AM-PAC Short Forms TM for Inpatient and Outpatient settings. Instruction Manual v.3 (revised 8/2/2013)

AM-PAC Short Forms TM for Inpatient and Outpatient settings Instruction Manual v.3 (revised 8/2/2013) Developed by: Alan Jette, PT, PhD Stephen M. Haley, PT, PhD Wendy Coster, OT, PhD Peng Sheng Ni, MD, MPH Boston University School of Public Health Health and Disability Research Institute AM-PAC Short Forms and Conversion Tables are available for download at: www.am-pac.com

Acknowledgements: Work to develop the Boston University Activity Measure for Post-Acute Care (AM-PAC) was undertaken at the Boston University Research and Training Center for Measuring Rehabilitation Outcomes and supported, in part, by the National Institute on Disability and Rehabilitation Research (grant no. H133B990005) and the National Institute of Child Health and Human Development (grant no. R01 HD43568). Development of the AM- PAC Outpatient Short Forms for use in outpatient settings was supported in part by an Independent Scientist Award (K01 HD45354-01) received by Stephen Haley, PhD. The use of AM- PAC forms and software in a product that will be sold or distributed to others is prohibited. AM-PAC scores or item responses may not be shared with others for purposes of creating databases, prediction modeling or aggregate reports without the expressed written consent of CREcare 2

Table of Contents Section 1: Introduction... 5 1.1 Activity Measure for Post-Acute Care... 5 1.2 AM-PAC: Psychometric Properties... 6 1.2.1 Reliability and Validity... 6 1.2.2 Sensitivity to Change... 7 1.2.3 Responsiveness in Clinical Trials Research... 7 1.3 AM-PAC Items and Response Options... 9 1.3.1 Sample Items: Basic Mobility Domain... 9 1.3.2 Sample Items: Daily Activity Domain... 9 1.3.3 Sample Items: Applied Cognitive Domain... 10 1.4 Patient Self-Report vs. Proxy Respondent... 10 Section 2: AM-PAC Short Forms... 11 2.1 AM-PAC Inpatient Short Forms... 11 2.1.1 Item Types and Response Options... 11 2.1.2 Administering the AM-PAC Inpatient Short Forms... 12 2.1.3 Scoring AM-PAC Inpatient Short Forms: Standardized Scores and CMS G codes... 12 2.2 AM-PAC Generic Outpatient Short Forms... 14 2.2.1 AM-PAC Generic Outpatient Short Form: Content... 14 2.2.2 AM-PAC Generic Outpatient Short Forms: Item types and Response Options... 14 2.2.3 Administering the AM-PAC Generic Outpatient Short Forms... 14 2.2.4 Guidelines for selecting responses... 15 2.2.5 Guidelines: Patients with Questions on How to Respond to Items... 15 2.2.6 Scoring AM-PAC Generic Outpatient Short Forms: Standardized Scores and CMS G codes... 16 2.3 AM-PAC Medicare Outpatient Short Forms... 18 2.3.1 Content... 18 2.3.2 Item types and Response Options... 18 2.3.3 Administering the AM-PAC Medicare Outpatient Short Forms... 19 2.3.4 Guidelines for selecting responses... 19 2.3.5 Guidelines: Patients with Questions on How to Respond to Items... 20 2.3.6 Scoring AM-PAC Medicare Outpatient Short Forms: Standardized Scores and CMS G codes... 21 Section 3: Interpreting AM-PAC Standardized Scores... 23 3.1 Functional Stages... 23 3

Section 4: Frequently Asked Questions... 26 Appendix A... 27 AM-PAC Inpatient Short Forms 6 Clicks... 27 Appendix B... 34 AM-PAC Generic Outpatient Short Forms... 34 Appendix C... 47 AM-PAC Medicare Outpatient Short Forms (DOTPA)... 47 AM-PAC Bibliography... 64 4

Section 1: Introduction 1.1 Activity Measure for Post-Acute Care The Boston University Activity Measure for Post-Acute Care (AM-PAC) is an activity limitations instrument developed by researchers at Boston University s Health and Disability Research Institute. The World Health Organization s International Classification of Functioning, Disability and Health (ICF) is the conceptual framework for the AM-PAC. According to the ICF, an activity limitation is defined as difficulty in the execution of a task or action by an individual. 1 The AM-PAC was developed as a functional outcomes system that can be used across post-acute care settings and consists of a comprehensive list of 269 functional activities (i.e., the item bank). The AM-PAC measures functional outcome by using contemporary measurement techniques, such as Item Response Theory (IRT), to overcome the limitations of traditional functional outcome measures (Jette and Haley, 2005). Unlike traditional functional outcome measures which are disease, condition, or setting-specific, the AM-PAC was designed to be used across patient diagnoses, conditions and settings where post-acute care is being provided; therefore, the AM-PAC is the ideal measure for developing benchmarks and for examining functional outcomes over an episode of post-acute care, as patients move across care settings. The AM-PAC instrument examines a set of functional activities that are likely to be encountered by most adults during daily routines within the context of either an inpatient episode of care or outpatient post-acute services. We have defined functional activity as the execution of discrete daily tasks. Because functional activity is multidimensional, AM-PAC item banks are organized into three functional areas: Basic Mobility (131 items), Daily Activity (88 items), and Applied Cognitive (50 items). Items for the AM-PAC have been drawn from two sources: (1) a set of new items that examine the functional content domains listed above; and (2) items from existing outcome instruments used in rehabilitation and post-acute programs. The items in the AM- PAC assess multiple aspects (i.e., difficulty, assistance, limitations) of an individual's ability to perform specific daily activities. IRT analyses were used to scale individual items in different functional areas along a continuum of item difficulty. To develop the AM-PAC, test items were administered to a large sample of patients from different care settings with different diagnoses. Factor analytic work identified three distinct, interpretable factors that accounted for 72% of the variance: Applied Cognitive (44%), Daily Activities (19%) and Basic Mobility (9%). These factors were verified by a confirmatory factor analysis (Haley et al. 2004) and defined as the three AM- PAC domains. Using Item Response Theory (IRT), items in each domain were scaled along a continuum of item difficulty. Items that were redundant or did not fit the model were eliminated. The remaining items formed the AM-PAC item banks, which included a wide range of items calibrated along a continuum of difficulty. The AM-PAC is a comprehensive and precise point-of-care assessment of patient-related outcomes. For this reason, it is being used with increasing frequency in healthcare. 1. International Classification of Functioning, Disability and Health (ICF). Geneva, Switzerland: World Health Organization; 2001. 5

Two AM-PAC versions are currently available: 1.) Short forms for inpatient and outpatient settings: There are separate fixed forms for specifically designed for inpatient and outpatient settings. Since the items from the different short forms are drawn from the same item banks, the standardized scores for specific functional domains (i.e., basic mobility, daily activity and applied cognitive) can be compared across all versions of the AM-PAC instruments. The short forms can be administered without access to a computer. See Section 2 for more detailed information about AM-PAC Short Forms. 1.2 AM-PAC: Psychometric Properties 1.2.1 Reliability and Validity The AM-PAC s psychometric properties have been tested in inpatient as well as outpatient post-acute care patient samples with major medical, neurologic and orthopedic impairments. The AM-PAC short forms as well as its CAT versions have demonstrated a high degree of internal consistency and test retest reliability, known groups and construct validity and show a high degree of sensitivity to change across all three functional domains across 1-month, 6-months and 12-months of follow-up. This work was supported by the National Institutes of Health and the National Institute on Disability & Rehabilitation Research. Some details are provided below. Haley et al. (2004) presented results from an initial exploratory factor analysis of AM-PAC items. Three distinct, interpretable factors were identified and accounted for 72% of the variance: Applied Cognitive (44%), Daily Activity (19%), and Basic Mobility (9%); these 3 activity factors were verified by a confirmatory factor analysis. Scaling assumptions were met for each factor in the total sample and across diagnostic groups. Internal consistency reliability was high for the total sample (Cronbach alpha = 0.92 to 0.94), and for specific diagnostic groups (Cronbach alpha = 0.90 to 0.95). Rasch scaling, residual factor, differential item functioning, and modified parallel analyses supported the unidimensionality and goodness of fit of each unique activity domain. Coster et al (2004) examined the dimensional structure and content coverage of a Daily Activity item set and compared ADL and IADL items from existing instruments (FIM, MDS, MDS-PAC, OASIS, PF-10) to a set of new items (AM-PAC) as measures of this domain. ADL and IADL items from existing rehabilitation outcomes instruments that depend on skilled upper limb and hand were located along a single continuum, along with the new items from the AM-PAC that addressed gaps in content. Results support the validity of the proposed definition of the Daily Activity dimension of function as a guide for future development of rehabilitation outcome instruments, such as linked, setting-specific short forms and computerized adaptive testing approaches. Haley et al. (2004a) compared simulated short-form and computerized adaptive testing (CAT) scores to scores obtained from complete item sets for each of the three domains of the Activity Measure for Post-Acute Care (AM-PAC). Inpatient and community-based short forms and CAT applications were developed for each of 3 activity domains (Basic Mobility, Daily Activity, Applied Cognitive) using item pools constructed from new items and items from existing post-acute care instruments. Simulated CAT scores correlated highly with score estimates from the total item pool in each domain (4- and 6-item CAT r values ranged from.90-.95; 10-item CAT r values ranged from.96-.98). Scores on the 10-item short forms constructed for inpatient and community settings also provided good estimates of the AM- 6

PAC item pool scores for the Basic Mobility and Daily Activity domains, but were less consistent in the Applied Cognitive domain. Confidence intervals around individual scores were greater for the short forms than for the CATs. The strong relationship between CAT and item pool scores demonstrates the CAT's ability to select specific items to match individual responses. The CAT may have additional advantages over short forms in practicality, efficiency, and the potential for providing more precise scoring estimates for individuals. 1.2.2 Sensitivity to Change One of the advantages of using the AM-PAC is that it alone among existing outcome instruments has demonstrated a high degree of measurement sensitivity in both inpatient and community settings for a sample with major neurologic, orthopedic and medical impairments (Coster et al., 2006). Coster et al. examined the sensitivity of the Short Form Activity Measure for Post-Acute Care (AM-PAC) compared to the Functional Independence Measure (FIM) across a 12-month period after discharge from rehabilitation hospital. All 3 AM-PAC scales were sensitive to both positive and negative change across the follow-up period. Standardized response means for the AM-PAC were consistently larger than for the FIM across patient and severity groups. A greater percentage of patients showed positive change that exceeded the minimal detectable change on the AM-PAC than on the FIM at 6- and 12-month follow-ups. By using calibrated inpatient and outpatient AM-PAC short forms we were able to track patients as they moved from the inpatient to the community-based portion of their post- acute care. When looking at sensitivity in terms of Standardized Response Means (SRM), which relates the degree of change observed to the standard deviation of change in that sample, SRM values for the AM-PAC were very similar for patients with neurologic, major medical, and orthopedic impairments, and the SRM values exceeded a full standard deviation at six months and were somewhat lower at 12 months. In all cases the AM-PAC SRMs were as good as or superior to those seen using a legacy instrument (i.e., the Functional Independence Measure [FIM]) and consistently superior to the FIM for community-dwelling patients. The AM-PAC Short Forms were also able to classify patients into functional stages throughout their recovery process, and these functional stages were as good and in many cases better than the existing functional staging system developed for the FIM (Tao, et al., 2008). 1.2.3 Responsiveness in Clinical Trials Research To assess the responsiveness of two AM-PAC-CAT scales (basic mobility and daily activity scales) to performance measures of function, data were obtained from 108 patients participating in a randomized, double-blind, placebo-controlled, multi-center study of patients recovering from a unilateral hip fracture with non-complicated surgical repair (Latham et al, 2008). Data obtained from the baseline and Week 12 visits comprise the data for these analyses. Statistical analyses focused on comparing the responsiveness measurement characteristics of four functional performance measures (Physical Function Test (PFT), Short Physical Performance Battery (SPPB), a 6-minute walk test, and a timed gait speed test. Several statistics were calculated to determine the different measurements' sensitivity to change from baseline to week 12: Cohen's Effect Size Calculated as (M 1 M 2 )/S P, where M 1 is the mean score at week 12, M 2 is the mean score at baseline and S P = [(S 1 2 + S 2 2 )/2] is the pooled standard deviation. 7

Standardized Response Mean (SRM) Calculated as (M 1 M 2 )/S Δ, where M 1 is the mean score at week 12, M 2 is the mean score at baseline and S Δ is the standard deviation of (M 1 M 2 ). Standard Error of the Mean (SEM) Calculated as S B * (1- r) where S B is the standard deviation at baseline and r is the test-retest reliability coefficient for each measure from previous studies Minimal Detectable Change (MDC 90 ) Calculated as SEM*1.645* 2. The MDC 90 can be interpreted as the smallest detectable change which falls outside the measurement error of the instrument. Percent Who Exceed MDC 90 From Baseline to Week 12 Calculated as the percent of patients who had at least a minimally detectable change from baseline to week 12. The distribution-based statistics of responsiveness for each measure are shown in Table 1. Within this sample, the effect size and SRM for the AM-PAC-CAT scales matched or exceeded the levels found for all 4 performance-based measures. The percent that exceeded the MDC 90 from baseline to week 12 was 90.9% for the AM-PAC-CAT PM scale, greater than all the other functional measures. Table 1. Measures of Responsiveness (Baseline to Week 12) Effect Size (Cohen's d) Standardized Response Mean (SRM) SEM MDC 90 % Exceeding MDC 90 (Baseline to Week 12) AM-PAC- 1.28 1.43 1.73 4.02 90.9% Basic Mobility AM-PAC 0.93 1.22 1.60 3.72 74.0% Daily Activity PFT 0.95 1.13 2.21 5.14 75.0% SPPB Total 1.18 1.28 1.47 3.42 36.5% Gait Speed 0.85 1.04.075 0.17 69.1% (m/s) 6 Min Walk 0.99 1.11 23.0 53.51 75.7% 8

1.3 AM-PAC Items and Response Options The 269 AM-PAC items are divided into three domains. The AM-PAC Basic Mobility domain includes 131 items that address basic movement and physical functioning activities, such as bending, walking, carrying, and climbing stairs. The AM-PAC Daily Activity domain includes 88 items that address basic self care and instrumental activities of daily life. The AM-PAC Applied Cognitive domain includes 50 items that access higher level cognitive functions that are necessary to live independently. There are two types of items with different response options. Difficulty Items: How much difficulty does the patient (do you) currently have 1 Unable 2 A lot 3 A little 4 None Help from another person items: How much help from another person does the patient (do you) currently need 1 Total 2 A lot 3 A little 4 None 1.3.1 Sample Items: Basic Mobility Domain How much DIFFICULTY do you (or the patient) currently have moving from lying on your back to sitting on side of the bed getting up from the floor reaching overhead while standing, as if to pull a light cord using an escalator going up and down a flight of stairs inside, using a handrail walking around inside a building (50 ft, or 16 meters) on the same level going up and down three flights of stairs inside, using a handrail carrying something in both arms while climbing a flight of stairs (e.g., laundry basket climbing stairs step-over-step without a handrail (alternating feet) walking quickly indoors to answer the telephone How much HELP from another person do you (or the patient) currently need climbing a full flight of stairs without a railing moving from a bed to a chair 1.3.2 Sample Items: Daily Activity Domain How much DIFFICULTY do you (or the patient) currently have reaching behind your back to put a belt through the loop inserting a key in a lock and turning it to unlock the door unscrewing the lid off a previously unopened jar without using devices tying shoes 9

How much HELP from another person do you (or the patient) currently need putting on and taking off regular upper body clothing taking care of personal grooming 1.3.3 Sample Items: Applied Cognitive Domain How much DIFFICULTY do you (does the patient) currently have explaining how to do something involving several steps to another person following/understanding a 10 to 15 minute speech or presentation (e.g. a lesson at a place of worship, a guest lecture at a senior center) describing something that has happened to you so that others can understand you carrying on a conversation with a small group (e.g., family or a few friends) telling someone that what they are doing is bothering you (e.g., interrupting or making noise that is distracting) getting to know new people reading a long book (over 100 pages) over a number of days reading and following complex instructions (e.g., directions to operate a new appliance or for a new medication) looking up a phone number or address in the phone book or in your own address book filling out a long form (e.g., insurance forms or an application for services) writing down a short message or note planning for and keeping appointments that are not part of your weekly routine, (e.g., therapy, doctor appointment, or a social gathering with friends and family remembering to take medications at the appropriate time using a calendar, or weekly activity planner to keep track of appointments and events putting together a shopping list of 10 to15 items 1.4 Patient Self-Report vs. Proxy Respondent We recognize that not all patients have the cognitive and communication skills to complete the AM-PAC. It is for this reason that the AM-PAC was designed so that it could be completed by a proxy respondent as well as the patient him or herself. We have conducted two studies (Andres, 2003; Haley 2006) on two separate samples that have shown high correlations between patient self-report and proxy responses to the AM-PAC. These studies included both inpatients and community-dwelling patients. Of the three functional domains included within the AM-PAC, proxy responses are very consistent across the mobility and daily activity domains but are somewhat less consistent for the Applied Cognitive domain. AM-PAC Domain Andres et al (2003) Haley et al (2006) Mobility 0.86 0.92 Daily Activities 0.90 0.93 Applied Cognitive 0.68 0.77 10

Section 2: AM-PAC Short Forms AM-PAC Short Forms were created by carefully selecting subsets of AM-PAC questions from the scaled item pools for each of the three functional domains. Since individuals in inpatient and outpatient settings may function at a different level of ability and engage in different kinds of activities, separate short forms were developed for each setting. However, since the AM-PAC Inpatient and Outpatient Short Form items are derived from a larger bank of items, standardized scores generated from AM-PAC Inpatient and Outpatient Short Forms are comparable and can be used to assess patients function as they move from inpatient to outpatient settings. By using standardized AM-PAC scores, all versions of the AM-PAC instrument can be compared. Three versions of short forms available: 1. AM-PAC 6-Clicks Inpatient Short Forms 2. AM-PAC Generic Outpatient Short Forms 3. AM-PAC Medicare Outpatient Short Forms (DOPTA) Computer Adaptive Tests (CATs): The AM-PAC can be administered as a computer adaptive test (CAT). Access to a computer is required to use the AM-PAC CAT. Two versions of the AM-PAC CAT are available a freestanding computer version and web-based computer version. For more information on the AM-PAC CAT, visit the AM-PAC website: http://www.am-pac.com 2.1 AM-PAC Inpatient Short Forms AM-PAC Inpatient Short Forms are commonly referred to as 6 Clicks because functional scores are estimated with only 6 items. The 6 items included in the AM-PAC Inpatient Basic Mobility and Daily Activity Short Forms were selected because these items assess activities that are appropriate for inpatient settings and these items were best for predicating the functional scores derived from administering the entire set of AM-PAC items. There are separate AM-PAC Inpatient Short Forms for the Basic Mobility and Daily Activity domains. While there is an AM-PAC Applied Cognitive Short Form for outpatient settings, an Applied Cognitive Short Form is not currently available for the inpatient setting. 2.1.1 Item Types and Response Options The AM-PAC Inpatient Basic Mobility and Daily Activity Short Forms include two different types of questions and response options: Difficulty Items: How much difficulty does the patient currently have 1 Unable 2 A lot 3 A little 4 None 11

Help from another person items: How much help from another person does the patient currently need 1 Total 2 A lot 3 A little 4 None 2.1.2 Administering the AM-PAC Inpatient Short Forms Who should complete the AM-PAC Inpatient Short Forms? Clinicians select a response for the AM-PAC Inpatient Short Forms questions based on observed patient activity and/or professional judgment. Guidelines for selecting responses Responses should reflect the individual s ability to do the activity without special equipment or help from another person, unless specified in the question. Clinicians may find it helpful to use the guidelines outlined below to select functional item responses. Difficulty Items 1. Select can t do if the patient is not able to do the activity. 2. Select a lot if it is a struggle, requiring great effort and/or time. 3. Select a little if the patient can manage to do the activity, but it takes more effort and/or time than you think it should. 4. Select none if the patient does not experience any problems. Help from another person items 1. Total = Total/Dependent Assist 2. A lot = Max/Mod Assist 3. A little = Min/Contact Guard Assist/Supervision 4. None = Modified Independence/Independent 2.1.3 Scoring AM-PAC Inpatient Short Forms: Standardized Scores and CMS G codes Recently, the Centers for Medicare and Medicaid (CMS) announced a requirement for therapy functional reporting that uses severity modifiers to reflect the percentage of a patient s functional impairment, as determined by the therapist. The percentage of impairment includes 7 different percentage intervals, ranging from 0% - 100%. AM-PAC scoring conversion tables are available for the Basic Mobility and Daily Activity Inpatient Short Forms. These tables are used to convert AM-PAC raw scores to standardized scores, the percentage of functional impairment and CMS Functional Modifier G-codes. AM-PAC Inpatient Short Forms are scored by using the process outlined below. 1. Obtain the raw score by summing the number for the individual item responses (1 = lowest functioning; 4 = highest functioning). 12

Since there are 6 items in each domain, raw scores for the Basic Mobility and Daily Activity domains range from a low of 6 to a high of 24. 2. Use the appropriate conversion table to convert raw scores to standardized scores. Based on the Item Response Theory (IRT) item calibrations, we calculated the conditional distribution of each summed raw score as a function of the latent trait. 2,3 Then a Bayesian (Expected A Posterior) estimation was used to estimate the standardized score for each raw summed score. 4 The standardized z-score has a mean of zero and standard deviation of one. Positive values represent scores higher than the mean and negative values for scores lower than the mean. Due to the difficulty in interpreting negative scores, a linear transformation of AM-PAC z-scores was used to convert scores to a t-score scale with a mean of 50 and standard deviation of 10. Transformation from z-score to t- score is achieved by multiplying the z-score by 10 and adding 50 (t = z*10+50). When converted to standardized scores, AM-PAC Inpatient Short Form scores are directly comparable to AM-PAC Outpatient Short Form scores for the same functional domain. Standardized short from scores are also comparable to AM-PAC scores generated from AM-PAC computer adaptive test (CAT) scores. 3. Use the appropriate conversion table to determine percent of functional limitation ranging from 0% - 100% and to select the CMS Functional Modifier G-Code. Each percentage range is associated with a two-letter modifier code, as illustrated in the table below. Refer to Appendix A to view AM-PAC Inpatient Short Forms, Conversion Tables and scoring examples. CMS Modifier G-Code CH CI CJ CK CL CM CN Impairment Limitation Restriction Description 0 percent impaired, limited or restricted At least 1 percent but less than 20 percent impaired, limited or restricted At least 20 percent but less than 40 percent impaired, limited or restricted At least 40 percent but less than 60 percent impaired, limited or restricted At least 60 percent but less than 80 percent impaired, limited or restricted At least 80 percent but less than 100 percent impaired, limited or restricted 100 percent impaired, limited or restricted 2. Lord, F. M., and Wingersky, M. S. (1984). Comparison of IRT true-score and equipercentile observed-score "equatings." Applied Psychological Measurement, 8(4), 453 461. 3. Thissen, D., Pommerich, M., Billeaud, K., and Williams, V. S. L. (1995). Item response theory for scores on tests including polytomous items with ordered responses. Applied Psychological Measurement, 19(1), 39 49. 4. Thissen, D., and Wainer, H. (Eds.). (2001). Test scoring. Mahwah, N.J: Erlbaum. 13

2.2 AM-PAC Generic Outpatient Short Forms The AM-PAC Generic Outpatient Short Forms include separate forms for three domains: Basic Mobility, Daily Activity and Applied Cognitive. Items for each short form domain were selected because they include activities that are appropriate for outpatient settings and these items were best for predicating the functional scores derived from administering the entire set of AM-PAC items. 2.2.1 AM-PAC Generic Outpatient Short Form: Content The AM-PAC Generic Basic Mobility Outpatient Short Forms include 18 items that assess basic movement and physical functioning activities, such as bending, walking, carrying, and climbing stairs. The AM-PAC Generic Daily Activity Outpatient Short Forms include 15 items that assess basic self care and instrumental activities of daily life. The AM-PAC Generic Applied Cognitive Outpatient Short Forms include 19 items that assess the ability to communicate and process information. 2.2.2 AM-PAC Generic Outpatient Short Forms: Item types and Response Options The AM-PAC Outpatient Basic Mobility, Daily Activity and Applied Cognitive Short forms use the following format for the items and responses: How much difficulty do you currently have 1 Unable 2 A lot 3 A little 4 None 2.2.3 Administering the AM-PAC Generic Outpatient Short Forms Who should complete the AM-PAC Generic Outpatient Short Forms? The AM-PAC Generic Outpatient Short Forms were designed for adults at all levels of function who are receiving outpatient therapy. The AM-PAC Generic Outpatient Short Forms are completed by having patients answer the questions independently or with the assistance of a clinician or family member, if necessary. See the description of using a proxy or recorder respondent and the criteria for using each below. A proxy is someone who answers the questions on behalf of the patient if the patient is unable to do so for him/herself. The proxy respondent determines the content of the answer based upon his/her knowledge or direct observation of the patient. The proxy respondent can be a family member, close personal friend, or a clinician. A proxy should be used if the patient cannot concentrate for the time it takes to complete the assessment; or if the patient cannot give correct/accurate answers to questions about his/her health. In contrast, a recorder may also complete the assessment for the patient. A recorder is someone who enters responses provided by a patient who can respond reliably; even if the patient requires 14

assistance understanding the content, or giving an answer. A recorder should not influence or answer for the patient. A recorder should be used if the patient cannot read English or Spanish; the patient has difficulty reading, but can answer reliably verbally; the patient cannot write his/her own responses on the form (e.g., upper limb impairment, vision impairment); or the patent has difficulty understanding instructions. 2.2.4 Guidelines for selecting responses Responses should reflect the individual s ability to do the activity without special equipment or help from another person, unless specified in the question. The following introductory statement can be used to instruct patients or proxies on how to complete the AM-PAC Generic Outpatient Short Form: Please read each question listed below and put a check in the box next to the statement that best describes your (or the patient s) current level of difficulty in doing each task. For example, for the question: How much difficulty do you currently have doing light housework? Select can t do if you are not able to do light housework. Select a lot if it is a struggle to do light housework, requiring great effort and/or time. Select a little if you manage to do light housework, but notice that it takes more effort and/or time than you think it should. Select none if you do not experience any problems. 2.2.5 Guidelines: Patients with Questions on How to Respond to Items Patient is not sure how to respond If the patient responds to an item by saying I don t do that activity ask for more information: Do you not do the activity because it is too difficult for you to do? Because you don t know how to do the activity? Or is it that you have not tried the activity recently? If the patient does not do the activity, does not know how to do the activity or has not tried it recently but states: I know how much difficulty I would have or how much help I would need if I tried. Ask the patient to pick one of the choices listed with the item. Items seem repetitive In some cases there are items that patients might feel they have already answered (e.g., two questions about stairs). Inform the patient that some may involve the same kind of activity but focus on different levels of difficulty for that activity. Patient is confused about the scope of the activity: Within each section, the items vary from simple actions to complex activities. In some cases, the item is an action or activity that is part of a complex activity or task. For example, one item in the Daily Activities domain is inserting a key in a lock and turning it to unlock a door. If the person is focusing on the broader activity instead of the specific item (e.g., they say "my door is too heavy for me to open"), 15

redirect them to the item as written (e.g., the simpler activity of inserting a key in a lock and turning it, not the complex task of opening a locked door). Items that ask about the difficulty a person has doing an activity, and do NOT mention help from another person, but the person requires help. For items that ask about the level of difficulty, if the patient can ONLY do the activity with physical help from another person, the appropriate response is unable. For example, if the person says, I can only walk if someone is there to help me, s/he should select unable. 2.2.6 Scoring AM-PAC Generic Outpatient Short Forms: Standardized Scores and CMS G codes Recently, the Centers for Medicare and Medicaid (CMS) announced a requirement for therapy functional reporting that uses severity modifiers to reflect the percentage of a patient s functional impairment, as determined by the therapist. The percentage of impairment includes 7 different percentage intervals, ranging from 0% - 100%. Scoring conversion tables are available for the AM-PAC Outpatient Basic Mobility, Daily Activity and Applied Cognitive Short Forms. These tables are used to convert AM-PAC raw scores to standardized scores, percent of functional impairment and CMS Functional Modifier G-codes. AM-PAC Generic Outpatient Short Forms are scored by using the process outlined below. 1. Obtain the raw score by summing the number for the individual item responses (1 = lowest functioning; 4 = highest functioning). Since each domain has a different number of items, raw scores for the three domains differ in the range of possible scores as follows: Basic Mobility (low = 18; high = 72, Daily Activity (low = 15; high = 60), Applied Cognitive (low = 19; high = 76). 2. Use the appropriate conversion table to convert raw scores to standardized scores. Based on the Item Response Theory (IRT) item calibration, we calculated the conditional distribution of each summed raw score as a function of the latent trait. 2,3 Then the Bayesian (Expected A Posterior) estimation was used to estimate the standardized score for each raw summed score. 4 2. Lord, F. M., and Wingersky, M. S. (1984). Comparison of IRT true-score and equipercentile observed-score "equatings." Applied Psychological Measurement, 8(4), 453 461. 3. Thissen, D., Pommerich, M., Billeaud, K., and Williams, V. S. L. (1995). Item response theory for scores on tests including polytomous items with ordered responses. Applied Psychological Measurement, 19(1), 39 49. 4. Thissen, D., and Wainer, H. (Eds.). (2001). Test scoring. Mahwah, N.J: Erlbaum. 16

A standardized z-score generated from an IRT analysis, has a mean of zero and standard deviation of one. Positive values represent scores higher than the mean and negative values for scores lower than the mean. Due to the difficulty in interpreting negative scores, a linear transformation of AM-PAC z- scores was used to convert scores to a t-score scale with mean of 50 and standard deviation of 10. Transformation from z-score to t-score is achieved by multiplying the z-score by 10 and adding 50 (t = z*10+50). When converted to standardized scores, AM-PAC Inpatient Short Form scores are directly comparable to AM-PAC Outpatient Short Form scores for the same functional domain. Standardized short from scores are also comparable to AM-PAC scores generated from AM-PAC computer adaptive tests (CATs). 3. Use the appropriate conversion table to determine percent of functional limitation ranging from 0% - 100% and select the CMS Functional Modifier G-Code. Each percentage range is associated with a two-letter modifier code, as illustrated in the table below. Refer to Appendix B to view AM-PAC Generic Outpatient Short Forms, Conversion Tables and scoring examples. CMS Modifier G-Code CH CI CJ CK CL CM CN Impairment Limitation Restriction Description 0 percent impaired, limited or restricted At least 1 percent but less than 20 percent impaired, limited or restricted At least 20 percent but less than 40 percent impaired, limited or restricted At least 40 percent but less than 60 percent impaired, limited or restricted At least 60 percent but less than 80 percent impaired, limited or restricted At least 80 percent but less than 100 percent impaired, limited or restricted 100 percent impaired, limited or restricted 17

2.3 AM-PAC Medicare Outpatient Short Forms (DOTPA VERSION) The overarching purpose of the Developing Outpatient Therapy Payment Alternatives (DOTPA) project was to develop an alternative approach to reimbursement of outpatient therapy services covered under Medicare Part B. The patient reported functional assessment information collected with the DOTPA Continuity Assessment Record and Evaluation (CARE) instrument were based on the AM-PAC instrument. These forms may be more appropriate than the AM-PAC Generic Outpatient Short Forms for older persons or for younger adult outpatients with lower levels of functional ability. 2.3.1 Content AM-PAC Medicare AM-PAC short forms were specifically created for the Medicare Part B project in 4 content areas: Basic Mobility (13 items); Basic Mobility (for wheelchair users - 5 additional items); Daily Activity (12 items); and Applied Cognitive Function (13 items). The basic mobility short form is administered differently, based on the patient s use of a wheelchair. While 13 items are administered to all patients, individuals who indicate that they use a wheelchair answer 5 additional wheelchair items. It is important to note that different conversion tables are used depending on whether the 13 items or 19 items (13 plus 5 wheelchair items) are administered. 2.3.2 Item types and Response Options The AM-PAC Medicare Outpatient Basic Mobility, Daily Activity and Applied Cognitive Short Forms (DOPTA) use the following format for the items and responses: How much difficulty do you (does the patient) currently have 1. Unable 2. A lot 3. A little 4. None AND How much help from another person do you (does the patient) currently need 1. Unable 2. A lot of help needed 3. a little help needed 4. no help needed 18

2.3.3 Administering the AM-PAC Medicare Outpatient Short Forms Who should complete the AM-PAC Medicare Outpatient (DOPTA) Short Forms? The Medicare Outpatient Short Forms (DOPTA) were designed for individuals receiving outpatient care under the Medicare Part B program. The AM-PAC Medicare Outpatient Short Forms (DOPTA) are completed by having patients answer the questions independently or with the assistance of a clinician or family member, if necessary. See the description of using a proxy or recorder respondent and the criteria for using each below. A proxy is someone who answers the questions on behalf of the patient if the patient is unable to do so for him/herself. The proxy respondent determines the content of the answer based upon his/her knowledge or direct observation of the patient. The proxy respondent can be a family member, close personal friend, or a clinician. A proxy should be used if: the patient cannot concentrate for the time it takes to complete the assessment; or if the patient cannot give correct/accurate answers to questions about his/her health. In contrast, a recorder may also complete the assessment for the patient. A recorder is someone who enters responses provided by a patient who can respond reliably; even if the patient requires assistance understanding the content, or giving an answer. A recorder should not influence or answer for the patient. A recorder should be used if: the patient cannot read English or Spanish; the patient has difficulty reading, but can answer reliably verbally; the patient cannot write his/her own responses on the form (e.g., upper limb impairment, vision impairment); or the patent has difficulty understanding instructions. 2.3.4 Guidelines for selecting responses Responses should reflect the individual s ability to do the activity without special equipment or help from another person, unless specified in the question. The following introductory statement can be used to instruct patients or proxies on how to complete the AM-PAC Medicare Outpatient Short Forms (DOPTA): Difficulty Items Please read each question listed below and put a check in the box next to the statement that best describes your (or the patient s) current level of difficulty in doing each task. For example, for the question: How much difficulty do you currently have doing light housework? Select can t do if you are not able to do light housework. Select a lot if it is a struggle to do light housework, requiring great effort and/or time. Select a little if you manage to do light housework, but notice that it takes more effort and/or time than you think it should. 19

Select none if you do not experience any problems. Help from another person items Please read each question listed below and put a check in the box next to the statement that best describes how much help from another person you need to do the activities. For example, for the question about using a wheelchair to get around: 1. Select Total if you require total assistance 2. Select A lot if you require maximum to moderate assistance 3. Select A little if you require minimal assistance, contact guard assistance or supervision 4. Select None if you are independent 2.3.5 Guidelines: Patients with Questions on How to Respond to Items Patient is not sure how to respond If the patient responds to an item by saying I don t do that activity ask for more information: Do you not do the activity because it is too difficult for you to do? Because you don t know how to do the activity? Or is it that you have not tried the activity recently? If the patient does not do the activity, does not know how to do the activity or has not tried it recently but states: I know how much difficulty I would have or how much help I would need if I tried. Ask the patient to pick one of the choices listed with the item. Items seem repetitive In some cases there are items that patients might feel they have already answered (e.g., two questions about stairs). Inform the patient that some may involve the same kind of activity but focus on different levels of difficulty for that activity. Patient is confused about the scope of the activity: Within each section, the items vary from simple actions to complex activities. In some cases, the item is an action or activity that is part of a complex activity or task. For example, one item in the Daily Activities domain is inserting a key in a lock and turning it to unlock a door. If the person is focusing on the broader activity instead of the specific item (e.g., they say "my door is too heavy for me to open"), redirect them to the item as written (e.g., the simpler activity of inserting a key in a lock and turning it, not the complex task of opening a locked door). Items that ask about the difficulty a person has doing an activity, and do NOT mention help from another person, but the person requires help. For items that ask about the level of difficulty, if the patient can ONLY do the activity with physical help from another person, the appropriate response is unable. For example, if the person says, I can only walk if someone is there to help me, s/he should select unable. 20

2.3.6 Scoring AM-PAC Medicare Outpatient Short Forms: Standardized Scores and CMS G codes Recently, the Centers for Medicare and Medicaid (CMS) announced a requirement for therapy functional reporting that uses severity modifiers to reflect the percentage of a patient s functional impairment, as determined by the therapist. The percentage of impairment includes 7 different percentage intervals, ranging from 0% - 100%. Scoring conversion tables are available for the AM-PAC Medicare Outpatient Basic Mobility, Daily Activity and Applied Cognitive Short Forms (DOPTA). These tables are used to convert AM-PAC raw scores to standardized scores, percent of functional impairment and CMS Functional Modifier Gcodes. AM-PAC Medicare Outpatient Short Forms (DOPTA) are scored by using the process outlined below. 1. Obtain the raw score by summing the number for the individual item responses (1 = lowest functioning; 4 = highest functioning). Since each domain has a different number of items, raw scores for the three domains differ in the range of possible scores as follows: Basic Mobility (low = 13; high = 52), Basic Mobility Wheelchair Users (low=18; high = 72) Daily Activity (low = 12; high = 48), Applied Cognitive (low = 13; high = 52). 2. Use the appropriate conversion table to convert raw scores to standardized scores. Based on the Item Response Theory (IRT) item calibration, we calculated the conditional distribution of each summed raw score as a function of the latent trait. 2,3 Then the Bayesian (Expected A Posterior) estimation was used to estimate the standardized score for each raw summed score. 4 A standardized z-score generated from an IRT analysis, has a mean of zero and standard deviation of one. Positive values represent scores higher than the mean and negative values for scores lower than the mean. Due to the difficulty in interpreting negative scores, a linear transformation of AM-PAC z-scores was used to convert scores to a t-score scale with mean of 50 and standard deviation of 10. Transformation from z-score to t-score is achieved by multiplying the z-score by 10 and adding 50 (t = z*10+50). 2. Lord, F. M., and Wingersky, M. S. (1984). Comparison of IRT true-score and equipercentile observed-score "equatings." Applied Psychological Measurement, 8(4), 453 461. 3. Thissen, D., Pommerich, M., Billeaud, K., and Williams, V. S. L. (1995). Item response theory for scores on tests including polytomous items with ordered responses. Applied Psychological Measurement, 19(1), 39 49. 4. Thissen, D., and Wainer, H. (Eds.). (2001). Test scoring. Mahwah, N.J: Erlbaum. 21

When converted to standardized scores, AM-PAC Inpatient Short Form scores are directly comparable to AM-PAC Outpatient Short Form scores for the same functional domain. Standardized short from scores are also comparable to AM-PAC scores generated from AM-PAC computer adaptive tests (CATs). 3. Use the appropriate conversion table to determine percent of functional limitation ranging from 0% - 100% and select the CMS Functional Modifier G-Code. Each percentage range is associated with a two-letter modifier code, as illustrated in the table below. Refer to Appendix C to view AM-PAC Medicare Outpatient Short Form (DOPTA) Conversion Tables and scoring examples. CMS Modifier G-Code CH CI CJ CK CL CM CN Impairment Limitation Restriction Description 0 percent impaired, limited or restricted At least 1 percent but less than 20 percent impaired, limited or restricted At least 20 percent but less than 40 percent impaired, limited or restricted At least 40 percent but less than 60 percent impaired, limited or restricted At least 60 percent but less than 80 percent impaired, limited or restricted At least 80 percent but less than 100 percent impaired, limited or restricted 100 percent impaired, limited or restricted 22

Section 3: Interpreting AM-PAC Standardized Scores 3.1 Functional Stages It may be helpful to have a context for interpreting AM-PAC scores. This has been accomplished by dividing AM-PAC scale scores into ranges that we call functional stages. A functional stage represents a profile of the types of activities a patient might be able to accomplish at different scale score levels. These functional stages are hierarchical and represent functional activities that are increasingly more difficult as one proceeds from a lower to a higher functional stage. These stages were identified using data from over 1,000 cases across postacute care settings. The level of difficulty exhibited by most patients with scores in the specified ranges is described for each functional stage. Functional stages for each of the three AM-PAC scales are described below. Basic Mobility 84 104.9 Strenuous Recreation/Sports: Your score suggests a high level of independence in moving about both at home and in the community. You may be able to participate in most physical activities without much difficulty. 66 83 Moving Around Outdoors: Your score suggests that you are able to walk inside your home and other buildings without any difficulty. You may be able to move about outdoors without any limitations. You should be able to bend over and pick up things without much difficulty. Activities that might be difficult to manage without assistance include climbing a full flight of stairs, bending, kneeling or stooping. Vigorous activities such as playing sports or walking several miles may be very difficult to complete. 52 65 Moving Around Indoors: Your score suggests that you may be able to move about on the ground floor of your home where you are familiar with the environment. Activities that might be difficult to manage without assistance include sitting and standing from a low chair, climbing stairs, bending, kneeling or stooping. You may have some difficulty moving about outdoors and in the community. 34 51 Limited Mobility Indoors: Your score suggests significant difficulty in moving about independently and the need for assistance. You may be able to move about in a small area of your home that has been adapted to eliminate safety hazards. You may have difficulty moving from a sitting to standing position, climbing stairs and you may have a great deal of difficulty moving about outdoors and in the community. -11.95 33 Limited Movement: Your score suggests you may have a lot of difficulty or are unable to get out of your bed, to stand for several minutes and/or to walk short distances. You might have some difficulty completing the most basic mobility tasks including repositioning yourself in bed. 23

Daily Activity 84 115.4 On Your Own: Your score suggests that you may not be having any difficulty completing the daily tasks of bathing, dressing, grooming and eating independently. 62 83 Getting Things Done: Your score suggests that you may require some assistance with housekeeping and laundry, but otherwise you may be able to complete daily tasks of bathing, dressing, grooming and eating independently without much difficulty. 53 61 Difficult Activities: Your score suggests some difficulty in the ability to perform daily tasks. You may be struggling with things such as bathing and dressing. Housekeeping tasks may be too difficult for you to perform. They may experience some difficulties with your fine motor skills such as buttoning clothes, using utensils and combing your hair. 41 52 Daily Tasks are a Struggle: Your score suggests that you may experience significant struggles with performing most daily tasks. You may have significant difficulties in getting dressed and bathed. Tasks that require fine motor skills such as buttoning a shirt or tying your shoes may be too difficult to complete. -2.73 40 No Independent Tasks: Your score suggests daily tasks that require fine motor skills may cause considerable difficulty to the extent that you may be unable to complete them. Bathing and dressing may be so difficult that you may be unable to complete these tasks without assistance. You may be able to feed and groom yourself but with difficulty. You may be unable to tie your shoes or button your shirt. 24