How to Find the Best Neurosurgeon and Understand Surgical Quality

Jump To Section

• Postoperative Mortality Limitations
• Defining True Surgical Quality
• Clinical Outcomes Research Study
• Key Findings from Neurosurgery Data
• Future of Surgical Quality Metrics
• Full Transcript

Postoperative Mortality Limitations in Neurosurgery

Dr. Philip Theodosopoulos, MD, explains that postoperative mortality is not a useful metric for assessing neurosurgical quality. In cardiac surgery, risk-adjusted mortality rates between 1% and 4% provide a clear benchmark. However, Dr. Philip Theodosopoulos, MD, emphasizes that neurosurgery has an extremely low postoperative mortality rate. This metric fails to differentiate between the skills of individual surgeons or the quality of different hospitals.

Dr. Anton Titov, MD, and Dr. Theodosopoulos discuss the concept of risk-adjusted outcomes. Highly specialized academic centers treat the most complex and high-risk patients. Simple complication rates without risk adjustment are misleading. The expected morbidity and mortality for these difficult cases is inherently higher, making raw data an inaccurate measure of a surgeon's true skill.

Defining True Surgical Quality for Patients

True surgical quality is defined by a set of patient-centered outcomes beyond survival. Dr. Philip Theodosopoulos, MD, advises patients to ask specific questions about their potential recovery. These questions focus on functional outcomes and quality of life after a procedure. Patients need to understand their chances of returning to work and maintaining independence.

Dr. Theodosopoulos lists critical questions for a preoperative consultation. Patients should inquire about the risk of major infection and the expected hospital stay length. For neurosurgery, questions about cognitive function and mental ability are paramount. Dr. Anton Titov, MD, highlights that these factors are what truly matter to individuals facing brain surgery.

Landmark Clinical Outcomes Research in Neurosurgery

Dr. Philip Theodosopoulos, MD, led a major study reviewing outcomes for over 5,000 neurosurgical procedures. The research involved 19 neurosurgeons and aimed to accurately capture complication data. A key innovation was the prospective, point-of-service data collection method. Information was recorded by every member of the care team, not just the surgeon, to minimize bias.

Dr. Philip Theodosopoulos, MD, explains the database was rigorously audited for accuracy. The team reviewed a random sample of patient records to verify reported complications. This method proved that large-scale, accurate clinical outcomes research in neurosurgery is possible. The study challenged the long-held belief that such data could not be collected meaningfully.

Key Findings from Real-World Neurosurgery Data

The research provided realistic benchmarks for neurosurgical complication rates. Dr. Philip Theodosopoulos, MD, notes that the study found craniotomy complication rates are approximately 10%. This figure is significantly higher than the 1% rate many surgeons might assume or quote. Establishing these accurate baselines is crucial for honest patient communication and quality improvement.

Dr. Philip Theodosopoulos, MD, emphasizes the study demonstrated the scalability of this outcomes measurement model. The method can be applied to large and small practices alike. It allows for analysis of simple metrics like hospital length of stay and complex ones like specific surgical complications. Dr. Anton Titov, MD, notes this provides a framework for comparing surgeon performance against established norms.

The Future of Surgical Quality Metrics and Transparency

The future of assessing surgical quality lies in big data analysis and electronic health records. Dr. Philip Theodosopoulos, MD, states that accurate data transcription is imperative. Electronic records allow for the precise capture of patient comorbidities and complications. This data is essential for effective risk adjustment models.

Dr. Philip Theodosopoulos, MD, explains that large datasets enable powerful analysis of treatment outcomes. Combining data from multiple institutions provides the statistical power to define norms and outliers. This movement towards transparency empowers patients to make informed choices. Dr. Anton Titov, MD, concludes that this evolution is key to helping patients find the best neurosurgeon based on comprehensive quality data.

Full Transcript

Dr. Anton Titov, MD: How to find the best neurosurgeon for brain tumor treatment? What is "quality" in surgery? How to find the best surgeon in any specialty? These questions are most important for every patient who needs surgical treatment.

Dr. Philip Theodosopoulos, MD: In cardiac surgery, risk-adjusted postoperative mortality is a useful metric. But in many surgery subspecialties, postoperative mortality is very low. It is true in neurosurgery. The ability to complete surgery without killing a patient says little about the ultimate success of an operation. Postoperative mortality is not a useful metric of quality. It does not say much about the skills of an individual surgeon or the quality of a hospital.

Dr. Anton Titov, MD: A leading neurosurgeon with a special interest in clinical outcomes research shares his vast experience. How do you measure the quality of surgeons and the success of surgical operations? How do you find the best surgeon? How to find the best neurosurgeon? What is "quality" in surgery?

Find your surgeon or specialist. Knowing what quality means in surgery. A video interview with a leading expert in neurological surgery.

Dr. Philip Theodosopoulos, MD: Practical information on how patients should choose a surgeon. The results of surgery are the most important thing for patients. A medical second opinion confirms that a brain tumor diagnosis is correct and complete. A medical second opinion also helps to choose the best treatment for a brain tumor. Seek a medical second opinion on a brain tumor and be confident that your treatment is the best.

Neurosurgery has large variability in lesions and procedures. For neurosurgical operations, postoperative mortality is extremely low. The chance of dying after surgery is NOT a good indicator of the quality of neurosurgical care. Patients should ask questions.

Dr. Anton Titov, MD: What are my chances of complications? What if I have to be in the hospital for two weeks or two months? What are my chances of having a major infection? What are my chances of going back to work? Will my brain work the same way after surgery? Will I have the same mental ability after surgery? What are my chances of being independent and enjoying life? You need to understand what it means to have high quality in surgery.

Dr. Philip Theodosopoulos, MD: Only then can you find the best surgeon for any type of surgical operation. To find out surgical skills and results of surgery is not easy for a patient.

Dr. Anton Titov, MD: Get a medical second opinion on your diagnosis. It will help you to find the best surgeon. How good is the surgeon?

How to find a quality surgeon?

Another trend in modern medicine is increasing transparency of data on clinical outcomes. It means knowing surgery results both for hospitals and for individual surgeons.

Dr. Philip Theodosopoulos, MD: One of the goals is to provide patients with practical information on neurosurgery complications. It can help them to choose the best surgeon for cancer treatment. Patients can choose the hospital where they can have a surgical operation with fewer complications. Another goal of clinical quality metrics is to improve clinical outcomes for hospitals and for individual surgeons.

We discussed in Boston with Dr. Lawrence Cohn, MD. Risk-adjusted morbidity and mortality is very important information about a surgeon or a hospital. It is important because the surgeons at highly specialized academic centers treat the sickest, most difficult patients. Simply the numbers of death rate and frequency of side effects after a surgical operation do not tell you much.

Side effects of a surgical operation are morbidity and mortality. Because the risk for the very sick patients academic surgeons treat is much higher, the expected complication rates and even death rate are also higher.

Dr. Anton Titov, MD: The statistical expectation of clinical outcomes is different for low-risk patients. That is the concept of risk-adjusted morbidity and mortality.

In neurosurgery, there is a very high variability in lesions and procedures. It is easier to assess and standardize something like coronary artery bypass graft surgeries (CABG). It is harder to standardize neurosurgical operations. For example, the type of surgery that a skull base neurosurgeon does.

Dr. Anton Titov, MD: You have done recently a very interesting major study. It was a review of clinical outcomes for more than 5,000 neurosurgical procedures. Nineteen neurosurgeons did these surgical operations over the course of one year. Could you please talk about this study and what have you learned?

Dr. Philip Theodosopoulos, MD: Neurosurgeon, Director of Skull Base Tumor Program. Yes. Clinical outcomes are one of the most important things in medicine and surgery. Clinical outcome means the results of surgery and other treatments. Finally, the payers now systematically scrutinize clinical outcomes data. At least the government and private payers do that in the US.

I say this in talks and patients laugh at it. But it is true. You have more information about a steam iron you want to buy than about your neurosurgeon. There are a lot of reasons for that. The risk adjustment of results of surgery is only one part of quality metrics of surgeons and surgery.

I have trained many European neurosurgeons. I have seen Europe directly as a patient. I lived in Europe and studied complications after neurosurgery treatment. It is the same in the US. We have protected the independence of the medical profession for too long. We shielded clinical outcomes of how well our patients do. We protected doctors from reporting results of surgical operations publicly.

We protected doctors under many different disguises. It includes the disguise of the Hippocratic oath. "We would do the best that we can, et cetera. What happens is not only up to us, it is up to a lot of factors."

Dr. Anton Titov, MD: It is true. In reality, health is a very complicated equation to optimize. It starts probably with health and disease prevention. Optimization of health probably does not start with a patient coming into my office with a really large tumor in the base of the skull. At that point, the clinical outcomes are very limited. What outcomes could we expect?

Dr. Philip Theodosopoulos, MD: When a patient already has a large tumor. But when you look at clinical outcomes, there are so many different criteria to choose from. Surgery results can be very different. Clinical outcomes research has been my fascination over the past two decades. It fascinated me ever since I was in training. I had specialist training in clinical outcomes measurement techniques.

Dr. Anton Titov, MD: This is what turns out to be the case with complications of brain tumor surgery. In most of surgery, it is doubly difficult to assess the quality of an operation and the skill of a surgeon. It is more difficult than in any other part of medicine. You are right about cardiac surgery complications. It is also true about coronary artery bypass surgery complication rates.

Dr. Philip Theodosopoulos, MD: But cardiac surgery has one unique characteristic. It made it easy to apply clinical outcomes first in cardiac surgery. That characteristic of brain aneurysm treatment complications is the following. Cardiac surgery has a defined mortality. In most other surgery types, mortality is not a very well-defined number. It is so because of the many improvements. We have made many improvements in surgery quality assessment over many, many years and decades. We stand on the shoulders of many physicians and surgeons now.

For neurosurgical procedures, postoperative mortality is very, very low. It is something that wouldn't really be a good indicator of how good or bad the hospital is. Postoperative mortality in neurosurgery is not a good indicator of how good the neurosurgeon is.

In cardiac surgery, mortality is known. It is between 1% and 3% or 4%. Then it becomes easy to assess the quality of a heart surgeon. Because if you wanted to just rank everybody, you have this one determinant of quality. Everybody cares about the death rate after heart surgery. This factor is "I don't want to die". It is the risk of death after surgery. Postoperative mortality is available as data from the hospital. It is also available from Social Security databases. It is available from a lot of other factors that are very accurate. Because once you die, it is reported. You can have the ability to measure surgical quality by measuring the death rate after surgery. This works well for cardiac surgery complication rate assessment.

Cardiac surgery was the very first surgical field to have clinical outcomes studied in the 1980s.

Dr. Anton Titov, MD: Yes, it was in New York. They started doing clinical outcomes studies in cardiac surgery first. Then evidence-based medicine started being a true scientific discipline.

Dr. Philip Theodosopoulos, MD: But even in cardiac surgery, a lot of clinical outcomes analysis stops there. It stops at postoperative mortality data. Over the past several years, we all have thought that research on results of surgery shouldn't stop there. We thought that postoperative mortality in cardiac surgery is not good enough.

Most patients would want to know what is their chance of making it through a surgical operation. But they also want to know other factors that affect the results of their surgery.

Dr. Anton Titov, MD: What are my chances of being after surgery the same as I was before surgery? What are my chances of staying in the hospital for two weeks or two months? What are my chances of having a major infection or complication? What are my chances of being able to go back to the work that I used to do? What are my chances, in neurosurgery, of my brain working the same way? What are my chances of having the same cognitive function? What are my chances of being independent and enjoying life?

We have some measures for patients about important results of surgery. There have also been a lot of studies that focused on a small group of patients. They had a specific disease or specific surgical operations and complications after an operation.

Dr. Philip Theodosopoulos, MD: But there were very few big studies of complications in the surgical field. A couple of studies on complications after surgical operations were conducted in the Veterans Administration hospitals. They had hundreds of thousands of patients. But then again you get into these huge databases. They are not really directly connected to data on surgery complications from each individual patient. Data on treatment complications is extracted from databases and from medical records. Many studies in the quality of a surgeon had bad enough inaccuracy. You do not really know how robust your conclusions were.

We did research on results of surgery in Cincinnati. It was an initiative that we had done for about six or seven years. Then we actually published results of surgery complications. For all neurosurgery operations in the entire University of Cincinnati, we collected all of our data prospectively. We gathered data at the point of service. It means that I see you as a patient before surgery, after surgery, during surgery. Then I immediately record specific clinical outcome parameters for you. We record these data on a point-of-service prospective time. We expect that it would be as accurate as it can be.

Second, it wasn't just the surgeon who recorded results of surgery. At times, a surgeon can be biased. Bias could be positive or negative. Bias almost could be in either direction.

Dr. Anton Titov, MD: But everybody who interacted with the patient recorded the data about surgery results. Everyone recorded treatment results. It was medical assistants or anyone who cared for the patient. Finally, it was an audited database. The database had five thousand-plus cases after one year of study that we had already reported.

Dr. Philip Theodosopoulos, MD: But after my leaving for UCSF, the surgeons now have recorded many more thousands of patients into the database. We audited the data. We looked back at brain tumor surgery complication rates. We took, for example, 5% of Anton's patients and 5% of my patients. We looked at how accurately the data was transcribed.

Dr. Anton Titov, MD: Did this patient really have a postoperative infection after surgery? What exactly happened? The first lesson we found is that you can do a research study on complications after brain aneurysm surgery treatment. This is very important. This may sound simplistic to lay patients, even to some doctors.

Dr. Philip Theodosopoulos, MD: But I have been in clinical outcomes research now for almost two decades. Many times an argument was made that it is not possible to do such a study on complications of surgery. Many good doctors believed that accurate audited large research about results of surgery cannot be done accurately. They thought that such a study cannot be done well enough to really mean anything. We proved that such research on neurosurgery complications can be done well enough.

But how do you know it is done well enough? For many of the surgery results, we don't really know what that number of complications should be.

Dr. Anton Titov, MD: What's the rate of postoperative spinal fluid leaks after this or that procedure? You look at some of the indicators of surgery complications in our study. Side effects of surgical treatment were suddenly well defined.

Dr. Philip Theodosopoulos, MD: You can go to the national list of rates of complications and compare yourself to it. We could judge and compare our complication rates to the data that was very specific for specific diseases from good centers. We found that what we were reporting was actually very similar to that data. But not in a good way or in a bad way. For example, rates of complications following craniotomy are about 10% in real good studies. They are not 1% as many surgeons think.

The first thing we showed was that such large and rigorous study can be done. Research can be done well.

The second finding is that such a method of clinical outcomes research is scalable. You can really extend this type of clinical outcome study. You can use surgery complications research in a big clinical practice or small clinical practice. You can use complication rates for all types of diseases. Then you can utilize surgery results from the basic, simplistic things. For example, how long do patients stay in the hospital for a certain surgery procedure?

You can also ask more complicated questions. For example, what complications can you expect following a certain neurosurgery procedure? You can ask about costs of healthcare and procedures. You can ask a lot of other things that patients are very interested in. You can find complication rates in surgical practice.

Dr. Anton Titov, MD: Where does that leave us now? Where that leaves us is back to the risk adjustment concept. There is really no good method yet to risk-adjust very easily for patients and for co-morbidities. We need techniques of big data analysis to really start being applied to surgery complications research.

Dr. Philip Theodosopoulos, MD: We need very accurate transcription of data. We need the electronic medical record. Electronic medical records had just appeared when we started our study. The electronic medical record is imperative to have. Because it allows you to capture co-morbidities and other factors and surgery complications that happen to patients. You can do that with great accuracy. It is much better than somebody scribbling down a note in a chart. Then you may never be able to find the chart. A patient's chart may fall out.

Now that we have the data recorded in a much more accurate method, we need techniques of big data to do surgery complications research. Because we do 2,000 or more craniotomies each year here at UCSF. We are the busiest place in brain tumor surgery and many other neurosurgical treatments. Very quickly you amass the amount of data that individual physicians cannot really sort through.

Dr. Anton Titov, MD: For example, it is hard to manually analyze co-morbidities and various associations of surgery complications. Your data on side effects of treatment needs to be combined with data from other places. That is where you really can have the power of big data to analyze brain tumor treatment complication rates. That is where medicine is going. This is how to find out whether a patient or a surgeon are within the norm or outside the norm for complication rates after a surgical operation?

What is the norm in side effects of surgical treatment?

You have established the metrics that can be used to understand the skill and quality of surgeons and hospitals. You established a certain baseline for potential surgical complication rates. You found benchmarks for the length of stay of a patient in a hospital for surgical operations in neurosurgery. You are dealing with brain tumors and brain aneurysm ruptures as a neurosurgeon. That is very important to establish real neurosurgery complication rates. Because otherwise, it is "garbage in, garbage out".

Dr. Philip Theodosopoulos, MD: Correct. Finally, we are at the time where we can do research on surgical complications. It is not cheap. It requires a lot of cooperation by a lot of surgeons, nurses, administrators. It requires the qualified staff who actually do this for a living for your practice. You have to pay them to do this data collection and analysis. It requires development of electronic medical records and forms for surgery complications. There's a lot that goes into it. But there is a payout.

Dr. Anton Titov, MD: How to find the best neurosurgeon? What is "quality" in surgery? A video interview with a leading expert in neurological surgery. What are clinical outcomes in surgery?