Introduction
Since the first case of the novel 2019 coronavirus disease (COVID-19) was identified
in Wuhan, China on November 17, 2019, the disease has spread across the globe and
infected over 4.9 million individuals, resulting in over 328,000 deaths [1]. These figures likely underestimate of the true burden of COVID-19 due to uncounted
deaths outside of medical settings as well as the high prevalence of asymptomatic
or minimally symptomatic carriers, which facilitates rapid transmission in densely
populated areas [2]. Currently, the United States has the highest caseload worldwide, and since identifying
its first case on March 1, 2020, New York City has emerged as the world epicenter
of the pandemic [3]. COVID-19 has profoundly affected the care of patients with gastrointestinal diseases,
both through measures implemented to mitigate the spread of the virus, and through
diversion of human and material resources from gastroenterology services to areas
in greater need during this crisis.
Early in the course of the pandemic, endoscopy societies around the world recommended
restricting the practices of endoscopy units to protect patients, frontline healthcare
workers (HCWs), and to conserve personal protective equipment (PPE). In the United
States the major professional gastroenterology societies issued a joint statement
recommending that endoscopy units strongly consider rescheduling elective, non-urgent
endoscopic procedures [4]. These recommendations aligned with guidance from the US surgeon general calling
for cancellation of elective surgery. Additional recommendations were issued regarding
the pre-screening of all patients undergoing endoscopy with history of high-risk exposure,
including travel to areas with a high prevalence of COVID-19. For endoscopy units
located in New York City, these multi-society guidelines implied that essentially
all patients should be considered high risk for COVID-19 exposure.
In light of evidence that SARS-CoV-2, the virus that causes COVID-19, is present in
the GI tract and potentially transmissible through secretions [5]
[6], the use of enhanced contact and droplet splash precautions as well as the regular
use of aerosol filtering face-piece respirators has been recommended [7]
[8]. It is believed that both upper and lower endoscopy have the potential for aerosol
generation due to the insufflation and aspiration of digestive secretions required
during routine procedures. Repici et al reported their experience from Italy, once
the epicenter of the pandemic, and proposed universal use of respirator masks, a hairnet,
goggles/face shield, two pairs of gloves, and a gown for high risk patients (or patients
in high-prevalence areas). Due to shortages of PPE, particularly of N95 respirators,
the extent to which these recommendations are being followed in practice remains unknown.
In New York, HCWs have reported significant difficulty securing masks and other PPE,
and it is unknown to what extent these issues have affected endoscopy staff and units
in this region [9].
Despite the cancellation of elective procedures, there remains a steady influx of
patients requiring urgent inpatient or outpatient endoscopy. Our aim was to describe
the approach taken by endoscopy units operating during the COVID-19 pandemic, both
in terms of characterizing the type of procedures performed during this period, as
well as the operational measures implemented including endoscopy suite staffing, PPE
utilization, and pre-procedure COVID-19 testing.
Methods
Survey design
A web-based survey was distributed via the New York Society for Gastrointestinal Endoscopy
(NYSGE) list-serve. Each individual NYSGE member had the opportunity to designate
one member of their endoscopy faculty to contribute institutional based data. In addition,
direct emails were sent to all endoscopy directors and division chiefs in the NY region.
As our goal was to provide time-sensitive information, distribution was limited to
this area given the early clustering of cases.
The surveys were exempt from Institutional Review Board approvals as they were entirely
anonymous and did not contain identifiable information. However, each respondent was
asked to confirm this at their respective institutions. Subsequently each respondent
was asked to enter the data in a REDCap database that was housed at Columbia University
Irving Medical Center.
An institutional identifier was used for the sole purpose of recognizing multiple
entries from the same institution. As several hospital systems comprise multiple larger
hospitals, data entered separately for each hospital was treated as distinct.
Given the daily fluidity of the numbers, we asked respondents to average the values,
when appropriate, based on the time period of the questionnaire.
The first questionnaire was distributed on 3/30/2020 and data was collected until
April 12, 2020 (representing the first 2 weeks of the pandemic). This survey is found
in the Supplementary Methods, Appendix. An updated survey targeting the changes in
response to the pandemic at its peak in this region was distributed on April 20, 2020
and data were collected through May 2, 2020.
Survey data reporting and analysis
We collected data on the number of attending and trainee physicians, nurses and other
staff members in the endoscopy unit prior to and during the pandemic.
Baseline data on the weekly number and type of endoscopies performed was collected.
Subsequently, data regarding procedures performed at the two time points in the endoscopy
unit and the ICU was collected. Patients with confirmed COVID infection at the time
of endoscopy were also assessed. Continuous variables such as procedure volume were
summarized using means and interquartile ranges (IQR) and were compared between different
time points using a paired sample t-test. Categorical variables such as procedural indications were reported as the percentage
of institutions performing a specific procedure for that indication.
In addition, data were collected regarding staff exposure, guidance on procedural
prioritization and PPE use.
Results
Characteristics of surveyed institutions
Eleven large primarily academic endoscopy units in the New York City region responded.
Data were collected from a total of seven hospital systems; four systems contributed
data from one hospital endoscopy unit each (Montefiore Medical Center, Mount Sinai
Hospital, Memorial Sloan Kettering, Northwell Health) while the remaining three systems
contributed data from multiple units (New York University: Tisch Hospital, Bellevue
Hospital, and Ambulatory Care Center; New York Presbyterian: Weill Cornell Medical
Center and Columbia University Irving Medical Center; State University of New York:
Downstate and Stony Brook Medical Centers). We estimated the number of COVID + inpatients
and critically ill (ICU) patients based on an average at two time points and based
on publicly available reporting from each institution to their faculty and staff.
The data are summarized in [Table 1].
Table 1
Institutional burden of COVID infections.
|
Patient census N (n)[*]
|
Percent of total beds
|
Mean census per institution
|
Range per institution/IQR
|
|
Total hospital beds
|
9376 (10)
|
|
937.6
|
330 – 2650/599
|
|
COVID + inpatients (weeks 1 – 2)
|
5112 (9)
|
56.5 %
|
568
|
92 – 2162/172
|
|
COVID ICU patients (weeks 1 – 2)
|
1365 (9)
|
15.1 %
|
151.7
|
23 – 613/133.5
|
|
COVID + inpatients (weeks 3 – 5)
|
3139 (7)
|
57.6 %
|
448.4
|
107 – 1538/234
|
|
COVID ICU patients (weeks 3 – 5)
|
1121(7)
|
20.6 %
|
160.1
|
35 – 571/134
|
* N, total number of patients and staff; n, number of institutions able to provide
data.
Staffing
Endoscopy physician staffing levels were cut markedly during the pandemic. These data
are summarized in [Table 2]. The number of attendings performing endoscopy during the first 3 weeks of the pandemic
decreased from 19.2 to 4.4 per unit, representing a 76 % reduction from baseline staffing
levels (P = 0.014). The number of fellows performing endoscopy decreased from 13.4 to 0.6 per
unit, or a decrease of 96 % (P < 0.001). At the time of the initial survey, 80 % of attendings were not performing
procedures with trainees. At the time of the second survey, 60 % of attendings were
not performing procedures with trainees.
Table 2
Staffing levels in gastrointestinal endoscopy units during the COVID pandemic.
|
Total staff across all institutions N (n)[*]
|
Mean staff per institution
|
Range per institution/IQR
|
|
Gastrointestinal attendings pre-COVID
|
173 (9)
|
19.2
|
8 – 50/21.5
|
|
Gastrointestinal fellows pre-COVID
|
134 (10)
|
13.4
|
8 – 18/7
|
|
Gastrointestinal attendings performing procedures (weeks 1 – 2)
|
42 (10)
|
4.2
|
2 – 10/4
|
|
Gastrointestinal fellows post-COVID (weeks 1 – 2)
|
6 (10)
|
0.6
|
0 – 3/1.25
|
* N, total number of patients and staff; n, number of institutions able to provide
data.
Overall endoscopy procedures volumes
There was a dramatic and significant reduction in total volume of all procedures between
the first survey and baseline data ([Fig. 1a]), with total procedure volumes declining from a mean of 192 cases per week per unit
to a mean 19 per week (P < 0.001). There was a modest, non-significant increase in total number of cases between
the first and second questionnaire (P = 0.42). Of note, there was a significant increase in the number of COVID-positive
patients undergoing endoscopy between the first and second time points, 0.9 vs 4.8
patients per week per unit respectively (P < 0.001), ([Fig. 1b]).
Fig.
1 Proportion of COVID + cases and average number of endoscopic procedures performed
at surveyed institutions. a Percentage of COVID + cases and COVID + patients in ICUs among all hospitalized patients
in institutions surveyed. b Average number of endoscopies performed per week. The percentage of COVID + cases
is indicated.
Procedure indications
The most common indication for both upper and lower endoscopy, which were performed
at all institutions during the pandemic, was bleeding. All sites reported performing
EGD and colonoscopy for this indication at the second time point. In addition, the
majority of centers reported performing ERCP for cholangitis and obstructive jaundice
during both time periods. In addition, at the time of the second survey, foreign body
extraction, feeding tube placement and upper endoscopic stent placement was performed
in > 60 % of the centers. Other common gastrointestinal procedures were performed
at less than half of the centers surveyed ([Fig. 2]).
Fig. 2 Endoscopic procedures during the COVID Pandemic. a Percentage of EGD indications performed at each institution. b Percentage of colonoscopy indications performed at each institution. c Percentage of ERCP indications performed at each institution. d Percentage of EUS indications performed at each institution.
COVID testing and availability of personal protective equipment
Pre-procedure COVID-19 testing increased between the first and second surveys, with
44 % and 60 % of centers offering pre-procedure testing for both inpatient and outpatient
procedures at the time of the second survey. Most endoscopy units, 90 % in our study,
reported the use of a structured protocol to identify appropriate patients for endoscopy.
None of the participating institutions reported a shortage of PPE ([Fig. 3]).
Fig. 3 COVID testing and Personal protective equipment (PPE) use at each institution. a Percentage of institutions adopting practices and PPE use during the pandemic.
Discussion
In this survey of endoscopy leaders from 11 large academic centers in the New York
region, we describe the operations of endoscopy units from every major hospital system
in the epicenter of the COVID-19 pandemic during the early to peak pandemic period.
Our study demonstrates that while some aspects of the response to the pandemic have
been consistent across centers, there is significant variation in key areas, including
pre-procedure testing for COVID-19 among patients who require endoscopy. A consistent
finding in this study is that endoscopy volumes declined markedly from baseline in
accordance with directives from government and professional societies to postpone
and cancel all non-urgent, elective cases. By surveying unit leaders on the types
of procedures performed during the pandemic, we were able to characterize the indications
that were considered sufficiently urgent to warrant intervention during the crisis,
which may be instructive to other gastroenterology departments who find themselves
in an earlier stage in the natural history of the pandemic.
Overall procedure volumes experienced a marked decline compared to baseline across
all procedure types due to cancellation of elective cases. These findings highlight
the fact that a large proportion of the work done in endoscopy units, even in large
academic medical centers, consists of screening, surveillance, and other elective
and semi-elective indications that can be safely deferred during a crisis to conserve
resources. These are profound alterations in practice that are essentially without
precedent in the modern era of gastrointestinal endoscopy. Our findings with regard
to procedure volume – while extreme – are consistent with reports from other parts
of the world, including China [10], Italy [11], as well as a multicenter survey from the United States and Canada, who describe
reductions in overall procedure volume of 75 % to 99 %, with most centers operating
at ≤ 10 % of their normal endoscopy volume [12]. These findings are also consistent with a survey of 69 individual gastroenterologists
(as opposed to endoscopy unit leaders) from the New York area who report individual
procedure volumes falling to zero cases in the preceding 7 days for 62 % of respondents,
and weekly endoscopy volumes declining by 57 % to 96 % across all procedure types
[13].
Although there has been general consensus among professional societies that elective
cases should be postponed, it remains unclear how restrictive the threshold for urgent
and emergent cases should be, and which particular indications should continue to
be performed. Our survey found several indications that were consistently considered
urgent enough to warrant proceeding during the pandemic across all centers: primary
inpatient general endoscopic procedures for an indication of gastrointestinal bleeding,
gastrostomy tube placement, and biliary endoscopy for the management of obstructive
jaundice and cholangitis. These findings are consistent with a study that surveyed
gastroenterologists via social media on indications for endoscopy during the COVID-19
pandemic, where significant variability was found regarding recommendations on timing
of endoscopy for the majority of indications, however relative consensus was present
for performing endoscopy for suspected upper gastrointestinal bleeding, symptomatic
choledocholithiasis or cholangitis [14]. In this survey, the majority of respondents thought only patients with hemodynamic
instability due to bleeding should get inpatient colonoscopy. Our findings reveal
colonoscopy was performed for the indication of gastrointestinal bleeding at roughly
half the rate of upper endoscopy for this indication, suggesting that clinicians in
New York continue to evaluate hematochezia with both endoscopy and colonoscopy at
reasonable frequency. Another study attempted to reach a consensus amongst 14 invited
expert endoscopists on the degree of urgency of 41 common indications for advanced
endoscopy [15]. High rates of consensus were obtained for proceeding within 1 week in patients
with indications of jaundice with or without cholangitis, dysphagia, gastric outlet
obstruction, or bowel obstruction requiring intervention, as well as endoscopic drainage
of symptomatic fluid collections. Gastrointestinal bleeding was not evaluated in this
study.
Significant variability was found in this survey between centers regarding pre-endoscopy
COVID-19 testing procedures. This variability may reflect the initial scarcity of
testing resources during the early phase of the pandemic in New York. We found a slight
increase in testing rates between the first and second survey, which coincides with
increased testing capability. Of note, testing was not associated with an increase
in procedure volumes, likely because testing was instituted prior to resumption of
elective procedures. There is significant evidence that COVID-19 is frequently associated
with gastrointestinal symptoms, in particular diarrhea—with one-third of patients
reporting digestive symptoms in a large cohort from New York City [16]. This raises the possibility that COVID-19 may represent an etiology of gastrointestinal
symptoms that may prompt investigation, and possibly endoscopy, in patients who present
with extra-pulmonary COVID-19 symptoms, especially as units begin to resume semi-elective
and elective procedures. These facts support the practice of universal testing for
patients undergoing endoscopy, assuming the risk of transmission is significant and
that deferring endoscopy until such time as the patient is no longer polymerase chain
reaction-positive is feasible. The true risk of exposure to gastroenterologists and
endoscopy staff on COVID-19 positive patients is unknown. One study of patients and
HCWs practicing at 41 endoscopy units in Italy during the pandemic found that 4.3 %
(including 23 physicians) contracted COVID-19, of whom 14.3 % required hospitalization
for interstitial pneumonia [17]. The majority of centers in this study (71 %) reported no cases of infection amongst
their endoscopy teams. While the authors interpret these findings to suggest a low
risk of COVID-19 infection from performing endoscopy, we argue that this rate of infection
was unacceptably considering the marked reduction in endoscopy volume and the 10 %
baseline infection rate among HCWs in Italy at the time of that survey. As we consider
how to reopen endoscopy units and resume the care of less urgent but nevertheless
important indications for endoscopy, testing will undoubtedly be a central part of
the strategy to minimize the risk of infection of staff and COVID-19 negative patients
who are present in the facility, and we suspect rates of testing will increase over
time.
The centers who participated in this study were overwhelmingly large academic institutions,
and this provided information regarding the involvement of fellows in procedures during
the pandemic. The majority of centers did not allow fellows to participate in procedures
during the pandemic, however these restrictions appear to have eased somewhat between
the two survey timepoints. These findings are consistent with recommendations from
endoscopy societies to limit trainee involvement in order to minimize procedure duration,
reduce the number of individuals exposed in the procedure room, and to conserve PPE
[18]. Given the uncertain timeframe of the pandemic and restrictions required to mitigate
its spread, the impact on skills of fellows and competency acquisition may be considerable.
These restrictions are particularly important for advanced endoscopy fellows whose
training time is finite and who have little opportunity to recoup training opportunities,
even as endoscopy units begin to open at a fraction of their previous capacity. This
secondary effect warrants further study.
There are several implications of our study for what needs to come next: reopening
and resuming normal gastrointestinal endoscopy operations. The American Society for
Gastrointestinal Endoscopy (ASGE) has made several key recommendations, including
pre-procedure screening of all patients with a dedicated COVID-19 screening questionnaire,
however they stopped short of recommending universal viral testing for all patients
prior to endoscopy. Our study demonstrates that a significant proportion of centers
are already testing for COVID-19 in all patients undergoing endoscopy, and as availability
of testing materials increases, we anticipate a move toward universal pre-procedure
testing. The ASGE also recommends universal N95 respirator use for all members of
the endoscopy team. Our study indicates that these measures are currently in place
in all the surveyed centers and supports their universal implementation as feasible.
Given the significant discrepancies we have demonstrated between baseline and pandemic
era procedure volumes across all procedure types, significant challenges will no doubt
arise with the scaling up of operations to fulfill the backlog of deferred cases and
return to volumes approaching baseline.
Our study has several strengths. Most importantly, all major academic hospital systems
in New York City were included, hence the survey is able to present a comprehensive
picture of city-wide endoscopy activities in the epicenter of the COVID-19 pandemic
during the beginning and peak of the pandemic. The cooperative involvement of multiple
institutions in the survey reduces the likelihood that procedure volume was diverted
from one institution to another, and not captured in the survey. There are several
limitations however. The survey targeted large academic hospitals, and did not include
community hospitals, ambulatory surgery centers, or office-based procedures. During
the beginning and peak of the pandemic, these centers were likely to have been shut
down, hence it is unlikely they would account for a significant proportion of endoscopy
taking place in the New York City area, however, as centers reopen, survey results
that exclude these centers will likely be biased. The survey requested aggregate,
de-identified data and hence specific demographic and clinical information for each
patient was not available.
Conclusion
In summary, our survey demonstrates the profound impact COVID-19 has had on the care
of patients with gastrointestinal diseases, with marked reductions in endoscopy volumes.
As the peak of the pandemic passes and we plan ahead to resuming elective procedures,
we expect a large backlog of cases that will present challenges given the novel additional
burdens of pre-screening, testing, social distancing, and personal protective equipment
that now accompany the operations of endoscopy units.
