Sedation Administration Timing: Intermittent Dosing Reduces Time to Extubation

Background:

Sedation is a significant component in the comprehensive treatment of a critically ill patient. The presence of anxiety and pain can often be left unexpressed, particularly in patients on mechanical ventilation. Opioids are primarily used for the treatment of acute pain and sedative agents (i.e. benzodiazepines, propofol) are used to achieve sedation. Sedation is required in mechanically ventilated patients, yet the optimal strategy for its management is unclear. Daily interruptions of continuous intravenous (IV) sedation (i.e. sedation holiday) have shown to result in a shorter duration of time on mechanical ventilation and intensive care unit (ICU) stay. However, sedative agents have the potential to accumulate despite daily interruptions, leading to drug redistribution from the tissues into the circulation. This may delay mental status recovery and cause abrupt agitation despite decreased or discontinued sedatives. Therefore, daily sedation holidays do not always result in awakened and communicative patients. Optimal management for sedation is necessary to avoid accumulation of drug and prevent sedation-induced delirium.

According to the guidelines on the sustained use of sedatives and analgesics in the critically ill adult, published by the Society of Critical Care Medicine, maintenance of sedation with intermittent or "as needed" doses of diazepam, lorazepam, or midazolam may be adequate to accomplish the goal of sedation. Continuous infusions must be used cautiously due to accumulation of the parent drug and inadvertent over-sedation.

There are two main clinical trials that studied the use of continuous sedation compared to intermittent sedation in the critical care setting. In one study, a prospective, randomized, multicenter, open-labeled trial compared intermittent IV lorazepam with continuous infusion midazolam. The investigators concluded that intermittent lorazepam and continuous midazolam had similar levels of sedation and anxiolysis as measured by objective hemodynamic and laboratory data, along with patient and nurse scoring systems. However, it was determined that higher doses of continuous infusion midazolam were required to obtain adequate sedation and anxiolysis. In another prospective, cohort study, mechanically ventilated patients received either continuous or intermittent doses of sedation using a targeted Ramsay Scale of 3. Patients in the intermittent group required less hours on mechanical ventilation compared to the continuous IV sedation group, p<0.001. The authors of this trial concluded that continuous sedation may be associated with prolongation of mechanical ventilation. These two clinical trials conclude that the use of intermittent sedation may provide additional benefit to sedated patients in decreasing the days on mechanical ventilation and cumulative doses of sedative agents.

Using these studies as a starting point, the investigators conducted a prospective pilot study demonstrating that the use of intermittent sedation in medical intensive care unit patients significantly decreased amount of sedatives used, the hospital length of stay, and most importantly reduced duration of mechanical ventilation. In this study it was shown that intermittent sedation administered within twenty-four hours of mechanical ventilation reduced ventilation time by twenty-four hours (p<0.001). However, much like the studies that formed the basis of this pilot, this study suffered from some lack of control as well as small sample size. More importantly, this study analyzed only medically critically ill patients in whom large doses of narcotics are often not routinely employed. In surgical patients narcotics are often used concomitantly with a sedative in order to achieve anxiolysis and pain control while ventilated.

The investigators have already demonstrated in a small cohort that the use of intermittent sedation protocol is not only non-inferior but significantly reduces mechanical ventilation time in medical intensive care patients. The investigators hypothesize that an intermittent administration of both sedative and narcotic medications will reduce mechanical ventilation time in multiple critically ill patient populations requiring ventilation.

Purpose:

This prospective, randomized cohort study is designed to compare continuous infusions to intermittent boluses of sedative agents (i.e. opioids, benzodiazepines) and narcotic agents (i.e. fentanyl, dilaudid) for sedation and pain control in mechanically ventilated patients in the surgical trauma unit and neurologic intensive care unit (STU, NCC).

The primary end point measured will be the number of hours on mechanical ventilation. Secondary end points will include length of ICU stay, length of overall hospital stay, cumulative amount of sedative agents used per day, and time within target sedation levels.

Methods:

At the investigators institution, there are two critical care units that serve the majority of surgical, traumatic, and neurovascular critically ill patients (The surgical trauma unit and neurologic intensive care unit). These two units work in tandem and their patient populations are distributed throughout both units (hereafter referred to as the ICU). Patients requiring mechanical ventilation will be randomized into a control arm or trial arm of the study. The following demographic data will be collected: patient age, gender, race, indication for mechanical ventilation, APACHE II score, injury severity score (if patient is a trauma patient), and serum creatinine. All data will be recorded without patient identifiers to maintain confidentiality. Specific methods are detailed below.

Study Design: The S.A.T.I.R.E. trial will be a two year, randomized control trial performed on intubated patients in the ICU (surgical and neurovascular patients). The control group will consist of intubated patients that will be sedated and given pain control with using current practice: continuous infusions of a narcotic and sedative. The trial group will consist of intubated patients that will be sedated and given pain control with an investigational timing of dosing: Q1 hour bolus dosing given in a sliding scale fashion. The target levels of sedation and pain control will not differ between the two arms of the study. Upon admission to the ICU the surgical resident on call (hereafter referred to as ICU resident) is required to assess a patient. During that time, if the patient is intubated and does not fall into exclusion criteria, the ICU resident will then assign the patient to the control arm of the study or the trial arm of the study randomly as described below. The ICU residents will use designated order sets with specific protocols for each arm of the study which are described in more detail below. Each day during the week the clinical pharmacist on staff will record patient demographics (age, gender, race, indication for mechanical ventilation, serum creatinine, APACHE II, injury severity score if trauma) and MRN#, as well as time of intubation and time of extubation if applicable. During the weekend the ICU resident will record new patients enrolled into the study (MRN only); on Monday, the investigators or clinical pharmacist will review the weekend data and retrospectively record demographics, time of intubation, and time of extubation from patients admitted to the ICU over the weekend. This Monday data collection will be assigned on a rotating weekly schedule to the investigators. Data will be collected for approximately two years; aggregate data will be analyzed every six months.

Randomization: Patients will be randomized upon admission to the ICU based on birth year. Patients with a birth year ending in an even number will be randomized to the control arm and patients with a birth year ending in an odd number will be randomized to the trial arm. This randomization will occur during placement of initial sedation/pain medication orders and will be done via an order-set with two initial options (even birth year vs. odd birth year) which will open up to have sub-orders containing the medications dosed in either a continuous or intermittent fashion, respectively.

Study Medications: The control and trial arms will use the same medications. These will be ordered using a specified order set detailed below. Study participants will be evaluated daily by the resident, surgical critical care attending, and clinical pharmacist (weekdays only for clinical pharmacy) for appropriateness and adjustment for patient specific requirements.

If the patient's birth year ends in an even number, the order set will open the control arm. Here, the ICU resident will be able to order a midazolam continuous infusion using existing orders, and a fentanyl infusion. The midazolam infusion order is standardly dosed at 1 mg/hr and titrated to RASS -1. This order can be found under "MIDAZOLAM Infusion 50 mg/50 mL 0.9% sodium chloride". Fentanyl is standardly dosed at 25 mcg/hr and will be titrated to the CPOT for pain control per nursing protocol. This order can be found under "Fentanyl Infusion Order Set". This differs slightly from current practice with the addition of the CPOT scale in that previously nursing had titrated both sedative and narcotic to RASS and now fentanyl will be titrated specifically to pain; however, this is a change being adopted hospital wide and will be the current practice for the institution at the time of the study or shortly after launch of the study. Additionally, current practice is for nursing to administer infusions and titrate to RASS at their discretion by either increasing/decreasing the rate, bolus/pushing additional medication, or both. To standardize tracking of medication doses an as needed order has also been created to be used in conjunction with the infusion order to support this practice and accurately capture medication administered in the control arm. Patients that are admitted to the ICU on a continuous propofol infusion will be converted to the midazolam infusion or intermittent protocol within twenty-four hours or be excluded from the study. During weekdays this will be performed by the clinical pharmacist; during weekends, this will be performed by the ICU resident. If a patient must be taken off a midazolam protocol they will be excluded from the study.

If the patient's birth year ends in an odd number, the order set will open the trial arm. Here, the ICU resident will be able to order midazolam and fentanyl in a sliding scale fashion. Doses will be started at the same doses as the infusion medications; for midazolam 1 mg/hr, for fentanyl 25 mcg/hr and titrated to target RASS -1 or CPOT = 0. Dosing may be modified for creatinine clearance, hemodynamics, and opiate naivety.

Nursing Protocol: Nursing protocol for titrating sedation to RASS and pain to CPOT have been adopted institutionally. Patient RASS levels and CPOT levels are checked hourly.

Statistical Analysis: Demographics will be reported as aggregate means; if a larger than anticipated sample size is achieved we will also perform sub-analysis of individual groups (i.e., trauma vs. general surgical patients vs. neurovascular patients, etc). End points will be reported in means with a 95% confidence interval and will undergo standard significance testing (using α and β that are 0.05 and 0.20, respectively.) Continued Inclusion: Each patient will be assessed on a daily basis by a surgical critical care attending, resident, and pharmacist to determine appropriateness of his/her sedation regimen. If a patient continues to require significant additional sedative or narcotic outside the pre-established protocol, the patient may be withdrawn from the study and switched to a continuous infusion at the discretion of the attending physician in conjunction with the clinical pharmacist. In addition, a 24-hour on-call support for nurses and residents will be available to provide optimal, intermittent doses. At any time the attending critical care physician may remove a patient from the study at their discretion for any reason should it be required.