Barcode Medication Administration: Supporting Transitions in Articulation Work

Abstract

Articulation work is that which enables coordinated activity among colleagues distributed in time and space. Despite its important role in clinical settings, this work remains largely invisible in process flowcharts. When process-oriented information systems are implemented, the informal, flexible, contingent activities of participants that enable coordinated work are suddenly placed in a new context. Articulation work must adapt to new contexts of automation, and there are opportunities for clinical systems to better support coordination activities. This research explores the articulation work involved in medication administration, how it is affected by the implementation of barcoding, and strategies for support and problem resolution in this arena.

INTRODUCTION

The roles of Bridgers, or people who serve as liaisons between informatics and clinical functions, were first described in the context of CPOE by Ash and colleagues [1]. Bridgers were characterized as a category of “Special People” who significantly contributed to the success of CPOE implementations through their unique knowledge of informatics, understanding of clinical work and respect for the clinicians trying to learn and use the system. This paper expands the previous characterization of Bridgers by examining how they support changes in informal coordination work that result from the implementation of a clinical informatics application.

This analysis is part of a larger study of how nurses integrate barcode medication administration (BCMA) and an electronic medication administration record (eMAR) into everyday clinical practice in an inpatient setting. The Bridgers, in this case referred to as the “Informatics Support Team” or IST, were responsible for ensuring the success of the implementation of BCMA and eMAR through planning, training, problem resolution and go-live support. Observations of IST meetings and support of nurses during go-live revealed strategies for resolving highly complex problems, particularly related to the support of nurses’ coordination-related tasks, also referred to as articulation work.

BACKGROUND

Implementers of health informatics applications have learned that, when systems are implemented, forms of work are discovered that were not depicted in pre-implementation flowcharts. This work represents the many organizational and medical contingencies that must be managed to maintain the patient’s health care trajectory [2–4]. Among other contingent tasks, nurses, pharmacists, respiratory therapists, physicians and other health care workers must develop coordination mechanisms that smooth the interactions of their individual, highly complex arcs of work [5].

Articulation work has been described as that which “manages the consequences of the distributed nature of… work”[6]. Workers have been observed using everyday communicative actions such as talk, gesture, and shared manipulation of artifacts to coordinate their work in tightly enclosed spaces such as control rooms and towers [7, 8] and an airline cockpit [9]. However, special challenges may exist in environments requiring coordination at larger temporal and spatial scales [4], such as hospitals and health systems.

For nurses, articulation work is multi-faceted. The bedside nurse is the ultimate coordinator of care for any given patient, and is responsible for ensuring that a range of nursing activities, including medication administration and documentation, takes place within a context that includes 1) the patient getting appropriate rest and receiving care from specialist providers, 2) a dynamic profile of orders from the physicians, 3) the operation of a plethora of monitoring, diagnostic and other computer equipment, 4) the errors of others, such as missing medications or documentation errors, and 5) all of the above activities for the other patients to which she is also assigned. In other words, the nurse must juggle her own activities as well as the busy days of her patients.

When an informatics system is implemented, a disruption occurs in the articulation work that was previously successful in coordinating activity across shifts or between affected departments. For example, when a complex system of paper charts and sticky notes is replaced with an electronic clinical messaging system, the participants must adapt their work practices to this new genre of communication and make it an effective enabler of coordinated activity. This adaptation is real work, and sometimes fails to optimally integrate the new tool into practice.

METHODS

The study site was a multi-hospital, tertiary medical center. IRB approval from the subject institution was obtained prior to initiating the study. Data included fieldnotes from observations, documents such as meeting minutes and reports, and email communications. Interviews were scheduled in a later phase of the project, therefore interview data are not included in this report.

The Informatics Support Team (IST)

The IST was a group of approximately 20 registered nurses that existed as an interface between the informatics developers and the clinical staff. The team, growing over the years, had successfully implemented CPOE and nursing documentation in the hospitals. The IST had a number of responsibilities:

• Working with developers to customize vendor products and design in-house solutions for specific needs.

• Working with project managers, vendors, informatics staff, hospital management and clinical staff to design implementations, including detailed equipment considerations and rollout scheduling.

• Training new staff on clinical systems and training all users on new systems being rolled out

• Implementation support, including 24/7 coverage in the early weeks for each unit.

• Initiating and participating in problem resolution for new systems being rolled out, including interface issues with other systems.

• Ongoing support of implemented systems.

Observation

The principal investigator, a PhD-trained health care anthropologist, conducted ethnographic observations of IST and nurse interactions during the weeks that the system was being brought online. Observations described here occurred on ten patient care units, which went live over the course of 6 months in 2007. Approximately 50 hours of observation on the clinical units occurred primarily during the times of medication passes in the morning and evening. Participants included nurses, IST staff and other related informatics and clinical staff. Twenty users were observed interacting with the system. Fieldnotes occasionally included comments from patients, but no patient identifiers were collected in observation. Information about the study was provided to all participants, and all agreed to be observed.

Additionally, 48 meetings of IST staff were observed, including implementation team meetings (8–15 participants) and special problem-solving meetings (2–6 participants). Other participants in these meetings included IST staff, informatics staff, nursing coordinators and managers, hospital administrators, pharmacy staff and representatives from respiratory therapy. In initial encounters, the principal investigator provided a brief verbal overview and hardcopy information about the study to the meeting participants.

Observations were recorded as fieldnotes by the principal investigator and subsequently entered in expanded form into computer text files. Particular attention was placed on capturing issues the team encountered and strategies for problem resolution.

Email and Shift Reports

The IST used email extensively for informal communication as well as semi-structured shift reports. When it became clear that this was a potentially valuable source of data, IRB approval was obtained to analyze email communications as part of the study. Email communications contained detailed descriptions of problems and problem resolution strategies. Emails (170) were received on an email client and each message was saved as a text file for importation into the data analysis software.

Data Analysis

Files containing fieldnotes and email content were imported into QSR NVivo Version 7, a qualitative data analysis software program. In NVivo, each file becomes a “source” document. Sources are then opened and text units (usually paragraphs) are coded (assigned one or more labels, or nodes). The principal investigator performed all of the coding and analysis of data. It is in the coding process that theory is applied in qualitative research. In this case coding was focused on two major areas of inquiry 1) the nature of articulation work before and after BCMA and 2) support/problem resolution strategies engaged by the IST. Themes describing articulation work resulted in nodes such as “interacting with non-BCMA units” and “Nurse-Nurse Communication”. Themes describing support and problem resolution strategies of the IST included nodes such as “Training”, “Customizing”, and “Communication”.

Reviewing the content of the nodes, the principal investigator extracted recurring themes related to articulation work, some exemplary vignettes to describe the challenges faced by the nurses and problem resolution and support strategies of the IST. The Results and Discussion section reviews these themes in the context of the 5 Rights of medication safety (right patient, right drug, right dose, right time, right way).

RESULTS AND DISCUSSION

Medication Administration Prior to BCMA

Themes from the analytical coding are organized according to the nurses’ practice goals: the familiar “Five Rights” of medication administration.

Right Patient: to be accredited by the Joint Commission, hospitals must ensure that patients are identified in two ways prior to receiving care or treatment. While many nurses agree that patient identification is important, the process can be cumbersome and socially awkward in situations where certainty is high, for example in the ICU where the nurse may be assigned only one patient. In her ethnography about nursing work [10], Zane Robinson Wolf describes medication administration as a “therapeutic ritual”. Repeatedly asking the patient’s name or other identifier becomes an aspect of the nurse’s performance of her work, and Robinson argues that the routines involved in this ritual serve to prevent error. Individualized routines involved in patient identification reflect articulation strategies for communicating with patients and families.

Right Drug: Nurses were responsible for ensuring that the drug they were about to administer was the right one for the patient identified. Drugs primarily came from two sources, including the automated drug dispensing machine, which contained frequently used items stocked on the unit (“floorstock”) and also drawers labeled for specific patients. The pharmacy stocked the machine with floorstock and labeled patient-specific medications.

Nurses double-checked the order for the medication prior to giving it by comparing the drug name to the information on a printed order summary they carried in their pockets. The order summary was a “snapshot” of orders for the patient at the time it was printed, therefore the nurses also checked the patient chart prior to giving medications to ensure there were no new orders or discontinuances of orders. Nurses developed individualized routines for acquiring the medications for a specific medication pass, as well as strategies for dealing with medications that were missing from floorstock or the patient-specific drawer. Some nurses sent an electronic message to the pharmacy, some called the pharmacy, some called and messaged the pharmacy, and in some situations a nurse would designate a specific amount of time to wait to see if the medication showed up on the unit.

Right Dose: In comparison to “right drug”, achieving the right dose involved much more complexity and uncertainty. In many cases, doses are changed during the course of a nursing shift, and nurses must stay abreast of the latest information coming from the physicians. Through discussion at the bedside, the nurse may be aware that a dose will be changed, even though it has not been entered into the CPOE system. In this case, the nurse has more up-to-date information than the pharmacy, and is aware that pharmacy-sourced information and artifacts, such as a pre-printed medication administration record (MAR) or prepackaged syringes, inaccurate. Nurses develop tools to enable coordination of these various sources of information. Typically paper-based and kept handy in a pocket, nurses are able make changes on the tool and see the history of changes over time.

Right Time: Given the primacy of medication administration among nursing functions, nurses are generally excellent at cobbling together a schedule for the patient that meets the physician’s intent. In the current study a patient was scheduled to take a medication at 0800, but was asleep at that time. With barely a mutual glance, the nurse and the care technician agreed that the patient should get sleep, and the medication could wait until she was awakened for a procedure later in the morning. Making these tradeoffs requires expertise, including knowledge about the effect of the drug, the rest of the patient’s schedule (for example, is the medication required prior to the procedure?), the latitude allowed by the physician, and enough experience to properly estimate nursing workload and ability to administer the medication later.

Right Way: This aspect of safe practice is sometimes referred to as “right route”, suggesting the proper route of entry into the body (intravenous, oral, topical, etc.). Expanding the category to “right way” implies that it also includes such considerations as taking with food, not mixing with certain other medications, and other specific requirements. In intensive care units (and increasingly in acute care units) complex instructions accompany the orders. Instructions may include information about specific sequencing of medications, tapering of doses or staggering of doses to get a patient onto the standard administration schedule. Medication instructions were written by nurses on the medication administration record (MAR).

One aspect of nursing work is to find creative ways to get patients to take their medications. Nurses were observed using techniques with adult patients such as bargaining, nagging and authoritative language to get patients to comply. However, children are much more shrewd than adults in their negotiation tactics, so nurses frequently put medications in applesauce or peanut butter. One nurse described a child who would only take Tylenol in the form of a gelcap, which was not on the hospital formulary. In a good example of “making it work”, someone was sent to a retail pharmacy to pick up the Tylenol. The ability to transmit these detailed medication instructions to the next shift was a critical aspect of coordination for the nurses.

After BCMA/eMAR: New Forms of Articulation

Right Patient: BCMA alters patient identification by introducing a barcode scan into the process. In this case, the overall process was: 1) obtain medications, 2) scan patient, 3) scan medications, 4) scan patient, 5) give medications, 6) confirm administration. Two forms of patient identification were still required, so the nurse still needed to ask the patient’s name. The barcode technology altered the process by introducing an artifact that 1) needed to be explained to the patient and/or family, 2) could fail (sometimes barcodes do not scan properly or have to be re-scanned due to software failure or error), and, 3) was more physically invasive to the patient (scanning the armband is more likely to involve touching or waking the patient than simply reading the armband).

Right Drug: Nurses were instructed to scan the medication packaging to ensure the right drug and dose. This was a significant change in practice, requiring the nurse to devise an efficient process (at the bedside, in the hallway, or in the med room) of scanning and preparing the medications. Alerts fired frequently. For example, if 20mg of a medication were ordered, resulting in (2) tablets of 10mg each, an alert would fire after the first pill was scanned indicating a wrong dose. After the second pill was scanned the alert would disappear. This is not a problem for the experienced user, but the novice users needed to read all of the alerts carefully and interpret them to ensure there was no real error.

The BCMA system introduced new articulation work related to medication packaging. When packages did not properly scan, the procedure was to place the medication in the pharmacy “bin” on the nursing unit and send a message to pharmacy to send up a new medication. If the nurse confirmed through other means that the medication was appropriate, she could go ahead and give the medication and place the opened package in the bin. This troubleshooting and follow-up activity was critical for quality assurance in the pharmacy, which was also undergoing a massive process change to implement barcoding. However, keeping track of blister packs and empty vials (previously considered trash) was new, additional work for the nurses and required some adaptation.

Right Dose: BCMA introduced a new form of dose uncertainty that required a change in articulation work on the part of nurses. The eMAR system performed a double-check dose calculation, and rounded the results to 3 decimal places. The pharmacy system rounded doses to 2 decimal places because, as one pharmacist put it, “you can’t draw up 3 decimal places [in a syringe]”. When the eMAR came up with a different number from the pharmacy system, which produced the barcode labels, a “wrong dose” alert fired when the medication was scanned. This initiated a research process for the nurse, who had to confirm the content of the original physician order, and calculate the proper dose.

Right Time: The BCMA system impacted the way nurses prioritized their work. With a paper MAR, reasons for a late medication administration were noted, but overdue medications were not a significant performance indicator monitored by nursing management. With BCMA, the timeframe for giving medications was expanded from 30 minutes before or after the scheduled time to 1 hour before or after. However, the fact that late medications were now viewable on a report sensitized the nurses to the medication schedule in a new way. Late medications and the “Overdue Medication Report” became perceived as negative, despite the fact that late medications are frequently the result of valid logistical issues such as the patient being off the unit or asleep. Nurses felt the need to “address” the overdue meds with charting, which often removed a dose from the eMAR that still needed to be given. This was known as “stealing doses”, and it created problems for managing doses in downstream shifts.

Right Way: As described earlier, the “right way” was often communicated across shifts via written medication instructions on the MAR. This functionality was not included in the eMAR, creating great concern, especially for pediatric and chemo units that were further down in the rollout schedule.

New Articulation Work: Because computer systems are generally useful for scheduling, the eMAR was effective at providing the nurse with an overview of the number of medications due for all of her assigned patients at any point throughout the day. An exception was the case of “once” orders. With the paper MAR, a medication ordered for “one-time” administration was printed along with all of the scheduled meds. In the eMAR, after a specified period of time, the “one-time” dose was automatically discontinued by the system. This was a great concern for obvious reasons – if a logistical problem caused a delay in passing medications, the nurse may assume that the physician had discontinued the order and not give the medication. Increased vigilance and cross-referencing with the CPOE system was required to ensure that medications were not missed.

IST Support and Problem Resolution Strategies

Several strategies used by IST management and staff proved useful in supporting nurses as they developed new coordination mechanisms in response to BCMA.

Organization. In addition to go-live support, the IST staff rotated through shifts as trainers. This cross fertilization of training and support resulted in training classes that were highly relevant to real-world practice, ensured that training reflected the latest policies/procedures that had been developed to accommodate practice, and gave insights to go-live teams about what the nurses had been taught in class. An example of this is how they dealt with the creation of new routines in patient identification. The classes provided an opportunity to role-play and coach the nurses on how to talk to patients and families about BCMA. Similarly, unavoidable practice changes such as keeping track of medication packaging could be discussed in class and reinforced in practice.

Another aspect of organization was a direct linkage to hospital and informatics management through the BCMA team meetings and steering committee. This enabled rapid communication of critical problems and ensured that issues requiring policy decisions were dealt with quickly.

Communication. Extensive use of email communications and shift reports kept everyone up-to-date on the latest issues, news and approaches to problem resolution. Shift reports went to builders, pharmacists, the project manager, and other IST members, and most replies went to all recipients. Due to the large amount of email traffic this approach produced, the construction of the shift report emails was not taken lightly by most IST staff. One staff member was asked about this as he repeatedly edited the shift report email he was constructing (often the email was initiated early in the shift, saved as a draft and expanded as the shift went on). He said you have to “write the email so people don’t get overwhelmed, because there are so many of these emails going out”. He felt that an email that was written in a way that was too complicated would result in all the readers saying, “Somebody else will get that”. Problem resolution involved selling your problem.

Problem Resolution. The RNs on the IST were able to quickly assess the patient safety risk posed by a BCMA problem. A “swarming” approach was used, which involved rapidly convening a multidisciplinary team to analyze the critical issues and develop possible solutions. In the case of the medication messaging problem, a solution was quickly devised using the CPOE system and the creation of medication instruction and lab draw messaging functions. The collaborative engagement of informatics and pharmacy staff made this approach possible. This strategy also resulted in system solutions to the dose calculation rounding problem described above.

Follow-up – Rounding on live units. Another useful strategy was to go back to units already live on the system and identify problems that persisted. In this way the IST discovered the ongoing problem with overdue medications described above. They made changes to coaching strategies and the “Overdue Meds Report” was renamed the “Outstanding Meds Report”.

CONCLUSION

The implementation of new information technology in the clinical setting can be disruptive to existing patterns of articulation work, or work that coordinates the activities of people across time and space. Implementation teams must familiarize themselves with articulation work and support users in developing new ways of coordinating with colleagues on other shifts and in remote physical spaces.

Through strategies of integrated coaching and teaching, inclusive communication, swarming problems, and follow-up on live units, the IST was able to resolve many problems related to articulation work before they became organizational or patient safety crises.