Thresholds for Health-related Quality of Life measures: reality testing

Minimal important difference estimates for patient-reported outcomes: A systematic survey

OBJECTIVES

The objective of the study was to develop an inventory summarizing all anchor-based minimal important difference (MID) estimates for patient-reported outcome measures (PROMs) available in the medical literature.

STUDY DESIGN AND SETTING

We searched MEDLINE, EMBASE, CINAHL, PsycINFO, and the Patient-Reported Outcome and Quality of Life Instruments Database internal library (January 1989-October 2018). We included primary studies empirically calculating an anchor-based MID estimate for any PROM in adults and adolescents. Pairs of reviewers independently screened and selected studies, extracted data, and evaluated the credibility of the MIDs.

RESULTS

We identified 585 eligible studies, the majority conducted in Europe (n = 211) and North America (n = 179), reporting 5,324 MID estimates for 526 distinct PROMs. Investigators conducted their studies in the context of patients receiving surgical (n = 105, 18%), pharmacological (n = 85, 15%), rehabilitation (n = 65, 11%), or a combination of interventions (n = 194, 33%). Of all MID estimates, 59% (n = 3,131) used a global rating of change anchor. Major credibility limitations included weak correlation (n = 1,246, 23%) or no information regarding the correlation (n = 3,498, 66%) between the PROM and anchor and imprecision in the MID estimate (n = 2,513, 47%).

CONCLUSION

A large number of MIDs for assisting in the interpretation of PROMs exist. The MID inventory will facilitate the use of MID estimates to inform the interpretation of the magnitude of treatment effects in clinical research and guideline development.

Counseling Guidelines for Anticipated Postsurgical Improvements in Pain, Function, Mental Health, and Self-image for Different Types of Adult Spinal Deformity

STUDY DESIGN

Retrospective analysis of a multicenter prospective adult spinal deformity (ASD) database.

OBJECTIVE

Quantify postoperative improvements in pain, function, mental health, and self-image for different ASD types.

SUMMARY OF BACKGROUND DATA

Medical providers are commonly requested to counsel patients on anticipated improvements in specific health domains including pain, function, and self-image following surgery. ASD is a heterogeneous condition; therefore, health domain improvements may vary according to deformity type. Few studies have quantified outcomes for specific ASD types.

METHODS

Surgically treated ASD patients (≥4 levels fused) prospectively enrolled into a multicenter database, minimum 2-year follow-up, were categorized into ASD types according to Scoliosis Research Society-Schwab ASD classification (THORACIC, LUMBAR, DOUBLE, SAGITTAL, MIXED). Demographic, radiographic, operative, and patient reported outcome measures (NRS back and leg pain, SRS-22r, SF-36) data were evaluated. Preoperative and last postoperative values for pain, physical and social function, mental health, and self-image were evaluated, improvements in each domain were quantified, and domain scores compared to generational normative values. Postoperative improvements were also calculated for three age cohorts (<45 yr, 45-65 yr, and >65 yr) within each deformity type.

RESULTS

359 of 564 patients eligible for study (mean age 57.9 yr, mean scoliosis 43.4°, mean SVA 63.3 mm, mean 11.7 levels fused) had ≥2 yr follow-up. Domain improvements for the entire ASD population were 45.1% for back pain, 41.3% for leg pain, 27.1% for physical function, 35.9% for social function, 62.0% for self-image, and 22.6% for mental health (P < 0.05). LUMBAR, SAGITTAL, and MIXED had greatest improvements in pain and function, while THORACIC and DOUBLE had greatest improvements in self-image. Self-image was the most impacted preoperative domain and demonstrated the greatest postoperative improvement for all ASD types.

CONCLUSION

ASD patients demonstrated quantifiable postoperative improvements in pain, self-image, physical and social function, and mental health; however, improvements differed between ASD types. Further research is needed to understand specific patient expectations for ASD treatment.

LEVEL OF EVIDENCE

3.

Can a Successful Outcome After Surgery for Lumbar Disc Herniation Be Defined by the Oswestry Disability Index Raw Score?

STUDY DESIGN

Prospective multicenter cohort study.

OBJECTIVE

To investigate (1) the discriminative ability and cutoff estimates for success 12 months after surgery for lumbar disc herniation on the Oswestry Disability Index (ODI) raw score compared with a change and a percentage change score and (2) to what extent these clinical outcomes depend on the baseline disability.

METHODS

A total of 6840 patients operated for lumbar disc herniation from the Norwegian Registry for Spine Surgery (NORspine) were included. In receiver operating characteristic (ROC) curve analyses, a global perceived effect (GPE) scale (1-7) was used an external anchor. Success was defined as categories 1-2, "completely recovered" and "much better." Cutoffs for success for subgroups with different preoperative disability were also estimated.

RESULTS

When defining success after surgery for lumbar disc herniation, the accuracy (sensitivity, specificity, area under the curve, 95% CI) for the ODI raw score (0.83, 0.87, 0.930, 0.924-0.937) was comparable to the ODI percentage change score (0.85, 0.85, 0.925, 0.918-0.931), and higher than the ODI change score (0.79, 0.73, 0.838, 0.830-0.852). The cutoff for success was highly dependent on the amount of baseline disability (low-high), with cutoffs ranging from 13 to 28 for the ODI raw score and 39% to 66% for ODI percentage change. The ODI change score (points) was not as accurate.

CONCLUSION

The 12-month ODI raw score, like the ODI percentage change score, can define a successful outcome with excellent accuracy. Adjustment for the baseline ODI score should be performed when comparing outcomes across groups, and one should consider using cutoffs according to preoperative disability (low, medium, high ODI scores).

Follow-up score, change score or percentage change score for determining clinical important outcome following surgery? An observational study from the Norwegian registry for Spine surgery evaluating patient reported outcome measures in lumbar spinal stenosis and lumbar degenerative spondylolisthesis

Background

Assessment of outcomes for spinal surgeries is challenging, and an ideal measurement that reflects all aspects of importance for the patients does not exist. Oswestry Disability Index (ODI), EuroQol (EQ-5D) and Numeric Rating Scales (NRS) for leg pain and for back pain are commonly used patients reported outcome measurements (PROMs). Reporting the proportion of individuals with an outcome of clinical importance is recommended. Knowledge of the ability of PROMs to identify clearly improved patients is essential. The purpose of this study was to search cut-off criteria for PROMs that best reflect an improvement considered by the patients to be of clinical importance.

Methods

The Global Perceived Effect scale was utilized to evaluate a clinically important outcome 12 months after surgery. The cut-offs for the PROMs that most accurately distinguish those who reported ‘completely recovered’ or ‘much improved’ from those who reported ‘slightly improved’, unchanged’, ‘slightly worse’, ‘much worse’, or ‘worse than ever’ were estimated. For each PROM, we evaluated three candidate response parameters: the (raw) follow-up score, the (numerical) change score, and the percentage change score.

Results

We analysed 3859 patients with Lumbar Spinal Stenosis [(LSS); mean age 66; female gender 50%] and 617 patients with Lumbar Degenerative Spondylolisthesis [(LDS); mean age 67; 72% female gender]. The accuracy of identifying ‘completely recovered’ and ‘much better’ patients was generally high, but lower for EQ-5D than for the other PROMs. For all PROMs the accuracy was lower for the change score than for the follow-up score and the percentage change score, especially among patients with low and high PROM scores at baseline.

The optimal threshold for a clinically important outcome was ≤24 for ODI, ≥0.69 for EQ-5D, ≤3 for NRS leg pain, and ≤ 4 for NRS back pain, and, for the percentage change score, ≥30% for ODI, ≥40% for NRS leg pain, and ≥ 33% for NRS back pain. The estimated cut-offs were similar for LSS and for LDS.

Conclusion

For estimating a ‘success’ rate assessed by a PROM, we recommend using the follow-up score or the percentage change score. These scores reflected a clinically important outcome better than the change score.

Criteria for failure and worsening after surgery for lumbar disc herniation: a multicenter observational study based on data from the Norwegian Registry for Spine Surgery

PURPOSE

In clinical decision-making, it is crucial to discuss the probability of adverse outcomes with the patient. A large proportion of the outcomes are difficult to classify as either failure or success. Consequently, cutoff values in patient-reported outcome measures (PROMs) for "failure" and "worsening" are likely to be different from those of "non-success". The aim of this study was to identify dichotomous cutoffs for failure and worsening, 12 months after surgical treatment for lumbar disc herniation, in a large registry cohort.

METHODS

A total of 6840 patients with lumbar disc herniation were operated and followed for 12 months, according to the standard protocol of the Norwegian Registry for Spine Surgery (NORspine). Patients reporting to be unchanged or worse on the Global Perceived Effectiveness (GPE) scale at 12-month follow-up were classified as "failure", and those considering themselves "worse" or "worse than ever" after surgery were classified as "worsening". These two dichotomous outcomes were used as anchors in analyses of receiver operating characteristics (ROC) to define cutoffs for failure and worsening on commonly used PROMs, namely, the Oswestry Disability Index (ODI), the EuroQuol 5D (EQ-5D), and Numerical Rating Scales (NRS) for back pain and leg pain.

RESULTS

"Failure" after 12 months for each PROM, as an insufficient improvement from baseline, was (sensitivity and specificity): ODI change <13 (0.82, 0.82), ODI% change <33% (0.86, 0.86), ODI final raw score >25 (0.89, 0.81), NRS back-pain change <1.5 (0.74, 0.86), NRS back-pain % change <24 (0.85, 0.81), NRS back-pain final raw score >5.5 (0.81, 0.87), NRS leg-pain change <1.5 (0.81, 0.76), NRS leg-pain % change <39 (0.86, 0.81), NRS leg-pain final raw score >4.5 (0.91, 0.85), EQ-5D change <0.10 (0.76, 0.83), and EQ-5D final raw score >0.63 (0.81, 0.85). Both a final raw score >48 for the ODI and an NRS >7.5 were indicators for "worsening" after 12 months, with acceptable accuracy.

CONCLUSION

The criteria with the highest accuracy for defining failure and worsening after surgery for lumbar disc herniation were an ODI percentage change score <33% for failure and a 12-month ODI raw score >48. These cutoffs can facilitate shared decision-making among doctors and patients, and improve quality assessment and comparison of clinical outcomes across surgical units. In addition to clinically relevant improvements, we propose that rates of failure and worsening should be included in reporting from clinical trials.

Validation of a consensus-based minimal clinically important difference (MCID) threshold using an objective functional external anchor

BACKGROUND CONTEXT

The minimal clinically important difference (MCID) is defined as the smallest change in an outcome that a patient would perceive as meaningful. The Initiative on Methods, Measurement and Assessment in Clinical Trials (IMMPACT) group proposed defining the MCID as a 30% improvement in self-reported pain or function. However, this MCID threshold has not been validated against an objective physical measure.

PURPOSE

To test the validity of the IMMPACT-based MCID threshold, using an objective physical measure as an external anchor.

STUDY DESIGN/SETTING

Prospective study of chronic disabling occupational lumbar disorder (CDOLD) patients completing a functional restoration program.

PATIENT SAMPLE

A consecutive cohort of 743 CDOLD patients.

OUTCOME MEASURES

Self-report measures of pain-related function were compared with an objective lifting measure, the progressive isoinertial lifting evaluation (PILE), obtained after treatment.

METHODS

The association between reporting 30% or greater improvement (the IMMPACT's MCID key criterion) and the PILE score after treatment was assessed.

RESULTS

A 30% or greater improvement on the self-report measures was significantly associated with improvement in physical function on the PILE task.

CONCLUSIONS

Despite extensive use of the MCID to evaluate effects of treatment in spinal disorders, this is the first empirical documentation of the validity of the IMMPACT's 30% change criterion compared with an objective physical anchor.

Effects of preoperative education on spinal surgery patients

Background

Preoperative patient education (PE) has been used by many institutions to deal with patient anxiety, pain control, and overall satisfaction. Although the literature suggests PE's effectiveness in joint reconstruction, data are missing in spinal surgery.

Methods

We retrospectively analyzed patients having elective spinal surgery who underwent PE (spine pre-care class) from October 2009 to March 2010. Of the 155 patients surveyed, 77 (49.7%) attended the class whereas 78 (50.3%) did not.

Results

Of the participants in the pre-care class, 96% were satisfied with their pain management versus 83% in the control group (P =.02). There was also a trend for better overall satisfaction in the pre-care class group (91% vs 85%; P > .05, multiple regression analysis). Elderly women tend to be less satisfied with pain management and overall treatment.

Conclusions

Implementation of PE has had a positive impact on patient satisfaction, especially in terms of pain management.

Instrumentation constructs in pediatric patients undergoing deformity correction correlated with Scoliosis Research Society scores

STUDY DESIGN

Data collected prospectively from the Prospective Pediatric Scoliosis Study (PPSS) were analyzed statistically to address the hypothesis that covered specific aspects of treatment and its outcome.

OBJECTIVE

To assess and contrast Scoliosis Research Society (SRS) outcome scores for patients assigned to one of three types of spinal instrumentation constructs. The study hypothesis was that the instrumentation strategy that provides the best curve correction will be associated with the best SRS scores.

SUMMARY OF BACKGROUND DATA

Surgical treatment of scoliosis has evolved over time using implants and surgical techniques; however, quality of life indicators have not typically been analyzed to assess whether surgery and instrumentation will improve quality of life in pediatric patients.

METHODS

Patients were assigned to one of three instrumentation groups depending on the type of construct used. The Scoliosis Research Society's SRS-30 survey was used to measure patient outcomes comparing preoperative results to a 2-year follow-up.

RESULTS

Changes in the SRS Pain, Activity, Appearance, Mental, Satisfaction, and SRS Total domains did not differ significantly among instrumentation groups for any time intervals. However, analysis of SRS Pain did show a significant change over time for all instrumentation patterns. The analysis of SRS Activity showed a significant change over time for all instrumentation patterns preoperatively to 2 years postoperatively. Analysis of SRS Appearance showed a significant change over time for all groups but no difference between instrumentation groups. The analysis of SRS Mental based on instrumentation types showed a significant change over time, but only the pedicle screw group's change was statistically significant. Finally, analysis of SRS Satisfaction by instrumentation type showed a statistically significant change over time for all instrumentation patterns.

CONCLUSION

There were no statistically significant baseline differences among the three instrumentation construct groups based on mean scores for the six SRS domains. None of the SRS domains had differences among the instrumentation constructs in change scores or significant differences among the instrumentation constructs.