SPINE CENTER

UCDAVIS

Adult Spinal DeformitySurgical Complications and

Classification

Eric Klineberg, MDProfessor and Vice Chair

Department of OrthopaedicsUniversity of California, Davis

9th Annual UCSF Practical Course in Advanced Spinal Techniques

SPINE CENTER

UCDAVIS

I have no financial interest with any company

regarding this subject

Eric Klineberg, MD

Consulting: Depuy/Synthes, Stryker, Medicrea

Speaking: AO Spine

Fellowship Funding: AO Spine

SPINE CENTERUCDAVIS

Introduction

Surgical intervention can have a significant impact

Complications can be significant

SPINE CENTERUCDAVIS

Introduction

Deformity Surgery

– Considered to have higher risks

– Perioperative complications are frequent (up to 40%)

Glassman et al. Spine 2007

SPINE CENTERUCDAVIS

What is a complication?

com·pli·ca·tion noun \ˌkäm-plə-ˈkā-shən\ : something that makes something

harder to understand, explain, or deal with medical : a disease or condition that happens in addition to

another disease or condition : a problem that makes a disease or condition more dangerous or harder to treat

SPINE CENTERUCDAVIS

What is a complication?

Does it matter?

Complication List Infection Majoro Deep, Pneumonia, Sepsis Minoro Superficial, UTI, C Diff infection Implant Majoro Hook dislodgement, Interbody fracture/migration, Rod fracture/dislodgement, Screw fracture Minoro Painful/promininent, Screw malposition/loosening, Interbody subsidence/dislodgement

Radiographic Majoro DJK, PJK, Pseudoarthrosis Minoro Coronal/Sagittal imbalance, Curve decompensation, HO, Adjacent segment degeneration Neurologic Majoro Visual deficit/blindness, Brachial plexus injury, CVA/Stroke, Spinal cord injury, Nerve root injury with weakness, Retrograde ejaculation, Bowel/Bladder deficit Minoro Neuropathy or sensory deficit, Pain (radiculopathy), Peripheral nerve palsy, Delirium Mortality All majorCardiopulmonary Majoro Cardiac arrest, PE, Respiratory arrest, DVT, MI, Reintubation, ARDS Minor o Coagulopathy, Arrhythmia, Pleural effusion, Hypotension, CHF

Gastrointestinal Majoro Obstruction, Perforation, Bleed requiring surgery, Pancreatitis/Cholecystitis requiring surgery, Liver Failure, SMA Syndrome Minoro Ileus, Bleed not requiring surgical intervention, Pancreatitis/Cholecystitis no surgery

Renal Majoro Acute Renal failure requiring dialysis Minoro Acute Renal failure requiring medical intervention Operative Majoro Retained sponge/instrument, Wrong surgical level, Unintended extension of fusion, Vascular injury, Visceral injury, EBL >4L Minoro Dural tear, Fixation failure (hook/screw), Pedicle fracture, Posterior element fracture, Vertebral body fracture Wound Problems Majoro Dehiscence requiring surgery, Hematoma/seroma requiring surgery +/- neurological deficit, Incisional hernia Minoro Hematoma/seroma not requiring surgery, Hernia

SPINE CENTERUCDAVIS

INTRODUCTION

Glassman et al– major and minor complications did not adversely

effect the improvement found in the HRQOL measures

– except for deterioration in the SF-12 for major complications.

Theorized that outcome instruments were not sensitive enough to detect a difference

Perioperative complications may not have a continued impact at one year.

SPINE CENTERUCDAVIS

What is a complication?

Physician and patient dependent

SPINE CENTERUCDAVIS

Prevention

Medical Optimization– Cardiac

– Pulmonary

– Nutritional

– Metabolic

– Bone Quality

– What about consent?

SPINE CENTERUCDAVIS

Informed Consent

Despite ranking the consent process as important, patient recall was only 41% immediately after discussion and video re-enforcement.

Recall subsequently declined to 20% at 6 months post-operatively.

SPINE CENTERUCDAVIS

Prevention

Medical Optimization– Cardiac

– Pulmonary

– Nutritional

– Metabolic

– Bone Quality

SPINE CENTERUCDAVIS

Surgical Strategy

SPINE CENTERUCDAVIS

Surgical Strategy

SPINE CENTERUCDAVIS

Surgical Strategy

SPINE CENTERUCDAVIS

Complication CategoryPeri-op (<6wks)

minor/major(%)

Delayed (>6wks)minor/major

(%)

Totalminor/major

(%)

Implant 3/8 (3.8) 11/59 (24.1) 14/67 (27.8)Radiographic 4/10 (4.8) 25/42 (23.0) 29/52 (27.8)Neurologic 21/24 (15.5) 16/20 (12.4) 37/44 (27.8)Operative 41/32 (25.1) 0/1 (0.3) 41/33 (25.4)Cardiopulmonary 31/20 (17.5) 1/3 (1.4) 32/23 (18.9)Infection 11/20 (10.7) 5/7 (4.1) 16/27 (14.8)Gastrointestinal 24/1 (8.6) 0/0 (0) 24/1 (8.6)Wound (excluding infection) 3/7 (3.4) 0/5 (1.7) 3/12 (5.2)Vascular 4/0 (1.4) 1/0 (0.3) 5/0 (1.7)Musculoskeletal 0/0 (0) 3/0 (1.0) 3/0 (1.0)Renal 1/2 (1.0) 0/0 (0) 1/2 (1.0)Other 2/1 (1.0) 0/0 (0) 2/1 (1.0)Total (minor/major) 270 (145/125) 199 (62/137) 469 (207/262)Mean # complications/patient (minor/major)

0.93 (0.50/0.43) 0.68 (0.21/0.47) 1.61 (0.71/0/90)

Number of patients affected (%)150 (51.5) 124 (42.6) 203 (69.8)

Results: 246 patients with 2 year f/u

SPINE CENTERUCDAVIS

Complication CategoryPeri-op (<6wks)

minor/major(%)

Delayed (>6wks)minor/major

(%)

Totalminor/major

(%)

Implant 3/8 (3.8) 11/59 (24.1) 14/67 (27.8)

Radiographic 4/10 (4.8) 25/42 (23.0) 29/52 (27.8)

Neurologic 21/24 (15.5) 16/20 (12.4) 37/44 (27.8)

Operative 41/32 (25.1) 0/1 (0.3) 41/33 (25.4)

Cardiopulmonary 31/20 (17.5) 1/3 (1.4) 32/23 (18.9)

Infection 11/20 (10.7) 5/7 (4.1) 16/27 (14.8)

Gastrointestinal 24/1 (8.6) 0/0 (0) 24/1 (8.6)

Wound (excluding infection) 3/7 (3.4) 0/5 (1.7) 3/12 (5.2)

Vascular 4/0 (1.4) 1/0 (0.3) 5/0 (1.7)

Musculoskeletal 0/0 (0) 3/0 (1.0) 3/0 (1.0)

Renal 1/2 (1.0) 0/0 (0) 1/2 (1.0)

Other 2/1 (1.0) 0/0 (0) 2/1 (1.0)

Total (minor/major) 270 (145/125) 199 (62/137) 469 (207/262)

Mean # complications/patient (minor/major)0.93 (0.50/0.43) 0.68 (0.21/0.47) 1.61 (0.71/0/90)

Number of patients affected (%) 150 (51.5) 124 (42.6) 203 (69.8)

SPINE CENTERUCDAVIS

Can We Develop A Better Complication Score?

We rely on AE/Minor/Major determination– No consensus

– Severity of complication may be biased

Can a less biased score better predict HRQoL outcomes?

SPINE CENTERUCDAVIS

CMS

Increased interest in complications and when they occur

All complications that occur within 30 days from the operation

All readmission/reoperations that occur within 90 days

May have significant impact with bundling of payments

SPINE CENTERUCDAVIS

Canadian (SAVES)

Have led the way with the development of a intervention severity score– Use a scale from I-IV to determine severity (or grades 1-6)

– Also assign a Length Of Stay modifier

– Do not have specific score for neurology, readmission or reoperation

– How we obtain the information is critical

SPINE CENTERUCDAVIS

AO Spine/Scoli-Risk-1

Gathers info for non-neurologic complications

Granular information regarding neurologic injury

Defines the neurologic injury – (cord, motor, sensory, incontinence etc…)

– Level of injury

Describes timing, intervention, and outcome

SPINE CENTERUCDAVIS

Factor that predicted 2-year SF-36PCS – Age (p < .001), ASA grade (p < .001)– Maximum preoperative Cobb angle (p = .007)– Number of three-column osteotomies (p = .049)– Type of neurologic complication (p = .068)

Factors predictive of 2-year SRS-22R Total scores Maximum preoperative Cobb angle (p = .001) Number of serious adverse events (p = .071)

SPINE CENTERUCDAVIS

Do Complications Effect HRQoL?

355 pts prospectively enrolled in the ISSG multicenter study

202 met the inclusion criteria

Mean age 57.4, levels fused 12

Four groups identified:– No Complications N=84

– Minor Complications N=87

– Major Complications N=65

– Both Major and Minor N=35

SPINE CENTERUCDAVIS

Baseline Pre-OP Demographics

Similar distribution for Age, BMI, and ASA, as well as Pre-OP spinopelvic parameters.

Sig lower Charlson Comorbidity Index for the no complication group.

No Complications Minor Major Both p‐value Age 55.2 57.7 61.1 58.8 0.072 BMI 26.9 27.3 28.1 28.4 0.487 ASA 2.2 2.4 2.4 2.4 0.06 Charlson 1.2 1.9 2.0 1.9 0.015* Smoker (%) 6 11 8 11 0.693 SVA (mm) 45.6 53.9 68.6 68.5 0.217 Max Cobb (Degrees) 41.5 45.0 41.9 44.2 0.689 Prior Spine Fusion Surgery (%)

75.0 73.0 80.6 70.6 0.853

SPINE CENTERUCDAVIS

Operative Summary

Trend towards > PSO for Major and Both complication groups

No complication group also had the lowest percent of BMP, anterior approach, EBL and Time in the OR.

– May be a surrogate for surgical complexity.

No Complications Minor Major Both p‐value Levels Fused 12.0 11.9 12.3 12.4 0.825 Osteotomy (%) 71.1 55.6 71.4 73.0 0.997 PSO/PVCR (%) 22.9 21.1 31.7 29.7 0.413 BMP (%) 51.8% 86.7% 86.5% 69.8% 0.0001 Anterior (%) 14.5 30.0 30.2 40.5 0.013 EBL (cc) 1783 2061 2698 2704 0.005* OR Time (min) 412 494 517 533 0.0001** Length of Stay (Days) 8.0 8.9 10.5 9.9 0.073

SPINE CENTERUCDAVIS

Baseline/1 Year HRQoL

All

ComplicationNo

ComplicationMajor Minor Both Pvalues

BaselineODI(Std)

42.5(19.6)

41.3(19.5)

46.4(17)

39.5(19.5)

42.5(16.9)

NS

1yearODI(Std)

28.3(20.2)

26.6(18.6)

29.9(20)

26.9(20.0)

28.1(19.5)

NS

BaselinePCS(Std)

32.9(10.3)

32.9(9.75)

31.1(8.8)

33.9(10.3)

31.8(9.9)

NS

1yearPCS(Std)

39.5(11.1)

41.3(10.9)

38.0(12)

40.7(10.8)

39.8(11.3)

NS

Significant improvement in All groups from Baseline to 1 year

No differences between groups for any of the outcome measures, regardless of complication

SPINE CENTERUCDAVIS

1 Year HRQoL

Significant impact on ODI and PCS for readmission, reoperation and no complication resolution.

NoReadmission Readmission PValue1yearODI 24.5 39.5 P<0.011yearPCS 41.3 31.9 P<0.01 NoReoperation Reoperation 1yearODI 24.8 37.1 P<0.011yearPCS 41.1 33.9 P<0.01 ResolutionOf

Complication NoResolution

1yearODI 24.5 39.5 P<0.011yearPCS 41.3 31.9 P<0.01

SPINE CENTERUCDAVIS

ISSG/AO/ESSG

Working to develop a comprehensive score

Using: – 1. Complication Category

– 2. Intervention severity

– 3. Complication Severity

– 4. Neurologic severity

– 5. Reoperation/readmission

– 6. Resolution of complication

– 7. Timing/Effect on LOS

SPINE CENTERUCDAVIS

Complication Grading System

Complication Score

0 1 2 3

Severity Adverse Event

Minor Major Death

Intervention None Non-Invasive Invasive Surgical

Neurologic Sensory Motor Bowel/BladderSpinal Cord

InjuryImpact on Length

of StayNone <2 days 3-7 days >7 days

Readmission No Yes

Revision Surgery No Yes

Resolution Resolved Unresolved

Timing Intra-op In HospitalEarly Post-Op

(<90 days)Late Post OP (>3 mo - 1 yr)

SPINE CENTERUCDAVIS

Application of the system

Different components of the score could be used for different outcome metrics

Ie: LOS vs HRQoL

SPINE CENTERUCDAVIS

Complication Impact on LOS

Univariate analysis identified factors that correlated with increase over predicted LOS: – cumulative complication severity (OR 1.23, p=.0001)

– cumulative intervention severity (OR 1.15, p=0.0001)

– number of complications (OR 1.26, p=.02)

Development of a model to predict hospital LOS based on complications– Actual LOS was sigificantly higher than predicted LOS (10.7 days vs

8.3 days, p=0.0001)

SPINE CENTERUCDAVIS

Impact on HRQoL at 2 years

Minimum one complication had lower 2-yr improvements in HRQL– (SF-36 PCS 6.91 vs 9.48, p=.012, and SRS-22r 0.79 vs 0.95,

p=.03).

Number of complications – (PCS -0.1159, p=.016, SRS -0.0929, p=.048)

SPINE CENTERUCDAVIS

Impact on HRQoL at 2 years

Severity Score: – maximum severity score (PCS -0.1157, p=0.016)

– cumulative severity score (PCS -0.1223, p=.011, SRS -0.1487, p=.03)

Intervention Score:– Maximum intervention score (PCS -0.16, p=.001, SRS -0.125, p=.008)

– Cumulative Intervention Score (PCS 0.1245, p=.0096)

Complication resolution:– resolved complication PCS -2.22, p=.048,

– unresolved complication PCS -3.12 p=.012

SPINE CENTERUCDAVIS

Spine Complication Classification

A simple classification system with discrete data points

A more comprehensive one with additional data points and subgroups that captures more granular data.

Determining what data points need to be acquired is our first challenge

SPINE CENTERUCDAVIS

Complication Category

Each complication receives a categorical letter and sub-letter to define its primary complication category

Then each complication is stratified into the four complication modifiers: – neurologic

– timing

– intervention severity

– resolution

SPINE CENTERUCDAVIS

Universal Spine Complication Classification

Neurological

Timing

Intervention

Resolution

Classification

SPINE CENTERUCDAVIS

Neurological

LEMS Score

Severity

No Deficit

Sensory only +/- Pain

Motor+/- Impact on ambulatory

status

Spinal Cord Injury

+/- Impact on ambulatory

status

+/- Impact on bowel/bladder

function

Neurologic Sub Score

SPINE CENTERUCDAVIS

Timing (of complication diagnosis)

Intraoperative

In-Hospital

+/- Reoperation

+/- <30 days

Post-discharge

<30 days

+/-Readmission

+/- Reoperation

30-90 days

+/-Readmission

+/- Reoperation

>90 days

+/-Readmission

+/- Reoperation

Timing Sub Score

SPINE CENTERUCDAVIS

Intervention

(choose highest)

Mild

consultation, lab values, diagnostic imaging, small

needle intervention (contrast, epidural, transfusion)

Moderate

large needle intervention (PICC line, chest tube,

angiocath, dialysis), cardioversion

Severe

Surgical treatment (knife intervention)

Intervention Sub Score

SPINE CENTERUCDAVIS

Resolution status

Completeresolution

Partial resolution

Unresolved (unchanged)

Death

Resolution Sub Score

SPINE CENTERUCDAVIS

Universal Spine Complications Classification

Medical

Neurological

Timing

Intervention

Resolution

Surgical

Neurological

Timing

Intervention

Resolution

ISSG/AO/ESSG

SPINE CENTERUCDAVIS

Validation

Identification and classification of complications can be difficult, and simple categories will improve our ability to classify and quantify the impact of complications.

Intrinsic surgeon bias may increase accuracy of reporting for some complications more then others– Ie better reporting of surgical vs medical complications

SPINE CENTERUCDAVIS

Methods

10 randomized cases were sent to participants, and they were asked to identify the complications and complete a standardized data collection form.

There were 34 events that occurred: – 25 events with only one complication

– 5 with 2 complications

– 4 with 2 or more complications

Cat N % Cases with Mean StD

Gastro 2 20% 0.2 0.421637

Musculoskeletal 3 30% 0.3 0.483046

CNS 3 30% 0.3 0.483046

Cardiac 5 20% 0.5 1.080123

Pulmonary 1 10% 0.1 0.316228

Renal 1 10% 0.1 0.316228

Radiographic 4 30% 0.4 0.699206

Neurologic 7 50% 0.7 0.948683

Operative 8 70% 0.8 0.632456

Wound/Approach 2 10% 0.2 0.632456

Implant 1 10% 0.1 0.316228

SPINE CENTERUCDAVIS

Results

17 people filled out all questionnaires: – 10 attending surgeons, 5 trainees, and 2 research

coordinators.

Overall accuracy– 87.4% high level (i.e. neurologic vs gastrointestinal vs cardiac

etc.)

– 75.7% with more granular data (i.e. motor deficit vs ileus vs MI etc).

Accuracy for medical and surgical complications is similar– (87.6% vs 87.1% for high level, 77.4% vs 74.3% for detail).

SPINE CENTERUCDAVIS

Results

Highest overall accurate rate– CVA, gastrointestinal and

radiographic (above 94%)

Lowest overall accurate rate – renal (44.8%), pulmonary

(54.5%) cardiac (55%).

Overall event accuracy (combination of complications occurring simultaneously) is 57.1%.

 

HighLevel 2 Detail Level 3

 Gastro  94.1% 94.1%

 Musculoskeletal 80.6% 80.6%

 CNS  100.0% 98.1%

 Cardiac  88.2% 55.0%

 Pulmonary 54.5% 54.5%

 Renal  81.3% 44.8%

Radiographic 96.2% 96.2%

Neurologic 77.9% 66.9%

Operative  89.7% 79.6%

Wound/Approach 100.0% 65.6%

Implant  81.3% 81.3%

 

SPINE CENTERUCDAVIS

Results

Neurologic impairment per event was accurate for 79.1%.

Intervention severity is 79.6% accurate, with the highest rate for severe intervention (98.6%).

Resolution was accurately reported for 70.3% of the events– 80.1% for Resolved

– 42.9% for Unresolved

SPINE CENTERUCDAVIS

Conclusions

Accurate reporting and gathering of complications is difficult to standardize.

In this cased based survey, complex complications were categorized accurately 87%, neuro deficits accurately 79%, intervention accuracy of 80% and resolution accuracy of 70%.

Surgeons need to be actively involved in complication reporting to enhance accuracy.

SPINE CENTERUCDAVIS

Does this system help us?

What is the effect / incidence of timing?

What is the effect on HRQL?

Can it predict LOS?

SPINE CENTERUCDAVIS

Background: Timing of complication

The timing and impact of complications over time is important to understand for patients, payors and providers. While most medical and operative complications occur proximate to the index surgical intervention, complications may occur at any time point during the care of our adult spinal deformity patients.

Understanding the timing of specific complications may be helpful to guide patients and surgeons. The impact of those complications on health outcomes at 2 years is also critically important.

SPINE CENTERUCDAVIS

CompsIntra Operative Before discharge After discharge

Sum % Event Sum % Event Sum % Event

Adverse Event 43 28.9% 73 47.1% 71 23.3%

Cardiopulmonary 14 9.4% 41 26.5% 20 6.6%

Gastrointestinal 3 2.0% 37 23.9% 4 1.3%

Implant 6 4.0% 1 0.6% 111 36.4%

Infection 1 0.7% 21 13.5% 31 10.2%

Neurologic 19 12.8% 19 12.3% 61 20.0%

Operative 97 65.1% 7 4.5% 6 2.0%

Other 0 0.0% 1 0.6% 2 0.7%

Radiographic 0 0.0% 1 0.6% 130 42.6%

Renal 0 0.0% 3 1.9% 0 0.0%

Wound 1 0.7% 2 1.3% 7 2.3%

Before 30D Between 30 and 90D After 90D

Sum % Event Sum % Event Sum % Event

Adverse Event 82 55.0% 12 7.7% 55 18.0%

Cardiopulmonary 51 34.2% 7 4.5% 3 1.0%

Gastrointestinal 39 26.2% 1 0.6% 1 0.3%

Implant 6 4.0% 6 3.9% 100 32.8%

Infection 39 26.2% 11 7.1% 3 1.0%

Neurologic 27 18.1% 18 11.6% 36 11.8%

Operative 10 6.7% 1 0.6% 2 0.7%

Other 2 1.3% 0 0.0% 1 0.3%

Radiographic 11 7.4% 21 13.5% 99 32.5%

584/732 patients met inclusion criteria (mean age 58.6yrs, 78% female, mean BMI 27.5, mean CCI 1.64, mean ODI 43.5).70.9% had least one complication event over the 2-year period, with an average of 1.45 events per patient.

SPINE CENTERUCDAVIS

Complications over time….

Early

Continue

SPINE CENTERUCDAVIS

All Complications: Relationship to Timing

All Complications

SPINE CENTERUCDAVIS

Propensity Matching

grp Mean p

BE

FO

RE

MA

TC

HIN

G

BL_ODI 0 40.340.001

1 45.85

BL_PCS 0 33.920.000

1 30.62

demo_AgeBase0 55.86

0.0011 60.31

BL_Frailty_Index0 2.909

0.0001 3.504

LATpre_PI_LL0 12.54

0.0031 18.08

AF

TE

R M

AT

CH

ING

BL_ODI 0 42.150.791

1 41.66

BL_PCS 0 32.840.878

1 33.00

demo_AgeBase0 57.69

0.9381 57.81

BL_Frailty_Index0 3.054

0.7901 3.013

LATpre_PI_LL0 14.44

0.5871 15.62

grp Mean p

Y2_ODI 0 23.430.104

1 26.72

Y2_PCS 0 42.600.043

1 40.24

Y2_MCS 0 51.170.925

1 51.29

Y2_SRS_ACTIVITY 0 3.750.168

1 3.61

Y2_SRS_PAIN 0 3.650.055

1 3.44

Y2_SRS_APPEARANCE 0 3.780.143

1 3.64

Y2_SRS_MENTAL 0 3.900.933

1 3.91

Y2_SRS_SATIS 0 4.300.093

1 4.13

Y2_SRS_TOTAL 0 3.820.133

1 3.70

Y2_SF36_PF 0 41.090.015

1 38.20

Y2_SF36_RP 0 42.530.522

1 41.72

Y2_SF36_BP 0 45.830.053

1 43.56

Y2_SF36_GH 0 49.300.347

1 48.19

Y2_SF36_VT 0 49.730.372

1 48.66

Y2_SF36_SF 0 45.880.680

1 45.37

Y2 SF36 RE 0 46 62

• With only one complication: regardless of type had worse final outcomes then no comps

• No complication:• ODI (40 to 22, p=0.01)• PCS (33.9 to 43, p=0.05)

• One complication • ODI (45.8 to 30.5, p=0.05)• PCS (30.6 to 38, p=0.05)

When we sub-analyze for type of complication, those that occur early have

minimal effect, while those that occur later have a much more significant effect.

SPINE CENTERUCDAVIS

Conclusion: Timing

Complications occur over-time and can be predicted by type.

Implant, radiographic and neurologic complications continue to occur over time, and need to be followed closely.

Complication type is critical, and those complications that occur later, and increase over-time are more impactful for our patients at 2 years.

Determining the relationship of the timing of complications and its impact to our patients is critical to understand.

SPINE CENTERUCDAVIS

LOS: Hypothesis

Investigate the role of complications that occur during the initial hospitalization to predict LOS based on a novel classification that includes treatment severity.

SPINE CENTERUCDAVIS

LOS: distribution

This parameter is not normally distributed

– Kolmogorov-Smirnov p = 1.1737E-27

Comparison with poison distribution

– Kolmogorov-Smirnov p = 0.103964

SPINE CENTERUCDAVIS

LOS Parameters

List of surgical parameters simplified for abstract analysis– Posterior length of fusion:

“short” versus “medium” versus “long” fusion

– Threshold for short < 5

– Threshold for long > 13

– Major osteo versus no

– IBF versus no IBF

– Primary versus revision

– Stage yes/no

– Posterior only vs combined

N Mean StD Min Max 25th 50th 75th

Length of fusion

Short 102 4.24 2.09 1 11 3 4 5

Medium 407 6.43 2.01 2 12 5 6 7

Long 138 6.88 1.93 2 12 6 7 8

3 column ostotomy

No Major 574 6.07 2.24 1 12 5 6 7

Major 86 6.57 1.84 3 12 5 6 8

Interbody fusion

No 242 5.95 1.71 1 12 5 6 7

Yes 414 6.22 2.41 1 12 4.75 6 8

revision

Primary 476 6.17 2.24 1 12 5 6 7

Revision 184 6.05 2.10 1 12 5 6 7

Approach

Posterior Only 426 5.95 1.79 1 12 5 6 7Anterior-Posterior (APSF) 223 6.58 2.72 1 12 4 7 9

Stage

Same Day 492 5.79 1.97 1 12 5 6 7

Staged 109 8.11 2.13 2 12 7 8 9.5

SPINE CENTERUCDAVIS

LOS Demographics

494 patients included in the analysis– Mean age: 61 year

– 73.8% female

– 28.07 kb/m2

Mean ASA grade was 2.44– 45.3% Grade 2

– 45.7% Grade 3

Mean number of levels fused posterior:– 11.6 +/- 3.9

Mean number of levels fused using IBF– 2.5 +/- 1.6

77.3% underwent some type of osteotomy– 26.1% underwent a major osteotomy (PSO / VCR)

78.7% same day surgery

SPINE CENTERUCDAVIS

LOS by Intervention

During the hospital stay– 65.1% of the maximum intervention where minor

– 10.1% had at least one moderate intervention

– 10.5% had at least one severe intervention

Reop rate: 9.5%

Number of events per patient between surgery and discharge– Mean Number events: 1.7 +/- 1.1

SPINE CENTERUCDAVIS

Multivariate Analysis

4 independent predictors identified – Group posterior fusion (short being reference)

Medium (p = 5.0798E-10)

Long (p = 8.7728E-12)

– Major Osteo (No Major being reference) Use of Major (p = 0.000986)

– Stage (Same day being reference) Stage (p = 0.0E0)

– Intervention (No complication being reference) No intervention (p = 0.000219)

Minor intervention (p = 0.000004)

Moderate intervention (p = 0.000006)

Severe intervention (p = 0.006724)

SPINE CENTERUCDAVIS

Predicators of LOS

3 parameters are significant independent predictor of LOS– Posterior fusion length group

– Stage yes/no

– Intervention

SPINE CENTERUCDAVIS

Conclusion: LOS

LOS is correlated to in-hospital complications and to complication intervention severity.

Surgical factors that affect LOS included length of fusion, major osteotomy or need for staged surgery.

Increased invasiveness of complication treatment was identified by a novel complication severity assessment scale as the only non-surgical factor that independently predicted increased hospital LOS following ASD surgery.

SPINE CENTERUCDAVIS

Much left to understand

Relationship of complication to HRQoL measure (ie timing)– Likely a correlation, and effect of complication weaken with time

– Complication that has no effect now, but does later

Cost of complication– May use scoring system

Consensus for component score– AO Spine, ISSG, ESSG, Canadians, others….

Development of a complication score

SPINE CENTERUCDAVIS

• Although surgical treatment for ASD can improve pain and disability, it is associated with high rates of complications.

• Many complications likely have minimal or no impact on ultimate patient outcome at 2 years• But may have impact on LOS, cost, recovery time

• No classification is currently able to predict LOS or HRQoL

• Can a comprehensive scores better classify complications for us and our patients?

Conclusion