| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Special Report |
1 From the Departments of Radiology (S.J.S., S.M., J.R., G.T., D.G., J.A., S.S., G.W.S.) and Anaesthesia (N.B.), McMaster University Medical Centre, Hamilton, Ontario, Canada; and St Vincent’s University Hospital, Dublin, Ireland (D.E.M.). From the 1998 RSNA scientific assembly. Received November 29, 1998; revision requested January 14, 1999; final revision received February 3, 2000; accepted February 22. Address correspondence to D.E.M. (e-mail: d.malone@st-vincents.ie).
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
MATERIALS AND METHODS: Ninety-one adult patients underwent 113 abdominal interventional procedures. Fentanyl citrate and midazolam hydrochloride were administered in one to five steps (A, B, C, D, E) until the patient was drowsy and tranquil at the effective loading dose (ELD). Doses per step were as follows: A, fentanyl 1 µg per kilogram of body weight; B, midazolam 0.010–0.035 mg/kg; C, repeat dose in A; D, repeat half the dose in B; and E, midazolam 1–2-mg boluses (maximum, 0.15 mg/kg).
RESULTS: The ELD was reached in no procedure after step A, in 70 after B, in 23 after C, and in 18 after D. Step E was needed in two procedures. The mean maximum pain score (scale of 0 to 10) was 3.4; pain scores in 85 (75%) procedures were 4 or less (discomforting). Severe pain occurred in seven (6%) procedures. Hypoxia (oxygen saturation < 90%) occurred in 11 (22%) procedures performed in patients breathing room air and four (6%) performed in those breathing supplemental oxygen (P = .04). All patients responded to supplemental oxygen.
CONCLUSION: This stepwise "ABCDE protocol" allows safe and effective sedation of patients. It is easy to use and may be useful in training radiology residents, staff, and nurses in the techniques of sedation and analgesia. Supplemental oxygen should be used routinely.
Index terms: Abdomen, interventional procedures • Anesthesia • Special Reports
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
In this article, we describe a simple and safe protocol for sedation and analgesia with fentanyl citrate and midazolam hydrochloride that is based on the patient’s body weight. This protocol was developed by an anesthesiologist (N.B.) and a radiologist (D.E.M.) for use by radiologists who perform procedures in a busy radiology department in patients who have varying degrees of preanesthetic risk. The term "ABCDE protocol" was coined to describe the five steps of drug administration in the protocol. The protocol was developed in an attempt to optimize and standardize existing practice, which varied substantially among radiologists before the introduction of the protocol. The purpose of this study was to evaluate the effectiveness and safety of this standardization of practice.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
These procedures were classified as minor, intermediate, or major to reflect the anticipated level of pain with each procedure (Table 1). In essence, minor procedures were those that involved simple needle insertion (eg, biopsy) or tube manipulation without dilation. Intermediate procedures typically involved those in which a needle, guide wire, dilator, and tube were used in relatively superficial organs (eg, abscess drainage). Major procedures were often prolonged and involved deep visceral stimulation close to the celiac plexus (eg, biliary drainage).
|
Patients in this study ate no solid food from midnight before the procedure. They were allowed to consume clear fluids up to 4 hours before the procedure, and they fasted for 4 hours prior to the procedure. Patients provided written informed consent for each procedure. The chairman of our institutional review board indicated that approval was not required for this study, as this procedure was simply a standardization of existing practice. No medication was administered before the procedure. No anesthesiologists were present during the procedures. At our hospital, there are no considerations regarding privilege to preclude radiologists from performing this procedure.
The ABCDE Protocol: Design and Usage
The desired end point of sedation is drowsiness, tranquility and reduced sensitivity to pain. The right level of sedation has been achieved when the patient rests with their eyes closed but responds to verbal commands or mild physical stimulation (2). The classical sign of drooping of the eyelid represents unnecessarily deep sedation. The best end point depends on the body language of the patient and their subjective feeling of relaxation and loss of anxiety (3). The effective loading dose (ELD) is defined as the dose required to reach this level of sedation.
Fentanyl and midazolam may be the ideal drugs for use in many interventional procedures because they have rapid clearance rates and short elimination half-lives (2). Since they act synergistically, it is unwise to administer both simultaneously (2,4,5). In one study (5), fentanyl administered in a dose of 2 µg/kg produced hypoxia (oxygen saturation < 90%) in six of 12 healthy volunteers. The addition of midazolam at a dose of 0.05 mg/kg increased the incidence of hypoxia to 11 (92%) volunteers. There is great interpatient variation in their susceptibility to these drugs, and the concept of stepwise administration of drugs with a calculated loading dose was developed (by D.E.M. and N.B.) as a logical way to avoid hypoxia in more susceptible patients.
Step A of the ABCDE protocol involved the administration of 1 µg/kg fentanyl. This was based on half of the dose given to subjects in the study by Bailey et al (5). The time between injection and clinical effect is the lead time for that patient. With fentanyl, most patients report feeling light-headed, more relaxed, or simply different. They may also notice facial pruritus. The initial dose of midazolam tends to cause further relaxation and more obvious drowsiness. Later in the procedure, knowledge of the lead time is useful in timing the administration of subsequent doses.
Midazolam was administered in step B by using drug synergy to achieve anxiolysis. After step A, the nurse administered a dose of midazolam that varied from 0.010–0.035 mg/kg, depending on the age and clinical condition of the patient. For example, a patient older than 65 years (particularly one with coexistent cardiac or respiratory illness) would receive 0.010 mg/kg, and a young healthy adult, 0.035 mg/kg. If the initial dose in step B was less than the maximum dose (0.035 mg/kg) for that step, then step B was continued in divided doses, if necessary, until the patient was adequately sedated (ELD, therefore, reached) or until 0.035 mg/kg of midazolam had been administered.
Step C involved an injection of 1 µg/kg fentanyl to reach the calculated loading dose of 2 µg/kg. If the ELD still had not been reached after step C, the patient underwent step D, which was the administration of half of the dose of midazolam administered in step B. This approximated 0.015 mg/kg of midazolam, which brought the total dose of administered medication to the calculated loading dose of 2 µg/kg fentanyl and 0.05 mg/kg midazolam used in the study of Bailey et al (5).
Clearly, any patient not adequately sedated after step D was relatively resistant to the drugs. Additional doses of fentanyl and midazolam were arbitrarily tried during the pilot phase, and midazolam (1–2 mg repeated to the maximum dose agreed on by the radiologists and anesthesiologist on the basis of their clinical experience) was selected for use in step E, the extra dose. As shown in Table 2, we allowed for the use of doses of up to approximately three times the calculated loading dose of midazolam in step E.
|
To facilitate implementation of the protocol, a wall chart that showed the dose per body weight for each step was placed in each procedure room (Table 2). A 50 µg/mL solution of fentanyl was drawn into a 2-mL syringe, and a l mg/mL solution of midazolam was drawn into a 10-mL syringe to minimize the opportunity for error. Each patient had an intravenous line in place; each drug dose was administered into the intravenous tubing over about 30 seconds and slowly flushed with saline. Slow administration is generally advisable, and it is especially important for the avoidance of apnea due to chest wall rigidity, a recognized complication of bolus administration of fentanyl.
Steps A and B were usually administered over 5–10 minutes while the puncture site was being localized, cleansed, and draped, so that the administration of sedative did not prolong the procedure. At this point, if the patient was drowsy and tranquil and if he or she had a minimal reaction to subcutaneous local injection of anesthetic (which we found to be a good marker of an ELD), the procedure proper began. Any patient who reported that the local anesthetic injection was distressing proceeded to the next step in the protocol before further needle insertion was performed.
Patient Monitoring and Data Collection
A registered radiology nurse (D.G., J.A.) was designated to monitor each patient and record clinically important data on a standard form. In addition to recording vital signs, the primary role of the nurse was to observe the patient for hypoventilation or apnea. A pulse oximeter was used to monitor oxygen saturation and heart rate, and an automatic monitor was used to intermittently obtain the blood pressure. As doses of fentanyl and midazolam were administered, individual and cumulative doses were entered on the procedural record form. Standard equipment for cardiopulmonary resuscitation was available. Antagonists to fentanyl (Narcan [naloxone hydrochloride]; DuPont Pharma, Mississauga, Ontario, Canada) and midazolam (Anexate [flumazenil]; Roche, Mississauga, Ontario, Canada) were available during all procedures.
Pain
The patient’s experience was recorded by using a visual analog scale that was graded from 0 to 10. The interventional radiology nurse questioned the patient about his or her pain during the procedure. She asked the patients to describe their pain with standardized adjectives that corresponded to a numerical score as follows: 0, none; 1–2, mild; 3–4, discomforting; 5–6, distressing; 7–8, horrible; and 9–10, excruciating. The highest score reported in each time period was noted on the procedural record form. Procedural maneuvers such as needle insertion or tract dilatation that caused the maximum pain score in each time interval were also recorded.
Hypoxia and Use of Supplemental Oxygen
Patients with clinically important cardiac, respiratory, or hepatic disease and those already receiving supplemental oxygen were all considered to be at high risk and received supplemental oxygen at a rate of 2 L/min. Other patients received supplemental oxygen during the procedure at the discretion of the radiologist. If the oxygen saturation decreased to between 90% and 95%, the patient was asked to take deep breaths. If the saturation decreased to 90% or less, the patient was given oxygen at a rate of 4 L/min. The protocol for the management of abnormal pulse oximetric readings in patients already receiving supplementary oxygen is shown in Table 3.
|
2 test. Differences in pain scores and drug doses between the minor, intermediate, and major procedural groups were tested by using a one-way analysis of variance. For all tests, a P value of less than .05 indicated a statistically significant difference. |
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
Dose of Fentanyl and Midazolam
The mean total dose of fentanyl used in each procedure was 143 µg (range, 50–400 µg). The mean dose of midazolam was 2.7 mg (range, 0.5–9.5 mg). The mean doses required for each procedural category are shown in Table 4. Significantly more fentanyl was required in the major procedural group than in the minor or intermediate procedural groups (P < .05). The dose of midazolam was not significantly different between the groups. The number of steps in the ABCDE protocol required to reach the ELD is illustrated in Figure 1. No patient was adequately sedated after step A, while the majority were sufficiently drowsy after step B to commence the procedure. In addition to the ELD, top-up doses were given in 93 procedures (82%). On average, 2.8 top-up doses of fentanyl (range, one to eight doses) and 2.2 top-up doses of midazolam (range, one to six doses) were required in those procedures in which a top-up dose was necessary.
|
|
|
The minimum intraprocedural oxygen saturation was significantly lower (mean, 93.7%) in the room-air group than in the supplemental-oxygen group (mean, 95.5%; P = .006). The maximum intraprocedural oxygen saturation was also significantly higher in the supplemental-oxygen group (mean, 99.1%) than the room-air group (96.3%, P = .006). All patients with arterial oxygen desaturation in both the room-air and supplemental-oxygen groups responded promptly an increase in inspired oxygen concentration. None of the patients in this study required pharmacologic reversal of sedation because of hypoxia.
Hemodynamic Stability
The mean minimum pulse rate in all patients was 80.6 beats per minute, with a mean maximum pulse rate of 93.5 beats per minute. The pulse changed by more than 20 beats per minute in only 12 patients. In eight of these patients, it increased during needle insertion or track dilatation. In two patients, the pulse increased, and in one, it decreased after combined sedation and analgesia but before any potentially painful stimulus occurred. In one patient, the pulse rate increased transiently from 80 to 120 beats per minute when rigors occurred during abscess drainage.
The mean preprocedural systolic blood pressure in all patients was 137.6 mm Hg, with a mean postprocedural systolic blood pressure of 133.3 mm Hg. Mean preprocedural diastolic blood pressure was 77.8 mm Hg, which decreased to a mean of 75.8 mm Hg after the procedure. The mean range of systolic blood pressure during a procedure was 24.9 mm Hg. In 16 patients, the systolic blood pressure changed by more than 35 mm Hg during the procedure. In six of these patients, a reduction in blood pressure was temporally related to drug administration. In five patients, the blood pressure initially decreased when sedatives were administered and then increased during a painful part of the procedure. In one patient, blood pressure increased substantially during a painful stimulus, while in another, it decreased. No cause was apparent in only three of the 16 patients who had clinically important changes in the systolic blood pressure. All of the previously described hemodynamic changes responded rapidly to increased intravenous fluid and/or to an incremental dose of sedative or analgesic (for those patients in whom a painful stimulus was the cause of the change).
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
General guidelines on the use of sedation by nonanesthesiologists have been published recently (since this study was performed) (7), but no consensus exists on a single drug administration protocol that is suitable for the majority of interventional radiologic cases. Many drugs have been used, including diazepam, midazolam, pentazocine, morphine sulfate, meperidine hydrochloride, fentanyl, nalbuphine hydrochloride, propofol, pentobarbital, and nitrous oxide (8–15). General or epidural anesthesia has been advocated for use in more painful procedures such as biliary interventions (16,17).
In response to these difficulties, we have developed a protocol for sedation and analgesia that is safe for radiologists to use in almost all patients without direct assistance from an anesthesiologist and that takes into account the wide and generally unpredictable interpatient variation in susceptibility to these drugs. Elderly patients and those with cardiorespiratory disease are more susceptible to the respiratory depressant effects of these drugs than are healthy volunteers (18). The rationale for this protocol is to allow patients with a low threshold of response to the drugs to be safely sedated, while patients with a higher tolerance to the drugs can still obtain effective analgesia.
We used a combination of fentanyl and midazolam in this protocol. Both drugs share similar advantages compared with older agents in that their prompt effect facilitates titration of the dose. Fentanyl is a synthetic opioid narcotic. It has a rapid onset of action (usually less than 3 minutes and, rarely, up to 10 minutes) and short duration (40 minutes at these low doses), and it produces potent analgesic and euphoric effects with minimal depression of the cardiovascular system, compared with other commonly used narcotics (19). Its principal adverse effect is profound respiratory depression due to depression of the response to carbon dioxide and due to elimination of the response to hypoxemia. The effects of fentanyl are rapidly reversed with naloxone hydrochloride.
Midazolam is a synthetic imidazobenzodiazepine derivative with a prompt onset of action (2–4 minutes) (20). The duration of the clinical effect is approximately 30–40 minutes. It provides an excellent anxiolytic effect and sedation with more retrograde amnesia and less postprocedural drowsiness than diazepam. It has no intrinsic analgesic effect. It produces some respiratory depression by reducing the ventilatory response to carbon dioxide. Apnea may occur with rapid injection of midazolam. The effects of midazolam can be quickly reversed with flumazenil.
When used in combination, fentanyl and midazolam provide good analgesia, sedation, and amnesia (4,6,8,9). As noted earlier, the effects of the two drugs on respiratory drive are also synergistic (5). This experimental synergism translates into the following important clinical problem: As of 1989, there were reports of 49 deaths due to the use of midazolam combined with an opiate (most commonly fentanyl or meperidine hydrochloride) in the United States (5).
Effectiveness
Patients in 85 (75%) procedures described maximal pain scores of 4 or less on a scale of 0 to 10 (Fig 2). The verbal equivalent of a score of 4 was discomforting pain. This score represented an acceptable level of analgesia while avoiding the risks and costs of general anesthesia. Seven patients reported a pain score of 7 (horrible) or greater. This level of pain is clearly unacceptable for both the patients and those caring for them. Further investigation revealed that severe pain was associated with tract dilation in five of these patients. Tract dilation is often the most painful part of a procedure. This pain needs to be anticipated during a procedure so that a top-up dose can be administered prior to dilation. It is important to wait at least one lead time between the administration of the top-up dose and the performance of the dilation.
During the early phases of development of this protocol, we found that administration of an additional dose of fentanyl during a painful part of a procedure often resulted in poor pain control, while, paradoxically, midazolam provided much better pain relief. This observation is supported by study findings in healthy volunteers that demonstrated a statistically significant reduction in the affective and motivational components of the pain experience with midazolam (21). This was the rationale behind the choice of midazolam for the extra dose, step E.
One patient who underwent transjugular intrahepatic portosystemic shunt insertion complained of severe pain both during the transhepatic puncture and during dilatation of the transhepatic tract. In our experience, the pain associated with these insertions can be difficult to control, but since only three patients who were undergoing transjugular intrahepatic portosystemic shunt insertions met the inclusion criteria for this study, further evaluation of analgesia for this procedure is warranted. In a recent survey of North American interventional radiologists (22), 20% of respondents expressed a preference for general anesthesia for transjugular intrahepatic portosystemic shunt procedures. Once the ELD has been administered and the procedure has begun, we now tend to use subsequent steps of the protocol (eg, step C in a patient whose ELD was achieved in steps A and B) instead of calculating a top-up dose of 25% of the ELD. This practice simplifies drug administration.
Table 4 demonstrates a progression in the mean maximal pain score from 3.1 in minor procedures to 3.6 in major procedures, although the difference in pain scores between groups was not statistically significant. This confirms the effectiveness of the protocol in procedures of varying complexity, although the lack of a difference in pain scores between those who underwent minor procedures and those who underwent major procedures is still somewhat surprising. It is possible that the group of patients who underwent minor procedures had expressed a very low threshold for pain, which influenced the radiologist’s decision to use a combination of sedation and analgesia. Significantly more fentanyl was required in major procedures than in minor or intermediate procedures due to their increased potential for discomfort and longer duration.
Few other studies report on the effectiveness of analgesia in interventional radiologic procedures. Ayre-Smith (9) described "favourable results" with fentanyl and midazolam in a preliminary report of 12 patients who underwent unspecified interventional radiologic procedures. Miller and Wall (15) reported effective analgesia in all 100 patients who underwent sedation and analgesia for radiologic procedures with a combination of fentanyl and diazepam. Most of the patients in their study were undergoing diagnostic angiography, which, in general, would not be considered to be particularly painful. In one large study of pain control during interventional biliary procedures (16), epidural anesthesia was effective in 91% of cases, while intravenous sedation was successful in only 50%. This study may have been biased against intravenous sedation, since most patients received single-agent therapy with a narcotic, such as meperidine or morphine, without the synergistic benefit of a benzodiazepine.
It is important to stress that effective use of local anesthetic is a vital component of any protocol for analgesia in interventional radiology. It is possible that optimal use of local anesthetic may have substantially reduced the pain associated with tract dilatation in the seven patients in whom pain was poorly controlled.
Safety
The principal concerns with the combination of fentanyl and midazolam are hypoxemia and/or respiratory arrest. No patient in our study became apneic. It is important to restate that the primary role of the designated observer (a registered nurse in our department) is to detect bradypnea and apnea. This is in keeping with the guidelines of the American Society of Anesthesiologists that recommend that "ventilatory function should be continually monitored by observation" (7). We chose an oxygen saturation level of 90% or less to indicate clinically significant hypoxemia. Below this level, further desaturation occurs along the steep portion of the oxyhemoglobin dissociation curve, resulting in a greater risk of hypoxia-related complications such as cardiac or cerebral ischemia, and arrhythmias. It is our practice to stop the procedure when the oxygen saturation decreases to 90% and to assist ventilation with the placement of an airway and/or with use of a ventilator bag. The oxygen saturation decreased to 90% or lower in only 15 (13%) patients. This finding attests to the safety of the staged approach in drug administration. The safety of the ABCDE protocol is also confirmed by the fact that a increase in the inspired oxygen concentration was the only measure required to reverse hypoxemia in all 15 patients.
Our data show that the use of supplemental oxygen from the start of a procedure reduces the risk of clinically important hypoxia. Patients in 11 (22%) of 51 procedures performed in the room-air group became hypoxic compared with patients in four (7%) of 62 procedures performed in supplemental-oxygen group. This difference was statistically significant (P = .04). There was also a statistically significant difference between the two groups in mean minimum and mean maximum intraprocedural oxygen saturation. Therefore, we now routinely use supplemental oxygen in all patients in accordance with the recommendations of the American Society of Anesthesiologists (7).
Hemodynamic changes were seen in a minority of patients. The changes in pulse rate were not clinically significant in that no intervention was required. A reduction of more than 35 mm Hg in systolic blood pressure that could be directly attributable to drug administration was identified in 11 procedures. The clinical importance of these changes was minimal; all patients were asymptomatic and responded to increased intravenous fluids.
The ABCDE protocol is a safe and effective means of providing sedation and analgesia for most interventional radiology procedures. An ELD can be achieved with stepwise administration of drugs, without risking severe respiratory complications. Continuous monitoring of oxygen saturation, pulse, and blood pressure are essential in all patients, since clinically important hypoxemia and minor hemodynamic changes will occur in a minority. Ideally, drug administration and patient monitoring should be performed by an independent observer, usually a trained radiology nurse, to allow the radiologist to fully concentrate on the procedure. The administration of supplemental oxygen at a rate of 4 L/min via a nasal cannula helps to prevent clinically important hypoxemia and is recommended for all patients. Particular care should be taken to administer top-up doses prior to potentially painful aspects of a procedure, such as tract dilatation.
Since this study was performed, the ABCDE protocol has been introduced (by D.E.M.) at a second institution (St Vincent’s University Hospital, Dublin, Ireland) and is now used in all appropriate interventional cases. We have found that this protocol, with the wall chart in the interventional room, allows new interventional radiology fellows and others who are relatively inexperienced with sedation and analgesia to provide safe and effective pain control at an early stage in their training.
|
FOOTNOTES |
|---|
Author contributions: Guarantor of integrity of entire study, D.E.M.; study concepts, D.E.M., S.J.S., S.M., G.W.S., N.B.; study design, D.E.M., N.B.; definition of intellectual content, S.J.S., D.E.M., G.W.S., N.B.; literature research, S.J.S., D.E.M., S.M., G.T.; clinical studies, D.E.M., S.M., J.R., G.T., D.G., J.A., S.S., G.W.S.; data acquisition, S.J.S., S.M., D.G., J.A.; data analysis, S.J.S., S.M.; statistical analysis, S.J.S.; manuscript preparation, S.J.S.; manuscript editing, S.J.S., D.E.M., J.R., G.W.S.; manuscript review, D.E.M., J.R., S.S., G.W.S., N.B.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  D. Gianfelice, A. Khiat, M. Amara, A. Belblidia, and Y. Boulanger MR Imaging-guided Focused US Ablation of Breast Cancer: Histopathologic Assessment of Effectiveness—Initial Experience Radiology, June 1, 2003; 227(3): 849 - 855. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
