Safety of Outpatient Single-Level Cervical Total Disc Replacement

A Propensity-Matched Multi-Institutional Study

Patawut Bovonratwet, BS; Michael C. Fu, MD, MHS; Vineet Tyagi, MD; Nathaniel T. Ondeck, MD, MHS; Todd J. Albert, MD; Jonathan N. Grauer, MD


Spine. 2019;44(9):E530-E538. 

In This Article


Spine procedures that were once only performed on an inpatient basis are now being considered in the outpatient setting. This shift has been reported to be associated with high levels of patient satisfaction and reduced health care costs.[13,18] Single-level anterior cervical procedures such as ACDF and CTDR are two procedures for which this is the case.[1,5] ACDF has been relatively well studied in this setting.[10,19] However, although previous studies have supported the safety of outpatient CTDRs, their results were derived from limited sample sizes and lacked pertinent quality outcome measures such as 30-day readmissions.[1,5] To address these prior limitations, the current study utilizes the NSQIP database, a national database with a large patient population and wide breath of recorded perioperative variables, to investigate any variances in perioperative complications between outpatient and inpatient single-level CTDRs.

After propensity score matching, the current study revealed no statistical difference in rates of individual or aggregated adverse events between outpatient and inpatient single-level CTDR. In addition, the current study demonstrated similar complication profiles between outpatient single-level CTDR and outpatient single-level ACDF, which is already being routinely performed. These findings endorse the safety of outpatient single-level CTDR procedures and corroborate findings reported in previous smaller cohort studies.[1,5]

A potential concern of outpatient surgical procedures is also the possibility of delayed diagnosis and treatment if a SAE was to occur following discharge.[20,21] Complications such as cervical hematoma, venous thromboembolism, and other life-threatening complications are particularly worrisome.[22] Therefore, in order to provide more comprehensive insight on safety of outpatient CTDRs, further analysis on rates of postdischarge complications was also performed. After propensity score matching, no statistical difference was found for the rate of individual or aggregated adverse events that occurred following hospital discharge between outpatient and inpatient CTDR. These findings further support the safety of outpatient single-level CTDR procedures.

Yet, another concern regarding outpatient surgical procedures is the potential for higher rates of early readmission, particularly in the first 2 to 3 days after the procedure.[23] Common reported reasons for early readmission include insufficient pain control or symptomatic anaemia requiring blood transfusion.[23,24] The current study is the first to demonstrate no difference in rates of 30-day readmissions between outpatient and inpatient single-level CTDR procedures. This finding signifies that current enhanced surgical protocols and perioperative care pathways are able to handle postdischarge complications without requiring hospital readmission.

Results from the current study demonstrate that outpatient CTDRs have similar complication profiles to inpatient procedures. However, in order to minimize perioperative adverse events and ensure patient safety, careful patient selection is crucial. A previous study has identified certain patient populations as unsuitable candidates for outpatient surgery due to increased risk for postdischarge complications. These patients include those with ASA of 3 or higher or those with advanced age (>80 years).[25] Although the current study did not attempt to determine selection criteria for patients to undergo outpatient CTDR, the authors of the current study generally agree with the conclusions of Sher et al[25] in that patients with advanced age and high comorbidity burden are not appropriate candidates for outpatient CTDR procedures. In addition, it is the author's opinion that patients who lack an adequate support system at home or access to post discharge care are not suitable candidates for outpatient CTDR surgery as well. In the current study, patients who underwent outpatient CTDR procedures tended to have lower rates of hypertension and lower ASA classification than those who underwent inpatient procedures. Even though this was controlled for using propensity score matching, this variance highlights the need for guided patient selection when recommending outpatient surgical procedures for patients. Further, meticulous perioperative care pathways, which should include patient education and pain management regimens, may help to further minimize complication and hospital readmission rates.[18]

However, it is worth pointing out that in the current study, some patients who underwent outpatient CTDR did experience postoperative SAEs (three patients) and readmission (two patients), albeit at a low rate. Of note, one of the patients from the outpatient CTDR cohort who was readmitted had to undergo a reoperation as well. Due to these findings, it is the authors' opinion that providers still should have privileges at an acute care hospital so that patients who develop complications not suitable for intervention at a surgery center can be treated in a timely and effective manner. In addition, it is the authors' opinion that surgery centers should have transfer agreements with an acute care hospital so that serious postoperative anesthesia and medical problems can be managed safely. All practices should center on the patient's best interests and ensure that patient safety is always the top priority.

Strengths of the current study include the high-quality data from the NSQIP database, which includes a wide breath of recorded perioperative variables.[26] Limitations of the current study stem from those related to the use of a national database. These include the lack of spine-specific outcome measures and the short follow-up period of only 30 perioperative days. In addition, even though propensity score matching was employed to minimize selection bias for those who undergo outpatient procedures, the algorithm is dependent upon available patient demographic data. Further, missing data are a potential limitation of large database studies, as different missing data treatments can affect study outcomes.[27] However, in the current study, only 11 (0.55%) patients were excluded due to missing preoperative characteristics. This low percentage of missing data is not expected to bias any results.[27] In addition, the NSQIP database does not record the type of prostheses used for CTDR procedures, so these data could not be analyzed in the current study. However, as the NSQIP database records data from over 500 hospitals in the United States over the span of a decade, the effects of and biases toward a particular type of prosthesis should have been statistically insignificant. Lastly, as there is no specific CPT code for bone morphogenetic protein (BMP) use, this variable could not be analyzed in the current study. Although there is a CPT code for "Allograft, morselized, or placement of osteopromotive material," 20930, this code is not specific for BMP. We felt that if we utilized this code to categorize cases as having used BMP, we would report erroneous BMP usage rates. From previously published studies using the National Inpatient Sample database by Ong et al[28] and Guzman et al,[29] the reported rates of BMP usage for anterior cervical fusion range from approximately 10% in 2006 to 5% in 2013. As the patient population in the current study is also from across the United States, we postulate the BMP usage rate for ACDF cases to be similar to those reported by these previous studies. For reference, the United States Food and Drug Administration issued an official warning in 2008 regarding life-threatening complications associated with recombinant human BMP in cervical spine fusion.[29]

In summary, the current study compared the rates of perioperative complications that occurred anytime during the 30-day perioperative period and that occurred following hospital discharge between outpatient and inpatient single-level CTDR procedures. After propensity score matching to control for patient selection biases, there were no differences in any individual complications, aggregated adverse events, or rate of 30-day readmissions between the two groups. These results provide evidence toward the safety of outpatient single-level CTDR procedures performed in cautiously selected patients.