Should We Train Non-Physicians To Do Endoscopy? No

This is half of a two-part debate — read the opposing argument. The opinions expressed here are that of the author and do not represent the position of AGA.

The use of non-physicians to provide clinical services is widespread. Mid-level providers can act both independently without supervision, i.e. nurse practitioner, or act as surgical assistants, i.e. physicians assistants.

First, which procedures are we considering for non-physicians: sigmoidoscopy, upper endoscopy, colonoscopy, ERCP or endoscopic ultrasound? Second, within those procedures, are we considering training in diagnostic or therapeutic procedures? One could debate each of these points but because of space limitations I would like to restrict this discussion to screening colonoscopy.

In the United States, we should not train non-physicians to perform colonoscopy.

The reasons for this position are the following:

1. We don’t need more endoscopists to perform screening colonoscopy as part of our colorectal cancer (CRC) prevention program.

The Division of Cancer Prevention and Control of the Centers for Disease Control recently published their detailed analysis of the need, volume and capacity for CRC screening using the Microsimulation Screening Analysis-Colon (MISCAN-Colon) Center.1 They concluded that the estimated colonoscopy capacity is sufficient to screen 80 percent of the population using a mix of colonoscopy, fecal immunochemical test (FIT) or both. This may not be true for other countries or capitated systems in the United States where the cost of training is high or the availability of manpower is limited. Even in Europe and Australia where resources are more limited, mid-level providers are not widely used to replace trained physicians.

2. We will not save money by using non-physician endoscopists.

CRC screening by colonoscopy is a significant financial burden on the health care system. However, it is not clear that using less expensive providers will reduce the cost significantly. Costs vary widely depending on geographic location, hospital versus ambulatory surgery settings, availability of services and insurance contracts. Thus, major variances in cost are not solely based on the relative cost of physicians versus non-physicians performing the procedure. In fact, the physician fee may account for as little as 20 percent of the total cost. Medicare data for 2015 published by the U.S. Department of Health
and Human Services Digestive Health Network Inc. calculated that $416.5 million was spent for professional fees against a total program cost for outpatient colonoscopy of $1.88 billion. Costs including facility, pathology, anesthesia, evaluation and management vary and would reduce the financial impact of a reduction in reimbursement associated with the use of non-physician endoscopists.2

An English cost-benefit analysis compared nurses to consultants with regard to upper EGD and sigmoidoscopy did not demonstrate an improvement in cost-effectiveness.3 In contrast, the Dutch identified a savings in personnel costs of approximately 15 percent per procedure. The Dutch model required a senior endoscopist to supervise three nurse endoscopists per session. Yes, the productivity of a unit could go from approximately 10 to 30 procedures per day. However, how many trained endoscopists would want to assume responsibility for that work? How would that work? What would be the ratio between the non-physician endoscopist and supervising physician? What would the cost and reimbursement model be? What are the liability issues? Would this be acceptable to a patient population with the ability to choose services? Would this be an acceptable career track for a gastroenterologist? There are too many unknowns for what is likely to be a marginal cost benefit.

Establishing the infrastructure and processes will be a very complex undertaking and the payoff is not clear.

3. The elaborate training, certification and quality assurance process will be too time-consuming, expensive and complicated by unforeseen consequences.

The training of non-physician personnel must include several complex processes. First, multiple professional societies must establish specific criteria for training, practice and certification. Second, institutions must be willing to establish training programs that are then certified by a governance body with the appropriate jurisdiction. Third, state boards must approve an expanded scope of practice for non-physician practitioners. Fourth, federal and state insurers must develop mechanisms for coverage of services. This is not a trivial undertaking and is likely to take many years, significant resources and multiple studies to demonstrate the clinical equivalence of non-physicians and trained physician endoscopists.

It cannot be assumed that guidelines for physicians would simply be extended to non-physicians. It is likely that to develop a high level of skill requires many more procedures than the 250 to 300 that have been proposed. In addition, it is not clear that time to cecum and adenoma detection rates would be adequate measures of skill and quality in this setting. What would happen if a supervising endoscopist is not satisfied with the performance of the non-physician endoscopist? How would this be managed? Medical directors of current endoscopy units already have difficulty monitoring and assuring the performance of physician colonoscopists. Directors recognize the significant variation among physicians who perform colonoscopy. How will quality problems be managed when the supervising physician assumes responsibility for the behavior of a non-physician? Again, establishing the infrastructure and processes will be a very complex undertaking and the payoff is not clear.

4. The future of colonoscopy as the mainstay of CRC screening is too uncertain to commit the resources necessary to train non-physician endoscopists.

Currently approved or under investigation are an array of non-invasive stool and blood tests designed to screen for CRC. The application of these technologies may significantly reduce the demand for colonoscopy as a primary screening option. If these studies demonstrate a non-invasive test performance profile that is epidemiologically and clinically acceptable as a method to reduce population mortality, then the number of colonoscopies could decline significantly. Does the uncertainty justify the investment in program development? What will happen to non-physician endoscopists then? What will happen to physician endoscopists and the economic model created to support CRC screening by colonoscopy?

In summary, I do not believe that non-physicians should provide endoscopic services. The risks are too great and the payoff too uncertain to invest the time, energy, intellectual and financial resources at this time. The only exception would be those care settings where manpower resources are so limited that small, targeted programs under local governance may be justified.

Luke John Day, MD, provides a different view on training non-physicians to do endoscopy.

Dr. Korman owns shares in Capital Digestive Care and Metropolitan Gastroenterology Group. Metropolitan Gastroenterology Group has an interest in our Ambulatory Surgery Center and Capital Anesthesia Partners.

1. Joseph, D.A., Meester, R.G., Zauber, A.G. et al, Colorectal cancer screening: Estimated future colonoscopy need and current volume and capacity. Cancer. 2016; 15;122(16):2479-86.
2. U.S. Department of Health & Human Services. Digestive Health Network Additional Information. Available at: Accessed Dec. 2, 2017.
3. Richardson, G., Bloor, K., Williams, J. et al, Cost effectiveness of nurse delivered endoscopy: findings from randomised multi-institution nurse endoscopy trial (MINuET). BMJ. 2009; 338:b270.
4. Massl, R., van Putten, P.G., Steyerberg, E.W. et al, Comparing quality, safety, and costs of colonoscopies performed by nurse vs physician trainees. Clin
Gastroenterol Hepatol.
2014; (3):470-7.

5. Hamzehzadeh, L., Yousefi, M., Ghaffari, S.H. Colorectal Cancer Screening: A Comprehensive Review to Recent Non-Invasive Methods. Int J Hematol Oncol Stem Cell Res. 2017; 11(3):250-261.


  • Totally agree with Dr. Korman’s point of view. Theoretically it’s logical to look for additional Endoscopists even among mid level providers but it will not work from practical stand point. Physician extenders will require more extensive training that physicians to perform Endoscopy so we should look into increasing fellowship positions nationwide and encourage internist to pursue subspecislties instead of becoming general practitioners or hospitalists.

  • I do not think we should train no physician to do endoscopy. There is no need for that; if anybody think that there is a need, we should be talking about increasing GI trainning programs. In the 80’s, a lot of programs were closed for not good reason and today we see a shortage of slots for future gastroenterologist.

  • You are right and I agree, since I learned and wasn’t useful… 20 yrs ago, except giving a wider perspective about interpretation of symptoms

  • I have a couple of comments as a PA. First, I in no way feel like I am equal to an MD nor do I want to be, however I do feel like everyone in the medical field should practice at the top of their licence. Second, not to be overly sensitive but to compare a PA to a surgical assistant is a huge insult. (“The use of non-physicians to provide clinical services is widespread. Mid-level providers can act both independently without supervision, i.e. nurse practitioner, or act as surgical assistants, i.e. physicians assistants.”) This is 2018 and to not understand that a PA is a an APP and plays a huge role in medicine in this country is crazy. I think it is important to note that NP’s and PA’s are both APP’s and both can perform the same duties. I also think that this article is a huge over generalization. What about in medically undeserved communities like where I work? We have a part time GI MD who is amazing but overwhelmed and we are booking cases eight months out with elective cases not to mention bleeds and other urgent cases. We are currently trying to find way that I can help decompress this situation. Again I want to stress I am not trying to replace her, nor do I think I can replace her training with only 300 cases. It would be great if I could somehow decompress the easier cases so she can do more of the harder more urgent cases. I do want to thank you for the article and bringing light to some great points, I hope that this conversation will continue.

Join the discussion

Your email address will not be published. Required fields are marked *