Outpatient surgery in the cervical spine: is it safe?

Study Design Systematic review. Study Rationale As the length of stay after cervical spine surgery has decreased substantially, the feasibility and safety of outpatient cervical spine surgery come into question. Although minimal length of stay is a targeted metric for quality and costs for medical centers, the safety of outpatient cervical spine surgery has not been clearly defined. Objective The objective of this article is to evaluate the safety of inpatient versus outpatient surgery in the cervical spine for adult patients with symptomatic or asymptomatic degenerative disc disease. Methods A systematic review of the literature was undertaken for articles published through February 19, 2014. Electronic databases and the bibliographies of key articles were searched to identify comparative studies evaluating the safety of inpatient versus outpatient surgery in the cervical spine. Spinal cord stimulation, spinal injections, and diagnostic procedures were excluded. Two independent reviewers assessed the strength of evidence using the Grading of Recommendations, Assessment, Development, and Evaluation (GRADE) system, and disagreements were resolved by consensus. Results Five studies that met the inclusion criteria were identified. One study reported low risk of hematoma (0% of outpatients and 1.6% of inpatients). Two studies reported on mortality and both reported no deaths in either group following surgery. Dysphagia risks ranged from 0 to 10% of outpatients and 1.6 to 5% of inpatients, and infection risks ranged from 0 to 1% of outpatients and 2 to 2.8% of inpatients. One study reported that no (0) outpatients were readmitted to the hospital due to a complication, compared with four inpatients (7%). The overall strength of evidence was insufficient for all safety outcomes examined. Conclusion Though the studies in our systematic review did not suggest an increased risk of complication with outpatient cervical spine surgery, the strength of evidence to make a recommendation was insufficient. Further study is needed to more clearly define the role of outpatient cervical spine surgery.

Study Rationale and Context

As the length of stay after cervical spine surgery has decreased substantially, the feasibility and safety of outpatient cervical spine surgery come into question. Although minimal length of stay is a targeted metric for quality and costs for medical centers, the safety of outpatient cervical spine surgery has not been clearly defined. Because of the location of surgery, complications that may be innocuous in other parts of the body can literally be fatal in the neck. A compressive hematoma after cervical spine surgery can compromise the airway and if not addressed rapidly, can result in hypoxic complications. A patient in a home setting will not receive medical attention as quickly as an inpatient. In the context of airway compromise, this difference may be critical. This study seeks to examine the safety of outpatient cervical spine surgery.

Objective

The objective of this article is to evaluate the safety of inpatient versus outpatient surgery in the cervical spine for adult patients with symptomatic or asymptomatic degenerative disc disease.

Materials and Methods

Study design: Systematic review.

Search: PubMed, Cochrane Collaboration, EMBASE, National Guideline Clearinghouse databases; bibliographies of key articles.

Dates searched: The dates were searched through February 19, 2014.

Inclusion criteria: The inclusion criteria for the study include the following: (1) patients 18 years of age or older with symptomatic or asymptomatic degenerative disc disease (including myelopathy and radiculopathy), (2) single- or multi-level surgery in the cervical spine (including discectomy, decompression, fusion, arthroplasty), and (3) comparative studies published in peer-reviewed journals (see Table 6 in the online supplementary material).

Exclusion criteria: The exclusion criteria for the study include the following: (1) spinal cord stimulation, spinal injections, diagnostic procedures; (2) studies with fewer than 10 subjects per treatment arm; and (3) case series.

Outcomes: The outcomes of the study included the following: hematoma, death, and other adverse events.

Analysis: Descriptive statistics. Details about methods can be found in the online supplementary material.

Overall strength of evidence: Risk of bias for individual studies was based on using criteria set by the Journal of Bone and Joint Surgery 1 modified to delineate criteria associated with methodological quality and risk of bias, based on recommendation from the Agency for Healthcare Research and Quality.2 3 The overall strength of evidence across studies was based on precepts outlined by the GRADE Working Group4 and recommendations made by the Agency for Healthcare Research and Quality (AHRQ).2 3

Five comparative studies5 6 7 8 9 met the inclusion criteria (Fig. 1). All compared anterior cervical discectomy and fusion (ACDF) with allograft (one also included autograft)7 in an inpatient versus outpatient setting; two included plate fixation.5 7 Three studies included single- or multi-level surgeries,6 7 8 one study included single-level surgeries only,5 and one study did not report on number of surgical levels9 (Tables 1 and 2; see also Table 5 in the online supplementary material).

Fig. 1

Flowchart showing results of literature search.

Table 1

Detailed characteristics and results of included studies

Author (year)LoE	Study design	Demographics/patient characteristics	Surgical procedure(inpatient and outpatient)	Surgery characteristics	Follow-up (% followed up)	Safety outcomes (n [%])
Liu et al (2009)III	Retrospective cohort study	N = 109Age (mean): NRMale: 51.5% Inpatient ( n  = 64) (58.7%) Age (mean): 56.06 y (range: 22–89)Male: 58% Comorbidities Total major medical comorbidities = 120 (1.875 average per patient)• Hypertension (44%)• High cholesterol (31%)• Arthritis (28%)• Diabetes (20%)• Depression (14%)• Migraines (14%)• Heartburn/reflux (11%)• Obesity (8%)• Cancer (6%)• Stroke (5%)• Hepatitis (3%)• Osteoporosis (3%) Outpatient ( n  = 45) (41.3%) Age (mean): 48.73 y (range: 23–64)Male: 69% Comorbidities Total major medical comorbidities = 63 (1.4 average per patient)• Hypertension (33%)• High cholesterol (20%)• Arthritis (7%)• Diabetes (4%)• Depression (20%)• Migraines (13%)• Heartburn/reflux (20%)• Obesity (4%)• Cancer (2%)• Stroke (2%)• Hepatitis (7%)• Osteoporosis (7%)	ACDF with plate fixation (using structural allograft and titanium plating, one level)	Inpatient ( n  = 64) Preoperative symptoms • Myelopathy (42%)• Radiculopathy (41%)• Combination (14%)• Neck pain (3%)• Previous cervical spine surgery (13%) Levels (n [%]) C5–6: 20 (31%)C6–7: 20 (31%)C4–5: 14 (22%)C3–4: 8 (13%)C7–T1: 2 (3%) Outpatient ( n  = 45) Preoperative symptoms • Myelopathy (24%)• Radiculopathy (60%)• Combination (16%)• Neck pain (0%)• Previous cervical spine surgery (4%) Levels (n [%]) C6–7: 20 (44%)C5–6: 16 (35%)C4–5: 5 (11%)C3–4: 3 (7%)C7–T1: 1 (2%)	Mean: 62.4 d (range: 7–208)F/U: 100% (109/109)	Inpatient ( n  = 64) Complications (n [%]) • Any: 4 (6%)• Symptomatic cerebrospinal fluid leak: 1 (1.6%)• Moderate dysphagia: 1 (1.6%)• Syncope: 1 (1.6%)• Hematoma: 1 (1.6%)• Death: 0 (0%) Outpatient ( n  = 45) Complications (n [%]) • Any: 0 (0%)• Symptomatic cerebrospinal fluid leak: 0 (0%)• Moderate dysphagia: 0 (0%)• Syncope: 0 (0%)• Hematoma: 0 (0%)• Death: 0 (0%)
Silvers et al (1996)III	Cohort study (outpatients prospective, inpatients retrospective)	Outpatient ( n  = 50) Age (mean): 45.9 ± 1.5 y (range: 28–69)Male: 25 (50%) Inpatient ( n  = 53) c Age (mean): NRMale: NR	ACDF (with dowel allograft, one or two levels)	Diagnosis: cervical disc herniation Outpatient ( n  = 50) Preoperative symptoms • Significant neck pain: 80%• Significant arm pain: 74%• Weakness and/or atrophy: 50%• Significant impairment of ability to perform normal activities or work: >80% Surgical data • One level: 31 (62%)  • C3–4 (2%)  • C4–5 (6%)  • C5–6 (38%)  • C6–7 (16%)• Two levels: 19 (38%)  • C4–5 + C5–6 (4%)  • C5–6 + C6–7 (32%)  • C4–5 + C6–7 (2%) Inpatient ( n  = 53) c Preoperative symptoms NR Surgical data • One level: 36 (68%)• Two levels: 17 (32%)Specific levels: NR	F/U: all >1 yearMean F/U:Outpatient: 1.3 yInpatient: 1.6 yOutpatient: 100%Inpatient: 88–100%	Outpatient ( n  = 50) Complications (n [%])• Vocal paralysis: 1 (2%)• Death: 0 (0%)• Admitted for complication:0 (0%) Inpatient ( n  = 53) Complications (n [%])• Infection: 1 (2%)• Death: 0 (0%)
Stieber et al (2005)III	Retrospective cohort study	N = 90Male: 43 (48%) Outpatient ( n  = 30) (33.3%) Age (mean): 43.2 yMale: 18 (60%)BMI: 27.4 Inpatient 1 a ( n  = 30) (33.3%) Age (mean): 43.6 yMale: 13 (43%)BMI: 30.4 Inpatient 2 b ( n  = 30) (33.3%) Age (mean): 44.9 yMale: 12 (40%)BMI: 29.2	ACDF with plate fixation (left-sided anterior approach, with allograph or autograph, one or two levels)	Diagnosis: cervical disc herniation Outpatient ( n  = 30) Preoperative diagnoses • Weakness: 11• Numbness: 17• Arm Pain: 26• Neck Pain: 26• Myelopathy: 0 Surgical data • Autograft: 0 (0%)• Allograft: 30 (100%)• Primary: 30 (100%)• Revision: 0 (0%)• One level: 13 (43%)• Two levels: 17 (57%)• C3–4: 0 (0%)• C4–5: 9 (30%)• C5–6: 22 (73%)• C6–7: 16 (53%)• C7–T1: 0 (0%) Inpatient 1 a ( n  = 30) Preoperative diagnoses • Weakness: 18• Numbness: 24• Arm Pain: 29• Neck Pain: 30• Myelopathy: 4 Surgical data • Autograft: 14 (47%)• Allograft: 16 (53%)• Primary: 27 (90%)• Revision: 3 (10%)• One level: 12 (40%)• Two levels: 18 (60%)• C3–4: 1 (3%)• C4–5: 6 (20%)• C5–6: 25 (83%)• C6–7: 15 (50%)• C7–T1: 1 (3%) Inpatient 2 b ( n  = 30) Preoperative diagnoses • Weakness: 7• Numbness: 12• Arm Pain: 24• Neck Pain: 25• Myelopathy: 3 Surgical data • Autograft: 22 (73%)• Allograft: 8 (27%)• Primary: 27 (90%)• Revision: 3 (10%)• One level: 15 (50%)• Two levels: 15 (50%)• C3–4: 2 (7%)• C4–5: 5 (17%)• C5–6: 20 (67%)• C6–7: 18 (60%)• C7–T1: 0 (0%)	Evaluated on first postoperative day, 7–10 d and 3 wk after surgeryF/U: 96% (86/90)	Outpatient ( n  = 30) • Dysphagia: 3/30 (10%)• Admitted for complication: 0 (0%) All inpatients ( n  = 56) • Any complication: 7 (13%)• Dysphagia: 3 (5%)• Donor site pain: 4 (7%)• Increased LOS: 4 (7%)• Airway swelling: 1 (2%)• Readmitted for complication: 4 (7%)
Trahan et al (1997)III	Retrospective cohort study	N = 117Age (mean): 50.2 yMale: 52.1% Outpatient (n  = 59; 50%) Age (mean): 47.6 y (range: 28–73 y)Male: 36 (61%) Comorbidities 1.8 average per patient Inpatient ( n  = 58; 50%) Age (mean): 52.8 y (range: 23–79 y)Male: 25 (43%) Comorbidities 2.9 average per patient	ACDF (with allograft, one or two levels)	Outpatient (n  = 59) One level: 38/59 (64%) C2–3: 0 (0%)C3–4: 2 (5%)C4–5: 7 (18%)C5–6: 19 (50%)C6–7: 10 (26%) Two levels: 21/59 (36%) C3–5: 0 (0%)C3–4, C5–6: 0 (0%)C4–6: 4 (19%)C5–7: 17 (81%) Inpatient (n  = 58) One level: 30/58 (51%) C2–3: 1 (3%)C3–4: 4 (13%)C4–5: 7 (23%)C5–6: 9 (30%)C6–7: 9 (30%) Two levels: 28/58 (49%) C3–5: 4 (14%)C3–4, C5–6: 1 (4%)C4–6: 5 (18%)C5–7: 18 (64%)	F/U: 6 hours after procedure (100%; 117/117)	Outpatient ( n  = 59) a • Neck swelling and difficulty breathing (required readmission) 1/59 (1.7%) Inpatient ( n  = 58) a • Any complication: 0/58 (0%)
Walid et al (2010)III	Retrospective cohort study	N = 675Age (mean): 53.4 yMale: NR Outpatient (n  = 97; 15%) Age (mean): 49 yMale: NR Comorbidities d • Obesity: 42.9%• DM: 10%• Hypertension: 8.8%• COPD: 13.5%• CHD: 1.3%• CABG: 3.8%• Stroke: 2.5%• Knee arthritis: 7.5%• Hip problems: 1.3%• Antidepressants: 11.6% Inpatient (n  = 578; 85%) Age (mean): 55 yMale: NR Comorbidities d • Obesity: 46.9%• DM: 19%• Hypertension: 9.7%• COPD: 11.8%• CHD: 19.7%• CABG: 15.9%• Stroke: 1.9%• Knee arthritis: 39.8%• Hip problems: 1.7%• Antidepressants: 25.4%	Three surgeries were assessed:• ACDF• Lumbar microdisectomy (LMD)• Lumbar decompression with or without fusion (LDF)	NR	F/U: NR	Outpatient ( n  = 97) • Infection: 1/97 (1%) Inpatient ( n  = 578) • Infection: 16/578 (2.8%)7 patients were readmitted due to infection; all were obese (NR by treatment group)

Abbreviations: ACDF, anterior cervical discectomy and fusion; CABG, history of coronary artery bypass graft, stent or balloon angioplasty; CHD, congestive heart disease; CHF, congestive heart failure; COPD, chronic obstructive pulmonary disease; DM, diabetes mellitus; HT, hypertension; LoE, level of evidence; LOS, length of stay; NR, not reported.

Admitted patients who had surgery before commencement of patient selection for outpatient group.

Admitted patients who had surgery concurrently with outpatient group.

Inpatients were not significantly different (p < 0.05) from outpatients on presenting symptoms, time before initial consultation, sex, age, and operative parameters.

Prevalence of DM, CHD, CABG, knee problems, and depression was significantly higher in the inpatient group, and patients in the inpatient group were significantly older.

Table 2

Summary of safety outcomes in studies comparing outpatient and inpatient surgery in the cervical spine

Author (year)Study designLoE	Diagnosis	Surgical interventionlevels	Hematoma (n [%])	Mortality (n [%])	Other safety outcomes (n [%])
Liu et al(2009)Retrospective cohort studyIII	NR	ACDF with plate fixation (using structural allograft and titanium plating)One level	Outpatient ( n  = 45) 0 (0%) Inpatient ( n  = 64) 1 (1.6%)	Outpatient ( n  = 45) 0 (0%) Inpatient ( n  = 64) 0 (0%)	Outpatient ( n  = 45) • Any complication: 0 (0%)• Symptomatic cerebrospinal fluid leak: 0 (0%)• Moderate dysphagia: 0 (0%)• Syncope: 0 (0%) Inpatient ( n  = 64) • Any: 4 (6%)• Symptomatic cerebrospinal fluid leak: 1 (1.6)• Moderate dysphagia: 1 (1.6)• Syncope: 1 (1.6)
Silvers et al(1996)Retrospective cohort studyIII	Cervical disc herniation	ACDF (with dowel allograft)One or two levels	NR	Outpatient ( n  = 50) 0 (0%) Inpatient ( n  = 53): 0 (0%)	Outpatient ( n  = 50) • Vocal paralysis: 1 (2%)• Admitted for complication: 0 (0%)• Infection 0 (0%) Inpatient ( n  = 53) • Infection: 1 (2%)
Stieber et al(2005)Retrospective cohort studyIII	Cervical disc herniation	ACDF with plate fixation (left-sided anterior approach, with allograph or autograph)One or two levels	NR	NR	Outpatient ( n  = 30) • Dysphagia: 3 (10%)• Admitted for complication: 0 (0%) Inpatient ( n  = 56) • Any complication: 7 (13%)• Dysphagia: 3 (5%)• Donor site pain: 4 (7%)• Increased LOS: 4 (7%)• Airway swelling: 1 (2%)• Readmitted for complication: 4 (7%)
Trahan et al(1997)Retrospective cohort studyIII	NR	ACDF (with allograft)One or two levels	NR	NR	Outpatient ( n  = 59) • Neck swelling and difficulty breathing (required readmission) 1/59 (1.7%) Inpatient ( n  = 58) • Any complication: 0/58 (0%)
Walid et al(2010)Retrospective cohort studyIII	NR	• ACDF• Lumbar microdisectomy (LMD)• Lumbar decompression with or without fusion (LDF)Levels: NR	NR	NR	Outpatient ( n  = 97) • Infection: 1/97 (1%) Inpatient ( n  = 578) • Infection: 16/578 (2.8%)7 patients were readmitted due to infection; all were obese (NR by treatment group)

Abbreviations: ACDF, anterior cervical discectomy and fusion; CABG, history of coronary artery bypass graft, stent or balloon angioplasty; CHD, congestive heart disease; CHF, congestive heart failure; COPD, chronic obstructive pulmonary disease; DM, diabetes mellitus; HT, hypertension; LOS, length of stay; NR, not reported.

Overall, a total of 281 outpatients and 813 inpatients were represented in the included studies, with approximately the same number of males and females ranging in age from 22 to 89 years. Detailed characteristics of the included studies are provided in Table 1.

All five studies were retrospective cohort studies determined to be of moderately high risk of bias (Class of Evidence [CoE] III). Further details on the CoE rating for these studies can be found in the online supplementary material in Table 2a.

In one study,5 hematoma occurred in zero patients in the outpatient group and in one patient (1.6%) in the inpatient group. The other four studies did not report on risk of hematoma.

Two studies5 6 reported on mortality. Both studies reported zero deaths in both the inpatient and outpatient groups.

Two studies reported data on dysphagia, with one study7 reporting higher risk than the other.5 One study5 reported low risk of moderate dysphagia (0 and 1.6%) in the outpatient and inpatient groups, respectively. The other7 reported higher risk, with 10% of outpatients and 5% of inpatients experiencing this complication.

Two studies reported on postoperative infection risk. One study9 reported that 16 patients (2.8%) who underwent surgery as inpatients developed infection, compared with 1 patient (1%) in the outpatient group. The other6 reported that one patient (2%) in the inpatient group experienced infection following surgery.

Three studies reported data on patients who needed to be readmitted following surgery due to a complication. All three studies reported data for outpatients: two studies6 7 reported that zero outpatients were readmitted and one study8 reported that one outpatient was readmitted (due to neck swelling and difficulty breathing). One study reported data for inpatients and found that four patients (7%) were readmitted for a complication.7

Data on additional safety outcomes are provided in Table 2, and a summary of the strength of evidence is provided in Table 3 (additional details on the strength of evidence are located in Table 4 of the online supplementary material).

Table 3

Evidence Summary

Outcome	Strength of evidence	Conclusions/comments
Hematoma		One small LoE III study reported hematoma in one patient (1.6%) in the inpatient group compared to zero patients (0%) in the outpatient group.
Mortality		Two small LoE III studies reported zero deaths in both the inpatient and outpatient groups.
Dysphagia		Two small LoE II studies reported on dysphagia, ranging from 0–10% of outpatients and 1.6–5% of inpatients.
Infection		Two small LoE III studies reported similar risks of infection in the inpatient group compared with the outpatient group (2.8% versus 1% in one study and 2% versus 0% in the other, respectively).
Other complications		All five LoE III studies reported on risks of other complications, including vocal paralysis, syncope, CSF leak, and airway swelling, but there was no overlap in these additional outcomes across studies. The complication risks were very low in both inpatient and outpatient groups, with the exception of hospital re-admission, which in a single study was higher in the inpatient group (7%) compared with the outpatient group (0%). All studies had small sample sizes (∼50 patients per treatment group), except one study of 97 outpatients and 578 inpatients.

Abbreviations: CSF, cerebral spinal leak; LoE, level of evidence.

Clinical Guidelines

None found.

The overall strength of evidence for safety of outpatient surgery in the cervical spine is insufficient for all outcomes examined, that is, the limited evidence available prevents any definitive conclusions. There are few studies that have been done, and they are of poor quality, indicating a need for further well-designed, prospective studies (Table 3). Details on the methodology for determining the strength of evidence, criteria used for upgrading and downgrading, and a more detailed evidence summary table are provided in the online supplementary material.

Illustrative Case

A 50-year-old woman presented with severe cervical myelopathy. She did have significant spondylosis diffusely along her spinal column, but only one site of cord compression (Figs. 2 and 3). She underwent a single-level anterior cervical discectomy and fusion. Meticulous hemostasis was obtained and the wound was closed over a drain. On the evening of surgery, she became unresponsive and the emergency response team had difficulty establishing an airway. Fullness was noted at the incision site and the resident physician on call emergently opened the incision at the bedside and evacuated the suspected hematoma. The airway was secured and the patient was safely ventilated. The patient was returned to the operating theater where the surgical site was examined. No obvious source of bleeding was identified. The wound was irrigated and hemostasis was re-achieved and the wound was closed over a drain. The patient subsequently recovered without hypoxic or neurologic deficit.

Fig. 2

T2 sagittal view of severe C4-5 stenosis.

Fig. 3

T2 axial view of severe C4-5 stenosis.

The overall strength of evidence for the primary outcomes is insufficient, meaning that we have very little confidence in the effect estimates, making it difficult to draw firm conclusions (Table 3).

Patient characteristics (e.g., comorbidities, operative indications, severity of condition) were generally different between the inpatient and outpatient groups so selection bias and confounding bias are important limitations of studies available for this systematic review.

There are five comparative studies that address safety of inpatient versus outpatient surgeries; however, four have small sample sizes and only a small number have overlapping outcomes (hematoma [one study], mortality [two studies], dysphagia [two studies], infection [two studies], and hospital readmission [three studies]). Small sample sizes may preclude detection of rare events and demonstration of statistical difference between groups.

Further study is required to more clearly define the role of outpatient cervical spine surgery. In addition, acceptable thresholds for complication have yet to be defined for outpatient cervical spine surgery. Acceptable thresholds for a complication like dysphagia are not comparable to acceptable thresholds for complications such as airway compromise, hypoxia, and death. Finally, clear identification and quantification of risk factors for complication will be useful in determining which patients may potentially be candidates for outpatient cervical spine surgery.