Determination of minimum clinically important difference (MCID) in pain, disability, and quality of life after revision fusion for symptomatic pseudoarthrosis

Department of Neurosurgery, Vanderbilt University Medical Center, Nashville, TN, USA
The spine journal: official journal of the North American Spine Society (Impact Factor: 2.8). 11/2012; 12(12). DOI: 10.1016/j.spinee.2012.10.006
Source: PubMed

ABSTRACT BACKGROUND CONTEXT: Spinal surgical outcome studies rely on patient reported outcome (PRO) measurements to assess the effect of treatment. A shortcoming of these questionnaires is that the extent of improvement in their numerical scores lacks a direct clinical meaning. As a result, the concept of minimum clinically important difference (MCID) has been used to measure the critical threshold needed to achieve clinically relevant treatment effectiveness. Post hoc anchor-based MCID methods have not been applied to the surgical treatment for pseudoarthrosis. PURPOSE: To determine the most appropriate MCID values for visual analog scale (VAS), Oswestry Disability Index (ODI), Short Form (SF)-12 physical component score (PCS), and European Quality of Life 5-Dimensions (EQ-5D) in patients undergoing revision lumbar arthrodesis for symptomatic pseudoarthrosis. STUDY DESIGN/ SETTING: Retrospective cohort study. METHODS: In 47 patients undergoing revision fusion for pseudoarthrosis-associated back pain, PRO measures of back pain (BP-VAS), ODI, physical quality of life (SF-12 PCS), and general health utility (EQ-5D) were assessed preoperatively and 2 years postoperatively. Four subjective post hoc anchor-based MCID calculation methods were used to calculate MCID (average change; minimum detectable change; change difference; and receiver operating characteristic curve analysis) for two separate anchors (health transition index (HTI) of SF-36 and satisfaction index). RESULTS: All patients were available for a 2-year PRO assessment. Two years after surgery, a significant improvement was observed for all PROs; Mean change score: BP-VAS (2.3±2.6; p<.001), ODI (8.6%±13.2%; p<.001), SF-12 PCS (4.0±6.1; p=.01), and EQ-5D (0.18±0.19; p<.001). The four MCID calculation methods generated a wide range of MCID values for each of the PROs (BP-VAS: 2.0-3.2; ODI: 4.0%-16.6%; SF-12 PCS: 3.2-6.1; and EQ-5D: 0.14-0.24). There was no difference in response between anchors for any patient, suggesting that HTI and satisfaction anchors are equivalent in this patient population. The wide variations in calculated MCID values between methods precluded any ability to reliably determine what the true value is for meaningful change in this disease state. CONCLUSIONS: Using subjective post hoc anchor-based methods of MCID calculation, MCID after revision fusion for pseudoarthrosis varies by as much as 400% per PRO based on the calculation technique. MCID was suggested to be as low as 2 points for ODI and 3 points for SF-12. These wide variations and low values of MCID question the face validity of such calculation techniques, especially when applied to heterogeneous disease and patient groups with a multitude of psychosocial confounders such as failed back syndromes. The variability of MCID thresholds observed in our study of patients undergoing revision lumbar fusion for pseudoarthrosis raises further questions to whether ante hoc or Delphi methods may be a more valid and consistent technique to define clinically meaningful, patient-centered changes in PRO measurements.

  • [Show abstract] [Hide abstract]
    ABSTRACT: The European Quality of Life-5 Dimensions Questionnaire (EQ-5D) is the most common instrument to value health outcomes under the patient's perspective. Several studies have investigated whether observed changes are meaningful to patients, using a variety of approaches to estimate the minimum clinically important difference (MCID). This study provides an overview of the state of art of the estimation of the MCID for the three-level EQ-5D index based on the UK scoring algorithm, critically assessing the available evidence. The interest in estimation of MCID for the EQ-5D has been increasing in recent years. However, some additional standardization in the estimation procedures may be of value, in order to enhance the ability to make comparisons across measures and disease areas. Further methodological research might also contribute to reducing gaps between theory and practice.
    Expert Review of Pharmacoeconomics & Outcomes Research 04/2014; 14(2):221-33. DOI:10.1586/14737167.2014.894462 · 1.87 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: Opioids are commonly used for preoperative pain management in patients undergoing spine surgery. The objective of this investigation was to assess whether preoperative opioid use predicts worse self-reported outcomes in patients undergoing spine surgery.METHODS: Five hundred and eighty-three patients undergoing lumbar, thoracolumbar, or cervical spine surgery to treat a structural lesion were included in this prospective cohort study. Self-reported preoperative opioid consumption data were obtained at the preoperative visit and were converted to the corresponding daily morphine equivalent amount. Patient-reported outcome measures were assessed at three and twelve months postoperatively via the 12-Item Short-Form Health Survey and the EuroQol-5D questionnaire, as well as, when appropriate, the Oswestry Disability Index and the Neck Disability Index. Separate multivariable linear regression analyses were then performed.RESULTS: At the preoperative evaluation, of the 583 patients, 56% (326 patients) reported some degree of opioid use. Multivariable analyses controlling for age, sex, diabetes, smoking, surgery invasiveness, revision surgery, preoperative Modified Somatic Perception Questionnaire score, preoperative Zung Depression Scale score, and baseline outcome score found that increased preoperative opioid use was a significant predictor (p < 0.05) of decreased 12-Item Short-Form Health Survey and EuroQol-5D scores, as well as of increased Oswestry Disability Index and Neck Disability Index scores at three and twelve months postoperatively. Every 10-mg increase in daily morphine equivalent amount taken preoperatively was associated with a 0.03 decrease in the 12-Item Short-Form Health Survey physical component summary and mental component summary scores, a 0.01 decrease in the EuroQol-5D score, and a 0.5 increase in the Oswestry Disability Index and Neck Disability Index score at twelve months postoperatively. Higher preoperative Modified Somatic Perception Questionnaire and Zung Depression Scale scores were also significant negative predictors (p < 0.05).CONCLUSIONS: Increased preoperative opioid consumption, Modified Somatic Perception Questionnaire score, and Zung Depression Scale score prior to undergoing spine surgery predicted worse patient-reported outcomes. This suggests the potential benefit of psychological and opioid screening with a multidisciplinary approach that includes weaning of opioid use in the preoperative period and close opioid monitoring postoperatively.LEVEL OF EVIDENCE: Prognostic Level II. See Instructions for Authors for a complete description of levels of evidence.
    The Journal of Bone and Joint Surgery 06/2014; 96(11):e89. DOI:10.2106/JBJS.M.00865 · 4.31 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: Background Improvement in quality of life (QOL) is a long term goal of drug treatment. Although some brief interventions have been found to reduce illicit drug use, no trial among adult risky (moderate non-dependent) drug users has tested effects on health-related quality of life. Methods A single-blind randomized controlled trial of patients enrolled from February 2011 to November 2012 was conducted in waiting rooms of five federally qualified health centers. 413 adult primary care patients were identified as risky drug users using the WHO-ASSIST and 334 (81% response; 171 intervention, 163 control) consented to participate in the trial. Three-month follow-ups were completed by 261 patients (78%). Intervention patients received the QUIT intervention of brief clinician advice and up to two drug-use health telephone sessions. The control group received usual care and information on cancer screening. Outcomes were three-month changes in the Short Form Health Survey (SF-12) mental health component summary score (MCS) and physical health component summary score (PCS). Results The average treatment effect (ATE) was non-significant for MCS (0.2 points, p-value = 0.87) and marginally significant for PCS (1.7 points, p-value = 0.08). The average treatment effect on the treated (ATT) was 0.1 (p-value = 0.93) for MCS and 1.9 (p-value = 0.056) for PCS. The effect on PCS was stronger at higher (above median) baseline number of drug use days: ATE = 2.7, p-value = 0.04; ATT = 3.21, p-value = 0.02. Conclusions The trial found a marginally significant effect on improvement in PCS, and significant and stronger effect on the SF-12 physical component among patients with greater frequency of initial drug use.
    Drug and Alcohol Dependence 09/2014; 142. DOI:10.1016/j.drugalcdep.2014.06.034 · 3.28 Impact Factor