Surgery nurses' telephone communication: a mixed methods study with a special focus on newcomers' calls

1.1 Background

Studies on verbal telephone communication at the hospital have mostly concentrated on calls between healthcare professionals and patients. A few studies have focused on calls between hospital coworkers, investigating for instance the telephone communication of physicians (Aziz et al., 2005; Ortega, Taksali, Smart, & Baumgaertner, 2009; Soto, Chu, Goldman, Rampil, & Ruskin, 2006), among the staff of general internal medicine units (Lo, Wu, Morra, Lee, & Reeves, 2012; Whitlow, Drake, Tullmann, Hoke, & Barth, 2014; Wu et al., 2010), between floor nurses and on-call physicians (Bernstam et al., 2007) and between nurses in medical emergency centres and physicians on duty (Tjora, 2000). The majority of the existing studies revolve around safety issues related to mobile phones (Myerson & Mitchell, 2003) and their pros and cons compared with pager-mediated contact (Lo et al., 2012; Ortega et al., 2009; Soto et al., 2006) and landline telephone use (Hanada, Fujiki, Nakakuni, & Sullivan, 2006). They also examine the usefulness of telephone communication between physicians in terms of continuity of care (Blankenship, Menapace, Fox, & Frey, 1999; Crone, 1987), ways remote and technology-mediated work communication may balance the nurse–physician professional relationship (Tjora, 2000) and the reasons for after-hours nurse-to-physician calls (Bernstam et al., 2007). Previous studies also highlight the importance of shared understanding among co-workers of telephone conversation terms of use to prevent interruptions (Solvoll, Scholl, & Hartvigsen, 2013), unnecessary calls (Bernstam et al., 2007; Wu et al., 2010) and communication errors (Rabøl et al., 2011).

Telephone communication is an important part of nurses' work and is closely connected to care organization. Nevertheless, research has repeatedly shown that new graduate nurses experience challenges when communicating with co-workers, handling organizational discussions with other hospital personnel and answering the telephone (Maben, Latter, & Macleod Clark, 2006). Wangensteen, Johansson, and Nordström (2008) note that novice nurses often lack the competences required to work as members of an interdisciplinary team, not knowing, for instance, whether they should call the physician or not. Fink, Krugman, Casey, and Goode (2008) report the need for novice nurses to work on assertiveness with physicians and to learn to give orders to assistant personnel with confidence. O'Shea and Kelly (2007) identified several tasks as problematic for novice nurses: organizing investigations and referrals, contacting specialized clinical nurses and other healthcare providers such as dieticians and physiotherapists and booking tests and ordering items from the pharmacy (pp. 1538 & 1540). In response to these issues, nursing training programmes now incorporate teaching on interprofessional communication (Chant, Jenkinson, Randle, & Russell, 2002), sometimes focusing specifically on telephone transactions (Krautscheid, 2008).

Most of the above-mentioned studies are based on simulated, experimental or intervention-driven communication situations and/or used questionnaires, interviews and in situ observations as data collection techniques. There is a lack of naturalistic studies and basic information like the identity of the nurses' most common interlocutors, the number, length and direction of the calls and the purposes they serve. The present study addresses nurses' telephone communication, as practiced in situ and in real time during recorded real-life hospital calls. The study specifically aims to: (i) document the main features of surgery nurses' telephone calls with hospital personnel, including a special focus on newcomer nurses' calls; (ii) identify the main activities accomplished during the newcomers' calls.

2.1 Design

The study was part of a larger research project on hospital personnel telephone communication adopting the interactional workplace studies framework (Arminen, 2005) and based on ethnographic fieldwork and systematic audio recording of calls. For this particular mixed methods study, we combined statistical descriptive analysis and conversation analysis-based coding of calls recorded in two surgery nursing stations.

2.2 Setting

Our study was conducted in the surgery department of an acute-care hospital with approximately 200 beds in Switzerland. The surgery department comprised four different care units, two stationary and two ambulatory, each of them with its own nursing station. The study concentrated on telephone calls made or answered using the landline phones of the two stationary units, which we called units one and two. Both units provided similar care services related to general, orthopaedic, trauma, vascular and visceral surgery. Each care unit had 15 double rooms. The nursing staff of each unit was composed of approximately 35 people (1 head nurse, 24–25 nurses, 4–6 aides, 2–3 nursing interns and nursing students). In both units, the landline telephone was used by all categories of nursing personnel and for all kinds of purposes. In both units, mobile phones were used by some of the nurses (for instance, the head nurse, the nurse in charge for the day and the nurses of the night shift), but the activity of these devices was not considered for the study.

2.3 Methods

We used a service of the hospital's telephone provider to collect (directly from the hospital switchboard centre) audio recorded calls between the landline telephones of the surgery department's care units one and two and 155 telephone numbers in the hospital. These telephone numbers corresponded to 111 mobile and 44 landline phones. They were attributed to 123 different interlocutors, one single interlocutor sometimes having two different numbers, one landline and one mobile. An interlocutor could designate an individual person, a physician for instance, or a collective entity, such as the laboratory. The interlocutors belonged to 19 hospital departments.

The 24/7 recording service automatically recorded calls from the time the caller finished dialling the number to the time the handset of the nursing station landline phone was hung up. Calls were collected this way over a period of 174 consecutive days. Based on the information provided automatically by the recording system, a researcher (Author 1) created an Excel^® file with 9,931 entries, each corresponding to a different recorded call; we named the resulting collection the NTH set. It includes information on the date, time, length and direction (incoming or outgoing) of the call, the caller's and the called party's telephone numbers and the devices used (landline or mobile phone).

All the calls were listened to twice, each time by a different member of the research team (Author 3, Author 4), to identify the ones made or answered by the three newly graduated nurses in their first year of employment who agreed to participate in the study. Two female newcomer nurses were working in unit one and the other, a male, was in unit two. This collection of 374 calls forms the NTH-3 subset.

2.4 Analysis

We analyse the NTH database, which includes the NTH-3 subset, with descriptive statistics (number of events, percentages, mean, median and standard deviation) using SPSS^® software. We examined the nurses' interlocutors and the devices they used, whether landline or mobile phone, as well as the day of the week and the time, direction and length of the calls and compared the calls of the general nursing staff with those of the newcomers. The aim was merely to describe the telephone activity of the general nursing staff and the three newcomer nurses during the 6 months of the study. We did not infer that the identity of the group—the fact of being newcomers specifically—was in itself the reason for any identified differences between the two sets, which in any case did not refer to two distinct homogenous populations.

We also fully transcribed the newcomers' calls—the NTH-3 subset containing 374 calls—applying the conversation analysis conventions (Jefferson, 2004). We then coded the calls manually according to the interlocutors' main activities (Robinson, 2007; Stivers, 2015), which resulted in numerical results. The coding of the transcripts followed the orientations of the call interlocutors, who exhibited their own understanding of the ongoing activities (Koenig & Robinson, 2014).

The research team's previous scientific publications (cf. Sterie, 2015; Sterie & González-Martínez, 2017) presented qualitative conversation analysis of specific interactional phenomena found in a few selected newcomers' calls. In this article, for the first time, we present the quantitative results of the statistical descriptive analysis and the conversation analysis-based coding in their entirety.

Several steps were taken to ensure the reliability of the audio recordings (Peräkylä, 2004); we discussed the details of the recording procedure with the staff in advance to secure their support; the automatic recording system provided for minimal interference in work routines and the recordings were made uninterruptedly over an extended period of 6 months, thus becoming a common feature of the setting. The descriptive statistical analysis was done by one researcher (Author 2) and reviewed by another (Author 1). The transcription of the recordings was done collaboratively by two researchers (Author 3 and Author 4). The coding was done independently by two researchers (Author 1 and Author 3). Discrepancies between transcribers and coders were subsequently examined during team data sessions until agreement was reached.

2.5 Ethics

The research protocol was approved by the hospital's board of directors, the body responsible for ethical reviews and included ethical requirements specifically related to the collection of recordings in a clinical setting (Broyles, Tate, & Happ, 2008). We collected only internal calls between a predetermined set of telephone numbers. All the employees involved in the study were informed of the research project and the fact that recordings would be made. They were given the opportunity to freely opt out of the study if they wished to. We kept only the calls of interlocutors who agreed to be recorded. Moreover, a pre-recorded voice message reminded callers that the calls were being recorded for research purposes. We anonymized all personal information related to the telephone interlocutors and any persons referred to during the calls. According to the research protocol, we transcribed only the calls made or answered by the three newcomers, who gave voluntary written informed consent, for further conversation analysis.

3.1 The main features of the calls

3.2 Main activities in the newcomers' calls

In most of the newcomers' calls (N = 289 out of 374), the dialling resulted in a single telephone connection where one single conversation with one single main activity took place. Interlocutors thus produced monofocal conversations (Wakin & Zimmerman, 1999) where they rapidly transacted a single business. However, some calls were initially made to one telephone number and then transferred to another number, resulting in two telephone connections. During a single connection, two distinct conversations sometimes took place, each of them corresponding to a spate of talk between different interlocutors. Moreover, some conversations had no main activity, for instance when the called person immediately stated that he or she was not available to talk. In contrast, a few conversations had more than one main activity, for instance if a nurse called her manager with two matters on the agenda: she first addressed a request for instruction, the two parties discussed it and then she moved on to produce an unrelated report on a patient. This diversity of organizations accounted for a total of 356 instances of main activities out of 374 calls (Table 4).

The coding of the main activities showed that requests, where one speaker asked his or her interlocutor, more or less overtly, to do something, were by far the most frequent main activity accomplished over the phone (N = 337 of 356 identified main activities; 94.67%). In contrast to directives, requests display at least some minimal orientation to the recipient having a say in the future conduct (Craven & Potter, 2010). In 85.76% of the requests, the newcomer was the requester and his or her interlocutor the requestee. Excerpt 1 provides an example of a request.³ A newcomer nurse (Amaryse/May) calls the main number of transportation services. A porter on duty (Ronaldo/Ron) picks up the phone (1).

Following self-identification, Amaryse addresses a request for transportation to Ronaldo (2–4). She displays entitlement to produce the request but orients towards the porter having a say about his future conduct using the interrogative format “can you.” Moreover, she frames her utterance as a request with a turn-final “please” token. Ronaldo commits to granting the request; Amaryse thanks him and immediately moves to closure. Sterie (2015) and Sterie and González-Martínez (2017) present additional request formats used by the newcomers.

A total of 108 requests were about medical treatments, related to adding/changing patients' medicine, changing their diet, inspecting and redoing surgical wound dressings and requesting a physician's intervention for problematic medical situations. 80.56% of these requests were produced while the newcomer was in contact with a surgery physician.

A total of 109 requests were about transfer issues, whether of patients (N = 74) or objects (N = 35) such as medical tools, drugs and food. The most frequent interlocutors for patient transfer requests were transportation services (63.51%) and the recovery room (27.03%) since the nurses needed to go there and take the patients back to their own rooms.

Requests for contact (N = 35) were oriented towards asking the call taker to get in contact with or to facilitate getting in contact with a third party. 68.57% of such requests were produced by the newcomers calling the hospital reception, but many also involved the nurse answering incoming calls from interlocutors trying to get in touch with hospital staff or patients through the nursing station.

Most of the requests about patient discharges (N = 29) concerned preparing the discharge papers and organizing home transportation and home-care services. The newcomers produced this activity with physicians (51.72% of such requests) and, in a lesser proportion, with other hospital staff like social workers and dieticians.

Requests related to laboratory tests (N = 16) were made either by the newcomers to the laboratory or the other way around, for instance when the laboratory called about a missing blood tube for a previously ordered test. Regular orders were usually placed using paper forms. Telephone communication was used mostly in the event of unusual, urgent or out-of-hours requests or when nurses needed to amend or cancel a previous order.

Requests about appointments (N = 12) involved the nurses calling mainly physicians' secretaries. Requests for information for relatives (N = 8) happened mainly when the nurses called the recovery room to obtain news about a patient. The newcomers produced and handled requests about patient entry (N = 6) when the emergency department sought to allocate a new patient to the surgery care unit. Other requests (N = 14) concerned, among other topics, administrative issues, such as asking for a patient's identification number or card, technical assistance or help from other nurses.

Besides requests, the parties to the calls produced informings (Heritage, 1984) as main activities (N = 15). In this case, the business of the conversation was transferring a piece of information. The types of information and the nurses' interlocutors were highly varied. Excerpt 2 reproduces a call whose main activity is transferring information. A newcomer nurse (Amaryse/May) calls the Intensive Care Department; Anaïs (Ana) picks up the phone.

It transpires from the conversation that a patient was previously transferred from surgery to intensive care. Amaryse calls intensive care and informs Anaïs that she has now sent her the patient's laboratory results (5–10). Anaïs shows appreciation and Amaryse responds and moves to closure. The excerpt exemplifies a common use of the phone at the hospital to notify the called person of something sent through the pneumatic tube system or by fax, computer or other means.

The main activity of a few conversations was transferring another call to the call taker (N = 3). This happened for instance when the hospital reception contacted the nursing station with a call for a newcomer coming from a third party. In other calls, it was the nurse who acted as a switchboard operator. In Excerpt 3, a newcomer nurse (Leandra/Lea) calls the recovery room main number and Maeva (Mae) answers the phone.

Amaryse displays intent to direct a call from a patient's wife waiting on another line (2–4). Maeva tells Amaryse that she herself was about to call, discusses the patient's return to the care unit (not transcribed) and then resumes with the reason for the call, prompting the nurse to proceed with the transfer. The nurse replies affirmatively and moves to closure.

In only one conversation was the main activity giving an order, which is a directive consisting of telling an individual to do something (Craven & Potter, 2010). In this conversation, Samuel (Sam), a physician, calls the nursing station, a newcomer nurse (Leandra/Lea) replies and the physician enquires whether a patient has already arrived in the unit (non-transcribed), which is a preliminary to the main activity. In Excerpt 4, following the nurse's answer, the physician enters into the business of the call: giving an order consisting of serving food to the patient on her arrival (4–5).

In Excerpt 4, the physician formats his utterance as an informing that nevertheless does more than that he orders the nurse to follow his instructions. Physicians' instructions regarding patients' diets are entered and signed in the medical record and thus have an imperative character. The physician uses informing as a vehicle for his order and provides a reason for serving food but nevertheless shows little orientation towards the nurse having a say about the instructed conduct. On her side, Leandra produces several compliance markers (7) that Samuel may be simply acknowledging in line 8. Then, Leandra responds to Samuel's appreciation token and moves to closure.

In 39 of 380 conversations, the interlocutors did not engage in any proper main activity. During the opening, they realized that the call was no longer appropriate because for instance, the called person was about to reach the nursing station or the caller had failed to reach the intended interlocutor. Once this became known, the interlocutors proceeded to close the call. In Excerpt 5, a newcomer nurse (Amaryse/May) calls a physician on his mobile phone, but it is Franco (Fra), a member of the operating room staff, who answers.

Amaryse receipts Franco's response to the call as conveying that the physician is unavailable and, following self-identification, tells him that she will call back later (2–3). Franco acknowledges the announced future action and shows acceptance of it; Leandra displays appreciation and moves to closure.

The number of the conversations (N = 39) without any proper main activity (inappropriate interlocutors or interlocutors lacking suitable conditions for proceeding with the call), added to the requests for contact (N = 35) and activities consisting of transferring a call (N = 3), underscores the fact that the newcomers spent a significant part of their time on the phone trying to get in touch with relevant/available interlocutors or putting them in touch with their intended interlocutors.

4.1 Limitations

One limitation of the study is that it captures only a part of the telephone activity of the two participating nursing stations. The nurses also used mobile phones and were in contact with additional hospital staff as well as interlocutors outside the hospital and these calls were not recorded. The set of numbers included in the study was nevertheless significant in size and relevant for the nursing staff itself. Another limitation is that the automatically generated data do not make it possible to distinguish between calls, connections and conversations. The system would create a call entry in the database as soon as the caller finished dialling a number regardless of whether, for instance, the call was answered by one interlocutor and then automatically transferred to another number, resulting in two different connections, or answered by one interlocutor and then handed to another co-present person, resulting in two different conversations. We used the automatically generated information on connected telephone numbers as a basis for attributing interlocutors' identities, except for the coding of the main activities of the newcomers' calls, based on transcripts of the actual conversations. As we did not compare two distinct homogenous populations, we did not identify statistically significant specificities of newcomer calls compared with, for instance, experienced nurses. We merely describe the telephone activity of the general nursing staff, which included a diverse array of personnel and of the three newcomer nurses during the time of the study. Finally, there is a possibility that some interlocutors may have altered their telephone habits because they knew that the calls were being recorded. However, routine telephone practices in a fast-paced work environment are difficult to alter unilaterally and consistently, for an extended period, in a statistically significant way.