Previous Boston ROC Pilot Studies

Older Adult Safety in Surgery (OASIS) II.  Alok Kapoor, MD, Boston University School of Medicine

This project will measure the relative performance of the Late Life Function (LL-F) part of the Late Life Function and Disability Instrument (LL-FDI) compared with Frailty Criteria and the American Heart Association score in predicting an adverse postoperative course in older adults.

Content Analysis and Refinement of the AM-PAC Basic Mobility Domain Item Bank for Computer Adaptive Testing.  Andrew D. Lynch, PT, PhD University of Pittsburgh School of Medicine

The pilot project will address content of the item bank and measurement precision and accuracy of the AM-PAC Basic Mobility CAT for patients returning to high levels of functional mobility

Project Aims:

  1. Analyze the content of the AM-PAC Basic Mobility item bank in comparison to the content of lower extremity-specific legacy PRO measures.
  2. Conduct cognitive interviews with highly functioning individuals about activities and life situations important to them to identify gaps in content coverage of the AM-PAC Basic Mobility item bank.
  3. Conduct cognitive interviews with highly

 Reliability and Validity of a Developmental Feeding Assessment for the High Risk Infant. Bobbi G. Pineda PhD OTR/L, Washington University School of Medicine

The focus of this proposal is to develop content validity, investigate reliability, and investigate some aspects of construct validity of the Neonatal Eating Outcome assessment. It is hoped that this tool will serve to aid early identification of feeding problems, inform early therapeutic interventions and assess the efficacy of early feeding interventions.

Project Aims:

  1. Establish content validity for the NEO Assessment.
  2. Determine inter-rater and intra-rater reliability of the NEO Assessment
  3. Examine whether NEO Assessment scores discriminate good and poor feeders

 Tools to quantify gait behaviors in individuals recovering from anterior cruciate ligament reconstruction. Susan Sigward PhD PT, ATC, University of Southern California

This project aims to improve clinical outcome measures in patients following anterior cruciate ligament reconstruction (ACLr) and provide new information regards the effects of daily gait behaviors on the development of maladaptive gait patterns.

Project Aims:

  1. To validate the ability of the inertial measurement unit (IMU) to capture spatio-temporal and kinematic measures of gait in ACLr patients
  2. To quantify the development of gait strategies using in-home activity monitors during the early post ACLr time period

Identifying motor performance outcome measures that underlie speech intelligibility in Parkinsonʹs disease. Cara E. Stepp, Ph.D., Boston University

The goal of this research is to develop motor outcome measures that correlate with functional measures of speech intelligibility in individuals with Parkinson’s disease.

Project Aims:

  1. Determine which motor tasks (speech vs. non‐speech) best correlate with speaker intelligibility in PD.
  2. Determine signal features of motor function that best correlate with speaker intelligibility in PD



“Development of a Low-Cost Computerized Clinical Assessment Tool for Movement Quality”

Na Jin Seo, PhD    University of Wisconsin-Milwaukee

The primary goal of Dr. Seo’s pilot project is to develop a computerized clinical assessment tool for
patients’ limb movements using a low-cost motion capture device, Microsoft Kinect.

  • Aim 1. Develop a computerized version of the Mallet test using a low-cost motion capture device
  • Aim 2. Evaluate accuracy of the computerized Mallet test with patients
  • Aim 3. Determine usability of the computerized Mallet test for clinicians


“Development of a performance-based measure of executive function: the Complex Task Performance Assessment (CTPA)”

Timothy J. Wolf, OTD, OTR/L    Washington University School of Medicine

The focus of this pilot project is to further develop the reliability and validity of the CTPA in order for it to be used as a performance-based outcome measure in cognitive rehabilitation

  • Aim 1: Establish the discriminant validity of the Complex Task Performance Assessment.
  • Aim 2: Establish the alternate form reliability of the Complex Task Performance Assessment


Title: Identifying disability in older adults using accelerometery:  Can we do better than self-report?

Daniel K. White, PT, ScD

Disability in daily life activities, such as difficulty caring for one-self or going out to public places, are an important and meaningful health outcomes in aging research.  To date, there are few self-report measures that demonstrate satisfactory psychometric properties to capture disability, such as the Pepper Assessment Tool for Disability (PAT-D) or the Late-Life Disability Instrument (LLDI), and at present there are no objective measures of disability.  There are several notable disadvantages to self-report, including bias due to contextual influences such as cognitive impairment and limited precision, which decreases generalizability across different patient populations and sensitivity to change.  One promising complementary approach to measuring disability may be to use accelerometery.  Many important daily life activities have unique movement patterns which can be captured by an accelerometer-enabled monitor. For this reason, activity data recorded using triaxial accelerometers may contain features, or unique patterns of activity which are indicators of disability.  For instance, the Actigraph is a triaxial accelerometer enabled monitor that quantifies free-living movement in units of activity counts, which is a measure of total energy expenditure.  In a pilot study, we recently compared Actigraph data with disability as measured by the LLDI in 70 older adults.  We found low variability of sustained bouts of activity counts and low peak 5 and 10 minute activity counts to be correlated with disability as defined as perceived limitation with important daily life activities.  Now, we are seeking to employ more sophisticated classification techniques using data sets from larger older adult cohorts to examine disability from accelerometer data.

The overall objective of this proposal is to examine the utility of using accelerometry as a measure of disability in older adults.  We will first identify unique features from accelerometry data that demonstrate concurrent validity with self-report measures of disability.  Since there is no gold standard measure of disability, we will next examine which measure of disability (self-report or accelerometry) is more sensitive to change over two years.  We acknowledge that disability covers a broad spectrum of behaviors many of which may or may not be captured by accelerometery.  Therefore, in addition to examining disability as defined by global summary scales, we propose to examine disability behaviors that involve some physical mobility where accelerometery may be of greatest use, such as limitations with recreational activities and visiting family and friends.


Specific Aims and Hypotheses

  1. To validate features from accelerometry with self-reported measures of disability in older adults. Features derived from accelerometry will be correlated with self-reported measures of disability defined from summary scales as well as from individual items of disability that involve physical mobility.
  2. To compare sensitivity to change in disability over two years measured with accelerometry to self-reported measures of disability in older adults.

Measures of disability from accelerometry will demonstrate larger effect sizes and standardized response means and a higher proportion of older adults will meet minimum detectable change (MDC) criteria measured by accelerometry compared with self-reported measures of disability.


Title: Measuring Participation in veterans with blast related mTBI; a comparison of CRIS and PART-O

PI: J. Kay Waid-Ebbs, PhD, BCBA-D; Sub-investigator: Pey-Shan Wen, PhD

Specific Aims:

Aim 1: Determine which participation measure is best suited to veterans with blast-related mTBI.

Aim 2: Investigate the association between the participation measure determined from Aim 1 (the CRIS or the PARTO) and the comparable domains of the legacy quality of life measure, Veteran SF-36 Heath Status Questionnaire (VSF-36).

Aim 3: To evaluate the accuracy and precision (standard error) of the Computer Adaptive Testing (CAT)-based scores of the CRIS compared to the Rasch generated scores based on total item pool of the CRIS (paper and pencil version).

Title: Development of a Quantitative Spasticity Measure for Routine Clinical Practice

Key Personnel: YingChih Wang PhD, Jay Kapellusch PhD, Roger O. Smith, PhD

Specific Aims:

The aims of this project are to: (1) develop a reliable, easy to use, quantitative spasticity measurement device that is portable, cost efficient, and suitable for use in routine clinical practice, (2) validate the kinematic measurements of the device, and (3) perform a pilot study to determine the device’s measurement repeatability potential and to compare the inter-rater reliability of the device and the modified Ashworth Scale.


Links to the Key Personnel:




To develop, validate, and disseminate computer adaptive testing (CAT) technology for assessing function and quality of care for rehabilitation clinical trial research.

The proposed project will examine the responsiveness of the new CAT, developed specifically for osteoarthritis, to changes in function among persons with OA undergoing surgery for hip or knee joint reconstruction.



To identify a valid, quantitative measure of physical activity intensity in youth with CP to be used in clinical trials.

The proposed project will:

  1. Validate existing or develop new youth-specific calibration equations to estimate physical activity intensity for three physical accelerometers with indirect calorimetry as the criterion measure.
  2. Determine whether one or more accelerometers is valid in estimating levels of physical activity intensity based on energy expenditure in comparison to actual energy expenditure.

The central hypothesis for this proposal is that all of the three physical activity accelerometers will provide valid estimates of physical activity intensity in youth with CP.



    1. To develop a prototype CRIS-CAT User Interface that incorporates a touch screen interface as well as audio-assisted computer interface (ACASI).
    2. Conduct a formative evaluation to refine the features of the CRIS-CAT User Interface.
    3. Evaluate test-retest reliability and calculate minimal detectable change (MDC) of the CRIS-CAT.