Circulation 94,429 • Volume 26, No. 3 • Winter 2012   Issue PDF

Scientific Papers Address Patient Safety

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

Over 1,700 abstracts were presented at the 2011 American Society of Anesthesiologists annual meeting in Chicago, Illinois. As in previous years, a number of these abstracts examined issues directly related to patient safety. This brief review will highlight several abstracts discussed at the meeting.

Patient Handover Communication

checklistMark et al. from the Durham VA Medical Center, North Carolina, performed a review of the literature on postoperative patient handovers (A1528). Twenty-three articles identified factors leading to poor handover communication, while also providing recommendations for improving the handover process. Some factors identified for ineffective handovers were poor teamwork and communication, patient instability on arrival, unclear procedures, technical errors, unstructured processes, interruptions and distractions, lack of central information repositories, and nurse inattention. Some of the broad recommendations included standardizing the transfer of care and training in team skills and communication. Bready et al. from UTHSC, San Antonio, Texas, devised an 18-element checklist and performed a process improvement study over a 4-month period. With education and implementation of the checklist, the rate of communication of essential elements improved from 50% at baseline to 98% post-implementation (A1530). Greilich et al. from UT Southwestern Medical Center in Dallas, Texas, designed a handoff checklist for patient transfers from the operating room to the intensive care unit (A1529). After a 20-week study period, the average provider satisfaction following implementation increased by 51% from baseline measures. The average time to complete the checklist was 11 ± 4 minutes. Central line associated blood stream infections fell from 2.9/1000 catheter days to zero in the 5 months following implementation. Further studies are needed to investigate and identify best practices in postoperative patient handovers (A1528).

Database Studies and Perioperative Complications

Several institutions utilized database information to examine factors that may be associated with perioperative complications. Mehta et al. from the University of Washington in Seattle reviewed 85 claims for cautery-related surgical fires from 7031 total surgical claims over a 23-year period. Cautery related fires increased from <1% in 1985-1994 to 4% of all surgical claims in 2000-2008. Ninety-six percent of fires occurred in high risk procedures (which were not defined by authors) and 84% of fires occurred during sedation (A1722). Investigators (A435) performed a cross-sectional study to identify cases of maternal cardiac arrest among admissions for delivery from 1994-2006 utilizing the Nationwide Inpatient Sample (NIS) of the Healthcare Cost and Utilization Project. Six hundred and ten women arrested during 7,503,155 admissions for delivery (an event rate =1:12,300). Women who experienced an arrest had the following co-diagnoses in descending order of frequency: hypertensive disorders of pregnancy, postpartum hemorrhage, antepartum hemorrhage, amniotic fluid embolism (AFE), cardiomyopathy, anesthetic complications, sepsis, aspiration pneumonitis, venous embolism, stroke, trauma, myocardial infarction (MI), pulmonary edema, magnesium toxicity, local anesthetic toxicity, status asthmaticus, anaphylactic shock, and vascular catastrophes. Cardiac arrest secondary to antepartum hemorrhage, MI, AFE, stroke, embolism, trauma or vascular catastrophe was associated with low survival rates (from 46 to 0%) (A435). Roth et al. (A716) from the University of Chicago, utilized the NIS database to examine trends in obstetric anesthesia complications in the US from 1993-2008. While the total number of pulmonary, cardiac, or CNS complications appeared to be decreasing (958 to 376), the number of diagnoses that were other or unspecified complications has increased significantly (4,159 to 16,138). The database did not identify the nature of these other or unspecified complications.

Another study (A415) examined closed claims associated with epidural and spinal anesthesia in non-obstetric surgical cases (1990-present). Of 6894 claims, 443 were associated with spinal or epidural anesthesia in the surgical setting. While most of the injuries were temporary (45%), 37% were associated with death and brain damage. Sixteen percent of injuries were associated with permanent nerve injury. The 4 most common causes for injury were block technique (primarily with nerve damage), neuraxial associated cardiac arrest, dural puncture, and high spinal epidural block. Sviggum et al. (A416) at the Mayo Clinic performed a retrospective cohort study to investigate whether peripheral nerve blocks for upper extremity joint surgery (UEJS) increased the risk of peripheral nerve injury (PNI). Out of 3044 patients who underwent UEJS, 53 cases of PNI were identified (1.7% incidence). A multivariate analysis suggested that peripheral nerve blockade was not associated with PNI (OR=0.72; 95% CI 0.39 to 1.32). Those patients who developed PNI had complete neurologic recovery in 62% of cases, while 34% of cases experienced partial recovery.

Some authors investigated risk factors for reintubation, prolonged intubation, and postoperative respiratory complications. Investigators (A549) performed a case-controlled study to identify the incidence of tracheal reintubation and associated risk factors. The authors reviewed medical records of 58,854 anesthetic cases in a tertiary care hospital in Thailand over a 5-year period. The reintubation rate was 1.86 per 1000 intubations. Risk factors identified were height, preoperative anemia, preoperative hypoalbuminemia, preoperative hypokalemia, site of operation (airway and cardiac), non-official time (4:30 pm – 8:30 pm), operation time > 3 hours, and use of any muscle relaxants. Friedman et al. from New York Presbyterian-Columbia University Medical Center in New York, performed a retrospective study to identify risk factors for prolonged intubation (> 24 hours) after multi-level spine surgery (A1502). A multivariate analysis suggested that longer anesthetic times (677.8 ± 102.5min) and more blood product administration (1799 ± 1924.8cc) were associated with prolonged intubation times. Henneman et al. from the Massachusetts General Hospital reviewed 57,100 surgical cases requiring intubation during a 4.5 year period to investigate whether neuromuscular blockade (NMBA) is associated with adverse postoperative respiratory events. Results suggested that NMBA was associated with an increased risk of hypoxic events after extubation (OR-1.49 95% CI: 1.36-1.62), and an increased risk of reintubation/unplanned ICU admission (OR-2.12 95% CI: 1.71-2.63). The use of neostigmine was independently associated with an increase in hypoxic events (OR-1.09 95% CI: 0.98-1.21). Neuromuscular monitoring was documented in only 50% of patients receiving intermediate-acting NMBA (A437). The relationship between anesthesia time and risk for postoperative pulmonary complications (PPC) in patients undergoing general anesthesia for orthopedic surgery was examined in abstract #1498. Out of 162,247 discharges, 8,966 patients developed a PPC (5.53%). This study suggested that a 15-minute interval increase in anesthesia time was associated with an 8% increased risk of PPC, a $974 increase in average total hospital cost and 3.1 hours in mean total length of stay.

beachchair

Rohrbaugh et al. from the University of Pittsburgh (A235) examined 13,512 cases of shoulder surgery in the beach-chair position (99% performed under interscalene block and propofol sedation) over a 9.5-year period. The authors identified 37 total adverse events. All of these events were rare (occurred in < 0.07% of cases) and included emergency airway intubation, acute respiratory distress without need for intubation, seizures, persistent nerve injury, CNS injury (or stroke within 24 hours of surgery), cognitive dysfunction, headache, myocardial infarction, dysrhythmia, hypotension, drug reaction, or unexpected admission. Investigators examined 136,371 moderate-high risk surgical patients over a 5-year period to determine whether postoperative troponin I values predict 30-day in-hospital mortality (A1578). Among 9516 cases where troponin I levels were measured, 912 patients had troponin elevations. Approximately 80% of troponin I elevations occurred during the first 3 postoperative days. Forty-two percent of all deaths were associated with elevations in troponin levels.

Gan et al. from Duke University Medical Center performed a retrospective database study of 18,961 non-cardiac surgical procedures to investigate whether “Triple Low,” (low bispectral index (BIS), low mean arterial pressure (MAP), and low anesthetic concentration) increases postoperative mortality (A1574). The authors suggested that the “triple low” combination was associated with a 2.5-times increased risk of 1-year mortality compared to those patients with normal values. Investigators (A003) reported patient experiences with awareness who had general anesthesia and those that had not received general anesthesia (GA). Among 183 patients who had enrolled in the Registry with awareness experiences from 1990 or later, most respondents had psychological sequelae related to awareness regardless of anesthetic type (88% receiving GA, 65% receiving non-GA). Patients in both the non-GA and GA groups reported paralysis even when the medical records indicated that neuromuscular blockade was not given. The authors suggest that improved communication and education may help patients who are not receiving GA understand that some degree of patient awareness should be expected. Bhavani et al. (A1024) at the Cleveland Clinic performed a retrospective review of all patients undergoing spine surgery from 1995-2010 at their institution to identify the incidence and associated risk factors for postoperative visual loss. Out of 2532 potential controls, 6 cases of visual loss were identified. Cases with visual loss had a significantly greater blood loss (P=0.002) and a greater amount of red blood cells transfused (P=0.006). No other intraoperative risk factors were identified.

Postoperative Delirium

Several abstracts this year addressed the incidence of and associated risk factors for the development of postoperative delirium (POD) and cognitive dysfunction (POCD). Investigators (A1617) utilized a large database of inpatient surgical discharges (Premier Perspective Database®) to examine the incidence, risk factors, and cost related to postoperative cognitive complications (POCC). Of 1,043,647 inpatient surgical discharges, 1% of the patients had a diagnosis of POCC. Episodes of POCC were associated with a significant increase in hospital mortality, mean total cost, and mean length of stay. The risk of developing POCC was associated with male gender, Caucasian, diseases of the Charlson comorbidity index, emergency admissions, ICU stay, and general anesthesia. Brewbaker et al. from Wake Forest University School of Medicine performed a retrospective study to determine the prevalence of postoperative delirium (POD) among patients undergoing hip fracture repair. Out of the 72 patients included in the review, 22 patients (30.6%) showed strong evidence of POD (A1527). Abstract #1497 evaluated the influence of delirium on mortality and quality of life 6 months after carotid endarterectomy. This prospective observational study evaluated 70 patients admitted to the post-anesthesia care unit; delirium was assessed using the Intensive Care Delirium Screening Checklist (ICDSC). Seventeen percent of the patients developed POD. Mortality rates at 6 months were higher for patients with POD (25% vs. 3%, p=0.023). However, POD did not influence quality of life at 6 months after surgery. Wagner et al. from Vanderbilt University examined 200 consecutive patients from the Cardiovascular Intensive Care Unit (CVICU) to determine the prevalence and risk factors associated with the development of delirium. The overall prevalence of delirium was 26%. The duration of delirium was 0.5 ±1.1 days. Risk factors associated with an increase rate of delirium included use of statins, dexmedetomidine, benzodiazepines, and physical restraints (A088). Another study investigated the incidence and risk factors related to POD (A083). Out of 775 adult patients admitted to the post-anesthesia care unit (PACU), 128 patients developed POD (18.9%). Patients with delirium were more severely ill, had longer hospital and PACU stays, and had higher mortality rates. Independent risk factors for delirium included age, ASA physical status, emergency surgery, and the total amount of FFP administered during surgery (A083).

Chapman et al. (A1516) from Duke University performed a prospective study of 1274 patients undergoing non-cardiac surgery to determine whether obstructive sleep apnea (OSA) predisposes patients to postoperative cognitive dysfunction (POCD). The incidence of POCD in this study was 35%. A multivariate analysis suggested that age > 50 years old was a predictor of POCD at hospital discharge and 3 months after surgery. A BMI > 35 was also a predictor of POCD at one week. Subjects who had a positive STOP-Bang questionnaire (suggestive of those with OSA) had higher 1-year mortality. However, the study could not definitively conclude whether there was an association between OSA and POCD. Behrends et al. from UCSF in San Francisco, California, utilized an internal database in older patients undergoing major non-cardiac surgery to determine whether blood transfusion was associated with early POD in the elderly (A1090). Of 577 patients examined for POD on postoperative day 1, 31.9% developed POD. A multivariate regression analysis suggested that age and blood transfusion were independent risk factors for early POD in older patients. Larger amounts of transfusion were associated with a further increase in the risk of developing early POD.

Use of Etomidate & Outcomes

Sunshine et al. from the University of Washington performed a retrospective study involving 824 mechanically ventilated patients to determine whether etomidate administration is associated with an increase risk of hospital mortality in the critically ill (A089). After adjusting for age, gender, simplified acute physiology score (SAPS II), the relative risk (RR) of death among etomidate recipients was 20% higher than that of patients given an alternative agent. Another study (A638) analyzed 329 postoperative cardiac surgical patients who had a cortisol level and/or corticotrophin (ACTH stimulation test) drawn during a 2-year period. Adrenal insufficiency occurred in 43.4% of patients, and etomidate was given to 57% of the patients. Etomidate use was associated with a significant increased risk of adrenal insufficiency (53.7% of recipients vs. 29.4% of non-recipients). In a multivariate analysis, etomidate was the only independent risk factor for developing adrenal insufficiency (OR-3.05).

Obstructive Sleep Apnea (OSA)

maskSeveral abstracts investigated perioperative issues facing patients with OSA or at risk for OSA. Sharma et al. (A1501) from the University of Buffalo, New York, reviewed 3,593 patients undergoing surgery under general anesthesia and indentified 306 patients who were at high risk for OSA (HR-OSA). During the postoperative period, HR-OSA patients had a higher incidence of hypoxia, reintubation, and postoperative use of CPAP. This group also had a longer PACU/ hospital length of stay and had an increase in overall postoperative complications. Mehta et al. from the Toronto Western Hospital, (A223) performed a systematic review of the literature to define the incidence of postoperative complications in patients with OSA versus those without the disease. Of the 12 studies selected, 10 reported significantly higher postoperative complications in OSA patients than those without OSA. Mehta et al. also performed a retrospective study to determine the long-term health benefits of screening patients for OSA in a preoperative clinic (A038). The investigators contacted 156 patients, and 82% of these patients had OSA established by polysomnography. Sixty-nine percent of these patients were prescribed continuous positive airway pressure (CPAP), but only 45% of these patients were compliant with using CPAP. The CPAP compliant patients had a greater reduction in medication dosage for comorbidities than the non-compliant group. Eight percent of the study population had long-term health benefits from OSA screening and compliance with CPAP (A038). Abstract A224 investigated whether postoperative oxygen therapy can improve oxygen saturation without worsening OSA. Out of 168 patients, 58% were on oxygen therapy on the first postoperative night. Compared with those without O2 therapy, those that had oxygen applied had a significantly higher SpO2 and a lower oxygen desaturation index. Those patients who had oxygen therapy also experienced a lower time percentage with SPO2 < 90% (CT90). However, the average duration of apnea/hypopnea episodes was longer in the patients who had O2 therapy.

Miscellaneous

Kawashima et al. (A186) from Wakayama Medical University in Japan performed a systematic review to investigate the anti-emetic effect of perioperative fluid loading on postoperative nausea and vomiting (PONV). Eight randomized controlled studies (851 patients) were selected. Results suggested that perioperative fluid loading of crystalloids (1000 ml or more) improved PONV (OR-0.48) and reduced the use of anti-emetics when compared with control groups. Subramaniam et al. (A787) from the Beth Israel Deaconess Medical Center in Boston, MA, enrolled 986 patients undergoing cardiac surgery in a prospective observational trial to investigate whether a preoperative hemoglobin A1C (HbA1C) level may be associated with postoperative glycemic variability and adverse outcomes. A multivariate analysis suggested that the HbA1C ≥ 6.5 gm/dl was associated with a higher risk of perioperative complications (OR-1.7). Glycemic variability was also significantly greater in the HbA1C ≥ 6.5 gm/dl. Lastly, Abrishami et al. from the University of Toronto, performed a meta-analysis to determine the minimal period of smoking cessation before surgery that could reduce postoperative pulmonary complications (PPC) (A227). While 62 articles were analyzed, only 13 papers (15,297 patients) reported results on different times of smoking cessation. Results suggested that short-term smoking cessation prior to surgery (2-4 weeks) was not associated with an increased risk of PPC when compared to current smokers. However, smoking cessation for > 4 or 8 weeks was associated with a 23-47% decreased risk.

This brief review summarized only a small number of abstracts on patient safety presented at the 2011 Annual Meeting. This is not an endorsement of the methods, results, or conclusions of any particular abstract. 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.

Dr. Greenberg is Director of Critical Care Services, Evanston Hospital and Co-Director for Resident Education Department of Anesthesia NorthShore University HealthSystem and Clinical Assistant Professor, Department of Anesthesiology Critical Care University of Chicago, Pritzker School of Medicine.

Dr. Vender is the Harris Family Foundation Chairman Department of Anesthesia / Critical Care Services and Vice President, Physician & Programmatic Development at NorthShore University HealthSystem and Clinical Professor Anesthesiology University of Chicago Pritzker School of Medicine.