Introduction
Sedation and analgesia are common practice in ICU. However, evolutions
in concepts and practices have motivated the French Intensive
Care and Anesthesia Societies to launch prospective work in order
to propose to their members a set of clinical guidelines in this
field.
This work was completed at a previous "consensus conference"
managed by the same societies, which needed to be updated, despite
being held in 1993. Methodology used was that recommended by the
Agency for Health Care Policy and Research and in France by the
Agence Nationale pour l'Accréditation et l'Evaluation de
la Santé. Work was conducted by a group of 14 experts who
gave their conclusions in January 99. We summarize them here.
What were the questions to answer ?
Organization committee addressed several basic questions to the
experts which were :
a) What are the goals of sedation?
b) How to evaluate patient needs?
c) What are the means?
d) How to evaluate the effects of the treatments? e) What strategy
and feasibility?
These questions had been addressed in a close form during the previous consensus conference but had led to very "weak" answers with few practical recommendations.
Guidelines
The goals of sedation
No scientific evaluation of French practice was available in the
literature in this field. Historical analysis of foreign literature
shows an homogenization of practices among clinicians. They include
synergistic use of drugs. The conclusions of the expert conference
held by the American College of Critical Care Medicine were mentioned:
1) analgesia shall be based on morphine sulfate. Hydromorphone
is an alternative to morphine. Fentanyl is preferred in terms
of hemodynamic stability.
2) Propofol and midazolam are the drugs of choice for short term
sedation, whereas iv lorazepam is the one to prefer for long-term
use. Noteworthy, this drug is not available in France. Also to
be mentioned, the price of midazolam is much less in France than
in the USA.
3) Pancuronium is the first step drug for paralysis except in
case of instability or when tachycardia is not acceptable.
4) Monitoring of paralysis is made by observations of any residual
respiratory activity or by twitch ("train of four").
The importance and evidence of the disturbances induced by intensive care have been reassessed. They advocate for a careful evaluation of patients at risk for pain, anxiety and agitation which should take advantage of sedation and analgesia. Situations at risk are highlighted : extensive nursing procedures and physiotherapy, all kinds of puncture and catheter placement, tracheal intubation, wounds care when large, tracheal suction, endoscopic procedures, invasive radiologic procedures, transport of ventilated patients, cardioversion, tracheotomy.
Several clinical conditions have been highlighted for sedation-analgesia: ARDS to allow ventilation without strain or mechanical stress to patient; ventilated asthma; severe head trauma and other conditions with cerebral hypertension; Guillain Barré and tetanus ; postoperative major pain in intensive care where analgesia is the main goal; delirium tremens; burned patients, especially during nursing cares.
Evaluation of needs
Unidimensional scales shall be used in ICU, whenever its possible.
Despite being poorly evaluated, the most relevant sedation scale
is that by Ramsay. One should aim a level of 2-3 and reach 4-6
only in specific cases (cerebral edema, ARDS, invasive procedures).
The means
Non pharmaceutical means are not considered to be efficient enough
to be used alone. Optimization of physical nursing aspects and
psychological attention are essential.
Drugs have been extensively discussed. Pharmacology gives rationale to define best choices. Especially situations at risk have been reassessed: hypovolemia, unstable hemodynamics, depletion in protein binding molecules (albumin, a1 AGP), renal and hepatic insufficiencies, role of CYP450, sepsis and inflammation, drug interaction, and limits of models in the condition of long-term use.
Basic and contextual pharmacology of drugs has also been reported in order to explicate practical choices retained in these guidelines.
Briefly, the benzodiazepine family induce accumulation with variable effects among patients. Midazolam is preferred for a lesser risk accumulation which is not negligible in long-term use. Adaptation of doses is mandatory in both extremes of life. Lorazepam cannot be included in practice as it is still not available in France. Propofol seems to be much simpler to use as reproducibility of effects is good and accumulation minimal. However, lipidic emulsion and price limit its use for long periods. Thiopental is restricted to states of cerebral edema. Etomidate, ketamine and inhalational agents are not suitable. Neuroleptics represent an alternative to benzodiazepines in case of predominant agitation. Morphine is the cheapest central analgesic with few drawbacks. It is recommended as the first level drug, like in American guidelines. Fentanyl is an alternative. Sufentanyl is pharmacologically acceptable but expensive. Data are missing for this use. Clonidine is an adjuvant of interest as it has analgesic and sedative effects. Regional anesthesia shall be considered each time it is possible.
The guidelines in practice
Short-term sedation (<24h) can be based on either propofol
(3-6 mg.h-1) or midazolam (1-3 mg.h-1). Cost of propofol justifies
its use only when extubation is probable in a short delay. It
is not suitable for children <15 year old. Morphine is the
drug of choice, and fentanyl is another issue. Renal insufficiency
requires that regimen should be adapted.
Long-term sedation is based on midazolam with specific attention on regimen in renal insufficiency. Oral medication can be an alternative: lorazepam (3-15 mg/24h), hydroxizine (50-100 mg/24h). Agitation with psychiatric symptoms is an indication for haloperidol iv or oral. Analgesia is also based on morphine or fentanyl. Ketamine is interesting in case of bronchospasm. Paralysis shall be considered only when needed, under close monitoring. Pancuronium is not deprived from drawbacks and vecuronium is probably better. In case of hepatic or renal insufficiency, atracurium and cisatracurium are preferred.
Withdrawal is not codified. Withdrawal syndrome to opiates is accessible to clonidine. When benzodiazepines are involved, either reintroduction of small doses and gradual withdrawal or neuroleptics can be used.
Conclusion
These guidelines converge to what has been edited by North-American
societies. The major differences can be explained by differences
in costs (propofol and midazolam) and availability (lorazepam)
compared to the USA. Both stress the need to monitor sedation
and paralysis and to carefully consider pharmacological specificity
of each patient to adapt treatment to its condition.