Disappearing expertise in clinical automation: Barcode medication administration and nurse autonomy

Abstract

Objective

Using the case of barcode medication administration (BCMA), our objective is to describe the challenges nurses face when informatics tools are not designed to accommodate the full complexity of their work.

Materials and Methods

Autonomy is associated with nurse satisfaction and quality of care. BCMA organizes patient information and verifies medication administration. However, it presents challenges to nurse autonomy. Qualitative fieldwork, including observations of everyday work and interviews, was conducted during the implementation of BCMA in a large academic medical center. Fieldnotes and interview transcripts were coded and analyzed to describe nurses’ perspectives on medication safety.

Results

Nurses adopt orienting frames to structure work routines and require autonomy to ensure safe task completion. Nurses exerted agency by trusting their own judgment over system information when the system did not consider workload complexity. Our results indicate that the system’s rigidity clashed with adaptive needs embodied by nurses’ orienting frames.

Discussion

Despite the fact that the concept of nurse as knowledge worker is foundational to informatics, nurses may be perceived as doers, rather than knowledge workers. In practice, nurses not only make decisions, but also engage in highly complex task-related work that is not well supported by process-oriented information technology tools.

Conclusions

Information technology developers and healthcare organization managers should engage and better understand nursing work in order to develop technological and social systems to support it.

INTRODUCTION

In her 1994 American Anthropological Association paper presentation, Diana Forsythe discussed the problem of “Disappearing Women in the Social World of Computing.”¹ For Forsythe, a fundamental issue among the computer scientists she studied was the assumption that their science was culturally neutral—a notion that social scientists who focus on science and technology have repeatedly shown to be wrong.^2,3 Technology design embodies cultural assumptions about the world.³ Focusing her talk on the field of biomedical informatics, Forsythe argued that women, their knowledge, and their expertise are ignored through “ongoing cultural work” of mostly male informaticians. From her fieldwork, Forsythe offered as evidence the comments of a senior physician working to develop a list of relevant medical information sources. Included in the list were doctors, research literature, databases, patients and families, and communities. When Forsythe asked, “What about nurses?” the informatician replied, “Nurses don’t count.”¹

Forsythe continued conceptualizing this problem further in two papers^4,5 that both draw on Star’s notion of “deletion” in scientific practice, that Forsythe describes as “instances in which particular pieces or types of work are rendered invisible.”^4,6 With this article, we argue that fundamental assumptions underlying the design of traditional electronic health record (EHR) systems delete important aspects of nursing work. One result of this deletion is an observable loss of autonomy in nursing practice. We conclude with recommendations for both technology developers and healthcare managers to recognize and support the complex work of nursing.

Background and theoretical framework

Nurses are essential healthcare providers who work closely with patients, families, and physicians and carry out, among other processes, the important task of administering medication. The key goal of patient safety is often accomplished by maintaining the practice of the “rights” of medication administration, including the “5 rights” (right medication, dose, time, route, and patient), with the more recent additional rights of “drug form” and “response.”^7,8

Autonomy is a key element of nursing practice,⁹ encompassing patient advocacy^10,11 and the qualification and accountability for decision making related to direct patient care.^9,10,12 For the purpose of this article, nurse autonomy is defined as the professional capacity for decision making based on knowledge and clinical judgment, and within the full scope of nursing professional practice.¹³ Nurse autonomy is illustrated in the organizational structure and pattern of nurses’ daily medication administration work by determining the safety of a medication or action while also ensuring efficient workflow for individual patients and the entire unit. Previous studies demonstrate the importance of nurse autonomy in nurse satisfaction and quality of care,¹⁴ and patient advocacy.¹⁵

With the complex environment of new technologies, personalized therapeutic interventions, and gaps in nurse supply and experience,¹⁶ medication safety requires nurses to think beyond the “5 rights” and use the nursing process to prepare for situations that may compromise patient care. The interplay between autonomous nursing practice and the medication administration “rights” can yield insight into how nurses approach their workload while still acting as patient advocates. Here, we examine nurses’ perspectives on medication safety and the implementation of a barcode medication administration (BCMA) system in an inpatient unit to describe its impact on medication safety practices and nurse autonomy.

Nurse autonomy manifests in preexisting organizational systems of accountability that outline scope of practice and include setting-specific policies and procedures regarding patient care, healthcare workers’ expectations of each other, and technology. The interplay of these systems produces situations that require nurse autonomy in decision making to ensure patients’ needs are safely met. For example, when a patient is being emergently treated but has not been through the admission process and therefore has no electronic chart, nurses are still responsible for administering and documenting urgently needed medications.

Hazlehurst and McMullen’s¹⁷ concept of “orienting frames” provides a useful lens to understand the relationship between autonomy and patient safety. An orienting frame is a “conceptual structure that is common to a group of actors who share a history of interaction”; orienting frames reflect the rationale for how nurses “organize, prioritize, conduct, and evaluate their work,” steps that correspond to the nursing process steps of assessment, diagnosis, planning, implementation and evaluation.¹⁷ Orienting frames incorporate both the organizational regulations and the individualized methods used to accomplish tasks or guard against adverse events. Accordingly, these frames facilitate nurse decision making, enhance medication safety, and are enacted by nurses using clinical judgment based on training, experience, and professional relationships. In their study of intensive care unit nursing practices, Hazlehurst and McMullen¹⁷ identified 3 key frames: “being prepared for emergencies,” “being organized,” and “being responsible and accountable” that each demand autonomy, especially within the dynamic healthcare environment.

BCMA technology

Healthcare organizations adopt technologies to ensure patient safety, such as clinical decision support (CDS), computerized provider order entry, or BCMA technologies.¹⁸ Specifically, BCMA is intended to promote patient safety by double-checking the “5 rights of medication administration,” potentially reducing medication error.¹⁹ It uses patient-specific barcoded wristbands and barcoded medication containers to electronically ensure the correct medication, dose, and route are present at the right time for the right patient.

The addition of new technological systems may enable, constrain, or alter nurses’ daily routines and impact patient safety. Adjusting regular activities to meet a computerized system’s structure can alter decision-making authority and other prior working arrangements.²⁰ Previously identified negative effects of BCMA include “nurses dropping activities to reduce workload during busy periods,” “increased prioritization of monitored activities during goal conflicts,” and “decreased ability to deviate from routine sequences.”²¹ As such, a BCMA system’s design can diminish nurse autonomy by altering the range of actions available in medication administration.¹⁸ This study aimed to identify how BCMA may accommodate or compromise existing cultural structures, and how nurses adapt and respond to such changes. This report focuses on how nurse autonomy can be altered in the face of technology-induced policy and practice changes.

MATERIALS AND METHODS

The project was approved by the institutional review board for the research site, a tertiary academic medical center. The project took place from July to November 2011. We conducted a qualitative study in an inpatient unit that included 20 hours of observation and 9 nurse interviews. We conducted an additional 18 interviews with nurses, recruited via an email campaign, from a variety of clinical settings, including inpatient, outpatient, and research. See Table 1 for participant training and experience characteristics. As a group, the participants reported having current or previous experience in a total of 27 different areas of practice. For the qualitative study, 3 trained researchers (L.L.N., C.L.S., J.Y.H.) recorded field notes on an acute care cardiology step-down unit using notepads or iPads. Field notes documented unit activities, focusing on nurse workflow, the process of safe medication administration, and interactions with the BCMA system. We accompanied nurses through medication management activities, asking about the processes of patient care and medication distribution to better understand daily activity on the unit, nurses’ task prioritization, and general perspectives on medication safety.

Table 1.

Training and experience of participants

Training	Experience					Unknown	Observed and Interviewed	Interview Only	Total
	0–5 y	6–10 y	11–20 y	21–30 y	31+ y
LPN	1	0	0	1	1		3		3
RN	6	1	1	1	1		6	4	10
RN, BSN	5	2	2	0	0	1	5	5	10
RN, MSN	0	0	1	0	1		2		2
RN, MS (other)	0	1	0	0	1		2		2
Total	12	4	4	2	4	1	18	9	27

BSN: bachelor of science in nursing; LPN: licensed practical nurse; MS: master of science; MSN: master of science in nursing; RN: registered nurse.

Interviews took place in a private location. Probes focused on themes that arose in the observational data, including: “What does medication safety mean to you?”; “What do you do to ensure medication safety?”; “How does patient turnover rate affect medication safety?”; “How do you deal with complexity in patient care?”; and “What role do you think patient education has in medication safety?” Interviews ranged from 19 to 30 minutes. Each interview was recorded and then transcribed.

Data were analyzed using the qualitative data analysis tool Dedoose (Dedoose, Manhattan Beach, CA). All field notes and transcripts were imported into the tool. The data were coded by 3 trained coders using an open coding approach.²² Two of the researchers (L.L.N., C.L.S.) who analyzed the data were experienced organizational ethnographers; the other (J.Y.H.) was a student who had received training on rigorous methods of data analysis. In this process, text excerpts were labeled with 1 or more codes, reflecting themes represented in the excerpts. The 3 researchers collaborated on the development and definition of the codes. Autonomy emerged as a key theme. We searched the literature on nursing autonomy to better understand how to analyze the data. The concept of orienting frames¹⁷ provided a way to organize the data analytically, and we sought additional themes that helped illuminate the relationship between autonomy and BCMA.

RESULTS

Our findings highlight situations in which nurses preferred their own judgment over the BCMA system and how this choice adheres to previously established orienting frames. Over half the participants responded to the question “What does medication safety mean to you?” initially by summarizing the “5 rights of medication safety” (right patient, right drug, right dose, right time, and right route). Of these 5 rights, the ones mentioned more commonly were right patient, right medication, and right dose. One nurse added that ensuring medication safety was a key aspect of the nurse’s role:

“We play a huge role in making sure that the patient is safe and also making sure that the correct medication is written even by the physician. Errors can be made, and so we’re next in line to make sure that the patient is safe.”—Research nurse, reflecting on her experience in intensive care

The BCMA system verified these 3 rights by displaying an error if scanning the patient’s barcode did not produce an order for a medication that matched the medicine scanned. Nurse routines involved several actions to ensure patient safety: (1) scan the medication and patient; (2) attend to system alarms; (3) verify orders in the system as needed, by consulting the physician, pharmacy, or other nurses. Nurses needed to “be aware of orders coming in” and relied on familiarity with the patient’s history and medication schedule. One cardiology nurse reported: “This is on you to go into some sort of [medication reference] that you have access to and check to make sure if medications are compatible or not.” Thus, the BCMA system does not guarantee the correct medication for the patient has been ordered. Nurse autonomy and decision making are relevant to verifying the system’s instructions themselves, as the nurse assesses that the ordered medication is appropriate for the intended patient.

Hazlehurst and McMullen’s¹⁷ frames of “being prepared,” “being organized,” and “being responsible and accountable” were particularly relevant to our data. To account for their patients individually while maintaining unit workflow, nurses created their own set of practices to remain accountable, relying on their own knowledge rather than the BCMA system. Autonomy was essential for patient and nursing schedule accommodation, time management, and medication verification.

Being prepared

Nurses in the cardiology unit cared for up to 4 patients at a time, with each patient often on multiple medications. The unit experienced an especially high patient turnover rate (ie, frequent admissions to and discharges from the unit) that, when coupled with individual patient complexity, led to significant cognitive load on the nurses. Nurses had to be prepared for changes in patient schedules as they planned the distribution of medications around each patient’s availability. Nurses prepared for the schedules of each patient and communicated accordingly with other physicians and nurses.

When asked about working around patient schedules, one nurse replied:

“With patients going downstairs to the Cath lab, they pretty much always have to have their aspirin and Plavix, so if I know that my patient is going to the Cath lab, then I will make sure that I do that early in the morning…I always try to talk to the [provider] team and say, ‘Do you want me to give them this dose now, or do you think they can wait until the next dose,’ and so it’s all about communication.”—Cardiology Nurse 1

Similarly, another nurse reported keeping track of key pieces of information from previous nurses’ shift report.

“I start out in the beginning of the day and getting report, and I get a report that the patient’s going to stress test, and I’m like, “File that,” you know, in the back of my brain because you can’t give certain meds if they’re going to a stress test, right? Okay, so I know that, but the meds are always still ordered, by the way, that’s another thing… hardly ever do the doctors DC those orders for patients going for stress.”—Cardiology Nurse 2

Effective preparation for these situations therefore required nurses to directly apply their knowledge of patients’ histories and schedules as they planned medication administration activities.

Being organized

To manage the “when” and “where” of medication administration, nurses developed specific schematics in order to complete their work effectively. To be considered “on time,” the BCMA system required medications to be administered within 1 hour of the time specified in the order. Several nurses described grouping medications together because they would not be able to work effectively if they adhered to the BCMA medication schedule. For example, some nurses grouped all daily routine medications that were not time-sensitive into a specific medication distribution time: if medications were supposed to be distributed hourly from 7 am to 11 am, nurses would give the medications all at 8 am or 9 am. As one nurse explained,

“So [nurses] either block it forward or block it back, because [giving all medications at their scheduled times] is ridiculous, there’s no way.”—Cardiology Nurse 2

Nurses also remarked that to manage time, they prioritized medications according to urgency:

“If someone has a med that’s due at nine o’clock that’s a multivitamin and someone has an antibiotic due, then you know which one’s going to get priority, so I just…prioritize things like pain and safety issues…prioritize according to keeping patients comfortable and keeping them safe.” —Cardiology Nurse 3

Nurses who worked in oncology described the delicate balance they managed, given the many medications prescribed to patients. In this excerpt, a nurse describes her own system:

“I actually have a trick book. I call it my little bible. And I have a list of meds that most of my patients are on that cannot be taken together. So when I go through either a new patient med list or a patient med list that’s changing their chemo or a patient that I haven’t seen in 6 or 8 months, updating their med list, I take that bible in with me because all it takes is one slip. All it takes is “oh you’re on Compazine and Phenergan together.” That’s all it takes… and you’ve screwed them up.”—Oncology Nurse

Nurses also reported strategies for medication safety during peak workload and critical situations. In the following excerpt, the nurse describes using physical artifacts (medication vials and a paper towel) to keep track of meds given, because the BCMA system did not easily accommodate such events:

“That’ll happen a lot where you’re having a critical situation for us and you have to override a medication that they need. In that instance, it’s basically just like writing things down at the bedside. I usually keep every medication. Usually it’s going to be an IV medication you push, so I’ll keep the vials on the computer, because you don’t have time to go and chart it yet. You can’t even really scan it because it’s not in the system yet, so you just have gotten the verbal order and you’re just giving it based off of that. I’ll leave all the vials so I can scan them later and double check and make sure “okay—how many milligrams did I give? Well, there are three vials and they’re all empty, so I must have given all three of these.” I usually have a paper towel on the cart or something like that and just write down everything as I’m doing it. That way I can sit down and get reorganized later.”—Neurology ICU Nurse

Being responsible

Nurses did not solely rely on the BCMA system to ensure that the medication prescribed for the patient was accurate. Experienced nurses often double-checked orders placed by physicians, and if any discrepancy was present, contacted other nurses, physicians, or pharmacies for verification or clarification. When an order seems incorrect, nurses may seek out nursing colleagues to support or refute the appropriateness of an order before contacting the order’s author. One nurse recalled,

“[If] it’s a dose that I think seems really weird…I verify that order with the actual… order. And this is something that I usually verify with another nurse.”—Cardiology Nurse 4

Nurses reported on their actions when risk to the patient was high, for example when transitioning from one unit to another:

“That to me is a key point—a transition that is at high risk for having medications not given or given twice because the ED uses a different documentation system. So it’s a little bit harder to tell—well did they get this medication that got ordered down there or did they not? Because when you pull up the charting, it has all these overdue meds and if they were in the ED for several hours, you’re like, “Well I hope they this got med.” This was ordered three hours ago. They don’t necessarily always tell you that in the report. So then you have to call back down or you have to ask the patient if they can talk. If they can talk, they might not remember what they got.”—Cardiology Nurse 5

Furthermore, information in the system could not be followed indiscriminately:

“…This time of year when there are all these brand new interns still figuring out how to do their job, much less communicate with their cohorts, sometimes, you’ll see the same order put in by two people but the dose is off. Whether the dose is off or not…you can’t give both of those [doses ordered].”—Cardiology Nurse 6

Other times, the system would not update the doctors’ orders in time and would therefore have 2 dosages of the same medication prescribed. In such instances, nurses exerted autonomy in order to ensure patient safety by deviating from the orders in the system.

Conversely, nurses sometimes felt comfortable enough to skip what they considered to be minor verification steps in order to save time. For example, for less risky medications:

“A lot of people—and I’ve been guilty of this as well—carry their scanner around and scan the patient and then give the med and have no idea what happened on the computer when they do it…[but] I feel pretty confident that I’ve already checked the meds…well part of the issue also is that I don’t want to be hauling that COW [Computer on Wheels] back and forth up and down the hallways all day, and…if that can save me a minute or two, then I’m going to do it because I feel confident enough.”—Cardiology Nurse 7

As such, nurses could take advantage of gaps in the BCMA system to save time but effectively undermined the system’s ability to verify that the right patient was receiving the right medications. Thus, various additional situations required nurses to make decisions regarding patient safety. These situations often rely on nurse autonomy, judgment, and familiarity with individual patient schedules.

DISCUSSION

Medication administration: A task that requires expertise and critical thinking

Medication administration is a complex process that includes often-conflicting goals and demands that shape nurses’ work.²³ While previous studies report that BCMA systems overall have improved medication safety for up to a year after implementation, these results excluded time overrides and acknowledge possible user noncompliance.²⁴ BCMA was intended to serve as a verification system to ensure multiple aspects of medication safety. However, an inflexible system (eg, one that views an antibiotic and a multivitamin with the same priority) can create constraints that challenge autonomy.²⁰ This has been referred to as brittleness in the human factors engineering literature on EHRs and patient safety.²⁵ Further, BCMA was designed to capture the medication administration process for 1 single nurse and 1 single patient at a time.²⁶ As such, the implementation of new health informatics technology in the inpatient setting requires adjustment to routines and can disrupt the unit’s workflow to accommodate the new technology.²⁷

Organizational characteristics such as the availability of mobile BCMA workstations or the need to maintain isolation precautions for a patient may lead to bypassing of important steps in the BCMA process.²⁸ Furthermore, our data show that when caring for patients in transition from one unit to another, nurses are the safeguard to recover any information lost in the transition. Our results therefore indicate that for numerous reasons, the system’s primary function of verification was not always achieved.

Nurse autonomy is a key element of nurse satisfaction and patient safety.^9,29 The concept of “orienting frames” has previously been suggested to be a useful method of understanding how nurses enact medication safety.¹⁷ However, the themes in our data show that focusing purely on medication safety using a strict interpretation of the 5 rights is not feasible in real-world practice. Our data showed a misalignment of the BCMA system and the orienting frames nurses used to complete their work. BCMA often disrupted nursing work schedules and the temporal and spatial coordination with other clinical units.^23,27 Nurses exerted their autonomy by overriding the instructions of the BCMA technology, based on their expertise, to maintain an effective workflow. Indeed, the development of safe workarounds to this specific technology is an additional form of expertise that the nurses needed and developed. Our findings are consistent with previous studies regarding how nurses organize their temporal schedule, including: grouping meds together or scheduling medications depending on hospital meal times.²³

Fundamental assumptions about clinical work

The vast majority of CDS development in biomedical informatics is aimed at supporting physicians and other providers who were seen as sole decision makers required to enter orders for medications, tests, and interventions.³⁰ Only in recent years have researchers recognized the importance of nurses as decision makers rather than task doers and have focused on CDS for nurses.³¹ Forsythe problematized this hierarchy in her research on knowledge construction. She described how technology developers made a distinction between “expertise” and “common sense,” describing expertise as a “particular way of thinking that is possessed by a certain category of human beings labeled “experts.”⁴ She referenced a passage from Olson and Rueter³² that is particularly relevant for the present argument:

Experts have rich structures and reasoning abilities… Experts have stored rich representations of facts, objects and their attributes, as well as a set of inference rules that connect constellations of facts for use in problem-solving situations.

This description of expertise lends itself well to physicians’ medical knowledge (eg, of the body, the impact of medications, surgical techniques, and other “facts and objects” of treatment). We assert that nurses’ extensive expertise and nursing knowledge may be minimized by informatics tools that automate their tasks with the goal of improving safety and efficiency. Despite the fact that the concept of nurse as knowledge worker is foundational to informatics,^33–35 nurses may be perceived as doers, rather than knowledge workers. Thus, informatics tools designed for their use may conceptualize nursing work as the accomplishment of tasks. Clearly, administering a medication is indeed a task. However, closer inspection of nursing work reveals that nurses continually assess and modify care based on the needs of patients and they often engage in patient education during medication administration. Informatics tools could be useful in, for example, helping the nurse communicate by visually demonstrating a dosing regimen for a patient during each administration, so that the patient can be better prepared for discharge.

In alignment with other critiques of user-centered design in EHRs and applications such as BCMA,^36,37 we recommend that developers and researchers engage with nurses to understand and observe the nature of their expertise, the nursing process, and how both are used in everyday work, to design and implement ways that the nursing process and nursing expertise can be supported through technology. The creation of new Safety-Enhanced Design Standards by the Office of the National Coordinator for Health Information Technology is a step in the right direction; however, the user-centered design standards are required for health information technology elements primarily used by physicians, including order entry, CDS, and electronic prescribing, among others.³⁸ Potentially useful strategies for the field to supplant the required actions and improve design for nursing practice include:

• The creation of clinical immersion experiences for technology developers to gain access to the everyday working lives of technology end users.

• The use of simulated clinical environments for testing new technologies with real users.

While our findings are not novel regarding the shortcomings of health information technology application design, our focus on autonomy and expertise have additional implications for management’s role in maintaining nurse satisfaction and consequent retention. Regular assessments of nursing workflow to assess constraints on autonomous practice will reveal sources of dissatisfaction and increased risk for patient safety events. Constraints found in everyday activities such as medication administration, whether technological, political, or both, inhibits nurses from accessing the full potential and value of their scope of practice and experience. Management has the opportunity, through policy, to endorse or reject the rigid constraints of a technological system.¹⁸

Limitations

The key limitation to this report is that the data collection was conducted in 2011. Other findings have been reported from the study.^18,27,39,40 However, we feel that this subanalysis of the data is of continued relevance because the elements discussed here have changed little since the data were collected. Task-based systems such as BCMA continue to demonstrate how the expertise involved in nursing work can be diluted in the process of automating a task. Furthermore, the recommendation of user-centered design remains critical in new technological system development and implementation in healthcare settings. A second limitation in the study is potential bias in qualitative coding. None of the coders had clinical experience or were EHR developers, which minimized bias of training or professional perspective. However, because coders were not clinicians, nursing scope of practice may not have been considered in the context of themes, such as “being responsible.” Our study participants included licensed practical nurses (LPNs) as well as registered nurses (RNs). There are instances during the medication administration process when scope of practice regulations require the LPN consult with the RN prior to administration of a medication. However, because coders were not clinicians, nursing scope of practice may not have been considered in the context of themes, such as “being responsible.” Our study participants included LPNs as well as RNs. There are instances during the medication administration process when scope-of-practice regulations require that the LPN consult with the RN prior to administration of a medication. The coding team had regular meetings to review the emerging themes and explore theoretical frameworks that helped explain the data.

CONCLUSION

Using orienting frames as a lens for understanding nurse work, our analysis indicates that nurse autonomy can be compromised by inflexible informatics tools leading to system brittleness, potentially increasing the likelihood of workarounds and subsequent risks to patient safety. Only when new technological systems are designed with the user’s everyday working experience in mind can they be used as intended.

FUNDING

This work was supported by National Library of Medicine grants 1 K99 LM 010038-01 A1 through 4 R00 LM010038-02 (to LLN).

AUTHOR CONTRIBUTIONS

JYH contributed to conceptualization, data collection and analysis, and writing. LLN and CLS contributed to data collection, interpretation, and writing. CV and CHI contributed in interpretation, writing, and editing. All authors approved of the final version.

CONFLICT OF INTEREST STATEMENT

None declared.