Link to Figures Show Figures
	Article Text Figures References
	Send Link Send HTML Article

Orthotics and Prosthetics National Office Outcomes Tool (OPOT): Initial Reliability and Validity Assessment for Lower Extremity Prosthetics

Dennis L. Hart, PhD, PT

ABSTRACT

The Orthotics and Prosthetics National Office Outcomes Tool (OPOT) was developed to assess health status, client satisfaction, and prosthetist's perception of function for clients with lower extremity prosthetic needs. The data set comprised 840 adults requiring lower extremity prostheses, as seen in 56 practices by 278 prosthetists from 25 states Clients completed standardized health status and satisfaction questions and prosthetists completed questions pertaining to client demographics and functional abilities at the initial fitting and at the eight-week follow-up. Data support the reliability and construct validity of the OPOT for clients with lower extremity prosthetic needs.

Key Words: Health-related quality of life, outcomes, SF-12, SF-36, lower extremity prosthetics, rehabilitation

The prosthetic industry, like other health care services, is increasingly asked to quantify its quality and value. Accountability for clinical decisions and the associated costs of services is under increasing scrutiny. As a result, the need to measure the results of treatment has been the catalyst for steady progress in the development of outcomes tools appropriate for clients seeking prosthetic services. These outcomes tools must have good reliability, validity, and responsiveness to clinical change to be useful for future research efforts. The tools commonly provide outcomes from the perspectives of the client or the clinician.

Health-related quality of life (HRQL) is assessed via client self-report generic and condition-specific (also referred to as disease-specific) outcomes tools. HRQL tools have been used to investigate clients with lower extremity prosthetic needs. The "gold standard" of generic HRQL tools is the MOS SF-36^1-3 with its shortened version SF-12.⁴ Generic HRQL tools measure multiple constructs of health and well being, including general health, physical functioning, bodily pain, role physical, mental health, social functioning, vitality, and role emotional.^1-7 Generic HRQL tools generally have strong psychometric properties including reliability and validity (both the SF36 and SF-12 have been shown to be reliable and valid).^1-7 These tools are designed to measure HRQL across a broad spectrum of disease states, and have been shown to be able to differentiate functional abilities across interventions and groups of culturally diverse clients. Because the SF-36 and SF-12 were designed to provide a strong assessment of ambulatory adults, they appear suitable for clients with lower extremity prosthetic needs.

Dagum et a1.⁸ assessed outcomes following reconstructive surgery for clients with severe lower extremity trauma by comparing the SF-36 to several other tools. The SF-36 was found to differentiate clients with successful reconstructive surgery compared to salvage surgical procedures. Results could be compared directly to the normal population and other people with severe medical problems. They found that clients with severe lower extremity trauma had reduced physical summary scores, but their mental health summary scores were similar to the normal population. They concluded that, in these clients, disability is primarily physical rather than psychological. Duggan et a1.⁹ studied elderly clients with limb salvage vascular surgery secondary to limb-threatening ischemia and found no difference in generic HRQL outcomes between clients with successful vascular surgery compared to clients with amputations after failed vascular surgery. By using another generic HRQL tool, the Sickness Impact Profile, Greive and Lankhorst¹⁰ found that functional ability decreased with age. Kegel et al.¹¹ used an SF-36 to show that clients with below-knee amputations were more independent than clients with either above-knee or bilateral amputations. Smith et a1.,¹² by using the SF-36, found that clients with isolated, traumatic below-knee amputation had reduced physical functioning and role limitations secondary to physical concerns and pain, but their psychological constructs were not different than those of published aged-matched norms. Therefore, several HRQL tools were able to differentiate functional abilities of clients with lower extremity prosthetic needs.

Disease-specific HRQL outcomes tools have been designed to assess more relevant attributes of clients with specific clinical syndromes. ¹³ These tools generally have stronger content validity and responsiveness, that is, the ability of a scale to detect change, ¹⁴ compared to the generic HRQL tools. ^13,15,16 .tee Prosthetic Profile of the Amputee (PPA)^17-20 is an example of a condition-specific outcomes tool. It was designed to determine which factors are potentially related to prosthetic use by a person with a lower extremity amputation after discharge from rehabilitation. It has been shown to be reliable and valid. Imbedded within the PPA is the Locomotor Capabilities Index.¹⁸ The index provides a comprehensive profile of locomotor capabilities of clients who use lower extremity prostheses. Eleven content items directly related to these clients are used to assess their locomotor abilities. Legro et a1²¹ developed another condition-specific functional tool called the Prosthesis Evaluation Questionnaire (PEQ). They confirmed the reliability of the PEQ, and comparisons with the SF-36 physical functioning scales confirmed its construct validity.

As results become available from outcomes research, many researchers recommend a combination of condition-specific and generic outcomes tools to assess the multitude of content and construct areas reflective of changes in HRQL. Generic tools include measures of psychosocial function and accommodate comparisons between a variety of constructs of interest; condition-specific tools commonly increase responsiveness.^13,22 It is hypothesized that the combination of types of tools will facilitate a more comprehensive assessment of outcomes.^13,22

Clinician-reported functional questionnaires have also been used to assess outcomes in clients with lower extremity prosthetic needs. Melchiorre et al²³ used the Functional Independence Measure (FIM)²⁴ to compare differences in outcomes between a limited number of clients who participated in an inpatient acute-rehabilitation program following traumatic versus vascular unilateral lower limb amputations. There was no difference between FIM scores and functional abilities between the two groups of clients.

As in other clinical trials, ^25,26 if clients can be classified into logical functional categories, stratification of outcomes would be expected. For clients with lower extremity needs, Durable Medical Equipment Regional Carriers (DMERC) use the Health Care Financing Administration's Common Procedure Coding System (HCPCS) code modifiers (K0, K1, K2, K3, and K4)^27,28 adopted by Medicare in 1995 to describe clients with lower extremity prosthetic needs for billing purposes. The modifiers allow classification of clients according to their functional abilities and potential. To my knowledge, there are no data demonstrating that prosthetists can reliably classify clients by the HCPCS modifiers. However, because the classifications relate to potential functional abilities of the clients, the DMERC functional levels form a common, widely used condition-specific client classification system that can be used to assess construct validity for outcomes tools. The lowest functional level represents clients with no ability or potential to ambulate or transfer safely with or without assistance, and a prosthesis would not enhance their quality of life or mobility. The highest functional level represents clients with the ability or potential for prosthetic ambulation that exceeds basic ambulation skills, exhibiting high impact, stress, or energy levels typical of the prosthetic demands of a child, active adult, or athlete.^27,28

No studies were found that compared client self-reported health status/functional abilities, client satisfaction, and prosthetist perception of functional abilities for clients requiring lower extremity prosthetic services. Therefore, the purpose of this investigation was to develop an outcomes tool for clients seeing a prosthetist for lower extremity prosthetic needs and to test the internal consistency reliability and construct validity of the tool.

Methods

Subjects

Subjects included 840 consecutive adults (29.8% females, 70.2% males) seen in 56 practices by 278 prosthetists from the 25 states that were participating in Focus On Therapeutic Outcomes data collection between January 1998 and January 1999 (Table 1 ). Prosthetists were instructed to enter all clients with new or replacement lower extremity prosthetic needs. The practices reported seeing an average of 10.2 clients per month with new or replacement lower extremity prosthetics needs. On average, the practices had 2.0 full-time equivalent (FTE) prosthetists who were certified by the American Board of Certification in Orthotics and Prosthetics (ABC), 0.2 FTEs with other prosthetist certifications, and 0.6 FTEs of noncertified prosthetists. The practices had an average of 0.6 FIE registered technicians certified by the ABC, 0.06 FIFE technicians with other registrations, and 2.3 FTE nonregistered technicians. Among the participating facilities, 88.7% were accredited by the ABC, 1.6% were accredited by the Commission on Accreditation of Rehabilitation Facilities (CARF), and 6.5% were accredited by the Joint Commission on Accreditation of Healthcare Organizations (JCAHO).

Of the 278 participating prosthetists, 85.9°/a were male, 26.4% were licensed, 66.9% had bachelor of science degrees, 12.5% had associate of science degrees, and 7.0% had masters of science degrees. On average, it had been 12.5 years since the prosthetists received ABC certification, and they had 14.6 years of direct client care. The prosthetists reported working 42.8 hours per week with 26.1 client visits per week. The prosthetists had the following credentials: 39.4% ABC CP, 38.8% ABC CPO, 10.4% ABC CO, 1.1% ABC Registered Associate, 4.0% BOC, and 2.5% physical therapists.

Procedures

Practices were identified by the Orthotics and Prosthetics National Office as practices that were interested in participating in the outcomes project and were seeing at least four clients per month with new or replacement lower extremity prosthetic needs. Once identified, Focus On Therapeutic Outcomes, Inc. (FOTO) (Knoxville, TN), an independent medical data management company, was consulted to register the practices and prosthetists, instruct the staff in the outcomes process, and collect and analyze the data. FOTO developed the outcomes instrument with the assistance of the Orthotics and Prosthetics National Office (Alexandria, VA).

All clients completed health status and satisfaction questionnaires. Prosthetists completed separate questionnaires. All questionnaires were completed at initial fitting and at follow-up eight weeks later. Questionnaires were mailed to FOTO for data entry and analyses. If errors in entry were identified, FOTO staff contacted the prosthetists, and the errors, if possible, were corrected and resubmitted.

Outcomes Instrument

The outcomes tool assessed three construct areas: the client's perception of their overall health-related quality of life (health status) and functional abilities; the clients' satisfaction; and the prosthetist's perception of the functional abilities related to ambulatory status.

The core health status questionnaire contained the entire acute version of the SF-12.⁴ Questions assessing physical functioning and bodily pain were expanded by including the complete physical functioning scale (PF-10) and the complete bodily pain scale (BP-2) from the MOS SF-36.^1-3,5 Eleven new questions were added to the physical functioning scale. The new questions were designed to reduce the expected ceiling and floor effects of the PF-10 and to determine if any of the new questions improved the construct validity of the PF-10. The SF-36,1-3, 5 PF-10, and BP-2 have been shown to be reliable and valid.^{1-3,5-6,29-30} There are no reliability or validity assessments of the additional questions to the physical functioning scale, and therefore the new questions were analyzed separately.

The client's responses to the health status questions were transformed into scores from 0 to 100 for each of the eight functional .scales. A measure of overall health status was calculated by averaging the eight functional scales including the PF-10, BP-2, and the six other scales from the SF-12.

Two component summary scores (SF-12 Physical Component Summary Score [PCS] and SF-12 Mental Component Summary Score [MCS]) were calculated from the SF-12 ^4,7 PCS and MCS scores were scaled to have a mean of 50 and a standard deviation of 10 for the general population of the United States following published algorithms.⁵ Responses to the health status questions represent the client's perception of their functional abilities/health and well-being at the time of questionnaire completion. The health status scales and summary scores are independent of the type of client, clinician, and treatment provided.

Client satisfaction was measured by using a new tool developed by the Orthotics and Prosthetics National Office Committee on Outcomes. The client completed 13 satisfaction questions covering the following perceptions: receiving an appointment within a reasonable time period, location of office, courtesy received from staff, waiting room time, waiting time beyond scheduled appointment time, training, overall care, prosthesis appearance, pain during use, ability to walk with the prosthesis, quality of life, insurance company coverage, and ability to express client concerns about the limb. These questions assessed the performance of the prosthetist and staff. The client was asked five satisfaction questions that assessed the importance to the client of the areas of satisfaction being questioned. These questions covered the following perceptions: receiving an appointment within a reasonable time period, location of office, courtesy received from staff, waiting room time, and ability to express client concerns about the limb. Each response was transformed into a score from 0 to 100, and the transformed values of the two groups of questions were averaged. This provided an assessment of two overarching constructs of client satisfaction: performance of the staff and importance of the satisfaction questions to the client.

To assess the prosthetist's perception of the functional abilities related to the client's ambulatory abilities, the prosthetist was asked to assess the client's ability to climb stairs, walk, and use assistive devices. The prosthetist's responses for each question were transformed into a score from 0 to 100, and the transformed scores were averaged to produce the prosthetist's perception of client's ambulatory abilities (PROS).

Change in functional abilities between initial fitting and follow-up for each functional scale, the two component summary scores, overall health status, the prosthetist's perception of client function, and client satisfaction were calculated as effect size scores (discharge-intake/standard deviation of the intake scores)^32-33 Change in overall health status was calculated by averaging the effect size scores from the eight functional scales. Standardized effect size scores from overall health status, two component summary scores, and the prosthetist's perception of function represent measures of quality, that is, the unit of functional improvement per episode.

Data Analyses

For each functional scale with two or more questions, internal consistency reliability was assessed by using a Cronbach's alpha. ³⁴ An alpha coefficient estimates the homogeneity of the items and represents the intercorrelation among different items in each functional domain of the tool.

Because there is no accepted "gold standard" measure of function in clients with lower extremity prosthetic needs, the concept of construct validity was used to assess the validity of the proposed functional scales, as described elsewhere. ¹³ Several theories were proposed. The extent to which the functional measures yielded results concordant with the theory provided support for the validity of the measures. ¹³ In this study, if certain correlations were observed, the validity of the functional measures would be supported (Table 2 ).

Correlations between all functional scales, summary scores, overall health status, PROS scores, and client satisfaction were assessed with a Pearson product moment correlation for initial fitting and follow-up. Analysis of covariance (ANCOVA) was used to examine the hypotheses about validity concerning the differences in improve ment in function as perceived by the client or the prosthetist across DMERC functional levels. ANCOVAs were run for each of the twelve effect size change scores: eight functional scales, PCS, MCS, overall health status, and PROS. Intake scores specific to each of the functional measures were used as the covariate. Scheffe post hoc pairwise mean comparisons were used to detect differences between levels of function as assessed by DMERC levels when necessary. Alpha was set to 0.05.

To describe the attributes of hierarchical order, construct validity, and internal consistency of the larger physical functioning construct, that is, PF-21 (the PF-10 plus the 11 new questions) and to determine if any new combinations of questions improved these attributes, the data were fit to the Rasch model by means of the WINSTEPS³⁵ computer program.

Results

Descriptive Statistics

Time between initial fitting and followup averaged 82.2 ± 44 days. Functional measures statistics are presented in Table 3 . Distribution of initial fitting responses, from which floor and ceiling effects can be determined, are presented in Table 4 . Normality plots were run on all transformed scales and summary component scores, and all were considered normally distributed.

Reliability

Cronbach's alpha for each functional scale with two or more questions are listed in Table 5 . In all cases, the internal consistency alphas exceeded alpha = .61. The scales representing physical functioning, bodily pain, and role physical had alphas exceeding 0.80. Alphas greater than 0.70 indicate that the items grouped well enough together to perform adequately as composite measures for discriminating between groups of clients.¹⁶

Construct Validity

Correlations between functional scales, summary scores, prosthetist's perceptions, and client satisfaction at initial fitting and follow-up are presented in Table 6 . Scales representing physical constructs were more closely correlated with other physical constructs, including the PCS score, as compared to mental constructs. Of the four scales representing physical functioning constructs, two at initial fitting and three at follow-up met our correlational cut-off of r > .6. Mental health constructs had stronger correlations to other mental health constructs, including the MCS score as compared to physical constructs. Of the four scales representing mental health constructs, three at initial fitting and three at follow-up met our correlational cut-off of r > .6. The PCS and MCS were not correlated (r = .11 at initial fitting and r = .15 at follow-up, P > .OS). These correlations confirm the SF-12 summary scores commonly measure orthogonal concepts as they were designed to do, which supports the idea that the tool functioned in the prescribed manner. Therefore, the data support the validity of the SF-12 summary component scores for these clients.^1-7

The correlations-between PROS and the PF-10 did not make the cut-off of r > .6 (initial fitting r = .58 and followup r = .55). Likewise, the correlations between PROS and the PCS were below 0.6 (initial fitting r = .38 and followup r = .42). Client satisfaction was unrelated to any other scale, but the two overarching constructs of satisfaction were related (r = .5 at initial fitting and r = .43 at follow-up).

Change in overall health status (F = 5.4, P < .001), PCS (F =13.6, P < .001), physical functioning (F = 17.06, P < .001), bodily pain (F = 4.48, P = .004), role physical (F = 10.02, P < .001), and the prosthetist's perception of function (F = 29.13, P < .001) were able to differentiate between DMERC functional levels. The higher the DMERC levels, the better the improvement in function. No clients were classified in DMERC KO classification for change scores, which was expected. According to the post hoc pairwise analyses, the PROS differentiated all six possible pairs of the four highest DMERC functional levels (Scheffe pairwise comparisons all P < .001). The PF-10 differentiated between five of the six possible pairs of DMERC functional levels (Scheffe pairwise comparisons all P < .001). The PCS and role physical scale differentiated between three of the six possible pairs of DMERC functional levels (Scheffe pairwise comparisons all P < .001). No scales or summary scores representing mental constructs were able to differentiate DMERC levels.

In general, clients with transtibial-BK amputations had slightly higher functional scales and PROS scales at initial fitting and follow-up compared to clients with transfemoral amputations. All change scores for transtibial-BK amputations were equal to or greater than clients with transfemoral amputations (Table 7 ). PCS (F = 6.4, P = .012), physical functioning (F = 7.3, P = .007), role physical (F = 4.2, P = .04) and PROS (F = 18.1, P < .001) differentiated between types of amputation. Younger (< 60 years) clients demonstrated greater improvement in overall health status (F = 6.6, P = .01), PCS (F = 15.8, P < .001), physical functioning (F= 25.1, P < .001), bodily pain (F= 9.9, P = .002), role physical (F = 7.3, P = .007), role emotional (F = 9.4, P = .002), and PROS (F= 29.1, P < .001).

The Rasch analyses confirmed clinically logical hierarchical ordering (construct validity) of the PF-10 and the PF21 questions at initial fitting and followup. Question order was similar between intake and follow-up (test-free)^29,36 There was strong internal consistency alpha = .91) of the question responses at fitting and follow-up. There was considerable overlapping of questions (standard error of calibration < 0.15 logits)³⁷ One question of each overlapping pairs of questions was deleted one at a time for data from initial fitting and followup, and Rasch analyses were rerun on the reduced data sets until a set of physical functioning items remained without question overlapping (standard error of calibration > 0.15 logits).³⁷ The remaining data sets (initial fitting and follow-up) contained 15 physical functioning questions including the two questions used in the calculation of the SF-12 PCS. The remaining 15 questions (PF-15) demonstrated clinically logical hierarchical ordering (construct validity) with similar ordering between fitting and follow-up (testfree). Internal consistency reliability at initial fitting alpha = .89) and follow-up alpha = .90) were strong. The PF-15, compared to the PF-10, had a more normal distribution with a slight ceiling effect, no overlapping of questions, and better fit of the data, thus presenting stronger construct validity than the PF-10.

Discussion

The results of this study confirm the internal consistency reliability, construct validity, and responsiveness of the OPOT. The tool was able to differentiate outcomes between several groups of clients with lower extremity prosthetic needs on the bases of functional abilities, level of amputation, and age. The tool was able to differentiate outcomes by different constructs, that is, mental versus physical constructs, and from different perspectives, that is, the client's perspective of functional abilities versus the prosthetist's perception of the client's functional abilities.

The core instrument of the OPOT is the SF-12 ^4,7 The strength of this generic HRQL questionnaire is the ability to quantify health status across different health constructs. The disability associated with lower extremity amputations has been reported as primarily physical compared to psychological.^8,12 In this study, the MCS at initial fitting was similar to the normal population (Table 4 ), but the PCS was 1.3 standard deviations below normal. Therefore, our data support the finding that disability associated with lower extremity prosthetic need is primarily physical. There was less than minimal³¹ improvement in the client's perception of improvement in either the physical or mental constructs (standardized effect size scores PCS = 0.15 and MCS = 0.1). In spite of the small improvements, the physical component summary scale and physical functioning scale change scores were able to differentiate across various groups of clients. Younger clients perceived better improvement than older clients, clients with transtibial-BKs perceived better improvement than transfemoral AK amputation, and clients in the higher DMERC functional levels perceived better improvement than clients in lower DMERC levels. To my knowledge, this is the first time DMERC functional levels, a billing coding HCPCS modifier system,^27-28 has been correlated to functional improvement. The findings justify the validity of the modifier classification paradigm. Whether these small yet statistically significant measures of improvement will prove to be changes that are clinically important and meaningful is a subject that awaits further study.

The findings support the validity of the SF-12 generic HRQL tool in a sample of clients with lower extremity prosthetic needs. The MCS and PCS were not correlated (r = .11 at initial fitting and r = .15 at follow-up, both P > .OS). This supports the fact that the SF-12 measures orthogonal constructs of health status (Table 6 ) as it was designed to do.^1-7

The prosthetist's perception of function (PROS) was the strongest measure for differentiating the functional levels of the client (DMERC). PROS was capable of distinguishing the six possible pairs of combination of the DMERC K1, K2, K3 and K4 functional levels (ANCOVA F = 29.1, P < .001, Scheffe post hoc analyses all P < .OS). No other outcomes measure was able to differentiate all six pairs of comparisons. The PF-10 differentiated five pairs, but was unable to differentiate between DMERC K1 and K2. (Scheffe post hoc P = .059) The acceptable correlation between variables was set a priori at r > 0.6 for construct validity tests of association. PROS was related to the PF-10 at initial fitting (r = .58) and follow-up (r = .55), but the correlations were not sufficient to support construct validity. It appears the prosthetists' questions pertain more to the physical functioning than to the psychosocial concerns of the client, which was expected.

As predicted, client satisfaction assesses different constructs compared to the client's perception of their functional abilities and the prosthetist's perception of the client's functional abilities (Table 6 ). Thus, the data support that the OPOT measures three different broad spectra of health constructs.

The overall health status and the physical functioning scale used the questions from the SF-36 PF-10.^1-3,5-6 Eleven additional questions were added in the hopes of improving construct validity of the physical functioning scale (Table 4 ). Of the 10 questions included in the PF-10, bending (including kneeling and stooping) and walking several blocks were eliminated because they overlapped (item calibration < 0.15 logits)³⁷ with other questions, and therefore did not differentiate levels of difficulty of the questions for clients with lower extremity prosthetic needs. By adding the seven new questions, six of which assessed the client's abilities to perform easier activities (excluding participating in recreational sports), stronger construct validity was obtained. There is still need to add questions that assess higher levels of abilities, but the PF-15 appears stronger than the PF-10^1-3,5 in quantifying the physical functioning of these clients.

There are similarities between the PF-10/PF-15 and the Locomotor Capabilities Index (LCI).¹⁸ As a condition specific tool, the LCI should have strong content validity, which it does. The PF-10 and PF-15 also have strong construct validity. However, the purposes of the two tools are different. The LCI was part of a larger tool designed to profile characteristics of clients that allow prediction of prosthetic use. The PF-10 and PF-15 were designed to assess perceived functional abilities of clients independently of clinical concerns. We await prospective studies comparing these two tools directly, so that in the future, we can determine how or if generic and condition-specific outcomes tools should be merged to develop a stronger outcomes tool. The inability of the PPA to measure generic health status eliminates the possibility of direct comparisons to national norms and measures of multiple health dimensions.

Previous research (Hart 1998 unpublished data) supports that the intake score (baseline health status) is one of the best predictors of change in health status outcome.^38-39 Therefore, the technique of controlling for the intake scores with an analysis of covariance by using the intake scores as the covariate was followed. Because we used the ANCOVA on standardized change scores, which also tend to control variability of the intake scores, the results should represent conservative statistical findings.

The change scores of generic HRQL measures quantify functional improvement and represent measures of clinical quality. The data measure HRQL across multiple health dimensions and have good reliability and construct validity. This supports previous literature.^l-7 Measures of health status/functional abilities have been recommended as measures of functional status for clients in rehabilitation⁴⁰ and medical clinical trials^5-6 because of the breadth of health status constructs assessed. Some report that condition-specific measures are more responsive to clinical change. ^13,15,41 The OPOT data demonstrate good ability to differentiate between different groups of clients, thus offering good construct validity, but the change scores were in a range between negligible and small.³¹ Thus, although the responsiveness to clinical change was low, the ability to differentiate between differences in improvement of health status in these clients was good. Work needs to be completed to determine if the changes are clinically meaningful.

Because there was such a difference between the physical and mental component summary change scores, the data suggest the need to differentiate effects related to physical and mental constructs. This supports previous literature with clients with lower extremity prosthetic needs^8,12 and patients with cervical syndromes.¹⁵

Limitations

There was no randomization of clients or attempt to randomly select the participating clinics. The results represent analyses on a clinical database where providers were solicited to participate, so that their outcomes could be assessed and compared in a standardized manner with national external benchmarks. Because this is a clinical database from practices electing to participate for many reasons, there are threats to external validity including lack of a priori planning, missing observations, selection bias, and referral bias.^42-44

Because we could not monitor the effect of the training process of data collection for the FOTO outcomes process, we have no assessment of reliability of the method of following the data collection process. No test-retest study was done. Rigorous data quality procedures were in place and used for automated monthly data checking for completeness. No attempt was made to assess the level of vigor with which the provider followed the proposed data quality standards and recommended implementation processes.⁴⁵

The clients in this study came from a sample of practices across the United States covering several prosthetic clinical settings. Participating prosthetists had a wide diversity of clinical skills, techniques, and experiences. The clients represent a sample with a wide variety of characteristics, including clinical, demographic, and administrative variables. The sample does not represent a well-defined population, but represents a common client population in prosthetic settings. This is an inherent strength as well as a threat to external validity for this type of project.

In spite of these limitations, the design and statistical techniques allow for a robust analysis of the differences in outcomes associated with prosthetic needs. The author hopes the analyses will facilitate discussion, improved research designs, implementation, and interpretation of future studies on functional abilities and health and well-being in clients who receive prosthetic services.

Conclusion

A new prosthetic outcomes tool was developed and tested for initial reliability and construct validity. The tool quantified three constructs, including the client's perception of function/health and well-being, client satisfaction, and the prosthetist's perception of the functional abilities of the client. The tool was shown to have good internal consistency reliability, construct validity, and responsiveness to functional change and is ready for use in future research efforts.

Acknowledgements

The author would like to express his gratitude and appreciation to the Orthotics and Prosthetics National Office whose financial support made this outcomes study possible. Appreciation is also extended to Mr. Jim Alaimo, CPO, whose review and constructive criticism throughout the project strengthened the final product.

References:

1. Ware JE, Sherbourne CD. The MOS 36item short-form health survey (SF-36): I. Conceptual framework and item selection. Med Care. 1992;30:473-183.
2. McHorney CA, Ware JE, Raczek AE. The MOS 36-item short-form health survey (SF-36): II. Psychometric and clinical tests of validity in measuring physical and mental health constructs. Med Care. 1993;31: 247263.
3. McHorney CA, Ware JE, Lu JFR, Sherbourne CD. The MOS 36-item short-form health survey (SF-36): III. Tests of data quality, scaling assumptions, and reliability across diverse client groups. Med Care. 1994; 32:40-66.
4. Ware JE, Kosinski M, Keller SD. A 12item short-form health survey: construction of scales and preliminary tests of reliability and validity. Med Care. 1996;34:220-233.
5. Ware JE, Snow KK, Kosinski M, Gandek B. SF-36 Health Survey Manual and Interpretation Guide. Boston, MA: The Health Institute, New England Medical Center, 1993.
6. Ware JE, Kosinski M, Keller SK. SF-36 Physical and Mental Health Summary Scales: A User's Manual. Boston, MA: The Health Institute, 1994.
7. Ware JE, Kosinski M, Keller SD. SF-12: How to Score the SF-12 Physical and Mental Health Summary Scales Boston, MA: The Health Institute, New England Medical Center, 1995.
8. Dagum AB, Best AK, Schemitsch EH, Mahoney JL, Mahomed MN, Blight KR. Salvage after severe lower-extremity trauma: are the outcomes worth the means? Plast Reconstr Surg. 1999;103(4):1212-1220.
9. Duggan MM, Woodson J, Scott TE, Ortega AN, Menzoian JO. Functional outcomes in limb salvage vascular surgery. Am J Surg. 1994:168(2):188-191.
10. Greive AC, Lankhorst GJ. Functional outcome of lower-limb amputees: a prospective descriptive study in a general hospital. Prosthet Orthot Int. 1996;20(2):79-87.
11. Kegel B, Carpenter ML, Burgess EM. Functional capabilities of lower extremity amputees. Arch Phys Med Rehabil. 1978; 59(3):109-120.
12. Smith DG, Horn P, Malchow D, Boone DA, Reiber GE, Hansen ST, Jr. Prosthetic history, prosthetic charges, and functional outcome of the isolated, traumatic belowknee amputee. J Trauma 1995;38(1):44-47.
13. Binkley JM, Stratford PW, Lott SA, Riddle DL. The lower extremity functional scale (LEFS): scale development, measurement properties, and clinical application. Phys Ther 1999;79(4):371-383.
14. Kirshner B, Guyatt G. A methodological framework for assessing health indices. J Chronic Diseases. 1985;38:27-36.
15. Riddle DL, Stratford PW Use of generic versus region-specific functional status measures on clients with cervical spine disorders. Phys Ther 1998;78:951-963.
16. Patrick DL, Deyo RA, Atlas SJ, Singer DE, Chapin A, Keller RB. Assessing healthrelated quality of life in patients with sciatica. Spine. 1995;20(17):1899-1909.
17. Gauthier-Gagnon C, Grise MC. Prosthetic profile of the amputee questionnaire: validity and reliability. Arch Phys Med Rehabil. 1994;75(12):1309-1314.
18. Gauthier-Gagnon C, Grise MC, Lepage Y The locomotor capabilities index: content validity. J Rehabil Outcomes Meas.1995; 2(4):40-46.
19. Gauthier-Gagnon C, Grise MC, Potvin D. Predisposing factors related to prosthetic use by people with a transtibial and transfemoral amputation. Journal of Prosthetics and Orthotics. 1998;10(4):99-109.
20. Grise MC, Gauthier-Gagnon C, Matineau GG. Prosthetic profile of people with lower extremity amputation: conception and design of a follow-up questionnaire. Arch Phys Med Rehabil. 1993;74(8): 862870.
21. Legro MW, Reiber GD, Smith DG, del Aguila M, Larsen J, Boone D. Prosthesis evaluation questionnaire for persons with lower limb amputations: assessing prosthesis-related quality of life. Arch Phys Med Rehabil. 1998;79(8):931-938.
22. Kantz ME, Harris WJ, Levitsky K, et al. Methods for assessing condition-specific and generic functional status outcomes after total knee replacement. Med Care. 1992;30: MS240-MS252.
23. Melchiorre PJ, Findley T, Boda W Functional outcome and comorbidity indexes in the rehabilitation of the traumatic versus the vascular unilateral lower limb amputee. Am J Phys Med Rehabil. 1996;75 (1): 9-14.
24. Hamilton BB, Granger CV, Sherwin FS, Zielezny M, Tashman JS. A uniform national data system for medical rehabilitation. In Fuhrer MJ, ed. Rehabilitation Outcomes: Analysis and Measurement. Baltimore: Brooks, 1987.
25. Delitto A, Cibulka M, Erhard R, Bowling RW, Tenhula JA. Evidence for use of an extension-mobilization category in acute low back syndrome: a prescriptive validation pilot study. Phys Ther 1993;73: 216222.
26. Werneke M, Hart DL, Cook D. A descriptive study on the centralization phenomenon: a prospective analysis. Spine. 1999;24(7):676-683.
27. AdminaStar Federal. Region B DMERC Supplier Manual. AdminaStar Federal: Indianapolis, IN, 1993.
28. St. Anthony's HCPCS Level II Code Book. 1999 Edition. St. Anthony Publishing, Inc.: Reston, VA, 1999.
29. Haley SM, McHorney CA, Ware JE. Evaluation of the MOS SF-36 physical functioning scale (PF-10): I. Unidimensionality and reproducibility of the Rasch item scale. J Clin Epidemiol. 1994;47(6):671-684.
30. Hart DL, Dobrzykowski EA. Effect of exercise history on outcomes. North American Clinics in Physical Therapy. In press 1999.
31. Cohen J. Statistical Power Analysis for the Behavior Sciences New York, NY: Academic Press Inc. 1977.
32. Kazis LR, Anderson JJ, Meenan RE Effect sizes for interpreting changes in health status. Med Care. 1989;27:Sl78-Sl89.
33. Stratford PW, Binkley JM, Riddle DL. Health status measures: strategies and analytic methods for assessing change scores. Phys Ther 1996;76:1109-1123.
34. Cronbach LJ. Coefficient alpha and the internal structure of tests. Psychometrika. 1951;16:297-334.
35. Linacre JM, Wright BD. A User's Guide to WINSTEPS: Rasch-Model Computer Program. Chicago, IL:MESA Press, 1998.
36. Hart DL. Assessment of Unidimensionality of Physical Functioning in Patients Receiving Therapy in Acute, Orthopedic Outpatient Centers MESA Press, in press 1999.
37. Silverstein B, Kilgore KM, Fisher WP, Harley JP, Harvey RF. Applying psychometric criteria to functional assessment in medical rehabilitation: I. Exploring unidimensionality. Arch Phys Med Rehabil. 1991;72: 631-637.
38. Deyo RA, Diehl AK. Psychosocial predictors of disability in patients with low back pain. J Rheumatol. 1988;15:1557-1564.
39. Hart DL, Dobrzykowski EA. Effect of clinical specialist certification on clinical outcomes. J Orthop Sports Phys Ther Submitted 1999.
40. Jette AM. Using health-related quality of life measures in physical therapy outcomes research. Phys Ther. 1993;73:528-537.
41. Patrick DL, Deyo RA. Generic and disease-specific measures in assessing health status and quality of life. Med Care. 1989; 27: S217-S232.
42. Pryor DB, Lee KL. Methods for the analysis and assessment of clinical databases: the clinician's perspective. Stat Med. 1991; 10:617-628.
43. Jette DU, Jette AM. Physical therapy and health outcomes in patients with knee impairments. Phys Ther. 1996;76:1178-1187.
44. Jette DU, Jette AM. Physical therapy and health outcomes in patients with spinal impairments. Phys Ther 1996;76:930-945.
45. Hart DL, Geril A, Pfohl R. Outcomes process in daily practice. PT Magazine. September 1997; 68-77.