Advertisement

Factors Influencing the Use of Outcome Measures for Patients With Low Back Pain: A Survey of New Zealand Physical Therapists

Janet M Copeland, William J Taylor, Sarah G Dean

Abstract

Background: Rehabilitation of patients with low back pain forms an important component of musculoskeletal physical therapist practice, yet treatment outcomes often are poorly measured.

Objective: The study examined the methods used to evaluate treatment outcomes and factors influencing the use of outcome measures by New Zealand physical therapists.

Design: This cross-sectional study used qualitative and quantitative methods for data collection.

Methods: Two focus groups were conducted: one in a private practice (n=6) and one in a public hospital (n=6). A survey questionnaire was mailed to all private practices listed in a telecommunication database and to outpatient physical therapy departments at public hospitals (n=579).

Results: The mail survey achieved a 65% response rate and showed that physical therapists use improvements in person-specific functional activities as their main outcome measure. Only 40% of the respondents reported using back-related outcome measures. The statistically significant factors determining their use were having a master's degree and an increased level of knowledge of outcome measurement, but these factors explained only 22% of the variance in the logistic regression model. Lack of time, frequently mentioned as a reason for not using standardized outcome measures, did not reach statistical significance.

Limitations: The data collected relate to the physical therapists’ reported or perceived behavior, which may be different from reality.

Conclusion: Physical therapists do not routinely use outcome measures in their clinical practice. A master's degree and increased knowledge were statistically significant factors supporting increased use of outcome measures. Further research is needed on how to convey to practitioners that the information they provide can be useful and can improve patient outcomes.

Outcome measures are tools for measuring the outcomes of health care interventions over time. For physical therapists, important outcomes include changes in patient impairments, activity limitations, participation restrictions, and quality of life,1 as evaluated with patient self-report measures. Outcome measures have been used in research settings on low back pain (LBP) for more than 20 years to evaluate the effectiveness of treatment techniques.24 Studies in New Zealand, Canada, Scotland, England, the United States, and Australia,513 however, have indicated that their use by physical therapists in routine practice is limited. At the same time, there is increasing pressure on physical therapists to demonstrate that their practice is evidence-based and to clearly document the improvements in a person's health status. In addition to these pressures, there has been a gradual change in health outcome philosophies that could have an impact on physical therapy intervention and the choice of outcomes measured. The International Classification of Functioning, Disability and Health (ICF)14 is a framework that promotes a more holistic model of patient care, with the focus on enabling patients to participate in society, in contrast to the previous focus on pathology and impairments. For physical therapists, this approach means a move away from focusing on pain, muscle strength (force-generating capacity), or movement patterns toward a greater emphasis on the individuals’ goals based on activities and participation.

The New Zealand health environment recognizes physical therapists as primary health care providers, and patients with LBP can access a physical therapist directly, without referral or medical prescription. These physical therapists, therefore, tend to see far more acute presentations of LBP than do physical therapists in countries where a referral system is in place. New Zealand also has a comprehensive accident insurance scheme, the Accident Compensation Corporation (ACC), which is a no-fault insurance scheme for all kinds of personal injuries. Most physical therapy intervention, particularly in private practices, is funded by the ACC. The treatment cost, for up to a maximum of 10 sessions for patients with LBP, is covered by the ACC without any prior approval being required, although under some contracts, physical therapists can add a surcharge to the payment.

Low back pain is an increasing problem, especially in developed countries, and a significant health care cost. The ACC estimates that LBP claims costs it around $280 million a year15 to provide services to the New Zealand population, given as 4.1 million (2006 census).16 The treatment of LBP also is an important component of the caseload for many physical therapists. However, the effectiveness of physical therapy interventions for LBP is frequently questioned. This was evident in New Zealand following the publication of the New Zealand Acute Low Back Pain Guide,17 where the only physical therapy intervention to receive an A grade (using the SIGN grading system) for evidence of improved clinical outcomes apart from advice to stay active was manipulation in the first 4 to 6 weeks. For other commonly used physical therapy modalities, there was A grade evidence that they did not provide improvement in clinical outcomes, and for other physical therapy modalities, there was insufficient evidence to comment.

Previous research on the use of outcome measures by physical therapists513 was an important factor in the design of this study. Research in Canada by Cole and colleagues5 into physical therapists’ use of outcome measures was the first physical therapy–based research in this area. They identified lack of knowledge, time constraints, and the perception that outcome measures did not meet the needs of their patients as the main reasons for the failure by physical therapists to use outcome measures. Following this research, a resource book titled Physical Rehabilitation Outcome Measures5 was published; however, follow-up surveys 6 years later6,7 still identified a low level of use. A later Australian study12 was particularly relevant to this research, as Victoria, Australia has a state-run accident insurance scheme, the Transport Accident Commission (TAC), that is similar to New Zealand's ACC in regard to the treatment of injury following motor vehicle accidents. In Victoria, the cost of physical therapy intervention for people involved in motor vehicle accidents is covered by the TAC. The Australian study consisted of 2 mail surveys conducted 6 months apart. During that time, the TAC introduced the Clinical Justification Model, which recommended the use of standardized outcome measures for treatment that was taking longer than 6 weeks. To promote this, the TAC, in conjunction with the Australian Physiotherapy Association, ran an extensive education program. The study showed an increase in use of outcome measures, from 30% in the March survey to 66% in the September survey.12

This study aimed to evaluate the beliefs and attitudes of New Zealand physical therapists relative to their use of outcome measures, specifically their use of 3 commonly used outcome measures: the Oswestry Low Back Pain Disability Index (OLBPDI),18 the Roland-Morris Disability Questionnaire (RMDQ),19 and the Quebec Back Pain Disability Scale (QBPDS).20 A New Zealand–based study was thought important, given the unique funding environment that exists and the increasing pressure by funders of physical therapy for greater accountability. A secondary aim was to explore whether the level of exposure to outcome measures that New Zealand physical therapists have experienced over the past decade has influenced their use of outcome measures, given changes that have occurred in undergraduate and postgraduate training and changes in health outcome philosophy, in particular the ICF framework, with its emphasis on activity and participation.

Method

This was a cross-sectional study using qualitative and quantitative methods of data collection. The focus groups were held in May 2006, and the survey questionnaires were mailed in September 2006. All of the participants were physical therapists working in New Zealand, and the study was limited to those physical therapists treating patients with LBP. A covering letter explaining the study was sent to all participants. Written informed consent was obtained from the participants in the focus groups; for the mail survey, consent was presumed if the survey questionnaire was returned.

Stage 1: Focus Groups

Purposeful sampling was used to identify 2 possible groups to participate in the focus group discussion. One group consisted of physical therapists working in a private practice, and the other group consisted of physical therapists working in the outpatient department of a public hospital. There were 6 participants in each group. We anticipated the 2 groups might reflect different issues, as physical therapists working in the public health system are not under the same immediate commercial pressures as physical therapists working in private practice. A list of preset but open-ended questions was developed to guide the discussions, which were audiotaped for future analysis.

The audiotapes were transcribed verbatim. An inductive approach was taken to allow patterns and categories to emerge, and those that emerged were coded to determine whether any conceptual categories were apparent.21 The categories then were regrouped, based on the nature of the responses and their intensity, to facilitate further analysis. Four themes emerged, which were verified by one of the researchers (SD). Following analysis of these results, the information obtained was used in the design of the mail survey; this enabled the themes to be explored further.

Stage 2: Mail Survey

The 2 main influences in the design of the mail survey were previous research and the information obtained from the focus groups. Close attention was paid to the surveys completed in Canada57 and the survey conducted in Australia.12 The survey conducted by Abrams and colleagues12 contained useful questions relating to physical therapists’ attitudes regarding the use of outcome measures, and the questionnaire specifically targeted physical therapists treating patients with musculoskeletal problems. The most important influence, however, was the information obtained from the focus groups. Questions were included, therefore, on the methods physical therapists are currently using to measure outcomes and their level of satisfaction with these methods. Barriers to the use of outcome measures identified in the focus groups were incorporated into a 23-item questionnaire scored using 5-point Likert scales. Space was left following some questions for free-text comments so that the respondents’ attitudes regarding the use of outcome measures could be more fully captured (see Appendix for a copy of the survey questionnaire).

A power calculation was performed to obtain a point estimate of the proportion of respondents who routinely use standard LBP-related outcome measures with acceptable precision (95% confidence interval [CI]), rather than in comparison with a hypothesized proportion. We assumed that 40% of the respondents would report that they routinely use standard LBP-related outcome measures, with a desired CI of 35% to 45%; Epi Info* was used to obtain a required sample of 372 contacts. This was adjusted to 572 contacts to account for an anticipated response rate of 65%.

The sample population was based on the practice or department, and there was only one response from each sampling unit. The survey questionnaire was sent to the owner or manager of each of the 547 private practices identified from a Telecom database and the senior physical therapist at each of the 32 physical therapy outpatient departments in the public hospitals (which covered all of the district health boards), for a total of 579 primary contacts.

To improve the response rate, a reminder postcard was sent to nonrespondents 2 weeks after the original mailing, and another copy of the survey questionnaire was mailed to nonrespondents after 4 weeks. Data collection stopped 6 weeks after the final mailing.

The data were entered onto Stats Direct (version 2.5.5) for all the analyses except the factor analysis and logistic regression analysis, which were performed using SPSS (version 15.0). Initially, descriptive analysis of data was undertaken. Frequencies and 95% CIs were calculated for appropriate items. Comparisons were made with data from the New Zealand Health Information Service,22 where the data were available, to test whether the sample was representative of the larger population. The information considered to be most relevant to the research question was whether physical therapists were using any of the standard LBP-related outcome measures. A new variable, therefore, was created from the responses to question 9 (Appendix) and included all respondents who had indicated they had used one of the following LBP-related outcome measures in the past 6 months: RMDQ, OLBPDI, or QBPDS. This variable was labeled “Use of outcome measures” and became the dependent variable for the remaining data analysis.

Chi-square tests were carried out on nominal explanatory variables, and odds ratios (ORs) were calculated. The variables that indicated a possible significant association with the use of outcome measures were used for further analysis. A principal component factor analysis with varimax rotation was done on the 23 subitems from question 10. Factors with an eigenvalue of at least 1 were selected. Items with factor loadings with an absolute value of at least 0.5 were aggregated, and factor scores were calculated by simple summation of the relevant item groups, which resulted in the creation of 5 new variables.

To further determine which variables were independently associated with the use of outcome measures, a multivariate logistic regression analysis was performed using the same dependent variable and the other variables that had previously shown a P value of ≤.10 in the univariate analysis. Logistic regression identifies the independent relationship between the dependent variable and the explanatory variables when the dependent variable is dichotomous (ie, use or not of outcome measures). The coefficients of the regression function are expressed as ORs, which indicate the strength of the explanatory variable to the dependent variable, in this case, “Use of outcome measures.”

Role of the Funding Source

Funding for this study was provided by the Wellington Branch of the New Zealand Society of Physiotherapists (Searchwell Grant). They had no input into the study design or analysis. The views expressed in this article do not reflect the views of the New Zealand Society of Physiotherapists.

Results

Focus Groups

Of the 12 participants (1 male, 11 female), 5 were 25 to 29 years of age, 2 were 30 to 39 years of age, and 5 were 40 to 49 years of age. Seven of the participants had completed bachelor's degree programs for their undergraduate training, and they had a range of postgraduate diplomas and certificates.

Four themes emerged from the focus groups. The dominant theme related to possible barriers to the use of outcome measures. Time was the main barrier identified, and in the private practice, there was a strong link between time and the financial pressures that existed to keep the practice viable: Oh, it takes a huge amount more time to do.It comes down to the time component … it's not just time for us, it's time for the patient as well.

However, lack of knowledge also was a major issue, and one focus group could not even name any of the commonly used outcome measures for the treatment of LBP: “Os” something.“Oswes” something, yeah, any other ones?Isn’t there an SPF-12?

There also was a general feeling that outcome measures would not provide them with any useful information. The physical therapists felt that all their patients were individuals with their own unique set of problems, and, therefore, it was irrelevant to look for standardized outcomes: I don’t think we tend to think about the patients as being a group of patients when we see them, because they’re always so completely different.

The other dominant theme related to assessment and measurement, which was overwhelmingly based on the individual clients’ problems, so that goals were set accordingly. There was a strong emphasis on visual observation and the patients’ interpretation of their condition: Just looking at the way they’re getting on and off the bed or in and out of the chair.

Outcome measures were not used by any of the participants in the focus groups.

The other 2 themes related to any possible benefits of the use of outcome measures and future possibilities. Participants found it difficult to identify any possible benefits.

The lack of knowledge about outcome measures was perceived as a possible future threat to the profession, as the participants acknowledged that funders of health services are continually looking for more robust outcomes in order to justify or approve ongoing intervention: I think it is a bit threatening to the profession.

Mail Survey

Three hundred sixty-nine usable survey questionnaires were returned, giving a 65% response rate. This was consistent with the previous power calculations.

Initially, descriptive analysis of the demographic data was undertaken (Tab. 1). The survey identified a higher percentage of men in the profession than data collected by the New Zealand Health Information Service22 and more respondents in the 40- to 49-year-old age group. This may be because the survey was addressed to the practice owner or manager, and these individuals are more likely to be male and older (information from the 2006 New Zealand Society of Physiotherapists database). There were no previous data available on undergraduate or postgraduate qualifications.

Table 1.

Demographic Data (n=369)a

Because the focus groups had identified that outcome measures were not used, participants were asked to give information about the methods they currently used for recording outcomes. The most common treatment outcomes recorded were: patient reports of change in pain levels (95%), changes in range of motion (although the question did not specify whether this was measured or observed) (93%), observed improvement in function (91%), and patients’ individual goals (80%). Participants then were asked about their use of outcome measures, and Table 2 shows the level of reported use of outcome measures during the past 6 months.

Table 2.

Reported Use of Outcome Measures During the Past 6 Months (n=368)a

The relationship between the physical therapists’ qualifications and the variable “Use of outcome measures” is shown in Table 3. A principal component factor analysis with varimax rotation was done on the 23 subitems from question 10 (Appendix), and the results are shown in Table 4. Five factors with eigenvalues greater than 1 were identified:

  • Factor 1: value of outcome measures (as perceived by the physical therapists)

  • Factor 2: patients (the perception that all patients were individuals with a unique set of problems)

  • Factor 3: knowledge (the physical therapists’ level of knowledge about outcome measures)

  • Factor 4: time (the possible role of time in the use of outcome measures)

  • Factor 5: choice (the ability to choose which measures to use)

Table 3.

Relationship Between the Variable “Use of Outcome Measures” and Possible Explanatory Variablesa

Table 4.

Factor Analysis of Responses to Question 10 of the Mail Survey Questionnaire (see Appendix)

To further determine which variables were independently associated with the variable “Use of outcome measures,” a multivariate logistic regression analysis was performed (Tab. 5) using explanatory variables identified as potentially important by the univariate analysis (P value of ≤.10).

Table 5.

Multivariate Logistic Regression Analysis for Predictors of Outcome Measuresa

Discussion

In the multivariate logistic regression analysis, having a master's-level qualification was identified as the variable that exhibited the strongest association with the use of outcome measures (OR=2.5, 95% CI=0.99–6.36). This has not been identified in previous research, although earlier American research had identified that those most likely to claim to have the skills needed to implement evidence-based practice had “professional master's or advanced degrees.”23 The other variable that was important was knowledge of outcome measures (OR=1.75, 95% CI=1.47–2.09), and this has consistently been mentioned in the literature57,12 as one of the factors influencing the use of outcome measures by physical therapists. There may be a link between having a master's degree and knowledge of outcome measures. For a master's degree, a research project usually is undertaken, and if this involves a clinical area, it is likely that outcome measures will be used to evaluate outcomes. This may result in a greater familiarity with outcome measures so that it becomes easier for the physical therapist to integrate them into routine clinical practice. However, having a master's degree contributed an additional association with use of outcome measures, over and above knowledge, indicating that familiarity with the research process itself was a factor that might explain the use of outcome measures.

However, apart from a master's degree, further postgraduate qualifications were not associated with the use of outcome measures. These included specific manual therapy qualifications, which encompassed all of the physical therapists who indicated they had completed courses related to manual therapy (eg, Mulligan, McKenzie, or courses offered by the New Zealand Manipulative Physiotherapists Association). This was the most common postgraduate qualification for New Zealand physical therapists (with 42% reporting some qualification in that field). Time, often given as the reason for failing to use outcome measures, did not reach either statistical or clinical significance in the final analysis (OR=1.11, 95% CI=0.97–1.30) and neither did the variable “patients,” which related to the concept that all their patients were individuals (OR=1.11, 95% CI=0.97–1.28).

Although an increased level of knowledge was an important factor determining the use of outcome measures, the survey confirmed there was still a widespread lack of knowledge and understanding of outcome measures. This is illustrated by some free-text responses in the mail survey: I don’t understand them or how to use them.What they are, why I should use them, how to use them, when to use them.

Some physical therapists expressed concerns regarding the reliability and validity of outcome measures; others questioned whether outcome measures were suitable for their patients because of cultural differences. However the 3 LBP-related outcome measures used in the mail survey have been tested for reliability and validity in numerous studies, and it is generally accepted that they are sound measures to use for evaluating treatment outcomes in people with LBP. It could be that these concerns relate more to the individual physical therapist's level of understanding and knowledge of outcome measures, resulting in a misunderstanding of what outcome measures demonstrate. Some therapists may perceive outcome measures as tools that will replace their assessment and the patient's individual goals rather than as tools for evaluating treatment outcomes. An important concept that emerged in relation to the knowledge variable was how to access the measures and, in particular, how to score them: Often, you can get hold of the questionnaires but not the scoring scales to interpret the results.

The ability to interpret the results from outcome measures could be the component of the knowledge variable that differentiates physical therapists who really understand how to use them effectively and, therefore, include them in their practice from those who do not.

The other variable that was widely discussed was time. Time, as the reason for the failure to use outcome measures, has been identified in previous studies5,6,9,11,12 and was frequently given by the participants of the focus group as the reason for the failure to use outcome measures. One theme closely associated with the time variable that emerged from the qualitative data was the link between time and the financial viability of the physical therapist practice. The majority of physical therapy treatment for LBP in New Zealand is covered by payments from the ACC, and the discontent felt by physical therapists toward the low level of reimbursement was obvious. They felt pressured to see as many patients as possible and did not want to engage in any activity that may have an impact on treatment time. However, it was evident in one focus group and from free-text survey comments that many participants knew practically nothing about outcome measures and had no idea where to find them or how to use them. It is difficult, therefore, to accept time as a valid excuse. Time was being used as the reason for justifying their lack of knowledge, as one participant in the mail survey identified: Time is a factor—maybe an excuse.

New Zealand physical therapists, however, are recording outcomes and are confident in the robustness of these measures. The main finding to emerge from the free text in the mail survey, supported by the focus group findings and the quantitative data, was the widespread use of informal measures. These measures included the observation of movement and function and the use of patients’ goals as methods for measuring the outcome of treatment intervention. The emphasis on patients’ goals is in line with the ICF, but, unfortunately, the physical therapists still tend to use nonstandardized questions during assessment of the patient rather than standardized methods (such as patient self-report questionnaires) to determine outcomes. These quotes from the mail survey are representative of the views expressed by many physical therapists: Patients goals and observed improvement of function are often the best indicators to return to functional ability and for patient satisfaction.My measurements are very patient specific and, therefore, accurate to that patient's circumstances and needs.

These quotes highlight the physical therapists’ view of the patient as an individual and their emphasis on patient-centered care. The physical therapists may see this view as being in conflict with the use of guidelines and outcome measures, which they think assume there is a high level of similarity among people. They do not perceive that outcome measures will add further value to the outcomes they are already collecting. This viewpoint was frequently expressed in the free-text comments and is succinctly summarized in one of the more negative comments on outcome measures from the mail survey: Clinical outcome measures invite recipe-orientated practice.

The view that outcome measures are seen as part of a drive toward “recipe-orientated practice” may be widely held by New Zealand physical therapists. This view is similar to that reported in an American study,23 where the implementation of clinical guidelines was referred to as “cookbook medicine.” It is this attitudinal problem that could be more difficult to change when trying to encourage physical therapists to use outcome measures more routinely.

An Australian study showed that the main factor influencing the routine use of outcome measures was a funding body's requirement for the mandatory reporting of outcome measures.12 The information obtained in the current study would appear to support that view. It seems likely that, if New Zealand funding bodies mandate the reporting of outcome measures, especially if this is attached to payment for treatment, the routine use of outcome measures will increase.

Limitations

There were a number of limitations in this study. One limitation is a weakness inherent in all research undertaken using a survey. All of the data collected related to the physical therapists’ reported or perceived behavior, which may be different from the reality. In a Canadian study,8 where note audits had been used for data collection, the authors found the use of outcome measures was much lower than expected.

Although the free text from the mail survey did give a greater understanding of the physical therapists’ perception of outcome measures, most of the comments were negative. This may have been because those physical therapists not routinely using outcome measures felt their practice was being challenged by the mail survey, with questions effectively putting them on the defensive when answering.

Physical therapists were asked whether they had used any of the named outcome measures within the past 6 months and at any period of time during a patient's treatment. From this information, it could not be identified whether outcome measures were being used routinely or whether they had been used for one patient over that time period. Therefore, although nearly 40% of the respondents reported using LBP-related outcome measures, it is probable that the percentage using them routinely in clinical practice is much lower.

The mail survey was addressed to the practice owner or manager, and, although there was only one response per sampling unit, it often was not clear whether their reply represented the views of their staff or was their individual opinion. Finally, the usable response rate was only 65%, although a further 4% of the questionnaires were returned with the option “I am not involved on the treatment of patients with low back pain” selected. No information is known about the remaining 31%. It would be plausible to suggest that the respondents would be more likely than the nonrespondents to be using outcome measures. Consequently, our results on the use of outcome measures could be more optimistic than what is actually occurring.

Future Research

Several possible areas for further research into the reasons New Zealand physical therapists fail to use outcome measures in routine clinical practice were identified. There is one measure, the Patient-Specific Functional Scale (PSFS),24 that captures information similar to what physical therapists are already recording. It also is quick and easy to administer, and interpretation of the results is straightforward. As the items are patient-specific, it is likely to be more acceptable for people of different cultures. The original study on the PSFS24 was on patients with LBP and showed promising results for reliability, validity, and sensitivity to change. Another study involving patients with LBP showed the PSFS to be more responsive than either the RMDQ or physical impairment measures.25 Studies also have demonstrated positive results for the validity and reliability of the PSFS across a range of musculoskeletal conditions commonly treated by physical therapists.2628 This could be the measure of choice for the patient who makes a rapid and uncomplicated recovery. However, in the mail survey, only 13% of the physical therapists indicated they had used it.

Further research is needed to improve the acceptance of outcome measures by New Zealand physical therapists, for example, by validating outcome measures with New Zealand populations, especially Maori and Pacific Islanders. Maori are the indigenous people of New Zealand and make up 13.6% of the population,16 and a further 6.4% of the population are immigrants from the Pacific Islands.16 Many of these people continue to maintain their own language and a respect for their traditional culture. For health, this involves a wellness or holistic model with a strong spiritual dimension; any outcome measures need to be worded carefully to encompass these cultural dimensions. This issue of cross-cultural adaptation of outcome measures is common to physical therapist practice in all countries with an indigenous population or immigrants from different cultures.

In addition, there is still confusion over the interpretation of the results from outcome measures, and this is an area where more research may be required. Any future education package needs to focus on making sure this aspect of outcome measurement is understood so clinicians learn to interpret the information obtained from outcome measures and alter their practice if needed. Finally, qualitative research may help to identify more clearly what factors differentiate the user from the nonuser, enabling future education to be more focused and to target the factors that are important in changing attitudes as well as behaviors.

Conclusion

This study showed that the routine use of outcome measures for the treatment of LBP among New Zealand physical therapists is still low, despite more than a decade of exposure promoting reasons for their use. Although 40% of the survey respondents indicated they had used LBP-related outcome measures in the past 6 months, it is likely that the routine level of use is far lower, and 22% indicated they did not use any outcome measures. Statistical analysis and free-text comments identified a lack of knowledge and understanding of outcome measures and a lack of time, respectively, as the reasons for failing to use them. This presents a huge challenge to the profession, which needs to (1) convey to practitioners how information obtained from outcome measures can be useful to them and can support improved outcomes for patients and (2) convey that the use of outcome measures does not invite recipe-oriented practice. Otherwise, the profession is likely to have an important component of its practice—the methods used to evaluate treatment outcomes—dictated by external sources.

Appendix.

Appendix.

Questionnaire for Mail Surveya

aLBP=low back pain, SF-36=Medical Outcomes Study 36-Item Short-Form Health Survey questionnaire, ACC=Accident Compensation Corporation, NZSP=New Zealand Society of Physiotherapists.

Footnotes

  • All authors provided concept/idea/research design and data analysis. Ms Copeland provided writing, data collection, project management, fund procurement, and subjects. Dr Taylor and Dr Dean provided facilities/equipment, institutional liaisons, and consultation (including review of manuscript before submission).

  • This article is the result of research that Ms Copeland undertook while completing a postgraduate master's degree at the Rehabilitation, Teaching and Research Unit, Wellington School of Medicine and Health Sciences, University of Otago.

  • Study approval was obtained from the University of Otago Human Ethics Committee.

  • Funding for this study was provided by the Wellington Branch of the New Zealand Society of Physiotherapists (Searchwell Grant). They had no input into the study design or analysis. The views expressed in this article do not reflect the views of the New Zealand Society of Physiotherapists.

  • * Centers for Disease Control and Prevention, 1600 Clifton Rd, Atlanta, GA 30333.

  • StatsDirect Ltd, 9 Bonville Chase, Altrincham, Cheshire WA14 4QA, United Kingdom.

  • SPSS Inc, 233 S Wacker Dr, Chicago, IL 60606.

  • Received March 18, 2008.
  • Accepted July 30, 2008.

References

View Abstract