Introduction
As the third leading cause of cancer-related deaths in both men and women in the United
States, colorectal cancer (CRC) is a significant health concern [1]. In 2018, there were over 140,000 estimated new cases of CRC and over 50,000 CRC
related deaths, with an overall lifetime risk of 4 % to 5 % [1]. Fortunately, CRC incidence and mortality rates have been steadily declining over
the past 30 years, primarily due to increased screening rates but also in part due
to advances in screening. Colonoscopy is currently considered the preferred screening
test, especially for high risk individuals, with studies showing colonoscopy reduces
CRC incidence by 40 % and mortality by 50 % to 60 % [2]
[3]
[4].
Use of colonoscopy has steadily increased over the past decade, becoming the most
commonly used test for CRC screening in the United States [5]
[6]. There is concern about whether there are adequate resources to meet the need for
CRC screening, with several predictive models showing a shortage of gastroenterologists
[6]
[7]
[8]
[9]. In addition, with total US healthcare costs exceeding $3.5 trillion annually, there
is increasing focus on providing high-value care [10]
[11]. Therefore, innovative solutions are needed in gastroenterology both to drive down
costs and to bridge the growing gap between supply and demand for colonoscopy. Increasingly,
less invasive screening methods are being considered to address these needs. However,
another proposed solution is extending endoscopic training to non-physicians, such
as nurse practitioners (NPs) [12]. This approach would potentially maintain the diagnostic and therapeutic benefits
of colonoscopy at a lower cost while also addressing workforce gaps. This alternative,
however, is contingent on the ability of non-physicians to provide adequate and comparable
quality colonoscopy to gastroenterologists.
Current literature on the performance of non-physicians in endoscopy is limited, with
the majority of studies focused on non-physician performance of flexible sigmoidoscopy.
There is evidence to support that non-physicians can perform flexible sigmoidoscopy
safely and effectively, with comparable adenoma detection rates (ADRs), adverse events
(AEs), and patient satisfaction compared to physicians [13]
[14]. However, much less is known about the performance of non-physicians in colonoscopy.
This lack of evidence prompted a 2009 statement by the American Society for Gastrointestinal
Endoscopy (ASGE) that “there is insufficient data to support the use of non-physician
endoscopists to perform colonoscopy” [15].
The aim of our study was to assess the quality of colonoscopy performed by three fellowship-trained
NPs ([Table 1]) after endoscopic training using quality indicators for colonoscopy as defined by
the ASGE/American College of Gastroenterology (ASGE/ACG) Taskforce. The three NPs
were selected from a pool of candidates who applied for the yearlong nurse practitioner
fellowship program. All three of the selected NPs completed the fellowship and remained
employed at the same tertiary medical academic center for a minimum of 3 years post-fellowship
completion.
Table 1
Fellowship-trained nurse practitioner demographics.
|
Characteristics
|
N = 3
|
|
Age (yrs)
|
|
|
3
|
|
|
0
|
|
Gender
|
|
|
3
|
|
|
0
|
|
Highest Education
|
|
|
0
|
|
|
2
|
|
|
1
|
|
Race
|
|
|
2
|
|
|
1
|
Methods
This was a single-center retrospective study analyzing consecutive screening colonoscopies
performed by three NPs after completion of endoscopic training. The NPs were trained
during a 1-year NP gastroenterology fellowship program, structured with the same didactic
and endoscopic training as first-year medical gastroenterology fellows. Each NP completed
a minimum of 140 supervised colonoscopies during training, which is equivalent to
the minimum number of colonoscopies recommended by the ASGE to achieve competency
[16]. The NPs then underwent a formal performance evaluation to demonstrate competency,
including skills of biopsy and polypectomy, before being credentialed and privileged
to perform colonoscopy independently within the institution. The first NP successfully
completed her endoscopic training in 2010, the second NP in 2011, and the third NP
in 2012.
The study was conducted at a large urban outpatient endoscopy center that was part
of an academic medical center between September 2010 and June 2016. Inclusion criteria
were all patients undergoing colonoscopy for CRC screening purposes. Exclusion criteria
were patients undergoing diagnostic colonoscopy and procedures aborted due to an extremely
poor bowel preparation. All patients received procedural sedation with the use of
either anesthesia-administered propofol or proceduralist-administered fentanyl and
midazolam.
Outcome measures included quality indicators for colonoscopy as defined by the ASGE/ACG
Taskforce [17]
[18]. The primary outcome was ADR. Secondary outcomes included cecal intubation rate,
mean withdrawal time, and complication rates of colonic perforation and post-polypectomy
bleeding.
Data collection was performed by review of electronic medical records and endoscopy
procedure reports. Data were entered into a secured Excel spreadsheet, from which
statistical analysis (mean, range and percentages) was calculated using Excel 2016. Institutional
review board (IRB) approval was obtained prior to initiating data collection. Adenoma
detection was histologically confirmed. Cecal intubation was verified by photo documentation
of cecal landmarks. Withdrawal times were determined using time stamps of cecal images
and retroflexion images. Incidence of complications were obtained from procedure report
documentation and chart review of any post-procedure hospitalizations.
Results
A total of 1,425 colonoscopies were performed by the three NPs during the study period.
Of these, 413 colonoscopies were excluded due to indications other than screening
(n = 374) and procedures aborted due to poor bowel preparation (n = 39), resulting
in 1,012 colonoscopies included for analysis. Patients had a mean age of 56.2 years
(range 41 – 83), 51.5 % were female, and 73.9 % were African American ([Table 2]). Informed consent was obtained from all patients.
Table 2
Patient demographics.
|
Characteristics
|
N = 1,012
|
|
Age (yrs)
|
56.2 ± 6.7 (41 – 83)
|
|
Gender
|
|
|
521 (51.5 %)
|
|
|
491 (48.5 %)
|
|
Race
|
|
|
748 (73.9 %)
|
|
|
201 (19.9 %)
|
|
|
24 (2.4 %)
|
|
|
19 (1.9 %)
|
|
|
20 (2.0 %)
|
N = number of procedures
NP colonoscopy performance outcomes were compared to the proposed standards for colonoscopy
quality indicators recommended by the ASGE/ACG Taskforce ([Table 3]). Overall, adenomatous polyps were detected in 360 procedures, equating to an overall
ADR of 35.6 %, with the proposed standard being ≥ 25 %. More specifically, the ADR
was 39.3 % in males and 32.1 % in females, with the proposed standard being ≥ 30 %
males and ≥ 20 % in females. Cecal intubation was successful in 997 subjects (98.5 %),
with the proposed standard being ≥ 95 %. Procedures where the cecum was not reached
was due to excessive looping or angulation based on documentation. If colonoscopies
that were aborted due to poor bowel preparation are included in analysis of cecal
intubation rate, the number of colonoscopies analyzed would increase to 1,051 and
would equate to a 94.9 % success rate. Mean withdrawal time was 18.9 minutes (range
5.8 – 66.7), with the proposed standard being a mean of ≥ 6 minutes. There were no
complications of colonic perforation or post-polypectomy bleeding, with the proposed
standard being < 1:1000 perforations for screening colonoscopies and < 1 % incidence
of post-polypectomy bleeding. When analyzing NP performance separately, each individual
NP also met and exceeded all of the proposed quality standards ([Table 4]).
Table 3
Overall nurse practitioner colonoscopy performance.
|
ASGE/ACG Quality Indicator
|
NP Outcomes
(N = 1,012)
|
Proposed ASGE/ACG Standard
|
|
Adenoma detection rate
|
|
|
35.6 %
|
≥ 25 %
|
|
|
39.3 %
|
≥ 30 %
|
|
|
32.1 %
|
≥ 20 %
|
|
Cecal intubation rate
|
98.5 %
|
≥ 95 %
|
|
Withdrawal time
|
18.9 min (5.8 – 66.7)
|
≥ 6 min
|
|
Complications
|
0
|
Perforation < 1:1000
Post-polypectomy bleeding < 1 %
|
ASGE/ACG, American Society for Gastrointestinal Endoscopy/American College of Gastroenterology;
NP, nurse practitioner.
N = number of procedures
Table 4
Individual nurse practitioner colonoscopy performance.
|
ASGE/ACG Quality Indicator
|
NP #1
(N = 533)
|
NP #2
(N = 137)
|
NP #3
(N = 342)
|
Proposed ASGE/ACG Standard
|
|
Adenoma detection rate
|
|
|
38.3 %
|
44.5 %
|
27.8 %
|
≥ 25 %
|
|
|
40.9 %
|
45.3 %
|
34.0 %
|
≥ 30 %
|
|
|
35.4 %
|
44.1 %
|
22.2 %
|
≥ 20 %
|
|
Cecal intubation rate
|
98.9 %
|
95.6 %
|
98.5 %
|
≥ 95 %
|
|
Withdrawal time
|
19.6 min (6.7 – 66.7)
|
32.3 min (7.9 – 57.3)
|
12.6 min (5.8 – 43.0)
|
≥ 6 min
|
|
Complications
|
0
|
0
|
0
|
Perforation < 1:1000
Post-polypectomy bleeding < 1 %
|
ASGE/ACG, American Society for Gastrointestinal Endoscopy/American College of Gastroenterology;
NP, nurse practitioner.
N = number of procedures
Additional data regarding polyp characteristics and polypectomy technique were also
examined ([Table 5]). In total, 1,471 polyps/lesions were detected. Of these, 847 (57.6 %) were benign,
611 (41.5 %) were adenomas, nine9 (0.6 %) were advanced adenomas, and four (0.3 %)
were adenocarcinomas. Benign polyps included hyperplastic polyps, inflammatory polyps,
and non-diagnostic pathology; adenomas included tubular adenomas and sessile serrated
adenomas; and advanced adenomas included adenomas with villous features or high-grade
dysplasia. The majority of polyps were small, with 1,185 (80.6 %) being 0 to 5 mm,
179 (12.2 %) being 6 to 9 mm, 83 (5.6 %) being 10 to 20 mm, and 24 (1.6 %) being > 20 mm.
In total, 1,183 polyps (80.4 %) were removed via cold forceps polypectomy and 276
(18.8 %) were removed via snare polypectomy. Twelve polyps/lesions (0.8 %) were not
removed, with eight polyps being referred to an advanced therapeutic endoscopist for
advanced polypectomy and four polyps not removed because they appeared to be cancerous
and not amenable to endoscopic resection.
Table 5
Polyp characteristics and removal technique.
|
Polyps/Lesions
|
N = 1,471
|
|
Pathology
|
|
|
847 (57.6 %)
|
|
|
611 (41.5 %)
|
|
|
9 (0.6 %)
|
|
|
4 (0.3 %)
|
|
Size
|
|
|
1185 (80.6 %)
|
|
|
179 (12.2 %)
|
|
|
83 (5.6 %)
|
|
|
24 (1.6 %)
|
|
Polypectomy Technique
|
|
|
1183 (80.4 %)
|
|
|
276 (18.8 %)
|
|
|
12 (0.8 %)
|
N = number of polyps/lesions
Discussion
Overall, the NPs met and exceeded all of the proposed quality ASGE/ACG Taskforce quality
standards for colonoscopy, in line with the performance expected by trained physicians.
As a result, this study demonstrates that adequately trained NPs can perform screening
colonoscopy safely and effectively.
Use of screening colonoscopy has continued to rise over the past decade [5]. One major reason for this is increased evidence to support that colonoscopy reduces
CRC incidence and mortality, leading to endorsements for screening colonoscopy by
all major gastroenterology societies [19]. There has also been increased access to colonoscopy, with Medicare mandating coverage
for all beneficiaries in 2001 and private insurers shortly following suit [20]. There has also been a significant rise in the demand for colonoscopy due to the
increasing aging population. By 2030, the US population aged 65 and older is projected
to grow by 55 %, being a large portion of individuals who will require screening [21]. Finally, there has also been an increase in patient awareness efforts, such as
the 80 % by 2018 initiative by the National Colorectal Cancer Roundtable (NCCRT),
as well as celebrity involvement such as Katie Couric’s televised colonoscopy, which
were found to effectively boost screening rates [22].
Increased use of colonoscopy raises the question whether there are adequate resources
to meet the demand for CRC screening. One study on the projections for demand and
capacity for colonoscopy published in 2009 estimated an additional 1,050 gastroenterologists
would be needed by 2020 to meet the demands for colonoscopy if screening rates were
to remain unchanged, increasing to an additional 1,550 gastroenterologists if screening
rates were to increase by 10 % [7]. A similar study published in 2004 estimated that 1,000 additional endoscopists
would be needed to meet the demand for colonoscopy if 70 % of the 2004 population
were to be screened. Advocacy groups have called for even higher screening rates which
would widen this gap. [6]. With the number of trained gastroenterologists having remained steady for the past
30 years and expected to remain unchanged in upcoming years, this increase in the
number of gastroenterologists in practice is unlikely to occur [23]. Therefore, training NPs to perform colonoscopy could be a solution to ensure the
needs for CRC screening are adequately met.
The role of Advanced Practice Registered Nurses (APRNs), including NPs, has expanded
to address the national physician shortage in the United States. Expansion of APRN
scope of practice was specifically recommended in the Patient Protection and Affordable
Care Act of 2010 to increase health care access and affordability [24]. There are several examples of successful expansion of APRN scope of practice including
Certified Registered Nurse Anesthetists (CRNAs) who administer anesthesia services
and Certified Nurse Midwives (CNMs) who specialize in ob/gyn, including the delivery
of babies. A large systematic review published in 2011 that included 20 randomized
controlled trials and 49 observational studies found that APRNs in collaboration with
physicians could achieve outcomes and patient satisfaction that were comparable, and
in some cases superior, to care by a physician alone [25].
There is precedence for NPs performing colonoscopy in the United Kingdom (UK), where
the practice has already been adopted as means to meet the need for increased CRC
screening [26]. By 2005, there were approximately 200 NPs performing diagnostic and therapeutic
endoscopy procedures, including skills of polypectomy and management of gastrointestinal
bleeding [27]. Nurse endoscopists in the UK have demonstrated the ability to perform colonoscopy
safely and effectively, with high patient acceptability and improved patient care
[27]
[28]
[29]
[30]. One UK study analyzing 100 colonoscopies performed by a nurse endoscopist immediately
after training found the nurse endoscopist achieved a cecal intubation rate of 92 %,
with no procedure-related complications and a high degree of patient satisfaction
[27]. The nurse endoscopist also demonstrated competency in polypectomy skills and administration
of conscious sedation. A similar subsequent study was conducted in the UK with comparable
results, with findings that nurse endoscopists had similar efficacy, safety, and patient
satisfaction scores compared to physicians [30]. In 2005, the British Society of Gastroenterology published a position report supporting
the use of nonmedical endoscopists to perform endoscopy procedures once they have
completed a full training program and are able to demonstrate sufficient knowledge
and competence to satisfy established standards [28]. NPs continue to be trained in colonoscopy in the UK, where they have also established
strategic guidelines for training and monitoring of non-physician endoscopists [27].
Studies on nurses trained to perform colonoscopy have also been conducted in the Netherlands
with results similar to the UK studies. One study at a Dutch center comparing the
performance of colonoscopy by two endoscopy nurses, one gastroenterology fellow, and
one experienced gastroenterologist, all completing 150 colonoscopies each, found similar
cecal intubation rates, time to cecum, and AE rates between groups [31]. Patients also reported similar degrees of pain, levels of satisfaction, and willingness
to undergo a future colonoscopy among groups. Additionally, a larger prospective multicenter
study performed in the Netherlands analyzing 10 nurse endoscopists performance of
100 consecutive colonoscopies after training found that all of the nurses met the
international quality standards with an ADR of 26.7 %, average cecal intubation rate
of 94 %, AE rate of 0.2 %, and high patient satisfaction of 95 % [32].
A Canadian study also demonstrated an NP’s ability to perform colonoscopy safely and
effectively. The study evaluated 225 independent colonoscopies performed by an NP
after completing a 2-year training program and found the NP was as effective as gastroenterologists
in performing colonoscopy based on quality measures, with a cecal intubation rate
of 92 %, polyp detection rate of 39 %, and one minor adverse post-polypectomy bleeding
event [33].
In the United States, a single-center, randomized controlled trial demonstrated that
an NP was able to perform colonoscopy as safely and effectively as her physician colleague.
The NP had an ADR of 42 %, cecal intubation rate of 100 %, mean withdrawal time of
8.5 minutes, and no procedure-related AEs [34]. Another single-center prospective study in the United States evaluating the performance
of an NP’s first 300 consecutive screening colonoscopies after training found the
NP satisfied all of the proposed quality indicators for colonoscopy, with an ADR of
35.0 %, cecal intubation rate of 99.0 %, mean withdrawal time of 19.3 minutes, and
no procedure-related AEs [35].
The results of our study are consistent with previous studies, supporting properly
trained NPs as a possible solution to filling the gap between the supply and demand
for screening colonoscopies in the United States. CRC screening rates vary widely
across the country, with lower screening rates and higher CRC-related deaths in rural
and underserved areas [36]. NPs may be especially useful in these under-resourced settings where conventional
access to a gastroenterologist is limited.
Use of NPs performing colonoscopy may also help to reduce overall health care costs.
This would result from lower professional fees, reimbursement rates, and salaries
compared to physicians. Using the annual salary differential for the same endoscopy
case load, we calculated a per-case savings of $102. Total savings when multiplied
by the number of procedures exceeded $100,000 annually. Indirect cost savings have
not been calculated, however, having NPs perform colonoscopy can allow gastroenterologists
additional time to perform more complex procedures that generate higher revenue. In
addition, downstream revenue can be generated for other hospital departments, such
as pathology, if more colonoscopies are performed due to the need for additional pathology
review. A more robust direct and indirect cost-analysis would be beneficial; however,
that was beyond the scope of this paper, which primarily sought to evaluate the quality
of fellowship-trained NP colonoscopy when compared to physician colonoscopy.
Limitations of this study are that it focused on a small number of fellowship-trained
NPs in a single center. The total number of nurse-endoscopists in the United States
at present is unknown and those data are not readily available. In addition, no known
complications are reported, however, we cannot rule out complications that may have
occurred post-procedure discharge that were unreported by the patients. Of note, patients
who underwent biopsy or polypectomy were notified of their pathology results post-procedure
and none of the individuals contacted reported complications. A final limitation of
this study is that it did not address how many NP colonoscopies required physician
support for cecal intubation and/or polypectomy. Future studies providing subjective
patient feedback about their experience with a nurse endoscopist may also be of value.
A potential barrier to expanding colonoscopy training to NPs may be a lack of physician
acceptance. However, NPs performing colonoscopy may benefit physicians by allowing
them to perform more therapeutic and complex cases as discussed above. They may also
help to decrease patient wait times and improve overall patient satisfaction which
could benefit their overall practice. A more likely scenario is adoption in underserved
areas or in large health systems and accountable care organizations where value-based
care plays a larger role, however, appropriate training programs would need to be
in place in these institutions to support successful NP endoscopic training.
For non-physician endoscopists to effectively scale to meet the supply gaps, colonoscopy
training programs with appropriate training guidelines would need to be established.
In the setting of our study, three NPs were trained using the same curriculum as first-year
gastroenterology fellows, suggesting that some infrastructure for training programs
may already be in place. In addition, continual performance evaluation of non-physician
endoscopists and guidelines for physician assistance would be needed for long-term
viability of this practice
Conclusion
In conclusion, NPs present a viable option for delivering high-quality screening colonoscopy.
This, in turn, may help to reduce the overall health burden of CRC in the United States.
