 to 1.10; 1 RCT, 72 children; moderate‐certainty evidence); however, this was not the case at two hours (SMD −0.27, 95% CI −0.76 to 0.22; 2 RCTs, 861 children; high‐certainty evidence).

Where the research disagrees

Whether adjunctive corticosteroids help in tuberculous meningitis when the person is HIV-positive

• Thwaites and colleagues, 2004, Randomised placebo-controlled trial in 545 patients with and without HIV, with an HIV-stratified relative risk of death of 0.78 (0.59 to 1.04): Adjunctive treatment with dexamethasone improves survival in patients over 14 years of age with tuberculous meningitis but probably does not prevent severe disability. (Source 2)
• Donovan and colleagues (ACT HIV), 2023, Randomised placebo-controlled trial restricted to 520 HIV-positive adults: Among HIV-positive adults with tuberculous meningitis, adjunctive dexamethasone, as compared with placebo, did not confer a benefit with respect to survival or any secondary end point. (Source 29)

Whether corticosteroids should be used in acute bacterial meningitis everywhere, or only in high-income settings

• Brouwer and colleagues, Cochrane 2015, overall analysis, Systematic review and meta-analysis of 25 randomised trials, 4,121 participants: There was insufficient evidence that corticosteroids caused a reduction in mortality overall (17.8% versus 19.9%; risk ratio (RR) 0.90, 95% confidence interval (CI) 0.80 to 1.01; P = 0.07) (Source 8)
• Brouwer and colleagues, Cochrane 2015, authors' conclusions, Same review's subgroup analyses by country income and study quality: Data support the use of corticosteroids in patients with bacterial meningitis in high‐income countries. We found no beneficial effect in low‐income countries. (Source 22)

How much

• Reference intake: Dosing is set by the prescriber, not by the reader. As a position, the generic oral tablet label states an initial range of 0.75 to 9 mg a day depending on the disease being treated, and insists in capitals that dosage requirements are variable and must be individualised on the basis of the disease under treatment and the response of the patient. The same section gives the initial range in pediatric patients as 0.02 to 0.3 mg/kg/day in three or four divided doses. (Source 19)
• Upper limit: The generic oral tablet label sets no single maximum dose. The ceiling it states is the top of the initial range, 9 mg a day depending on the disease being treated, and it separately states that in acute exacerbations of multiple sclerosis daily doses of 30 mg of dexamethasone for a week, followed by 4 mg to 12 mg every other day for one month, have been shown to be effective. (Source 19)
• Studied: RECOVERY gave oral or intravenous dexamethasone 6 mg once daily for up to 10 days to hospitalised COVID-19 patients. (Source 32)
• Studied: The surgical noninferiority trial gave 8 mg of intravenous dexamethasone while patients were under anaesthesia. (Source 33)
• Studied: The croup review compared 0.15 mg/kg with 0.60 mg/kg dexamethasone in children. (Source 34)

A common belief, and what the research shows

The belief: Dexamethasone was proven to treat COVID-19, so it helps anyone who catches it.

What the research shows: The trial that made dexamethasone standard care found benefit only where there was already respiratory failure. In RECOVERY, among people needing no respiratory support at randomisation, 28-day mortality was 17.8% vs. 14.0%; rate ratio, 1.19; 95% CI, 0.91 to 1.55) — a point estimate pointing the wrong way. The authors wrote that the results were consistent with possible harm in this subgroup. The benefit was also larger in people treated more than seven days after symptoms began.

Questions and answers

What is it?

Dexamethasone is a synthetic adrenocortical steroid — a laboratory-made version of the glucocorticoid hormones the adrenal glands produce. The label describes it as a white to practically white, odourless crystalline powder that is practically insoluble in water. It is formulated as tablets, oral solution, injection, eye preparations and implants. (Source 35)

What does it do in the body?

It is absorbed from the gut and acts on cells across the body, damping inflammation and modifying immune responses while also changing metabolism. At anti-inflammatory doses it has almost none of the salt-retaining effect of hydrocortisone. In myeloma cells it triggers programmed cell death, though the label says the mechanism there is not known. (Source 1)

Is it good or bad for you?

Both, depending entirely on the situation. In hospitalised COVID-19 patients on a ventilator, dexamethasone cut 28-day mortality from 41.4% to 29.3%. In the same trial, among patients needing no oxygen, mortality was 17.8% on dexamethasone against 14.0% on usual care — no benefit and results the authors called consistent with possible harm. The same drug, the same dose, opposite directions. (Source 6)

How do you get more of it?

Does not apply in the way it would for a nutrient. Dexamethasone is a prescription medicine, not something the body needs or stores, and dosing is set by a prescriber for a particular condition. As a position, the oral label's stated initial range is 0.75 to 9 mg a day depending on the disease treated, and the label stresses that requirements are variable and must be individualised. The same section states a far higher regimen for acute multiple sclerosis relapses (30 mg a day for a week, then 4 to 12 mg every other day for a month) and a separate weight-based range for children of 0.02 to 0.3 mg/kg a day, so the 0.75 to 9 mg figure is an adult starting range and not a ceiling. (Source 19)

If it is harmful, what reduces it?

When the harms outweigh the benefit, the answer in the literature is dose reduction or withdrawal rather than anything that removes the drug faster. The label instructs that treatment should be withdrawn gradually rather than abruptly after long-term use, because abrupt stopping can expose adrenal suppression. Cholestyramine is recorded as increasing corticosteroid clearance, and CYP3A4-inducing drugs speed metabolism, but neither is used as an antidote. (Source 19)

Why might someone be low in it or missing it?

There is no such thing as being low in dexamethasone, because it is a drug rather than a body constituent. The nearest real phenomenon is the reverse: after taking it, the body's own cortisol production can be suppressed, and a meta-analysis of 74 studies found this happens often after stopping, with no dose or duration that excludes it. (Source 16)

Which whole foods contain it or feed it?

No whole food contains dexamethasone; it is a synthetic steroid made in a laboratory. Food does affect how it is absorbed: a high-fat, high-calorie meal left total exposure unchanged but lowered the peak blood concentration by 23% after a 20 mg tablet, and that label states the tablet can be taken with or without food. (Source 14)

What happens if you do not have it?

Not taking dexamethasone has no consequence in itself. What the trials show is what is forgone in specific situations: without it, 28-day mortality among ventilated COVID-19 patients was 41.4% rather than 29.3%, and in low-resource hospitals neonatal death after threatened early preterm birth was 23.5% rather than 19.6%, a stated 25 women needing treatment to prevent one death. (Source 9)

How can you test for it?

There is no routine blood level test used to guide dexamethasone treatment. What is tested is its effects: blood glucose, because the label lists hyperglycaemia and unmasking of latent diabetes, and adrenal function after stopping. The adrenal-insufficiency meta-analysis concluded the threshold for testing corticosteroid users should be low, especially for people with vague symptoms after stopping; it also noted that no route or duration excludes the problem, so a normal result does not settle the question for all time. (Source 16)

References

1. DailyMed (U.S. National Library of Medicine). DEXAMETHASONE tablet — CLINICAL PHARMACOLOGY (Pangea Pharmaceuticals, LLC). 2026. Read the source
2. New England Journal of Medicine. Dexamethasone for the treatment of tuberculous meningitis in adolescents and adults. 2004. PMID 15496623, DOI 10.1056/NEJMoa040573. Read the source
3. DailyMed (U.S. National Library of Medicine). DEXAMETHASONE tablet — PRECAUTIONS, Neuro-Psychiatric (Pangea Pharmaceuticals, LLC). 2026. Read the source
4. DailyMed (U.S. National Library of Medicine). HEMADY (dexamethasone) tablet — 2 DOSAGE AND ADMINISTRATION (Edenbridge Pharmaceuticals LLC). 2026. Read the source
5. DailyMed (U.S. National Library of Medicine). DEXAMETHASONE tablet — INDICATIONS AND USAGE (Pangea Pharmaceuticals, LLC). 2026. Read the source
6. New England Journal of Medicine. Dexamethasone in Hospitalized Patients with Covid-19 — Preliminary Report. 2020. PMID 32678530, DOI 10.1056/NEJMoa2021436. Read the source
7. Cochrane Database of Systematic Reviews. Glucocorticoids for croup in children. 2023. PMID 36626194, DOI 10.1002/14651858.CD001955.pub5. Read the source
8. Cochrane Database of Systematic Reviews. Corticosteroids for acute bacterial meningitis. 2015. PMID 26362566, DOI 10.1002/14651858.CD004405.pub5. Read the source
9. New England Journal of Medicine. Antenatal Dexamethasone for Early Preterm Birth in Low-Resource Countries. 2020. PMID 33095526, DOI 10.1056/NEJMoa2022398. Read the source
10. New England Journal of Medicine. Dexamethasone and Surgical-Site Infection. 2021. PMID 33951362, DOI 10.1056/NEJMoa2028982. Read the source
11. Cochrane Database of Systematic Reviews. Calcium and vitamin D for corticosteroid-induced osteoporosis. 2000. PMID 10796394, DOI 10.1002/14651858.CD000952. Read the source
12. DailyMed (U.S. National Library of Medicine). DEXAMETHASONE tablet — PRECAUTIONS, Drug Interactions (Pangea Pharmaceuticals, LLC). 2026. Read the source
13. Microchemical Journal. Core shell stationary phase for a novel separation of some COVID-19 used drugs by UPLC-MS/MS Method: Study of grapefruit consumption impact on their pharmacokinetics in rats. 2022. PMID 35855210, DOI 10.1016/j.microc.2022.107769. Read the source
14. DailyMed (U.S. National Library of Medicine). HEMADY (dexamethasone) tablet — 12 CLINICAL PHARMACOLOGY (Edenbridge Pharmaceuticals LLC). 2026. Read the source
15. European Journal of Gastroenterology & Hepatology. Which patients taking non-aspirin non-steroidal anti-inflammatory drugs bleed? A case-control study. 1995. PMID 7614104. Read the source
16. Journal of Clinical Endocrinology and Metabolism. Adrenal Insufficiency in Corticosteroids Use: Systematic Review and Meta-Analysis. 2015. PMID 25844620, DOI 10.1210/jc.2015-1218. Read the source
17. DailyMed (U.S. National Library of Medicine). DEXAMETHASONE tablet — PRECAUTIONS, Endocrine (Pangea Pharmaceuticals, LLC). 2026. Read the source
18. DailyMed (U.S. National Library of Medicine). DEXAMETHASONE tablet — PRECAUTIONS, Information for Patients (Pangea Pharmaceuticals, LLC). 2026. Read the source
19. DailyMed (U.S. National Library of Medicine). DEXAMETHASONE tablet — DOSAGE AND ADMINISTRATION (Pangea Pharmaceuticals, LLC). 2026. Read the source
20. BMJ. Short term use of oral corticosteroids and related harms among adults in the United States: population based cohort study. 2017. PMID 28404617, DOI 10.1136/bmj.j1415. Read the source
21. BMJ. Short term use of oral corticosteroids and related harms among adults in the United States: population based cohort study. 2017. PMID 28404617, DOI 10.1136/bmj.j1415. Read the source
22. Cochrane Database of Systematic Reviews. Corticosteroids for acute bacterial meningitis. 2015. PMID 26362566, DOI 10.1002/14651858.CD004405.pub5. Read the source
23. DailyMed (U.S. National Library of Medicine). DEXAMETHASONE tablet — WARNINGS, Immunosuppression and Increased Risk of Infection (Pangea Pharmaceuticals, LLC). 2026. Read the source
24. DailyMed (U.S. National Library of Medicine). DEXAMETHASONE tablet — PRECAUTIONS, Musculoskeletal (Pangea Pharmaceuticals, LLC). 2026. Read the source
25. DailyMed (U.S. National Library of Medicine). DEXAMETHASONE tablet — PRECAUTIONS, Neuro-Psychiatric, acute myopathy and psychic derangements (Pangea Pharmaceuticals, LLC). 2026. Read the source
26. DailyMed (U.S. National Library of Medicine). DEXAMETHASONE tablet — ADVERSE REACTIONS (Pangea Pharmaceuticals, LLC). 2026. Read the source
27. New England Journal of Medicine. Dexamethasone in Hospitalized Patients with Covid-19 — Preliminary Report. 2020. PMID 32678530, DOI 10.1056/NEJMoa2021436. Read the source
28. New England Journal of Medicine. Dexamethasone in Hospitalized Patients with Covid-19 — Preliminary Report. 2020. PMID 32678530, DOI 10.1056/NEJMoa2021436. Read the source
29. New England Journal of Medicine. Adjunctive Dexamethasone for Tuberculous Meningitis in HIV-Positive Adults. 2023. PMID 37819954, DOI 10.1056/NEJMoa2216218. Read the source
30. Cochrane Database of Systematic Reviews. Glucocorticoids for croup in children — Plain Language Summary. 2023. PMID 36626194, DOI 10.1002/14651858.CD001955.pub5. Read the source
31. Cochrane Database of Systematic Reviews. Glucocorticoids for croup in children — Main results: Dexamethasone given in different doses. 2023. PMID 36626194, DOI 10.1002/14651858.CD001955.pub5. Read the source
32. New England Journal of Medicine. Dexamethasone in Hospitalized Patients with Covid-19 — Preliminary Report (Abstract: Methods). 2020. PMID 32678530, DOI 10.1056/NEJMoa2021436. Read the source
33. New England Journal of Medicine. Dexamethasone and Surgical-Site Infection — Methods. 2021. PMID 33951362, DOI 10.1056/NEJMoa2028982. Read the source
34. Cochrane Database of Systematic Reviews. Glucocorticoids for croup in children. 2023. PMID 36626194, DOI 10.1002/14651858.CD001955.pub5. Read the source
35. DailyMed (U.S. National Library of Medicine). DEXAMETHASONE tablet — DESCRIPTION (Pangea Pharmaceuticals, LLC). 2026. Read the source