Keywords
Consultation - WhatsApp - secure messaging application - length of stay - overcrowding
- emergency department
1. Background
Emergency department (ED) crowding has been recognized as a serious concern and is
associated with prolonged Emergency Department Length of Stay (ED LOS). Consultations
occur frequently in the EDs ([1]). It is the process by which emergency physicians request other specialists (consultants)
to participate in the care of the ED patients. Generally, as do most of the countries,
the consultant should provide one of the following outcomes with this process: admission
to hospital, discharge with or without consultant follow-up, completion of procedures
or investigations, or consult another specialty in Turkey ([2]).
Hospitals in Turkey have different ED care systems due to type of hospital such as
university hospital, affiliated hospital, training and research hospital and state
hospital. Our hospital is affiliated with a state hospital. In our hospital, if primary
care ED residents request consultations for patients, consultants (faculty or senior
physicians) always return consultation requests. Residents from other departments
such as surgery or internal medicine participate in ED patient care. Residents in
the other departments can report to ED and take care of the patients.
Although many factors contribute to prolonged ED LOS, consultation difficulty and
increased consultation time are the most important problems ([1]). In order to eliminate these problems, consultation methods including the use of
telemedicine and SMS/MMS are beginning to be evaluated ([3], [4]). However, these methods have major disadvantages such as a prolonged consultation
process for telemedicine methods and per message fees applied for the use of SMS/MMS
([5]–[7]). Therefore, healthcare providers and clinicians are increasingly exploring alternative
communication technologies in order to facilitate and accelerate the consultation
process ([8]). Smartphones are a common mode of communication in daily life. Almost all health
professionals in the United States of America and the United Kingdom now carry smartphones
([9]). The widespread use of smartphones represents a new opportunity to integrate mobile
technology into daily clinical practice. Despite concerns regarding security and privacy
issues with certain applications, smartphones can have tremendous benefits in clinical
practice ([10]). In addition, the development of new applications facilitates communication among
users by improving the capture, storage, and transmission of images. The secure messaging
application (WhatsApp Messenger) is a prominent example of this type of application
([11]). It, the most popular messaging application for smartphones worldwide, provides
real time communication through text messages, videos, voice messages, and photographs.
The secure messaging application has several advantages such as rapid transmission
and no cost to the user ([9]). The secure messaging application has been used for clinical consultation, collaboration
and communication between physicians in healthcare ([9], [11]–[15], [16]).
The aim of this study was to evaluate the effect of secure messaging application usage
for consultations on ED LOS and consult time.
2. Methods
2.1 Study design and setting
This prospective, randomized controlled clinical trial was conducted according to
a detailed protocol that conforms to the CONSORT (Consolidated Standards of Reporting
Trials) statement for reporting randomized trials. This study was planned as a superiority
trial.
This study was conducted in the emergency department (ED) of a tertiary care university
hospital (affiliated with state hospital) with an annual census of approximately 180,000
visits between 01, November 2015 and 01, February 2016. Consultations in this hospital
are requested via telephone or secure messaging application. Consultants who are faculty
or senior physician and residents from other departments are out of the hospital in
many times in this hospital during night shifts. The pediatric ED is in a different
location in the province. All diagnosis and treatment algorithms are standardized.
All participants provided written informed consent.
The study was approved by the Medical Ethics Committee. Patient confidentiality was
maintained by minimizing mention of patient identifiers. All precautions were taken
to ensure that protocol of this study was compliant with the EU Data Protection Directive
95/46/EC which protects individuals’ privacy and personal data use. This study was
registered to clinicaltrial.gov (Clinicaltrial ID of NCT02586779).
Consultations were randomized into two arms. The intervention group [secure messaging
application group (WhatsApp)] included the consultations that were requested via the
secure messaging application. The control group (Telephone group) included consultations
that were requested by telephone as a verbal report.
2.2 Selection of participants
Adults (age>18 years) who were referred for consultation in the ED were eligible for
the study. Participants were enrolled consecutively 24 hrs a day, 7 days a week. Study
eligibility was determined by a resident working in the ED with attending physician.
The consulted and consulting physicians who have a smartphone, who are familiar to
use of secure messaging application, who are known as active secure messaging application
user from previous communications and who have the connection to the internet network
were enrolled in study. All consulted and consulting physicians were blinded to the
purpose of the investigation.
Consultations, which was not determined real “consult time” and the “ED LOS”, meeting
any of the following criteria were excluded from the study: an unresponsive consultation
(consultations not received response after two calls), an undetermined response time
from the consulting physician, more than one consultation, re-admittance to the ED
on the same day due to consultation failure, dead on arrival, left against medical
advice, patients stayed in the ED after consultation.
2.3 Randomization and interventions
The randomization model was generated by a computer program. Opaque sealed envelopes
were used to allocate the consultations requested into two groups. The randomization
model and preparation of opaque sealed envelopes were performed by a person. The consultant
specialist and the data collector were blinded to the purpose of the investigation.
The intervention group included consultations that were requested via secure messaging
application. In intervention group, all communications between consultant and the
ED physician were only generated via secure messaging application. A patient history,
brief report and vital findings (blood pressure, heart and breathing rate, fever)
by writing text messages (►[Figure 1]), and x-rays, ultrasound results, electrocardiograph, laboratory results, wound
photographs, computerized tomography and magnetic resonance images, monitor rhythm
sounds were sent to consultants as secure messaging messages. The control group included
consultations that were requested by telephone. A brief patient history, blood pressure,
heart and breathing rate, fever, Glasgow Coma Score, sensory-motor findings, neurovascular
and movement examinations, and relevant investigations were transmitted to consultants
as a verbal report. Eligible patients were enrolled in the study in a consecutive
manner. After a patient was enrolled in the study, the times were noted by two medical
secretaries, one of whom is responsible for record arrival time and one of whom is
responsible for initiating consults and leave time from ED recorded as minutes to
standardized study form.
Fig. 1 Example of one of the communication via secure messaging application * (text message
that emergency physician sent to consultant) (complaint): shortness of breath, COPD
(chronic obstructive pulmonary disease), CHF (congestive heart failure), HT (hypertension),
(TA): 190/70 mmHg, (Heart rate): 135/min, Bilateral rales, ** (consultant’ response)
I’m coming to the ED.
2.4 Outcome measurements
Primary outcome measurement was the difference in total ED LOS between the two groups.
Total LOS in the ED, was calculated as the ED arrival (hour and minutes when the patient
first arrived) minus ED discharge (hour and minutes when the patient left the ED)
([17]) and was presented as minutes.
The secondary outcome measurement was the difference in consult time and termination
of consultation between the two groups. Consult time was calculated as the first consultation
time (time when the first consultation was requested) minus the time of disposition
decision (time the bed was requested for admitted patients or time of ED discharge
for discharged patients) ([17]) and was presented as minutes. In the time between the initiation and end of consultation,
the ED physicians performed the advices of consultant (such as treatment, laboratory
follow up, etc.).
2.5 Statistical analysis
The numeric data were expressed by mean± standard deviation (SD) or median (interquartile
range (IQR)) where applicable. Categorical data are expressed as rates. The Kolmogorov–Smirnov
test was used to assess normality of continuous variables. The Mann-Whitney U test
was applied for the comparison of data that did not conform to a normal distribution.
The study was designed as a superiority trial requiring 42 patients per group with
an alpha critical value of 0.05, 97% power, 12 minutes difference between groups and
standard deviation of 15. All of the hypotheses were constructed as one tailed tests.
A 95% confidence interval (CI) was used for expressing the study data. The study data
were analyzed using MedCalc Statistical Software for Windows, v12.7.0.0 (Ostend, Belgium)
and G*power v3.1.9.2. For all analyses, statistical significance was defined as p
< 0.05. All the analyses were performed according to the intention to treat analysis.
3. Results
We assessed 439 consecutive consultations for study eligibility. From the eligible
consultations, 94 (21.4%) consultations were excluded due to following reasons; 46
consultations (48.9%) were more than one consultation, 18 (19.2%) were stayed in the
ED after consultation, 13 consultation (14.9%) had undetermined response time, 11
patients (11.7%) declined to participate 3 (3.2%), patient had re-admittance to ED
on the same day, 2 patients (2.1%) left against medical advice, 1 patient (1.1%) was
dead on arrival. A total of 345 consultations were randomized for the study: 173 patients
for the secure messaging group and 172 patients for the Telephone group. All the patients
allocated to a study arm were evaluated by a consultant physician and included in
the final analysis (►[Figure 2]). Consultant responded the consultation during second call for two patients in the
intervention group and one patient in control group. However, these patients were
included into the final analysis. The mean age of the study participants was 48.5±22.1
years; 67.3% (n=231) of the study participants were men. There was no statistical
difference was between two groups according to age, gender and numbers of consultation
during night shifts (p>0.05 for each other). The baseline characteristics are detailed
in ►[Table 1].
Fig. 2 CONSORT Patient flow diagram
Table 1
Baseline characteristics of the requested consultations for secure messaging application
and telephone groups
|
Variables
|
WhatsApp Group (n=173)
|
Telephone Group (n=172)
|
|
Age (years) ± SD
|
47.1 ± 20.2
|
46.5 ± 22.9
|
|
Male, no. (%)
|
111 (64.2)
|
120 (70.6)
|
|
Consultation during night shifts, no. (%)
|
100 (57.8)
|
96 (55.8)
|
|
Most common clinics requested consultation, no. (%)
|
|
Orthopaedics
|
45 (26.0)
|
43 (25.3)
|
|
Cardiology
|
44 (25.4)
|
42 (24.4)
|
|
Internal Medicine
|
36 (20.8)
|
34 (19.8)
|
SD: Standard deviation
Two hundred and seventy-three secure messaging application messages were sent to consultant
specialists by Emergency physicians. Emergency physicians receipted from consultant
specialists the 367 response messages as text message. Of 367 response messages, 58.0%
(n=213) were instructiongiving comments, 30.2% (n=111) were the clinical questions,
%9.2 (n=34) response messages were administrative questions and %2.5 (n=9) were other
questions. Any problem was not reported about secure messaging application by consulted
physicians.
Median total ED LOS was 270 minutes (IQR:240 to 279) for all patients included in
study. The median ED LOS was 240 minutes (IQR:230 to 270, 95%CI: 240 to 255.2) for
patients in the Secure messaging application group and 277 minutes (IQR:270 to 287.8,
95%CI:277 to 279) for patients in the Telephone group. The median total ED LOS of
the Secure messaging application Group was significantly lower than the Telephone
Group (median dif: –30, 95%CI:-37 to –25, p < 0.0001, ►[figure 3a]).
Fig. 3 Boxplot graphics of time differences between two groups; 3a. Difference of total
Emergency Department Length of Stay between secure messaging application and Telephone
Groups; 3b. Difference of Consult Time between Secure messaging application and Telephone
Groups
The median consult time was 170 minutes (IQR:150 to 180) for all patients included
in the study. The median consult time was 158 minutes (IQR:133 to 177.25, 95%CI:150
to169) for patients in the Secure messaging application group. The median consult
time was 170 minutes (IQR:165 to 188.5, 95% CI:170 to 171) for patients in Telephone
group. The median total consult time of the secure messaging application Group was
significantly lower than the Telephone Group (median dif.: –12, 95%CI: –19 to –7,
p<0.0001, ►[figure 3b]).
Consultations completed without ED arrival was 61.8% (n=107) in the Secure messaging
application group and 33.1% (n=57) in the Telephone group (dif: 28.7, 95% CI:48.3
to 66, p<0.001). Consultation with Secure messaging application eliminated more than
half of in-person ED consultation visits.
4. Discussion
Our findings suggest that the Secure messaging application reduces ED LOS and consult
time for ED patients who require consultation. In traditional ED consultations, verbal
report by telephone is the most commonly used method. This method does not allow for
the transfer of X-rays, wound image, CT and MR images to the consulting physician
([12]).
Earlier studies have shown that consultation by telephone and/or other methods such
as pager systems can result in a number of important limitations ([8], [14], [18], [19]). Tashkandy et al ([21]) found that consultation by telephone increased ED LOS.
To the best of our knowledge, this is the first randomized, double-blind clinical
trial of the effect of secure messaging application usage on ED LOS and consultation
time. The median total ED LOS was lower among consults conducted using secure messaging
application relative to consults conducted by telephone. This may be explained by
the view that the usage of secure messaging application lowers barrier to consultation
and allows the resident to consult earlier, because more data can be shown via pictures
and images to the consultant instead describing over the telephone.
Other studies have evaluated the effectiveness of secure messaging application implementation
and use in health care. A study by Giordano et al ([22]) investigating consultations requested using the secure messaging application Messenger
reported on the use of secure messaging application for initial documentation and
transmission of images in the emergency department to more experienced consultants;
the authors concluded that this approach is safe for clinical use. Another well-designed,
prospective, study by Khanna et al ([23]) compared the effectiveness of communication prior to the introduction of secure
messaging application and after the introduction of secure messaging application in
an orthopedic clinic. They reported that the secure messaging application allows residents
to function efficiently, improves communication and awareness regarding admitted patients,
and facilitates handovers without significant disruptions to daily working routines.
A systematic review by Lee et al. ([24]) reported that consult time is the most time consuming procedure in the ED, and
reducing consultation time may improve ED patient flow. A study by Samuel et al ([25]) found that the mean total consultation time for all consultants was greater than
2 hours. In our study, median consult time was 170 minutes for all consultations and
consult time for consultations that were requested using the secure messaging application
was shorter compared with consultations by Telephone.
Increasing the number of consultant visits to the ED leads to the increasing ED LOS
and the ED crowding ([2], [16], [20], [22]). The most time consume process in consultation is the time arriving of consultant
to the ED. A prospective cohort study by Brick et al. found that consult time was
most influenced by time from arrival to consultation request. ([26]). In present study, the use of secure messaging application resulted in more consultations
ending without the specialist attending the ED. A recent report by Qureshi et al ([27]) reported that consultations that required the presence of the consulting physician
in the ED were the most time consuming. In the present study, 61.8% of consultations
were completed using secure messaging application alone. Thus, consultation via secure
messaging application reduced ED LOS by reducing the number of occasions on which
a consultant needed to report to the ED.
Consulted physicians are able to look at radiographies in an electronic system in
hospital. But, the consultant may not be able to connect to hospital database or may
experience an inability in accessing patient information systems anywhere, if especially
a clinician is mobile or out of hospital as do our country ([2]). However, secure messaging application is almost always with us. Therefore, consultants
reach to all patient data anywhere.
The novelty of using app on the outcome is effect that advantage of secure messaging
application send to consultant large amount of patients’ data and consultant’ decision
is fasten and thus reduced to ED LOS. We thought that it may be preferred in patients
who has not complex clinical condition and when specific radiographic images and laboratory
data are ready at the time of consultation time. Thus, the consultation time reduction
might get less.
There are some concerns about security and privacy issues associated with the use
of communication technologies. However, until now, there is no scientific evidence
that patient information which was sent to consultant via Secure messaging application
reached the other person ([10], [28]) and this concerns is only a prediction. Secure messaging application has end-to-end
encryption and when end-to-end encrypted, tthe messages, photos, videos, voice messages,
documents, status updates and calls are secured from falling into the wrong hands.
([16]) Therefore, Secure messaging application is compatible with Health Insurance Portability
and Accountability Act (HIPAA), which is United States legislation that provides data
privacy and security provisions for safeguarding medical information ([29]). It ensures the protection of personal and contact data with a password. All necessary
measures in communication via secure messaging application can be taken by the usage
of password protection for both the smartphones and the application, screening and
securing the devices with antivirus software along with prior instructions to the
participating residents ([23]). Furthermore, to address concerns regarding the security of communications containing
health data, health care providers or Health Ministry officials working in collaboration
with information governance services may choose to create a new mobile application
similar to secure messaging application program.
5. Limitations
This study has several limitations. First, despite high attendance number of our hospital,
this study was conducted with small sample size because of study design. Patients’
chronic illness, elapsed times for sending a secure messaging application message
and ED LOS in patients requiring more than one consultation could not be evaluated.
Second, how long it take physicians to do a consultation in telephone and Secure messaging
application groups and consultation response times of consulted physicians were not
compared between two groups. Third, there are a number of exclusions from this study,
some of which may have had an impact on the results had they been included or are
important outcomes when evaluating a new system for making referrals e.g. non responsive
consultation, consultation failure. Because we could not calculate ED LOS and consult
time of these consultations, we had to exclude these patients. A large multi-center
study would help in the establishment role of secure messaging application on the
ED LOS and consult time.
6. Conclusion
Duration of ED LOS and consult time for consultations requested via secure messaging
application is shorter in comparison to conventional Telephone consultation. Consultations
via Secure messaging application reduce the need for a consultant to report to the
ED by more than half. The use of secure messaging application for consultations may
significantly improve ED LOS and consult time.
ED crowding and prolonged ED LOS continues to be a major problem. Use of mobile Health
applications has gained the momentum in recent years. This study demonstrates the
effectiveness of simple messaging software to improve ED LOS and consult time. Further
studies are needed for a more detailed description of secure messaging application
use in the ED as new consultation method.
Multiple Choice Questions
Multiple Choice Questions
What is benefit of using secure messaging application (WhatsApp) compared to telephone
consultation?
Answer: D. Consult time and ED LOS is the shorter in consultations via secure messaging
application than consultations via standard telephone and reduce the need for a consultant
to report to the ED by more than half.
