Advanced Search

  • Topic
  • Specialty
  • Content Type


Also Related

< Back To Patient Safety
0 dislikes

< Hide

Comments For

Outpatient Diagnosis Insights


< Shrink

Add Your Voice

All comments are posted anonymously. Your comment will be attributed to: "Anonymous user."

post comment


Are you sure you want to delete this comment?

Outpatient Diagnosis Insights

By Tom A. Augello, CRICO

Related to: Ambulatory, Diagnosis, Emergency Medicine, Primary Care, Obstetrics, Other Specialties

Researchers identify multiple causes of diagnostic error and targets for intervention.


  • Tejal Gandhi, MD, MPH; Brigham and Women’s Hospital; Boston, MA
  • Marjie Harbrecht, MD; Colorado Clinical Guideline Cooperative; Golden, CO


Medical errors related to diagnosis continue to bedevil patients, doctors, and malpractice insurers. At the national level, lawsuits for alleged negligent misdiagnosis represent the top category of malpractice claims. This is reflected in the Harvard system, where diagnostic errors account for a quarter of all claims naming a physician defendant, and 40 percent of incurred losses. Between 1995 and 2005, almost a third of all the physicians named in a malpractice case faced diagnosis-related allegations. Eighty-two percent of the diagnosis-related claims are in the ambulatory setting.

What can these liability cases teach the medical community about the cause of diagnostic mistakes and the best way to intervene?

In a recent study of the ambulatory setting in the Annals of Internal medicine, Dr. Tejal Gandhi and colleagues reviewed more than 300 closed malpractice claims from five insurance companies across the United States. Nearly 60 percent of the malpractice claims were judged to involve a diagnostic error that harmed patients. The same percentage of those claims resulted in serious harm. Two-thirds of the outpatient diagnostic errors in the study involved missed cancer, led by breast and colorectal cancer.

Dr. Gandhi, an Assistant Professor of Medicine at Harvard Medical School, and an internist and Director of Patient Safety for Brigham and Women’s Hospital in Boston, says the patient safety themes emerged quickly.

“When thinking about where things broke down, why these errors occurred, we found that more than half of the errors involved a failure to order a test and almost half involved a failure to create a proper follow up plan. But in looking at again the systems and the reasons for these errors probably one of the most interesting and new findings was that these errors were found to be very complicated.”

Most of the cases involved multiple providers and several breakdowns in the diagnostic process. Failing to order a test or create a proper follow-up plan weren’t the only mistakes.

“…failing to interpret a test correctly or failing to do a proper history and physical or failing to make a referral… and we found on average these cases involved not one, not two, but three of those kinds of failures in each case.  So, again, it is not a single point failure, but really multiple breakdowns in the course of events that led to the final outcome.”

Not only did each diagnostic error involve multiple breakdowns, each breakdown can involve multiple contributing factors. The leading contributing factors in the study include cognitive issues, such as judgment, memory, and knowledge. Other leading factors involve system issues, such as supervision, workload, or interruptions. Issues around communication were also prevalent, including handoffs, failure to establish clear lines of responsibility, and conflict.

Dr. Gandhi says this complexity suggests that efforts to reduce diagnostic errors will require a multi-level effort that goes after process breakdowns and contributing factors.

“We saw that the same contributing factors occurred across many of these process breakdowns and…things like judgment and vigilance and those kinds of things were common throughout.  However, I think that when you try to tackle these, tackling overall judgment and vigilance, I think, is actually difficult.  I think you would have to tackle it in the context of a specific breakdown area.  So, I think knowing that failing to order the test is really the key breakdown and then thinking about vigilance, memory, judgment, etc. as it relates to ordering tests I think that is the right approach as opposed to just overall, let’s think about vigilance and memory and judgment.”

Dr. Gandhi’s own hospital system has leveraged computerized order entry and electronic patient records by using reminders and decision support tools to reduce diagnostic errors. It can help with tracking test results, and it gets at the kind of cognitive issues identified in Gandhi’s malpractice research. Alerts, algorithms and clinical guidelines, for example, are imbedded in the electronic patient record and physician order entry system. These kinds of resources are aimed at reducing errors in judgment, memory lapses and lack of knowledge.

“I think the main challenge really is that designing these kinds of interventions in the ambulatory setting—which is so diverse; not everybody is on computerized systems, for example—is complicated and so we’ll have to target these certain diagnoses that are the most common, such as breast and colon, but then we realize that we might be missing things like the missed MI or the fracture diagnoses that also came up.  So, I think it is challenging in that this is such a wide range of diagnoses and a wide range of practice settings that finding a single silver bullet so to speak is going to be very challenging.”

In Colorado, Dr. Marjie Harbrecht has long been involved in efforts to reduce diagnostic errors, first as an executive with the largest physician insurer in that state, and now as a  promoter of decision support tools at the non-profit Colorado Clinical Guidelines Collaborative. Dr. Harbrecht says electronic tools have obvious advantages, but paper can also be effective. She says her experience has shown that some ambulatory practices have a lot to overcome just to set up one solution to one type of breakdown in the diagnostic process, such as tracking systems for test results.

“The problem is that it is a bit cumbersome when you are ordering thousands of labs in a month. Or you’re asking patients to do things, and there are some physicians that believe it is not their responsibility to be tracking these things, that it is really the patient’s responsibility.  So, you have that aspect to it and in addition, unfortunately, when a patient does fall through the cracks the courts often decide—especially if it is a serious problem—that that it is the physician’s responsibility.”

Courts and juries hearing testimony in medical malpractice suits often assume that if a provider orders a diagnostic or screening test or referral, they wanted the results in order to reach an appropriate diagnosis. Dr. Harbrecht recommends certain key components to a reliable follow-up system. Whether paper-based or electronic, it should be built on redundancy, and start with clear, written instructions and information to the patient.

“Once that occurs, we should also be writing that in the patient’s chart in a prominent location.  So, if you have something that is a high risk problem, particularly if you have a tracking sheet on the front of the chart that says that ‘here is the problem that needs to be followed up.’ Let’s say we sent them for a mammogram or we sent them for a chest X-ray or something for a high risk problem, let’s say that ‘here’s the test we sent them for and here’s the due date when we want those results back by.’  If the test does not come back by that particular date then we need to make sure that somebody calls the patient to find out what happened.”

The third main component is to have that information in a centralized location apart from the patient’s chart in case the patient doesn’t come back in or the individual chart is not seen for some reason. Someone must be assigned to review all the orders that are due to ensure the referral report or lab report is back on time. How that is accomplished can vary.

Typically, an electronic system can tie the patient’s referrals and lab orders to the patient’s record and the electronic calendar. Paper-based systems can too, by using logs or tickler files, but sorting is difficult.

“But if you don’t have that type of sophisticated system then at least you can couple this paper based system with an Excel spreadsheet that basically uses the same things that we’ve been talking about.  It would have the patient’s name, date of birth or some other identification, what tests you ordered, what event or test that you ordered that you want to have followed up, the due date of when you want it back, the ordering physician, the priority, and then the date completed.  If you have that on an Excel spreadsheet you can then sort by the ordering physician and say I just want to have all of Dr. Jones’ test or I just want to have all the patients with a high priority test.  That makes it much easier to actually use your system on a regular basis.”

Dr. Harbrecht says practices can start such a system with high risk cases. But once a system is in place it might as well be used for everything that the doctor orders. According to Dr. Gandhi, deciding which intervention for diagnostic failures to start with requires prioritization. This is because of the complexity of the diagnostic process and the complexity of the breakdowns seen in her study.

“We understand which are the most common, but in understanding of those three failures per case was there one that was particularly most critical?  We weren’t able to get at that with the study that we did. The other thing is that this study was using medical charts and malpractice claims to get the information so it wasn’t able to kind of drill down to that level of detail in terms of thinking about which was the most critical failure. But ideally work in the future would really help us to understand further where we should focus. I think right now based on what we know, the ones that are the most common are where we will probably need to start.”

January 1, 2007
0 dislikes

< Back To Patient Safety