Vascular surgery was consulted for placement of a dialysis catheter in a patient on the medical floor. The surgical resident examined the patient, an elderly woman with dementia and renal insufficiency receiving IV hydration. The resident called the family to obtain informed consent for the procedure. The daughter was very surprised by the call, stating that no one had discussed initiating dialysis; she insisted that it was a mistake. The surgical resident attempted to convince the daughter that the patient's life was in danger, but the daughter refused to give consent. The next morning, the surgeon returned to the bedside, where the family again refused to provide consent. The medical attending caring for the patient was ultimately called, and he verified that the patient had prerenal azotemia related to dehydration. He was unaware of any request for dialysis catheter placement in this patient. Because they were worried about her safety, the family signed the patient out of the hospital and returned her to the nursing home. Upon further review, the catheter placement request was for another patient on the same floor with the same unusual last name, who had chronic renal failure progressing towards dialysis.

by Darrell Campbell, Jr., MD

The case presented represents a near miss. In the absence of an involved family, a surgical procedure might have been performed on the wrong patient. Two previous and thorough discussions on this Web site (1,2) have focused on the problems of patient misidentification, the most obvious issue here. As Chief of Clinical Affairs at my institution and also a general surgeon, I want to share my perspective on the problem of wrong patients, wrong side, or wrong site surgery, and the daunting task of initiating a successful patient safety policy in a busy surgical environment.

Wrong side, wrong site, or wrong patient surgery is no small problem. The most complete data shows a rate of 1 in 15,500 surgical procedures(3), assuming all events were reported. A review of such cases reported to JCAHO showed that in 36% of cases the wrong patient was operated on, in 44% right and left mix-ups occurred, 14% involved the incorrect implant, and 7% involved the wrong site (not right versus left but, for example, the wrong level of the spine).(4) The sites of errors, in order of decreasing frequency, were eye, groin or genitals, chest, and leg. It has been suggested that a properly performed "time out" would have prevented 85% of the errors, proper patient identification and discussion with the patient would have prevented 75%, proper marking of the site would have prevented 65%, and a properly completed consent form would have prevented 45% of errors (J.P. Bagian, MD, of the National Center for Patient Safety, written communication, March 2004).

Tackling this problem at an institutional level by developing and implementing a wrong site policy is far from easy. Early attempts suffered from inconsistency. At our institution, Orthopedics marked an "X" on the non-operative site, while other groups marked the operative site, which was worse than no site marking at all. A colleague of mine at a different hospital produced "NO" stickers to place on the non-operative site, but soon realized that "NO" looked like "ON" from the other end of the table. Many groups marked the right ("R") or left ("L") shoulder to indicate laterality regardless of the type of case, but this mark usually couldn't be seen after the patient was prepped.

In 2003, JCAHO published "The Universal Protocol for Preventing Wrong Site, Wrong Procedure, Wrong Person Surgery."(5) Endorsed by 47 professional societies, including the American College of Surgeons, this document directly addresses JCAHO's 2004 Patient Safety Goal #4: "Eliminate wrong-site, wrong-patient, wrong-procedure surgery."(6)

The policy is good, but the ramp up period is difficult. Long-standing policies and procedures have become ossified, and surgeons resist change, particularly when the changes aim to prevent relatively low-frequency events. Yet the need for an unyielding policy is obvious when one looks at what really happens in a fast-paced, hectic environment. For example, both our medical center and JCAHO have a strict policy to use an "active" ("What is your name?") manner to identify patients. One day not too long ago, an incognito observer noted that in our preanesthesia care unit (PACU) a minority of patients was identified in the "active" manner. Instead, most were identified in the "passive" manner ("Are you Mr. Jones?"), an approach that probably contributed to the near miss in this case.

Many hospitals, particularly large ones, are having difficulty implementing the universal policy. At our hospital, the major hindrance to implementation was deciding on who should mark the site preoperatively. The universal policy seems difficult to apply to a menu of highly complex procedures, such as spinal surgery requiring intraoperative x-ray confirmation of levels. We predicted much confusion in the PACU, many unnecessary pages, and a loss of efficiency. There was also consternation among attending surgeons. Surgeons juggling clinics, rounds, lectures, and paperwork felt that marking the site personally in the PACU was an inefficient use of their time, since a long gap often transpires between when the patient is evaluated in the PACU and when the patient actually rolls into the OR.

The JCAHO Universal Policy allows for some flexibility in this matter, in that it states that the attending surgeon "should" (not "must") mark the site preoperatively. We developed a new algorithm that we believe is very safe and consistent with JCAHO expectations, in which the PACU nurse is allowed to mark the site preoperatively depending on the circumstances. So far, we have not seen a decrease in efficiency nor any worrisome events.

The algorithm is based on a newly designed operative permit (Figure 1). The permit includes diagrams of a patient on which the operative site can be marked by the team at the time of initial evaluation in the office. The attending surgeon must initial a box indicating agreement with the site marking at that time. The patient signs a statement that says, "I understand the approximate location of my surgical incision as indicated on the illustration." When this procedure has been followed during the preoperative office visit, the PACU nurses need only reconcile the permit with the patient (and family) and the OR schedule before marking the site themselves and sending the patient to the OR. Any discrepancy in the above requires the attending to come to the PACU to resolve the issue. In other cases, indicated by checking the respective boxes on the form (Figure 2), the attending must personally mark the operative site (lymph node biopsy, breast biopsy), the attending localizes the site intraoperatively (eg, spinal surgery), or preoperative site localization is not indicated (eg, midline sternotomy, C-section).

This procedure has worked very well, but the lost operative permit has now become the bane of the busy OR, especially when the permit cannot be found on the morning of surgery. In the past, the most junior member of the OR team was stat paged to fill out a new one, and usually this team member knew next to nothing about the case. This was a major patient safety vulnerability. We are transitioning to an electronic medical record (after which time we will be able to scan the permit at the time of the clinic visit and enter it into the electronic record), but at this point we still rely on the chart. In any case, because of the vulnerability of the old system, the days of the stat page to the intern are gone. If the permit can't be found, the attending must redo the permit in the PACU.

Back to this case: if the hapless resident had actually sent this patient to our OR, I believe that several checkpoints in our new system would have worked to prevent a wrong patient, wrong procedure disaster. We now require an active patient identification, made by three separate individuals (PACU nurse, circulating nurse, and anesthesiologist). The PACU nurse also is required to confirm the procedure with the family; this step is particularly valuable in cases like this one in which dementia diminishes the value of active identification. Finally, the attending surgeon's presence would be required in the PACU to fill out the permit, since it had not been completed during a preoperative clinic visit. Hopefully, the result from all this effort would be a happy one—a patient sent back to the floor with an apology, but no operation.

Take-Home Points

• Of wrong side, wrong site, and wrong patient mistakes, 35% involve operating on the wrong patient.
• A properly performed "time out" prior to surgery can prevent the vast majority of wrong patient errors (85% in some estimates), and it is thus a critical piece of any institutional safety policy.
• A national policy/guideline to prevent surgical error can be implemented at the institutional level. Key recommendations need to be retained, but logistical details should be modified at each institution to create a system that addresses wrong site, wrong procedure, and wrong patient surgery without a loss in efficiency.

Darrell Campbell, Jr., MD
Henry King Ransom Professor of Surgery
Chief of Clinical Affairs
University of Michigan Health Systems

1. Shojania KG. Patient mix-up. AHRQ WebM&M [serial online]. February 2003. Available at: [ go to related commentary ]. Accessed May 14, 2004.

2. Kaplan H. Transfusion "Slip". AHRQ WebM&M [serial online]. February 2004. Available at: [ go to related commentary ]. Accessed May 14, 2004.

3. NYPORTS. The New York patient occurrence reporting and tracking system annual report 2000/2001. The New York State Department of Health website. Available at: [ go to related site ]. Accessed May 14, 2004.

4. Sentinel event alert. Joint Commission on Accreditation of Healthcare Organizations Web site. December 5, 2001. Available at: [ go to related site ]. Accessed May 14, 2004.

5. Universal protocol for preventing wrong site, wrong procedure, wrong person surgery. Joint Commission on Accreditation of Healthcare Organizations Web site. Available at: [ go to related site ]. Accessed May 14, 2004.

6. 2004 national patient safety goals. Joint Commission on Accreditation of Healthcare Organizations Web site. Available at: [ go to related site ]. Accessed May 14, 2004.

Figure 1. Sample Operative Permit

Click on the thumbnail for a full view of Figure 1.

Figure 2. Sample Form for Preoperative Checklist

Click on the thumbnail for a full view of Figure 2.

To view the figures, you must have Abobe Acrobat Reader. Get Adobe Acrobat Reader