Introduction
In March 2020, the World Health Organization (WHO) declared the SARS-CoV-2/Novel Coronavirus-19
(COVID-19) a global pandemic. As of March 23, 2020, more than 300,000 cases have been
reported worldwide [1 ]. Patients can present with varying degrees of symptoms and according to one report,
86 % of all infections were undocumented prior to January 23, 2020 [2 ]. Therefore, risk of infection reported in healthcare workers is substantial [3 ]. In particular, while performing gastrointestinal endoscopy, there is risk of exposure
to the endoscopists as well as the endoscopy team, including nurses, endoscopy technicians
and anesthesia staff [4 ]. While upper gastrointestinal endoscopy is an aerosol-generating procedure, there
are now data to suggest that the risk may not limited to upper endoscopy alone, as
recent reports have detected SARS-CoV in stool samples [5 ]. This has led to recommendations that all elective and non-urgent endoscopic procedures
be cancelled or postponed at this time [4 ]
[6 ].
However, important questions have emerged regarding how to define an urgent procedure
vs a non-urgent procedure, or a procedure that can be deferred for a discrete period
of time. In some clinical scenarios, the decision to perform or delay a procedure
is evident. For example, there is clear consensus that procedures for indications
such as suspected variceal bleeding, non-variceal upper gastrointestinal bleeding,
acute cholangitis, foreign body removal, and cancer-related care (i. e. tissue acquisition
for diagnosis, loco-regional staging, and palliative procedures) are urgent and should
continue to be performed [7 ]
[8 ]
[9 ]. Similarly, endoscopic evaluations of chronic symptoms such as diarrhea and gastroesophageal
reflux disease (GERD), or screening for colorectal cancer in average-risk individuals,
are considered non-urgent and should be deferred. Between these definitions exists
a large array of potentially time-sensitive but not technically urgent or emergent
endoscopic interventions. These grey areas or “semi-urgent” indications pose a clinical
dilemma for the gastroenterologist with regard to proceeding with or deferring the
procedure during this unprecedented time.
We aimed to survey gastroenterologists worldwide using Twitter to help elucidate these
definitions using commonly encountered clinical scenarios during the COVID-19 pandemic.
We hypothesized that there would be significant variability regarding procedures and
indications considered urgent or non-urgent, highlighting the need for further guidance
and standardization in identifying time-sensitive procedures.
Methods
A 12-question survey was designed by the authors. The goal was to choose common clinical
scenarios that do not have clear guidelines regarding the timing or urgency of endoscopic
evaluation or treatment during the current COVID-19 pandemic [4 ]
[6 ]. These questions were posted on Twitter using the “Twitter poll” option (by the
author MB). This author was chosen given that that he has more than 2,500 followers,
with the majority being gastroenterology fellows or gastroenterologists from across
the world. The initial tweet described the framework of the survey in the context
of the current pandemic. The questions were posted under the comments section of the
initial tweet. Numerous gastroenterologists and gastroenterology fellows from across
the world with prominent presence on Twitter were tagged to help disseminate the survey.
Four additional questions were added based on request from other gastroenterologists.
The questions were open for polling to Twitter audience for 48 hours, after which
Twitter automatically closes the survey to polling. The survey was completely anonymous.
The results were analyzed using the “Tweet Activity” function on Twitter.
Definitions
Regarding Twitter analytics, two definitions are important to understand. Impressions
are the number of times a tweet appears on the timeline or “feed” of Twitter users.
Engagements refers to the number of times a user becomes involved in a tweet. These
engagements include retweets, likes, replies, or as is important to this study, poll
answers.
For the purpose of this manuscript, endoscopies that were not classified urgent/emergent,
or elective were described as “semi-urgent.” Semi-urgent endoscopy was defined as
a procedure that could reasonably be deferred for at least 8 weeks without negatively-impacting
an important patient outcome (e. g. upstaging of a new cancer diagnosis).
Results
The initial Tweet had 38,795 impressions and a total of 2,855 engagements. The details
of the tweet were expanded 2,291 times. The number of votes received on the initial
polls ranged from 338 to 834, providing an estimate response rate ranging from 11.8 %
to 29.2 %. The four additional questions added later had a lower response rate as
expected with an average of 44 votes polled. The summary of the results is outlined
in [Table 1 ]. The actual Twitter analytics report can be accessed at: https://twitter.com/BilalMohammadMD .
Table 1
Results of survey regarding timing of endoscopic procedures during COVID-19 pandemic.
Indication
Most common decision
Percentage of highest votes, % (n)
Gastrointestinal bleeding
Everyone should have endoscopy
44.5 (438)
Scope patients with hemodynamic instability
61.9 (389)
Everyone should have EGD
87.6 (410)
EGD for acute onset of dysphagia
55.8 (407)
Evaluation of dysphagia
Defer endoscopic evaluation for now
58.3 (48)
Defer endoscopic evaluation for now
84.7 (359)
Concerns for dysplasia/cancer
Defer procedure for now
70 (834)
Defer chromoendoscopy for now
52.6 (19)
Defer colonoscopy for now
63.3 (355)
Defer EGD for now
58.5 (340)
Defer EGD for now
50.4 (421)
Defer EUS for now
53.5 (368)
Defer endoscopic resection for now
53.5 (383)
Benign [pancreatobiliary indications]
ERCP only if abdominal pain/jaundice
35.7 (338)
Defer ERCP for now
44.1 (59)
Defer EGD/ERCP for now
66 (50)
EGD, esophagogastroduodenoscopy; FIT, fecal immunochemistry testing; EMR, endoscopic
mucosal resection; ESD, endoscopic submucosal dissection; EUS, endoscopic ultrasound;
ERCP, endoscopic retrograde cholangiopancreatography
Scenario 1 focused on patients with positive fecal immunohistochemistry testing (FIT)
or fecal FIT-DNA test, and 70 % of respondents suggested that colonoscopy was semi-urgent.
Scenario 2 focused involved Barrett’s esophagus with dysplasia and/or nodularity needing
endoscopic treatment, and 50.4 % of respondents deemed this as semi-urgent. Scenario
3 included patients with a benign ampullary adenoma needing endoscopic resection,
and 53.5 % of respondents voted this as semi-urgent in the current setting. Scenario
4 questioned respondents regarding patients with melena, and 44.1 % thought that “any
melena” needs urgent upper endoscopy, while 53.8 % thought only patients with “ongoing
melena” or hemodynamic instability” should undergo endoscopic evaluation. Scenario
5 discussed patients presenting with hematochezia, and the majority (62 %) thought
only patients with hemodynamic instability should get inpatient colonoscopy. In patients
with cirrhosis who had symptoms of upper gastrointestinal bleeding (scenario 6), 87.6 %
thought that they should get urgent upper endoscopy. For patients presenting with
dysphagia (scenario 7), 55.8 % of participants suggested performing esophagogastroduodenoscopy
(EGD) during this time only if dysphagia was acute in onset. Scenario 8 was regarding
patients with a double duct sign on cross-sectional imaging, but without a discrete
mass seen, and 53.5 % suggested it was semi-urgent. In patients presenting with isolated,
unexplained weight loss (scenario 9), 84.7 % thought this was semi-urgent. 63.5 %
thought that all colonic endoscopic mucosal resections (EMRs- scenario 10) could be
deferred at this time. In regard to endoscopic submucosal dissection (ESD) for early
gastric cancer (scenario 11), 58.5 % of respondents thought this was semi-urgent.
For patients with common bile duct (CBD) stones without cholangitis (scenario 12),
35.8 % thought that urgent endoscopic retrograde cholangiopancreatography (ERCP) is
only needed if patients have symptoms or jaundice, 29 % felt that ERCP was not urgent
if no cholangitis, while 25.1 % deemed all CBD stones as urgent. Scenario 13 discussed
planned endoscopic removal or exchange of plastic biliary stents previously placed
for scenarios currently resolved or for which the patient was currently asymptomatic.
45.6 % thought this was semi-urgent and can be deferred ([Fig. 1 ]). For patients who had a pancreatic duct (PD) stent placed during a prior ERCP (scenario
14), 66 % respondents favored deferring this as semi-urgent. In patients with iron
deficiency anemia (scenario 15) without overt gastroinestinal bleeding, 58.3 % thought
this was semi-urgent. In a patient with long-standing ulcerative colitis with recent
diagnosis of dysplasia (scenario 16), 57 % of respondents voted to defer performing
chromoendoscopy at this time.
Fig. 1 Demonstration of variation in gastroenterologists regarding timing of procedures for
semi-urgent procedural indications.
Discussion
Our results show that there is significant variability among gastroenterologists in
regard to the timing of endoscopic procedures for semi-urgent indications during the
COVID-19 pandemic. There were only three of 16 scenarios in which greater than 70 %
of gastroenterologists agreed on procedure timing. These scenarios were deferring
colonoscopy for patients who had fecal FIT-DNA or FIT positive testing, performing
urgent endoscopy for patients with cirrhosis presenting with melena and hematemesis,
and deferring endoscopic evaluation for unexplained isolated weight loss. In regard
to other scenarios such as the endoscopic treatment for patients with dysplastic Barrett’s
esophagus, patients with ampullary adenoma needing resection and patients with double
duct sign on imaging (without a focal mass), approximately 50 % to 55 % of the gastroenterologists
thought these procedures were semi-urgent and should be deferred. While approximately
half of respondents thought that procedures for these indications should be performed
in select situations during this time.
Gastrointestinal bleeding is routinely considered an indication for urgent endoscopy.
However, in our survey regarding patients presenting with hematochezia, the majority
of respondents indicated that colonoscopy should only be pursued if a patient has
hemodynamic instability. Similarly, in patients with melena, fewer than half (44 %)
of respondents thought that every patient with melena warranted endoscopy, while the
remainder opted for endoscopy only if ongoing signs of melena or hemodynamic instability
were present. These findings are interesting since they might represent a change in
typical gastroenterology management pathways necessitated by the COVID-19 pandemic.
The highest degree of variability was seen in answers related to ERCP in patients
with CBD stone without cholangitis. One-fourth of respondents indicated that all CBD
stones are urgent, while 45 % suggested performing ERCP only if symptoms were present,
and approximately 30 % indicated that ERCP was semi-urgent if cholangitis was not
present. This highlights that there is currently no consensus on optimal timing for
ERCP in patients with asymptomatic choledocholithiasis, even prior to the COVID-19
pandemic.
Another interesting finding was that ESD for early gastric cancer was suggested to
be deferred by 58 % of respondents at this time, but 35 % of respondents indicated
this was an urgent procedure. One hypothesis for this could be that ESD can be a long
procedure and it might concern gastroenterologists that longer procedure times might
increase risk of COVID-19 transmission. Secondly, ESD also carries a higher AE event
rate than routine endoscopy [10 ], so there could be concerns that in case of complications such as perforation, there
would be utilization of the operating room services in an already resource-constrained
environment. It is interesting to note that the highest degree of variation in responses
was for premalignant conditions such as Barrett’s esophagus, ampullary adenomas, and
ulcerative colitis with dysplasia. This suggests that there is some uncertainty among
gastroenterologists regarding deferring treatment of premalignant conditions in the
current environment.
There was also variation in answers in regard to patients who underwent prior ERCP
with biliary and PD stent placement. While most respondents suggested these procedures
should to deferred during this time, some indicate that ERCP for biliary stent exchange
and PD stent removal should be prioritized and performed during this period.
There could be several reasons for variations in responses seen in our survey. The
respondents are from all over the world and there may be practice variation in different
regions of the world. In addition, the COVID-19 pandemic is in different phases throughout
the world and as the crisis worsens, the definition of semi-urgent endoscopy may narrow.
It’s plausible that respondents from hard-hit Western European countries such as Spain
and Italy have a stricter definition of what warrants endoscopic evaluation during
this pandemic. This cross-sectional analysis captures the opinion of the respondents
at a specific time in the pandemic, the severity of which varies by locality. As the
crisis worsens, a longitudinal study may well show that these opinions on what constitutes
“semi-urgent” endoscopy narrow over time as disease prevalence increases. The variability
of responses could also be driven by the fact that currently there is no defined duration
for this pandemic and some gastroenterologists might be concerned about the uncertainty
of how long a deferred patient would have to wait until their procedure is finally
performed. This becomes especially important in patients with premalignant conditions.
The Joint GI Society Message on COVID-19 does state that some non-urgent procedures
are higher priority and examples included prosthesis (e. g. stent) removal and evaluation
of significant symptoms [11 ]. However, as the number of COVID-19 cases are exponentially rising in many regions,
another major concern for performing any endoscopy is the amount of personal protective
equipment (PPE) that is needed to perform a single procedure safely. PPE is an important
resource at this time which needs to be judiciously used. Hence, many gastroenterologists
are opting to defer many of the procedures above, even though in the “non-pandemic”
situation these are likely to be performed sooner.
Our study has several limitations. Given that this survey was conducted on Twitter,
we do not have an exact response rate. We did, however, use the engagements and the
number of times the details of tweets were expanded on the initial Tweet as the denominator
and number of votes as the respondents to estimate the response rate. Our estimated
response rate is low, however, previously reported response rates are applicable to
traditional methods like email and regular mail, and may not apply to social media
platforms like Twitter. This highlights the need to develop new standards of data
acquisition/surveys as social media platforms are going to be increasingly and effectively
used for this purpose in the future. Also, we cannot tell how many of the people
who cast votes were from which countries and what level in their career (gastroenterologists,
gastroenterology fellows, internal medicine physicians). Expanding on this, there
were poll questions regarding procedures such as ERCP and ESD, which a minority of
the respondents actually perform in clinical practice. It is unclear whether answers
would vary if only physicians credentialed in these techniques were allowed to respond.
As previously mentioned, this crisis is in different phases throughout the world.
At the time of this manuscript, Spain and Italy continue to have dire situations regarding
PPE availability, and inpatient volume has exceeded the critical care threshold capacity
in many of these centers. We were not able to stratify responses by geographical region
to assess for variability, which would have certainly provided important information.
In addition, releasing a poll on Twitter will only capture physicians who are currently
engaged in using this particular social media platform, and the popularity of Twitter
use varies among different countries. However, we did find significant engagement
from across the world (North America, South America, Europe, Asia, Africa and Australia),
and participation from gastroenterologists from both the academic and community setting.
Conclusion
Despite these limitations, our analysis provides a real-time snapshot of the current
thoughts of gastroenterologists around the world during the COVID-19 pandemic. It
also highlights that there is currently a lack of consensus regarding how to prioritize
certain potentially time-sensitive endoscopic procedures. Although each patient is
unique and many clinical decisions must be made on a case-by-case basis, our analysis
will provide some perspective and guidance to gastroenterologists while dealing with
these clinical scenarios. Our findings also strongly support the need for developing
societal guidance in these “semi-urgent” scenarios to assist during the current COVID-19
pandemic, and we are aware of numerous gastroenterology societies that are engaged
presently in this endeavor. Finally, this study shows how social media platforms can
be positively used to gain instantaneous and clinically useful information from around
the globe in response to rapidly changing situations.