Better ways to document the reasons for doing a procedure (the indication)

Endoscopy reports traditionally include the indication for the procedure. An important practical issue is how to express this succinctly and meaningfully, especially in electronic reporting systems, which demand structured data sets. For the reasons argued above, simply ticking a box of ASGE-approved indications is insufficient. It is preferable to define and document all of the individual elements that go into making a procedural recommendation in a matrix:

• •
  

  Overall goal

  
  
• •
  

  Planned procedures (strategy)

  
  
• •
  

  Symptoms

  
  
• •
  

  Known and suspected diseases

  
  
• •
  

  Recent laboratory tests and images

  
  
• •
  

  Prior surgeries

  
  
• •
  

  Prior interventions (eg, prior stents)

  
  
• •
  

  Comorbidities and risk factors.

This matrix would be comprehensive, allow some detailed research analysis, and might help in medicolegal conflicts, whereby the need for the (complicated) procedure is often at issue. The data entered could also easily be mapped to any other proposed list, because it will contain all of the elements. However, these lists would be long, and most endoscopists will argue that the details should already be available in their preprocedure clinical report (assuming that there is one). A simpler approach is to note the general clinical context, as is used in the ERCP Quality Network ( Box 2 ).

A similar approach is taken in the latest iteration of the Minimal Standard Terminology (MST 3.0), where indications are replaced by “reasons for” procedures. The endoscopist has simply to pick a suspected (or known) diagnosis, and add whether the goal is to make or refine a diagnosis, perform therapy, take specimens, or do surveillance or screening.

Report cards and benchmarking performance

Endoscopists cannot be expected to populate their report cards with all of the data elements that have been listed in various well-meaning publications. Items should be selected based on ease of data collection, and by assumed relative importance. Some items are easily recorded, and already appear in most procedure reports (eg, anatomic extent, duration, diagnosis, treatments, immediate adverse events). Other items are more subjective (eg, lesion interpretation), or more difficult to record (eg, delayed events, endoscopist-specific patient satisfaction). Some items would appear to be more important markers of quality than others. For ERCP, selective cannulation and stone extraction rates are obvious key parameters.

Once several practitioners agree to share their data, they can compare their practices and levels of performance with those of their peers and “competitors,” which requires organization as well as motivation.

With the support of Olympus America, the author’s group set up a pilot project, the ERCP Quality Network, to test the practicality and acceptability of collecting and comparing data on the practice and quality of ERCP procedures by individual endoscopists. Baseline information included the experience and practice environment of the endoscopists. Data on each procedure are loaded onto a secure Web site. The data points include the clinical context (indications), complexity grade, American Society of Anesthesiologists (ASA) grade, sedation/anesthesia, admission policy, scope and fluoroscopy times, and success rates for individual technical procedures such as deep biliary cannulation, sphincterotomy, stenting, and adverse events. There are no patient identifiers. The data are analyzed automatically, and results posted immediately on the Web site. Contributors can view a summary of their own performance (report card), and compare it with that of all other contributors to the system (benchmarking), not identified by name. To date, more than 150 ERCPists from several countries have entered data on more than 25,000 cases. The variation in individual biliary cannulation rates is shown in Fig. 1 .

Average biliary cannulation rates by individual endoscopists in the ERCP Quality Network.

The Quality Network project shows that data can be collected, shared, and compared by a small number of enthusiastic volunteers, with some commercial support. What might motivate the main body of practitioners? Ultimately, informed consumers will drive this agenda. Practitioners with poor outcomes, or those who choose not to provide data, will be disadvantaged. Keeping a report card will provide a competitive advantage, as intelligent patients will increasingly wish to have greater control over their treatments. It should also provide some medicolegal protection, and will eventually be a crucial tool in the evolving use of “pay for performance” in the United States. The increasing use of electronic endoscopy reporting systems will make this process easy, even automatic.

Is ongoing volume important?

It is logical to assume that anyone offering ERCP should be doing it on a regular basis, to maintain skills and, hopefully, to continue to improve. Several studies have shown that more is better, but none have data sufficient to define a minimum number (mainly because the lower-volume practitioners do not publish their results). The number probably varies according to the extent of prior practice and the volume of the institution. The British Society of Gastroenterology (BSG) has recommended a minimum of 75 per year. A recent survey of ERCP practice among more than 1000 ASGE members showed that 40% were doing less than 50 per year, data that stimulated an editorial entitled “Are low-volume ERCPists a problem in USA?”

Teamwork and environment

As ERCP has become almost completely therapeutic its needs have become more complex, and increasingly resemble those of traditional surgery, that is, purpose-designed rooms with high-quality fluoroscopy, complex equipment, sterility (where possible), anesthesia, and specially trained staff. Doing complex procedures on uncomfortable patients in unfamiliar territory with untrained staff is a recipe for failure and potential disaster, and this is more likely to be the case in centers with low volumes of ERCP. Recognizing this issue, the BSG has suggested that ERCP should only be performed in centers with more than 200 cases per year. In such light it is striking that a recent survey suggests that about half of all hospitals offering ERCP in the United States do less than 50 per year.

Defining adverse events

There is a substantial literature describing individual adverse events, and many large collected series. A key issue is the need for a standardized nomenclature and agreed definitions for adverse events. For example, what is meant by “hypoxia,” or “bleeding,” or “infection”? At what level do they become significant enough to be “counted”? Another problem is how to classify and report delayed events, which may or may not be attributable to the procedure. This lack of standardization has many consequences. It hampers the comparison of data from different research and quality-improvement studies. It makes individual studies suspect, because practitioners may not be consistent with their own perceptions and definitions. Furthermore, it makes it impossible to compare endoscopic outcomes with those from other disciplines such as surgery. The need for standardized nomenclature has come into closer focus recently with the increasing use of electronic report writers, which demand a lexicon.

Some attempts were made years ago to address this problem, and some 20-year-old definitions for adverse events in ERCP have been widely used, but there has not been consensus across the whole range of endoscopic procedures. The ASGE Quality Task Force convened a workshop in September 2008 to explore the current situation and to make recommendations. Invitations were issued to representatives from the ASGE, the American College of Gastroenterology, and the Society of American Gastrointestinal and Endoscopic Surgeons, and also to those familiar with related lexicons, such as NSQIP (Surgery), HI-IQ (Radiology), NIH, CCTAE (National Cancer Institute), Snomed Clinical Terminology, and the MST. The working party made several recommendations.

• 
  

  Recommendations of the ASGE Working Party on Adverse Events

  
  
• 1.
  

  Definitions and severity criteria should be generic across the different endoscopic procedures, although certain events will occur only after some of them (eg, cholangitis after ERCP).

  
  
• 2.
  

  An adverse event is one that prevents completion of the planned procedure, and/or results in admission to hospital, or prolongation of existing hospital stay, or (for management of the event) another procedure (one needing sedation/anesthesia), or consultation with another specialty.

  
  
• 3.
  

  Timing. Events should be recorded as happening preprocedure, intraprocedure (from entering the preparation area through leaving the endoscopy room), postprocedure (up to 14 days), and late (any time after 14 days). For delayed events, the number of days after the procedure should be documented.

  
  
• 4.
  

  Events should be documented as: definitely attributable, probably, possibly, and not attributable.

  
  
• 5.
  

  Severity is graded by the degree of consequent disturbance to the patient and any changes in the plan of care. Adverse events should be followed until their conclusion to assess severity, and which should include the outcome of any new procedures required to treat them. Thus death after surgery for debridement of post-ERCP pancreatitis would be attributable to the ERCP. However, this should not apply to adverse events occurring after second procedures done solely because of the technical failure of the initial endoscopy (ie, without any adverse event prompting failure or abortion of the procedure).

  
  
• 6.
  

  Event tracking. Attempts should be made to contact patients at about 14 days after procedures to determine whether any adverse events have occurred, and whether they are attributable. Report generators should allow these data to be included as an addendum to the endoscopy report, including the statement that there were definitely no adverse events, when this had been confirmed by patient contact.

  
  
• 7.
  

  Reporting statistics. When reporting “complication rates,” only definite and probably attributable events occurring within 14 days should be included. Rare adverse events that present after 14 days and are clearly attributable can be recorded as a separate category. Examples include a proven nosocomial infection, or stent migration causing a new clinical problem, not just failure of the original treatment goal.

  
  
• 8.
  

  Incidents are unplanned events that do not interfere with completion of the planned procedure or change the plan of care, that is, do not fulfill the stated criteria for adverse events. Examples include bleeding that stops spontaneously or with endoscopic therapy, and transient hypoxia that resolves with or without reversal agents, supplemental oxygen, or bagging. These incidents should be recorded for the purposes of quality improvement, and perhaps to see which incidents may predict later adverse events. Incidents may occur during the procedure or in the immediate recovery period (ie, while still under supervision).

Risk factors for adverse events

It is self-evident that adverse events are less likely to occur when both the endoscopist and the patient are well prepared for the specific procedure, and if performed with the support of a competent team in an appropriate environment. Issues relating to endoscopist and unit competence have been addressed earlier.

Understanding the risks inherent in specific contexts, and how to defuse them, was addressed recently by another working party of the ASGE. The goal was to examine the evidence for specific risks and to suggest a list of factors that might be included in endoscopy reporting/database systems. The conclusions of a very comprehensive review by Romagnuolo and colleagues were published in 2 parts, relating respectively to the risks for cardiopulmonary events and the remainder, which relate mainly to individual procedure types.

The factors increasing the risk of cardiopulmonary events are obvious, and include advanced age, established cardiac and pulmonary disability, other comorbidities, and obesity. These considerations and a few others are nicely wrapped up in the ASA score, which is a standard entry in most reporting systems. Risk is also affected by the setting for the procedure (eg, intensive care unit), its duration, and who administers the sedation/anesthesia.

There are many generic risk factors that apply to all types of endoscopy, including the type and complexity of any therapeutic activity, active infection, immunosuppression, and spontaneous or therapeutic coagulation deficits. ERCP carries other very specific risks, such as pancreatitis, the predictors for which have been widely researched and documented.

Managing adverse events

There are several rather obvious but important management strategies when things “go wrong.” The author believes that it is important to:

• 1.
  

  Recognize the problem quickly, and take appropriate diagnostic and therapeutic actions. Consult other specialists quickly as needed, especially surgeons when there is possibility of perforation (preferably one who knows that not all perforations need surgery).

  
  
• 2.
  

  Behave professionally, just as you would if things had gone well. Visit regularly, but not excessively. Explain carefully to the patient and family what has happened, referring to the consent process: “The x-ray shows a small perforation. You remember that we discussed that possibility beforehand. I’m sorry that it happened to you. This what we plan to do.”

  
  
• 3.
  

  Show that you care. Saying “sorry” is simply to show sympathy, not to suggest that you did something wrong. Do not speculate that “I must have pushed too hard or cut too far.” Keep in touch even if a patient is transferred to another service or hospital. Failure to do so may look like abandonment, and can make people angry and out for revenge. One way to show that you care is to give the patient your cell-phone number.

  
  
• 4.
  

  Ask whether there are other family members in the background (eg, nurses, paralegals) that would appreciate a phone call.

Physical harm

Radiation is an obvious hazard, although no definite damage has been described. It is wise to wear special spectacles, thyroid collars, and wrap-around aprons, and to monitor exposures. Endoscopists performing large numbers of complex ERCPs, and their assistants, should seek professional help from radiation safety officers to further minimize their exposure with customized barriers, such as lead-lined curtains and glass dividers on wheels. The chance of being infected by patients should be eliminated by standard shielding methods. Several surveys have shown significant musculoskeletal issues in endoscopists. Neck problems are more common in older endoscopists who practiced before videoendoscopy, caused by having to crouch over the eyepiece. Some ERCP practitioners have suffered from “cannulator’s thumb” from working the elevator lever.

Medicolegal hazard

ERCP is the most dangerous procedure that endoscopists perform on a regular basis, and some bad outcomes are inevitable. Whether any of these lead to lawsuits is determined by how well endoscopists follow the advice outlined earlier. By far the biggest issue is communication, or lack of it, at all stages of the process. Informed consent is not a piece of paper; it means that the patient fully understands the potential benefits, risks, limitations, and alternatives, and has developed sufficient trust in the people involved to accept the entire package on offer, including the chance of a poor outcome. This process can be assisted by brochures, videos, and interactive Web sites, and by nursing assistants, but must involve an unhurried face-to-face meeting with the endoscopist.

Endoscopists who practice within the accepted standard of care and who communicate well before and after procedures may still get sued, and lose some sleep, but not the suit.