The Outcome of Using Closed Suction Wound Drains in Patients Undergoing Lumbar Spine Surgery: A Systematic Review.

Study Design Systematic review. Objective Determine whether closed suction wound drains decrease the incidence of postoperative complications compared with no drain use in patients undergoing spine surgery for lumbar degenerative conditions. Methods Electronic databases and reference lists of key articles were searched up through January 22, 2015, to identify studies comparing the use of closed suction wound drains with no drains in spine surgery for lumbar degenerative conditions. Outcomes assessed included the cumulative incidence of epidural hematoma, superficial and deep wound infection, and postoperative blood transfusion. The overall strength of evidence across studies was based on precepts outlined by the Grades of Recommendation Assessment, Development and Evaluation Working Group. Results Five heterogeneous studies, three randomized controlled trials, and two cohort studies form the evidence basis for this report. There was no difference in the incidence of hematoma, superficial wound infection, or deep infection in patients with compared with patients without closed suction wound drains after lumbar surgery. The upper bounds of the 95% confidence interval for hematoma ranged from 1.1 to 16.7%; for superficial infection, 1.0 to 7.3%; and for deep infection, 1.0 to 7.1%. One observational study reported a 3.5-fold increase in the risk of blood transfusion in patients with a drain. The overall strength of evidence for these findings is considered low or insufficient. Conclusions Conclusions from this systematic review are limited by the quality of included studies that assessed the use of closed suction wound drains in lumbar spine surgeries for degenerative conditions. We believe that spine surgeons should not routinely rely on closed suction wound drains in lumbar spine surgery until a higher level of evidence becomes available to support its use.

Study Rationale and Context

Wound suction drains have been used to decrease the rate of postoperative hematoma formation and thus wound infections for many years throughout all surgical subspecialties. Although the use of surgical drains dates back to the years of Hippocrates,1 in the orthopedic literature these drains have not been shown to be beneficial in decreasing the rates of these complications, especially in orthopedic procedures including fracture fixation or arthroplasty surgeries.2 3 4 5 However, these drains are still commonly used throughout the orthopedic community, including spine surgery.

Debate in this area remains, as proponents of its use in the immediate postoperative period believe that it will prevent fluid collection in the surgical dead space and thus eliminate the media for bacterial growth. On the other hand, opponents believe that they are considered a foreign body that promotes inflammation and even sometimes provokes an infectious response.

In spine surgery, the controversy is even more profound because it decreases the rare but devastating complication of postoperative epidural hematoma, but it may have a hypothetical increase in the risk of infection.6 7 In the spine literature, the incidence of epidural hematoma in the postprocedural period ranges from 0.2 to 2.9%,8 9 10 and the incidence of postoperative wound infection is 0.7 to 16%,11 12 thus identifying a method that can decrease the incidence of these complications would be of great benefit.

To help address this debate, we conducted a systematic review of the use of these drains and extensively explored the efficacy and safety of closed suction wound drainage in spine surgery in the postoperative period.

Clinical Question

In patients undergoing spine surgery for lumbar degenerative conditions, does the use of closed suction wound drains decrease the incidence of postoperative complications compared with no drain use?

Materials and Methods

Study design: Systematic review.

Search: PubMed, Embase, the Cochrane Library, Web of Science, Scopus, and bibliographies of key articles

Dates searched: through January 22, 2015.

Inclusion criteria: (1) comparative studies in peer-reviewed journals; (2) adult patients undergoing spinal surgery for degenerative conditions in the lumbar spine receiving postoperative closed suction drains or no drains; (3) outcomes included at least one of the following: epidural hematoma, superficial wound infection, deep wound infection, or postoperative blood transfusion.

Exclusion criteria: (1) skeletally immature patients (<18 years of age); (2) surgery for intradural pathology, dural tears, tumor, trauma, fracture, or infection; (3) cervical or thoracic spine surgery; (4) nonclinical studies, case reports, and case series.

Outcomes: (1) epidural hematoma; (2) superficial wound infection; (3) deep wound infection; and (4) postoperative blood transfusion.

Analysis: Due to heterogeneity in patient populations (including differences in patient demographics, diagnoses, and surgical procedures) and differences in study design, a meta-analysis was not performed. We calculated the cumulative incidence by dividing the number of patients with an event by the number at risk for the event and the associated 95% confidence interval. When zero events were reported for an outcome, the confidence interval was found using the “rule of three” estimation.13

Details about the methods can be found in the online supplementary material.

Overall strength of evidence: The overall strength of evidence across studies was based on precepts outlined by the Grades of Recommendation Assessment, Development and Evaluation Working Group. Study critical appraisals and reasons for upgrading and downgrading for each outcome can be found in the online supplementary material.

Three randomized controlled trials (RCTs) and two cohort studies met the inclusion criteria and form the basis for this report (Fig. 1). Their characteristics are described in Table 1. A list of studies excluded and the reason for exclusion can be found in the online supplementary material.

Fig. 1

Flowchart showing results of literature search.

Table 1

Characteristics of included studies

First author (year)	Study type	Population	Treatment	Diagnosis	Surgery type
Payne (1996)7	RCT	n = 200Age: NR% male: NR	Drain (n = 103)No drain (n = 97)	• Herniated disk (% NR)• Degenerative stenosis (% NR)	• Singe-level hemilaminectomy (% NR)• Decompressive laminectomy (% NR)
Brown (2004)14	RCT	n = 83Age: 67.4 y% male: NR	Drain (n = 42)No drain (n = 41)	• Herniated nucleus pulposus: n = 13 (16%)• Spinal stenosis: n = 73 (88%)• Degenerative spondylolisthesis: n = 40 (48%)• Postlaminectomy syndrome: n = 15 (18%)	• Decompression (2–4 levels)Primary: n = 15 (35%)Reoperation: n = 12 (15%)• Fusion with instrumentation: n = 16 (19%)• Fusion without instrumentation: n = 30 (36%)
Mirzai (2006)15	RCT	n = 50Age: 46.7 ± 7.6 y58% male	Drain (n = 22)No drain (n = 28)	• Herniated disk: n = 50 (100%)	• Virgin single-level microdiskectomy with hemipartial laminectomy and flavectomy
Kanayama (2010)16	Retro cohort	n = 560Age: 46 y64.3% male	Drain (n = 298)No drain (n = 262)	NR	• Single-level laminoplasty (medial facetectomy) (% NR)• Diskectomy (% NR)
Walid (2012)17	Retro cohort	n = 402Age: 57.3 ± 11.5 y43% male	Drain (n = 285)No drain (n = 117)	• Spondylosis: n = 31 (8%)• Disk displacement: n = 145 (36%)• Disk degeneration: n = 84 (21%)• Disk disease with myelopathy: n = 23 (6%)• Lumbar spinal stenosis: n = 79 (20%)• Acquired spondylolisthesis: n = 29 (7%)• Spondylolisthesis: n = 11 (3%)	• Lateral fusion: n = 301 (75%)• Posterior fusion: n = 101 (25%)

Abbreviations: NR, not reported; RCT, randomized controlled trial.

One RCT included patients undergoing multilevel decompression or fusion as a primary or reoperation procedure for herniated nucleus pulposus, spinal stenosis, degenerative spondylolisthesis, or postlaminectomy syndrome in an older population (mean age 67.4 years; Table 1).14 Two RCTs included patients undergoing single-level laminectomy for herniated disk or degenerative stenosis, though one study placed the drains in the epidural space (mean age 46.7 years),15 and the second placed the drain in the lumbodorsal fascia (mean age not reported).7

Two retrospective cohorts included patients undergoing single or multilevel laminoplasty, diskectomy, or fusion for varying diagnoses (mean age 46 to 57.3 years).16 17

The evidence base included three RCTs and two retrospective cohorts.

The method of identification of hematoma varied among studies. One small RCT reported 89% of the patients with drains and 36% of the patients without drains had epidural hematomas as detected by magnetic resonance imaging. Most were minimal in size with only 7% prominent in the no-drain group compared with 0% prominent in the drain group.15 All other studies assessed hematomas that required drainage or reoperation. The risk in these studies ranged from 0.0 to 0.7%.7 14 16 17

There was no difference in risk for hematoma comparing drains with no drains across studies of different design, surgeries of varying complexity, patients of different ages, or complexity of diagnosis.7 14 15 16 17 The upper 95% confident limit of risk for hematoma in the patients with drains was 13.6% compared with 16.7% for the patients without drains (Table 2).

Table 2

Hematoma, superficial wound infection, deep infection, and postoperative blood transfusion among patients receiving closed wound drains compared with those who do not receive closed wound drains following lumbar surgery

First author (year)	Treatment	Hematoma		Superficial wound infection		Deep infection		Postoperative blood transfusion
		n	% (95% CI)	n	% (95% CI)	n	% (95% CI)	n	% (95% CI)
Payne (1996)7	Drain (n = 103)No drain (n = 97)	0a 0a	0.0 (0.0–2.9)0.0 (0.0–3.1)	21	1.9 (0.0–4.6)1.0 (0.0–3.0)	NR	NR	NR	NR
Brown (2004)14	Drain (n = 42)No drain (n = 41)	00	0.0 (0.0–7.1)0.0 (0.0–7.3)	00	0.0 (0.0–7.1)0.0 (0.0–7.3)	0/420/41	0.0 (0.0–7.1)0.0 (0.0–7.3)	NR	NR
Mirzai (2006)15	Drain (n = 22)No drain (n = 28)	0b 2b	0.0 (0.0–13.6)7.1 (0.0–16.7)	NR	NR	NR	NR	NR	NR
Kanayama (2010)16	Drain (n = 298)No drain (n = 262)	2c 0c	0.7 (0.0–1.6)0.0 (0.0–1.1)	00	0.0 (0.0–1.0)0.0 (0.0–1.1)	00	0.0 (0.0–1.0)0.0 (0.0–1.1)	NR	NR
Walid (2012)17	Drain (n = 285)No drain (n = 117)	0d 0d	0.0 (0.0–1.1)0.0 (0.0–2.6)	103	3.5 (1.4–5.7)2.6 (0.0–5.4)	NR	NR	688	23.9 (18.9–28.8)6.8 (2.3–11.4)

Abbreviations: CI, confidence interval; FU, follow-up; NR, not reported.

Reported as hematoma requiring drainage.

“Prominent” hematoma detected with magnetic resonance imaging. Minimal hematoma was found in 31.8% of patients with drains versus 60.7% without drains, and moderate in 4.5% patients with drains versus 21.4% without drains.

Reported as epidural hematoma causing neurologic compromise requiring surgical evacuation.

Reported as hematoma requiring reoperation.

The evidence base included two RCTs and two retrospective cohorts.

There was no difference in risk between the patients receiving closed suction wound drains (range of risks, 0.0 to 3.5%) compared with the patients with no drains (range of risks, 0.0 to 2.6%; Table 2).7 14 16 17 The upper 95% confident limit of risk for superficial wound infection in the patients with drains was 7.1% and for the patients without drains was 7.3%.

The evidence base included one RCT and one retrospective cohort.

There are no events recorded in the two studies that report risk for deep infection in the patients treated with versus patients treated without closed suction wound drains after lumbar surgery (Table 2).14 16 The upper limit of risk for deep infection in the patients with drains was 7.1% and in the patients without drains was 7.3%.

The evidence base included one retrospective cohort.

One observation study reported a greater risk of blood transfusion after lumbar surgery in the patients with drains (24%) versus patients without drains (7%; risk ratio 3.5; 95% confidence interval 1.7 to 7.0).17

Evidence Summary

There was no difference in the incidence of hematoma, superficial wound infection, or deep infection in the patients with versus patients without closed suction wound drains after lumbar surgery. The overall strength of evidence for these findings is considered low or insufficient (Table 3). There is insufficient evidence to draw conclusions on the effect of closed suction wound drains on the risk for postoperative blood transfusion.

Table 3

Evidence summary: Do closed suction wound drains decrease the incidence of postoperative complications?

Outcome	Overall quality of evidence	Studies	Effect size
			Range of upper bound of 95% confidence intervala
Hematoma	Low due to risk of bias and imprecision	3 RCTs (n = 333)	Drain: 2.9–13.6%No drain: 3.1–16.7%
		2 retro cohorts (n = 962)	Drain: 1.1–1.6%No drain: 1.1–2.6%
Superficial wound infection	Insufficient due to risk of bias and imprecision	2 RCTs (n = 283)	Drain: 4.6–7.1%No drain: 3.0–7.3%
		2 retro cohorts (n = 962)	Drain: 1.0–5.7%No drain: 1.1–5.4%
Deep infection	Insufficient due to risk of bias and imprecision	1 RCT (n = 83)	Drain: 7.1%No drain: 7.3%
		1 retro cohort (n = 560)	Drain: 1.0%No drain: 1.1%
Postoperative blood transfusion	Insufficient due to risk of bias	1 retro cohort (n = 402)	Drain: 28.8%No drain: 11.4%RR: 3.5 (1.7, 7.0)

Abbreviations: RCT, randomized controlled trial; retro, retrospective cohort; RR, risk ratio.

Calculated using Hanley's rule of three when zero events are reported for a given outcome.

Clinical Guidelines

No clinical guidelines were found.

In the setting of spinal surgery, one of the biggest concerns is an epidural hematoma; some spine surgeons use drains to decrease the incidence of postoperative hematomas and subsequently neurologic complications. However, there is no clear consensus whether is it beneficial by decreasing hematomas or harmful by increasing the risk of infection and the rate of blood transfusion.

Five studies assessing the use of closed suction wound drains in spine surgeries for lumbar degenerative conditions were identified and included in this systematic review.

One strength of our study was the thorough and systematic search of the evidence performed. The limitations included the following: the methodologies of studies included in this review were heterogeneous, and all studies had small sample sizes. There was a lack of random sequence generation or allocation concealment in the RCTs. In addition, there was a lack of observer blinding and lack of controlling for confounding factors. (See Table 3 in the online supplementary material.)

Contrary to previous beliefs, there is no difference in the incidence of hematoma, superficial wound infection, or deep infection in the patients with versus patients without closed suction wound drains after lumbar surgery.

As this systematic review was of low or insufficient evidence, more randomized trials of the use of closed suction wound drains in lumbar spine surgery are needed with sufficient sample sizes and appropriate methodology protocols to adequately assess the effect of drain use on these rare outcomes.

Conclusions

Conclusions from this systematic review are limited by the quality of included studies that assessed the use of closed suction wound drains in lumbar spine surgeries for degenerative conditions. The small sample sizes, lack of random sequence generation, or allocation concealment in RCTs and lack of controlling for confounding factors may not have allowed us to detect differences between treatment groups, particularly for rare outcomes. We believe that spine surgeons should not routinely rely on the use of closed suction wound drains in lumbar spine surgery until a higher level of evidence becomes available to support its use.