Introduction
Critical care is the area where patients are sickest and where the highest costs are incurred (Thungiaroenkul & Kunaviktikul, 2006). But, at a time when intensive care units (ICU) are rapidly adopting electronic medical record systems (AHIMA, 2011; Office of the National Coordinator for Health Information Technology [ONC], 2011), little has been done to examine the effect of HIT on the workflow and performance of the ICU nurse. There is an evolving awareness that HIT is bringing with it extraneous non-care-related activities or “HIT barriers” which add to the RN workload, and it is known that in response to these actual or perceived barriers to workflow, nurses have developed “workarounds” (Debono, Greenfield, Black, & Braithwaite, 2011; Halbesleben, Wakefield & Wakefield, 2008; Tucker & Edmondson, 2002). Workarounds are described as a nursing response when mismatches between HIT and the nurses’ work-patterns occur (Tucker, 2010; Tucker & Edmondson, 2002).
While workarounds have been described, they are poorly defined and understood. Current thinking identifies some type of barrier or perceived barrier as a precursor to the workaround (Ferneley & Sobreperez, 2006; Tucker, 2004). A preponderance of literature cites workarounds as safety risk behaviors (Alper & Karsh, 2009; Halbesleben et al., 2008). Workarounds can bypass safety mechanisms (Koppel, Wetterneck, Telles, & Karsh, 2008) and interfere with organizations’ attempts at standardization (Ferneley & Sobreperez, 2006).
Despite evidence supporting the potential of HIT to improve the productivity of nurses and reduce medication errors, there is evidence of HIT safety and risk trends (American Health Information Management Association, 2011; Institute of Medicine, 2011; ONCHIT, 2011). It is also becoming apparent that as organizations layer HIT policies and procedures on registered nurses (RNs) in an attempt to meet regulatory requirements and divert error, there is potential for these small protocol-based behavior changes to cascade collectively, disrupting and compromising the integrity of the system it was trying to protect (Haraldsson, Sverdrup, Belyazid, Holmqvist, & Gramstad, 2008).
The call for more research into the use of HIT is particularly important to front line critical care nurses because it is estimated that 37% of nurses in hospitals work in critical care (American Health Information Management Association [AHIMA], 2011). In 2008, The U.S. Department of Health and Human Services (2010) reported that there were 328,392 critical care nurses practicing in U.S. intensive care units. One hour of misused time per nurse would equate to 13,683 non-productive days or 37 lost years. The identification of HIT performance obstacles, nursing workflow and workaround behaviors is an underexplored phenomenon which, if better understood, is a potential source of nursing workflow, work efficiency and patient outcome improvements. This paper presents findings from a mixed methods dissertation study that explored a proposed Health Information Technology Workaround Model (HITW). The confirmation of preliminary workaround definitions and hypothesized relationships are presented.
Background
A number of shortcomings have been identified in researching a nurses’ interactions with HIT. Prior studies on workarounds have relied on known predictors of workarounds (i.e, workflow blockages) and nurse’s verbal reports of causes (technology malfunctions). While taking advantage of existing research is considered a cornerstone of scientific rigor, utilizing a priori variables may limit the identification of other, equally important variables. Finding ways to allow the data to uncover new variables, patterns and relationships can support the development of more accurate frameworks for study.
A second issue with workaround research has been the linear research approach to investigation. Often, healthcare systems and workflows are viewed mechanistically whereby specific parts interact in a linear manner to produce definitive, predictable outcomes (i.e., workflow diagramming). In the United States, Donabedian’s structure-process-outcome linear model is the foundation from which quality outcomes are assessed; however, workaround variables behave in precisely the manner that is expected in a complex adaptive system, not a linear model (Donabedian, 1980; Dykes & Collins, 2013). In linear systems, there are predictable outcomes from actions, but in non-linear systems the same action could result in a multitude of possible outcomes (Lipsitz, 2012).
Complex systems interactions will produce unforeseen results (Lipsitz, 2012). This would help to explain the unexpected consequences that can result between nursing activities and HIT. Complex system science is not commonly found in nursing research, but is being explored in such fields as leadership and management, military analysis, and agricultural sciences. (Green, 2011; Moore, 2009; Rammel, Stagl, Wilfing, 2007; Schneider & Somers, 2006). Taking a cue from other professions, combining qualitative and quantitative methods in order to explore the behaviors of complex systems can result in data driven variable identification and more accurate workaround research models.
Research from a linear perspective assumes the rules of cause and effect. If components of a system are understood and improved, then the entire system will work better. Complexity theory however assumes that systems have complex relationships and patterns that are highly changeable with emerging patterns and iterations. From this perspective, nursing work and behaviors are influenced by many different agents that learn and are non-linearly interdependent. As a result, traditional quantitative research methods may be unable to accurately describe nursing behaviors. For example, in a complex system, variables can act as independent, dependent or both simultaneously. (McDaniel) Finally, when exploring the relationships between workarounds and patient safety events, researchers often select outcomes that can be quantified (i.e., bar code scanning, # patient falls). By relying on only quantifiable measures, we may be excluding patient safety risks and nursing workaround behaviors.
Purpose
The purpose of this paper is to present general findings of a nursing workaround study conducted in collaboration with the American Association of Critical Care Nurses (AACN). The preliminary workaround definitions were developed in a pilot study with the San Antonio Chapter of AACN. This article will present the confirmed workaround definitions with examples of the types of workarounds used by critical care nurses. Relationships between workload, turbulence and patient safety will be explored and reasons for workarounds will be presented.