Postoperative ischemic optic neuropathy (ION) is a devastating complication that can occur after a variety of procedures including cardiac surgery, spine surgery, head and neck surgery, prostatectomy, liver transplantation, major vascular surgery, liposuction, and other miscellaneous procedures. The highest reported incidence of ION after spine surgery is 1:1000.1 Patients who may already be debilitated from spinal disease can develop complete blindness in both eyes, greatly compounding their disability. Because of its low incidence, existing data on this complication are retrospective, making identification of risk factors problematic. Suggested risk factors from case reports and case series are unsupported and numerous, and include prolonged surgery in the prone position, venous congestion of the head, large blood loss with large fluid resuscitation, hypotension, anemia, use of vasopressors, emboli, and susceptible patient vascular anatomy and physiology.1–7 Identifying patient or perioperative anesthetic and surgical predictors for postoperative ION might lead to therapeutic modifications designed to minimize its occurrence.
Postoperative ION should be distinguished from other types of postoperative visual loss (POVL) including central retinal artery occlusion and cortical blindness. When patients complain of POVL, urgent ophthalmologic examination with dilated fundoscopic examination is recommended to rule out extremely rare, but potentially treatable conditions such as direct globe injury, acute angle closure glaucoma, and retinal detachment. In addition, fundoscopic examination will also diagnose untreatable retinal vascular disorders such as central retinal artery occlusion (CRAO) related to emboli or sustained ocular compression. If an obvious ocular cause is not found, neuroimaging with computed tomography and/or magnetic resonance imaging of the head is helpful in ruling out occipital infarction and pituitary apoplexy.6 Other ancillary tests such as visual evoked potentials, electroretinograms, transcranial Doppler to detect embolic phenomena, and carotid duplex may provide further confirmation or exclusion of specific diagnoses.
ION can be divided into anterior ION (AION) and posterior ION (PION). AION is an injury to the optic nerve that occurs anterior to the lamina cribrosa, a semirigid, sieve-like piece of connective tissue through which the optic nerve and central retinal vessels pass just before their entry into the globe (Figure 1).6–8a AION is further divided into arteritic and nonarteritic forms. Arteritic AION is caused by temporal arteritis, and is almost never found perioperatively. Nonarteritic AION more commonly occurs spontaneously in the community, typically in patients with coexisting vascular disease, though it can occur in patients who are otherwise healthy. It is also the type of AION that can occur perioperatively. Perioperative nonarteritic AION occurs primarily after cardiac bypass surgery and spine surgery but has also been reported after prostatectomies, major vascular procedures, liposuction, and miscellaneous procedures.6,7 It has been associated with vascular disease, hypotension, and anemia in case reports and case series, though these factors may be confounders related to the patient population undergoing these procedures. After surgery, patients may present with AION immediately on awakening, but often may have an initial period of normal vision lasting several days and then suddenly experience an abrupt decrement in their vision that progresses over a few days before stabilizing.6,7 Pupillary examination reveals poorly reactive pupils and a relative afferent pupillary defect if unilateral or asymmetric. More often than not, both eyes will be involved. Visual field deficits include altitudinal field defects, central scotomas, or complete blindness with no light perception. Fundoscopic examination reveals a swollen optic disc with or without peripapillary flame-shaped hemorrhages or splinter hemorrhages at the optic disc margin (Figure 2B).6 Weeks later, the swelling starts to subside, peripheral hemorrhage is reabsorbed, and the optic disc becomes pale (Figure 2C). At this point, AION and PION may be indistinguishable as PION results in late optic nerve pallor as well. Postoperative AION has poor recovery of vision and there is no known beneficial treatment.
PION occurs posterior to the lamina cribrosa (Figure 1) and can also be arteritic or nonarteritic. The postoperative nonarteritic form of PION typically manifests on awakening from anesthesia. PION is the most common cause of POVL after procedures with venous congestion of the head including prone spine surgery, bilateral radical head and neck procedures, and other procedures.6,7 It also occurs in association with cardiac bypass procedures.6 Its spontaneous occurrence in the community is much less common than AION. Both eyes are affected in at least two thirds of reported perioperative cases.4 Visual loss from PION is usually more severe than that from AION, but it does not typically worsen beyond the initial presentation.2 Similar to AION, pupillary examination demonstrates poor reactions to light and a relative afferent pupillary defect if unilateral or asymmetric. Visual field deficits include central scotomas, altitudinal field defects, or no light perception. Fundoscopic examination initially reveals a completely normal appearing fundus (Figure 2A), with optic nerve pallor and atrophy ensuing only after 4 to 6 weeks (Figure 2C).6,7 Recovery is poor and there is no demonstrated beneficial treatment.
Two other common causes of persistent POVL that are important to distinguish from ION are central retinal artery occlusion (CRAO) and cortical (or cerebral) blindness.6 CRAO that occurs perioperatively is most commonly unilateral with severe visual loss in the affected eye. Presentation is almost always on awakening from anesthesia after cardiac surgery, prone spine surgery, and occasionally other procedures. Periorbital trauma is commonly associated with CRAO after prone spine operations and likely indicates the mechanism of globe compression.4,6 The pupillary light reflex will be absent or sluggish, and there will be a relative afferent pupillary defect. Fundoscopic examination demonstrates the classic findings of retinal whitening with attenuated retinal vessels and a pathognomonic cherry red spot at the macula (Figure 2D). This cherry red spot is caused by the fact that the macula is supplied by the choriocapillaris which continues to perfuse this area after the central retinal artery has become occluded.6,7 Recovery is poor and there is no known beneficial treatment. Four to 6 weeks later, a pale optic disc will be seen as the death of the inner retinal layers will cause antegrade degeneration of the optic nerve. The mechanism of postoperative CRAO may be either embolic (especially in the setting of cardiac bypass) or because of sustained ocular globe compression with consequent elevation of intraocular pressure and occlusion of the inner retinal circulation. Hollenhorst et al first described this complication in 1954 in 8 patients after prone spine surgery on a horseshoe headrest.8 This scenario has subsequently been referred to as the “headrest syndrome.”
Cortical blindness usually presents on awakening from procedures with a high risk of emboli (such as cardiac bypass procedures) to the bilateral posterior cerebral artery territory, or when there has been profound hypotension and bilateral watershed infarctions.6 Theoretically, cortical blindness could occur in spinal instrumentation cases in which numerous emboli are released during pedicle screw placement. Pupillary light reflexes will always be normal as will the fundoscopic examination. If unilateral, the patient will have a contralateral homonymous hemianopsia. If bilateral, the patient may be completely blind or retain only a small island of central vision. Head CT or MRI will demonstrate occipital infarctions in the posterior cerebral artery territory, or parieto-occipital infarctions in a watershed distribution (typically bilateral). Cerebral infarctions can improve, usually within the first few weeks, but the prognosis for complete visual recovery is poor.
The purpose of this review is to attempt to answer the following 2 clinical questions:
- Are there predictors of postoperative ischemic optic neuropathy (ION) associated with major spine surgery?
- Can we recommend preventative measures regarding postoperative ION?
Materials and Methods
Electronic Literature Database
The literature search is outlined in detail elsewhere.8b Briefly, we conducted a systematic search in MEDLINE, EMBASE and the Cochrane Collaboration Library for literature published from 1990 through December 2008 reporting on ION following spine surgery. We limited our results to humans and to articles published in the English language. Reference lists of key articles were also systematically checked. We excluded studies evaluating cases of postoperative visual loss that did not include cases of ischemic optic neuropathy in spine patients. Because ION is a rare event, we included case reports describing this complication.
Each retrieved citation was reviewed by 2 independently working reviewers (J.R.D and N.J.D). Most articles were excluded on the basis of information provided by the title or abstract. Citations that appeared to be appropriate or those that could not be excluded unequivocally from the title and abstract were identified, and the corresponding full text reports were reviewed by the 2 reviewers. Any disagreement between them was resolved by consensus. From the included articles, data on study design, patient demographics, and perioperative anesthetic and surgical characteristics were collected and analyzed.
Level of evidence ratings were assigned to each article independently by 2 reviewers using criteria set by The Journal of Bone and Joint Surgery American Volume (J Bone Joint Surg Am)9 for therapeutic studies and modified to delineate criteria associated with methodologic quality and described elsewhere.8b
Data from case reports are displayed for each patient found in the literature. We summarized case report data by pooling information from each case when present. Qualitative analysis was performed considering the following 3 domains: quality of studies (level of evidence), quantity of studies (the number of published studies similar in patient population, condition treated and outcome assessed), and consistency of results across studies (whether the results of the different studies lead to a similar conclusion).10 We judged whether the body of literature represented a minimum standard for each of the 3 domains using the following criteria: for study quality, at least 80% of the studies reported needed to be rated as a level of evidence I or II; for study quantity, at least 3 published studies were needed which were adequately powered to answer the study question; for study consistency, at least 70% of the studies had to have consistent results. The overall strength of the body of literature was expressed in terms of the impact that further research may have on the results. An overall strength of “HIGH” means that further research is very unlikely to change our confidence in the estimate of effect. The overall strength of “MODERATE” is interpreted as further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate. A grade of “LOW” means that further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate, whereas “VERY LOW” means that any estimate of effect is very uncertain.8a
We identified 360 articles from our literature search potentially evaluating ION following spine surgery. From these potential articles, we judged 28 to undergo full text review. After full text review, we excluded 9 of the articles for the following reasons: 5 articles were review articles without any reporting of primary data or cases, 1 was a case report of non-ION perioperative vision loss, 1 was a case-control study of postoperative vision loss of which an unknown proportion were non-ION cases, 1 was a retrospective case series without any ION cases associated with spine surgery, and 1 was a case study with contradictory results. Of the remaining 19 articles, 14 are case reports11–24 that provide descriptive information on ION cases and are included in this review. Five articles are retrospective observational studies.1,2,4,5,25 Four of the 5 are retrospective case series that provide descriptive data for ION cases that were graded as a level of evidence IV. Descriptive data for the 14 case reports and 3 of the 4 retrospective case series are summarized in Table 1,1,2,11–25 whereas Table 2 contains descriptive data for the largest retrospective series of patients with ION associated with spine surgery.4 Overlap between these 2 tables could not be ruled out. One of the 5 retrospective observational studies was a retrospective cohort study assessing factors associated with ION5 that was graded level of evidence III. We retained one article that is a practice advisory from the American Society of Anesthesiologists Task Force on Perioperative Blindness.26 The selection process is summarized in Figure 3.
Question 1. Are There Predictors of ION Following Spine Surgery?
Factors thought to be associated with postoperative ION are primarily obtained from case report data and include intraoperative anemia, hypotension, long duration surgery, type and amount of intraoperative fluid administration, the prone position, and blood loss, Table 3. The largest case series of ION found that 94% of ION cases after spine surgery had an anesthetic duration of 6 or more hours and 82% had an estimated blood loss of 1 or more liters (Figures 4 and 5).4 Consistent with these data, 85% of cases in our literature review had a surgical duration of 5 or more hours, and 83% had an estimated blood loss of 1 or more liters. One retrospective cohort study evaluated 271 patients with ION following spine surgery.5 In that study, patients 45 to 64 years of age were 3.7 times more likely than patients 18 to 44 years of age to develop ION after spine surgery. Other factors associated with ION following multivariate analysis included peripheral vascular disease (OR = 6.4; 95% CI = 2.2–18.6), diabetes (OR = 2.3; 95% CI = 1.1–4.8), hypertension (OR = 1.7; 95% CI = 0.99–2.9), anemia (OR = 5.9; 95% CI = 3.2–11.1), blood transfusion (OR = 4.3; 95% CI = 1.7–10.8) and hypotension (OR = 10.1; 95% CI = 2.9–35.8). This study was based on data taken from the Nationwide Inpatient Sample (NIS) administrative database and is limited by the inconsistent data collection discussed below.
Question 2. Can We Recommend Preventative Measures Regarding Postoperative Visual Loss?
Current recommendations for preventing ION following spine surgery lack evidence from quality interventional trials because of the low incidence of this complication. Unsupported suggestions include minimizing blood loss, positioning the head equal or above the heart level to decrease venous congestion of the head, transfusing to maintain HCT ≥ 30%, minimizing large crystalloid infusions by using both colloid and crystalloid, staging very prolonged surgeries, maintaining MAP closer to patient's baseline, and avoiding hypotension (Table 3).3,11,15,16,20,26,27 Recommendations from The American Society of Anesthesiologists Task Force on perioperative blindness include using direct continuous blood pressure monitoring, using colloids along with crystalloids to maintain intravascular volume, positioning patients with their head at the level or higher than the heart with the head in a neutral forward position, and consider staging spine procedures in high-risk patients.27 The task force defined high-risk patients as those patients anticipated before surgery to undergo procedures that are prolonged, have substantial blood loss, or both, and further recommended considering informing these patients of the small and unpredictable risk of perioperative visual loss.
The overall strength of evidence to identify predictors of postoperative ischemic optic neuropathy is “Very Low,” that is, any estimate of effect is very uncertain. The overall strength of evidence for current recommended preventative measures is also “Very Low,” Table 4.
This review is limited by the quality and quantity of research on postoperative ION following spine surgery. Articles on this rare adverse event are limited to many case studies, 4 retrospective case series, and 1 retrospective cohort study taken from an administrative database. Data from the American Society of Anesthesiologists' (ASA) Postoperative Visual Loss (POVL) Registry contains the largest series of these patients (n = 83) yet reported and provides a profile of their perioperative characteristics.4 Ninety-four percent of patients in this database with ION after spine surgery had an anesthetic duration in the prone position of more than 6 hours and 82% of patients with a postoperative diagnosis of ION had an estimated blood loss larger than 1000 mL (Figures 4 and 5). Consistent with these findings, 85% of cases in our literature review had a surgical duration of 5 or more hours, and 83% had an estimated blood loss of 1 or more liters. Because these studies lack a control group, specific risk factors can not be identified outside of major spine surgery. However, the majority of major spine surgery patients do not develop ION, suggesting there may be other contributory risk factors.
Variability in patient optic nerve vascular anatomy or physiology may account for unusual complications under similar perioperative conditions. For example, in the majority of the population, the ophthalmic artery originates from the internal carotid artery and is considered an intracranial vessel. However, the ophthalmic artery arises from the middle meningeal artery in 2%–3% of patients, and another 2%–3% of patients may have a very enlarged anastomosis between these 2 vessels.28 This contribution from the extracranial circulation may impair the ability of the ophthalmic artery to autoregulate, or maintain a constant blood flow over a wide range of perfusion pressures, as occurs in the brain. Pillunat has shown that 2 of 10 healthy volunteers were unable to autoregulate the blood flow to the anterior optic nerve head when the intraocular pressures were elevated to values typically seen in prolonged spine surgery.29 Moreover, Drummond has shown that there is significant variation in the autoregulatory capacity of the brain with the lower limit of autoregulation (i.e., the lowest blood pressure at which the brain is able to maintain stable blood flow) varying from a mean arterial pressure of 31 mm Hg to 113 mm Hg in different studies.30
Outflow obstruction of the head and neck has been suggested as a more significant contributory cause of ION than inflow problems, particularly for PION.4 The prone position is associated with elevated central venous pressure caused by increased intrabdominal and intrathoracic pressure. Other procedures associated with the development of ION with elevated venous pressure in the head are bilateral radical head and neck surgery, where bilateral external and internal jugular veins are ligated, and radical prostatectomy in the head down position.31–34 Some investigators postulate that perioperative ION occurring after spine surgery in the prone position is the result of a compartment syndrome of the optic nerve that primarily develops because of elevated venous pressures with subsequent edema formation.4 It has been speculated that perfusion and oxygen delivery may be further compromised by relative hypotension during general anesthesia, severe anemia, prolonged use of vasopressors, and/or hypovolemia from blood loss, and decreased venous return in the prone position.
The type of surgical headrest has also been suggested to contribute to the development of ION. Approximately half of the ION cases associated with spine surgery in the ASA POVL Registry used soft foam cushions with cutouts for the eye, nose and mouth, whereas 19% had their heads suspended in Mayfield pins.4 These data clearly demonstrate the occurrence of ION without globe compression. Moreover, it has been suggested that specific surgical spinal frames promote venous congestion in the head and may predispose to the development of ION. Without denominator data, it is unclear if the type of headrest or surgical frame is associated with the development of ION.
One retrospective cohort study reported on risk factors associated with postoperative ION; however, the data for this study were taken from an administrative database.5 In general, administrative databases contain data that have been gathered for a process other than research. Data are collected by many individuals at different sites without verification.35–37 For example, it is unclear how the outcome variables such as anemia and hypotension were defined, or the duration at specific values of hematocrit and blood pressure. Because they were not intended for research purposes, data on many variables may be missing, and there is inconsistent examination of different records. Perhaps the biggest misuse of the NIS database is that researchers may use it to examine variables that are not typically evaluated on all patients. It is highly probable that ION cases were examined more thoroughly for suggested risk factors including hypotension, anemia, and coexisting vascular diseases that have been associated with spontaneous ION such as atherosclerosis, peripheral vascular disease, diabetes, and hypertension compared with controls. The high odds ratio for hypotension is likely the result of intense examination of blood pressure only for cases with an adverse ischemic event (e.g., ION), and not for the controls. This inconsistent examination of variables between cases and controls will lead to erroneous association results. These issues surrounding large databases led to the controversy over their use in epidemiologic and health services research, and point to the need to consider validity and reliability issues.38,39 Without data on duration of surgery, blood loss, degree of anemia, and other variables in the NIS, correction for confounding factors cannot be performed.
The lack of any class I evidence on postoperative ION limits the strength of suggested interventions for prevention. These interventions are primarily based on the assumption that elevated venous pressure in the head reduces perfusion pressure to the optic nerve and contributes to inadequate oxygen delivery to the optic nerve.4 Maintaining the head in a neutral, nondependent position, and limiting the duration in the prone position to minimize the venous congestion of the head has been recommended by many experts.27 Using colloids along with crystalloids to maintain euvolemia, and avoiding extremes of anemia or hypotension have also been advocated to reduce edema formation and improve oxygen delivery to the optic nerve. As with any intervention, there may be unintended adverse side effects. For example, the simple maneuver of elevating the head of the bed may result in decreased venous return of blood to the heart, and lead to increased use of fluids and vasopressors to maintain a suitable perfusion pressure. The potential risk: benefit ratio of suggested interventions should be individualized for each patient.
There is currently no reliable monitor of intraoperative optic nerve function. Visual-evoked potentials have been shown to have a very high-false abnormality rate as a result of the degradation of the signal by general anesthesia, even with the use of total intravenous anesthesia.40 Further, the sensitivity of visual-evoked potentials for perioperative ION associated with spine surgery will be difficult to assess given the low incidence of this complication. It is also unclear if intraoperative visual-evoked potentials would prove useful in cases of AION where symptomatic visual loss may not manifest until the second or third postoperative day. Despite these limitations, further refinement of this monitor, or development of another optic nerve function monitor, with high sensitivity and specificity, would be the most ideal intervention for preventing this complication.
The clinical presentation of postoperative visual loss has always been diagnosed after the fact, typically in the recovery room, or in some cases the intensive care unit or patient rooms. Patients have been known to ask “When will they take the bandages off my eyes?” or have assumed their lack of vision is a result of residual anesthetic. This varied length of time to diagnosis may delay the workup and potential treatment, and would make implementation of a more formalized patient evaluation appear desirable.
For instance, routine postoperative neurologic checks can be enhanced by visual screening evaluations such as pupillary responses to light and ability to see hand motion or count fingers. However, the usefulness of such tests in an early postoperative phase may be hampered by side effects of residual anesthetic and narcotic analgesics on patient cooperation. Abnormality on examination or unusual patient visual complaint should trigger an urgent ophthalmological consultation where more formal testing of vision with Snellen cards, confrontation visual fields, and a fundoscopic examination can be performed. For example, one extremely rare cause of POVL, acute angle closure glaucoma, is amenable to treatment, if instituted early.
Based on the available studies we recommend that preoperative counseling include discussion of the very rare potential risk of blindness for major spine operations in the prone position. Discussing the risk of POVL as a remote risk would be similar to discussing the risk of heart attack or stroke. It is important to emphasize that the etiology of postoperative blindness remains unknown. Introduction of this topic by the surgical team before surgery would allow the anesthesiologist to discuss blindness with the patient without causing undue duress immediately before surgery, and would provide adequate time for consideration of the risks and benefits of the surgical procedure by the patient. Documentation of the preoperative risk counseling, would ideally be accessible to the anesthesiology team, and would also verify appropriate discussion of risks and benefits for any subsequent medicolegal claims. Both the discussion and the documentation are suggested for good patient care and for liability issues.
Treatment strategies for perioperative ION are currently unsupported by the literature. High-dose steroids, hyperbaric oxygen treatment, and mannitol have been used, but no consistent benefit has been demonstrated. Many neuro-ophthalmologists currently recommend restoration of baseline or near baseline blood pressure, and transfusion to correct severe anemia. As these treatments are also nonvalidated, the practitioner must weigh the potential benefits against the potential risks of increased fluid administration, use of vasopressors, and/or blood transfusion.
There was very low evidence that the following factors may be associated with postoperative ION: hypotension, anemia, blood transfusion, elevated venous pressure, prolonged duration in the prone position, and peripheral vascular disease.
For patients anticipated to undergo prolonged spine surgery exceeding 5 hours in the prone position with blood loss expected to exceed 1000 mL, we recommend discussing the remote risk of perioperative blindness before surgery.
Recommended interventions for major spine surgery patients to minimize the risk of perioperative vision loss include:
- The ASA Practice Advisory for Perioperative Visual Loss Associated with Spine Surgery
- Consider informing patients of risk of perioperative visual loss.
- Use direct continuous blood pressure monitoring.
- Use colloids along with crystalloids to maintain euvolemia.
- Position the head at the level of heart or higher, and in a neutral position.
- Consider staging very prolonged procedures.
- Frequent eye checks in the prone position to prevent central retinal artery occlusion caused by globe compression.
- There is very low evidence regarding postoperative ION associated with spine surgery.
- Unsupported associated factors include prolonged duration in the prone position, large blood loss, venous congestion of the head, anemia, hypotension, prolonged use of high-dose vasopressors, peripheral vascular disease, type and amount of fluid replacement, diabetes, and blood transfusion.
- Surgeons should consider counseling all patients undergoing anticipated prolonged spine surgery in the prone position with predictable substantial blood loss regarding the remote risk of perioperative blindness.
- Unsupported recommendations to minimize the risk of perioperative ischemic optic neuropathy associated with major spine surgery include: The ASA Practice Advisory for Perioperative Visual Loss Associated with Spine Surgery.
- Consider informing patients of risk of perioperative visual loss.
- Use direct continuous blood pressure monitoring.
- Use colloids along with crystalloids to maintain euvolemia.
- Position the head at the level of heart or higher, and in a neutral position.
- Consider staging very prolonged procedures.
- Frequent eye checks in the prone position may help prevent central retinal artery occlusion and other direct eye injuries from globe compression.
The authors are indebted to Ms. Nancy Holmes, RN, for her administrative assistance, to Mr. Jeff Hermsmeyer, BS, and Erika Ecker for their assistance in searching the literature, abstracting data, and proofing, and to Valerie Biousse, MD, for her assistance with Figures 1 and 2A–D.
1. Stevens WR, Glazer PA, Kelley SD, et al. Ophthalmic complications after spinal surgery. Spine
2. Dunker S, Hsu HY, Sebag J, et al. Perioperative risk factors for posterior ischemic optic neuropathy. J Am Coll Surg
3. Ho VT, Newman NJ, Song S, et al. Ischemic optic neuropathy following spine surgery. J Neurosurg Anesthesiol
4. Lee LA, Roth S, Posner KL, et al. The American Society of Anesthesiologists Postoperative Visual Loss
Registry: analysis of 93 spine surgery cases with postoperative visual loss
5. Patil CG, Lad EM, Lad SP, et al. Visual loss after spine surgery: a population-based study. Spine
6. Newman N. Perioperative visual loss after non-ocular surgeries. Am J Ophthalmol
7. Roth S. Postoperative blindness
. In Miller RD, ed. Miller's Anesthesia
, 6th ed. New York, NY: Elsevier; 2005:2991–3020.
8. Hollenhorst RW, Svien JH, Benoit CF. Unilateral blindness
occurring during anesthesia
for neurosurgical operations. AMA Arch Ophthalmol
8a. Biousse V, Newman NJ. Neuro-ophthalmology illustrated. Thieme, New York-Stuttagart, 2009, Fig. 8.16, p. 194, Illustration by Karl Wesker.
8b. Dettori JR, Norvell DC, Dekutoski M, et al. Methods for systematic reviews on patient safety during spine surgery. Spine
9. Wright JG, Swiontkowski MF, Heckman JD. Introducing levels of evidence to the journal. J Bone Joint Surg Am
10. van Tulder M, Furlan A, Bombardier C, et al. Updated method guidelines for systematic reviews in the cochrane collaboration back review group. Spine
11. Abraham M, Sakhuja N, Sinha S, et al. Unilateral visual loss after cervical spine surgery. J Neurosurg Anesthesiol
12. Alexandrakis G, Lam BL. Bilateral posterior ischemic optic neuropathy after spinal surgery. Am J Ophthalmol
13. Brown RH, Schauble JF, Miller NR. Anemia and hypotension as contributors to perioperative loss of vision. Anesthesiology
14. Chalam KV, Shah VA. Severe bilateral posterior ischemic optic neuropathy as a complication of spinal surgery. Eye
15. Dilger JA, Tetzlaff JE, Bell GR, et al. Ischaemic optic neuropathy after spinal fusion. Can J Anaesth
16. Heitz JW, Audu PB. Asymmetric postoperative visual loss
after spine surgery in the lateral decubitus position. Br J Anaesth
17. Kamming D, Clarke S. Postoperative visual loss
following prone spinal surgery. Br J Anaesth
18. Katz DM, Trobe JD, Cornblath WT, et al. Ischemic optic neuropathy after lumbar spine surgery. Arch Ophthalmol
19. Kim JW, Hills WL, Rizzo JF, et al. Ischemic optic neuropathy following spine surgery in a 16-year-old patient and a ten-year-old patient. J Neuroophthalmol
20. Lee A. PION after lumbar spine injury. Ophthalmology
2004;111:612; author reply 612.
21. Lee AG. Ischemic optic neuropathy following lumbar spine surgery. Case report. J Neurosurg
22. Murphy MA. Bilateral posterior ischemic optic neuropathy after lumbar spine surgery. Ophthalmology
23. Reddy A, Foroozan R, Edmond JC, et al. Dilated superior ophthalmic veins and posterior ischemic optic neuropathy after prolonged spine surgery. J Neuroophthalmol
24. Roth S, Nunez R, Schreider BD. Unexplained visual loss after lumbar spinal fusion. J Neurosurg Anesthesiol
25. Chang SH, Miller NR. The incidence of vision loss due to perioperative ischemic optic neuropathy associated with spine surgery: the Johns Hopkins Hospital Experience. Spine
26. Gill B, Heavner JE. Postoperative visual loss
associated with spine surgery. Eur Spine J
27. Practice advisory for perioperative visual loss associated with spine surgery: a report by the American Society of Anesthesiologists Task Force on Perioperative Blindness
28. Hayreh SS, Dass R. The ophthalmic artery. I. Origin and intra-cranial and intra-canalicular course. Br J Ophthalmol
29. Pillunat LE, Anderson DR, Knighton RW, et al. Autoregulation of human optic nerve head circulation in response to increased intraocular pressure. Exp Eye Res
30. Drummond JC. The lower limit of autoregulation: time to revise our thinking? Anesthesiology
31. Pazos GA, Leonard DW, Blice J, et al. Blindness
after bilateral neck dissection: case report and review. Am J Otolaryngol
32. Foroozan R. Shock-induced anterior ischaemic optic neuropathy after radical prostatectomy. Clin Exp Ophthalmol
33. Sadaba LM, Garcia-Layana A, Maldonado MJ, et al. Bilateral ischemic optic neuropathy after transurethral prostatic resection: a case report. BMC Ophthalmol
34. Weber ED, Colyer MH, Lesser RL, et al. Posterior ischemic optic neuropathy after minimally invasive prostatectomy. J Neuroophthalmol
35. Baron JA, Weiderpass E. An introduction to epidemiological research with medical databases. Ann Epidemiol
36. Lange LL, Jacox A. Using large data bases in nursing and health policy research. J Prof Nurs
37. Cowper DC, Hynes DM, Kubal JD, et al. Using administrative databases for outcomes research: select examples from VA Health Services Research and Development. J Med Syst
38. Connell FA, Diehr P, Hart LG. The use of large data bases in health care studies. Annu Rev Public Health
39. Flood AB. Peaks and pits of using large data bases to measure quality of care. Int J Technol Assess Health Care
40. Wiedemayer H, Fauser B, Sandalcioglu IE, et al. Observations on intraoperative monitoring of visual pathways using steady-state visual evoked potentials. Eur J Anaesthesiol