The most common way of measuring safety is by considering negative outcomes. Negative outcomes include adverse events, sequelae, and failure to cure. Adverse events are defined as any deviation from the normal postoperative course. A sequela is an aftereffect of surgery that is inherent to the procedure. A failure to cure is when surgery is well executed, there are no complications, but the original purpose for surgery has not been realized or achieved. Sequelae and failure to cure are not discussed in this chapter (2,3).

The words “complication” and “adverse events” are often used interchangeably. There is no consistent definition of an adverse event, and many definitions, although having elements in common, are contradictory. For the purpose of this discussion, an adverse event is considered any unfavorable and unintended sign, symptom, or disease, associated with the use of a medical product or procedure, whether or not causally related to the product or procedure.

Hence, any unplanned event is a deviation from the expectations of the surgeon and the patient and therefore meets this definition of an adverse event. If these adverse events result in negative impacts for patients, they can be called complications. Most adverse events do not affect long-term outcomes, and so the term complication is inappropriate for these events. The decision of when an adverse event becomes a complication is somewhat arbitrary but is usually based on the level of severity of the event, whether it is associated with the procedure or device, and whether it requires treatment in a hospital. Complications are thus the subset of the most serious adverse events.

Health care professionals strive for an error-free practice, and failure to do so is often perceived as a failure of character, expertise, or technique. There can be a general reluctance to report what is considered to be a “complication” for fear of litigation or questions of competence, while less severe adverse events may be considered too trivial to report. However, all adverse events must be documented and audited for quality assurance, especially in the context of clinical research.

Adverse events specific to spine surgery have often been categorized as approach related or device related. They have also been described as serious or major events and even preventable events. What is clear is that the terminology associated with the categorization of adverse events is varied, and therefore, the ability to compare results within a specialty through adverse events remains difficult. This profoundly influences the ability to interpret and compare published results (4, 5, 6 and 7).

In 1992, Clavien et al. proposed a classification system, and this has been expanded upon by Dindo et al. (2,3,8). These particular classification systems are developed from
the perspective of medical providers. Application of these classifications by payers or patients is generally not viable.

We propose here a classification system intended to reduce subjective interpretation of adverse events and, therefore, their downgrading (Table 118.1). The system presents a simple, objective, easily remembered, yet comprehensive, severity score that can be used by surgeons and hospitals in a prospective rather than retrospective manner, in hopes of improving uniformity in classifying adverse events in spine surgery. It takes into account World Health Organization recommendations, as well as those published by Dindo et al. in Classification of Surgical Complications (8,9):

Murrey et al. (10) reported on the prospective randomized trial of ProDisc-C versus anterior cervical diskectomy and fusion (ACDF) at one level. They defined adverse events success as the absence of adverse events related to either the implant or its implantation. Adverse events success was achieved in 97.1% (100/103). The three patients who did not achieve adverse events success had two implant-related and one implantation-related events. The two implant-related patients had removal of their prosthesis. Otherwise, in that study, there were no other revisions, reoperations, or supplemental fixations for ProDisc-C total disk replacement (TDR). The implantation-related complication was in association with a dural tear. In this study, device success was 98.1% (101/103).

Anderson et al. (11) analyzed the adverse events related to Bryan artificial cervical disk insertion versus anterior cervical arthrodesis. This was a prospective randomized multicenter trial where 463 people were studied, 242 of which received the Bryan disk. They found there was a 35.1% incidence of medical events occurring 6 weeks or more after surgery in the group receiving the Bryan disk. The majority of these adverse events were related to gastrointestinal and genitourinary issues. They found that 14.9% of medical events within the first 6 weeks of surgery were thought to be possibly or directly related to the operation.

Overall, they reported that adverse events related to the disk replacement group amounted to 33.9%. Most of these related to superficial wound infections (N= 7 vs. 1), dysphasia (N = 26 vs. 16), cardiovascular events (N = 6 vs. 0), anesthesia related (N = 8 vs. 5), and cerebrospinal fluid leak (N = 2 vs. 3) that occurred during decompression, not instrumentation. Serious adverse events occurred twice in the fusion group—one who had a hematoma and the other who had a spinal cord injury. The reporting of adverse neurologic symptoms at all time periods revealed a postoperative incidence of sensory or motor symptoms in 20.7% of individuals. The vast majority of these were grade I/II adverse events.

In terms of grade III and grade IV events, there was a reasonably low incidence in the group receiving cervical total disk replacement (cTDR) versus ACDF, and this was thought to be mainly due to reoperations for treatment of persistent symptoms and/or pseudarthrosis. This study reported on the medical and surgical adverse events that occurred to patients whether they were unrelated or related, directly or indirectly, to the surgery. It showed
that more surgically related adverse events occurred in the investigational group, mainly due to more complaints about dysphasia and superficial wound infections. This was possibly related to the use of the special frame that is required at insertion of this device and the longer operating times that were reported. The reoperation rate in this group was 5.4%. The authors commented that the high rate of additional surgery in both groups within the study was worrisome and indicated that surgeons should be more comprehensive in their preoperative assessment.

Mummaneni et al. (12) reported on the clinical and radiographic results of cervical disk replacement with the Prestige-ST artificial device and compared it to ACDF. Of 541 patients, 276 received the Prestige-ST prosthesis. They reported any adverse event with an incidence of 81.9% in the investigational group versus 80% in the control group. The majority of these adverse events related to neck and arm pain and neurologic symptoms that had no effect on outcome. The incidence of these adverse events in relation to any surgical procedure is consistent across the class 1 studies. They reported device-related adverse events in the investigational group amounting to 3.3%. This included one case of anatomical or technical difficulty with implantation, two cases of implant displacement or loosening, one case of neck and arm pain, four cases of neurologic symptoms, and one case of subsidence. The total incidence of perioperative adverse events associated with the investigational group amounted to 17 events or 6.2%.

A statistically significant lower rate of secondary surgeries (implant revision and supplemental fixation) occurred in the investigational group. The study classified secondary surgical procedures as revision, removals, reoperations, and supplemental fixations. In the investigational group, five prostheses were removed, mainly because of persistent radiculopathy or neck pain. One occurred at 6 months, three at 12 months, and one at 24 months. In the control group, nine reoperations occurred, giving an incidence of 3.5%. It is important to note that this study defined a revision as any procedure in which the original implant configuration was adjusted or modified. There were four reoperations in the investigational group versus two in the control group. Supplemental fixation was not required in the investigational group, whereas eight cases required supplemental fixation in the control group. As a result, the total incidence of secondary surgical procedures was 9 (3.2%) in the investigational group versus 24 (9.2%) in the control group.

The overall success of the study was 96.8% for the investigational group versus 95.5% for the control group. In this study, the Prestige disk was shown to be statistically noninferior to ACDF in all three primary outcome measures (Neck Disability Index (NDI), neurologic status, and functional spinal unit (FSU) height). The neurologic status outcome reached the prespecified level for statistical superiority. However, neither the NDI nor the FSU did so.