J Neurosurg / Volume 121 / August 2014

J Neurosurg 121:307–312, 2014

307

©AANS, 2014

The introduction of CT imaging in the 1970s has profoundly altered the practice of medicine in gen-eral and neurosurgery in particular; its inventors

were justifiably awarded the Nobel Prize in 1979.1 Rou-tine postoperative head CT scanning has since become a common practice; however, the origins of this habit are unclear, as is the evidence supporting it.9 Significant con-cerns over radiation exposure, costs, and hospital logis-tics involving excessive CT scans have arisen in the last 10 years.1 It has been recently estimated that patients with aneurysmal subarachnoid hemorrhage receive a mean cu-

mulative radiation dose of 2.76 Gy during the index hos-pitalization.23 Stein et al. reported that 26.3% of patients with newly diagnosed hydrocephalus received doses > 150 mSv to the ocular lens through CT scanning within 3 years of diagnosis.20 On the other hand, recent evidence involving head trauma and other general neurosurgical indications have suggested that a large fraction of CT ex-aminations may be unnecessary.2,9,18

Based on these concerns over unnecessary costs and radiation exposure, we sought to analyze our practice of obtaining early postoperative CT scans to determine if this led to early detection of complications and treatment. Our hypothesis is that early postoperative CT scans per-formed in intact patients did not detect pathological enti-ties requiring emergency surgical intervention.

Relevance of early head CT scans following neurosurgical procedures: an analysis of 892 intracranial procedures at Rush University Medical Center

Clinical articleRicaRdo B. V. Fontes, M.d., Ph.d., adaM P. sMith, M.d., LoRenzo F. Muñoz, M.d., RichaRd W. ByRne, M.d., and Vincent c. tRayneLis, M.d.Department of Neurosurgery, Rush University Medical Center, Chicago, Illinois

Object. Early postoperative head CT scanning is routinely performed following intracranial procedures for de-tection of complications, but its real value remains uncertain: so-called abnormal results are frequently found, but active, emergency intervention based on these findings may be rare. The authors’ objective was to analyze whether early postoperative CT scans led to emergency surgical interventions and if the results of neurological examination predicted this occurrence.

Methods. The authors retrospectively analyzed 892 intracranial procedures followed by an early postoperative CT scan performed over a 1-year period at Rush University Medical Center and classified these cases according to postoperative neurological status: baseline, predicted neurological change, unexpected neurological change, and se-dated or comatose. The interpretation of CT results was reviewed and unexpected CT findings were classified based on immediate action taken: Type I, additional observation and CT; Type II, active nonsurgical intervention; and Type III, surgical intervention. Results were compared between neurological examination groups with the Fisher exact test.

Results. Patients with unexpected neurological changes or in the sedated or comatose group had significantly more unexpected findings on the postoperative CT (p < 0.001; OR 19.2 and 2.3, respectively) and Type II/III interven-tions (p < 0.001) than patients at baseline. Patients at baseline or with expected neurological changes still had a rate of Type II/III changes in the 2.2%–2.4% range; however, no patient required an immediate return to the operating room.

Conclusions. Over a 1-year period in an academic neurosurgery service, no patient who was neurologically intact or who had a predicted neurological change required an immediate return to the operating room based on early postoperative CT findings. Obtaining early CT scans should not be a priority in these patients and may even be can-celled in favor of MRI studies, if the latter have already been planned and can be performed safely and in a timely manner. Early postoperative CT scanning does not assure an uneventful course, nor should it replace accurate and frequent neurological checks, because operative interventions were always decided in conjunction with the neuro-logical examination.(http://thejns.org/doi/abs/10.3171/2014.4.JNS132429)

Key WoRds      •      computed tomography      •      radiography      •      craniotomy      •      reoperation      •      traumatic brain injury

Abbreviations used in this paper: ICU = intensive care unit; RUMC = Rush University Medical Center; TBI = traumatic brain injury.

See the corresponding editorial in this issue, pp 305–306.

R. B. V. Fontes et al.

308 J Neurosurg / Volume 121 / August 2014

MethodsThis study is a retrospective chart review of all intra-

cranial procedures performed at Rush University Medical Center (RUMC) by the 10 surgeons in the Department of Neurosurgery between 01/01/2010 and 12/31/2010. This study was submitted to the RUMC Institutional Review Board and considered exempt. Our institutional protocol mandates that patients who have undergone any intracra-nial procedure be cared for in the first 24 hours postop-eratively within our Neurosciences Intensive Care Unit (ICU), a 28-bed unit staffed with specialized intensive care physicians and nurses. Hourly neurological exami-nation checks are performed during these 24 hours and early postoperative CT scans (< 4 hours) are obtained.

Two of the authors (R.B.V.F. and A.P.S.) reviewed the electronic medical record system to include all pro-cedures performed in the main operating rooms and at bedside and listed under our staff member names. After initial screening, a total of 953 open intracranial proce-dures were identified. Sixty-one procedures (57 patients) were excluded from the study. The main reason for exclu-sion was absence of immediate postoperative CT scans due to surgeon’s preference (59 cases). These 59 cases comprise a mix of deep brain stimulation (in patients who underwent MR imaging), suboccipital decompression for Chiari malformation, and transsphenoidal resection of pi-tuitary tumors, for which no imaging was deemed neces-sary. As for the 2 remaining cases, one patient died before CT scans could be obtained and one patient with glioma underwent immediate MRI studies.

In all, 892 procedures were identified. Patient and procedure data are listed in Table 1. Both elective and emergency/urgent cases were included; a significant por-tion (332 procedures, 37.2%) also involved placement of an intracranial device, most frequently a ventricular catheter. We used a methodology for analysis that was slightly modified from that of Khaldi et al.13 Preoperative and postoperative progress notes and physical examina-tions performed by the residents and faculty members of our department were reviewed and patients grouped in one of four categories according to immediate post-operative neurological examination findings: A, baseline; B, expected changes (including only focal deficits consis-tent with manipulation of eloquent cortical area or cranial nerve); C, unexpected changes; and D, comatose or se-dated (the latter also included patients already comatose at baseline).

Early postoperative CT images were reviewed, along with the progress notes documenting their interpretation by the members of our service, their subsequent actions, and the outcome. The CT findings were classified into ex-pected or unexpected (Fig. 1). Unexpected findings were further subdivided based on the actions then taken by the neurosurgery team: Type I, no active intervention except for heightened observation measures through additional imaging or longer ICU stay; Type II, active nonsurgical intervention, such as medical intracranial pressure man-agement; and Type III, active surgical intervention, either at bedside or in the operating room. It is important to note that no imaging interpretation was performed by the au-

thors of this study because retrospective analysis could introduce bias, but this last classification was purely based on the treating physicians’ interpretation at the time of imaging. The patients’ hospital course was reviewed un-til discharge, and delayed complications were also noted. The occurrence of unexpected CT results and Type II or III interventions in Groups B, C, and D was compared with Group A by using the Fisher exact test, at a signifi-cance level of 0.05 (Prism 6, GraphPad Software).25

ResultsOur results are briefly summarized in Fig. 1 and Tables

1 and 2. The vast majority of our patients were grouped into the A (neurological baseline) or D (sedated or coma-tose) categories. No difference was found in the incidence of unexpected CT findings between Groups A and B (Table 2). The incidence of unexpected changes was significantly higher in Groups C (unexpected change) and D (sedated or comatose). A sedated or comatose patient was 2.3 times more likely to have a postoperative CT with unexpected results; on the other hand, an unexpected postoperative

TABLE 1: Demographic and clinical data in 892 patients with intracranial procedures

Factor Value (%)*

no. of ops 892no. of patients 596sex male 296 (49.7) female 300 (50.3)surgery† craniotomy 405 (45.4) bur hole 379 (42.5) craniectomy 59 (6.6) cranioplasty 29 (3.3) transsphenoidal 25 (2.8)location‡ supratentorial 820 (91.9) infratentorial 73 (8.2)pathology or reason for op CSF diversion§ 328 (36.8) tumor 252 (28.3) vascular 144 (16.1) trauma 81 (9.1) reconstructive 36 (4.0) functional 27 (3.0) other 52 (5.8)device inserted 332 (37.2)

* Percentages may exceed 100% because two procedures may have been performed concomitantly.† Four procedures consisted of cranioplasty + bur hole.‡ One procedure was supra- and infratentorial. § Twenty-eight of the CSF diversion procedures were performed in the context of another pathology.

J Neurosurg / Volume 121 / August 2014

Relevance of postoperative head CT

309

change in neurological examination was 19.2 times more likely to result in an unexpected CT result.

Approximately 90% of patients in Groups A and B had the expected postoperative findings on CT scans; most patients with unexpected findings required a longer ICU stay and/or additional CT imaging, but no active in-tervention was necessary. Any interventions performed in Groups A and B were done in an elective and/or delayed fashion (following additional imaging and/or a period of clinical observation) or consisted of minor bedside adjust-ments—either retracting a few centimeters or completely removing a misplaced external ventricular drain (Fig. 1). No patient in Groups A or B had imaging findings that demanded an immediate return to the operating room.

Groups C and D, however, both contained cases for which CT findings dictated an immediate return to the operat-ing room; furthermore, unlike Group A, bedside insertion of new ventricular catheters or intracranial pressure mon-itors (procedures considered active surgical intervention in our service and this study) was also required.

DiscussionRoutine postoperative CT imaging is obtained for

two reasons: assessment of surgical technique and detec-tion of complications. It also allows for the establishment of a new “anatomical baseline” in the postoperative pa-tient, to which subsequent scans will be compared. Prior

Fig. 1. Flow chart showing summary of study results. EVD = external ventricular drain; ICP = intracranial pressure; RTO = return to operating room.

TABLE 2: Case distribution per neurological examination group*

Neurological ExaminationCT Results

p Value OR (95% CI)Unexpected Expected

expected change 4 38 0.552 1.30 (0.44–3.80)unexpected change 14 9 <0.001 19.20 (7.89–46.69)sedated or comatose 31 169 0.001 2.26 (1.39–3.68)baseline 47 580

* Each group was analyzed on a 2 × 2 format against the baseline group; p value is for the Fisher exact test.

R. B. V. Fontes et al.

310 J Neurosurg / Volume 121 / August 2014

to the advent of the CT imaging modality, neurosurgeons had considerable trouble diagnosing postoperative com-plications due to the lack of specific angiographic fea-tures of some common postoperative findings, such as intracerebral hemorrhage, edema, and necrotic brain tis-sue.14 Postoperative CT scans tend to be obtained as early as possible following cranial surgery (sometimes imme-diately upon arrival to the ICU), theoretically to enable early treatment of significant abnormalities. As a method to assess surgical technique, postoperative CT scanning can be highly valuable, particularly within the context of a training program. However, for this specific reason it need not be performed urgently following an interven-tion, especially when there are other medical priorities. Furthermore, it has been supplanted by MRI in specific situations such as surgery for intrinsic tumors, where ex-tent of resection cannot be accurately estimated with CT scanning and MRI studies may also be used to plan for adjuvant therapy.15,21

Complications are inherent to surgery and can be ex-pected to occur in a percentage of neurosurgical patients. Kalfas and Little reported 40 cases of postoperative he-matomas requiring medical or surgical intervention out of 4992 cranial procedures (0.8%).12 Interestingly, they reported a similar rate for both craniotomies and smaller procedures involving bur holes only. The cases reported by Kalfas and Little represent true complications because management was modified and surgical intervention was required in most of them. The occurrence of postopera-tive CT findings considered “abnormal” by different cri-teria (such as interpretation by a radiologist) is far higher. Horowitz et al. reported that 25% of patients undergoing vestibular schwannoma resection had abnormal postop-erative findings on CT scans, but only 8% of the changes resulted in symptoms.10 Fukamachi et al. reviewed 168 epidural hematomas that developed after 1055 intracra-nial procedures (15.9%); the vast majority (152 of 168) were small and regional, that is, located under the cra-niotomy flap or bur hole. None of these small regional hematomas had clinical implications and therefore none required evacuation.8 Overall, only 10 of the 168 hema-tomas required surgical intervention, which is a figure dramatically different from those found in any trauma se-ries.19,24 Deterioration of imaging features has been dem-onstrated with repeat imaging in traumatic subarachnoid hemorrhage, subdural hematomas, and intraparenchymal hematomas, but not necessarily related to a change in neurological condition.6,7,16

Whereas so-called abnormal postoperative CT find-ings are a common occurrence, postoperative neurologi-cal deterioration is not. Unexpected clinical deterioration in neurosurgical patients is known to correlate highly with progression of prior imaging abnormalities and also new pathology. This has been demonstrated in unselected neu-rosurgical patients,5 traumatic brain injury (TBI),19,24 vas-cular pathology, and in the postoperative period.13 When neurological deterioration is not verifiable—for example, as in the case of a patient with poor results on baseline examination or under sedation—the value of routine im-aging has also been demonstrated, with detection of new pathology and worsening of preexisting findings.16 Khaldi

et al. have recently demonstrated that early postoperative scans may have a very low predictive value, but when per-formed in the context of neurological deterioration, 30% required active intervention.13

Based on the evidence reviewed above, the concept of obtaining routine head CT scans at predetermined points in time has been challenged, particularly for the awake and cooperative patient with TBI. Sifri et al. studied a group of 161 patients with mild head injury; 99 of these patients had normal results on neurological examination at the time of their predetermined repeat head CT, and in none of them was a change in management indicated based on the repeat CT.18 Brown et al. further demon-strated that, of a group of almost 274 patients with TBI who underwent predetermined repeat head CT, none of the patients with mild TBI and unchanged neurological examination required a surgical intervention, resulting in 80 excess CTs.2 Garrett et al. have recently analyzed CT scans ordered for all indications within a large academic neurosurgical service; of 126 CT scans ordered postoper-atively, only 1 (0.8%) resulted in a return to the operating room. It is not specified, however, if this was performed in an early manner or if there was a change in neurologi-cal examination.9

The results presented here support the concept that routine postoperative CT scanning is not a priority for the awake and cooperative patient or those with new but predicted neurological deficits. Patients with unexpected neurological changes were 19.2 times more likely to have an unexpected finding on postoperative CT; on the oth-er hand, no patients in the A and B categories required emergency neurosurgical intervention based on the re-sults of early postoperative CT. This is to be expected be-cause the decision to perform any surgical intervention is defined by the binomial of neurological examination and imaging findings—if one of these criteria is considered satisfactory, surgeons are reluctant to return to the operat-ing room to simply “fix the CT.” It also corroborates what has been shown in TBI literature and smaller series of postoperative CTs. That is also the reason why we refrain from analyzing the sensitivity, specificity, and predictive values of postoperative CT findings for active postopera-tive management—it is an inherently flawed contingency analysis because the postoperative CT (“the test”) is one of the defining factors for which patients return to the op-erating room (“the disease”).

Due to the nature of our service and the referral pat-tern in our area, most of the procedures included in this study were elective. Although RUMC is only a Level II trauma center, there is a streamlined transfer system in place to allow patients within the greater Chicago area to be quickly transferred to RUMC for complex neuro-surgical care. It has been estimated that approximately 80% of these emergency transfers would be unrelated to cranial trauma.3,4 This explains the elevated vascular/trauma ratio of emergency and urgent cases in our se-ries and also the number of CSF diversion procedures. A significant number of Type III CT changes in our series consisted of improperly placed ventricular catheters (20 of 51 [39.2%]). In any large series with freehand place-ment of intracranial catheters, particularly if these hap-

J Neurosurg / Volume 121 / August 2014

Relevance of postoperative head CT

311

pen at bedside where ultrasound or navigational guid-ance is unavailable, misplacement will account for a significant number of reoperations. Inaccurate placement of ventricular catheters utilizing the freehand technique has been reported in the 15%–45% range.11,17,22 Misplaced catheters were managed in our service according to the patient’s neurological condition—if the patient was awake and stable, external catheters were simply removed and shunts were repositioned electively. If, on the other hand, the patient was sedated or comatose, the catheter was im-mediately repositioned at bedside (external catheters) or in the operating room (shunts).

A delayed postoperative CT scan has been suggest-ed as being more useful than an early scan, because it would allow for hematoma accumulation and the need to return to the operating room to be more evident.13 The retrospective nature of our study does not allow for con-clusions regarding timing; we did, however, immediately intervene in comatose or sedated patients when early CT scans revealed unexpected changes. It is unlikely that a retrospective study could answer this question, even if it included paired scans (early and delayed). Probably the only way to completely address the question of ideal timing is through a prospective randomized trial. We thus tend to favor a neurological examination–based ap-proach: postoperative CT scans should not be a priority in the immediate postoperative period, as long as the patient has a satisfactory neurological examination. A CT scan is probably not even necessary in the majority of patients in whom an early MRI study is planned, as in the case of surgery for intraaxial tumors and assessment of extent of resection. Within this database, 168 procedures for the resection of such tumors were identified after which the patient was awake and oriented and, accordingly, still un-derwent early postoperative CT. As described above, CT scans did not reveal alterations that demanded an imme-diate return to the operating room, but, regardless of the CT findings, these patients also underwent MRI to assess extent of resection. A minor change in protocol (MRI within 8 hours, for example) could have thus resulted in the cancellation of these 168 CTs in this study (18.9% of the total CTs performed) with no added cost or burden, since hourly neurological checks and MRI were already part of our postoperative routine. A final point to be con-sidered is whether there are additional patient character-istics besides the neurological examination results that would affect the decision to obtain early imaging, such as age or comorbidities; future prospective studies should be encouraged to look into this question.

ConclusionsOur review of the data presented here has demon-

strated that early postoperative CT scans performed in awake patients or those with expected deficits did not result in a single emergency surgical intervention, thus confirming the study hypothesis. On the other hand, new, unexpected neurological findings were frequently associ-ated with important new findings on CT scans, requiring emergency neurosurgical action. Early postoperative CT is thus not required on an emergency basis in awake pa-

tients, nor does it represent an “insurance policy” for the surgeon that all is well postoperatively. The inestimable value of the old-fashioned neurological examination per-formed by diligent, specialized nurses and physicians is reaffirmed once again.

Disclosure 

Dr. Byrne is a consultant to Stryker and Integra.Author contributions to the study and manuscript prepara-

tion include the following. Conception and design: Fontes, Smith, Traynelis. Acquisition of data: Fontes, Smith. Analysis and interpre-tation of data: Fontes. Drafting the article: Fontes. Critically revising the article: all authors. Reviewed submitted version of manuscript: all authors. Approved the final version of the manuscript on behalf of all authors: Fontes. Statistical analysis: Fontes. Administrative/technical/material support: Muñoz, Byrne, Traynelis. Study supervi-sion: Traynelis.

References

1. Brenner DJ, Hall EJ: Computed tomography—an increasing source of radiation exposure. N Engl J Med 357:2277–2284, 2007

2. Brown CVR, Zada G, Salim A, Inaba K, Kasotakis G, Had-jizacharia P, et al: Indications for routine repeat head comput-ed tomography (CT) stratified by severity of traumatic brain injury. J Trauma 62:1339–1345, 2007

3. Byrne RW, Bagan BT, Bingaman W, Anderson VC, Selden NR: Emergency neurosurgical care solutions: acute care surgery, re-gionalization, and the neurosurgeon: results of the 2008 CNS consensus session. Neurosurgery 68:1063–1068, 2011

4. Byrne RW, Bagan BT, Slavin KV, Curry D, Koski TR, Origi-tano TC: Neurosurgical emergency transfers to academic cen-ters in Cook County: a prospective multicenter study. Neuro-surgery 62:709–716, 2008

5. Dauch WA, Szilagyi KD: “Medical decision making” for re-peated computed tomography in neurosurgical intensive care patients. Neurosurg Rev 17:141–144, 1994

6. Durham SR, Liu KC, Selden NR: Utility of serial computed tomography imaging in pediatric patients with head trauma. J Neurosurg 105 (5 Suppl):365–369, 2006

7. Fainardi E, Chieregato A, Antonelli V, Fagioli L, Servadei F: Time course of CT evolution in traumatic subarachnoid haem-orrhage: a study of 141 patients. Acta Neurochir (Wien) 146: 257–263, 2004

8. Fukamachi A, Koizumi H, Nagaseki Y, Nukui H: Postoperative extradural hematomas: computed tomographic survey of 1105 intracranial operations. Neurosurgery 19:589–593, 1986

9. Garrett MC, Bilgin-Freiert A, Bartels C, Everson R, Afsar-manesh N, Pouratian N: An evidence-based approach to the efficient use of computed tomography imaging in the neuro-surgical patient. Neurosurgery 73:209–216, 2013

10. Horowitz SW, Leonetti JP, Azar-Kia B, Anderson D: Postop-erative radiographic findings following acoustic neuroma re-moval. Skull Base Surg 6:199–205, 1996

11. Hsieh CT, Chen GJ, Ma HI, Chang CF, Cheng CM, Su YH, et al: The misplacement of external ventricular drain by freehand method in emergent neurosurgery. Acta  Neurol  Belg 111: 22–28, 2011

12. Kalfas IH, Little JR: Postoperative hemorrhage: a survey of 4992 intracranial procedures. Neurosurgery 23:343–347, 1988

13. Khaldi A, Prabhu VC, Anderson DE, Origitano TC: The clini-cal significance and optimal timing of postoperative comput-ed tomography following cranial surgery. Clinical article. J Neurosurg 113:1021–1025, 2010

14. Lin JP, Pay N, Naidich TP, Kricheff II, Wiggli U: Computed tomography in the postoperative care of neurosurgical pa-tients. Neuroradiology 12:185–189, 1977

R. B. V. Fontes et al.

312 J Neurosurg / Volume 121 / August 2014

15. Meyding-Lamadé U, Forsting M, Albert F, Kunze S, Sartor K: Accelerated methaemoglobin formation: potential pitfall in early postoperative MRI. Neuroradiology 35:178–180, 1993

16. Oertel M, Kelly DF, McArthur D, Boscardin WJ, Glenn TC, Lee JH, et al: Progressive hemorrhage after head trauma: pre-dictors and consequences of the evolving injury. J Neurosurg 96:109–116, 2002

17. Patil V, Lacson R, Vosburgh KG, Wong JM, Prevedello L, An-driole K, et al: Factors associated with external ventricular drain placement accuracy: data from an electronic health re-cord repository. Acta Neurochir (Wien) 155:1773–1779, 2013

18. Sifri ZC, Homnick AT, Vaynman A, Lavery R, Liao W, Mohr A, et al: A prospective evaluation of the value of repeat cranial computed tomography in patients with minimal head injury and an intracranial bleed. J Trauma 61:862–867, 2006

19. Stammler U, Frowein RA: Repeated early CT examinations in traumatic intracranial hematomas and in closed head injuries. Neurosurg Rev 12 (Suppl 1):159–168, 1989

20. Stein EG, Haramati LB, Bellin E, Ashton L, Mitsopoulos G, Schoenfeld A, et al: Radiation exposure from medical imag-ing in patients with chronic and recurrent conditions. J Am Coll Radiol 7:351–359, 2010

21. Warmuth-Metz M: Postoperative imaging after brain tumor resection. Acta Neurochir Suppl 88:13–20, 2003

22. Wilson TJ, Stetler WR Jr, Al-Holou WN, Sullivan SE: Com-parison of the accuracy of ventricular catheter placement using freehand placement, ultrasonic guidance, and stereotactic neu-ronavigation. Clinical article. J Neurosurg 119:66–70, 2013

23. Wong JM, Ho AL, Lin N, Zenonos GA, Martel CB, Frerichs K, et al: Radiation exposure in patients with subarachnoid hemorrhage: a quality improvement target. Clinical article. J Neurosurg 119:215–220, 2013

24. Yoshimizu N, Hiramoto M, Notani M: Study of those cases which showed rapid deterioration within a few hours after head injury—importance of follow-up CT scans at an early stage. Neurosurg Rev 12 (Suppl 1):175–177, 1989

25. Zar JH: Biostatistical Analysis,  ed  4. Upper Saddle River, NJ: Prentice Hall, 1999

Manuscript submitted December 9, 2013.Accepted April 24, 2014.Please include this information when citing this paper: pub-

lished online May 30, 2014; DOI: 10.3171/2014.4.JNS132429.Address correspondence to: Ricardo B. V. Fontes, M.D., Ph.D.,

Department of Neurosurgery, Rush University Medical Center, 1725 W. Harrison, Ste. 855, Chicago, IL 60612. email: ricardo_fontes@rush.edu.