Secondary Logo

Journal Logo

Clinical Investigations

Effect of Daytime Versus Night-time on Outcome in Patients Undergoing Emergent Neurosurgical Procedures

Qadri, Ali H. BSc*; Sproule, Stephanie BSc, MMath; Girling, Linda BSc; West, Michael BSc, MD, PhD, FRCSC§; Cappellani, Ronald MD, FRCPC; Chowdhury, Tumul MD, DM, FRCPC

Author Information
Journal of Neurosurgical Anesthesiology: October 2020 - Volume 32 - Issue 4 - p 315-322
doi: 10.1097/ANA.0000000000000600


Neurosurgical procedures are considered inherently high-risk1 and are associated with high rates of perioperative morbidity and mortality.2 Several factors are believed to influence patient outcomes after neurosurgical procedures. These include patient-related factors, such as age,3 sex,4 and comorbid diseases5,6; disease and surgery-related factors, such as intracranial hypertension and emergent versus elective surgery7; physician-related factors, such as experience8,9; and institution-related factors, such as availability of resources.10,11

The start time of surgery is a contentious surgery-related factor. Several challenges have been identified with night-time surgery including physician fatigue10,12,13 and sleep deprivation,11,14 decreased staff and resources,10,15 and decision fatigue.16 Physician fatigue and excessive duty hours have been associated with a higher rate of medical errors.16,17 Further, night shifts disturb the circadian rhythm, alertness, and cognitive functions.16,18 Higher functions remain at their lowest levels between 3 and 5 am,18 whereas decision fatigue is more pronounced later in the day and can lead to diagnostic errors.16 Burnout and depression have also been observed to be more common among physicians and residents who work long hours, and this may affect patient care and safety.17

Previous studies have reached different conclusions regarding an association between night-time surgeries and adverse outcomes.6,9,10,12,13,15,19–21 Some found no difference in perioperative complications or mortality in night-time compared with daytime surgery,6,9,12,15,19 whereas others demonstrated that night-time surgery is associated with higher complication rates.10,12,22,23 In some surgical specialties operations that were routinely performed overnight are now being considered for urgent or elective scheduling during the day in an attempt to address patient safety concerns.22,23 The impact of timing of adult neurosurgical procedures on neurological outcomes remains a topic of debate.

The aim of this study is to investigate whether the timing of emergent neurosurgical procedures (daytime vs. night-time) affects neurological outcomes.


The research ethics board of the University of Manitoba and Health Sciences Center, Winnipeg, Canada (HS20635 H2017: 113) approved this retrospective study in March 2017. The study was conducted on the basis of robust methodological principles for retrospective chart reviews as outlined by Matt and Matthew.24

All adult patients who had undergone emergent neurosurgical procedures, defined as E1 (surgery performed immediately or in the next available operating room in <1 h) or E2 (critically emergency surgery performed within 6 h), between January 1, 2010 and December 31, 2016 were included in the study. Patients with multiple trauma, spine surgeries, re-do procedures, cerebrospinal fluid diversion procedures, pediatric patients (age younger than 18 y), and pregnant patients were excluded from the study. Patients with missing primary outcomes data were excluded from the study, whereas those with missing data for other variables (secondary outcomes) were included in the final analysis.

Data Collection and Quality Control

The list of patients, procedures and charts were retrieved from our operating room database, and medical records department, respectively. The data collected are outlined in the Supplementary Figure 1 (Supplemental Digital Content 1, The data collector (A.H.Q.) performed data review and collection under the supervision of the primary investigator (T.C.) for the first 5 cases to ensure accuracy in data abstraction. Any discrepancies or uncertainties on the data collector’s part were clarified. After this training, data were reviewed by the coinvestigators (T.C., R.C.) after every 30 cases to ensure accuracy and reduce data collection bias. The data manual/guidelines document included the general rules and guidelines for data abstraction from the medical chart, a list of synonyms and abbreviations, inclusion and exclusion criteria, and codes for different variables.

The identified patient cohort was divided into 2 groups based on the start time of surgery: a daytime group in which surgery started between 7:00 am and 6:59 pm, and a night-time group in which surgery started between 7:00 pm and 6:59 am. The 7:00 am cutoff was chosen to coincide with the handover time for both operating room nursing and physician staff, and the 7:00 pm cutoff to divide the day into 2 equal periods.


The primary outcome was neurological outcome at discharge, as assessed by the Glasgow Outcome Scale (GOS). The GOS is a validated and standardized assessment of neurological outcomes with 5 grades: 1: death, 2: persistent vegetative state, 3: severe disability, 4: moderate disability, and 5: good recovery.24 For the purposes of this study the GOS was dichotomized into favorable and unfavorable outcomes: “favorable” outcome, GOS 4 or 5 (moderate disability or good recovery), and “unfavorable” outcome, GOS 1, 2, or 3 (death, persistent vegetative state, or severe disability).25 Neurological outcomes at 1 month postdischarge were collected for a secondary sensitivity analysis. Secondary outcomes included the effect of perioperative variables on neurological outcome, differences in length of hospital stay and comparison of the perioperative anesthetic and surgical complications between the 2 groups. The intraoperative mean arterial pressure difference (ie, the difference between highest and lowest mean arterial pressure during surgery) was used as an assessment of intraoperative blood pressure control. All the variables outlined in Supplementary Figure 1 (Supplemental Digital Content 1,, except hospital length of stay and anesthetic and surgical complications, are the perioperative variables that were used for secondary outcome analysis.


The majority of procedures were performed under general anesthesia with endotracheal intubation. Local anesthetic and sedation were used for awake craniotomies for procedures involving brain tumors near the eloquent area, for burr holes for evacuation of chronic subdural hematoma (when indicated by clinical status), and brain abscess aspiration. Standard monitoring, including an arterial line for invasive blood pressure monitoring, was applied. After surgery patients received a trial of extubation on the basis of their neurological and respiratory status, and were transferred to the postanesthesia care unit or intensive care unit as appropriate.

Statistical Analysis

A sample size calculation was performed for the univariate comparison of surgery time of day with risk of unfavorable outcome assessed by the GOS (the primary outcome). On the basis of the following assumptions—(i) outcome risks of 15% and 30% for unfavorable outcome for day surgeries and night surgeries, respectively,12 (ii) an equal number of night and day surgeries, (iii) 80% power, and (iv) a type-1 error rate of 5%—a minimum sample size of 242 was calculated using PROC POWER of SAS, version 9.3 (SAS Institute, Cary, NC). Our patient database, which includes all neurosurgeries for the 6-year period of the study, exceeded this minimum total sample size. However, after applying the predefined inclusion and exclusion criteria, the study did not have equal numbers of patients in each group. As only a few additional patients (16) were required for the night-time group based on the power analysis, we decided to proceed with the final analysis using the data from this 6-year period.

For the primary outcome (neurological outcome at hospital discharge and repeated for sensitivity analysis at 1 month postdischarge), the Pearson χ2 test was used to compare daytime versus night-time surgeries. For secondary outcomes, comparisons were performed using the Fisher exact test, the Pearson χ2 test, 2-sample t tests or nonparametric Mann-Whitney tests, depending on whether the outcome was categorical or continuous, the normality of the distribution, and the presence of outliers. Normality assumption was assessed through visual inspection of histograms and normal quantile plots. Continuous outcomes were analyzed using t test or Mann-Whitney nonparametric test, as appropriate. The Pearson χ2 test or the Fisher exact test were used for categorical variables. The logistic multiregression analysis was performed in R to evaluate the effects of various factors on unfavorable neurological outcome. All factors evaluated in the model are summarized; no factor selection was performed. In cases where categorical variables had no or low sample sizes for ≥1 categories, similar categories were grouped for analysis. When 2 factors were highly correlated, only 1 factor was included (eg, procedure was included in the model and not diagnosis). Data are summarized using mean and SDs for continuous variables, whereas counts and percentages are presented for categorical variables. Odds ratios with corresponding 95% confidence intervals (CIs), estimated from the logistic regression model, are also presented. A P value <0.05 was considered significant. No adjustments for multiple comparisons were performed.


A total of 620 patient charts were screened and, based on predefined exclusion and missing data omission criteria, 304 were included in the final analysis (Fig. 1). There were 199 patients in the daytime surgery group and 105 in the night-time surgery group. The mean age of the patients was 58.2 years (43.1% were over the age of 65 y), and 66.4% were male. The 2 groups were comparable with respect to age, sex, body mass index, and American Society of Anesthesiologists physical status score; 64.8% had an American Society of Anesthesiologists score of II (Table 1).

Screening of charts. CSF indicates cerebrospinal fluid.
TABLE 1 - Demographic and Perioperative Characteristics
n (%)
Variables Total (n=304) Daytime (n=199) Night-time (n=105)
Age (y)
 Mean (SD) 58.2 (19.29) 58.6 (18.59) 57.4 (20.62)
 ≥65 131 (43.1) 84 (42.2) 47 (44.8)
 <65 173 (56.9) 115 (57.8) 58 (55.2)
 Male 202 (66.4) 133 (66.8) 69 (65.7)
 Female 102 (33.6) 66 (33.2) 36 (34.3)
 <18.5 7 (5.3) 4 (4.6) 3 (6.7)
 18.5-24.9 50 (37.9) 36 (41.4) 14 (31.1)
 25.0-29.9 38 (28.8) 26 (29.9) 12 (26.7)
 30.0-34.9 26 (19.7) 16 (18.4) 10 (22.2)
 35.0-39.9 5 (3.8) 3 (3.5) 2 (4.4)
 ≥40.0 6 (4.5) 2 (2.3) 4 (8.9)
 Missing* 172
ASA score
 I 67 (22.0) 44 (22.1) 23 (21.9)
 II 197 (64.8) 126 (63.3) 71 (67.6)
 III 39 (12.8) 28 (14.1) 11 (10.5)
 IV 1 (0.3) 1 (0.5) 0
 V 0 0 0
*Percentages are calculated excluding patients with missing values.
ASA indicates American Society of Anesthesiologists; BMI, body mass index.

Diagnostic and Surgical Characteristics

The daytime and night-time surgery groups were comparable for type and duration of primary symptoms, diagnosis, day (weekend, weekdays) and season (spring, summer, autumn, winter) of the surgery. Overall, 36.5% of patients presented with altered sensorium, and 31.9% with headache; the mean duration of the primary symptoms was 7.6 days. More than half of the patients (59.5%) had a subdural hematoma. The night-time surgery group had lower median (range) baseline Glasgow Coma Scale (GCS) scores (14 [3, 15]) compared with the daytime surgery group (14 [3, 15]) (P=0.0013). In addition, a greater percentage of the patients in the night-time group presented with midline shift and had larger midline shift compared with the daytime group (P=0.0009 and 0.0127, respectively). Strikingly, the majority of patients in the night-time surgery group came to the operating room directly from the emergency department (63.8%), whereas 57.7% of those having surgery in the daytime came from a ward or postoperative care unit (P<0.0001) (Table 2 and Supplementary Table 1, Supplemental Digital Content 2,

TABLE 2 - Diagnostic, Surgical, and Intraoperative Data
n (%)
Variables Total (n=304) Daytime (n=199) Night-time (n=105)
Primary symptom
 Altered sensorium 111 (36.5) 67 (33.7) 44 (41.9)
 Headache 97 (31.9) 67 (33.7) 30 (28.6)
 Nausea and/or vomiting 9 (3.0) 6 (3.0) 3 (2.9)
 Visual changes 3 (1.0) 2 (1.0) 1 (1.0)
 Seizure 18 (5.9) 12 (6.0) 6 (5.7)
 Hemiparesis or hemiplegia 29 (9.5) 17 (8.5) 12 (11.4)
 Facial palsy 2 (0.7) 2 (1.0) 0
 Balance and/or gait 15 (4.9) 13 (6.5) 2 (1.9)
 Sensation loss and/or cranial nerve issue 5 (1.6) 4 (2.0) 1 (1.0)
 Other 15 (4.9) 9 (4.5) 6 (5.7)
Symptom duration (d)
 Mean (SD) 7.6 (18.74) 8.6 (22.08) 5.9 (9.5)
 Intracerebral bleed 27 (8.9) 16 (8.0) 11 (10.5)
 Subdural hematoma 181 (59.5) 115 (57.8) 66 (62.9)
 Epidural hematoma 20 (6.6) 12 (6.0) 8 (7.6)
 Unruptured aneurysm 5 (1.6) 4 (2.0) 1 (1.0)
 Subarachmoid hemorrhage 17 (5.6) 13 (6.5) 4 (3.8)
 Brain tumor 31 (10.2) 21 (10.6) 10 (9.5)
 Brain abcess 7 (2.3) 5 (2.5) 2 (1.9)
 Other 16 (5.3) 13 (6.5) 3 (2.9)
 Base systolic blood pressure, mean (SD)* 142.5 (26.35) 142.3 (25.49) 142.9 (28.03)
 Base diastolic blood pressure, mean (SD)* 77.6 (14.31) 76.7 (13.89) 79.5 (14.97)
 Base mean arterial pressure, mean (SD)* 99.3 (16.46) 98.5 (15.99) 100.7 (17.3)
 Base heart rate, mean (SD) 78.1 (18.17) 76.7 (17.28) 80.7 (19.56) §
 Base Glasgow Coma Scale, median (range) 15 (3, 15) 15 (3, 15) 14 (3, 15) §
 Base hemoglobin, mean (SD) 129.3 (20.57) 129.4 (20.45) 129.0 (20.89)
 Base hemoglobin<90 14 (4.7) 8 (4.1) 6 (5.8)
 Base hemoglobin≥90 283 (95.3) 186 (95.9) 97 (94.2)
 Missing† 7
Primary location
 ICU 27 (9.0) 13 (6.6) 14 (13.3) §
 ED 137 (45.5) 70 (35.7) 67 (63.8)
 Ward/PACU 137 (45.5) 113 (57.7) 24 (22.9)
 Missing† 3
Secondary location
 ICU 56 (18.4) 31 (15.6) 25 (23.8)
 PACU 248 (81.6) 168 (84.4) 80 (76.2)
Intraoperative hemodynamics
 Systolic blood pressure difference, mean (SD) 54.2 (23.21) 52.4 (22.23) 57.8 (24.81)
 Diastolic blood pressure, mean (SD) 30.4 (13.05) 28.6 (11.63) 34.0 (14.96) §
 Mean arterial blood pressure, mean (SD) 37.6 (14.45) 36.3 (13.17) 40.1 (16.39) §
 Sedation 35 (11.7) 24 (12.3) 11 (10.6)
 General anesthetic 264 (88.3) 171 (87.7) 93 (89.4)
 Missing† 5
 Craniotomy with tumor excision 23 (7.6) 15 (7.5) 8 (7.6) §
 Craniotomy with aneurysmal clipping 26 (8.6) 21 (10.6) 5 (4.8)
 Burr holes and evacuation of hematoma 106 (34.9) 74 (37.2) 32 (30.5)
 Craniotomy and hematoma evacuation 115 (37.8) 63 (31.7) 52 (49.5)
 Other‡ 34 (11.2) 26 (13.1) 8 (7.6)
 E1 134 (44.1) 77 (38.7) 57 (54.3) §
 E2 170 (55.9) 122 (61.3) 48 (45.7)
Length of hospital stay (d)
 Mean (SD) 14.3 (28.6) 13.7 (30.5) 15.3 (24.9) §
 Missing† 1
Intraoperative complications
 Present 143 (47.0) 92 (46.2) 51 (48.6)
 Absent 161 (53.0) 107 (53.8) 54 (51.4)
Medical complications
 Present 70 (23.0) 44 (22.1) 26 (24.8)
 Absent 234 (77.0) 155 (77.9) 79 (75.2)
Surgical complications
 Present 112 (36.8) 74 (37.2) 38 (36.2)
 Absent 192 (63.2) 125 (62.8) 67 (63.8)
*One patient with missing values for systolic and diastolic blood pressure and a recorded value of 0 for mean arterial pressure is excluded.
†Percentages are calculated excluding patients with missing values.
‡Other category includes depressive caniotomy or craniectomy, incision and drainage, cranioplasty, and others.
§Nominal P<0.05 based on χ2 test (categorical variables) or Mann-Whitney test (numeric variables).
df indicates degrees of freedom; ED, emergency department; ICU, intensive care unit; PACU, postanesthesia care unit.

Intraoperative and Postoperative Characteristics and Complications

Overall, 54.3% of patients in the night-time surgery group were designated surgical acuity E1 compared with 38.7% in the daytime group (P=0.0092). The most common procedure in the night-time group was craniotomy with hematoma evacuation (49.5%), whereas burr holes and evacuation of hematoma was the most common procedure in the daytime group (37.2%) (P=0.0244). General anesthesia was the most common anesthetic method (88.3%), and the majority of patients (93.2%) were operated in the supine position. The duration of surgery was >2 hours in 60.5% of patients. There was a larger intraoperative mean arterial pressure difference in patients the night-time compared with daytime surgery group (40.1 vs. 36.3 mm Hg, respectively; P=0.0434). Night-time surgery patients also had a longer hospital length of stay (15.3 d) compared with those in the daytime group (13.7 d; P=0.0013). The majority of patients (94.4%) did not require readmission to hospital after discharge.

Intraoperative complications occurred in approximately half of the patients (47%). Overall 23% and 36.8% of patients had anesthetic and surgical complications, respectively, but there was no significant difference between the groups in the proportion who developed either complications (Table 2 and Supplementary Table 1, Supplemental Digital Content 2,

Neurological Outcomes

Overall, 50 patients (16.4%) had an unfavorable neurological outcome at hospital discharge and 48 (15.8%) unfavorable neurological outcome at 1 month postdischarge. There was no difference in the primary outcome (unfavorable GOS score) between patients in the daytime and night-time groups at hospital discharge. Overall, 14.1% of patients in the daytime surgical group had unfavorable outcome compared with 21.1% of those in the night-time group (P=0.12) (absolute difference, −6.9%; 95% CI: −16.04, 2.28) (Table 3). At 1 month after discharge, 14.6% in the daytime group and 18.1% in the night-time group had unfavorable outcome (P=0.42) (absolute difference, −3.5%; 95% CI: −12.37, 5.32).

TABLE 3 - Neurological Outcomes
n (%)
Variables Total (n=304) Daytime (n=199) Night-time (n=105) Absolute Difference (95% CI) χ2 (df) P
Neurological outcome at discharge
 Favorable 254 (83.6) 171 (85.9) 83 (79.1) −6.9 (−16.04, 2.28) 2.3689 (1) 0.1238
 Unfavorable 50 (16.4) 28 (14.1) 22 (21.0)
Neurological outcome at 1 mo postdischarge
 Favorable 256 (84.2) 170 (85.4) 86 (81.9) −3.5 (−12.37, 5.32) 0.6414 (1) 0.4232
 Unfavorable 48 (15.8) 29 (14.6) 19 (18.1)
CI indicates confidence interval; df, degrees of freedom.

In the logistic regression model only 5 among 20 variables were associated with unfavorable neurological outcome at discharge; these were age, baseline GCS score, procedure type, surgery acuity (E1, E2 status), and intraoperative complications (Table 4). McFadden’s R2 was 0.405, indicating that the model is a good fit to the data.

TABLE 4 - Logistic Regression for Outcome at Discharge
Variables β SE OR (95% CI) P
Intercept −1.8193 2.5637 0.4779
Night-time surgery −0.2616 0.4612 0.77 (0.31, 1.89) 0.5705
Age (y) 0.057 0.0159 1.06 (1.03, 1.09) <0.001
Female 0.2409 0.4743 1.27 (0.50, 3.22) 0.6115
ASA score* 0.1283 0.6089 1.14 (0.33, 3.68) 0.8331
Symptom duration (d) −0.01 0.0258 0.99 (0.94, 1.03) 0.6986
Midline shift present −0.1192 0.574 0.89 (0.29, 2.78) 0.8356
Weekend 0.4076 0.4798 1.50 (0.58, 3.88) 0.3956
 Spring 0.26 0.6439 1.30 (0.37, 4.72) 0.6863
 Summer 0.83 0.6479 2.29 (0.66, 8.55) 0.2002
 Winter 1.1322 0.7039 3.10 (0.80, 12.97) 0.1077
Baseline mean arterial blood pressure 0.0083 0.0132 1.01 (0.98, 1.04) 0.5312
Baseline heart rate −0.0004 0.0116 1.00 (0.98, 1.02) 0.9709
Baseline GCS −0.1923 0.0631 0.83 (0.73, 0.93) 0.0023
Baseline hemoglobin −0.0144 0.0116 0.99 (0.96, 1.01) 0.2148
Primary location†
 ED 0.7722 0.7402 2.16 (0.54, 10.13) 0.2968
 Ward/PACU 0.5152 0.8076 1.67 (0.36, 8.72) 0.5235
Intraoperative mean arterial blood pressure difference 0.0062 0.0157 1.01 (0.98, 1.04) 0.695
 Craniotomy with aneurysmal clipping −2.8484 1.3512 0.06 (0.00, 0.65) 0.0350
 Burr holes and evacuation of hematoma −2.6721 0.9967 0.07 (0.01, 0.48) 0.0073
 Craniotomy and hematoma evacuation −1.2329 0.9035 0.29 (0.05, 1.76) 0.1724
 Other −3.2086 1.3892 0.04 (0.00, 0.47) 0.0209
E2§ −1.2407 0.5853 0.29 (0.09, 0.89) 0.0340
Length of hospital stay (d) 0.0063 0.0058 1.01 (1.00, 1.02) 0.2809
Intraoperative complications present 1.5694 0.4978 4.80 (1.88, 13.52) 0.0016
Surgery length>120 min −0.5412 0.5649 0.58 (0.19, 1.75) 0.3381
Bold value indicates statistical significance.
McFadden’s R2=0.405. Likelihood-ratio test versus null model; P<0.001.
0=favorable outcome; 1=unfavorable outcome.
*0=I or II, 1=III or IV.
†Intensive care unit is the reference category.
‡Craniotomy with tumor excision is the reference category.
§E1 is the reference category.
ASA indicates American Society of Anesthesiologists; β, regression coefficient; CI, confidence interval; ED, emergency department; GCS, Glasgow Coma Scale; OR, odds ratio; PACU, postanesthesia care unit.


This study found that patients undergoing emergent neurosurgical procedures during the night-time had similar rates of unfavorable neurological outcomes at hospital discharge and 1 month after discharge as those whose surgery was performed during the day. Previous literature on this topic is limited and controversial; the results of a preliminary search of publications dealing with this issue over the past 10 years are shown in Supplementary Figure 2 (Supplemental Digital Content 3,,12–15,18,19,20,26–30 One study in adult neurosurgical procedures found a 50% increase in the odds of complications with night-time surgery.13 Another, in a pediatric neurosurgical population, also highlighted higher rates of intraoperative complications, morbidity and mortality in after-hours weekday compared with regular-hours weekday surgeries.19 Some studies in non-neurosurgical procedures similarly suggest that night-time surgeries are associated with more complications and increased mortality.20,26,27,31 In a large-scale study of 2,948,842 cases, Whitlock et al31 found that operative start time is associated with 48-hour mortality, after adjustment for comorbidities; specifically this study found higher mortality in surgeries starting after 6:00  pm. In contrast, a recent study found no link between physician fatigue and worse outcomes after cerebral aneurysm surgery.32 In that study there was no difference in in-patient mortality or hospital length of stay between patients undergoing elective cerebral aneurysm operations by surgeons who performed emergency procedures the night before, compared with those who did not. Several large-scale studies of non-neurosurgical operations also suggest no difference in perioperative complications and mortality in night-time compared with daytime surgeries.12,14,15,18,28–30,32,33,34 Our study supports this notion as it did not demonstrate an increase in intraoperative or postoperative complications in night-time surgeries. The slightly higher rates of unfavorable outcome and longer length of hospital stay in the night-time group may be explained by the perioperative characteristics of patients in the night-time surgery group who had lower baseline GCS scores and increased presence of midline shift and larger midline shifts compared with those in the daytime group. Further, a greater proportion of patients in the night-time group were classified as E1. Other studies have also found that decreased GCS scores4 and increased midline shift35 are prognosticators for poor outcomes.

The logistic regression model showed that age, baseline GCS, surgical acuity status (E1/E2), procedure type, and intraoperative complications influenced the incidence of unfavorable outcomes in our study. Of these factors, intraoperative complications and surgical acuity E1 seem to be strong predictors; the odds of an unfavorable outcome were 4.8 times higher if intraoperative complications were present and 3.45 times higher if the case was an E1 emergency. Cases classified as more emergent on admission are generally associated with higher morbidity and mortality.7,13 Older age was not a strong predictor for unfavorable outcome in our study (odds ratio=1.06), although other studies have identified older age as a predictor for poor outcome.3,15

There has also been considerable debate about weekend surgeries, which present some of the same challenges as night-time surgeries with regards to staffing and physician fatigue. Studies show different results regarding the so-called “weekend effect” in which patients admitted during the weekend are suspected to have worse outcomes.19,36 There is a similar contention around the “July effect”, which is the hypothesized increase in morbidity and mortality corresponding to the entry of new medical house staff.5,8 However, these associations were not found in our study.

Our study has several limitations. First, we were not able to identify several potentially confounding variables because the required data were not present in the patient charts. These included the experience level of the physicians and other staff, length of time that personnel had been on shift, and availability of resources in the hospital environment. Second, we used only the start time of each operation to allocate patients into daytime or night-time surgery groups. Thus, a 4-hour procedure that started at 6:30 pm would be categorized as daytime surgery even though the majority of the surgery took place during “night-time” hours. Third, the time categories chosen for this study (7 am to 6:59 pm and 7 pm to 6:59 am) do not truly correspond to turnover hours for operating room nursing staff and physicians. In our hospital, operating room nursing staff turnover occurs at 3:00 pm, 11 pm, and 7 am, and physician (staff and residents) shift changes occur at 3:30 pm and 7 am. Fourth, this is a retrospective study. Although every effort was made to ensure rigorous and accurate data collection, we cannot exclude the possibility of errors in the collection and transcription of data. Finally, although the overall sample size was achieved the number of patients in each group was not equal such that the number in the night-time group was slightly lower than the predefined minimum. The power to detect differences in the primary and some secondary outcomes may therefore be limited. Although it seems unlikely (though not certain) that inclusion of an additional small number of cases (16) in the night-time surgery group would have impacted the final results, our findings should be interpreted with this caveat in mind. In addition, the confidence intervals for unfavorable outcome in daytime and night-time surgery are quite broad, and cannot exclude the possibility that an effect is present. Evaluation of a larger patient population may also allow us to draw stronger conclusions regarding some secondary outcomes, such as the impact of body mass index and diagnosis, which were limited in our analysis by the number of missing data and number of categories, respectively.

In summary, our study found that patients undergoing night-time emergent neurosurgery were not at higher risk of unfavorable neurological outcome at hospital discharge or 1 month postdischarge compared with those undergoing emergent surgery during the day. However, this study does not provide a definitive answer to this contentious topic. Our findings cannot exclude the possibility of an association between timing of surgery and outcomes given the limitations of sample size, confounding variables that could not be assessed, and lack of congruence between handover times and our time category cutoffs. Further well-designed prospective trials are warranted to confirm our findings.


The authors thank Brenden Dufault, MSc (Biostats Consultant for Rady Faculty of Health Sciences, University of Manitoba, Winnipeg, MB, Canada) for assistance with initial statistical analysis, Roberta Giltrow for assistance with data retrieval, and Christine Hasse for assistance with accessing medical charts in the medical records department.


1. Schwarze ML, Barnato AE, Rathouz PJ, et al. Development of a list of high-risk operations for patients 65 years and older. JAMA Surg. 2015;150:325–331.
2. Fatigba HO, Tove MKS De, Tchaou BA, et al. Surgical management of head trauma: problems, results, and perspectives at the Departmental Teaching Hospital of Borgou, Benin. World Neurosurg. 2013;80:246–250.
3. Nakashima H, Tetreault LA, Nagoshi N, et al. Does age affect surgical outcomes in patients with degenerative cervical myelopathy? Results from the prospective multicenter AOSpine International study on 479 patients. J Neurol Neurosurg Psychiatry. 2016;87:734–740.
4. Munivenkatappa A, Agrawal A, Shukla DP, et al. Traumatic brain injury: does gender influence outcomes? Int J Crit Illn Inj Sci. 2016;6:70–73.
5. Nandyala SV, Marquez-lara A, Fineberg SJ, et al. Perioperative characteristics and outcomes of patients undergoing anterior cervical fusion in July: analysis of the “July effect”. Spine (Phila Pa 1976). 2014;39:612–617.
6. Turrentine FE, Wang H, Young JS, et al. What is the safety of nonemergent operative procedures performed at night? A study of 10,426 operations at an academic tertiary care hospital using the American College of Surgeons national surgical quality program improvement database. J Trauma. 2010;69:313–319.
7. Mullen MG, Michaels AD, Mehaffey JH, et al. Risk associated with complications and mortality after urgent surgery vs elective and emergency surgery implications for defining “quality” and reporting outcomes for urgent surgery. JAMA Surg. 2017;22908:768–774.
8. Bohl DD, Fu MC, Gruskay JA, et al. “July effect” in elective spine surgery. Spine (Phila Pa 1976). 2014;39:603–611.
9. Araujo RLC, Karkar AM, Allen PJ, et al. Timing of elective surgery as a perioperative outcome variable: analysis of pancreaticoduodenectomy. HPB (Oxford). 2014;16:250–262.
10. Chacko AT, Ramirez MA, Ramappa AJ, et al. Does late night hip surgery affect outcome? J Trauma. 2011;71:447–453.
11. Asfour L, Asfour V, McCormack D, et al. In surgeons performing cardiothoracic surgery is sleep deprivation significant in its impact on morbidity or mortality? Interact Cardiovasc Thorac Surg. 2014;19:479–487.
12. Lonze BE, Parsikia A, Feyssa EL, et al. Operative start times and complications after liver transplantation. Am J Transplant. 2010;10:1842–1849.
13. Linzey JR, Burke JF, Sabbagh MA, et al. The effect of surgical start time on complications associated with neurological surgeries. Neurosurgery. 2017;83:1–7.
14. Zafar SN, Libuit L, Hashmi ZG, et al. The sleepy surgeon: does night-time surgery for trauma affect mortality outcomes? Am J Surg. 2015;209:633–639.
15. Switzer JA, Bennett RE, Wright DM, et al. Surgical time of day does not affect outcome following hip fracture fixation. Geriatr Orthop Surg Rehabil. 2014;4:109–116.
16. Lee CS, Nagy PG, Weaver SJ, et al. Cognitive and system factors contributing to diagnostic errors in radiology. Am J Roentgenol. 2013;201:611–617.
17. de Oliveira GS, Chang R, Fitzgerald PC, et al. The prevalence of burnout and depression and their association with adherence to safety and practice standards: a survey of United States Anesthesiology Trainees. Anesth Analg. 2013;117:182–193.
18. McKenna H, Wilkes M. Optimising sleep for night shifts. BMJ. 2018;360:j5637.
19. Patel SV, Groome PA, Merchant S J, et al. Timing of surgery and the risk of complications in patients with acute appendicitis: a population level case-crossover study. J Trauma Acute Care Surg. 2018;85:341–347.
20. Desai V, Gonda D, Ryan SL. The effect of weekend and after-hours surgery on morbidity and mortality rates in pediatric neurosurgery patients. J Neurosurg Pediatr. 2015;16:726–731.
21. Kelz RR, Freeman KM, Hosokawa PW, et al. Time of day is associated with postoperative morbidity. Ann Surg. 2008;247:544–552.
22. Boomer LA, Cooper ÃJN, Anandalwar ÃS, et al. Delaying appendectomy does not lead to higher rates of surgical site infections a multi-institutional analysis of children with appendicitis. Ann Surg. 2016;264:164–168.
23. Emmanouilidis N, Boeckler J, Ringe BP, et al. Risk balancing of cold ischemic time against night shift surgery possibly reduces rates of reoperation and perioperative graft loss. J Transplant. 2017;2017:5362704.
24. Matt V, Matthew H. The retrospective chart review: important methodological considerations. J Educ Eval Health Prof. 2013;10:12.
25. Wilson JT, Pettigrew LE, Teasdale GM. Structured interviews for the Glasgow Outcome Scale and the extended Glasgow Outcome Scale: guidelines for their use. J Neurotrauma. 1998;15:573–585.
26. Abella A, Hermosa C, Enciso V, et al. Effect of the timing of admission upon patient prognosis in the intensive care unit: on-hours versus off-hours. Med Intensiva. 2016;40:26–32.
27. De Souza Barbosa L, dos Reis GS Jr, Chaves RZ, et al. Night admission is an independent risk factor for mortality in trauma patients—a systemic error approach. Rev Col Bras Cir. 2015;42:209–214.
28. Fernandes S, Carvalho AF, Rodrigues AJ, et al. Day and night surgery: is there any influence in the patient postoperative period of urgent colorectal intervention? Int J Colorectal Dis. 2016;31:525–533.
29. Kristiansen NS, Kristensen PK, Nørgård BM, et al. Off-hours admission and quality of hip fracture care: a nationwide cohort study of performance measures and 30-day mortality. Int J Qual Health Care. 2016;28:324–331.
30. Tischenkel BR, Gong MN, Shiloh AL, et al. Daytime versus nighttime extubations: a comparison of reintubation, length of stay, and mortality. J Intensive Care Med. 2016;31:118–126.
31. Whitlock EL, Feiner JR, Chen LL. Perioperative mortality, 2010 to 2014: a retrospective cohort study using the national anesthesia clinical outcomes registry. Anesthesiology. 2015;123:1312–1321.
32. Bekelis K, Missios S, Mackenzie TA. Outcomes of elective cerebral aneurysm treatment performed by attending neurosurgeons after night work. Neurosurgery. 2018;82:329–334.
33. Beard JH, Martin ND, Reilly PM, et al. Nighttime trauma fellow care is associated with improved outcomes after injury. Am Surg. 2018;84:365–370.
34. Drake FT, Mottey NE, Castelli AA, et al. Time-of-day and appendicitis: impact on management and outcomes. Surgery. 2017;161:405–414.
35. Chiewvit P, Tritakarn S. Degree of midline shift from CT scan predicted outcome in patients with head injuries. J Med Assoc Thai. 2010;93:99–107.
36. Gillies MA, Lone NI, Pearse RM, et al. Effect of day of the week on short- and long-term mortality after emergency general surgery. Br J Surg. 2017;104:936–945.

neurosurgical procedures; postoperative complications; hospital mortality; night care

Supplemental Digital Content

Copyright © 2019 Wolters Kluwer Health, Inc. All rights reserved.