Surgeon Wants FDA to Have More Accurate Understanding of Scope RIO Robotic Arm System Malfunction
FDA’s MAUDE Adverse Event database was updated last week and now includes the complete set of 120 MAKOplasty MDRs that MAKO Surgical Corp. disclosed it had filed in its 10-Q dated May 7, 2012. A recap of the disclosure appears in ITEM 5. OTHER INFORMATION (see below). Notably, of the 120 adverse events filed on May 7th, 85 are characterized as RIO system malfunctions, including a number that MAKO Surgical attributes to surgeon error. Below are samples of the malfunctions described in this set of 120 MAKOplasty adverse events – complete descriptions are available on FDA’s website.
Also of interest is an adverse event reported by a physician on September 4, 2012. In the event description field the physician states he did not report the subject event earlier because he/she “was led to believe it was a bizarre one-time event” but had recently learned of “software issues [that] have caused similar problems and that he/she was “reporting now so fda has more accurate understanding as to the scope of this machine malfunction.” The full text of the event is available on the FDA’s MAUDE Adverse Event database page.
Text of MAUDE Entry - MAKO RIO ROBOTIC ARM INTERACTIVE SYSTEM
“Pt was scheduled for lateral compartment mako plasty of right knee. Mako robotic unit malfunctioned causing a loss of tactile feedback and inaccurate resection volume. Due to malfunction of device, pt’s surgery had to be changed intra-operatively to full knee replacement. I had not reported it earlier because i was led to believe it was a bizarre one-time event. However, i have recently received notification from mako company that prior software issues have caused similar problems, so i am reporting now so fda has more accurate understanding as to the scope of this machine malfunction.”
85 of 120 Adverse Events Characterized as RIO Robotic Arm System Malfunctions
Notably, of the 120 adverse events filed on May 7th, 85 are characterized as RIO system “malfunctions”, including a number that MAKO attributes to surgeon error. Below are samples of the malfunctions described in this set of 120 MAKOplasty adverse events.
Don’t Blame it On RIO – Blame it on the Surgeon
- While burring the patellofemoral joint (pfj), the surgeon noted the cut was translated medially and deeper on the lateral side than planned.
- The results of the engineering eval confirmed the hardware failure of the tracking camera, which was discovered through visual inspection.
- Extension of the leg during pfj resection likely caused a small inadvertent rotational shift of the bone array, which caused a shift of the resection.
- Immediately following the event, mako field service inspected the system. It was found that the burr motor malfunctioned due to an electrical short. It was also determined that the issue was not related to user error.
- During the procedure, a “joint angle inconsistency” error message appeared. A few subsequent attempts to register the rio failed, and further troubleshooting attempts were not successful in clearing the error message.
- When the surgeon entered the bone preparation page and was preparing to burr the tibia, the green resection area appeared streaked with black lines throughout the tibial surface. He began burring the green area without any issue. However, when the makoplasty specialist (mps) changed views of the tibia, the black lines no longer appeared. As he continued to burr, the green portions of the bone model turned black once removed instead of white. The final resection of the tibia was on plan and the trial and implants fit well. After resecting the femur, the surgeon noticed a gap between the femoral implant trial and the bone in the posterior region. The surgeon filled the gap with cement and implanted the femoral component.
- More than (b)(6) months after the procedure, on (b)(6) 2011, mako was informed that the tibial onlay case was alleged to have had an unfavorable outcome and had been revised to a total knee.
- The results of the eval revealed that the likely cause of the event was either an inadvertently bumped tibial array or a tibial bone registration technique error. A review of labeling determined that proper set-up and instructions regarding the stability of arrays as well as the importance of using proper bone registration technique are prominently stated.
- The pt had undergone bilateral makoplasty partial knee replacements on (b)(6) 2010. More than four months later, he reported he was still having pain, and talked to his surgeon about it. He said it looks like the implants are misaligned, but did not indicate whether this was the surgeon’s assessment or his own.
- Review of the post-operative x-rays revealed a 10-degrees varus slope in the tibia; the case session file from the rio showed a 6. 5-degrees valgus measurement, overhang of the tibial component, and bone registration errors. Poor cement fixation contributed to the event, as the surgeon noted no cement was adhered to the component upon explantation. To date, there is no evidence that suggests the rio system or the restoris mck system contributed to the event. Mako customer service facilitated numerous discussions with the physician to follow-up on the results of the investigation in order to prevent recurrence of a similar event.
- After burring the tibia, the surgeon collected the femoral checkpoint with the burr and the value was unacceptable high. Troubleshooting procedures led to re-registering the rio which was completed successfully. The surgeon proceeded burr the femur with good results. When checking the tibial cut, the surgeon observed that the cut was deeper than planned, and determined the proper course of action was to convert to a total knee arthroplasty.
- The revision was required because the original procedure was not completed per the user guide. Essentially, the surgeon, using his medical expertise, decided to minimize the initial surgery by not replacing the patella. Unfortunately, this approach did not work and the patella was later replaced.
- The surgeon was scheduled to perform a patellofemoral (pf) partial knee arthroplasty using mako’s robotic arm interactive orthopedic system (rio) on (b)(6) 2010. After resecting the trochlea, the surgeon observed that the cut was 1-2 mm deep, although it had been planned 1-2 mm proud. The final placement was accepted and the surgeon implanted the component successfully.
- During joint balancing pose capture, the surgeon observed that the values for varus/valgus positioning of the knee differed from that he observed clinically by about 10 degrees. Even after re-registering the bones, the perceived discrepancy remained. The case continued without further incident and was reported to have a successful outcome with implants placed to plan [MAKO evaluation: It was concluded that the software performed as intended and reflected an accurate varus/valgus angle; therefore, the issue probably resulted from the surgeon's perception of the position of the bones which can be affected by the disease state of the knee and the tension of the ligaments.]
- After cutting the tibia, the surgeon observed that bone was burred shallower than planned and he stated the stereotactic boundary did not allow him to burr that area. The surgeon determined that the proper course of action was to remove the extra bone using manually and increase component size by one. [MAKO evaluation: After reviewing the session file from this case, it was noted that user error was the contributing factor with regards to this issue.]
- During pre-surgery checks, camera errors were observed; the camera was shutdown, cleaned, and rebooted, which appeared to resolve the issue. During the probe check, additional errors appeared. After rebooting the system and troubleshooting with the guidance of mako customer service, the probe checks passed. After inserting the tibial and femoral tracking arrays, the arrays could not be detected by the camera. Attempts to re-register the rio failed, as the camera could not detect any of the arrays. Due to the delays, the surgeon decided the proper course of action was to convert to an alternate unicondylar implant system.
- After performing a plunge cut to ensure depth accuracy in the patellofemoral joint (pfj), the surgeon began burring and felt the burr “lunged” unintentionally. The surgeon placed the burr at the bottom of the burred area and the system indicated 3 mm had been removed. [MAKO evaluation: The burr being activated while touching the bone surface was the cause of the "lunge" described by the surgeon. Physicians learn in training that after entering the stereotactic boundary, the burr must be activated off of the bone surface and then brought to the area of interest for resection.]
- The surgeon performed a partial knee arthroplasty using mako’s robotic arm interactive orthopedic system (rio) on (b)(6) 2011. After cutting the tibia and one peg hole, the surgeon was constrained by the rio and was unable to burr the second peg hole. The surgeon determined that the proper course of action was to complete the procedure using the manual instrument set and surgeon was satisfied with the outcome of the procedure. [MAKO evaluation - As part of normal complaint follow-up an eval was performed by mako surgical with regards to the robotic arm interactive orthopedic (rio). The results of the eval revealed no evidence of a system related error that would have prevented the system from entering the stereotactic boundary.]
- During the case, rio registration did not pass and the operating room staff decided to replace the end effector. After passing the accuracy check, the case proceeded. Subsequent to burring the femur, the surgeon noted that it appeared the posterior aspect of the femur had been burred beyond what was initially planned. The surgeon confirmed this using the implant trials. The accuracy check displayed an error, and it was observed that the end effector screws had not been completely tightened when the instrument was replaced. The surgeon elected to convert the pt to a total knee arthroplasty. [MAKO evaluation: The evaluation concluded that the event was caused by user error since the end effector was not properly assembled onto the robotic arm when replaced.] BioLogic comment: Who assembled the end effector? The surgeon or the MAKOplasty specialist employed by MAKO Surgical?
MAKO Surgical SEC Disclosure Regarding Late Filed Adverse Events
- We take matters relating to regulatory compliance very seriously. In 2012, as part of our ongoing internal quality management initiatives and systems enhancements, we undertook a retrospective review of all product complaints to determine if we may have inadvertently failed to file certain Medical Device Reporting (“MDR”) reports with the U.S. Food and Drug Administration (the “FDA”) during the period 2010 to present. Based upon criteria set by the FDA as well as our own internal MDR reporting criteria, our internal review preliminarily identified potential MDR reportability for 105 of such complaints. Significantly, no new or unknown product safety issues were discovered in this retrospective review.
- On May 4, 2012, we met with and solicited the advice of the FDA as to the reasonableness of our enhanced MDR reporting systems and retrospective review, which resulted in the submission of 120 MDR filings on May 7, 2012. The filing of these MDR reports could result in scrutiny of the MDR reports, or an inspection of our records and reporting procedures, by the FDA, which could result in issuance of a warning letter with respect to such procedures. We do not believe, however, based upon the nature of the MDR reports, our interactions with the FDA and all other information currently available to us, that the filing of these MDR reports and the potential regulatory and other consequences related to such filing, if any, will have a material adverse impact on our results of operations.
- We have implemented corrective and preventive actions, including revised internal reporting procedures, revised standard operating procedures and additional employee training, to address and prevent regulatory issues from occurring in the future. We believe that we have made significant progress in transitioning our organization to increase focus on regulatory compliance and in implementing solutions to enhance our quality systems. Notwithstanding our continuing efforts in these areas, scrutiny or inspection by the FDA could result in regulatory consequences to us as described in greater detail under Item 1A, “Risk Factors,” in our periodic filings with the Securities and exchange Commission, including our annual report on Form 10-K for the year ended December 31, 2011.
Other RIO Robotic Arm System Malfunctions Identified in FDA MAUDE Adverse Event Database
Event Date: November 14, 2012
Reported by: Physician
The surgeon was performing a MAKOplasty partial knee arthroplasty procedure using MAKO Surgical’s RIO Robotic Arm.
- Before starting to burr the femur, the surgeon could not pass the femoral checkpoint. Hospital staff checked optical tracking discs for debris and alignment.
- A re-check of the burr checkpoint then also failed, as did an attempt at RIO registration.
- The surgeon determined that the proper course of action was to convert to a total knee replacement procedure.
MAKO Surgical Narrative: “An evaluation of the event has been initiated at MAKO Surgical. No preliminary results are currently available.”
Event Date: December 20, 2012
Reported by: Manufacturer
Event Type: Injury
Patient Outcome: Disability
- The surgeon was performing a MAKOplasty procedure.
- The surgeon reported a 1-2 cm lateral shift from the medial side of the trial after resection.
- The surgeon attempted to correct the lateral shift with re-verification of checkpoints, re-registration of the bone and recollection of the patient landmark.
- The surgeon determined that the proper course of action was to convert to a total knee replacement procedure.
MAKO Surgical Narrative: An evaluation was conducted. The investigation determined that the root cause is a combination of the following:
- Non-optimal bone registration technique.
- Array shift compounded by non-compliant checkpoint check technique.
- A bone array shift is the primary root cause of the shifted implant; a shifted bone array could result in a resection location different than expected.
- The bone array shift was confirmed following review of the case files.
- The investigation also found that user error occurred, the surgeon manipulated the checkpoint, thereby circumventing the designed mitigation for an array shift.
MAKO Surgical Investigation Pins Some Blame on Surgeon – Array Shift
BioLogic Equity Research Comment: It’s interesting that MAKO Surgical found that the surgeon circumvented “the designed mitigation for an array shift.” A MAKOplasty specialist is assigned to every surgery. As described by MAKO, a specialist: “Provides surgeons and operating room staff with support in the operation of the company’s RIO® robotic arm system and associated equipment and instruments.”
A story recently appeared in the Daily Press on February 18, 2013 that describes a surgery performed at Bon Secours Mary Immaculate Hospital. It’s helpful in understanding part of the role of the MAKOplasty specialist:
“The team called a “timeout” to go over the patient information and the procedure. Seven people — including Carter, surgical techs, nurses, a physician’s assistant, and a MAKOplasty specialist — huddled around Ritch’s draped form. The information from the earlier CT scan and plan for the sizing and positioning of the implant were fed into the computer. Larry’s leg was fitted into a boot with a small metal ball on the end, ready to slide into a fixed track to keep it stable and aligned.”
Another article that appeared in a Colorado online news source reported on a procedure conducted at Sky Ridge Medical Center and is also helpful in understanding the role of a specialist:
“The more information you have, the better decisions you can make, just like any business,” said Sarah Scaringe, MAKOplasty specialist for MAKO Surgical Corp. The surgeon, in this case Dr. Harold Hunt, and Scaringe map out the specific areas of the tibia and femur that will be sawed down, and can tweak the parameters mid-surgery. The affected portion of the knee is resurfaced, sparing the patient’s healthy bone and surrounding tissue.”
So, if the surgeon circumvented a safety procedure one would think that the MAKOplasty specialist would have observed it in real time and possibly intervened. That’s not the case.
RIO Robotic Arm Freezes During MAKOplasty Surgery
Model Number 201000
Event Date 07/19/2013
Event Type Injury
Patient Outcome Hospitalization
The surgeon was preparing to perform a partial knee arthroplasty using mako’s robotic arm interactive orthopedic system (rio). During presurgery checks, the software appeared to be running slowly. When the surgeon was reviewing the implant plan, the system froze. The patient was under anesthesia. Multiple attempts to reboot the system were unsuccessful, and the case was cancelled.
As part of normal complaint follow-up, an eval of the event was completed by mako surgical. The root cause of the event is inconclusive. The eval indicates that a hardware error likely occurred, the rio hard drive has been replaced, and no further problems have been reported at the site to date.
Revision Due to Deep Resection from Original MAKOplasty Procedure
|Event Date 06/18/2013|
|Event Type Injury Patient Outcome Required Intervention|
|The surgeon had performed a partial knee arthroplasty using mako’s robotic arm interactive orthopedic system (rio) and the restoris multicompartmental knee system (mck) in 2010. Complaint of new knee pain developed, and the surgeon performed a revision to a total knee arthroplasty.|
|The surgeon explained that the tibial component was placed deeply into the bone, a technique that the surgeon has since modified. The surgeon currently places the implant more proudly, and has achieved good results. An augment was required on the medial side of the tibia during the revision to compensate for the deep resection from the original procedure.|