Relationship of professional involvement to clinical behaviors of physical therapists.

Physical therapy has been described as an occupation that is in the process of professionalization. [1-4] The focus of this study was the extent to which avowed behaviors of physical therapy clinicians reflect the professionalization of that field. The broad hypothesis was that clinicians who are involved in the formal and informal culture of the occupation are more likely than others to exhibit a professional role that utilizes high-skill procedures and allows the practitioner greater influence over the course of treatment.

Professionalization

Occupational sociologists identify professions as occupations having a highly specialized, esoteric knowledge base; a service ethic; and autonomy over the task to be accomplished. [5-9] Since the mid-1800s, those occupations considered professions have gained their high status through the association of their work with scientific research and higher education. Professionalization of an occupation may occur in degrees-, thus, a continuum of professional status exists among occupations. [5-9] Furthermore, professionalization occurs at two levels: occupational collective) and individual. [1,5,6] Complete professionalization requires changes at both levels to promote compatibility between the structure, values, and norms of the occupation and the attitudes, values, and behaviors of the individual practitioners. [1,5,6]

Professionalization of Physical Therapy. The Occupational Level

The notion of a continuum of professionalization suggests the existence of emerging or marginal professions"; physical therapy is moderately professionalized or a mid-level profession. [1,2] The main thrust toward professionalization of mid-level occupations is by associations organized by leaders in the field. [5] Senters noted a likelihood for physical therapy to accelerate toward professionalization in the 1970s.2 Morrow found that from 1975 to 1982, the American Physical Therapy Association was engaged in significant activity clearly related to professionalization. [1]

Senters examined the profession of physical therapy at both occupational and individual levels and reported that the variety of tasks comprising physical therapy has increased. [2] As the range of tasks increases, so does the likelihood that tasks will be ranked in terms of complexity and become associated with specific worker categories. [2] Additionally, as this range of tasks is extended, the potential for additional specialization develops, and where the newly emerging tasks encompass evaluation and decision making, the scope of responsibility is extended. [2] Both specialization and increased responsibility accelerate professionalization at the occupational level.

Professionalization of Physical Therapy. The individual Level

Professionalization at the individual level requires changes in the attitudes, values, and behaviors of the practitioner. [1,6] Concurrent changes expected include changes in physical therapists' perceptions of the varying degrees of complexity and levels of competence required for application of different physical therapy procedures. This task differentiation would be reflected in physical therapists' ability to rank procedures consistently in terms of the level of competency required to perform them. Certain procedures would be expected to be used solely by physical therapists, whereas other procedures would be used by support personnel as well as physical therapists. [2]

In a study by Uili et al, procedures that both physical therapists and physical therapist assistants were trained to perform were ranked by the authors as "technical," whereas those procedures that only a physical therapist was trained to perform were ranked as professional." [4] A ranking of physical therapy procedures according to level of complexity by physical therapists has been used by Ritchey et al. [10] Senters found that physical therapists used a wide range of procedures, that some procedures were used with greater frequency than others, and that physical therapists used support personnel more with some procedures than with others. [2]

Individuals in the mid-level professions show greater variability than other professionals in their individual professionalization. [5,6] Ritzer applied occupation characteristics of professionalization to individual characteristics of professionalization to ascertain where a specific individual lies on the individual professional continuum. [6] Included in Ritzer's six dimensions of individual professionalization are membership in occupational associations and the degree of active involvement in the formal and informal occupational cultures. [6] Ritzer stated that involvement in the formal occupational culture can be determined by an individual's membership and activity level in the occupational associations. Participation in the informal culture can be determined by an individual's formal and informal activities within the occupation and the frequency of contact with colleagues. [6] Fitzer hypothesized that the greater the individual's involvement in the profession, the greater will be the degree of individual professionalization. [6]

Morrow compared delegate and non-delegate members of the APTA and found that the general and specific attitudes of individuals who were more involved in the APTA (delegates) were uniformly and significantly more favorable toward the continuing professionalization of physical therapy than individuals who were less involved in the APTA (nondelegates)., Senters compared physical therapists who reported use of support personnel (utilizers) with physical therapists who did not report use of support personnel (nonutilizers). This comparison showed that during the two years immediately preceding the study, a majority of the utilizers had participated in continuing professional education programs compared with fewer than half of the nonutilizers. [2]

Senters suggested that physical therapists must be willing to delegate some responsibility to support personnel and that failure to relinquish a portion of their responsibility hinders both the physical therapy professional and the professionalization of physical therapy. [2] The findings of Ritchey et al suggest that physical therapists' actions are an important determinant of physician referral and will be the most important determinant of role expansion for physical therapy. [10] The reported studies imply that the ability of physical therapists to develop and implement strategies for change is dependent on having information about the attitudes, values, and behaviors of physical therapists. [1,2,4,10] The purpose of this study was to describe 1) the ranking of 24 physical therapy procedures according to five dimensions of complexity and 2) the relationship of the ranking of a procedure to delegation of the procedure to support personnel. The study also was designed to measure the level of professional involvement of the respondents and to examine the relationship between a high level of professional involvement and the use of high-complexity procedures. The following hypotheses were posed: 1) There would be consensus in the ranking of procedures according to the level of complexity, 2) ranking of a procedure according to level of complexity would be related to the delegation of the procedure to support personnel, and 3) physical therapists with greater APTA and continuing education involvement would use the high-complexity procedures with greater frequency than less-involved physical therapists.

Method

This study was approved by the institutional Review Board for Human Use at The University of Alabama at Birmingham (UAB) for exemption from review.

Sample

Two hundred twelve physical therapists currently employed at facilities in Alabama that have clinical affiliations with the Division of Physical Therapy, UAB, were the population targeted in this study. Names and addresses of the clinical facilities and the number of physical therapists employed at each facility were obtained from the Academic Coordinator of Clinical Education, Division of Physical Therapy, UAB. Questionnaires were administered by mail. The physical therapy department directors were contacted initially by a letter from the ACCE, providing information about the study and seeking cooperation in completing the survey. A packet of questionnaires and a cover letter were then mailed to each director. The director was asked to distribute the questionnaires to each of the physical therapists at the facility, collect the questionnaires after completion, and return the completed questionnaires to the primary investigator (EH). Each questionnaire was accompanied by a separate cover letter to the physical therapist and an envelope to be used for confidentiality in returning the completed questionnaire. Subjects were asked to complete the questionnaire and return it to the director within a two-week period. After the initial two-week collection period, reminder letters were sent to the directors in facilities from which no response had been received. Reminder cards that could be distributed to the individual physical therapists, if necessary, were enclosed. Data were collected over a total period of five weeks.

Questionnaire

The questionnaire was developed by the primary investigator and based in part on a questionnaire used by Ritchey et al. [10] Five individuals in their second year of study in the Division of Physical Therapy, UAB, pretested the questionnaire, and their responses were used to improve the clarity of the instrument.

The questionnaire contained three parts. Part 1 consisted of sociodemographic items that addressed involvement in the APTA and continuing education, as well as characteristics of the facility where a respondent was employed (Figure). Respondents were asked to indicate their job title and to rank on a four-point scale (1 = greatest frequency; 4 = least frequency) the frequency with which they received four types of referrals. The four referral types listed were referrals that 1) stated "evaluate and treat," 2) included a prescription but allowed modification as necessary, 3) requested a specific protocol following surgery, or 4) indicated a specific prescription with no variation allowed.

Part 2 included a list of 24 procedures, or types of intervention, used in physical therapy. Respondents were instructed to indicate their frequency of use of each of the 24 procedures with answers of often," "sometimes," and "never."

In Part 3, the physical therapists were asked to rank the 24 physical therapy procedures on five dimensions of complexity. Using response categories of "high ... .. moderate," and low," the respondents were asked the following: 1) What is the level or degree of expertise and knowledge required in order to perform the procedure or activity? 2) To what extent must judgment be used in making decisions during the procedure or treatment? 3) What level of supervision would you feel that you should give an assistant or an aide if the procedure is delegated? 4) If performed incorrectly, if outcomes are misinterpreted or recorded inaccurately, what is the potential for placing a patient in jeopardy? and 5) What level of psychomotor skill is required?

Scoring and Analysis

The professional involvement scale is a summated rating scale of the questionnaire items (items 9-14) pertaining to APTA and continuing education involvement (Figure). A maximum score of 3 indicates high involvement; a minimum score of 0 indicates low involvement. The internal consistency of the scale was established by computing a Student's t-test statistic between each item and the total scale score. Each item was statistically associated with the total scale score at the .001 level of significance. For Part 2-frequency of use of the 24 physical therapy procedures-an answer of "often" was given a score of 2, "sometimes" a score of 1, and "never" a score of 0. Mean frequencies of use were then computed for each answer category for each of the 24 procedures. In Part 3, an answer of "not delegated" received a score of 4, an answer of "high" a score of 3, an answer of "moderate" a score of 2, and an answer of low" a score of 1. The ranking of each dimension by an individual respondent constituted an ordinal-scale measurement that was then scored as an interval-scale measurement. Means were then computed for each procedure and each dimension. The factor analysis for the five dimensions used correlations of these means. Chi-square contingency tests were performed on the cross-tabulations of professional involvement with job title, referral type, and years in practice.

Treating the ordinal-scale individual rankings as interval-scale rankings should not enter undue bias for two reasons. First, aggregated measures (ie, the mean rankings) of sufficiently large sample sizes are more stable from one population to another from one time to another than the individual ranks themselves. These grouped scores, reflections of shared normative evaluations of the procedures, are what is of interest, Second, the Pearson product-moment correlation statistic (r), upon which factor analysis is based, has been shown to be robust. That is, under certain circumstances, results based on ordinal-scale data are not likely to differ from results based on interval-scale data. [11,12] The aggregation of the rankings and the sufficiently large sample size meet the criteria of robustness.

Results

One hundred sixty-five questionnaires (77.8%) were returned in the data-collection period and analyzed. Physical therapists from 54 of the 65 facilities (83%) responded; there was no evidence of systematic bias in the responding facilities. No statistical significant difference in response rate was found with respect to urban-rural location of the facilities (X[.sup.2] = 2.85, df = 3, p > . 10). Physical therapists in facilities with large and small staffs responded at the same rate (X[.sup.2] = 2.50, df = 3,p > .10).

Demographic data are presented in Tables 1 and 2. Table 2 also includes the professional involvement levels for the total sample. The summated ratings on the five dimensions of complexity, a factor analysis score for each dimension, and the cluster analysis results are shown in Table 3. To ascertain the test-retest reliability of the factor and cluster analysis of the dimensions of complexity of the 24 procedures, the scale was compared with the same rating scale used by Ritchey et al. [10] The correlation of the two lists of summary complexity factor scores was .92, and the sorting of the tasks into clusters was highly similar; only three tasks fell into different clusters, and these tasks ranked near the adjacent cluster. The means of each of the five dimensions in the low-complexity cluster, with the lowest mean being 1.70, indicate that all of the procedures contain some degree of complexity. High-complexity procedures are delegated less frequently ([.sup.-] X = 3.61) than the moderate- and low-complexity procedures ([.sup.-] X = 2.77 and 1.75, respectively) (Tab. 3).

Table 4 presents the frequency of use of each procedure for the physical therapists with low involvement, those with high involvement, and the total sample. There was variation in frequency of use of procedures, but the variation was not clearly related to the ranking of a procedure as being of high, moderate, or low complexity. For example, when a difference-between-proportions test was used to compare the percentages of low- and high-involvement groups that used each of the procedures often," there was no statistically significant difference for 17 of the 24 procedures (Tab. 4). Furthermore, where one of the groups did show a higher percentage of using a procedure often," the high-involvement group was the greatest user, irrespective of the degree of complexity of a procedure (eg, relaxation and cryotherapy) (Tab. 4).

There was a significant association between professional involvement and job title X[.sup.2] = 28.29, df = 2, p < .001) (Tab. 5). A large percentage (73%) of the low-involvement group were staff physical therapists, whereas large percentages of the high-involvement group were directors and clinical coordinators or supervisors (52% and 40%, respectively). Only 8% of the high-involvement group were staff physical therapists. The comparison of professional involvement with four types of referral received is shown in Table 6.

There was a significant association between a high degree of professional involvement and "evaluate and treat" as the type of referral received most frequently X[.sup.2] = 16.25, df = 3,p <.001). Approximately two of three physical therapists in the high-involvement group (64%) received "evaluate and treat" as their most frequent type of referral, whereas only one of five physical therapists in the low-involvement group (21%) checked this category. Similarly, a larger percentage of the physical therapists in the low-involvement group (29%) checked "prescription with no variation allowed" as their most frequent type of referral, as compared with the percentage of those in the high-involvement group (14%) who checked that category.

Discussion

The results of the survey lend strong support to two of the three research hypotheses. First, there was a clear consensus in the ranking of procedures in terms of the level of complexity (Tab. 3). According to the data, the high-complexity procedures were evaluative, requiring a high degree of knowledge and judgment, and were infrequently delegated to support personnel. These procedures were judged to have a moderately high risk of misuse and to require a high degree of psychomotor skill. The moderate-complexity procedures of gait analysis and postural evaluation are evaluative also but were judged to have a lower risk of misuse. The remaining procedures in the moderate- and low-complexity clusters are treatment modalities that generally require application rather than evaluative skills. According to the responding physical therapists, procedures in the high-complexity cluster require increased responsibility for the consequences of performance. The results of this study support the results of Uili et al [4] and Ritchey et al [10] in which procedures involving evaluation or judgment were rated higher than those requiring application.

Support was found for the hypothesis that the ranking of a procedure according to its level of complexity would affect the delegation of the procedure to support personnel. The respondents indicated that high-complexity cluster procedures are delegated infrequently, moderate-complexity procedures require a high degree of supervision, and low-complexity procedures require a moderately low degree of supervision (Tab. 3). These findings support Senters's contention that with a ranking of tasks by complexity, there will be an association of tasks with specific categories of worker. [2] The results of our study, which was completed in 1987, suggest a change from Senters's study of the early 1970s. Senters reported that over one half of the physical therapists surveyed indicated no use of "auxiliaries" in patient care; when auxiliaries were used, the number of procedures delegated was small. [2] In contrast, the physical therapists responding to our study indicated a willingness to delegate a wide range of procedures to support personnel with varying degrees of supervision.

With reference to whom a task might be delegated, no distinction was made between physical therapy aide and phy,sical therapist assistant in our questionnaire. With lack of such distinction, many of the respondents expressed difficulty in responding to the dimension of "supervision required if delegated" and either made the distinction themselves between delegation to an aide or an assistant or noted that they would delegate differently to an aide or an assistant.

Support was found for Senters's expectation that, as the variety of tasks comprising physical therapy practice increases, some practitioners may devote more time to one task level than to another (Tab. 4).2 Support was not found for the hypothesis that phy,sical therapists with greater APTA and continuing education involvement would use the high-complexity procedures with greater frequency. The variation of frequency of use of procedures by physical therapists may simply indicate that increasing individual focus in the use of physical therapy procedures is the inevitable result of increasing knowledge and skills rather than the result of other behaviors.

The question arose of how well the items on the professional involvement scale measure the professionalization of the sample. Confounding effects in the measurement of professionalization may have been present. Morrow found that significant differences exist between the major role responsibilities of the more involved APTA delegates and those of the less involved nondelegates. [1] individuals involved in roles or settings that allow the exercise of professional judgment, personal responsibility, and participation in decision making or self-direction tend to respond in more professional ways than those in subordinate and structured roles or settings. We found that professional involvement and job title are clearly related, with those in the high-involvement group tending to be directors and clinical coordinators or supervisors (Tab. 5). We also found that, in general, the greater the professional involvement, the more likely a physical therapist will receive the type of referrals that permit greater autonomy (ie, "evaluate and treat") (Tab. 6).

Several factors must be considered when evaluating professional involvement. A large number of the respondents were graduates of less than three years; therefore, a 0 response for certain items for the year that they were still in school would not be unusual. In addition, many facilities participating in the survey were understaffed because of the shortage of physical therapists in the work force, a situation that might restrict or curtail participation in activities of the APTA and in continuing education programs. Both of these factors may have lowered a respondent's professional involvement score. The higher level of involvement on the part of directors might be a function of a more flexible time schedule for directors than for other staff members. Finally, more valid ways of measuring professional involvement may exist.

The sampling method used for this study involved a nonprobabilistic sample of convenience, and the population to which the findings can be generalized is difficult to establish. The data provide important information on physical therapists' attitudes, values, and behaviors. Morrow concluded that successful professionalization of physical therapy will depend ultimately on the attitudes and behaviors of individual practitioners and the degree of acceptance accorded physical therapy by the public in response to those attitudes and behaviors. [1] More research is needed to provide a complete picture of the behaviors and attitudes of physical therapy practitioners. The following recommendations are made for future research in this area:

1. Further examination of delegation
  of procedures, distinguishing
  clearly between physical therapy
  aides and physical therapist
  assistants.
2. Examination of the implication of
  frequency of use of procedures for
  physical therapy education.
3. Further examination of the factors
  affecting frequency of use.
4. Study of a random sample of physical
  therapy clinicians.

Conclusion

The results of this study indicate that there is variation in the degree of individual professionalization and in the avowed clinical behaviors of physical therapists. Physical therapists reported the use of a wide range of procedures, use of procedures that require increased expertise and evaluative skills, and use of procedures that require acceptance of increased responsibility for the consequences. A clear relationship, however, was not found between professionalization and behavior. In addition, some of the results may be interpreted as increased focus by individual physical therapists and increased autonomy in evaluation and treatment.

The relationship between professional involvement and a more autonomous role was not clear. Those physical therapists who were more professionally involved did not report more frequent use of procedures requiring higher degrees of expertise as compared with less-involved physical therapists, but they were more likely to receive referrals allowing autonomy. The results of the study indicate that substantial role expansion and differential will occur irrespective of individual involvement in the formal and informal cultures of the occupation. The implications of the study, however, are that physical therapy educators and administrators must be aware of elements that enhance professionalization and must support changes necessary to enhance role expansion and differentiation.

References

1 Morrow JR: The Nature and Bases of Physical Therapists' Attitudes Toward the Continuing Professionalization of Their Occupation. Doctoral Dissertation. Bloomington, IN, Indiana University, 1984

2 Senters JM: Professionalization in a health occupation: Physical therapy. Phys Ther 52:385-392, 1972

3 Perry J: The contribution of the physical therapist to medicine. Phys Ther 45:1033-1041, 1965

4 Uili RM, Shepard KF, Savinar E: Physician knowledge and utilization of physical therapy procedures. Phys Ther 64:1523-1530, 1984

5 Sussman MB: Sociology and Rehabilitation. Washington, DC, American Sociological Association, 1966, pp 179-203

6 Ritzer G: Professionalism and the individual. In Friedson E: The Professions and Their Prospects. Beverly Hills, CA, Sage Publications Inc, 1971, pp 59-73

7 Pavalko RM: Sociology of Occupations and Professions. Itasca, IL, F E Peacock Publishers Inc, 1971

8 Goode WJ: Encroachment, charlatanism, and emerging professions: Psychology, sociology, and medicine. American Sociological Review 25:902-914, 1960

9 Wilensky H: The professionalization of everyone. American Journal of Sociology 60:137-158, 1964

10 Ritchey FJ, Pinkston D, Goldbaum JE, et al: Perceptual correlates of physician referral to physical therapists: Implications for role expansion. Soc Sci Med 28:69-80, 1989

11 O'Brien RM: The use of Pearson's r with ordinal data. American Sociological Review 44:851-857, 1979

12 Henry F: Multivariate analysis and ordinal data. American Sociological Review 47:2991-304, 1982

(Tables and other figures omitted)