Scientific Papers Focus on Patient Safety at the 2009 ASA Annual Meeting

Steven B. Greenberg, MD; Glenn S. Murphy, MD; Jeffery S. Vender, MD

Over 1,600 abstracts were presented at the 2009 American Society of Anesthesiologists Annual Meeting in New Orleans, Louisiana. As in previous years, a number of these abstracts examined issues directly related to patient safety. This brief review will highlight a few of the important abstracts discussed at the meeting.


Perioperative Medications and Morbidity & Mortality

There continues to be an interest in perioperative therapies that may contribute to improved postoperative patient outcomes. Investigators from Johns Hopkins performed a prospective study to examine the effect of preoperative aspirin, beta blockers, ACE inhibitors, and diuretics on mortality after cardiac surgery (A2). Data were prospectively collected on 9,129 patients from a single center over a 4-year period. Thirty-day mortality was significantly less in patients receiving preoperative aspirin (OR=0.73), beta blockers (OR=0.78), and ACE inhibitors (OR=0.72). The preoperative use of diuretics was associated with an increase in mortality (OR=1.56). Another large database study (20,000 patients undergoing 40,000 surgical procedures were analyzed over a 12-year period) from the San Francisco VA Medical Center (A705) examined the association between perioperative beta blockade and postoperative outcomes. A standard protocol guiding the administration of beta blockers in at-risk patients was used. The addition of beta blockers resulted in a significant reduction in 30- and 365-day mortality (OR=0.59 & OR=0.83, respectively). Withdrawal of beta blockers was associated with an increased risk of death at 30 and 365 days (OR=3.24 & OR=1.83, respectively).

The influence of perioperative statin therapy on patient outcomes was examined in several abstracts. Pan et al. performed a retrospective cohort study (A1) involving patients undergoing CABG surgery with cardiopulmonary bypass. Patients receiving preoperative statin therapy (N=2385) were compared to patients not receiving statin therapy (N=1609). Multivariate logistic regression revealed that preoperative statin therapy was independently associated with a significant reduction in hospital mortality (p<0.05), reduced hospital length of stay (p<0.05), reduced postoperative renal insufficiency (p<0.05), and reduced postoperative IABP support (p<0.01). Long-term survival was also significantly improved with statin therapy. Another investigation (A1583) assessed the relationship between preoperative and early postoperative use of statins and the incidence of acute kidney injury (AKI) in elective cardiac surgical patients. Twenty-one percent of 324 patients studied developed AKI. These patients stayed in the hospital longer and were more likely to develop pneumonia or die. Early postoperative statin use was associated with a lower incidence of AKI (p=0.03), while no association between preoperative statin use and AKI was noted. Further analyses such as the ones above may allow for the development of protocols that incorporate such therapies to reduce morbidity and mortality rates.


Acknowledgment and Reduction of Medical Errors

Medical errors harm thousands of patients in hospitals each year. Cooper et al. (A614) attempted to identify human factors contributing to medication errors by anesthesia care providers. A medication error rate of 1:150-200 anesthetics was observed in 10,574 anesthetics. The 3 most common factors that were associated with medication errors were distraction, haste/pressure to proceed, and misread or look-alike vials/labels. New technology for labeling medications may reduce medication errors. Levine et al. (A612) compared the “Smart Label” system (a bar code-assisted syringe labeling technology) with conventional labeling measures. The use of conventional methods revealed a 10.4% error rate (either labeling the drug with the wrong date/time or incorrect drug concentration), while the use of the “Smart Label” system was associated with a 0% error rate. Eight-six percent of the 64 subjects using the new technology reported it to be faster than conventional methods, and 98% of the subjects thought it improved safety by reducing drug labeling errors.

A retrospective study (A1061) identified possible risk factors for retained instruments and sponges after surgery. Patients with retained sponges or instruments (n=89) were more likely to have a higher BMI and had counts of sponges and instruments performed at the end of the procedure. Risk of retention of foreign bodies occurred most frequently in non-emergent cases and with unplanned changes to procedures.

Other abstracts addressed patient care “handoffs” as a source of iatrogenic errors. One study (A1164) analyzed the quality of PACU handoffs using 4 criteria: content of handover, patient condition, professional behavior, and outcome assessment. Quality handovers were achieved 90% of the time. Unsatisfactory handovers were more frequent after 5:00 pm and when the anesthesia provider changed during the operation. Poor quality handovers were associated with a 61-minute increase in PACU readiness time. Another investigation (A1179) evaluated results from a survey distributed to assess operating room handoff adequacy, location for best handoff, method of best handoff, and need for inclusion of the electronic medical record in handoff communication. Of the 70 surveys completed, 34% found the current handoff practice to be inadequate. Most of the surveys revealed that handoffs should occur in the OR and in person. In addition, most surveys reported an interest in incorporating handoff communication into the electronic medical record.


Nosocomial Infections and Prevention

Central venous catheter (CVC) bloodstream and surgical site infections account for a substantial proportion of in-hospital morbidity. A study from Massachusetts General Hospital (A1167) attempted to define the incidence of catheter-related bloodstream infections (CRBSIs) attributable to central lines placed by anesthesiologists in the OR. Thirty-three CRBSIs were identified (out of 3948 catheters placed) over a 9-month period. Only one of the catheters that resulted in a CRBSI was placed in the OR by an anesthesiologist, while 32 catheters were placed by other providers in other locations in the hospital. The authors provide reasons for the low rates of CRBSI in the OR which include provider experience, degree of supervision of trainees, familiarity with sterile technique, placement environment, patients’ underlying comorbidities, and the time to CVC removal. Another abstract (A1166) examined staff education and physician simulation training on CRBSI-related costs over a 4-year period. The mandatory steps included an intranet module on infection prevention, a procedure checklist, and a 4-hour skills training course. With the above measures, a 53% total cost reduction (mean cost savings of $211,968) occurred between the pre-training era (2005-2006) and the post-training era (2007-2008). This cost savings may be attributed to the reduction in CRBSIs during the same time period.

The lack of hand hygiene can contribute to nosocomial infections including CRBSI. An investigation from the Netherlands (A1174) examined the frequency of hand hygiene in an OR among perioperative staff members who did not perform a surgical scrub. Among 28 operations (60 hours) that were observed, only 2% of staff members performed hand hygiene practices upon entering the OR and 8.4% of staff performed hand hygiene upon leaving the OR. In addition, when performing radial arterial catheter placement, 0% of staff members wore gloves. Another study (A1170) surveyed health care providers regarding hand hygiene compliance. All of the 107 providers surveyed agreed that they should maintain hand hygiene, and most respondents believed that their own compliance was high. The author suggests that the low compliance problem associated with hand hygiene worldwide is a behavioral one among health care providers that requires acknowledgement and change.


Potential Complications of the Prone and Sitting Position

Several abstracts reviewed possible deleterious effects associated with intraoperative prone and sitting positioning. An abstract (A1013) examined a database of 43,410 spinal fusion operations among 17 academic centers. This study reported on 100 control patients randomly selected from the 320 controls without postoperative visual loss (POVL). Results revealed that intraoperative blood pressure and hematocrit for prone spinal fusion surgery in control patients without POVL varied substantially. Fifty-four percent of these subjects had mean arterial pressures ≥ 30% below baseline values for at least 15 minutes.

Two investigators examined the effect of prone positioning and general anesthesia on ocular physiology. Grant et al. recruited 10 healthy volunteers to lie prone for 2 separate 5-hour sessions on a Jackson table (A1014). During the study period, intraocular pressure, choroidal thickness, optic nerve diameter and MAP were all significantly increased. The reverse Trendelenberg position had no affect on these changes. An abstract from Nara Medical University (A1016) measured intraoperative changes in intraocular pressure (IOP) under sevoflurane and propofol anesthesia during spine surgery in the prone position. IOP increased from baseline values (8-11mmHg), and continued to be elevated 5 minutes after postural change to the supine position. The increase in IOP was comparable in patients exposed to either sevoflurane or propofol anesthesia. By further studying the physiologic changes and better identifying risk factors for postoperative visual loss, health care providers may eventually be able to mitigate the incidence of this devastating complication.

Several studies assessed cerebral physiologic changes during sitting position operations. Haas et al. (A1009) presented preliminary results of an ongoing case series investigating which level of hypotension is safe for patients undergoing shoulder surgery in the sitting position. In 28 subjects, systolic blood pressure (SBP) was lowered to the range of 90-100mmHg. All of the baseline normotensive patients and 9 out of 10 of the baseline hypertensive patients had unchanged EEGs throughout the procedures. However, significant EEG changes (burst suppression/electrical silence) were observed in one of the chronic hypertensive patients when the SBP was lowered to approximately 90mmHg. The EEG returned to baseline within minutes when the SBP was raised to 120mmHg. The author concluded that most chronically hypertensive patients can safely tolerate significant hypotension in the sitting position during general anesthesia. An abstract from NorthShore University HealthSystem (A620) compared regional cerebral oxygen saturation (rSO2) values in 110 consecutive patients presenting for elective shoulder surgery in the beach chair position versus the lateral decubitus position. More than 75% of patients in the beach chair group had at least one cerebral desaturation event (defined as a ≥20% decrease in rSO2 values from baseline). No episodes of critical desaturation events were observed in the lateral decubitus position during the entire intraoperative measurement period. Similarly, Lathouwers et al. (A1288) reported changes in cerebral oximeter (StO2) measurements in patients undergoing shoulder surgery either in the sitting position or side position. Thirty-eight out of 45 patients in the sitting position had StO2 values of ≤55% (critical cerebral desaturation threshold), while no patients in the side position group had critical desaturation events. Lastly, Tange et al. (A522) examined whether the sitting position during general anesthesia promotes changes in cerebral oxygen metabolism in surgical patients. Thirty patients were assigned to either a control group (n=8) versus a cardiovascular risk factor group (n=22) (patients with hypertension, diabetes mellitus, or hypercholesterolemia). Heart rate and blood pressure declined under general anesthesia, but the tissue oxygen index values remained normal in both groups. In contrast to the previous abstracts (A620, A1288), cerebral oxygenation was not significantly altered when MAP was maintained > 60mmHg (and measured at the patients’ upper limb). Additional studies are needed to validate the above findings and further investigate the changes in physiology that occur during the sitting position.


Miscellaneous
“Triple Low”

Three notable abstracts examined an association between the combination of low BIS levels, low blood pressure levels, low anesthetic levels, and postoperative morbidity and mortality. An abstract (A6) from the Cleveland Clinic investigated the interaction of the above factors and hospital length of stay and mortality. Data from 18,035 non-cardiac procedures were analyzed and revealed that the relative risk of mortality was significantly greater in patients with a combination of low MAC and low MAP. Low BIS levels further increased relative mortality. The same group (A880) reported that 48% of the 18,035 non-cardiac procedures had at least one “triple low” episode (low MAP, low BIS, and low MAC). Increased duration of low MAP, low BIS, and low anesthetic concentration increased the incidence of 30-day readmission and postoperative mortality, while impairing postoperative recovery (pain, complications, and excess length of stay). Lastly, abstract (A354) evaluated the effects of vasopressors on mortality in patients with the “triple low” combination. Among 17,067 patients who were evaluated, those who had a “triple low” and were rapidly treated with vasopressors had mortality similar to the reference group of 7%. However, when vasopressor treatment was delayed in patients with “triple low” episodes, mortality increased to approximately 20%. This may suggest that early intervention may attenuate the long-term effects of the “triple low” combination.


Perioperative Complications in Patients with Coronary Stents

Two studies examined patients undergoing surgery with bare metal (BMS) or drug eluting (DES) stents and associated morbidity and mortality. The POSTSTENT study (A5) measured the incidence of postoperative complications in patients undergoing cardiac surgery. Among 219 patients with stents, 34% of patients had complications. Twenty-two percent of patients had bleeding complications (after 94% of patients stopped clopidrogrel approximately 8 days prior to surgery and 8% of patients stopped aspirin approximately 5 days prior to surgery). Four postoperative coronary thromboses were detected (2 of which were on days 32 and 59). Sixty-day mortality was 2.3%. Risk factors for all postoperative complications included female sex, urgent surgery, cardiopulmonary bypass, and withdrawal of clopidogrel < 5 days. Another study from the same investigators (A1571), examined the incidence of perioperative complications associated with patients undergoing non-cardiac surgery with BMS or DES. Aspirin was discontinued in 53% of patients (with a mean duration of 6 days prior to surgery), while clopidogrel was stopped in 83% of patients (with a mean duration of 7 days prior to surgery). Twenty-three percent of patients experienced a postoperative complication. Excessive bleeding accounted for 17% of these complications. Sixty-day mortality was 4.7%, while age and vascular surgery were identified as risk factors for bleeding complications. These 2 large studies indicate that patients with coronary stents still have significant perioperative morbidity.

This brief review summarized only a small number of the important abstracts on patient safety presented at the 2009 Annual Meeting. To view other abstracts on patient safety, or to obtain further information on the abstracts discussed in this review, please visit the Anesthesiology website at www.anesthesiology.org.

Drs. Greenberg, Murphy, and Vender are affiliated with the Evanston Northwestern Healthcare Department of Anesthesiology. They also serve on the APSF Editorial Board.