Observations on use of wrong agent in an anesthesia agent vaporizer.

The creation of agent mixtures from the addition of the wrong agent to a vaporizer might pose a risk to the patient. Patient injury would be more likely if the anesthesia gas monitor displayed erroneous concentration values. Conventional inhalation agent monitors do not necessarily distinguish anesthetic agents. Some modern monitors have that ability but its clinical significance has not been determined. We wanted to simulate such an erroneous mixture in a laboratory setup. Six comparisons were made. Isoflurane, Enflurane, and Halothane vaporizers were first filled with the correct agent. They were run at 5 liters/minute fresh oxygen flow at a vaporizer dial setting of 5% until it reached the "refill" line. Then, one of two incorrect agents was added to the "full" line. Thereafter, the vaporizer continued at the same flow and the same dial setting until it was exhausted. Vaporizer output was recorded or calculated by using three methods of measurement: mass spectrometry, conventional infrared analysis (at 3.3 micrometer wave length), and piezoelectric crystal analysis. Additional calculations were used to estimate measurements that could not be made because of lack of available equipment. In a Halothane vaporizer: Enflurane added--not a significant problem; Isoflurane added--not a significant problem. In an Isoflurane vaporizer: Halothane added--not a significant problem; Enflurane added--not a significant problem. In an Enflurane vaporizer: Isoflurane added--not a significant problem; Halothane added--the sum of the delivered Halothane MAC and the delivered Enflurane MAC was twice the expected Enflurane MAC output from vaporizer, with conventional agent monitor reading which showed decreasing agent concentration. Patient injury could be more likely in this last case. In this last case and in all cases, piezoelectric crystal monitoring correctly displayed the sum of the two agent concentrations in volumes percent. Automatic agent identification can identify erroneous agents.

Optimization of alarms: a study on alarm limits, alarm sounds, and false alarms, intended to reduce annoyance.

Alarms in the operating room remain a major source of annoyance and confusion. Nearly all alarms result from a transgression of certain alarm limits. We surveyed manufacturers at a major meeting of anesthesiologists for their default alarm limits. We also conducted a mail survey of anesthesiologists in the United States, Finland, and The Netherlands, to learn their chosen alarm limits, reasons for turning off the alarms, and estimates of the number of false alarms. The surveys of medical monitoring manufacturers demonstrated a wide variety of default alarm settings. Anesthesiologists' chosen alarm limits tended to parallel the manufacturers' defaults. Fewer than 30% of anesthesiologists stated that they did not turn off the alarms. The leading reason for turning off alarms was the large number of false alarms. Estimates of the number of false alarms varied from 30% to 76%. The desired role of alarms in anesthetic practice and some suggestions for the general improvement of alarms are discussed.

The Diatek Arkive "Organizer" patient information management system: experience at a university hospital.

OBJECTIVE: To install and successfully use early commercial automatic anesthesia recordkeepers, the Diatek Arkive "Organizer" units, in the operating rooms at a major university medical center. METHODS: Because of the history of previous academic attempts at automatic anesthesia recordkeeping, the units were installed on a "surprise" basis, with hardly any discussion of the devices beforehand. RESULTS: The devices themselves had a number of minor difficulties at the start, most of which were promptly corrected. Eventually the units were in use in all non-cardiac general operating rooms. At one point, usage reached over 90% of possible cases. Continued opposition to the device on the part of certain individuals, coupled with the obsolescence of the present devices in light of new technology, led to the eventual abandonment of the system and removal of the devices from service. DISCUSSION: Total resistance to the new devices on the part of a few vocal faculty members was a major factor in the ultimate downfall of the system. The method of introduction, and especially the lack of involvement of faculty, residents, surgeons, operating room personnel, hospital computer personnel, and the hospital administration in the installation also played a role in the failure of the system. Lack of a workable training mechanism for new residents prevented that user group from rapidly gaining comfort with the systems.

Technology evaluation report: obtaining pulse oximeter signals when the usual probe cannot be used.

We studied the function of four different monitoring probes used with the Satlite pulse oximeter (Datex, Helsinki). The aim was to evaluate ease of use and compare the function of the probes and their attachment methods in different locations (finger, toe, ear, thumb web, instep, wrist, and ankle). Two similar pulse wave oximeters were used in the study. To select the best signal we determined the absolute height of the plethysmographic waveform for each probe. Probes were compared on awake normal adult volunteers (N = 13), in anesthetized adult patients (N = 12), and in neonates weighting 500-1000 g (N = 8). In all the adult trials, the clip-on finger probe was used as a reference, and probes taped or clipped onto the finger provided adequate signal from the ear or from the thumb web, however. In the neonates, taped-on probes were most satisfactory.

Pre-operative oximetry and capnometry: potential respiratory screening tools.

The growing number of patients admitted for outpatient surgery or for same-day admission makes it difficult to obtain thorough pulmonary evaluation. We wanted to evaluate the applicability of pre-operative pulse oximetry and capnography as possible pulmonary screening tools. In this preliminary study, 200 unselected, unmedicated adult patients who were being admitted for surgery were connected to a dual parameter patient monitor (Capnomac Ultima, Datex). A standard adult clip-on finger probe was used for pulse oximetric oxygen saturation. Sidestream capnometry documented the end-tidal carbon dioxide and the capnogram which was recorded for further analysis. In these unmedicated patients, the oxygen saturation ranged from 91 to 99% and was found to be 94% or less in five percent (N = 10) of the cases. The end-tidal carbon dioxide ranged from 21 to 48 mmHg. In five percent of the cases (N = 10) it was found to be 45 mmHg or higher, reflecting elevated arterial CO2. When the shape of the capnogram was rated, it was found normal in 54% of the cases. Slow rising capnogram, indicating mild (N = 84) or moderate (N = 8) airway obstruction was detected in 42% or 4% of the cases respectively. Since pulse oximeter and end-tidal carbon dioxide values are often not measured until after sedation or after induction of anesthesia, patients with pre-operative abnormalities might escape pre-operative detection. In unmedicated patients, routine pre-operative or pre-admission determination of oxygen saturation, end-tidal carbon dioxide and the capnogram may be a valuable screening tool.

Ankle blood pressure measurement, an acceptable alternative to arm measurements.

The use of automatic noninvasive blood pressure (NIBP) devices has become a common technique to monitor blood pressure intraoperatively. The usual cuff placement for these devices on the upper arm sometimes poses problems. As an alternative, many clinicians place the cuff on the ankle. This practice has not been previously investigated to determine its efficacy. The purpose of our study was to determine whether a noninvasive blood pressure cuff on the arm could be replaced by one on the ankle. We monitored 24 patients intraoperatively with two non-invasive blood pressure cuffs, one on the upper arm and one on the ankle. Systolic, diastolic, and mean pressures were obtained from each cuff placement at intervals of no shorter than 3 minutes. The time necessary to obtain the measurements and the presence of any artifact were also recorded. A total of 404 pairs of data were obtained and the systolic blood pressure ranged from 82 to 196 mm Hg. The mean and diastolic pressure readings were equivalent between the arm and ankle blood pressure readings. The systolic pressures were not equivalent, reflecting the fact that the ankle systolic blood pressure is physiologically higher than the arm systolic blood pressure. The difference between the times necessary to obtain the readings from arm or ankle was not statistically significant. Eight of the paired readings (2.0%) represented artifact, arbitrarily defined as a difference in mean blood pressure readings of 15 mm Hg between the arm and the ankle. Since the mean blood pressure readings obtained at the arm and at the ankle were statistically equivalent, we concluded that the ankle cuff placement provided a reliable alternative to the placement of the cuff on the arm.

Auditory alarms during anesthesia monitoring with an integrated monitoring system.

Alarms in the operating room remain a major source of annoyance and confusion. A previous study by Kestin et al. utilized a specific combination of distinct, separate monitors in 50 pediatric patients. He reported a mean of 10 alarms per case with a mean frequency of one alarm every 4.5 minutes. The alarms were classified as spurious (75%), change outside the alarm limits (22%), or patient risk (3%). We performed a similar study with 50 adult patients under general anesthesia with default alarm settings on an integrated monitor, (Cardiocap, Datex, Helsinki). In our study, the number of alarms averaged 3 per case with a mean frequency of one every 34 minutes. Spurious alarms (those caused by electrocautery, accidental patient movement, or other non-physiological reasons) represented only 24% of all alarms. Those alarms sounding that were outside the limits occurred at a rate of 53%, and those that were considered patient risks occurred at a rate of 23%. Of the alarms, 67% occurred during the beginning and end of anesthesia. The end-tidal carbon dioxide accounted for 42% of the alarms, mostly during intubation and extubation. Suggestions are made for further improvement in alarm systems.

An expert system to teach troubleshooting of common problems associated with the automated anesthesia recordkeeper.

Automated anesthesia recordkeepers have been used to monitor patients during surgery in up to 90% of cases at The Ohio State University. The record-keeping devices are complex and can be difficult to troubleshoot. The 1st-CLASS Fusion Program, an expert system "shell-program," has been programmed to allow the resident or nurse anesthetist to solve the two most common types of problems associated with the recordkeeper: printer problems and patient monitor problems. Use of this program allows the resident or nurse anesthetist to troubleshoot the recordkeeper quickly and accurately and promotes in the user a sense of competence and control over the technology.

Effect of automatic record keeping on vigilance and record keeping time.

We have evaluated the effect of an automatic anaesthesia record keeper (AARK) on record keeping time and vigilance. With informed patient consent and institutional approval, we videotaped the attending anaesthetist and his/her immediate surroundings during 66 surgical procedures. Thirty-seven cases were charted manually and the remaining 29 were charted with a commercially available AARK. In order to evaluate vigilance, a physician examiner entered the operating room unannounced once during 33 of the manually charted cases and during 22 of the automatically charted cases and asked the anaesthetist to turn away from the monitors and recall the current value of eight patient physiological variables. The examiner recorded the recalled values and also the actual current monitor values of these variables. The videotapes were reviewed and the anaesthetist's intraoperative time was categorized into 15 predefined activities, including intraoperative anaesthesia record keeping time. We compared recalled and actual variable values to determine if the recalled values were within clinically relevant error limits. There was no statistical difference between the mean percentage case time spent recording manually (14.11 (SD 3.98)%) and automatically (12.39 (3.92)%). Moreover, use of the AARK did not significantly affect vigilance. Despite major advances in monitoring technology over the past 14 years, record keeping still occupies 10-15% of the anaesthetist's intraoperative time. It appears that in using an AARK, the anaesthetist reallocates intraoperative record keeping time from manual charting to dealing with problems in the anaesthetist machine interface caused by inadequate design.

Clinical evaluation of the 'head-up' display of anesthesia data. Preliminary communication.

To solve the problem of monitoring the patient during administration of anesthesia, a commercially available head-up display (HUD) was evaluated during one day of surgery at the Ohio State University Hospitals. This monitor is mounted on a headband worn by the anesthesiologist. It projects a monochrome image of monitor data directly into one eye. Eleven anesthesiologists tested the device. Most users were able to adjust to the monitor in about fifteen minutes. Nine of the testers expressed a desire to evaluate the monitor further. Their major complaints were that the connecting cable between the HUD and its computer was too short, the resolution of the monitor was inadequate, and the data on the screen were not organized in a familiar way. If these problems could be corrected, most users believed that this HUD could be a valuable tool to aid the anesthesiologist in the operating room.

Evaluation of users' abilities to recognize musical alarm tones.

The problem of accurate identification of alarm sounds in the operating room, recovery room, and intensive care environment has persisted for many years. Monitors made by different companies may have different alarm sounds for the same monitored variable, and similar alarm sounds for different variables. In an effort to illustrate universal alarms sounds, a system of six musical alarm tones was designed with musical themes from popular songs used for oxygenation, ventilation, cardiovascular monitoring, temperature monitoring, artificial perfusion, and drug administration systems. These melodies were played for a group of anesthesiologists and others, who were initially asked to guess the organ system for each melody. The answers were then given to the participants, and after a short delay the melodies were played again in a different order. Seventy-nine response sheets were collected. The expected random score was 1.0 +/- 1.0 SEM correct. The observed score on the first hearing was 1.5 +/- 1.6 SEM, p = 0.01 compared with the random score. The observed score on the second testing was 4.3 +/- 2.2 SEM, p = 0.001 compared with the first hearing. Indeed, 42 of 79 (53%) respondents got all six answers correct on the second testing, versus three respondents for the first testing. The implications of these findings are discussed in the context of integrated alarm systems used in complex medical environments such as the operating room.

Can people hear the pitch change on a variable-pitch pulse oximeter?

The introduction of the variable-pitch feature on pulse oximeters in 1983 by the Nellcor Corporation (Hayward, CA) allowed users to rapidly detect changes in oxygen saturation by listening for changes in the pitch of the tones emitted by the pulse oximeter. A few individuals have reported that they have been unable to detect a change in pitch when oxygen saturation changes. To these individuals, the variable-pitch feature of these pulse oximeters has not been beneficial. Using the pitches from one manufacturer of oximeters, we created a computer program to simulate the pitches that accompanied various oxygen saturations. The pitches were recorded onto a tape player and played for 75 volunteer subjects unfamiliar with the pitches of a variable-pitch pulse oximeter. Of our sample, 67% were able to detect a single change in pitch corresponding to a 1% fall in oxygen saturation, and 11% of the population could not detect a change in pitch until there was a change in pitch with every beat. We suggested four alternative designs that may prove beneficial to this group of individuals.

The "Torrstick": a ruler calibrated in millimeters of mercury for measurement of hydrostatic pressure.

We have designed and constructed a ruler ("Torrstick") calibrated in millimeters of mercury for the measurement of hydrostatic pressure differences. Marks are placed on a measuring stick every 1.34 cm, to afford the best compromise between a column of 0.9% saline solution and a column of blood. Several uses for this ruler are proposed, including the measurement of central venous pressure.