Healthcare 7 (2019) 100331

Contents lists available at ScienceDirect

Healthcare
journal homepage: www.elsevier.com/locate/healthcare

Case report

Implementation of smart phones to facilitate in-hospital telephone T

communication: Challenges, successes and lessons from a neonatal intensive
care unit
Colleen A. Hughes Driscolla, , Samuel Gurmub, Ahad Azeemc, Dina El Metwallya,d
⁎

a
University of Maryland School of Medicine, Department of Pediatrics, Baltimore, MD, USA
b
University of Maryland Medical Center, Clinical Engineering, Baltimore, MD, USA
c
Northwell Long Island Jewish Hospital, New Hyde Park, NY, USA
d
Suez Canal University, Pediatrics Department, Ismailia, Egypt

ARTICLE INFO ABSTRACT

Keywords: Implementation Lessons 1. Mobile telephony use in the hospital setting is complex and sub-optimal im-
Clinical communication plementation of mobile communication technology can create inefficiencies in clinical workflow 2. Objective
Hospital communications systems measurement of mobile technology’s impact on clinical communication workflow is necessary to identify and
Smartphones remediate associated inefficiencies in real-time 3. Functionality between mobile applications and devices should
Software applications
be evaluated when implementing technology, particularly when an application is non-native to a device 4.
Telecommunications
Continual collaboration between front-line clinicians and technical teams allows for early identification of ad-
verse impacts from, and optimization of, mobile communication technology implementation.

1. Background telephone communication.

Collaborative patient management across a hospital setting requires 2. Organizational context

reliable and efficient communication between healthcare providers.1–5
Caregiver communication can take on many forms, with telephony This study was performed in a newly designed Level IV NICU which
being an important mechanism for sharing critical clinical informa- treats neonates throughout the state. The NICU is housed within an
tion.6,7 Despite the importance of telephony in hospital communication, urban academic medical center in the Mid-Atlantic, USA. It is staffed by
successfully connecting healthcare providers by telephone can be physicians, nurse practitioners, nurses, trainees, and various ancillary
fraught with inefficiencies.6 Mobile smartphone technology has the service providers. Staff generally remain present in the NICU physical
potential to facilitate critical communication among providers and space but may leave the NICU for conferences, transporting patients, or
improve patient care delivery.8–10 performing consults and procedures in other areas within the hospital.
The potential benefits of mobile communication have prompted the The new physical space houses 52 individual patient rooms and re-
emergence of devices, software platforms, and applications to enhance placed an older NICU with open bay patient care areas. As a result, the
mobile communication for clinicians.11 With the increasing number of size of the clinical care areas significantly increased. Nurses were
applications and devices available, successful integration of these tools spread among workstations rather than at centralized nursing stations,
are necessary to ensure efficient communication.1,12,13 A number of decreasing the opportunities for face-to-face communication.
technical factors, including how devices and applications are designed To reduce noise for patients and staff, an overhead paging system
to work together, can impact implementation; yet, few studies have was not included as a mechanism for clinical communication. Its ab-
evaluated the integration of multiple technologies on clinical commu- sence was a unique feature within the institution, as other units in-
nication.10,12,14–16 cluding the previous NICU, relied on overhead page notification to
In this case study, we describe the implementation of smartphones signal a need to respond to incoming telephone calls.
to facilitate telephony in a newly designed neonatal intensive care unit The unique physical design and absence of overhead paging in the
(NICU), and the optimization of device-application interaction for NICU highlighted the need for reliable mobile telecommunication. To

⁎
Correspondence to: University of Maryland School of Medicine, 110 South Paca Street, 8th Floor, Baltimore, MD 21201, USA.
E-mail address: cdriscoll@som.umaryland.edu (C.A. Hughes Driscoll).

https://doi.org/10.1016/j.hjdsi.2018.07.001
Received 1 March 2018; Received in revised form 5 July 2018; Accepted 18 July 2018
Available online 14 August 2018
2213-0764/ © 2018 Elsevier Inc. All rights reserved.
C.A. Hughes Driscoll et al. Healthcare 7 (2019) 100331

accommodate these needs, individuals are provided Motorola MC40 Table 1

smartphones (Zebra Technologies Corp., Lincolnshire, Illinois). The Pattern of observed unit secretary phone call outcomes.
devices are shared among staff and handed off at shift changes. Calls to Call Outcome Pre-intervention Post-intervention
devices can only be received from within the institution, as not to N (%) N (%)
disrupt clinical workflow with unscreened or non-urgent calls. Calls to
Successful transfer 813 (47.5) 336 (70.1)
the unit are screened by the unit secretaries and answered on Cisco desk
Text/Retransferred 441 (25.8) 53 (11.1)
phones (Cisco Systems, Inc., San Jose, CA). Unit secretaries transfer Unsuccessful 126 (7.3) 7 (1.4)
phone call from within and outside the hospital to individual devices, as Not applicable 328 (19.2) 79 (16.6)
appropriate to facilitate communication between parties. If phone calls Total 1708 475
are unanswered after a pre-determined number of rings, the call is
chi-square 95.6, p < 0.05 for overall groups.
automatically forwarded back to the unit secretaries.

3. Personal context between 9 a.m. and noon. Calls were assigned to one of four categories:
successful, text/ retransfer, unsuccessful and not applicable. Calls were
The re-organization of the NICU physical space was designed and considered successful if the call was forwarded by the unit secretary to
engineered by a multi-disciplinary project team that included re- the intended recipient's device on first attempt. Calls were categorized
presentation from the medical team, pharmacy, environmental services as text/retransferred if the secretary made more than one attempt to
and infection control. This group influenced the decision to transition forward a call to the intended recipient's device or if the secretary
from open-bay rooms to individual patient rooms, as individual rooms texted the recipient to inform them of the call. Calls were categorized as
were felt to best support intensive and developmental care delivery for unsuccessful if the call did not reach the intended recipient by any
our neonatal population.17 This decision was in keeping with current mechanism. Calls that did not require forwarding to an individual
trends to privatize care areas for neonates in our region.18 within the NICU were considered not applicable. To determine the time
The motivation to eliminate overhead paging was based on staff required to reach a bedside nurse by phone, two phone calls per ob-
dissatisfaction with the frequency and disruptive nature of overhead servation period were made to bedside nurses by a member of the study
pages that occurred in the previous NICU, as well as concerns related to team. These calls were categorized as successful (the bedside nurse was
noise levels experienced by patients and families. These pages were not reached within 180 s) or unsuccessful (disconnected or greater than
limited to communication regarding telephone calls but also broader 180 s to connect). The duration of time to successfully connect to a
clinical communication including bedside alarms and emergency alerts. bedside nurse was measured using a stopwatch application. Data were
For example, notification regarding the need for a neonatal resuscita- prospectively collected from 27 May 2016–4 October 2016 (see results
tion team alone generated 60–100 overhead pages per month. for pre-intervention data in Tables 1 and 2).
Neonatal ICU physician and nurse leadership determined the com- The technical team suspected that the difficulty with auto-launch
munication objectives, while the clinical engineering department was was because the voIP application was non-native to the device, meaning
essential in finding technology that would support these communica- the application was not specifically designed for the operating system.
tions and provide ongoing support for the chosen technology. Non-native applications are designed to function on a variety of devices
Cooperation from the information technology (IT) department was re- and operating systems, however, this flexibility may lead to compro-
quired for supporting Wi-Fi integration. mised integration with other applications on the device.19 Alter-
natively, native applications are built to function on a specific device or
4. Problem operating system, inherently providing smoother integration between
multiple applications on the device. Our goal was to improve telephone
To enable telephone functionality of our mobile devices the tech- communication efficiency by replacing the non-native voIP application
nical team installed Cisco Jabber version 8.4 (Cisco Systems, Inc. San with a native application. We hypothesized that the native application
Jose, CA) as the voice-over IP application (voIP). Shortly after im- would decrease the time required to reach bedside nurses by phone and
plementation staff reported that phone calls were not being successfully increase the percent of telephone calls that were made successfully on
transferred from the front desk to individual staff members. first attempt. We also anticipated decreased number of calls to the unit
Investigating further, the technical team discovered that when a staff secretary desk phone, as fewer unanswered calls would be re-routed
member had not used their phone after a period of time the device went back to the secretary.
into sleep mode, as expected. However, in sleep mode the voIP appli-
cation also unexpectedly logged off. When the device was taken out of 5. Solution
sleep mode by the user the voIP application attempted to launch with
an auto-login. The auto-login frequently failed so that the device could In December 2016, Jabber was replaced with Workforce Connect
not make incoming or outgoing phone calls without manual config- (Zebra Technologies), the native voIP of the MC40 device. While the
uration by a member of the technical team. Unfortunately, users were native application was expected to be associated with greater efficiency
unaware when a device was in need of such configuration, and the for communication, committing to the application required additional
unrecognized issue prohibited phone calls from being received by the investment from our institution. Since the telecommunication needs of
user. These phone calls were routed to the secretaries at the front desk the NICU were unique and not generalizable across the hospital, the
phone. Often the secretaries would attempt to transfer the phone call expense was heavily considered by hospital administration. Additional
back to the intended receiver's device which was still non-operational,
leading to a vicious cycle of call transfers until it was recognized that Table 2
the voIP application was logged off. Sometimes the secretaries would Comparison of applicable unit secretary phone calls.
use alternative methods to notify a recipient of an incoming phone call, Outcome Pre-intervention Post-intervention
such as text messaging or transferring calls to another desk phone N (%) N (%)
nearby the intended recipient. Anecdotally, this created frustration and
Successfully transferred on first attempt 813 (58.9) 336 (84.8)
inefficiency among NICU staff, consultants, patient families, and an-
Not successfully transferred on first 567 (41.1) 60 (15.2)
cillary support staff. attempt
To estimate the extent of the problem we monitored a convenience
sample of phone calls to unit secretaries during 2-h observation periods chi-square 90.6, p < 0.05 for overall groups.

2
C.A. Hughes Driscoll et al. Healthcare 7 (2019) 100331

investment in NICU communication technology had broader implica- integration with patient monitoring systems and electronic health re-
tions because the administration had not yet committed to supporting cords, support of personal devices, equipment durability, infection
our particular smart device across the hospital system. It was unclear control, battery life, asset management, customer support capabilities,
how the NICU technology could be adapted across other hospital units and interoperability of devices and applications.11,13 Above all, the
and workflows. Ultimately, our data on failed telephone communica- technology must support the various clinical processes associated with
tions among providers (see Tables 1, 2) influenced additional monetary patient care. If any of these technical factors cannot meet the needs of
investment in the native application. the clinicians’ workflow, healthcare delivery and patient outcomes may
To estimate the impact of our solution, we prospectively collected be adversely impacted.8,20
post-intervention data from 26 January 2017–22 March 2017. In our case, mobile technology that was meant to support commu-
Normality of the sample was assessed with Kolmogorov-Smirnov test. nication resulted in ineffective and inefficient communication pro-
Continuous data are expressed as median and range. Categorical data cesses. The frequent failures of our telephone application were likely
are expressed as number (%). Differences between the pre- and post- caused by application-device incompatibility. The burden imposed on
intervention groups were assessed using independent sample t-test and the staff was immediately felt. Clinicians found themselves using work-
Mann-Whitney U tests for non-parametric continuous data and chi- arounds (e.g. texting) and alternative devices (e.g. desk phones) to fa-
square tests for categorical data. Statistical significance was set at cilitate communication. Over time our data demonstrated the extent of
p < 0.05. Analysis was completed using Socscistatistics, the communication system's inefficiency, ultimately leading to the im-
Physics.cbsju.edu/stats and R Statistical Software. plementation of an alternative application for telephony. Following this
A total of 2183 observed calls were received by the unit clerk during implementation, we observed a reduction in the number of calls to the
the study period; 1708 during the pre-intervention period and 475 unit secretary per hour, an increase in successful call transfer on first
during the post-intervention period. The overall pattern of phone call attempt, and a reduction in time to reach a bedside nurse by phone.
categories significantly changed significantly between the pre- and Since telephony is our primary mechanism for communicating with
post-intervention period with an increase in the percentage of calls specialty consultants, operating room staff, and our regional neonatal
being successfully forwarded; chi square 95.6, p < 0.05 (Table 1). transport team efficient mobile communication has the potential to
Among applicable calls, the percentage that were successfully for- improve healthcare delivery for our patients. Future study of our
warded to their intended recipient on first attempt significantly in- communication system on healthcare delivery and patient outcomes is
creased during the post-intervention period; chi-square 90.6, p < 0.05 ongoing.
(Table 2). A principle highlighted by this case is the need for close compat-
The median number of calls per hour to the unit secretary desk ibility between software and hardware. While a variety of software
phone significantly decreased during the study period; 14 (range 3–26) applications can serve a similar communication function, each may be
during the pre-intervention period and 10.5 (range 3–19) during the markedly different with respect to design, cost, functionality, reliability
post-intervention period (p < 0.05). and user experience.19 In our case, the difference in reliability between
A total of 108 calls were placed by the research team to bedside the native and non-native application was important. While in many
nurses during the observation period; 70 during the pre-intervention cases native applications are user-friendly they can be associated with
period and 38 during the post-intervention period. The percent of greater expense and/or limited integration with other devices and op-
successful phone calls (within 180 s) was unchanged between the pre- erating systems.19 With multiple manufacturers and developers com-
and post-intervention periods (88.5% pre- vs 94.7% post-; chi square peting for market share in the healthcare sector, it can be difficult for
1.1, p = 0.29). However, the percent of calls that took longer than 60 s clinicians to know which applications will be most beneficial for
to reach the intended recipient significantly decreased during the post- workflow needs. Additionally, as in our situation, problems with in-
intervention period from 30% to 5.2% (chi square 8.9, p < 0.05). tegration may not be obvious in the pilot stage or even immediately
Among calls that were successfully placed, there was a 44% reduction after larger scale deployment. Therefore, we recommend that leader-
in the median duration to reach the intended recipient. (Fig. 1). ship teams assign professional technical experts to partner with clin-
icians to understand their needs and assist with decision-making for
6. Unresolved questions and lessons from the field software and hardware, prior to and following implementation.
Another difficulty that we faced was concern about financial in-
Hospital-wide implementation of mobile technology is complex and vestment in an additional application for telephony. Financial con-
problems with integration can be a source of inefficiencies in healthcare siderations are a common barrier to clinical mobility implementation in
delivery.20 Technical factors that must be considered prior to im- hospitals.12 The cost of mobile telephony not only involves the initial
plementation include Wi-Fi and cellular infrastructure, data security, purchase of devices but also annual license fees, any upgrades to the
supporting infrastructure, and technical support for users and for sys-
tems integration. Data on the long-term benefits of mobile technology
investment are limited, making it difficult to weigh against the risks of
monetary commitment.14–16 Helping us overcome that barrier was the
ability to provide data on the application failures. The high degree of
failure was able to be viewed by everyone as unacceptable, leading to
investment in the replacement application.
This experience has emphasized the importance of gathering data
on clinical workflows both before, during, and after implementation of
any healthcare technology. Outcomes related to clinical processes must
be measured to ensure that the deployed technology does not lead to
increased work for staff. The promise of mobile technology is efficient
connectivity among patients and healthcare providers, while being free
of multiple or static devices. Without objective assessment of the
technology's impact on communication processes, hospitals risk poorer
healthcare delivery and additional strain for clinicians. Engineering and
evaluating health information technology systems to optimize patient-
Fig. 1. Box plot of duration to reach intended recipients. centered outcomes was identified as a national goal almost 10 years

3
C.A. Hughes Driscoll et al. Healthcare 7 (2019) 100331

ago; yet, further research on clinical mobility and healthcare delivery clinical staff in the emergency department. Med J Aust. 2002;176(9):415–418.
outcomes is still needed.1 8. Standing S, Standing C. Mobile technology and healthcare: the adoption issues and
systematic problems. Int J Electron Healthc. 2008;4(3/4):221–235.
The commitment to support this technology in the NICU led to 9. Wu RC, Morra D, Quan S, et al. The use of smartphones for clinical communication on
broader questions about technology use in other areas of the hospital. internal medicine wards. J Hosp Med. 2010;5(9):553–559.
Whether the current technology can support the clinical workflows of 10. Wu RC, Tran K, Lo V, et al. Effects of clinical communication interventions in hos-
pitals: a systematic review of information and communication technology adoptions
other units and hospitals within our system is a question that remains to for improved communication between clinicians. Int J Med Inform.
be answered. Additionally, who bears the responsibility and ownership 2012;81(11):723–732.
of the technology can be difficult to determine for products that are 11. Healthcare Information and Management Systems Society. Complexities of
Developing a Solution for Mobile Care Team Communication. Available from: www.
used by clinicians, but require technical integration and, in some cases, himss.org/file/1321972/download?Token=I1OWcIvV.
clinical engineering. As the devices are used for expanding capabilities 12. Wu RC, Lo V, Rossos P, et al. Improving hospital care and collaborative commu-
(secure texting, photodocumentation, clinical alarm notifications, etc.), nications for the 21st century: key recommendations for general internal medicine.
Interact J Med Res. 2012;1(2):e9.
the need for close collaboration between these hospital groups is
13. Facchinetti T, Fernando A, Quoi R. This process is just beginning: connecting mobile
emerging rapidly within our institution. We believe that this colla- medical devices. Horizons. 2012;Fall:19–25.
boration will be an essential characteristic of future success with clin- 14. Prgomet M, Georgiou A, Westbrook JI. The impact of mobile handheld technology on
ical mobility. hospital physicians' work practices and patient care: a systematic review. J Am Med
Inform Assoc. 2009;16(6):792–801.
15. Free C, Phillips G, Watson L, et al. The effectiveness of mobile-health technologies to
References improve health care service delivery processes: a systematic review and meta-ana-
lysis. PLoS Med. 2013;10(1):e1001363.
16. Mosa AS, Yoo I, Sheets L. A systematic review of healthcare applications for smart-
1. Institute of Medicine. Engineering a Learning Healthcare System: A Look at the phones. BMC Med Inform Decis Mak. 2012;12:67.
Future: Workshop Summary. Rountable on Value & Sciece-Driven Health Care ed. 17. White RD. Recommended standards for the newborn ICU. J Perinatol. 2007;27(Suppl
Washington, D.C.: The National Academies Press. 2):S4–S19.
2. O'Daniel M, Rosenstein AH. Professional Communication and Team 18. Frank K, Hughes Driscoll C, Shapiro K, Donohue PK. Moving to a single patient Room
CollaborationApr; 2008. 〈http://www.ncbi.nlm.nih.gov/pubmed/21328739〉. Design: are there physical and psychological Effects on the neonatal intensive care
3. Zinn C. 14,000 preventable deaths in Australian hospitals. BMJ. unit staff? J Pediatr Neonatal Care. 2016;4:4.
1995;310(6993):1487. 19. White J. Going native (or not): five questions to ask mobile application developers.
4. Coiera E. Communication systems in healthcare. Clin Biochem Rev. 2006;27(2):89–98. Australas Med J. 2013;6(1):7–14.
5. Institute of Medicine. To Err is Human: Building a Safer Heatlh System. Washington, 20. Hughes Driscoll CA, Schub JA, Pollard K, El-Metwally D. A wireless text messaging
D.C.: National Academy Press; 2000. system improves communication for neonatal resuscitation. Am J Med Qual.
6. Coiera E, Tombs V. Communication behaviours in a hospital setting: an observational 2017;32(3):3007–3312 [May/Jun].
study. BMJ. 1998;316(7132):673–676.
7. Coiera EW, Jayasuriya RA, Hardy J, Bannan A, Thorpe ME. Communication loads on