Interruptions in Community Pharmacies: Frequency, Sources, and Mitigation Strategies

Abstract

Background:

Interruptions constitute a key part of the communication strategy for healthcare providers, including community pharmacy personnel. Previous research in other healthcare environments has shown that interruptions are very common and may present as patient safety 7 hazards. One 1999 study, conducted in community pharmacy settings, found that interruptions 8 were prevalent and associated with dispensing errors. In the last 20 years, pharmacies have expanded patient services and implemented new technology. Yet, it is unknown how these changes have affected interruption sources and frequency.

Objective:

The objective was to characterize the frequency and sources of pharmacist task interruptions that occur in community pharmacies.

Methods:

A cross-sectional qualitative study design was used to evaluate the frequency and sources of pharmacist task interruptions in Wisconsin community pharmacies. The participants included 9 pharmacists and 9 technicians working in 2 independently-owned, 2 mass merchandise, and 2 hospital/clinical-affiliated pharmacies.

Results:

Interruption rates ranged from 3 to 7 per hour across the 6 pharmacies. Sources of interruptions included: patients, technicians, self-initiated interruptions, technology used in the pharmacy, and a second pharmacist.

Conclusions:

Interruptions are common in community pharmacies and the causes are multifactorial. Prudent management of interruptions in these work environments may involve improved technician training, and better designs of pharmacy technology.

Introduction

It is no surprise that interruptions occur frequently in healthcare environments. A study by Weigl et al. found that hospital physicians were interrupted an average of 4 times per hour, most frequently by nursing staff, telephones, beepers, and fellow physicians.¹ In the 1999 seminal report on medical errors and patient harm in hospitals, To Err is Human, the Institute of Medicine identified interruptions as potential contributors to medical errors and made a recommendation for hospitals to minimize them whenever possible in order to improve patient safety.² Following the publication of this report, there has been a gradual rise in the availability of scientific evidence documenting the frequency of interruptions and their relationship to adverse patient outcomes. For example, Westbrook et al. analyzed data from observations of drug administration episodes by nurses in hospital wards and found that interruptions occurred in 53% of the episodes. They also found that each interruption was associated with a nearly 12% increase in procedural failures and a 13% increase in clinical errors.³

Although research has been conducted to measure the effects of interruptions on the error rates in hospitals, there is still no consensus on the actual definition of an interruption. Brixey et al. offers a well-cited definition of an interruption as “a break in the performance of a human activity initiated by a source internal or external to the recipient, with occurrence situated within the context of a setting or a location.”⁴ For the purposes of this paper, the Brixey et al. definition is used. Common sources of interruptions in healthcare include phone calls, requests from other healthcare providers, medical equipment failures, alarms, and questions from patients and their caregivers. It has been found that interruptions disrupt healthcare professionals’ workflow and increase their workload.⁵ Note that these can be classified as interruptions initiated by sources external to the recipient. There is significantly less research on interruptions initiated by sources internal to the recipient, which are referred to in this paper as “self-initiated interruptions.”

Although interruptions are common in healthcare environments, there is a dearth of research on the impact of interruptions in community pharmacy literature.^6,7 To date, only one published study in the U.S. has described interruption frequency and the association of interruptions with dispensing errors in ambulatory care pharmacies. In this study, conducted nearly two decades ago by Flynn et al., it was found that the mean number of interruptions per hour per pharmacist was 5.98 (standard deviation=2.7). The sources of 85% of the interruptions were related to prescription-processing questions.⁸

Since the Flynn et al. study was conducted, the practice of pharmacy has changed significantly, resulting in a need for more research on how interruptions may impact patient safety and how to combat its potentially negative effect. More medications have become available to patients, and the number of prescriptions being filled in community pharmacies has increased.^9,10 Furthermore, the workflow in community pharmacies has also changed with technological advances. Most pharmacies currently receive the majority of their prescriptions electronically. Recent studies reported that, despite their perceived benefit in improving patient safety,¹¹ electronic prescription technologies are contributing to workflow interruptions, forcing pharmacy staff to interrupt their ongoing task to remediate the problem.^12–15 Given the dynamic change in pharmacy practice, it is possible that the nature and extent of interruptions has also changed.

Objective

The objective of this paper is to provide a descriptive account of the sources and frequency of interruptions that occur in community pharmacies.

Methods

A qualitative approach using observations was used in order to understand how and why interruptions occur in community pharmacies. The common sources of interruptions and the frequency of interruptions were also recorded. Because so little work has been done on the impact of interruptions in pharmacies, a qualitative approach providing rich findings was ideal for this research study.^16–19

For the purpose of this research study, Brixey’s definition of an interruption was operationalized as: a break in the task of a pharmacist while the pharmacist is engaged in any activity related to dispensing medications in a community pharmacy, such as order transcription, pill counting, filling, labeling, verification, and counseling.

Participant Recruitment

Permission to conduct observations in community pharmacies was obtained from the managers and staff of 6 community pharmacies representing 3 different pharmacy work environments: 2 independent, 2 mass merchandise, and 2 hospital/clinic affiliated pharmacies. Permission to conduct observations in large, corporation-type chain pharmacies was not granted. A combination of purposive and convenience sampling was used to recruit a total of 18 participants (9 pharmacists and 9 technicians) from the 5 pharmacies included in the study. Two pharmacists were recruited from 3 of the 6 pharmacies. One pharmacist was recruited from each of the remaining 3 pharmacies. Technicians were recruited because of their role as a potential source of interruptions to pharmacists, with 1–2 technicians recruited from each pharmacy. Written informed consent was obtained from participants.

Pharmacists that had expressed interest in research involvement were approached with a request to participate in this study. Upon agreeing to participate, the pharmacists invited their staff to participate. The interested pharmacy staff were then enrolled in the study, and observations were scheduled at times when only enrolled staff members were present. Observations were scheduled at various times and days of the week in order to capture the full range of interruption characteristics that may vary based on the time and day of the week. Table 1 provides a summary of the characteristics of participating study sites, pharmacists, and technicians.

Table 1:

Characteristics of Study Participants. Study participants include pharmacists and technicians from the participating community pharmacies. Described below are the mean and range of study participant age, average hours worked per week, years active in practice, and years practiced in this pharmacy. The average daily prescription volume for all participants is also included.

	Overall	Pharmacist (n=9)		Technician (n=9)
Variable	Mean (SD)	Mean (SD)	Range	Mean (SD)	Range
Age	37 (12)	43 (11)	31–55	32(11)	19–48
Average hours worked per week	39 (3)	39 (2)	36–40	38 (4)	32–40
Years active in practice	12 (9)	16 (9)	6–32	8 (6)	1–18
Years practiced in this pharmacy	7 (6)	10 (7)	3–21	4 (6)	0.5–17
Daily prescription volume	215 (96)

In order to reduce bias and undue behavior modification stemming from presence of research observers (i.e., the Hawthorne effect), the term “interruption” was not used when describing the study. Instead, participants were told that the purpose of the study was to understand how pharmacy staff generally work together to communicate, make decisions, and problem-solve in the pharmacy, without specifying that we were focused on interruptions. Each participant was offered $50 for their participation in this study. The study was approved by the Institutional Review Board of the University of Wisconsin-Madison.

Data Collection Procedure

Two researchers (EA and CL), trained in qualitative methods, conducted the data collection. For the first community pharmacy, both EA and CL performed the observations together, taking their own field notes and then comparing post-observation in order to refine observation note-taking congruence between researchers. The remaining 5 pharmacies were divided up between EA and CL. Detailed, handwritten field notes were taken about what the participant was doing, especially making note of the instances of interruptions while the pharmacist was performing a certain task, as well as the circumstances such as what happened before and after the interrupting event. During the observations, the researchers positioned themselves within the pharmacy at a location convenient for observing the pharmacist’s ongoing tasks. Each observation period lasted for 3–4 hours. The timing of the observations (morning or afternoon) varied for each of the pharmacies. Some were observed both in the morning and the afternoon if it was indicated by participants that there may be a difference in number of customers or tasks that needed to be completed by pharmacy staff. Immediately after each observation session, the researchers reviewed their field notes to clarify and expand their notes with details that could not be written quickly enough while observation was in progress. The 2 researchers then typed their field notes so they could be reviewed by a third member of the research team who reviewed them and provided comments on clarity and consistency. Following this step, the observers addressed the comments by providing additional detail where possible. The additional review step served to ensure completeness and clarity of the observation data. Each participant was observed for a total of 10–12 hours over the course of three observation sessions.

Data Analysis

Completed field notes from the observations were imported into the qualitative data analysis software Nvivo® 10 (QSR International) for coding. Coding was done to quantify the number of interruptions that occurred within the timeframe of each observation period as well as to identify sources of interruptions. Each break in the task of a pharmacist was coded as an interruption. Some interruptions were nested within an interruption and these were also coded as their own interruptions. A nested interruption occurs when an individual’s secondary task is interrupted by another task, requiring the interrupted individual to resume the primary task after completing the interrupted secondary task.²⁰ Coding was done in this fashion for all observation field notes from the 6 pharmacies. The instances of interruptions per observation session were then averaged for each pharmacy and reported as the number of interruptions per hour. The range of instances of interruptions per observation session was also reported. The field notes were coded inductively in order to generate categories for the sources of interruptions. Any activity or event immediately preceding and causing an interruption in the task of the pharmacist being observed was coded as a source of interruption. Once coding was completed in this manner for all field notes across the 6 pharmacies, the codes describing a source of interruption were aggregated into categories and sub-categories.

Two researchers performed all the coding of the observation notes. The codes were then reviewed by both researchers and other members of the research team. Any variation in the coding was resolved through group discussion.

Results

Analysis of the observation data showed that the rate at which interruptions occur in community pharmacies varies considerably. On average, interruption rates ranged from seven to thirteen per hour. Table 2 shows rates of interruptions in each of the six studied pharmacies.

Table 2: Frequency of Interruptions of Pharmacist Tasks.

The interruption frequency-related characteristics are described for each of the six participating pharmacies.

Pharmacy Code	Type of Pharmacy	Average number of interruptions per hour	Range (of interruptions per hour)
Pharmacy 1	Independent with a drive-thru window	7	4–9
Pharmacy 2	Independent	8	5–13
Pharmacy 3	Mass merchandise	13	11–16
Pharmacy 4	Mass merchandise	9	8–12
Pharmacy 5	Hospital/Clinic affiliated	5	3–6
Pharmacy 6	Hospital/Clinic affiliated	8	8–9

Within the participating pharmacies, multiple factors contributed to interruptions of pharmacist tasks. Although the specific sources of interruptions vary, they can be grouped into five major categories: patients, technicians, self-initiated, technology, and a second pharmacist. Table 3 shows examples of representative scenarios for each of these 5 categories.

Table 3: Sources of Interruptions.

The sources of interruptions to pharmacist tasks are described along with representative example scenarios to illustrate the manner in which each source of interruption disrupts the pharmacist workflow. The percent contribution refers to the percent of total observed interruptions that were initiated by a particular source of interruptions.

Sources of Interruptions (% contribution)	Example Scenario
Patients (43%)
In-person	Patient walking up to a pharmacy counter or driving up to the drive- thru window.
Phone call	Patient calling in a prescription refill and insisting to speak to a pharmacist.
Technicians (25%)
Not part of their role	Technicians are not legally authorized to provide a patient counseling and thus have to interrupt a pharmacist whenever they have a customer in front of the counter.
Need pharmacist assistance	Technician having difficulty with prior authorization or transfer of prescriptions.
Non-work related/Miscellaneous	Technicians initiating a conversation with a pharmacist about non- work-related issues such as family life or pets.
Self-initiated (17%)
Need information to complete tasks	Pharmacist receives prescription with missing information and calls the physician’s office to retrieve missing information.
Anticipation of events	Pharmacist overhears a developing situation in the pharmacy and anticipates an upcoming interruption from a technician; pharmacist breaks and waits for a question.
Coordination/ task management/ Prioritizing	Pharmacist makes sure that other pharmacists and technicians know about a specific prescription that was problematic.
Miscellaneous/Personal	Pharmacist initiates a conversation on non-work-related issue with another pharmacy staff.
Technology (8%)
Suboptimal designs of technology	Drug utilization review alerts appearing on the computer must be attended by a pharmacist for a technician to proceed with prescription processing.
Telephone, Public announcement system	Phone rings continuously and pharmacist answers it. Public announcement system makes a loud voice through speakers embedded in the store ceiling informing pharmacy staff about an incoming call.
Second pharmacist (5%)
Handing off information	Pharmacist will be leaving pharmacy at the end of the shift and needs to handoff information about an unresolved prescription issue.
Clarifying information	Pharmacist is not sure about a specific drug’s dosage in the prescription and asks another pharmacist for clarification.

Patients as Sources of Interruptions

Patients contributed to the highest proportion of interruptions that occurred in the pharmacy. During the entire observation period, 447 instances of interruptions were recorded, and 43% of these were due to patients. Patient-induced pharmacist interruptions occurred in 2 ways: first, patients came to the pharmacy in person to drop off or pick up their prescriptions, and second, patients phoned the pharmacy and asked to speak to a pharmacist even though their issue may have been able to be resolved by a technician. Because a customer could enter the pharmacy at any time, pharmacists were frequently interrupted while performing safety critical tasks such as verifying prescriptions. This, however, largely depended on the personnel’s working style, preferences of each pharmacist, as well as the physical layout of the pharmacy itself. For example, some pharmacists preferred to complete the task at hand before looking up and offering to help patients who approached the front counter. Others, on the other hand, abruptly stopped what they were working on upon noticing a patient and asked if they could help the person.

Although the extent of pharmacist interruptions caused by patients varied considerably across the studied pharmacies, good customer service, which manifested as greetings and offering help to patients who approached the pharmacy, took a higher priority than any other task. In some pharmacies, pharmacists also worked at stations that put them near the counseling counter, thus making it more likely that they would feel the need to greet and help the patient customer right away, irrespective of the task that was being performed at that particular time.

As alluded to in the previous paragraph, the physical layout and types of services offered in the pharmacy also seemed to play a role in attenuating or amplifying interruptions by patients and their effect on tasks. For example, some of the observed pharmacies had 2 pharmacists working at once; one pharmacist was responsible for counseling and helping patients who approached the front counter, and a second pharmacist worked at a counter located behind the main front counter. Thus, the layout of the pharmacy allowed the second pharmacist to focus on verifying prescriptions with little or no interruptions due to the protection afforded by the layout of the pharmacy. On the other hand, pharmacies that only had one front counter had the pharmacists directly exposed to the patients as they walked into the pharmacy.

As described in Table 2, Pharmacy 1 had a drive-thru window in addition to a normal walk-in service. In such an arrangement, patients who appeared at the drive-thru window were generally prioritized for service over those who walked in to the pharmacy.

Technicians as Sources of Interruptions

During the observation periods, the technician staffing patterns of study site pharmacies ranged from 1–4 technicians. Pharmacy technicians played a key role in processing prescriptions and attending to the needs of patients, such as receiving phone calls and routing them appropriately, as well as offering help in locating and providing pricing information on over-the-counter products. In this role, technicians often acted as first line of contact for patients calling the pharmacy or making in-person visits.

Technician-related interruptions of pharmacists were primarily due to 3 reasons. First, technicians interrupted pharmacists because they were not allowed to perform certain tasks for which they did not have the formal training or legal authorization. For example, some areas of the United States require that patients receive counseling on medications in pharmacies from a licensed pharmacist. Technicians also interrupted pharmacists when they encountered an error message at their computer station that could only be overridden by a pharmacist due to organizational policies. The second category of interruptions from technicians occurred when technicians did not have the necessary knowledge to deal with a particular issue and needed the pharmacist’s help. For example, the pharmacist would be interrupted when a technician did not know how to transfer a prescription to another pharmacy, when the technician had difficulty processing an insurance-related issue, and when the technician was unfamiliar with the medication required for a particular prescription. The third category of technician-induced pharmacist interruptions were non-work related; they were the result of conversations about family and leisure activities.

Despite being a source of interruptions, technicians also played a key role in shielding pharmacists from events that had the potential to interrupt them. For example, technicians greeted patients who either called or walked into the pharmacy, and they listened to their questions to see if they could answer them before transferring the case to the pharmacist. In these cases, technicians were able to triage potential interruptions, preventing unnecessary disruptions to pharmacist workflow.

Self-initiated Interruptions

Interruptions could also occur as a result of internal stimuli. These types of interruptions, called self-initiated interruptions, refer to when a pharmacist leaves midway through a task to perform a second task without being triggered to do so by an observable, external stimulus. These types of interruptions, accounting for about one-fifth of the total recorded interruptions, resulted from interrupting thoughts of pharmacists and had 4 major underlying causes:

1. First, pharmacists interrupted their own task when they lacked the information necessary to complete the task at hand. Pharmacists would interrupt their work to obtain additional information by speaking to other pharmacy personnel or people outside of the pharmacy depending on the issue at hand.

2. Second, pharmacists engaged in self-interrupting behavior due to their supervisory role in the pharmacy. In addition to their routine functions of verifying, dispensing and counseling patients, pharmacists also played the managerial role of continuously scanning the pharmacy work environment. This way, pharmacists maintained a global awareness about what was going on around them, as was confirmed in comments made by pharmacists. For example, during the observations, pharmacists were seen to be continually monitoring the conversations of other pharmacy staff and were able to intervene in situations requiring their involvement. Pharmacists indicated that the motivation for this kind of behavior was generated in part by the pressure to prevent disruptions in the overall workflow of the pharmacy team and to provide fast customer service. Pharmacists monitored the work of technicians and stepped in to help whenever they sensed that the technician was experiencing difficulty with pharmacy technology, medications, or patient specific questions.

3. Third, pharmacists commonly interrupted their own work to relay information to other pharmacy staff, patients, physicians, nurses or insurance agents. For example, during observations, pharmacists asked their technicians to relay information to patients about the status of their prescriptions. Pharmacists were also observed calling physician offices to communicate prescription issues.

4. Finally, and least frequently, pharmacists interrupted themselves due to non-work-related issues. Some of these were related to personal/family issues that they had to deal with in the middle of their work. In most of the cases, however, these were mostly caused by non-work-related conversations that the pharmacists initiated.

Technology as a Source of Interruptions

The participating pharmacies in this study utilized various types of technologies, including hardware and software tools that directly or indirectly affected pharmacy personnel’s work. Despite their indispensable role in supporting the work of pharmacy staff, many of the technologies also contributed to a significant proportion of interruptions recorded during observations. These interruptions primarily resulted from sub-optimally designed technology.

For example, decision support alerts frequently appeared as pop-up windows, requiring pharmacists to input an override code, which interrupted their ongoing task. The lack of prioritization in presentation of alerts also resulted in technology-induced interruptions. During the observation period, pharmacists commented that many alerts represented drug interactions that were not necessarily clinically significant or were not applicable to the particular scenario and thus caused unnecessary work disruptions.

It should also be noted, however, that the dispensing software provided flexibility in the pharmacy workflow by allowing the technicians to place a few of the decision support alerts into a queue. Technicians frequently switched between computer stations, taking advantage of the multiple computers available in the pharmacy. When confronted with an alert requiring pharmacist attention, technicians would simply switch to a different computer to avoid interrupting the pharmacist for an override code. Technicians would continue doing this until all the available computers required pharmacist attention.

Another frequent form of technology-induced interruptions resulted from communication modalities including phone calls, voice messages, and public announcement (PA) systems. Communication via telephone was observed to be a key part of pharmacy operations. Consequently, interruptions frequently resulted from telephone calls or call announcement systems installed in the pharmacy. Of the 6 observed pharmacies, one had a public announcement (PA) system that alerted pharmacy personnel about an incoming call by way of a voice that is heard over a speaker embedded in the ceiling. Upon hearing the voice from the PA system, pharmacy staff would answer the call from a telephone closest to them. In the remaining 5 pharmacies, pharmacy staff simply answered the phone as calls came in to the pharmacy.

A Second Pharmacist as a Source of Interruptions

These types of interruptions represented a relatively small proportion of the interruptions recorded during the observation periods (about 5%). Common reasons why a pharmacist was interrupted by a fellow pharmacist included: need for communicating handoff information during shift changes, need for clarification of prescription information, as well as personal, non-work-related conversation that occurred occasionally. Overall, interruptions of these nature were minimal due to two key reasons. First, pharmacists mostly functioned independently of each other with most of their interactions being with technicians and patients. This situation was commonly observed in the pharmacies that had at least 2 pharmacists working during the same shift. Second, some of the observed pharmacies did not have overlapping pharmacist shifts, and as a result, there were very limited opportunities for interaction, and hence, interruptions.

Discussion

This study sought to characterize the frequency and sources of interruptions that occur in community pharmacies. Analysis of the data collected through observation methods revealed that pharmacists’ tasks were frequently interrupted due to multiple contributing factors. This is the first study on community pharmacy interruptions in the US since the publication of a study led by Flynn et al⁸ nearly two decades ago.

On average, the rate of interruptions across the studied pharmacies ranged from 5 to 13 interruptions over a period of 60 minutes. This rate is comparable to interruptions that occur in other interruption-prone work environments in healthcare such as in emergency departments. For example, Chisolm and colleagues⁷ showed that emergency physicians’ in emergency departments were observed to be interrupted a mean number of interruptions 30.9 ± 9.7 per 180-minute (or roughly 10 interruptions per 60 minutes).

The SEIPS (Systems Engineering Initiative for Patient Safety) model is used as a conceptual model that guides the discussion of this study. The SEIPS model uses a human factors systems approach to address issues in quality of care. It allows for the analysis of work systems in terms of technology, tasks, persons, environment, and organization to guide the identification of work system barriers and the design of interventions that improve patient and employee safety-related organizational outcomes. Thus, the suggestions for improvement with regard to pharmacy interruptions are appropriately framed by the aforementioned categories.²¹

Impact on Cognitive Workload

Given the complex and dynamic nature of pharmacist work in community pharmacies, these findings are understandable. In fact, during the observations, pharmacists seldom expressed surprise with being interrupted mid-task. In many instances, interruptions may have served a useful purpose in aiding communication among pharmacy staff and in preventing medication errors.⁵ However, frequent interruptions also have the potential to negatively impact the work and well-being of pharmacists. This may be explained by the high cognitive demand that results from constantly switching between different contexts for different tasks due to interruptions.²² For example, when a pharmacist is interrupted by the sight of a patient approaching a pharmacy counter or by a technician asking a question, the pharmacist must use his or her working memory to recognize the specific solution that will address the patient’s or technician’s specific needs (i.e., the new task). This, depending on the nature of the new task and level of pharmacist’s expertise and experience, can be cognitively demanding and time-consuming. As the pharmacist spends time away from the interrupted task, the time needed to complete the original task becomes longer and the working memory needed to complete the interrupted task begins to decay, thus making it difficult to resume the original, interrupted task.²³

Once the secondary task has been completed, the pharmacist must return to the original, interrupted task, complete it, and move on to the next task in line. For that to happen, the pharmacist must reload the context of the original task into his or her working memory and other necessary information used to complete the task. The delay in returning to the interrupted task, combined with the time for a pharmacist to recall his or her position in the task and other relevant task context can result in a resumption lag which, over many such events, may contribute to frustration and fatigue.²⁴

The non-appointment-based services and unrestricted access to a pharmacist are among the features of community pharmacies that distinguish them from other types of healthcare facilities. Although advantageous in terms of access to pharmacy services, this model inherently invites more interruptions due to the uncertain nature of incoming patient traffic in pharmacies. The most common type of interruptions were patient-induced interruptions, followed by technician-induced interruptions. It may not be possible to prevent all interruptions from happening, but important steps can be taken to minimize unnecessary interruptions and thus avoid the potential for medication errors. The solutions are divided into 3 salient categories: improved technology design, training of technicians, and improved workspace design.

Improved Technology Design

Some pharmacies already use mobile technology that notifies patients when their prescriptions are ready for pick-up. This serves to prevent interruptions from patients asking about the status of their prescriptions. However, during observations, pharmacy staff were observed to exhibit frustration about patients who would call to ask about the status of their prescription or would arrive at the pharmacy before their prescriptions were ready, despite being enrolled in the pharmacy’s mobile notification system. Pharmacists described feeling pressured to finish prescription-related tasks for these patients who came in unexpectedly. This suggests a need for improved patient education on the complexity of processing prescriptions and the safety checks that are incorporated into it. Furthermore, it may also suggest a need for improved design of the notification system, including use of tailored messaging to provide patients with information on the status of prescriptions. Ideally, patients would receive occasional updates on the status of their prescriptions so that they do not experience out-of-the-loop syndrome.²⁵ This serves as an opportunity to leverage mobile health technology, such as smart phones and smart watches.

Strategies to address technology-induced interruptions include better call management systems such as dedicating a pharmacy staff member (e.g. a technician) to handle all incoming calls and triage them according to perceived urgency. Some of the observed pharmacies used this strategy but did not have a formal protocol in place for it. This could be further developed and enforced through better organizational policies that foster reliable call-related interruption management.

Another strategy is to create and implement a new interface design of technology that would increase situation awareness in pharmacies. Situation awareness refers to one’s understanding or comprehension of the dynamic environment they either monitor or control.²⁶ In a study that sought to understand the cognitive processes underlying a nurse’s decision to interrupt other nurses, it was found that the interrupter would conduct a cost-benefit assessment to determine the interruptability of the interruptee and to decide whether to interrupt. When making such an assessment, the interrupter would consider (1) the conditional factors of the interruptee and interrupter, (2) the content of the interruption, and (3) the consequences of the interruption. The interrupter’s level of situation awareness can affect their ability to determine the interruptee’s interruptability.²⁷ Thus, a possible solution is to provide all staff with visibility of system status to increase situation awareness of pharmacy staff. Visibility of system status would mean that pharmacy staff could see the status of any prescription processed at the pharmacy as well as what prescription or pharmacy-related task each of their team members is working on at any given time. Both pharmacists and technicians can be aware of each other’s ongoing tasks and be more attuned to times when their team members need assistance or when they need to prevent each other from being interrupted.

Interruptions were also caused by high numbers of decision support alerts; and pharmacists were required to override them in order for their technicians to continue processing prescriptions. Such an excessive number of warnings can contribute to alert fatigue²⁸ among pharmacists. As a result, pharmacists may pay less attention to or even ignore some vital alerts, thus limiting these systems’ effectiveness. In fact, studies in other healthcare settings have shown that alert fatigue can present as a major risk for patient safety²⁹ and results in medication errors.³⁰ Improved design of decision support systems that allow only clinically relevant alerts to be displayed may minimize their negative impact on pharmacists’ and technicians’ work.

Organizational Training of Technicians

Technicians play a key role in handling customers and effectively triaging the requests of patients. As described earlier, most of the technician-induced interruptions occurred because (1) technicians were not authorized to perform a specific task due to legal or organizational policy or (2) they did not have the knowledge or expertise to perform a task and sought help from a pharmacist. Technician training and socialization into the pharmacy staff team could improve technicians’ ability to distinguish tasks requiring immediate interruption of the pharmacist from tasks that do not require this disruption to pharmacist’s workflow.

Technicians can be specifically trained to differentiate between urgent and non-urgent pharmacist tasks as well as between safety-critical and non-safety-critical pharmacist tasks. Some of these interruptions may be necessary for patient safety. However, in situations with non-urgent tasks, technicians may need to be trained on how to avoid or manage interruptions when a pharmacist is engaged in a safety-critical task such as verifying a prescription. One way to provide technicians with improved situation awareness as well as the ability to identify safety-critical pharmacist tasks is to provide them a way to view the pharmacist’s computer screen or be able to see updates to the status of a pharmacist’s ongoing tasks from their own computer monitor in the pharmacy. The increased situation awareness may help technicians improve their timing of interruptions to the pharmacist. Trainings that improve situation awareness can be further supported in a pharmacy culture that promotes continuous learning. Implementation of a post-training performance review in order to continuously improve interruption-related work processes may allow pharmacies to reap the benefits seen in continuously learning health systems.³¹

In conjunction with the technician trainings, pharmacists can be trained to write down their ongoing task on a notepad or to bookmark a particular screen set-up within the prescription processing system before addressing an interruption. These actions may help them resume their task after responding to an interruption.³² The aforementioned trainings could be targeted for use at peak business times in the pharmacy when interruption management would have the highest likelihood to prevent interruption of safety-critical tasks.

Improved Workspace Design

The design of the physical environment can help manage interruptions. Pharmacies can be built with a separate room in which the pharmacist can work alone without being interrupted by patients waiting to be helped or by technicians who need assistance. The pharmacist’s room can be equipped with a notification system that alerts the pharmacist when help is needed so that there is less pressure and social obligation³³ to immediately engage with the interruption source. This combination of workspace redesign and technology-mediation of interruptions has the potential to lessen the social pressure of responding mid-task and would allow the pharmacist to finish his or her ongoing task first. This is important because task performance can deteriorate as a result of the division of attention due to interruptions.³⁴

Additionally, when a patient is waiting, there could be a sign placed near the pharmacist that indicates that the pharmacist is engaged in a patient safety-critical task. The signage represents an attempt at providing situation awareness to patients and to show that the pharmacy is providing a patient-centered, safety-related service. Pharmacists are expected to provide both good customer service while also ensuring patient safety. These expectations are often at odds at times when a customer is waiting to be helped by a pharmacist who is in the middle of a patient safety-critical task (e.g. verifying a prescription). By providing situation awareness to patients, pharmacists can complete safety-critical tasks while also helping patients understand the reason they should not attempt to interrupt a pharmacist mid-task. Signage is just one example of a pharmacy design element that may improve patient interactions with pharmacists, but there are endless possibilities for the incorporation of physical design elements that subtly modify patients’ expectations of their interaction with their pharmacists. These design improvements would need to be balanced against their financial viability and any potentially negative impact on patient perceptions of the availability of pharmacists in answering their questions.

Another workspace change that may improve interruption management is a change to the color of the counterspace. A study on the effect of the implementation of a red “No Interruption Zone” during medication preparation in an intensive care unit found that, after a three-week period, a significant decrease (40.9%) in interruptions had occurred. The “No Interruption Zone” was modeled after the aviation industry’s sterile cockpit rule.³⁵ Different colored counter space (e.g. green, yellow, red) in pharmacies could be associated with doing work that can be interrupted, work that should be interrupted only if necessary, and work that should not be interrupted. Staff could be trained to use certain-colored space when working on tasks at different safety-critical levels, or alternatively, a feature can allow them to change the color of their workspace depending on whether they can be interrupted or not. This provides a clearer visual cue that may facilitate interruption management among staff members.

The above solutions require further study to determine their impact on interruptions. If they are successful in decreasing interruptions to pharmacists when they are in the middle of a task, they may greatly improve time-efficiency and perceived workload of pharmacy staff. More importantly, they have the potential to decrease the likelihood of medication errors that can put patients at significant risk for health-related issues.

Interventions to Prevent Self-Initiated Interruptions

Self-initiated interruptions occurred as a result of factors that were internal to pharmacists as opposed to external factors such as patient, technician, or technology-induced interruptions. A deeper examination of these types of interruptions may generate insight that can be used to inform better designs of work processes and technology that support the work of pharmacists thus minimizing the potentially taxing effects of interruptions. Researchers in cognitive psychology have identified 7 common categories of self-interruptions that are associated with computer use that may also be utilized to better understand interruptions in health care as well.²⁴ These include: changing the environment to increase productivity (Adjustment), switching because of frustration or fatigue (Break), seeking information to assist in the primary task (Inquiry), remembering the need to perform another task (Recollection), performing a task out of habit (Routine), starting a task triggered by some cue (Trigger), and performing a task simply to fill time (Wait). Salvucci³⁶ argues that three of these categories (Break, Recollection, and Routine) are purely internal while the remaining four are internal but also rely on some condition of the primary task and the work environment.

During observations, pharmacists were seen to self-interrupt because (1) they needed additional information to complete a task, (2) they anticipated technician confusion and preemptively provided information to technicians, or (3) they engaged in activities to fill time or bond with staff such as initiating non-work-related conversations. One potential solution to prevent the first type of self-interruption (needing additional information) would be to provide pharmacists with access to the patient health records used by the patients’ prescribers. This may reduce miscommunication regarding prescriptions because it would avoid the use of potentially error-prone translations of prescriptions between pharmacy and clinic or hospital systems. Access to the same health records may reduce the number of times pharmacists must directly contact physicians to clarify unclear information written on the prescriptions. In a recent study describing the use of health information technology to support community pharmacists in their management of patient medications, it was found that pharmacists valued the use of a tool that integrated secure messaging to clinicians, community health records, and a patient registry. The study concluded that health information exchange systems ease the process by which pharmacists access the information that allows them to identify and resolve medication-related issues.³⁷ Self-initiated interruptions may be prevented through proper implementation of any of the aforementioned solutions, which all play a role in improving communication in pharmacies and making work processes more efficient and effective.

Limitations

There are a few limitations in this study. First, despite our best efforts, we were unsuccessful in recruiting pharmacists and technicians working in large, corporation-type chains. Although some of the included pharmacies have similarities to these large corporation-type chain pharmacies, they are limited to fewer regions of the US and differ in organizational structure. Because it was not possible to examine interruptions in such environments, we do not know if such pharmacy environments are considerably different from those included in our study.

Second, observations of pharmacy personnel during the data collection occurred mostly during the daytime (mornings and afternoons) on weekdays. It is possible that interruption patterns may show variation by the time of the day and hence, attempts to generalize these findings to other settings should be viewed with a caution. Additionally, although participants were not told that the study focused on interruptions, the participant behavior may have been affected by the fact that they were informed that the study would focus on communication strategies. It should also be noted that the generalizability of the results from this study to other states with different laws and regulations is limited due to differences in pharmacy staff tasks. For example, some states do not require pharmacists to transfer prescriptions.

The observations in this study may not have captured the full extent of cognitive interruptions by participants. However, by addressing the aforementioned limitations, there is potential for future studies to achieve improved comprehensiveness with respect to the characterization of interruptions. Future studies should focus on the evaluation of harmfulness of interruptions, which was beyond the scope of this study.

Conclusions

This study showed that interruptions are ubiquitous in community pharmacies and can create potential patient safety hazards. The causes for these are complex and multifactorial but can be generally categorized into external interruptions and self-interruptions. Solutions guided by human factors engineering principles can be particularly useful in mitigating interruptions as the human factors approach takes into consideration the entire system with its nuanced complexities.