The Spine Journal 10 (2010) 313–320

Clinical Study

Minimum acceptable outcomes after lumbar spinal fusion

Eugene J. Carragee, MD*, Ivan Cheng, MDDepartment of Orthopedic Surgery, Stanford Medicine Outpatient Center, Stanford University School of Medicine, 450 Broadway St, Pavilion C-MC 6342,

Redwood City, CA 94063, USA

Received 1 February 2010; accepted 4 February 2010

Abstract BACKGROUND CONTEXT: Defining succ

FDA device/drug

Author disclosures

rants, Simpirica, Bioa

ture capital, start-ups

Point B/C B/S; trips/t

Orthopedics, Cytonic

fellowship support,

1529-9430/10/$ – see

doi:10.1016/j.spinee.2

ess after spinal surgery remains problematic. Theminimal clinically important difference (MCID) in pain or functional outcomes is a common metricoften calculated independent of perceived risk and morbidity, which is an important considerationin large procedures such as spinal fusion and instrumentation.PURPOSE: The purpose of this study was to describe a method of assessing treatment successbased on prospective, patient-reported ‘‘minimum acceptable’’ outcome for which they would un-dergo a procedure. These goals can then be compared at follow-up to gauge how frequently patientgoals are met and determine correlation with patient satisfaction.STUDY DESIGN: This is a clinical descriptive study of the patient-reported minimum acceptableoutcomes for spinal fusion surgery.OUTCOME MEASURES: Minimum acceptable outcomes were determined by patients on pre-operatively administered standard questionnaires regarding ultimate pain intensity, functional out-come (Oswestry Disability Index [ODI]), medication usage, and work status. Satisfaction withoutcomes was assessed at 2-year follow-up.METHODS: One hundred sixty-five consecutive patients undergoing lumbar fusion for either isthmicspondylolisthesis or disc degeneration were asked to preoperatively define on standard questionnairestheir minimum acceptable outcomes after surgery. Two-year outcomes and satisfaction were subse-quently reported and compared with the preoperatively determined minimum acceptable outcomes.RESULTS: Both the spondylolisthesis and the degenerative disc disease (DDD) groups reportedthat a high degree of improvement was the minimum acceptable threshold for considering spinalfusion. A large majority indicated that the minimum acceptable outcomes included at least a de-crease in pain intensity to 3/10 or less, an improvement in ODI of 20 or more, discontinuing opioidmedications, and return to some occupational activity. Achieving the minimum acceptable outcomewas strongly associated with satisfaction at 2 years after surgery. Patients with compensationclaims, psychological distress, and other psychosocial stressors were more likely to report satisfac-tion in the absence of achieving their minimum acceptable outcome.CONCLUSIONS: Patients with spondylolisthesis and DDD both have relatively high minimumacceptable outcomes for spinal fusion. In these cohorts, few subjects considered more commonlyproposed MCIDs for pain and function as an acceptable outcome and report that they would nothave surgery if they did not expect to achieve more than those marginal improvements. Althoughthere was good concordance between achieving the minimum acceptable outcomes and ultimatesatisfaction, patients with significant psychosocial factors (compensation claims, psychological dis-tress, and others) are less likely to associate satisfaction with outcomes with actually achievingthese improvements. � 2010 Elsevier Inc. All rights reserved.

Keywords: Minimum acceptable; Outcomes; Lumbar; Spinal fusion

status: not applicable.

: EJC (stock ownership, including options and war-

ssetts, Cytonics; private investments, including ven-

, Simpirica; consulting, Medtronic, Synthes, Well-

ravel, US Army; scientific advisory board, Intrinsic

s; other office, Bioassetts; grants, AO Foundation;

DePuy Spine; other relationships, Orthopaedic

Surgeon); IC (speaking/teaching arrangements, Stryker Spine; grants,

Medtronic, Stryker Spine).

* Corresponding author. Department of Orthopedic Surgery, Stanford

Medicine Outpatient Center, Stanford University School of Medicine,

450 Broadway St, Pavilion C-MC 6342, Redwood City, CA 94063,

USA. Tel.: (650) 723-7797; fax: (650) 723-9805.

E-mail address: carragee@stanford.edu (E.J. Carragee)

front matter � 2010 Elsevier Inc. All rights reserved.

010.02.001

ContextMinimal clinically important differences (MCID) based

upon post-hoc anchors or ‘‘standard error of measure-

ment’’ are commonly employed to determine successful

clinical outcomes following therapeutic interventions.

Often MCID levels put forth in the literature (generated

by various methods) do not appear to commensurate

with patient expectations or the risks/morbidity of surgi-

cal treatments.

ContributionAs an alternative to minimum clinically important dif-

ference (MCID), the authors described a patient and in-

tervention-specific ‘‘Minimum-Acceptable Outcome’’

(MAO) and found patients undergoing surgery for disc

degeneration or spondylolisthesis required much larger

improvements than suggested by commonly calculated

MCID’s. The authors report that achieving a MAO

across four clinical dimensions (pain, function, medica-

tion usage, and work status) was highly predictive of pa-

tient satisfaction in subjects without psychosocial

comorbidities.

ImplicationsThe applicability of commonly calculated MCIDs in

accessing surgical outcomes is still open to debate.

The validity of MCIDs based solely on a patient’s

remote assessment of satisfaction or the test’s standard

error has methodological shortcomings. The concept of

MAO, employing both subjective and objective outcome

goals established before the intervention, may be a better

measure of ‘‘success’’ and, possibly, a useful tool for

pre-operative patient education. Further study is needed

to substantiate the role of MAO in surgical decision-

making.—The Editors

314 E.J. Carragee and I. Cheng / The Spine Journal 10 (2010) 313–320

and so forth. These measurements, however, usually focus

Introduction

Although many measures have been developed to quan-tify objective outcomes (eg., fusion integrity, adjacent discdegeneration grades, etc) and subjective patient reports(eg., visual analog scores [VAS], satisfaction ratings, etc),the use of these measures to define clinical success remainsproblematic. Even global satisfaction ratings are often in-ternally inconsistent and are not specific as to what ‘‘satis-faction’’ means. Furthermore, there may be an inherent biastoward reporting satisfaction after choosing certain ‘‘pre-ferred’’ interventions, regardless of outcome. [1–4].

Some measures of ‘‘success’’ have been focused ondemonstration of statistical improvement in a measure suchas pain, function, medication usage, return-to-work status,

on the group level and not the individual level. It is difficultto interpret these results because calculated changes, whichappear statistically significant at the group level, may notnecessarily translate into significant changes for the indi-vidual (eg., one or two points on a 10-point VAS scalefor back pain).

More recent trends have suggested a ‘‘composite’’ scorecombining minimum thresholds for success in several di-mensions. The most common suggested dimensions arepain, function, medication requirements, and work status.The magnitude of improvement in each area that qualifiesas significant has been the subject of controversy: not ev-eryone will have the same ultimate goal and not all inter-ventions are equally morbid, hazardous, or costly to thepatient.

As an alternative, we propose here a method of assessingintervention success based on prospective minimum goals.In this method, patients individually establish what theyconsider to be a ‘‘minimum acceptable’’ outcome for whichthey would undergo a procedure using standard metrics.These ‘‘minimum acceptable goals’’ can then be comparedat follow-up to gauge how frequently patient goals weremet. Patient-determined minimum acceptable outcomescan also serve as a validity check against post hoc patientsatisfaction or surgeon assessment scores.

This is a descriptive study of this method using ‘‘mini-mum acceptable’’ outcomes for common lumbar fusionprocedures as assessed in a large cohort of patients withisthmic spondylolisthesis or presumed primary degenera-tive back pain preparing to undergo posterior lumbarfusions. The outcomes at 2 years were subsequently com-pared with the preoperative minimum acceptable outcomes.

Materials and methods

Patient demographics

Consecutive subjects were enrolled at a single institutionbetween 1995 and 2000 and underwent operations by thesenior author. There were 71 isthmic spondylolisthesispatients (40 men, 31 women) and 94 degenerative discdisease (DDD) patients (38 men, 56 women).

Patients were included only if they were seen in the out-patient spinal surgery clinic before surgery. Patients wereexcluded if they were admitted through the Emergency De-partment, were seen as inpatient consults and undergoingsurgery during the same admission, or received in transferdirectly from another hospital.

Baseline data

In addition to standard demographic data, a battery ofpreoperative baseline measures was performed on all pa-tients. Each patient completed the VAS for the severity ofback and leg pain [5] and the Oswestry Disability Index

315E.J. Carragee and I. Cheng / The Spine Journal 10 (2010) 313–320

(ODI) [6] and underwent psychometric testing using theZung Depressive Index and the modified somatic percep-tion questionnaire [7,8]. The distress risk assessmentmethod rating was assigned by combining the Zung andmodified somatic perception questionnaire scores to createfour clusters as described by Main [8]: normal, at risk, dis-tressed depressive, and distressed somatic. Each patientalso recorded a history of pain medication usage, preoper-ative symptom duration, work status, and amount of preop-erative work loss on a standard questionnaire. In addition,some patients had additional consultations regarding sur-gery; the specialty and the number of consultations weredocumented and analyzed for the effect of multiple opin-ions on expectations.

Outcome goals data

Subjects were given standardized forms to complete in-dicating minimum acceptable outcomes before surgery.These surveys were performed after consent and preopera-tive counseling had been completed. Preoperative informedconsent and counseling were usually given over one ormore visits before the immediate preoperative visit inwhich these expectation questionnaires were completed.The forms were mailed to the patient approximately 10days before surgery and generally submitted the day imme-diately before surgery but sometimes occurred up to 1 weekbefore surgery. Figure 1 depicts schematically the methodof data collection.

The set of forms contained specific instructions for com-pletion at this preoperative visit regarding the ‘‘minimumacceptable outcome’’: Patients were asked to indicate theabsolute worst level of function, work capacity, pain inten-sity, and medication requirement, which would be accept-able as an outcome of the operation. Patients wereinstructed to complete these metrics with the assumptionthat any score indicating a poorer outcome than this ‘‘min-imum acceptable’’ rating was not worth the risk and dis-comfort of surgery. Specifically, they were to indicate theindividual outcomes that if they ‘‘knew they would notachieve this level, [they] would not consent to have thesurgery.’’

Reliability of responses

To determine test reliability, the first 30 subjects (mixeddiagnoses) completed the entire set of questionnaires ontwo separate occasions before surgery. One set was admin-istered as above (1–7 days before surgery) and the otherwhen surgery was scheduled, usually several weeks beforesurgery.

Comparison with 2-year surgical outcomes

At 2 years postoperatively, subjects completed the stan-dard questionnaires (VAS, ODI, medication usage, work sta-tus) and a five point Likkert global satisfaction survey. This

was compared against the patient’s achievement of theminimum acceptable outcomes established preoperatively.

Statistical analysis

Descriptive statistics were used to summarize patients’sociodemographic and clinical characteristics. Means, stan-dard deviations, and medians were calculated for continu-ous variables; frequency distributions were generated forcategorical variables. Comparative statistics were per-formed using Student t test for continuous variables andchi square test for categorical variables. Statview statisticalprogram (SAS Institute, Cary, NC, USA) was used for allanalyses. Reliability of test responses was estimated usingpercentage of agreement and kappa statistics.

Results

Feasibility

Of the 185 subjects solicited to the study, all agreedto participate. Of these, 155 completed the mailed question-naires before arriving in clinic, 22 were given new forms inclinic, and 8 subjects brought in questionnaires on the dayof surgery. The median time before surgery for completion,as indicated by the patient’s dating the forms, was 3 days(range 1–7 days). No subject complained to the investiga-tors of the burden to complete the questionnaire.

Baseline characteristics

The baseline results for both groups are indicated inTable 1. There was significantly less psychological distressin subjects with spondylolisthesis compared with those withdisc degeneration alone.

Reliability

There was a high consistency of responses for all cate-gories: VAS (kappa 0.82); ODI (kappa 0.88); medication(kappa 0.73); and work (kappa 0.90). There was no signif-icant difference in percentage agreement between diagnosisgroups or between expected and minimal acceptableoutcomes.

Spondylolisthesis group: minimum acceptable outcomes

The preoperative questionnaires indicated average mini-mum acceptable ODI improvements of 29 points and painimprovement of nearly 4 points (on a 10-point scale). Lessthan 10% of spondylolisthesis patients indicated that anacceptable final occupational outcome would be disabledor part-time work status with work restrictions. Similarly,any continued narcotic pain requirement was also reportedunacceptable as a minimum requirement in nearly allsubjects (Table 2, Fig. 2, and Fig. 4).

Fig. 1. Study design schematic. Standardized questionnaires on pain, function, medication usage, and work status were determined at baseline and 2 years

postoperatively. Shortly before surgery, patients were asked to complete the same set of questionnaires, indicating what their minimum acceptable outcome

would be for each item. ODI, Oswestry Disability Index; VAS, visual analog scale.

316 E.J. Carragee and I. Cheng / The Spine Journal 10 (2010) 313–320

Degenerative disc disease group

Preoperatively, the minimum acceptable outcome for theDDD group also reflected a high level of improvement(Table 2, Figs. 2–4). The questionnaires indicated the meanminimum acceptable improvement to be 29 points on theODI and nearly 4 points on the VAS. More than 90% ofthese patients indicated that they would not consent to sur-gery unless they were confident that they would be working2 years after surgery at least in some capacity. Similarly,more than 90% of patients considered continued narcoticrequirements to be neither an expected nor acceptableoutcome.

Comparison data

Comparing the two groups, the minimum acceptablescores were significantly different for the VAS improve-ment (p5.04) and the medication levels (p5.02); that is,spondylolisthesis subjects expected slightly lower painand less medication usage at follow-up compared withDDD subjects.

Additional consultations

In 32% of the subjects, the only surgical opinion con-sulted was with the operating surgeon, without any secondopinions. In 68% of the subjects, at least one second opin-ion was obtained. This group with one or more secondopinions, therefore, had additional input regarding possibleoutcome expectations and procedure limitations. Despitethis additional information from disparate sources, therewas little difference in reported minimum acceptable out-comes between subjects with and without additional opin-ions solicited by the patient before surgery. The onlymarginally significant difference was the finding that sub-jects in the DDD group who obtained three or more opin-ions actually appeared to expect a better functionaloutcome (an ODI outcome of 14.1 for three opinions com-pared with 24.2 for one opinion, p5.06).

Validity testing with 2-year outcomes and satisfaction

Of the 185 enrolled subjects, 161 completed the 2-yearfollow-up standard questionnaires and satisfaction survey.Details of these outcomes, not related to the validation ofthe minimum acceptable outcome model, have in part beenpreviously reported.

Overall, there appeared to be relatively high validitytesting of the minimum acceptable outcomes methods. Incomparing the satisfaction with eventual outcomes withthe preoperative minimum acceptable outcomes, there ap-peared to be a high concordance in subjects achieving theirown ‘‘minimum acceptable outcome.’’ That is, subjectsmeeting their preoperative minimum in all four categoriesvery commonly reported postoperative satisfaction (posi-tive predictive value 86%, confidence interval 82%–90%).

Conversely, not achieving the minimum acceptable out-come was associated with satisfaction in only 27% of thecases (negative predictive value 73%, confidence interval69%–77%).

Satisfaction despite failure to achieve minimum accept-able outcomes was more common in certain subgroups:subjects with baseline psychological distress on distressrisk assessment method (p5.002), compensation claimants(p5.005), and other chronic pain problems (p5.002). Inother words, subjects with these variables at baseline weremore likely to have a dissociation of minimum acceptableoutcomes stated at baseline and satisfaction at follow-up(Fig. 5).

Discussion

Various tools have been devised to assess individualresponses with regard to pain level or functional capacity.Benchmarks for outcomes studies have included the VAS,Roland-Morris Disability Questionnaire [9], and the ODI.When comparing one outcomes measure to another, how-ever, markedly inconsistent results can be found. Somestudies have shown that despite measures indicatingimprovement in pain and purported patient satisfaction

Table 1

Average baseline data for isthmic spondylolisthesis and degenerative disc

disease groups

Spondylolisthesis

Degenerative

disc disease p Value

Number of patients 71 94

Age (y) 42.6612.2 40.9610.1 .35

%Male 56.3 40.4 .04

Total duration of

back pain (wk)

53.6687.9 64.3685.6 .47

VAS back 6.461.8 6.961.6 .06

VAS leg 3.763.5 3.463.3 .30

ODI 50.664.1 49.766.3 .69

MSPQ 5.064.3 7.866.6 .003

Zung 19.169.4 26.6610.8 !.0001

Smoker (%) 28.3 28.8 .53

Drug/alcohol abuse (%) 2.9 3.4 .89

Working in last year (%) 89 82 NS

Opioid medications (%) 70 75 NS

VAS, visual analog scale; ODI, Oswestry Disability Index; MSPQ, mod-

ified somatic perception questionnaire; Zung, Zung Depressive Index.

317E.J. Carragee and I. Cheng / The Spine Journal 10 (2010) 313–320

after fusion for lumbar spondylosis, there continued to beheavy medication usage, low rates of returning to work,and poor functional ratings as measured by the ODI orRoland-Morris Disability Questionnaire [10,11].

The concept of a minimal clinically important difference(MCID) has been developed to try to systematically exam-ine the threshold for when an individual has experienceda clinically relevant change, relative to the burden, costand morbidity of treatment, as measured by various out-comes measures. The MCID has been defined as being‘‘the smallest difference in score in the domain of interestin which patients perceive as beneficial and which wouldmandate, in the absence of troublesome side effects and ex-cessive cost, a change in the patient’s management’’ [12].

Many approaches have been taken to calculate anMCID. Some include the Delphi method, which relies ona panel of experts to predict clinically important changesin outcomes; statistical modeling of data to calculate theamount of change needed to exceed the statistical errorof the test, a post-hoc anchor based method; and variousempirical methods that use fixed common outcomes

Table 2

Minimum acceptable outcomes and improvement, means, median, and 90 perce

spondylolisthesis or degenerative disc disease

Spondylolisthesis

Mean6SD (range) Median 90 percentile

Raw outcome scores

Minimum acceptable outcome

VAS 2.6560.85 (1–4) 3 4

ODI 24.967.57 (10–44) 24 34

Improvement in pain intensity and function

Minimum acceptable improvement

VAS 3.7361.36 (1.5–7) 4 2

ODI 28.769.31 (6–53) 28 21

VAS, visual analog scale; ODI, Oswestry Disability Index; SD, standard dev

measures, which have been deemed to be significant bythe governing bodies or common usage. Minimum clini-cally important difference for low back pain problemshas been variously reported as being improvements inVAS of 1 to 2 points, ODI improvements of approximately15 points, and Roland-Morris Disability Questionnaire im-provements of as little as 2 to 3 points to 7 points. How-ever, these estimates have generally been established fornonsurgical interventions (eg., physical therapy, medicationinterventions, etc) with much less risk and discomfort thanspinal fusion.

This study describes a prospective minimum acceptableoutcome compared with risk method whereby patients indi-vidually estimate a minimum acceptable outcome for whichthey would undergo a procedure. The method incorporatesthe subject’s perceived risk and morbidity of the procedurecontemplated. In this study, patients with spondylolisthesisand DDD both indicate that the minimum acceptable out-comes for spinal fusion exceed the usually referenced MCIDcited above. Most patients indicate that they would refuse toproceed with surgery if only those small gains (eg., 2 pointson the pain intensity VAS or 15 points on the ODI) wereachieved. The feasibility of the method and reliability of re-sponses appeared good in our hands.

The reported minimum acceptable outcome after surgeryseems to have some variation depending on the diagnosisand nature of surgery. We have previously reported thismethod in patients undergoing lumbar decompressionalone, and those minimum acceptable outcomes weresomewhat less than that reported here. This may reflectthe anticipation of a much smaller surgery and recovery,as well as age-related comorbidities in spinal stenosis pa-tients limiting the expectations of recovery. In the presentstudy’s group, in which all patients were having spinal fu-sion, there were small differences based on diagnosis.There were statistically significant differences in minimumacceptable pain relief and medication usage, where in bothcases the spondylolisthesis group required better outcomes.Also, more patients in the spondylolisthesis group had min-imum goals of return to full-time work and no narcoticmedication use compared with the DDD group.

ntile for subjects intending to undergo lumbar spinal fusion for either

Degenerative disc disease

Mean6SD (range) Median 90 percentile

2.9260.81 (1–5) 3 4

23.565.37 (12–42) 22 30

4.0261.38 (1.5–7) 4 2

29.267.80 (8–48) 28 22

iation.

Fig. 2. Percentage distributions of anticipated pain (VAS) (Top) and func-

tion (ODI) improvement (Bottom).

Fig. 3. Percentage distributions of anticipated work (Top) and medication

outcomes (Bottom). PT, part-time; FT, full time; occ, occasional; NSAID,

nonsteroidal anti-inflammatory medications.

Fig. 4. Acceptable pain intensity 2 years after surgery in patients having

fusion for disc degeneration and isthmic spondylolisthesis. Note that there

are few patients who preoperatively consider a pain rating of more than 3

acceptable as an ultimate outcome after spinal fusion.

318 E.J. Carragee and I. Cheng / The Spine Journal 10 (2010) 313–320

There is evidence in the literature that the high expecta-tions of outcomes in the spondylolisthesis group may bewarranted. Madan and Boeree [13], reporting on surgicaloutcomes more than 2 years after posterior-lateral instru-mented fusion for low-grade isthmic spondylolisthesis,showed that more than 70% achieved a final ODI score lessthan 20 (the median acceptable outcome in our spondylolis-thesis group was 24). Similarly, Lee et al. [14] and Molinariet al. [15] reported more than 90% of patients returning tofull or modified employment or military duty after com-bined anterior and posterior fusion for isthmic spondylolis-thesis (more than 90% of our group had some employmentas a minimal acceptable outcome). Employment outcomesmay vary according to social expectation and necessity.Moller and Hedlund [16], in Sweden, reported more than50% continued disability after fusion for spondylolisthesisdespite most patients having good functional improvement.Potter et al. [17] reported a mean pain score (VAS) of 3.6after transforaminal lumbar interbody fusion for isthmicspondylolisthesis (median acceptable outcome in our spon-dylolisthesis group was 3) and no narcotic use in approxi-mately 60% of subjects (90% of our subjects felt takingno narcotics to be a minimal acceptable outcome).

Surgical outcomes of fusion for degenerative discchanges, as reported in the available randomized clinicaltrials (RCT), have not been as positive as those anticipatedby our patients. Fritzell et al. [18], reporting the results ofthe Swedish multicenter RCT, found mean VAS (2-year)outcomes between 4 and 4.5 (40–45/100) after three differ-ent fusion methods. This represented an improvement ofonly 2 to 2.5 above baseline scores. Oswestry Disability

Fig. 5. Satisfaction in discordant patient subgroups: Subjects (Top) with

normal psychometric findings or (Bottom) without compensation claims

at baseline were highly likely to report satisfaction only when preoperative

minimum acceptable outcomes were achieved. Those with abnormal psy-

chometrics (p5.002) or compensation claims (p5.005) were more likely

to report satisfaction without achieving their preoperative minimum ac-

ceptable outcomes.

319E.J. Carragee and I. Cheng / The Spine Journal 10 (2010) 313–320

Index scores improved to 34 to 39 at 2 years, a reduction ofonly 10 to 15 points above baseline. Only one-third of sur-gically treated subjects returned to work. The NorwegianRCT by Ivar Brox et al. [19] reported that the mean backpain VAS at 1 year was 4, an improvement of only 2.4.The ODI at 1 year was 26, an improvement of only 16points. Similarly, in the Nottingham RCT of interbody fu-sion techniques, ODI outcomes at 2 years were 42 to 48and represented an improvement of only 6 to 12 points.Fairbank et al. [20], reporting on the British Medical Coun-cil’s RCT, found 2-year ODI scores of 33, a 12-point im-provement after surgery. Although these cohorts cannotbe compared directly with the patients in the present study,the minimal acceptable outcomes for pain, function, andwork status were clearly only achieved in a small fractionof the subjects in these RCT.

This study has certain inherent limitations. The patientsformulate their opinion about the risk and benefits of surgeryfrom multiple sources. The generalizability of our results isnot known. However, Slosar et al. [1], reporting on spinal fu-sions for disc degeneration in our geographic area of North-ern California, found that only 7–10% of their patients mettheir expectations of outcome. Perhaps paradoxically, inSlosar’s study, 45% to 50% of subjects said that they wouldhave the surgery again for the same result. The minimum

acceptable outcome requires a different calculation on thepatient’s part and may have less to do with the perceived ef-ficacy of the surgery than with the morbidity of the surgerycompared with a perceived natural course of their illness. Inaddition, it is our impression that we more cautiously recom-mend surgery to patients with DDD than those with spondy-lolisthesis and give less optimistic outcome estimates.Despite this, both groups gave very similar minimum ac-ceptable outcomes. It is interesting to us that receivingtwo or more opinions did not affect the patient’s expecta-tions of outcome. This would seem to indicate that ourdiscussions with the patients regarding outcomes are quitesimilar to those with our colleagues in our community.

Finally, the ‘‘minimum acceptable outcome’’ methodpresented appears to be an adequately validated means ofdefining ‘‘success’’ for most patients having spinal surgery.Global ‘‘satisfaction’’ in patients not achieving preoperativeminimum goals is highly correlated with psychosocial is-sues and may indicate that ‘‘satisfaction’’ in these patientsis related to factors other than functional improvement,return to work, or discontinuing pain medication.

Conclusion

The minimum acceptable improvements for patients un-dergoing spinal fusion for isthmic spondylolisthesis andDDD of the lumbar spine are substantial for pain reduction,four or more points on a 10-point scale, good physical func-tioning (ODI improvement of more than 20), and resumingoccupational duties. Continued occupational disability orchronic dependency on opioid medications was not consid-ered acceptable by 90% of the surveyed subjects. In theselarge cohorts, few subjects considered the commonlyproposed MCID for pain and function as an acceptable out-come and report that they would not have surgery if theyexpected to achieve only those marginal improvements.Although there was good concordance between achievingthe minimum acceptable outcomes and ultimate satisfac-tion, patients with significant psychosocial comorbidities(compensation claims, psychological distress, and others)were less likely to associate outcomes satisfaction with ac-tually achieving these improvements.

References

[1] Slosar PJ, Reynolds JB, Schofferman J, et al. Patient satisfaction after

circumferential lumbar fusion. Spine 2000;25:722–6.

[2] Errico TJ. Lumbar disc arthroplasty. Clin Orthop Relat Res 2005;435:

106–17.

[3] Lafuente J, Casey AT, Petzold A, Brew S. The Bryan cervical disc

prosthesis as an alternative to arthrodesis in the treatment of cervical

spondylosis. J Bone Joint Surg Br 2005;87:508–12.

[4] Anderson PA, Rouleau JP. Intervertebral disc arthroplasty. Spine

2004;29:2779–86.

[5] Zanoli G, Stromqvist B, Jonsson B. Visual analog scales for interpre-

tation of back and leg pain intensity in patients operated for degener-

ative lumbar spine disorders. Spine 2001;26:2375–80.

320 E.J. Carragee and I. Cheng / The Spine Journal 10 (2010) 313–320

[6] Fairbank J. Use of Oswestry disability index (ODI). Spine 1995;20:

1535–7.

[7] Zung WW. A self-rating depression scale. Arch Gen Psychiatry

1965;12:63–70.

[8] Main CJ. The modified somatic perception questionnaire (MSPQ).

J Psychosom Res 1983;27:503–14.

[9] Roland M, Fairbank J. The Roland-Morris Disability Question-

naire and the Oswestry Disability Questionnaire. Spine 2000;25:

3115–24.

[10] Hodges SD, Humphreys SC, Eck JC, et al. Predicting factors of suc-

cessful recovery from lumbar spine surgery among workers’ compen-

sation patients. J Am Osteopath Assoc 2001;101:78–83.

[11] Katz JN, Lipson SJ, Chang LC, et al. Seven- to 10-year outcome of

decompressive surgery for degenerative lumbar spinal stenosis. Spine

1996;21:92–8.

[12] Jaeschke R, Singer J, Guyatt GH. Measurement of health status. As-

certaining the minimal clinically important difference. Control Clin

Trials 1989;10:407–15.

[13] Madan S, Boeree NR. Outcome of posterior lumbar interbody fusion

versus posterolateral fusion for spondylolytic spondylolisthesis.

Spine 2002;27:1536–42.

[14] Lee SH, Choi WG, Lim SR, et al. Minimally invasive anterior lumbar

interbody fusion followed by percutaneous pedicle screw fixation for

isthmic spondylolisthesis. Spine J 2004;4:644–9.

[15] Molinari RW, Sloboda JF, Arrington EC. Low-grade isthmic spondylo-

listhesis treated with instrumented posterior lumbar interbody fusion

in U.S. servicemen. J Spinal Disord Tech 2005;18(Suppl):S24–9.

[16] Moller H, Hedlund R. Instrumented and noninstrumented posterolat-

eral fusion in adult spondylolisthesis—a prospective randomized

study: part 2. Spine 2000;25:1716–21.

[17] Potter BK, Freedman BA, Verwiebe EG, et al. Transforaminal lum-

bar interbody fusion: clinical and radiographic results and complica-

tions in 100 consecutive patients. J Spinal Disord Tech 2005;18:

337–46.

[18] Fritzell P, Hagg O, Wessberg P, Nordwall A. SLSS Group. 2001 Volvo

Award Winner in Clinical Studies: lumbar fusion versus nonsurgical

treatment for chronic low back pain—a multicenter randomized con-

trolled trial from the Swedish Lumbar Spine Study group. Spine 2001;

26:2521–32.

[19] Ivar Brox J, Sorensen R, Friis A, et al. Randomized clinical trial of

lumbar instrumented fusion and cognitive intervention and exercises

in patients with chronic low back pain and disc degeneration. Spine

2003;28:1913–21.

[20] Fairbank J, Frost H, Wilson-MacDonald J, et al. Randomised

controlled trial to compare surgical stabilisation of the lumbar spine

with an intensive rehabilitation programme for patients with chronic

low back pain: the MRC spine stabilisation trial. BMJ 2005;330:

1233–41.