CORTICOSTEROIDS IN THE ICU:D. P. Laporta MD
the whole truth Pulmonary and Critical Care
June 1997 Sir MBD - Jewish General Hospital, McGill University
PROMontreal, Quebec
REFERENCES- REVIEWS
Steroids and biologic agents in the ICU. RP Allen, Crit Care Clinics, July 1991:695-712
Adrenal crisis. R Chin, Critical Care Clinics January 1991:23-42
The utility of parenteral glucocorticoids in the emergency department. Hoang KD , Pollack CV Jr, J Emerg Med 1994; 12:507-19
OBJECTIVE
To establish the evidence to justify the use of systemic corticosteroids (SCS) in specific clinical situations in the ICU. Specific recommendations are made from a critical review of the literature.
METHODS
Clinical Experience
SMMBD-JGH ICU Database
ICU Textbooks (Shoemaker, Civetta, Rippe)
Evaluation of Published Literature1: Human, Clinical Trials, Reviews (Systematic, Narrative)
Not Used: Animal / Physiologic Studies
1 EVALUATION OF PUBLISHED STUDIES ON SCS IN THE ICU
1. Internal validityRandomized, blinded.control group, similar at outset, treated similarly,
inclusion/exclusion criteria
2. Results:treatment effect (how large, incl burden (side effects)), outcome measures, followup, comorbid illness
3. Generalizability (external validity)
4. Evaluation of literature (grading2)
5. Recommendations3
2 GRADING of clinical studies3 RECOMMENDATION
Ievidence from 1 RCT, low FP & FN GOOD evidence to support
IIevidence from 1 RCT, high FP FN MODERATE evidence to support
IIInon-randomized cohort comparisons POOR evidence to support
IVnon-randomized historical cohort comparisons
Vcase-series (no controls)
*specific clause if high-quality overview (systematic review)
1
SCS IN THE ICU: CLINICAL INDICATIONS
GRADERECOMMENDATIONUSEDONT
Adrenal cortical insufficiencyVPOOR
Asthma I GOOD
COPD II-III MOD-POOR
Sepsis/septic shockI GOOD
Acute bacterial meningitis II MOD (adult)
Acute hepatitisviral I GOOD
ETOHII MOD
Croup I GOOD
EpiglottitisIII POOR
Post-extubation stridor1I-IIGOOD-MOD
AIDS-PCPI-II GOOD-MOD
Spinal cord injuryIIMOD
Head injury I GOOD
ARDSpre (FES) II MOD
early I GOOD
late V POOR
1in high-risk, PEDS
1
NOT COVERED
Peds: Antenatal, Bronchopulmonary Dysplasia
Oncology: brain/spine mets
Collagen/vascular (SLE pneumonitis, cerebritis, renal failure; vasculitides)
Hemato: TTP, AIHA, AI-Tpenia,
Dermatologic
GI: inflammatory bowel disease
Infections: typhoid, tetanus, selected TB issues
Misc: Hypercalcemia, Myaesthania, Transplant, anaphylaxis, radioiodine contrast agents
INDICATIONS FOR SYSTEMIC GLUCOCORTICOIDS
Unfortunately, evidence comes in shades of gray (the EBM Working Group JAMA 1993)
ADRENAL CORTICAL INSUFFICIENCY (ACI), or ADRENAL CRISIS
pathophysiology: adrenal hypoperfusion, cortical necrosis, thrombosis, hemorrhage
problem: - making the diagnosis: difficult in the ICU pt : no gold standard
- concept of relative, rather than absolute ACI
cortisol levels in the ICU (Drucker D CCM 85, Schein CCM 90, Bouachour IntCareMed95)
cg/dl X 27.59 nmol/l (SI)
surgical stress> 20 > 552
ACI:- no shock< 10 < 276
- shock 1.< 20< 552
2.wACTH <7 or value<20 <193 or value<552
prevalence (1+2): 0-37% ...ie RARE
* no predictive value (baseline or ACTH-stimulated)
* often higher in non-survivors, liver disease
how frequent is hypotension in steroid-treated patients who undergo acute stress without SCS supplementation ? The few studies available suggest that hypotension ...is uncommon...adrenal steroids can and should be administered, but other contributory causes for the hypotension should be sought Udelsman J Clin Endocrinol Metab 87
RISK FACTORS FOR ACI
1
- previous SCS1
- septicemia
Waterhouse-Friderichsen
Gm -ive (Pseudomonas)
- shock
- meds:anticoagulants
ketoconazole
dilantin, rifampin, phenobarb
- coagulopathy
- thrombosis
- postoperative
- malnutrition
- chronic adrenal disease
autoimmune (Schmidts syndrome)
TB & other granulomatous
AIDS
metastatic disease
amyloidosis
- hypothal/pituitary disease
1
1 In the last year. >25mg/d X 5 days: blunted ACTH response, may last X months
1
SYMPTOMS/SIGNS OF ACI
- flank/abdominal pain (adrenal hemorrhage)
- nausea/vomiting
- SIRS
- BP (poor response to b agonists)
- encephalopathy
LABS
Hb (sudden) (adrenal hemorrhage)
K
Na
glucose
1
ASTHMA
1956 BMRC: SCS are efficacious.
1992 Meta-analysis (Howe BH Am J Emerg Med )* NB: po = iv
30 RCTs( Odds Ratio for Admission.47, for Relapse .15)
1993 Systematic review (McFadden ER Jr, ARRD)? lowest effective dose...
mg/kg/day hydrocortisone-equivalent EffectConclusion
4niltoo low
10-15(ie 40 mg MP q6h)goodsufficient
250no differentno better
...does not support the concept that very large doses of [SCS] are more
efficacious than smaller ones ER McFadden, Jr
1995 Corbridge TC and Hall JBFurther studies are needed to establish the best dose and dosing frequency of [SCS] in status asthmaticus.
SYSTEMIC CORTICOSTEROIDS IN ASTHMA - EFFECT OF WEIGHT
mg Q6H MP / mg/day MP / mg/day OH-CORTISONE / mg/kg/d OH -CORTISONEWT (kg) / 40 / 50 / 60 / 70 / 80
40 / 160 / 900 / 22 / 18 / 15 / 13 / 11
60 / 240 / 1200 / 30 / 24 / 20 / 17 / 15
125 / 500 / 2500 / 62 / 50 / 42 / 36 / 31
Note the greater than 5-fold variation in dose over usually-suggested dosages !
COPD: The scripture according to Albert
1978Review (Sahn SA, Chest 78)
Conflicting results, burden of evidence disfavors SCS (of the 6 positive studies,
1controlled, none double-blind; of 11 negative studies, 8 controlled, 1 double-blind)
2 RCTs
* Albert RK, (AIM 80)MP .5mg/kg q6h
more rapid improvement in FEV1, FVC after 12 hrs, continues to 72hrs
Chest 87 91:289-90: letter to editor re critique of study
(stats, no clinically relevant outcome measures)
* In Emergency Dept (Emerman CL Chest 89) no difference at 5 hours
1
COPD (Continued)
Physiologic intervention (Rubini AJRCCM 94) measurement at 90 minutes
Civetta: We routinely use similar doses...but...for more than 3 days.
After...improvement,...switch to a tapering course of prednisone over several weeks.
SEPSIS/SEPTIC SHOCK
2 Meta-Analyses (10 RCTs, >1000 pts)
Lefering (CCM 95)
No effect on gi bleed, 2ndary infections, hyperglycemia)
Cronin (CCM 95)
Trend to increase in 2ndary infections
Current RCT (Annane D): An unexplored side
SCS attenuates NOS at physiologic (replacement) doses in severe sepsis.
ACUTE BACTERIAL MENINGITIS (ABM):
Animal studies: decrease mortality
PEDS: Meta-analysis Havens PL (Am J Dis Child 1989)
RCTs : DEX .6 mg/kg/d D1-4 given early with Abx, age 2yrs
no change in mortality, improvement in hearing loss
not recommended routinely
May be useful (H Flu, S Pneumoniae) for hearing
ADULTS:4 Narrative reviews (Townsend, Harvey, Lauritsen , Berkowitz)
Systematic review (Prasad J Neur, Neurosurg & Psych 95)
7 RCTs: DEX in ABMproblematic methodology
:
Use is unjustified
ALCOHOLIC HEPATITIS (AH)
5 Meta-analyses (89 - 91) outcome: hospital mortality
- Imperiale (Ann Int Med 90) 10 RCTs
results: protective efficacy of SCS 37% overall, provided active gi bleed excluded, and PSE present
- Christensen E (Gut 95) 12 RCTs, adjusted for confounding variables
conclusion: previous meta-analyses not valid (major flaws in RCTs)
no effect (even PSE). Interaction with gender
NB: ACUTE VIRAL HEPATITIS (FHF): 5 RTCs: contraindicated
1
CROUP
commonest cause of acute upper airway obstructionn in children.
1989 Suggested but unproven
1 Meta-analysis (Yates Drug Safety 97)
Mild- mod2mg budesonide neb = DEX .6mg/kg IM
Severe1mg/kg prednisolone q12 h*
* varicella occurence
EPIGLOTTITIS
no RCT.
1 Grade III: 50% LOS with observation + SCS (irrespective of whether intubated)
5 Narrative reviews. Used extensively empirically.
Widely recommended (may decrease inflammatory edema & contribute to recovery,
and avert trach if angioneurotic edema)
POST-EXTUBATION CROUP/STRIDOR
animals: efficacious
case series: Id
5 RCTsPEDS: Louser RJ (J Peds 92), Tellez DW (J Peds 91),Anene O (CCM 96)
ADULTS: Darmon JY (Anesth 93), Ho LI (Int Care Med 96)
CONCLUSIONS (ADULT & PEDS):
- Unwarranted in routinely ventilated pts
- Effective in selected (high-risk) patients
Female
tracheal trauma
reintubation
prolonged intubation
hemodynamic instability
young age (in PICU population)
AIDS-PCP
3 RCT (Montaner JS AIM 90, Gagnon S NEJM 90, Bozzette SA NEJM 90)
...Moderate-severe ( Pao2 (RA) >75, Pao2/Fio2 <350)
Outcome: respiratory failure (AOR=5.87), hospital mortality
effective
NIH Consensus (NEJM 90) and letter (NIAID 90):
...in mod-severe PCP, give steroids
Day 1-580mg/d (Severe PCP 320mg/d)
Day 6-1040mg/d
Day 11-2120mg/dNB: give early, taper slowly ie 1month (flareup)
Textbooks (Rippe, Civetta): as per NIH consensus
Authors unpublished observations:Less septra rash, butmore long-term
secondary opportunistic infections (CMV, MAI, aspergillus, mucor)
1
SPINAL CORD INJURY (SCI)hrs
preclinical studies: positive
Narrative reviews (Ducker Spine 94, Nguyen Adv Surg 96)
NASCIS Ilow-dose MP (100mg/d X 11) vs high-dose MP (1000mg/d X 11)NS
NASCIS II 3 groups: 1) MP 30mg/kg 5 mg/kg/hr X23
2) Narcan
3) Placebo
F/U: 6wks, 6 mo, 1 yr
Result: effective
Critique:different outcome of the 2 placebo groups
Recommendation: despite controversy and unresolved issues, advocate...
initiation of steroids ASAP after acute SCI,but not beyond the first 8 hours.
There is too much data available to arrive at any other conclusion
HEAD INJURY
2 RCTs in J Neurosurg (1979 Cooper PR, 1985 Brackman R)
DEX 16mg/kg/d, or 96mg/d or Placebo X 6 daysF/U 1, 6 months
DEX 100 mg/d (within 6hrs)
Results: No effect
Despite this...
2 Surveys of ICU practice in ICU/Trauma centers: SCS in head injury
UK (Jeevaratnam BMJ 96)SCS used in 19/39 ICUs
USA (Ghajar CCM 95)SCS used in >50% of the time in 64% centers
NB: SCS in head injury: increased infection rate (De Maria Ann Surg 85)
ARDS
Pre ARDS (Outcome: the occurrence of ARDS)
2 uncontrolled trials (Weigelt Arch Surg 85, Sprung NEJM 84)
Ineffective
2 RCT in pts at risk for Fatty Embolism Syndrome
(Schonfeld AIM 83, Lindeque BGP JBBJSS 87)
Effective
1
Early ARDS
2 RCTs (Bernard NEJM 87, Bone Chest 87) 30mg/kg MP q6 X 24hNS
saw natural course of disease
only large enough group: sepsis ...Ineffective
Late (fibroproliferative) ARDS
Grade V evidence (5 studies, 55 pts)
Physiological evidence (mediators/outcome) ... positive, ? promising
NOT DISCUSSED:
MYAESTHENIA GRAVISGrade V (case series -no control) (Arsura E Arch Neurol '85)
Plasmapheresis may offset initial steroid-related deterioration, enabling more rapid institution of therapy Rippe
...pulse ...[medrol]...may also benefit...appears to produce ledd inital worsening, and more rapid immprovement than conventional doses of prednisone. Further studies...should be awaited before widespread use is adopted in myasthenic crises Civetta
ANTENATAL- NIH ConsensusJUSTIFIED in fetuses 24-34 wks X 24-48 hrs
- 2 Meta-analyses (Crowley PA and Sinclair JC Am J OBGYN 95, )
Decrease risk of RDS, IVH, NEC. No strong evidence of increased infection
SEVERE HYPERCALCEMIA - Narrative review (CCClinics Jan 91: Endocrine crises)
useful if: 1) vitamin D intoxication/sarcoidosis, or 2) lymphoproliferative disease
not useful if solid tumors, hyper-PTH
- Text: useful if life-threatening, paraneoplastic (lymphoproliferative, breast)
CONCLUSION: CORTICOSTEROIDS IN THE ICU:
Documentation of rational therapeutic regimens does not exist for most disease states, due to lack of evidence... this has given rise to a variety of approaches that define timing, duration, quantity, tapering schedule Allen RP, CCC 91
Corticosteroids are powerful tools, their potential benefits and risks are known, but their precise indications and contraindications are not clear
1) give the lowest effective dose
2)for the shortest effective time.
3) treat the underlying disease vigorously
4) think physiologically
The ART of medicine: practicing with much evidence but few thresholds
1