Four APSF Grants Awarded for 2014

Steven K. Howard, MD

The APSF’s mission statement explicitly includes the goal to improve continually the safety of patients during anesthesia care by encouraging and conducting safety research and education. Since 1987, almost $9 million has been provided to investigators for patient safety research and the field of anesthesiology continues to be a shining example in health care in this area. In 2013, the APSF investigator-initiated grant program had the second greatest number of grant submissions in its 28-year history—a total of 45. We are pleased that there continues to be such an enthusiastic interest in the study of patient safety.


Scott C. Watkins, MD Vanderbilt University, Department of Anesthesiology

Over the summer, members of the Scientific Evaluation Committee (SEC) provided reviews of this year’s grant submissions and from these reviews a subset was chosen for further discussion during a convened meeting of the SEC on October 12, 2013, at the ASA National Meeting in San Francisco, CA. Of the 8 finalists, 4 were recommended to and approved by the APSF Executive Committee for funding. The principal investigators of this year’s APSF grant awardees provided the following description of their proposed work:

Dr. Watkin’s submission is titled “The Effect of Technical and Non-Technical Decision Support Tools on Team Performance in Simulated Perioperative Pediatric Crises.”

Background: During critical events, clinicians routinely deviate from evidence-based standards and omit critical actions when they depend upon memory alone. Reasons for this are thought to be multi-factorial in nature including decay in technical skills (TS) and knowledge over time and the negative influence of stress on performance during high-stakes events. There is also increasing evidence that failures in team-based non-technical skills (NTS) including inadequate leadership, failing to assign roles, poor task distribution, inadequate planning, and broken communication, contribute to poor adherence to guidelines and treatment algorithms by clinicians. These deficits in NTS lead to failures in the transfer of medical knowledge into appropriate clinical actions.

Efforts to improve clinician performance with the use of cognitive aids containing reminders of TS have yielded mixed results with some demonstrating improved, but not perfect, adherence to guidelines and others demonstrating no change and even potential harm. To understand the gap in team performance in critical events, research has turned towards assessing the impact of human factors and NTS on performance, such as using alternative methods of delivering information to providers and incorporating electronic prompts and real-time feedback into cognitive aids. The management of critical events requires clinicians to utilize both TS and NTS; thus measures aimed at improving clinician performance, such as cognitive aids or decision support tools, should focus on both skill sets.

Aims: The aim of this study is to assess the impact of different versions of an electronic decision support tool (e-DST) on team performance as compared to memory alone. The 3 versions of the e-DST include 1) prompts for TS only, 2) prompts for NTS only, and 3) prompts for both TS and NTS. The performance of pediatric operating room teams in the management of 4 in situ simulations of perioperative emergencies using the 3 versions of the e-DST and 1 using memory alone will be assessed by adherence to evidence based guidelines and by the Mayo High Performance Teamwork Scale (MHPTS). The inter-disciplinary operating room teams will consist of personnel from anesthesia, surgery, nursing, and allied health services. We hypothesize that the e-DST that emphasizes both non-technical skills (NTS) and technical skills (TS) will significantly improve team performance as measured by adherence to evidence based guidelines and the MHPTS.

Implications: The work described in this proposal will assess the impact of non-technical skills (NTS) on inter-disciplinary team performance in simulated events as defined by adherence to evidence-based guidelines. Furthermore, it will allow us to refine the way DSTs are designed and improve our understanding of how to educate and train interdisciplinary medical teams for high-stakes events. It will explore the impact of NTS on team performance as guided by a clinical decision support tool. The e-DST implemented in this study is also designed to aid teams in accurately documenting what occurred during the event and to allow video capture of team performance for improved post-event debriefing.

Funding: $149,892 (January 1, 2014-August 31, 2015). This grant was designated as the APSF/ASA President’s Research Award. Dr. Watkins is also the recipient of the Ellison C. “Jeep” Pierce, Jr., MD, Merit Award which provides an additional unrestricted amount of $5,000.


Karthik Raghunathan, MD, MPH Duke University Medical Center, Department of Anesthesiology

Dr. Raghunathan’s project is titled “Comparative Safety of Different Types of IV Fluids for Resuscitation in the OR and ICU: An Applied Pharmacoepidemiologic Approach.”

Background: Intravenous (IV) fluids are used routinely to maintain euvolemia and to correct overt or presumed hypovolemia during major surgical procedures in operating rooms (ORs) and in intensive care units (ICUs). Among ICU populations, meta-analyses of "crystalloid versus colloid fluid" trials suggest equipoise in efficacy but certain colloids (hydroxyethyl starches) have been associated with harm. However in ORs, the use of colloid solutions has been associated with improved outcomes in moderate and high-risk surgical patients (contrasted against conventional crystalloid therapy). More recently, studies have linked safety outcomes to the chloride content of commonly used solutions. When compared with chloride restrictive ("physiologic") balanced fluids, resuscitation with chloride liberal solutions (such as isotonic saline) appears to be associated with adverse consequences. Hence, fluids may be categorized along distinct axes with specific patterns of exposure in the OR and ICU potentially leading to different clinical safety outcomes. Currently, there is a natural "quasi-experiment" in progress as patterns of IV fluid use vary influenced by institutional or physician preferences. With such significant unwarranted variability, patients with similar clinical conditions may receive different qualitative and quantitative exposures making a pharmacoepidemiologic study of fluids feasible and desirable.

Aims: Using the nation’s largest inpatient drug utilization database (maintained by Premier Inc., Charlotte, NC), we propose to study the comparative safety of different types of IV fluids used in ORs and ICUs. As varying combinations of fluids are typically used, exposure may be defined based on the relative proportions of balanced fluids and/or colloids used on the day-of-surgery and in the ICU. Contrasting fluids primarily based on their chloride content, we will test the hypotheses that exposure to increasing proportions of "chloride liberal" fluids (e.g., isotonic saline) results in greater in-hospital mortality, morbidities (renal failure, infections, bleeding), and higher treatment costs when compared to the use of larger proportions of balanced fluids. Further, we will analyze heterogeneity of effects across different surgical procedures and patient populations. Premier Inc.’s large quality-assured database contains a prospectively collected itemized date-stamped log of patient-level charges including granular data on patient and hospital demographics, comorbidities and co-treatments (such as fluid type and amount), co-medications, diagnostic tests, and therapeutic services from over 500,000 patients across hundreds of acute care facilities nationwide. We plan to conduct propensity-score matched analyses creating comparable subgroups that differ in fluid exposures. In addition, as fluid choice is often influenced by where care is received rather than on specific patient or procedural characteristics, there is potential for analysis based on hospital-preference for specific fluids that emulates "random" treatment allocation. Such methods estimate associations while minimizing bias, a major limitation when using such secondary data. Finally, we will use network meta-analysis to combine estimates from observational data with other studies in the literature.

Implications: Fluid choice during major surgery and in ICUs may have enormous public health implications if the outcomes associated with commonly used combinations of fluids vary significantly. Our proposed study will re-orient the fluid debate using representative real-world national data and results will complement and inform future randomized trials.

Funding: $94,680.30 (January 1, 2014-December 31, 2015). This grant was designated as the APSF/ASA Endowed Research Award.


Karl Hammermeister, MD University of Colorado School of Medicine/ Colorado Health Outcomes Program

Dr. Hammermeister’s grant is titled “Neuromuscular Blockade and Perioperative Outcomes.”

Background: Residual postoperative neuromuscular blockade (NMB) is common with a reported incidence of between 4 and 50%. Murphy and Brull concluded in a recent review, “… residual neuromuscular block is an important patient safety issue and that neuromuscular management affects postoperative outcome.” In 2012, Grosse-Sundrup et al., reported a significantly increased risk of reintubation with return to the intensive care unit within 7 days of surgery in patients who received an intermediate acting NMB agent matched to an equal number who did not by propensity score. Qualitative monitoring of neuromuscular transmission did not decrease this risk, and neostigmine reversal increased the risk of reintubation. Cholinesterase inhibitors, such as neostigmine, are commonly used to reverse NMB at the conclusion of surgery; however, they have significant side effects. More importantly, they may actually increase NMB by creating very high concentrations of acetylcholine at the neuromuscular junction, which has an antagonistic effect.

Aims: The specific aim of this proposal is to test the primary hypothesis that intraoperative administration of a non-depolarizing NMB agent is associated with an increase in one or more of the following respiratory complications: failure to wean from the ventilator within 48 hours following surgery, reintubation within 30 days following surgery, or postoperative pneumonia within 30 days following surgery. We will also examine the effects of NMB on 30-day postoperative non-respiratory complications, all-cause mortality, and length of stay.

To accomplish this, we will analyze an existing data set containing patient-related risk factors, operative data, 30-day mortality and morbidity, and late survival on more than 19,000 major surgical procedures performed in VA medical centers between 1/1/2001 and 9/30/2005. We have merged this data set as part of an ongoing study of intraoperative predictors of adverse outcomes. This data set comes from 3 sources: 1) preoperative, intraoperative, and outcomes data from the VA Surgical Quality Improvement Program (VASQIP); 2) intraoperative data from anesthesia information monitoring systems (AIMS) from 4 VA medical centers; and 3) long-term vital status data from the VHA’s vital status files. The VASQIP collects a standardized set of risk and outcomes data for the majority of major surgical operations performed in the VHA health care system as part of its ongoing quality assessment and improvement programs in surgery. We will use both propensity matching and multivariable risk-adjustment to minimize the inevitable selection bias that occurs in an observational study like this; therefore, the results cannot be considered to definitively demonstrate or exclude a causal relationship between NMB and adverse surgical outcomes.

Implications: If this study shows a relationship between NMB and adverse postoperative outcomes, it should lead to the following: a greater caution in the use of intraoperative NMB, the development of better and wider application of postoperative monitoring for residual NMB, serve as a stimulus for a randomized trial comparing the common use of cholinesterase inhibitors and NMB that are reversed by other mechanisms.

Funding: $148,802 (January 1, 2014-June 30, 2015). The funding was made possible by a grant from Covidien and is designated the APSF/Covidien Research Award.


Deborah Culley, MD Brigham and Women’s Hospital, Harvard Medical School, Department of Anesthesiology, Perioperative and Pain Medicine

Dr. Culley’s grant is titled “Preoperative Cognitive Status in Elderly Surgical Patients: Feasibility of Routine Screening and Utility for Predicting Morbidity.”

Background: Anesthesiologists have been at the forefront of preoperative evaluation and patient safety initiatives. Thus, it is surprising that the function of one of the most vital organs—the brain—has been largely neglected. There are multiple reasons why preoperative cognitive screening in elders might be valuable. Elders have more surgical procedures than middle-aged adults; a high prevalence of undetected cognitive impairment; and high rates of postoperative morbidity and mortality. Pre-existing cognitive impairment is a known risk factor for postoperative cognitive morbidity such as delirium, which is associated with longer length of hospital stay, a higher likelihood of discharge to a place other than home, and greater 1-year mortality. Equally important but much less well recognized is that pre-existing cognitive impairment may be an independent predictor of serious in-hospital non-cognitive morbidity such as in-hospital falls. This suggests that poor preoperative cognitive function has major quality and safety implications and is likely to be an important, but largely unrecognized, determinant of morbidity and mortality and cost of care for geriatric surgical patients. The goal of this study, therefore, is to cognitively stratify elective geriatric surgical patients preoperatively and to determine whether cognitive status predicts adverse cognitive or non-cognitive events.

Aims: Our overriding hypotheses are that cognitive impairment is 1) common in elders presenting for elective orthopedic surgery and better identified with a brief structured cognitive screen than standard practice; and 2) an independent predictor of adverse postoperative events. In addition, we will test whether it is feasible to implement routine structured preoperative cognitive screening of elders in a busy preoperative clinic. Accordingly, based on preliminary data and a power analysis, we will recruit and preoperatively cognitively stratify 211 patients 65 years of age or older who are scheduled for lower extremity joint replacement surgery. Cognitive screening will be performed with a standard and widely accepted instrument that, in a pilot study, took 3 min or less to administer and proved reliable and easy to score. The primary cognitive and non-cognitive outcomes will be, respectively, delirium and discharge to a place other than home.

Implications: This work has considerable clinical implications. The ability to identify and risk stratify patients with pre-existing cognitive impairment may improve the quality of shared decision making between patient/family and physician; influence decisions about appropriateness of surgery and anesthetic management; and/or optimize allocation of scarce perioperative resources known to improve outcomes (e.g., postoperative geriatric care units). Indeed, assuming preoperative cognitive impairment is common, routine preoperative cognitive screening promises to be an excellent low-cost, patient centered, high impact proposition for enhancing surgical outcomes.

Funding: $149,997 (January 1, 2014-December 31, 2015).

Finally, the members of the SEC would like to thank all of the investigators who submitted their proposals to the APSF for this grant cycle. We continue to encourage submission of well-designed studies of safety-related clinical research as well as research on education and training in patient safety. It is through the mechanism of this research that we continue to strive for the vision that “that no patient shall be harmed by anesthesia.”

Dr. Howard is Staff Anesthesiologist at the VA Palo Alto HCS and Associate Professor of Anesthesia Stanford University School of Medicine as well as Chair of the APSF Committee on Scientific Evaluation.