Risk Adjustment of the Modified Low Back Pain Disability Questionnaire and Neck Disability Index to Benchmark Physical Therapist Performance: Analysis From an Outcomes Registry.

BACKGROUND: Patient-reported outcomes (PROs) have been touted as the ultimate assessment of quality medical care and have been proposed as performance measures after appropriate risk adjustment. Although spine conditions represent the most common orthopedic disorders, the most used PROs for disabilities related to the back and neck-the Modified Low Back Pain Disability Questionnaire (MDQ) and the Neck Disability Index (NDI)-have not been evaluated as performance measures. OBJECTIVE: The objective of this study was to benchmark physical therapists' performance in the management of spine conditions not involving surgery through the use of risk-adjusted MDQ and NDI outcomes. DESIGN: This was a retrospective observational study. METHODS: Data were accessed for patients seeking physical therapy with no history of related surgery for back or neck pain (315,274 treatment episodes) between January 2015 and June 2018. Patients with complete data, including initial and matched final MDQ or NDI, were considered for analysis (182,276 patients; 2799 physical therapists). Linear models controlling for baseline PRO and patient characteristics predicted PRO change for each patient. An aggregated performance ratio of actual PRO change to predicted PRO change was calculated for each physical therapist, and then empirical bootstrapping was used to develop the median performance ratio and its confidence intervals. Physical therapists who met a 40-patient threshold for either cohort (MDQ or NDI) were classified as "outperforming," "meeting expectations," or "underperforming" relative to predicted values using these 95% confidence intervals. RESULTS: Performance ratios indicated that 10% and 11% of physical therapists outperformed, 79% and 78% met expectations, and 11% and 11% underperformed relative to the risk-adjusted predicted change in the MDQ (1240 therapists; 97,908 patients) and NDI (461 therapists; 26,123 patients), respectively. To demonstrate the clinical importance of risk adjustment, clinical performance was evaluated in the seemingly homogeneous subset of 208 physical therapists within 0.5 SD of the median baseline MDQ and the median actual change in the MDQ. Following risk adjustment, 2 physical therapists were classified in each of the outperforming and underperforming cohorts. LIMITATIONS: The secondarily obtained observational data used were not collected for research purposes. Additionally, the analyses were limited by missing baseline information and follow-up PROs. CONCLUSIONS: The risk-adjusted performance ratios for the MDQ and NDI resulted in disparate conclusions regarding the quality of care compared with the raw, unadjusted change scores. According to the baseline and unadjusted change in the MDQ, even physical therapists in the most homogeneous sample were differentiated following appropriate risk adjustment. Clinically important improvements in actual PROs were observed in the outperforming but not in the underperforming physical therapists. Clinically meaningful differences in the performance ratio are unknown and are a limitation to clinical application and an opportunity for future research.

Tears of the ligamentum teres: prevalence in hip arthroscopy using 2 classification systems.

BACKGROUND: The ligamentum teres (LT) anatomy has been known for many years. While its functionality remains debatable, it is well recognized that the LT can be a source of pain in the hip joint. In 1997, a landmark publication by Gray and Villar established a classification for LT tears and increased the awareness of LT disorders. However, the incidence of LT tears and the various tear types is unknown. PURPOSE: The authors report the prevalence of LT tears in a population of patients who underwent hip arthroscopy, using both the Gray and Villar classification and a new descriptive classification. STUDY DESIGN: Case series; Level of evidence, 4. METHODS: Between February 2008 and January 2011, 616 hip arthroscopies were performed by the senior author. After excluding revision surgeries, a total of 558 surgeries (502 patients) were included in the study. Data were collected regarding patients' demographics, mechanism of injury, range of motion, magnetic resonance results, and intraoperative findings. Preoperative hip-specific questionnaire scores and pain level were recorded as well. Ligamentum teres tears were classified according to Gray and Villar's classification, and were also categorized using a descriptive grading system as follows: 0, no tear; 1, <50% tear; 2, >50% tear; or 3, 100% tear. RESULTS: A total of 284 (51%) of the 558 surgeries in this cohort revealed LT tears. According to the descriptive grading system, 22% were grade 1, 24% were grade 2, and 5% were grade 3. According to the Gray and Villar classification 3.7% had full rupture, 43% had a partial tear, and 4.5% had a degenerative tear. Patients with LT tears were significantly older and had worse preoperative functional scores; they did, however, have a greater range of motion. Intraoperatively, an association with larger labral tear size and acetabular chondral damage was found. Magnetic resonance arthrography was found to have low accuracy and sensitivity in detection of LT tears. No correlation to the pain level was found. CONCLUSION: Ligamentum teres tears had a higher prevalence in this study than was published in the past, most probably attributable to a lower threshold used in defining a tear. The incidence is defined both using the Gray and Villar classification, as well as a new descriptive classification system that categorizes the LT according to amount of tearing.

Complex adaptive systems: a response to Luchins.

Luchins (2003, Administration and Policy in Mental Health, 31(2), 183-186) distinguishes the mutual strengths and shortcomings of qualitative and quantitative approaches as applied to the complexities of evaluation psychiatric hospital functioning in particular, and behavioral healthcare practice in general. This brief response offers an examination of the complex interplay of factors within non-linear models in a conceptualization of complex adaptive systems for understanding causal factors in mental health administration as well as calling into question the myth of the relationship between measurement and hence control and understanding and prediction.