Content range and precision of a computer adaptive test of upper extremity function for children with cerebral palsy

The Cerebral Palsy Profile of Health and Function: Upper-Extremity Domain's Sensitivity to Change Following Musculoskeletal Surgery

PURPOSE

The Cerebral Palsy Profile of Health and Function (CP-PRO) Computerized Adaptive Tests (CAT) are quality of life measures developed specifically for use in children with cerebral palsy. This study examined the ability of the upper-extremity (UE) CP-PRO CAT to detect change in function after UE surgery compared with the Pediatric Outcomes Data Collection Instrument (PODCI), ABILHAND-Kids, and Box and Blocks test.

METHODS

From 2009 to 2013, children with cerebral palsy who had UE musculoskeletal surgery completed the UE CP-PRO CAT, PODCI-UE, ABILHAND-Kids, and Box and Blocks tests before surgery (97 children) and at 3 postoperative intervals: 6 months (80 children), 12 months (73 children), and 24 months (52 children). Mean, SD, effect size (ES), and standardized response mean (SRM) values for each measure at each time interval and each level of the Manual Ability Classification System were calculated and compared. Finally, the minimal detectable change at the 90% confidence level was determined.

RESULTS

Values for the ES (0.40) and SRM (0.53) for the UE CP-PRO CAT at baseline to 6 months were moderate and significantly greater than the PODCI-UE (ES, 0.18; SRM, 0.25). The ES and SRM for the PODCI-UE, ABILHAND-Kids, and Box and Blocks tests were not significantly greater than for the UE CP-PRO CAT at any period. From baseline to 6 months, the UE CP-PRO CAT detected a large and significant improvement for Manual Ability Classification System level II (SRM, 0.70; ES, 0.70). The minimal detectable change for the UE CP-PRO CAT was 5.20.

CONCLUSIONS

The UE CP-PRO CAT is significantly better in detecting change in UE function in the first 6 months after surgery and is comparable to other measures at 12 and 24 months.

TYPE OF STUDY/LEVEL OF EVIDENCE

Diagnostic II.

Computerized Adaptive Tests Detect Change Following Orthopaedic Surgery in Youth with Cerebral Palsy

BACKGROUND

The Cerebral Palsy Computerized Adaptive Test (CP-CAT) is a parent-reported outcomes instrument for measuring lower and upper-extremity function, activity, and global health across impairment levels and a broad age range of children with cerebral palsy (CP). This study was performed to examine whether the Lower Extremity/Mobility (LE) CP-CAT detects change in mobility following orthopaedic surgery in children with CP.

METHODS

This multicenter, longitudinal study involved administration of the LE CP-CAT, the Pediatric Outcomes Data Collection Instrument (PODCI) Transfer/Mobility and Sports/Physical Functioning domains, and the Timed "Up & Go" test (TUG) before and after elective orthopaedic surgery in a convenience sample of 255 children, four to twenty years of age, who had CP and a Gross Motor Function Classification System (GMFCS) level of I, II, or III. Standardized response means (SRMs) and 95% confidence intervals (CIs) were calculated for all measures at six, twelve, and twenty-four months following surgery.

RESULTS

SRM estimates for the LE CP-CAT were significantly greater than the SRM estimates for the PODCI Transfer/Mobility domain at twelve months, the PODCI Sports/Physical Functioning domain at twelve months, and the TUG at twelve and twenty-four months. When the results for the children at GMFCS levels I, II, and III were grouped together, the improvements in function detected by the LE CP-CAT at twelve and twenty-four months were found to be greater than the changes detected by the PODCI Transfer/Mobility and Sports/Physical Functioning scales. The LE CP-CAT outperformed the PODCI scales for GMFCS levels I and III at both of these follow-up intervals; none of the scales performed well for patients with GMFCS level II.

CONCLUSIONS

The results of this study showed that the LE CP-CAT displayed superior sensitivity to change than the PODCI and TUG scales after musculoskeletal surgery in children with CP.

Computer Adaptive Test Approach to the Assessment of Children and Youth With Brachial Plexus Birth Palsy

This study examined the psychometric properties of item pools relevant to upper-extremity function and activity performance and evaluated simulated 5-, 10-, and 15-item computer adaptive tests (CATs). In a multicenter, cross-sectional study of 200 children and youth with brachial plexus birth palsy (BPBP), parents responded to upper-extremity (n = 52) and activity (n = 34) items using a 5-point response scale. We used confirmatory and exploratory factor analysis, ordinal logistic regression, item maps, and standard errors to evaluate the psychometric properties of the item banks. Validity was evaluated using analysis of variance and Pearson correlation coefficients. Results show that the two item pools have acceptable model fit, scaled well for children and youth with BPBP, and had good validity, content range, and precision. Simulated CATs performed comparably to the full item banks, suggesting that a reduced number of items provide similar information to the entire set of items.

Validity of computer adaptive tests of daily routines for youth with spinal cord injury

OBJECTIVE

To evaluate the accuracy of computer adaptive tests (CATs) of daily routines for child- and parent-reported outcomes following pediatric spinal cord injury (SCI) and to evaluate the validity of the scales.

METHODS

One hundred ninety-six daily routine items were administered to 381 youths and 322 parents. Pearson correlations, intraclass correlation coefficients (ICC), and 95% confidence intervals (CI) were calculated to evaluate the accuracy of simulated 5-item, 10-item, and 15-item CATs against the full-item banks and to evaluate concurrent validity. Independent samples t tests and analysis of variance were used to evaluate the ability of the daily routine scales to discriminate between children with tetraplegia and paraplegia and among 5 motor groups.

RESULTS

ICC and 95% CI demonstrated that simulated 5-, 10-, and 15-item CATs accurately represented the full-item banks for both child- and parent-report scales. The daily routine scales demonstrated discriminative validity, except between 2 motor groups of children with paraplegia. Concurrent validity of the daily routine scales was demonstrated through significant relationships with the FIM scores.

CONCLUSION

Child- and parent-reported outcomes of daily routines can be obtained using CATs with the same relative precision of a full-item bank. Five-item, 10-item, and 15-item CATs have discriminative and concurrent validity.

Evaluation of the box and blocks test, stereognosis and item banks of activity and upper extremity function in youths with brachial plexus birth palsy

BACKGROUND

One of the greatest limitations to measuring outcomes in pediatric orthopaedics is the lack of effective instruments. Computer adaptive testing, which uses large item banks, select only items that are relevant to a child's function based on a previous response and filters items that are too easy or too hard or simply not relevant to the child. In this way, computer adaptive testing provides for a meaningful, efficient, and precise method to evaluate patient-reported outcomes. Banks of items that assess activity and upper extremity (UE) function have been developed for children with cerebral palsy and have enabled computer adaptive tests that showed strong reliability, strong validity, and broader content range when compared with traditional instruments. Because of the void in instruments for children with brachial plexus birth palsy (BPBP) and the importance of having an UE and activity scale, we were interested in how well these items worked in this population.

METHODS

Cross-sectional, multicenter study involving 200 children with BPBP was conducted. The box and block test (BBT) and Stereognosis tests were administered and patient reports of UE function and activity were obtained with the cerebral palsy item banks. Differential item functioning (DIF) was examined. Predictive ability of the BBT and stereognosis was evaluated with proportional odds logistic regression model. Spearman correlations coefficients (rs) were calculated to examine correlation between stereognosis and the BBT and between individual stereognosis items and the total stereognosis score.

RESULTS

Six of the 86 items showed DIF, indicating that the activity and UE item banks may be useful for computer adaptive tests for children with BPBP. The penny and the button were strongest predictors of impairment level (odds ratio=0.34 to 0.40]. There was a good positive relationship between total stereognosis and BBT scores (rs=0.60). The BBT had a good negative (rs=-0.55) and good positive (rs=0.55) relationship with the clinical category of BPBP and Raimondi classification, respectively. There was a moderate negative (rs=-0.36) and moderate positive (rs=0.47) relationship between total stereognosis and clinical category of BPBP and Raimondi, respectively. Individual stereognosis items had moderate (rs=0.34 to 0.45) to good (rs=0.52 to 0.74) correlation with total stereognosis score.

CONCLUSIONS

In children with BPBP, there was little to no DIF with item banks of activity and UE functioning. The BBT is a good predictor of degree of impairment. Stereognosis with 2 items may provide comparable information as 12 items.

LEVELS OF EVIDENCE

Level II.