Secondary Logo

Journal Logo


Concordance Between Patients’ and Surgeons’ Expectations of Lumbar Surgery

Mancuso, Carol A. MDa,b; Duculan, Roland MDa; Cammisa, Frank P. MDa; Sama, Andrew A. MDa; Hughes, Alexander P. MDa; Lebl, Darren R. MDa; Yang, Jingyan DrPHc; Ghomrawi, Hassan M.K. PhD, MPHd; Girardi, Federico P. MDa

Author Information
doi: 10.1097/BRS.0000000000003775
  • Free

The decision to undergo lumbar surgery is driven by patients’ desires to relieve symptoms, improve quality of life, and avoid further disability, and by surgeons’ perspectives that there likely will be benefit in these areas.1 However, patients may have unrealistic outcome expectations that are either too high or too low.2 Both scenarios are undesirable and predispose to poor outcomes if patients become discouraged with recuperation time, abandon rehabilitation, and ignore recommended lifestyle modifications that minimize disease progression.2,3 To achieve the maximum benefit from surgery, patients and surgeons need to mutually acknowledge what is possible, probable, and realistic and to work toward the same goals.4 From the perspective of process of care, successfully achieving shared goals demonstrates effective patient–physician communication and quality of care if patients are satisfied with outcomes and future health status is optimized.5,6

We conducted a series of studies demonstrating that patients expect physical and psychological improvement from lumbar surgery.7,8 We also showed that the proportion of expectations fulfilled is a measurable and valid patient-centered outcome that is associated with pre- and postoperative characteristics.9 The objectives of this longitudinal study were to measure concordance between patients’ and surgeons’ preoperative expectations of surgery and then to determine which member of the dyad more closely predicted fulfillment of expectations, defined as patient-reported status postoperatively.


This study was approved by the Institutional Review Board at Hospital for Special Surgery; all patients provided written informed consent.

Preoperative Information From Patients

Patients were eligible if they were scheduled for elective lumbar surgery by one of five spine-specialty surgeons. Patients were excluded if they did not speak English or refused to participate. Patients were enrolled and interviewed during routine preoperative assessments and completed standard questionnaires assessing spine-related disability,10,11 depressive and anxiety symptoms,12,13 and expectations based on a valid 20-item lumbar surgery-specific survey. Unlike previous surveys, this expectations survey was patient-derived based on qualitative–quantitative methodology.2,7,14,15 The survey asks patients “how much improvement do you expect” for symptoms (e.g., interference with sleep), physical well-being (e.g., stand more than half an hour), and psychological well-being (e.g., remove the control my spine has on my life).7 Response options for all items range from “complete improvement/back to normal” (4 points), to “I do not have this expectation/this expectation does not apply to me” (0 points). A score is the sum of points divided by the maximum possible points; range 0–100, higher score is greater expectations. The survey is valid and reliable and provides a patient-specific measure of outcome (described below).9 Unlike other measures,2,14,15 this survey also has evaluative properties that can distinguish amount of improvement received versus what was expected. At enrollment, patients also reported symptom duration, back/leg pain severity, previous spine therapies, and comorbidity.16

Preoperative Information From Surgeons

Before surgery, surgeons independently completed the physicians’ version of the survey which is identical to the patients’ version except the surgeon is asked “how much improvement do you expect for your patient” for each item. Surgeons completed the survey for each patient within several days of when patients completed their surveys; scores were similarly generated. Surgeons were blinded to their patients’ responses.

Clinical Information

We reviewed medical records for diagnosis and physical examination, and intraoperative reports to rate surgical complexity according to a standard index based on surgical approach, decompression, fusion, instrumentation, and number of vertebrae involved.17

Postoperative Information From Patients

After 2 years, patients were interviewed by telephone to ask about outcomes. The timeframe of two years was chosen to ensure recuperation and return to valued activities for most patients. For each expectations survey item cited preoperatively, patients were asked “how much improvement did you receive?” with response options of “complete improvement” (4 points), to “no improvement” (0 points). The proportion of expectations fulfilled was calculated by comparing the amount of improvement received postoperatively to the amount of improvement expected preoperatively (described below). This proportion is a valid patient-centered outcome when compared to pre- to postoperative change in symptoms, function, disability, and satisfaction with surgery.9

Patients were asked about current pain and disability, any interval spine-related events (i.e., repeat surgery, hospitalization), and new or worse comorbidity.

Preoperative: Primary Analysis

The main analysis was a comparison of survey scores within each patient-surgeon pair based on the intraclass correlation coefficient (ICC), which measures concordance when concurrent evaluations of the same phenomenon are compared. ICC values range from 0 to 1 and using a threshold value of 0.20 to demarcate different categories of concordance, ICC values are considered poor (0–0.20), fair (0.21–0.40), moderate (0.41–0.60), good (0.61–0.80), and excellent (0.81–1).18 Additionally, multivariable linear regression models were derived with survey scores as dependent variables and demographic and clinical characteristics as independent variables. After verifying linearity assumptions were fulfilled, variables associated in bivariate analyses were initially included, and final models were derived based on backward stepwise elimination. Goodness of fit was assessed using residual plots which, when distributed symmetrically, confirm a linear regression model is appropriate for the data.

The sample size was based on the hypothesis that the ICC would be greater for nondegenerative versus degenerative diagnoses by a difference of 0.20; this value was determined by comparing scores according to diagnoses during development of the survey.7 In another previous study assessing concordance between patients’ and surgeons’ expectations for total hip/knee arthroplasty, the ICC standard deviation was 0.95.19 We assumed a narrower value of 0.70 for the present study because the participating surgeons were attuned to expectations from their participation in a previous study.8,9 For a difference between groups of 0.20, alpha 0.05, and power 0.80, three hundred eighty-four patients were required. To allow for surgeons not completing some surveys before surgery and some patients not participating in the follow-up (total estimated 12%), the targeted enrollment was 430 patients.

Postoperative: Secondary Analysis

The secondary analysis was the proportion of patients expectations fulfilled defined as amount of improvement received versus amount of improvement expected.

To ascertain whether surgeons’ preoperative expectations more closely coincided with patients’ postoperative reports of improvement, we also assessed the surgeon proportion. The proportions were calculated according to the following similar equations.PatientProportion=patient-reportedpostopimprovementreceivedpatient-reportedpreopimprovementexpected=sumofpatients'postopsurveypointssumofpatients'preopsurveypointsSurgeonProportion=patient-reportedpostopimprovementreceivedsurgeon-reportedpreopimprovementexpected=sumofpatients'postopsurveypointssumofsurgeons'preopsurveypoints

A proportion of 0.60 was shown previously to have a sensitivity of 0.90 and a specificity of 0.79 compared to a five-point ranked global rating of satisfaction (i.e., very satisfied to very dissatisfied) and was considered a clinically important threshold for expectations fulfilled.9 In addition, the proportion varied by a mean value of 0.20 between each pair of adjacent satisfaction categories (P < 0.0001) and was considered a minimum clinically important difference (MCID).9

In the present study, proportions were assessed as dependent variables in multivariable linear regression models with demographic and clinical characteristics after verifying linearity of associations and then confirming symmetry in residual plots.


In total, 539 consecutive patients were screened from December 2013 through February 2016; of these, 38 were missed, 10 were not eligible, 54 refused, and 437 were enrolled and completed baseline assessments (Figure 1). Surgeons completed their version of the survey for 419 patients, four of these ultimately did not have surgery; thus there were 415 pairs of patient–surgeon expectations surveys for comparison. Postoperative follow-ups occurred from February 2016 to February 2018, six patients were lost to follow-up, 409 were contacted, and of these 402 participated. This report addresses pre- and postoperative findings for these 402 patients.

Figure 1
Figure 1:
Recruitment, patient and surgeon participation, and patient follow-up.


Mean age was 55 years (range 18–85), 55% were men, and 21% were self-described as non-White race/ethnicity, and 51% were working (Table 1). Among those work disabled (14%) there were associations with non-White race/ethnicity (odds ratio [OR] 2.9; 95% confidence interval [CI] 1.6–5.4; P = 0.0009) and not being a college graduate (OR 2.1; 95% CI 1.2–3.9; P = 0.01). One-third had major medical comorbidity and a positive screen for depression, and more than half had anxiety greater than population norms.

TABLE 1 - Preoperative Demographic and Clinical Variables (N = 402)
Variables Value
Age, y, mean ± SD 55 ± 14
Men 55%
Non-White race/ethnicity 21%
College graduate 69%
Work status
 Full or part time 51%
 Disabled 14%
 Retired 26%
 Other 9%
Marital status
 Married 69%
 Separated/divorced 13%
 Widowed 4%
 Never married 14%
Current smoker 7%
Past smoker 32%
Availability of help at home
 All help available 73%
 Some help available 24%
 No help available 3%
Major medical comorbidity 32%
Body mass index, kg/m2, (mean ± SD) 28 ± 5
 ≤25 29%
 >25 and <30 37%
 ≥30 34%
Positive screen for depression§ 31%
Greater anxiety than population norms || 58%
 Herniated nucleus pulposus 21%
 Degenerative condition 79%
Spine-specific circumstances
 Worker compensation 9%
 Accident, not work related 7%
 Lawsuit pending 3%
Duration of current spine symptoms >12 mo 68%
Previous spine surgery
 No 76%
 Yes, for the same condition 14%
 Yes, for a different condition 10%
Currently taking opioids 34%
Treatments received for this condition
 Epidural steroids 83%
 Physical therapy 79%
 Chiropractic manipulation 48%
 Acupuncture 33%
Severity of back pain#
 0–4 21%
 5–6 21%
 7–8 32%
 9–10 26%
Severity of leg pain# , ∗∗
 0–4 27%
 5–6 21%
 7–8 29%
 9–10 23%
Disability due to spine, mean ± SD †† 53 ± 14
Abnormal on physical examination
 Gait 23%
 Heel walk 19%
 Achilles tendon reflex∗∗ 42%
 Light touch sensation∗∗ 29%
 Hallucis strength∗∗ 19%
 Pain with flexion 42%
 Pain with extension 59%
 Pain with straight leg raising 31%
Black (non-Latino) 7%; Latino (either black or white) 9%; Asian, Pacific Islander, Native American 5%.
Homemaker (5%), unemployed (2%), student (2%).
Based on Charlson Comorbidity Index.
§Based on Geriatric Depression Scale.
||Based on state domain of Spielberger State-Trait Anxiety Inventory, age and sex norms.
Includes spinal stenosis, degenerative disc disease, degenerative spondylolisthesis, adult scoliosis, spondylolytic spondylolisthesis, reoperation.
#Possible range 0–10, higher is worse pain.
∗∗For more affected leg.
††Based on modified Oswestry Disability Index, possible score 0–100, higher is worse status.

Most patients had degenerative diagnoses with symptoms ≥12 months; about one-quarter had previous spine surgery and one-third were currently taking opioids. Leg and back pain were equally prevalent. Patients had multiple physical examination abnormalities, and spine-related disability according a standard scale was typical for a preoperative sample. Extent of surgery spanned lower complexity single-level herniated disc excision to higher complexity five-level anterior–posterior decompression, instrumentation, and fusion.

Patients cited multiple expectations of surgery (Figure 2) and their mean preoperative expectations survey score was 73 ± 19 (range 20–100) (Table 2). Similar to a previous study,8 in multivariable analysis, variables associated with higher scores (i.e., greater expectations) included younger age (P = 0.03), non-White race/ethnicity (P < 0.0001), more anxiety symptoms (P = 0.008), and more spine-specific disability (P < 0.0001). The mean preoperative surgeons’ expectations survey score was 57 ± 16 (range 16–100) (Table 2). In multivariable analysis, variables associated with greater surgeons’ expectations were younger age (P < 0.0001), non-White race/ethnicity (P < 0.0001), more depressive symptoms (P = 0.03), taking opioids (P = 0.03), nondegenerative diagnosis (P = 0.0002), and no previous spine surgery (P = 0.004). Eighty-four percent of patients had higher scores (i.e., greater expectations) than their surgeons (median difference 16, interquartile range 6–25). These differences were not due to patients expecting more items but instead were due to patients more often expecting complete improvement, whereas surgeons more often expected a lot/moderate/little improvement (Figure 3). The variables with the greatest differences between patients’ and surgeons’ scores were female sex (P = 0.03) and more spine-related disability (P < 0.0001).

Figure 2
Figure 2:
Percent of patients citing amount of improvement expected for each item from expectations survey.
TABLE 2 - Patients’ and Surgeons’ Expectations Survey Scores According to Preoperative Demographic and Clinical Variables
Variables Patients’ Survey Scores (Mean) Surgeons’ Survey Score (Mean) Difference Between Patients’ and Surgeons’
Scores P
Entire sample 73 57 16 <0.0001
Women 75 57 18 0.03
Men 72 56 15
Age < 55 75 60 15 0.13
Age ≥55 y 71 54ŧ 17
Non-White race/ethnicity 81 63 18 0.29
White race 71ŧ 55ŧ 16
College graduate 72 56 16 0.47
Not college graduate 74 57 17
Working 73 58 15 0.07
Not working 73 55 18
Work disabled due to spine 79 60 18 0.31
Not work disabled due to spine 72 56 16
Negative depression screen 71 55 16 0.10
Positive depression screen 77 59 18
Less anxiety than norms 70 55 15 0.06
More anxiety than norms 75 57 18
Taking opioids 77 59 16 0.28
Not taking opioids 71 55 18
Non-degenerative spine diagnosis 77 62 14 0.17
Degenerative spine diagnosis 72 55ʈ 17
Symptoms >12 mo 72 56 17 0.40
Symptoms ≤12 mo 74 57 16
Prior spine surgery 71 53 18 0.22
No prior spine surgery 74 58 16
Normal gait 72 55 17 0.76
Abnormal gait 76 60 17
Normal leg sensation 72 55 16 0.35
Abnormal leg sensation 76 60 18
More back pain§ 70 56 18 0.05
Less back pain 75 57 15
More leg pain§ 71 55 17 0.18
Less leg pain 75 58 15
Less disability|| 67ŧ 54 13 <0.0001
More disability 79 59 20
More complex surgery 73 55 17 0.22
Less complex surgery 73 58 16
Self-described: black (non-Latino), Latino (either black or white), Asian, Pacific Islander, Native American.
Based on Geriatric Depression Scale score ≥11 (positive depression screen) or <11 (negative depression screen).
Based on Spielberg State-Trait Anxiety Inventory score for state anxiety above (more anxiety) or below (less anxiety) population norms.
§Based on numeric rating scale (7–10 more pain; 0–6 less pain).
||Based on Oswestry Disability Index score above (more disability) or below (less disability) group mean.
Based on Surgical Invasiveness Index above (more complex) or below (less complex) group median.P < 0.05; P < 0.01; ʈP < 0.001; ŧP < 0.0001.

Figure 3
Figure 3:
Percent of surgeons citing amount of improvement expected for each item from expectations survey.

The primary analysis, that is, concordance in survey scores within the patient–surgeon pair measured by the intraclass correlation coefficient (ICC) was 0.31 (95% CI 0.22–0.39) for the entire sample, indicating fair agreement with moderate variability according to demographic and clinical characteristics (Figure 4). Spine-related disability was the variable with the greatest difference (0.36) in ICC values (i.e., less disability ICC = 0.46 (95% CI 0.34–0.51) versus more disability ICC = 0.10 (95% CI 0.00–0.23). Other variables with differences in ICC values meeting a threshold value of 0.20 (i.e., demarcating different categories of concordance) were non-White race/ethnicity and work disabled due to the spine; however, these variables were associated with more spine-related disability (i.e., OR 1.7; 95% CI 1.1–2.8; P = 0.03, and OR 5.2; 95% CI 2.4–10.9; P < 0.0001, respectively).

Figure 4
Figure 4:
Intraclass correlation coefficients (ICC) with 95% confidence intervals representing agreement between patients and surgeons for preoperative expectations survey scores. ICC ranges: 0–0.20 poor, 0.21–0.40 fair, 0.41–0.60 moderate, 0.61–0.80 good, 0.81–1.0 excellent.


Mean time between surgery and follow-up was 2.1 years (range 1.8–2.9, 96% ≥2.0 years), during which 11% were re-hospitalized for spine care and 10% underwent another spine surgery (Table 3).

TABLE 3 - Intra- and Postoperative Demographic and Clinical Variables (N = 402)
Intraoperative Variables
Vertebral levels involved
 L1, L2 28%
 L3 39%
 L4 70%
 L5 60%
Complexity of surgery
 0–2 26%
 3–5 27%
 6–10 19%
 11–50 28%
Total no. of vertebral levels involved
 <3 74%
 ≥3 26%
Any fusion performed 50%
2-Year Postoperative Variables
Subsequent epidural steroid injection 15%
Subsequent spine surgery 10%
Severity of back pain
 0–4 66%
 5–6 19%
 7–8 13%
 9–10 2%
Severity of leg pain ,
 0–4 83%
 5–6 9%
 7–8 6%
 9–10 2%
Disability due to spine, mean ± SD§ 19 ± 21
New or worse medical comorbidity 17%
Subsequent hospitalization for spine care 11%
Pre- to postoperative change:
 Disability due to spine, mean ± SD§ , || 34 ± 23
 Back pain, mean ± SD|| 3.7 ± 3.9
 Leg pain, mean ± SD , || 4.5 ± 3.9
Based on Surgical Invasiveness Index, sum of values for all vertebral levels, maximum of 10 per level.
Possible range 0–10, higher is worse pain.
For more affected leg.
§Based on modified Oswestry Disability Index, possible score 0–100, higher is worse status.
||Preoperative minus postoperative value, higher value means more improvement.

Pre- to postoperative disability improved by an MCID (i.e., 12 points)20 for 83% patients, and back and leg pain improved by an MCID (i.e., 3 grades)21 for 59% and 69% respectively; 14% reported new and 4% reported worse major medical comorbidity.

For the secondary analysis, that is, fulfillment of expectations, the mean proportion of expectations fulfilled was 0.79 (0–3.00) for patients and 1.01 (0–2.29) for surgeons (Table 4). Using the MCID of 0.20 (described above) around a proportion value of 1 to represent equivalence (i.e., .80–1.20), 7% of patients had high, 41% equivalent, and 52% had low proportions of expectations fulfilled (i.e., reflecting overestimation of postoperative improvement and less likelihood of meeting or surpassing expectations (OR 0.34; 95% CI 0.25–0.45), whereas 32% of surgeons had high, 41% equivalent, and 27% had low proportions (i.e., reflecting greater likelihood of meeting or surpassing expectations (OR 2.98; 95% CI 2.22–4.00). Multiple variables were associated with lower proportions of expectations fulfilled in bivariate analysis (Table 4). In multivariable analysis for patients, the variables that remained associated were non-White race/ethnicity, not college graduate, more preoperative disability, subsequent spine hospitalization, less change in back pain, and less change in disability. In multivariable analysis for surgeons, the variables that remained associated with lower proportion of expectations fulfilled were non-White race/ethnicity, positive depression screen, subsequent spine hospitalization, less change in back pain, and less change in spine-related disability.

TABLE 4 - Proportions of Expectations Fulfilled and Bivariate and Multivariable Models of Lower Expectations Fulfilled According to Demographic and Clinical Variables
Variables Patients Surgeons
Proportion Expectations Fulfilled Bivariate (P) Multivariable (P) Proportion Expectations Fulfilled Bivariate (P) Multivariable (P)
Entire sample 0.79 1.01
Women 0.78 0.69 1.03
Men 0.79 1.00 0.59
Age <55 y 0.81 1.02
Age ≥55 y 0.77 0.24 1.01 0.94
Non-White race/ethnicity 0.66 <0.0001 0.004 0.84 <0.0001 <0.0001
White race 0.82 1.06
College graduate 0.82 1.04
Not college graduate 0.72 0.007 0.03 0.96 0.10
Working 0.73 0.001 0.98 0.11
Not working 0.84 1.05
Negative depression screen 0.81 1.05
Positive depression screen 0.73 0.03 0.94 0.03 0.009
Taking opioids 0.73 0.01 0.96 0.05
Not taking opioids 0.82 1.04
Previous spine surgery 0.71 0.04 0.94 0.04
No previous spine surgery 0.81 1.04
Less preop disability 0.85 1.04
More preop disability 0.72 0.0001 <0.0001 0.98 0.14
Subsequent hospitalization for spine 0.61 0.0002 0.03 0.78 <0.0001 0.009
No subsequent hospitalization for spine 0.81 1.04
More change in back pain 0.92 1.17
Less change in back pain 0.62 <0.0001 <0.0001 0.84 <0.0001 0.0002
More change in disability 0.96 1.23
Less change in disability 0.62 <0.0001 <0.0001 0.81 <0.0001 <0.0001
Proportion: ≥0 to < 0.80 (improvement received is less than improvement expected); 0.80–1.20 (improvement received is equivalent to improvement expected); and >1.20 (improvement received surpasses improvement expected.
Dependent variable is lower proportion of expectations fulfilled.


In our study, concordance between patients’ and surgeons’ preoperative expectations for lumbar surgery was fair. The variable most associated with less concordance was spine-related disability. Patients had greater expectations than surgeons; however, this was not due to patients expecting more items but instead was due to patients more often expecting complete improvement, whereas surgeons more often expected a lot, moderate, or a little improvement. Compared to patients’ preoperative expectations, surgeons’ preoperative expectations more closely coincided with patient-reported 2-year postoperative fulfillment of expectations. Thus, most patients had unrealistically high expectations.

Our findings demonstrated that patients’ and surgeons’ perspectives about probable overall results of lumbar surgery differed; however, a detailed assessment revealed these differences were mostly due to amount of improvement expected and not to what items were expected (i.e., surgeons endorsed most items for most patients). This is consistent with clinical practice where patients’ frequently comment “my surgeon said I will be able to do everything”; our study supports this claim for most patients, but what is missing is a qualifier about amount of activity or function that will be doable.5 Returning to some level is probably attainable, but returning to “normal” or “complete improvement” is unrealistic for most patients as evidenced by surgeons infrequently choosing this response option.

Our findings also demonstrated that surgeons’ expectations were good predictors of patient-reported outcome. Surgeons were able to integrate clinical, surgical, and patient-centered factors into their preoperative assessments to predict what most patients ultimately came to report was the improvement they received. It should be noted that, although it may seem that surgeons can ensure their predicted improvement will be realized or surpassed if they understate their expectations preoperatively, this is not sound reasoning. Overly conservative or dour expectations from surgeons discourage patients, engender doubt about the surgeon's skill, and diminish justification for costly and frequently invasive surgery. Instead surgeons have incentive to identify and meet appropriate expectations, which could be indicators of the quality and value of care they provide.

Why did surgeons and patients have different expectations? Didn’t surgeons tell their patients what to expect? According to our patients, yes they did. As part of the enrollment interview, we asked patients how they came to have their expectations. As previously reported, most cited their surgeons, but they also cited their social networks, experiences with health care, the internet, and other health care professionals.22 Although potentially unreliable or inapplicable to their condition, the information received from these other sources collectively shaped what patients came to expect. Future work should focus on developing accessible and rigorous resources to reinforce surgeons’ advice about expectations.

A few other orthopedic studies compared patient-surgeon preoperative expectations, and our results confirm their concordance rates. One study preoperatively asked patients and their spine surgeons what they expected at 3 months for pain, strength, sensory, and physical function. Using a physician-derived six-item survey, patients had greater expectations than their surgeons, concordance rates were ≤0.20, and differences were mostly due to patients expecting complete benefit while surgeons expected a lot/moderate benefit.23 Other lumbar surgery studies assessed only patients’ expectations preoperatively using clinician-derived surveys with a few items addressing pain and function, and then postoperatively assessed fulfillment of expectations globally.2,14,15 Results showed patients’ had high expectations with two-thirds reporting surgery met their overall expectations. These studies, however, did not assess improvement received compared to improvement expected. Other studies compared patients’ and surgeons’ postoperative assessments of outcome based on a single global measure, and found disagreement in one-quarter to one-half of cases.4,24 Using an analogous methodology to the one we employed, other investigators compared patients’ and surgeons’ preoperative responses to expectations surveys developed specifically for hip and knee arthroplasty. More than 50% of patients had greater survey scores than their surgeons, ICC values were ≤0.20, and most of the disagreement was for high-level activities.19 In a follow-up study, surgeons’ scores accurately anticipated patient-reported pre- to postoperative change in symptoms and function for hip arthroplasty, but not for knee arthroplasty.25 Similar to the knee arthroplasty result, in another study, surgeons were less likely to preoperatively predict global improvement after lumbar surgery based on patient-reported change at 1 year.26 All these studies found differences in concordance based on demographic and clinical variables, but results were preliminary due to small sample sizes.

Our study has several limitations. First, we enrolled patients from a single tertiary center and their expectations and clinical characteristics may differ from patients in other settings. Second, our surgeons participated in previous expectations studies and were particularly attuned to the topic of patients’ expectations. Thus, they probably were more likely to address expectations with patients. Third, the survey does not account for importance of expectations. For some patients importance could disproportionally impact assessment of outcome, including fulfillment of expectations. In addition, reporting status during an in-person interview (i.e., at enrollment) may differ compared to a telephone interview (i.e., at follow-up). Fourth, although we had a large sample, our findings according to demographic and clinical characteristics are preliminary and need to be confirmed with larger subgroups.

In summary, concordance between patients’ and surgeons’ expectations of lumbar surgery was fair, due mostly to patients expecting complete improvement, whereas surgeons expected a lot, moderate, or a little improvement. At the assessment 2 years after surgery, surgeons’ expectations versus patients’ expectations more closely coincided with patient-reported fulfillment of expectations. For both patients and surgeons, realistic and fulfilled expectations demonstrate shared decision-making and an important patient-centered outcome of surgery.

Key Points

  • Concordance between patients’ and surgeons’ preoperative expectations of lumbar spine surgery was fair with 84% of patients having greater expectations than surgeons based on a 20-item validated expectations survey.
  • Patients’ greater expectations was not due to patients expecting more items but instead was due to patients more often expecting complete improvement, whereas surgeons more often expected a lot, moderate, or a little improvement.
  • The clinical variable most closely associated with less concordance in patients’ and surgeons’ expectations was greater spine-related disability.
  • Compared to patients’ preoperative expectations, surgeons’ preoperative expectations more closely coincided with patient-reported fulfillment of expectations 2 years postoperatively.
  • Patients’ expectations of complete improvement are often unrealistic as evidenced by surgeons infrequently endorsing this level of improvement preoperatively and patients infrequently attaining this level of improvement postoperatively.


1. Knor S, Lavallee D, Cizik AM, et al. Development and validation of a predictive model for pain and functional outcomes after lumbar spine surgery. JAMA Surg 2018; 153:634–642.
2. Yee A, Adjei N, Do J, et al. Do patient expectations of spinal surgery relate to functional outcome? Clin Orthop Relat Res 2008; 466:1154–1161.
3. Davidson D, Noonan VK, Dvorak MF, et al. The impact of patient expectations on outcome following treatment for spinal trauma: Part 1: What are spine surgeons telling their patients. Spine (Phila Pa 1976) 2010; 35:1807–1811.
4. Schwarz CE, Ayandeh A, Finkelstein JA. When patients and surgeons disagree about surgical outcome: investigating patient factors and chart note communication. Health Qual Life Outcomes 2015; 13:161DOI 10.1186/s12955-015-0343-0.
5. Chou R, Loeser JD, Owen DK, et al. Interventional therapies, surgery, and interdisciplinary rehabilitation for low back pain: an evidence-based clinical practice guideline from the American Pain Society. Spine (Phila Pa 1976) 2009; 34:1066–1077.
6. Staiger TO, Jarvik JG, Deyo RA, et al. Patient-physician agreement as a predictor of outcomes in patients with back pain. J Gen Intern Med 2005; 20:935–937.
7. Mancuso CA, Cammisa FP, Sama AA, Hughes AP, Ghomrawi HMK, Girardi FP. Development and testing of an expectations survey for patients undergoing lumbar spine surgery. J Bone Joint Surg Am 2013; 95:1793–1800.
8. Mancuso CA, Duculan R, Stal M, Girardi FP. Patients’ expectations of lumbar spine surgery. Euro Spine J 2015; 24:2362–2369.
9. Mancuso CA, Duculan R, Cammisa FP, et al. Proportion of expectations fulfilled: A new method to report patient-centered outcomes of spine surgery. Spine (Phila Pa 1976) 2016; 41:963–970.
10. Fairbank JC, Pynsent PB. The Oswestry Disability Index. Spine (Phila Pa 1976) 2000; 25:2940–2952.
11. Fairbank JC, Couper J, Davies JB, et al. The Oswestry Low Back Pain Disability Questionnaire. Physiotherapy 1980; 66:271–273.
12. Rule BG, Harvey HZ, Dobbs AR. Reliability of the Geriatric Depression Scale for younger adults. Clin Gerontol 1989; 9:37–43.
13. Spielberger CD. State-Trait Anxiety Inventory for Adults (Form Y): Manual, Test, Scoring Key. Redwood City, CA: Mind Garden Publisher; 1983.
14. Ronnberg K, Lind B, Zoega B, et al. Patients’ satisfaction wth provided care/information and expectations on clnical outcome after lumbar disc herniation surgery. Spine (Phila Pa 1976) 2007; 32:256–261.
15. Toyone T, Tananka T, Kato D, et al. Patients’ expectations and satisfaction in lumbar spine surgery. Spine (Phila Pa 1976) 2005; 30:2689–2694.
16. Charlson ME, Pompei P, Ales KL, et al. A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chron Dis 1987; 40:373–383.
17. Mirza SK, Deyo RA, Heagerty PJ, et al. Development of an index to characterize the “invasiveness” of spine surgery: validation by comparison to blood loss and operative time. Spine (Phila Pa 1976) 2008; 33:2651–2662.
18. Kramer MS, Feinstein AR. Clinical biostatistics. LIV. The biostatistics of concordance. Clin Pharmacol Ther 1981; 29:111–123.
19. Ghomrawi HM, Franco Ferrando N, Mandl LA, et al. How often are patient and surgeon recovery expectations for total joint arthroplasty aligned? Results from a pilot study. Hosp Spec Surg J 2011; 7:229–234.
20. Copay AG, Glassman SD, Subach BR, et al. Minimum clinically important difference in lumbar spine surgery patients: a choice of methods using the Oswestry Disability Index, Medical Outcomes Study questionnaire Short Form 36, and pain scales. Spine J 2008; 8:968–974.
21. Kovacs FM, Abraira V, Royuela A, et al. Minimal clinically important change for pain intensity and disability in patients with nonspecific low back pain. Spine (Phila Pa 1976) 2007; 32:2915–2920.
22. Mancuso CA, Duculan R, Cammisa FP, et al. Sources of patients’ expectations of lumbar surgery. Spine (Phila Pa 1976) 2019; 44:318–324.
23. Lattig F, Fekete TF, O’Riordan D, et al. A comparison of patient and surgeon preoperative expectations of spinal surgery. Spine (Phila Pa 1976) 2013; 38:1040–1048.
24. Lattig F, Grob D, Kleinstueck FS, et al. Ratings of global outcome at the first post-operative assessment after spine surgery: now often do the surgeon and patient agree? Euro Spine J 2009; 18: (suppl 3): S386–S394.
25. Ghomrawi HMK, Mancuso CA, Dunning A, et al. Do surgeons expectations predict clinically important improvements in WOMAC scores after THA and TKA? Clin Orthop Relat Res 2017; 475:2150–2158.
26. Graz B, Wietlisbach V, Porchet F, et al. Prognosis of “curabo effect?”: Physician prediction of and patient outcome of surgery for low back pain and sciatica. Spine (Phila Pa 1976) 2005; 30:1448–1452.

concordance; expectations fulfilled; lumbar surgery; patients’ expectations; patient-surgeon communication; surgeons’ expectations

Copyright © 2020 Wolters Kluwer Health, Inc. All rights reserved.