New Harvard Study Could Change Orthopedics Forever
In what is likely the one of the first-ever attempt to map a cycle of care using an advanced economic model, researchers from Harvard have joined together with business gurus at the same institution. Jon J.P. Warner, M.D. is chief of the Shoulder Service at Massachusetts General Hospital and co-director of the Boston Shoulder Institute Fellowship worked in collaboration with Laurence D. Higgins, M.D. chief of the Sports Medicine and Shoulder Service at the Brigham and Women’s Hospital. A former president of the American Shoulder and Elbow Society (ASES), Dr. Warner told OTW, “Dr. Higgins and I have worked with Robert Kaplan and his colleagues at the Harvard Business School, and have just completed a study where we looked at the actual cost of an episode of care for rotator cuff surgery. By mapping the entire cycle of care and carefully analyzing each step, we were able to reduce the price of the entire one year episode of care by 15%.
“The tool we used—Time-Driven Activity Based Costing (TDABC)—allowed us to map the cycle of care. This accounting methodology analyzes use of resources including individual caregiver’s time for each step in the care cycle. A cost-capacity rate for an individual is their cost per minute for their time based on their salary and benefits and number of minutes worked per year. This is also combined with cost for using equipment and space. Such accounting is more accurate in representing true costs than allocation of historical costs. Moreover, this method allows us to analyze a care map looking for bottlenecks in the process and thus improve the efficiency of the entire care process. Applying this methodology allows us to bring healthcare delivery on par with other industries.”
“The single most important factor in the cost of rotator cuff repair was personnel (73%). So if you take a look at the evidence based medicine you see that the patient does not need physical therapy (PT) in the first four weeks after rotator cuff repair and voilà…you just saved a substantial amount of money.”
“We also found that in the entire year of care the day of surgery accounts for 53% of the total cost of care; PT is 22%. As a field we need to understand the value of taking apart what we are delivering. With this work we were able to improve 15 different elements of the cycle of care. We started the episode of care on the day that the patient signed up for surgery. This involved: measuring how many places he or she went, how many stops they have to make (and the indirect and direct costs of those stops), how many people they see on the day of surgery, how much time they spent in preanesthesia, what sort of bottlenecks there are from an equipment standpoint. Two institutions—Massachusetts General and Brigham and Women’s ambulatory center (Faulkner Hospital)—were studied. When all was said and done the cost difference between having rotator cuff surgery at one hospital versus another was inconsequential. These institutions used the same pathways of care and came up with the same cost of a cycle of care!”
“In addition to this study, we have established an outcomes registry for rotator cuff surgery; I believe that we are the first researchers to calculate the recovery curve for rotator cuff and shoulder replacement. The process is such that patients go online and input their feelings and information about pain levels. Then we calculate the mean recovery rate of patients’ pain and functional improvement. The most amazing finding was that two different surgeons at two institutions achieved the exact same outcome! People who have shoulder replacement recover faster than rotator cuff and do not experience much pain as compared to those undergoing rotator cuff surgery. But if you look at the standard deviation it decreases as the patient gets further from the point of surgery. So five weeks after surgery we’ll do risk modeling, but I think it’s high likely that these numbers represent a serious complication. Cardiothoracic surgeons have done some of this work in the past. We need to catch up with them…and with industry.”
Which Spine Surgery LEAST Likely to Repeat? Riew and Hilibrand Tackle Question
Dan Riew, M.D. is the Mildred B. Simon Professor of Orthopedic Surgery, professor of Neurological Surgery, and the chief of the Cervical Spine Service for Washington University Orthopedics and director of the Orthopedic Cervical Spine Institute. He and his co-authors Jae Chul Lee, Sang-Hun Lee and Colleen Peters have just published the lead study in the November 5, 2014 issue of Journal of Bone and Joint Surgery (JBJS) looking at the incidence of reoperation for adjacent level pathology (ASP) following a variety of spine operations. With nearly 1, 400 consecutive patients, says Dr. Riew, it is the largest such series to date. He tells OTW, “All patients in this retrospective study had been diagnosed with radiculopathy, myelopathy, or myeloradiculopathy and underwent cervical spine surgery performed by a single surgeon. We looked at anterior and posterior surgery, both fusion and non-fusion operations, and calculated the annual incidence of ASP requiring surgery and found that the operation that is least likely to require repeat surgery was laminoplasty…this was pretty surprising. The surgery that was most likely to require an additional procedure was fusion done from back of the neck (a 7.5x greater risk of reoperation for ASP than laminoplasty).”
“This was the first study to look at reoperation rates following all types of cervical operations in large numbers, using Kaplan-Meyer survivorship analysis. My good friend Alan Hilibrand of the Rothman Institute did a study years ago where he defined ASP, not by reoperation rates, but by when a patient came back to see Henry Bohlman on two consecutive occasions complaining of problems at the adjacent level. Most surgeons misquote his paper, thinking that he looked at reoperation rates. He argues in this article that his is more objective in that it is patient and not surgeon-controlled. While that may be true, the surgeon can still minimize patient reported symptoms, or worse, not even diagnose them properly. We both trained with Dr. Bohlman, and we both saw that if a patient complained of neck pain after a myelopathy or radiculopathy operation, he would minimize it by saying, ‘don’t sweat the small stuff’ and usually not even mention it in his notes.”
“Also, if a patient complained of ulnar forearm and hand symptoms, he would diagnose a cubital tunnel syndrome instead of a C8-T1 radiculopathy. Despite all the C4, 5, 6 corpectomies that he did, I never saw him do a C7-T1 or T1-T2 level operation once during my year with him. Nor did he ever diagnose a C8-T1 radiculopathy in that year. So, looking at the office notes for patient reported symptoms and surgeon diagnoses can also be inaccurate. But reoperation is very objective. More importantly, in order to compare apples to apples, we all have to have a uniform criterion by which we determine ASP rates. Every arthroplasty study defines ASP as reoperation. In fact, other than his paper, I don’t know of a single paper in the literature that defines ASP as ‘two consecutive office visits to the surgeon with complaints related to the next level.’ Also, while I agree with him that patients and surgeons are reluctant to have surgery again, there is no reason that this reluctance should be lower with one type of surgery as compared to another. So when we compare ASP rates amongst different operations, the same reluctance should hold for all of them.”
Dr. Riew: “The surgeon also has control over the first operation. I agree that surgeons look better if they don’t have to reoperate. But if someone complains about the next level and you ignore it, they will go elsewhere. Also, if you are a surgeon that ignores complaints and doesn’t operate, you are also less likely to put it in the chart, and that is 100% under the surgeon’s control. We published a study in JBJS several years ago that showed that office notes are a poor reflection of the patient’s complaints about adverse events (“Accurate identification of adverse outcomes after cervical spine surgery, ” J Bone Joint Surg Am 2004;86(2):251-6). So we know office notes often aren’t objective or accurate. Also, if you ignore the complaints, the patient may go to another doctor…and then you have no follow up. Finally, regardless of whatever conflicts the surgeon may have with implants and studies that may discourage reoperation, possibly the largest financial conflict that the surgeon has that we never seem to talk about is the one where the surgeon gets paid to reoperate.”
Alan Hilibrand, M.D. is The Joseph and Marie Field Professor of Spinal Surgery and professor of Orthopaedic Surgery and Neurological surgery at Jefferson Medical College and the Rothman Institute. Dr. Hilibrand tells OTW, “Patients are often reluctant to have a second operation; sometimes the surgeon is reluctant as well because they may view it as their own ‘failure, ’ although in my opinion it should never be viewed as such. If a patient is failing nonoperative treatment and comes in a second time with persistent complaints, then he or she is probably appropriate for surgery…yet there are many patients who just won’t go through with it. Almost half of Dr. Bohlman’s patients who developed adjacent segment disease didn’t undergo surgery. For this reason, we ran the analysis on the basis of people coming in with symptoms, since nearly half of Dr. Bohlman’s patients—developed symptoms and decided to just live with them.”
“There are situations where the surgeon who did the first surgery is reluctant to do another one and puts the patient off, and it’s not until they see a different surgeon that they are told that they need an operation. For example, in the 1990s when we gathered the data for our study, there were a number of arthroplasty studies where the design surgeon performed the index operation and then decided whether the patient later needed a revision. Obviously, if they did not reoperate then their procedure would look like it had greater durability. So when I did my study, I did not want to ignore the possibility that a reviewer might criticize the study by asking whether Dr. Bohlman might have had that unconscious bias.”
“I think that the most remarkable and noteworthy aspect of Dr. Riew’s new paper is the comparison of the rates of adjacent level issues among the different surgeries, and I agree that basing this comparison upon reoperation rates is very appropriate and valid. And although I agree that the arthroplasty trials used reoperation rate rather than recurrent symptom rate as their determinant of adjacent level pathology, this does not mean that it’s the best way to evaluate for adjacent segment disease. It all depends on the question asked: If a patient wants to know what are the chances that they will develop ‘the same type of problem at another level, ’ I would quote them our data from Dr. Bohlman’s patients (about 3% per year). On the other hand, if they asked, ‘What are the chances that I will need another operation because of a problem at another level?, ’ then I will now quote them the data from Dan Riew’s study, which can answer that question for both ACDF [anterior cervical discectomy and fusion] as well as posterior procedures.”
Blood Transfusions (and Related Concerns) Are Up
Using a database with two million patients, researchers have found that there has been an increase in allogenic blood transfusion among total hip arthroplasty (THA) patients. This is concerning, says Anas Saleh, M.D., an orthopedic surgeon at Cleveland Clinic, because transfusions are not without risk. Dr. Saleh tells OTW, “Blood transfusion is common in total joint replacement (and more common in total hips than in total knees). Despite blood conservation efforts and the advent of commercial products aimed at reducing the perioperative bleeding and blood transfusions, the rate of allogenic blood transfusion is increasing and the patient is still running the risk of pulmonary complications, infection, and a prolonged hospital stay.”
“We used the National Inpatient Sample (NIS) database, which collects information on about 20% of all hospital discharges; it is weighted so that it can be extrapolated to the U.S. population. Along with our statistician colleagues at Case Western Reserve University, we designed the study to look only at primary total hip replacement. They performed sophisticated analyses and controlled for patients’ comorbidities, socioeconomic status, location, race, and insurance. While allogenic and autologous blood transfusions were noted, the focus was on allogenic because it is now more common than autologous.”
“We found that the rate of allograft transfusion has increased over the study period, and that this is associated with an increase in complications (especially deep vein thrombosis, infection, and pulmonary embolism). There is currently a push for preop and perioperative methods of preserving blood. Clinical pathways exist that are aimed at optimizing the patient’s blood counts preoperatively, which include providing iron and/or erythropoiten, to help ensure that their blood levels don’t drop to a level where a transfusion is required. There are also intraoperative tools such as hemostatic agents which can stabilize clots in the surgical wounds. These hemostatic agents are expensive, however, and we need to decide whether these agents are cost-effective, or if they’re effective at all. At Cleveland Clinic most primary hips and knee patients get intravenous or local tranexamic acid, as it has proved to be the most cost-effective option.”
