Oakes and Oakes
|
|
|
Oakes and Oakes Consulting |
Work Hardening and Work Conditioning Interventions: Do They Affect Disability? Deborah E Lochner
7he purpose of this article is to review the research on the effectiveness of work hardening and work conditioning programs Twelve studies of work hardening and work conditioning programs in the United States and abroad were reviewed One study produced convincing evidence in a randomized study that a work conditioning program was useful in producing a higher percentage of return to work and an earlier return to work in a group of patients of work for at least 2 months. Another study demonstrated that a work hardening program increased the rate of return to work by 52016 in patients of work for greater than 4 months. Most of the other studies reviewed suggested positive results, but more carefully documented, randomized, and controlled studies are needed to support the efficacy of these programs and to date?7nine the optimum and most cost-effective work hardening and work conditioning interventions [Lechner DE Work hardening and work conditioning interventions Do they affect disability Phys Tber 1994; 74..471-493.1 Key Words: Musculoskeletal system, Physical disability, Rehabilitation, Work conditioning, Work hardening.
Work hardening and work conditioning programs evolved in the late 1970s" in an effort to
minimize the economic and human costs of work related injuries. The aim of these program is to rehabilitate injured
workers,. returning them to productive work as soon as possible with minimal functional restrictions. Leonard Matheson,
PhD, was the first to describe a work hardening program, which he established in 1977 at Rancho Los Amigos Hospital
in Downey, Calif" Matheson et a19 published the first article describing work hardening in 1985. Matheson!
and colleagues described work hardening as a "work oriented treatment program that has an outcome which is
measured in terms of improvement in the clients productivity."9(p314) In 1986. the American Occupational Therapy
Association's (AOTA) Commission on Practice10 published work hardening guidelines. The AOTA commission defined
a work hardening program as one that is These guidelines also defined the occupational therapist's role in screening referrals, assessing
patients and jobs, and planning and implementing work hardening programs. In 1988, the Commission for Accreditation of Rehabilitation Facilities (CARF) appointed a multi disciplinary advisory committee to develop work hardening standards." The CARF expanded the AOTA definition by stating that the work hardening program must be "highly structured and goal oriented," interdisciplinary, and combine both work simulation and work conditioning." In order for work hardening programs to achieve CARF accreditation, they also must conform to CARF standards for administration and organization. These administrative and organizational standards "address such issues as organizational governance, management, information management, health and safety programs, physical plant, and client transportation."" Services Provided by a single practitioner are automatically excluded from obtaining CARF accreditation.8 Because the administrative and organizational standards are difficult and costly to achieve, a limited number of work hardening programs have become CARF accredited. In states in which workers' compensation boards have mandated that only CARF-accredited work hardening programs can be reimbursed, only 9% of the programs are CARF accredited.8 In 1991, responding to the effects that the CARF standards were having on work hardening programs, the Industrial Rehabilitation Advisory Committee of the American Physical Therapy Association (APTA) developed the Guidelines for Programs in Industrial Rehabilitation.8 These guidelines were intended to be an alternative for the small clinics that
did not want to incur the costs of providing the administrative and organizational structure required by CARE The
guidelines offered more flexibility and recognition that clinicians could provide more than one level of service
in their industrial medicine clinics. The APTA guidelines introduced the concept of work conditioning as a "separate
and distinct" program, which is an appropriate alternative to work hardening for patients with less complex
conditions and those with chronic conditions. Work conditioning was
defined as a program with an emphasis on physical conditioning that addresses the issues of strength, endurance.
flexibility, motor control, and cardiopulmonary function. The term work hardening was reserved for interdisciplinary programs that address the need of patients with 11 vocational and behavioral dysfunction," utilizing a graded work simulation
approach and psychosocial intervention.12 In Table 1, the AFTA's definitions of work conditioning and work hardening
programs are compared. The APTA Guidelines for Programs in Industrial Rehabilitation 12 state that individuals are eligible
for work conditioning if they fulfill the following criteria: 1 . Have a job goal. 2. Have a stated or demonstrated willingness to participate. 3. Have identified systemic neuromusculoskeletal physical and functional deficits that interfere with work. 4. Are at the point of resolution of the initial or principal injury at which participation in
the work conditioning program would not be prohibited. The guidelines do not specifically state whether all or only some of these criteria must be met.
The guidelines further state that work conditioning "generally follows acute medical care or may begin when
the client meets the eligibility criteria" and "should not begin after 365 days have elapsed following
the injury without a comprehensive interdisciplinary assessment." 12(p7O) The AFTA guidelines12 state that for individuals to be eligible for work hardening, they must
fulfill the following criteria. 1. Have a targeted job or job plan for return to work at the time of discharge. 2. Have a stated or demonstrated willingness to participate. 3. Have identified physical (systemic neuromusculoskeletal), functional, behavioral, and vocational deficits that interfere with work. 4. Be at the point of resolution of the initial or principal injury at which participation in
the work hardening program would not be prohibited. The guidelines do not specifically state whether all or only some of these criteria must be met.
The guidelines further state that work hardening "may begin only after the completion of the work hardening
assessment." 12(p7l) In 1991, Wyrick et aI13 published a descriptive study of the work hardening programs in existence
at that time. These investigators sampled 192 multi disciplinary work hardening and work adjustment programs from
all regions of the United States. Most of the programs surveyed were affiliated with hospitals or rehabilitation
centers (57%) and operated for profit (56%). Services most frequently provided were "graded work simulation
and exercise to increase employability" (91%), baseline assessment of functional limitations(91%), assessment
of the match between worker and job demands (90%), and education for injury, prevention (90%). Occupational therapists
and physical therapists were the principal deliverers of "hands-on" patient care. Therapists tended to
use job simulation and generic exercise equipment in their evaluations and treatment processes more frequently
than computerized isokinetic evaluation or work simulation equipment, static force-producing equipment, or Nautilus*
equipment. Most of the programs were initiated between 1980 and 1986. The number of patients seen weekly ranged
from 3 to 100. Most of the programs operated 7 hours per day, 5 days per week, with patients averaging approximately
5 weeks of treatment. Outcome data were being collected at 63% of the programs. and research was being conducted
at 19% of the programs. As can be seen from the few studies available for this review, much of the research that
was reported to be ongoing before 1991 has yet to be published in refereed journals. After over a decade of work hardening and work conditioning programs, questions remain whether
these interventions really make a difference in returning people to work. Many questions come to mind when critically
reviewing the literature on the effectiveness of work hardening interventions. The purpose of this article is to
review the empirical literature concerning the effectiveness of work Subject Characteristics An overview of the reviewed studies with regard to subject characteristics is presented in Table 2.14--25 For a majority of the studies, subjects were recruited consecutively from those referred during a period of time. In four studies, 15-17.24 the method of subject recruitment could not be determined because of inadequate descriptions. A variety of methods were used for assigning subjects to control or treatment groups. Mitchell and Carmen17 matched their comparison and treatment groups for age, sex, and type of injury. Lindstrom et aI14 and Aberg15 randomly assigned subjects to either a treatment group or a control group. Haig et a116 and Catchlove and Cohen23 divided subjects into groups according to whether they were treated before or after the initiation of a change in the program. Mayer and colleagues18-19 and Hazard et a120 used subjects whose entry into the program had been denied by their insurance company for their comparison groups. According to these authors, the insurance companies' denial was a matter of policy rather than a reflection of the companies' attitude toward the subjects' rehabilitation potential. The comparison group used by Sachs et a121 consisted of individuals who were referred for evaluation only, who were unwilling to participate in the program because of their time demands or because they lacked child care, or whose travel distance was too great to permit attendance. This comparison group likely included some subjects who were not motivated to return to work; therefore, any comparisons made between the treatment and comparison groups should be viewed with extreme caution. Subjects in the work hardening and work conditioning programs ranged from 2 days postinjury to "years of back pain." Eight of the programs were directed toward individuals who had been injured for at least 1 year, whereas four of the programs were begun early in the recovery process (<2 months after injury). interventions that are introduced at one stage of the rehabilitation process may be more or less effectual than those introduced in another stage. Seventy percent of those who experience back pain will recover within 1month. 26 Of those off work for greater than 6 months, however, only half will return to work.27 These statistics should be considered when comparing studies. Most of the studies involved industrial "blue-collar" workers whose mean age ranged
from the mid-30s to the mid-40s. Only half of the studies were conducted in the United States. Care must be taken,
therefore, when generalizing results from studies done in other countries with our own. Among other variables that
are different. we need to consider that workers* compensation systems differ from country to country. Work ethics,
attitudes toward work, and economic conditions also differ. In 9 of the 12 studies, patients with low back pain
were studied. Results from these studies may not be generalizable to patients with disabilities involving other
areas of the body. Program Goals In Table 3, a list of the goals of the work hardening and work conditioning programs is provided.
Many of the investigators did not explicitly state the goals of their programs. In many of these cases, however,
goals could be inferred from the outcome measures or program descriptions. In 10 of the studies, 14,16-21.23-25
return to work was a primary goal of the program. Improving the physical impairments of individuals was the goal
for 5 programs.17-21 Improving patient function was a goal for 3 programs.14.15,25 The goal was to return individuals
to work as quickly as possible for 2 programs.16.24 None of the programs had goals pertaining to decreasing the
time required for case closure for those who did not return to work. Matheson, who is generally given credit for
originating the concept of work hardening, identifies case closure as an appropriate goal for work hardening interventions.9
Given the nature of industrial rehabilitation, not all patients will return to work, despite the best rehabilitation
interventions. Thus, a realistic and meaningful goal is to determine whether there is an inability to return to
work as quickly as possible. Future research should document all the goals of the program being studied in order
to determine appropriate interventions and outcome measures. Program Content The content, type of program (ie, work hardening versus work conditioning), disciplines involved in implementing the program, length, intensity, and cost per patient are shown in Table 4. Six of the 12 programs studied seemed similar to work hardening as the term has been defined by the AYrA_8 Five of the 12 programs seemed similar to work conditioning. One study15 was not described in adequate detail to determine the type of program being administered. A recent trend is to reserve work hardening programs for the patient who has been out of work for several months and who appears to have complex psychosocial problems.8 This trend is reflected in the programs included in this review. Five of the 6 work hardening programs were directed toward patients who had been out of work for at least 3 months, whereas 3 of the 5 work conditioning programs were directed toward patients off work less than 3 months. Nearly all of the studies evaluating work conditioning programs were conducted outside the United States. The prevalence of work conditioning programs in other countries may reflect a trend toward earlier intervention in work-related injuries in these countries. This tendency to use work conditioning programs, however, could indicate an implicit attitude in these countries that the work simulation and psychosocial components of treatment are not useful. important. or affordable. Most of the programs studied in the United States were work hardening programs. Mayer et al.18 in 1985, were the first to describe the outcome of a work hardening approach, complete with work simulation and psychological intervention. At the time of that report, their approach was a novel one and served as a model for subsequent investigators such as Hazard et a.120 and a chs et al.21 Sachs et al modified the Mayer et al approach by decreasing the intensity and placing less emphasis on the psychological component of the program. A novel approach was implemented by Catchlove and Cohen23 in Canada in 1982. These investigators
explained to their patients at the outset of the program -that they would need to return to work within 1 to 2
months." and this expectation was reinforced throughout the program. Patients were informed that returning
to work was an integral part of the program rather than an outcome. When compared with a group of patients who
had not been "instructed" to return to work, this treatment strategy increased the percentage of those
who returned to work by 40%. The legal implications of using such a directive in the United States have not been
determined. Table 5 also includes a description of the reported evaluation and treatment components of the
programs studied. Authors varied greatly in the amount of description provided regarding their programs' components,
making comparisons difficult. In most of the programs, an evaluation was done on admission. The nature of this
evaluation. however, varied greatly and included various combinations of clinical measures of range of motion and
muscle force production, physical work performance, and psychological testing. Post program evaluations were reportedly
used much less frequently. Nearly every program included some type of "back school" or patient education
program. The most frequently mentioned components of patient education were instruction in body mechanics, lifting
technique, back protection, anatomy and pathology of the spine, medications, surgical techniques, and the theoretical
basis for treatment. Psychosocial intervention was included in seven of the programs and was most commonly found in
the work hardening programs. The only programs in which the psychosocial components were described in any detail
were those of Mayer et al,18 Hazard et al,20 and Sachs et al.21 In these programs, behavioral pain management,
cognitive behavioral skills training. and individual and family counseling were emphasized. Exercise was a component
in 83% of the programs and was used for work hardening and work conditioning. The types of exercise most frequently
mentioned were individualized, graduated strengthening, stretching, endurance, and cardiovascular training. Use
of exercise equipment that provides resistance was often mentioned as a means of strengthening. Most programs focused
on back extensor, abdominal, and extremity muscle strengthening. Lindstrom et al,14 in Sweden, emphasized the use of operant conditioning in the progression of
their exercise program. Patients exercised to preset quotas instead of exercising to pain tolerance. These investigators
felt that this approach -teaches the patient that it is safe to move while increasing his or her activity level."
Oland and Tveiten,22 in Norway, incorporated recreational activities such as paddling, skiing, hiking, and riding
into their exercise programs. Although the variety provided by the recreational activities is certainly desirable,
the feasibility of implementing such a program in the United States seems doubtful. Work simulation was, by definition,
exclusive to the work hardening programs. Work simulation, in most studies, appeared to be based on the patient*s
abilities and job demands, addressing lifting demands as well as position tolerance and repetitive motions. Many
of the programs incorporated the concepts of proper body mechanics and lifting techniques into the work simulation
activities. A comprehensive list of the disciplines providing the treatment was often omitted from the reports,
but physical therapy was the most frequently mentioned discipline, followed by medicine and psychology. Length
and intensity of the programs vaned, ranging from 3 weeks to 30 months with most program falling within the 4-
to 6-week range. Most programs were administered on an outpatient basis and operated 5 days per week. The cost per patient was reported for only four of the US programs. Costs ranged from $1,400
to $9,000 per patient (Tab. 4). Sachs et al,21 in New Hampshire, reported that their program cost $1,400 per patient.
They attributed their low costs to two factors: (1) They used a social worker rather than a psychiatrist or psychologist
for psychosocial counseling, and (2) their program consisted of three half-day sessions per week instead of five
full-day sessions, with no inpatient stays. The return-to-work rates for the New Hampshire program were similar
to those for the more expensive programs described by Mayer and colleagues 111-19
in Texas and Hazard et a120 in Vermont. Macer et al, however, appear to have worked
with a patient population who had been off work longer and perhaps needed a more intensive program than did the
patients seen by Sachs et al. Questions remain to be answered regarding the amount of treatment needed, the treatment components
that are most effective, and the most cost-effective method of delivering those services. Providing the details
of program components will be extremely important in reports of future research so that successful programs can
be duplicated and studied in a variety of settings and with a variety of patient populations. Study Design Developing a well-designed study for evaluating the of outcome of work hardening and work conditioning
programs is difficult but essential to understanding the effectiveness of work injury interventions. The best and
most generalizable studies should include either randomly assigned control groups or matched comparison groups.
A careful and detailed description of how the subjects were recruited and the inclusion and exclusion criteria
used ensures that the treatment group can be assessed to determine whether there was a bias. Control/comparison
groups should be evaluated at the time they enter the study to determine whether differences in severity and type
of injury, physical abilities, physical job demands, psychosocial profile, demographic characteristics, level of
pain, acuteness of injury, workers' compensation, or litigation must be considered. In addition to a description
of the evaluation process, the results of this initial evaluation should be reported, and treatment and control/
comparison groups should be compared. Optimally, no significant differences should exist between the two groups
before initiating the intervention. A specific and complete description of the intervention applied to the treatment and control/comparison
groups will allow for more meaningful comparison with other studies and will assist others who wish to replicate
the research or program at their facilities. The services offered; the disciplines of those delivering the services;
and the length, intensity, and cost of the program must be documented to identify the type of program that is most
successful and cost effective. The evaluation process, performed upon entry to the program, should be repeated
at the end of the program for both the treatment and control/comparison groups. Such a comparison will determine
whether the intervention has been effective in changing outcomes such as pain, psychosocial attributes, and physical
and functional abilities. Immediate and long-term outcome measures of the success of a work hardening or work conditioning
program obviously focus on the percentage of patients returning to work. These results, however, should be specific
as to the level of work to which these individuals are returning and should be compared with their previous work
level. Whether the patients were able to return to their previous level of work or required a lighter level should
be noted. Whether the patients required any job or workplace modifications is also an important factor to document.
The time required to closure of the case, for those not returning to work, should also be examined. The percentage
of patients who experience reinjury or require additional medical or surgical treatment during the follow-up period
provides another measure of success of a work hardening or work conditioning program. Finally, analysis of cost
versus benefit is essential if we are to know whether a program is economically feasible. All of these outcome
measures will be affected to some degree by factors outside the control or influence of the industrial rehabilitation
program (eg, job availability, client motivation to return to work, recurrence of pain, further reinjury). If a
carefully described control group or matched comparison group is used in the study design, however, differences
should be evident regardless of these factors. Table 6 presents an overview of the attributes of a well-designed study and indicates which of
the studies reviewed have these attributes. None of the studies included all of the attributes of a well-designed
study. Undstrom et aI14 described more of these attributes for both the treatment and control groups than did any
of the other authors, followed by Mayer and colleagues, 18-19 Hazard et al.20 and Sachs et al.21 All of the studies except those of Caruso et a124 and Edwards et a125
had either a control group or a comparison group. In none of the studies was subject recruitment adequately described.
Lindstrom et al, 14 Oland and Tveiten,22 and Catchlove
and Cohen,23 however, gave more information than did the other authors. Lindstrom et al and Hazard
et al provided the most comprehensive description of the results of the initial evaluation
of their subjects. Descriptions of the group interventions were brief and incomplete in all of the studies reviewed.
The program descriptions for treatment groups were particularly lacking in the reports by Aberg15 and Oland and
Tveiten. Post intervention evaluation data for the treatment group were supplied for 5 of the 12 programs. Hazard
et al were the only investigators to report post intervention evaluation data for both treatment and control groups.
Program Outcomes Table 7 provides an overview of the program outcomes of the studies reviewed. At the conclusion
of each work hardening or work conditioning program, there are four possible outcomes for the patient. The patient
can return to a former job, can obtain a new job, can seek vocational assessment or retraining for a new type of
work, or is deemed incapable of employment. The first three outcomes are viewed as the most desirable for society.
If the patient is deemed incapable of employment, the sooner this can be determined in the course of rehabilitation,
the less expense will be incurred and the fewer medical resources will be expended on unsuccessful rehabilitation.
Percentage of patients returned to work and time to case closure, therefore, can both be viewed as measures of
a program's success.9 The level of success of most of the programs reviewed, however, was judged primarily by the
percentage of patients returned to work.14,17-25 The time to case closure for those not returning to work was not
reported for any of the studies. Other outcome measures were varied. The need for additional medical or surgical care was examined
in four of the studies, 1-1. 18,19.23 and the percentage
of patients who experienced reinsure was documented in one study.19 For four studies, 14.16-1-.23 the time required
to return to work was reported. For four studies,20-22,25 changes in measures of physical impairment or pain were
reported. Workers' compensation costs were reported for two studies.17.23 Unresolved litigation was reported for
one study.',' The level of physical demand of the work to which individuals returned or any job or workplace modifications
that were necessary for them to return to work were not mentioned for any of the studies. Mitchell and Carmenl_
were the only investigators who attempted any type of cost analysis. Studies Without a Control or Comparison Group The lack of either a control or comparison group in the studies reported by Caruso et al2q and
Edwards et a125 severely limits the importance of any reported outcome results. These two studies demonstrated
return-to-work rates of 71% and 55%, respectively. Without control or comparison groups, however, there is no evidence
that the return-to-work rates were any better than they would have been without the program. Despite the lack of
controls, interesting comparisons were made in both of these studies between those patients who returned to work
and those who did not. Caruso et al found that return to work was negatively correlated to being male, performing
a job with heavy lifting demands, experiencing a long duration of symptoms, and receiving workers' compensation
benefits. Edwards et al found that those patients who returned to work had a greater decrease in pain and a greater
increase in upper-extremity strength, compared with those who did not return to work, on completion of the program.
Studies With a Comparison Group In six of the studies, 16,18-22 there were no randomized comparison groups. In the studies reported
by Mayer and colleagues "'. 19 and Hazard et al,20
the comparison groups generally consisted of individuals who had been referred to the program but who were not
admitted because their workers' compensation insurance did not cover the program. Sachs et a121 included in their
comparison group individuals who were referred for evaluation only, who were unwilling to participate in the program
due to their time demands or lack of child care, or whose travel distance was too great to permit attendance. These
admission criteria appeared to bias the comparison group toward failure in the study by Sachs et al. The results
of this study, therefore, should be interpreted with extreme caution. The main problem with most of the comparison groups in these studies
16,18-22 is that the patients' physical abilities (eg, strength, range of motion, function
), psychological profiles, and severity of injuries at the time of entry into the program often were not documented.
Without this information, determining whether the comparison group is significantly different from the treatment
group is difficult. For example, the comparison group subjects could have had a poorer rate of return to work simply
because, at the outset of the study, they had more severe injuries, were more depressed and anxious, were weaker,
or had less range of motion than the treatment group subjects. What happened to the comparison groups during the
treatment interval also was not documented clearly in these studies. if the comparison group subjects had no treatment
or fewer contacts with health professionals, their lack of improvement could be due to their having less individualized
attention. Of the six groups of investigators who used a comparison group, Hazard et aJ20 Were the only
investigators who described the physical abilities and psychological profiles of the treatment and comparison groups
at initiation of the program. This initial comparison makes their study the strongest and most important of the
studies using a comparison group and directed toward patients off work greater than 4 months. Hazard et al found
no significant differences between their two groups at the outset of the study. Their 3-week outpatient program
included the full spectrum of work hardening services. This program had the highest percentage of improvement in
rate of return to work of the treatment group when compared with the comparison group. Hazard et al also found
that upon completion of the program, the treatment group's pain and self-assessed disability decreased, whereas
the subjects' physical capacity increased. At the 1-year follow-up, subjects had maintained their improvements
in pain and self assessed disability, but had lost some of their improvements in physical capacities. The study
by Hazard et al provides the most convincing evidence that a work hardening program improves the rate of return
to work for those off work greater than 4 months. Macer and colleagues ". 19 also
reported favorable results with their work hardening program, 39% and 46% increases, respectively, in return to
work for the treatment groups. As mentioned earlier, their lack of description of the physical characteristics
of the treatment and control groups seriously limits the value of these studies. In both studies, the same program
was used, which consisted of 3 weeks of inpatient care and an average of 5 weeks of outpatient care. These two
programs were the most expensive of those programs for which costs were described, possibly due to the inpatient
component. In both of these studies. the comparison group had significantly more health care visits than did the
treatment group. in the 1987 publication. Mayer et all9 reported that the comparison group had a significantly
higher incidence of additional surgery. These findings may be effects of the program or simply evidence that the
subjects in the comparison group had more severe back problems. The 1985 publication',' reported that the comparison
group had a greater percentage of unresolved litigation cases than did the treatment group. Again. these results
may be a cause or effect of the subjects' poor performance. Mayer et all" also correlated an increase in trunk
extensor strength and trunk range of motion with an increased rate of return to work. Sachs et a121 reported a -75% rate of return to work at the time of discharge from the program
in the treatment group. but they failed to report a return to work rate for the comparison group. At the 6-month
follow-up, 59% of the treatment group and 33% of the comparison group \,.-ere working. At the 1-year follow-up,
63% of the treatment group and 36% of the comparison group were working. One problem with this study is that the
authors were able to contact 71% of the treatment group compared with 45% of the comparison group at 6 months and
compared with only 36% of the comparison group at 12 months. This low rate of follow-up in the comparison group
may have affected the results. No significant differences were found between the treatment and comparison groups
at 12 months due to this poor rate of follow-up, leaving the 6-month follow-up for the only valid comparison between
the treatment and control groups. Sachs et al2l reported that they modified the programs reported by Mayer and colleagues18.19
and Hazard et a120 such that their program is less costly, less time consuming, and involves less psychological
intervention. Is their program as effective as those reported by Mayer and colleagues and Hazard et al? To answer
this question, we must examine the types of individuals recruited into each program. Mayer and colleagues'81" studied individuals who had failed other conservative programs
or previous surgery, whereas Sachs et al21 studied patients who had not undergone previous treatment. Sachs and
coworkers' program. therefore. may have been more likely to include those patients for whom surgery, " - was
needed. When those patients who required additional surgery after discharge were removed from the treatment group
in the study by Sachs et al. the rate of return to work was 73%. When this figure is compared with the comparison
group's rate of return to work at 6 months. it represents a 31% increase in the rate of return to work in the treatment
group. Mayer and colleagues reported 39% increase in the return to work rate of their treatment group in their
1985 publication'8 and a 46% increase in their 19877 publication."' Hazard et a120 reported a 52% increase
in the return to work rate of their treatment group. Based on these figures. it seems that Sachs et al produced
return-to-work rates comparable to those produced by Hazard et al and Mayer and colleagues. The subjects studied
by Mayer and colleagues, however. had been off work for an average of I year longer than those studied by Sachs
et al, making them much less likely to return to work and therefore more difficult to rehabilitate. The more streamlined program described by Sachs et al 2l may not achieve the same results in
patients off work for more than 1 year. Additional findings reported by Sachs et al were that 75% of their treatment
group was discharged with a higher functional capacity, 78% had increased their "weight capacity," and
84% had increased their spinal range of motion. The program described by Sachs et al was the least costly of those
programs reviewed. The low cost was attributed to an outpatient program that was less time-intensive than the programs
described by Mayer and colleagues "l. 19 and Hazard
et a120 and to the use of a social worker, rather than a psychologist, for psychosocial intervention. Mitchell and Carmen'- conducted a multi clinic study of 703 subjects from 12 clinics in Ontario,
Canada. They compared consecutively recruited subjects treated with a work conditioning program with individuals
who were treated in the community. Their comparison group was matched for age, sex. body part injured, and type
of injury. Treatment for the comparison group was not controlled and included a variety of interventions selected
by each subject's physician. Treatment programs in the various clinics were similar in their principles of approach
(Tab. 5) but varied in intensity'. components, and personnel. These investigators reported return to-work rates
of 85% in the treatment group and 75% in the comparison group. The rate of return to work in their comparison group
was much higher than that reported for other comparison groups.111-21 The relatively high success rate of their
comparison group in the study by Mitchell and Carmen could be related to several factors. The patients they studied
had been off work for an average of only 41 days. whereas the patients in the other studies18--21 had been injured
for an average of 11 to 25 months. If patients receive intervention sooner after the onset of their injury'. they may return to work at a higher rate, regardless of the
intervention. Be cause there was no control of the intervention received by the comparison group. the subjects
could have received intervention similar to that of the treatment group, thus making them a poor comparison group.
In addition to the 10% increase in rate of return to work, the treatment group returned to work sooner and realized
more than million savings in total compensation costs. Unfortunately. the importance and generalize ability, of
this study is limited severely by the lack of intervention controls for both the treatment and comparison groups.
If neither treatment nor comparison group interventions were controlled, the likelihood of seeing a treatment effect
could either increase or decrease. Oland and Tveiten22 applied a work conditioning program for 4 weeks to a group of patients who
had been off work for an average of 13 months. The divided their subjects into a group who received their program
and a group who received their program combined with "pool traction." Pool traction was defined as "suspending
the patient in a floating rig, in water heated to 34'C wearing a vest with straps. Plastic-coated lead weights
attached to a hip belt gave a total traction of 4 to 8 kg for 3 to 7 minutes." The found no significant difference
in return-ro-work rates between the two groups. In both groups. approximately one third of the subjects returned
to work. Because there was no control group, who did not receive the work conditioning program, the importance
of this finding is extremely limited. The return-to-work rates reported by Oland and Tveiten were lower than those
of most of the other programs studied. The lower rate of return to work could be due to the program design. Oland
and Tveiten's program appeared to be similar to a work conditioning format, thus lacking the elements of in-depth
psychological intervention and work simulation. Their exercise program also seemed to rely, more heavily on recreational
activities than on individualized exercise programs. Perhaps the addition of psychological intervention, work simulation.
and specific exercise routines are needed to increase the rate of return to work for patients who have been off
work for an average of 13 months. Socioeconomic factors unique to Norway may also explain the low rate of return
to work. Catchlove and Cohen16 studied the effect of instructing a group of patients with chronic pain
who were receiving workers' compensation "to return to work" as an expected part of their program. Both
treatment and control groups received a work conditioning program, but the treatment group was also asked to set
time based return-to-work goals upon entry into the program. Return-to work discussions were held on a regular
basis throughout the program, and patients were told that the), would have to return to work within I to 2 months
after initiating the program. The treatment and comparison groups had rates of return to work of 65% and 25%, respectively.
This study suggests the importance of the program's philosophical approach and the clinician's expectations regarding
return-to-work outcome in patients with chronic pain. One limitation of this study, however, was that the comparison
group had a mean duration of symptoms for approximately 18 months longer than the treatment group. The comparison
group, therefore, may have been biased toward a poorer outcome. Another limitation was that, because the physical
and psychological characteristics were not reported for either group, the comparison group may also have had more
severe physical and psychosocial restrictions. Haig et a116 did not report on either a traditional work conditioning or work hardening program.
instead, they described intervention by a single physiatrics who took an active role in case management of hospital
workers 2 days postinjury. Their comparison group included patients who had been treated by community physicians
before the initiation of the program. Again, because documentation of the subjects' initial physical status is
lacking, the importance of the study is limited. The fact, however, that subjects, whose care was aggressively
managed by the physiatrics, were off work an average of 4 days less than those seen by physicians in the community
is of interest and warrants further investigation. Controlled Studies In only 2 of the 12 studies discussed were subjects randomly assigned to either a treatment or
a control group. In Sweden, Aberg15 applied a 6-week inpatient program to patients with 1.
years" of back pain. The program was described as an "educational course coupled
with a complete physiotherapeutic service" and appeared to focus on ergonomic training and body mechanics.
The activity of the control group subjects during the course of the study was not described. Aberg found that only
15% of the treatment group and 9% of the control group had "changed for the better" in one or more vocational
factors, including return to work, being "sick listed", level of income, and self-rated working capacity.
Aberg also reported no differences between the groups in the patients' perception of pain. The only significant
difference between the two groups was an increased level of income in the treatment group. Because the program
was not described, speculating as to the cause for this low rate of improvement is difficult. For these patients
with very chronic back pain, a more comprehensive interdisciplinary program including psychosocial intervention,
work simulation, and work conditioning may have improved the outcome. In one of the most important studies, Lindstrom et all-, randomly assigned 103 patients who were
8 weeks postinjury to either a control or treatment group. The physical characteristics of the treatment and control
group patients at the initiation of the program were not different. Improvements noted in the treatment group,
therefore, are most likely to be due to the effects of the intervention. Work conditioning was administered 3 days
per week until the patients returned to work. Of the patients in the treatment group, 55% had 5 or fewer appointments
with the physical therapist and 90% had 25 or fewer appointments. The intervention included functional capacity
evaluation, education using Swedish back school principles, graded exercise using operant conditioning to progress
patients, and a work-site visit. The control group subjects received "traditional care recommended by their
physician." Traditional care included, but was not restricted to, bed rest, analgesics, and "available
physical therapy." Within 6 weeks after the randomization process, 59% of the treatment group and 40% of the
control group had returned to work. Within 12 weeks after the randomization process. 80% of the treatment group
and 58% of the control group had returned to work. The treatment group required an average of 10 weeks to return
to work, whereas the control group needed 15 weeks. Twenn.-one percent of the treatment group had a recurrence
of low back pain during the year following the intervention as compared with 42% of the control group. Overall, Lindstrom et all- conducted a well-designed and well-executed study. Nelson, in his
commentary following the article by Lindstrom et al, suggested that the control group could be described more specifically
and that the treatment group may have been "destined for success" if they received more individualized
attention than the control group. He also noted that the physician who made the return-to-work decision for both
groups was not blinded as to the status of the subject. The physician could have been biased in his decision making.
Nelson was also concerned that there may have been some uncontrolled interactions between the control and experimental
groups magnifying the effect because all subjects worked for the same company. I believe that additional objective
outcome measures such as those listed in Table 6 would enhance the value of this and any future intervention studies.
Discussion After examining the available literature, is there evidence to suggest that work hardening and
work conditioning programs really work? Two of the most recent and well-designed studiesl,1.20 suggest that these
programs are indeed effective in returning a greater percentage of individuals to the workplace and do so in a
more efficient manner than alternative strategies. Design problems with the remaining 10 studies limit the strength
of their conclusions but identify some trends that merit further
investigation. Return to work was the most common1v evaluated outcome in the studies re viewed. Studies in the
United States involving patients with more chronic conditions indicate that the treatment groups demonstrated a
21% to 39% improvement in the rate of return to work over the comparison groups. Of these studies, that of Hazard
et a120 provides the most convincing evidence that work hardening programs are successful in a group of patients
with more chronic conditions, with a 52% increase in the rate of return to work in the treatment group. If this
improved rate of return to work is substantiated by future controlled studies. substantial savings in disabling,
and workers' compensation costs in the United States could be realized. even for those patients with more chronic
conditions. Studies involving patients with more chronic conditions conducted outside the United States indicated
mixed results. In Canada. Catchlove and Cohen23 reported findings similar to those noted in the United States.
In Norway. Oland and Tveiren22 reported much lower rates of return to work than did the reports of studies carried
out in the United States and Canada. These lower rates in the Nimwegen study may be related to the type of program
used, the severity of the patients' problems, or the ~workers' compensation system. The program used by Oland and Tveiten focused on patient education and exercise, without any
psychosocial or ~work simulation interventions. and might not have met the complex needs of patients who had been
off ~work longer than I year. Lindstrom et al, in Sweden, and Mitchell and Carmen,1- in Canada, examined return to work in
patients with less chronic conditions. The rates of return to work reported by these authors indicated that programs
directed toward patients with less chronic conditions are also successful in improving the rate of return to work.
The return-to-work rates reported in these two studies, however, were not dramatically higher than those reported
for programs in the United States treating patients with more chronic conditions. On the surface, one might be tempted
to reject the idea that earlier intervention results in a higher rate of return to work. Due to the lack of similarity
in study design and the differing socioeconomic environments, however, no meaningful comparisons can be made. Further
research is needed comparing intervention outcomes between groups of patients at different stages in the recovery
process. An implied goal of work conditioning and work hardening programs is to return individuals to
work sooner. Undstrom et al,14 Haig et al,16 and Mitchell and Carrnen17 examined the time required to return to
work. In all three studies, an earlier return to work in patients treated with either managed care or a work conditioning
program was observed if the treatment was begun at least 2 months after injury. Additional research must be conducted,
however, to confirm this trend. The time required to return to work in patients off work for more than 2 months
has not been examined. In only one study, by Mitchell and Carmen, 17 was a cost analysis of a work conditioning program for patients with acute injuries conducted. They reported
that their program resulted in an increase in medical costs of approximately $400 per subject. This cost was offset,
however, by a saving in workers' compensation expenses of approximately $2,000 per subject. resulting in a saving
of $1,600 per subject. Only two studies14.19 examined reinjury or recurrence of low back pain. Undstrom, et aP4 reported
that 48% of their treatment group and 79% of their control group experienced a recurrence of low back pain within
the second year of follow-up. An operational definition of "recurrence of low back pain" was not provided,
We do not know, therefore, whether these recurrences were severe enough to lead to time off work or whether they
were minor discomforts. Depending on their severity, these high rates of recurrence of low back pain are disturbing
and may point to the need for more ergonomic training or job-site modification. Mayer et aI19 reported reinjury
rates of 6% and 12% for their treatment and comparison groups, respectively, during a 2-year follow-up period.
This difference, however, was not statistically significant. Based on these two reports, the ability to prevent
further injury or recurrence of pain. by administering a work hardening or conditioning program, is undetermined
and may vary greatly between countries and patients. Literature examining the natural history of low back pain indicates that approximately 50% to
60% of persons with low back pain will have a recurrence within the first year after the initial episode of low
back pain.291-31 Although a certain percentage of this recurrence may be unavoidable, future studies need to design
interventions aimed at decreasing the rate of reinjury and to determine their ability to accomplish this goal.
Recurrence must be carefully defined in future research. For some individuals, back pain resolves completely and
then returns at intervals. The pain never resolves for other individuals, and they continue to experience periodic
increases in pain.32 Both of these patterns could be considered a type of recurrence, but one may be more preventable
than the other. Two studies, by Mayer and colleagues,18.19 examined the percentage of subjects requiring additional
health care visits or surgery. These researchers found that their comparison group had at least twice the rate
of additional health care visits as did the treatment group. In their 1987 publication,19 they reported that their
comparison group also required more than twice the number of additional surgeries. Because no data were reported
on the severity of the injuries at the beginning of the program, these results may reflect group differences.
Only Mayer et aI18 reported the percentage of their subjects who were involved in unresolved
litigation cases. They reported that 14% of their treatment group and 32% of their comparison group continued to
have unresolved litigation cases at 1 year follow-up. Does this mean that their comparison group was biased toward
failure? Comparisons regarding litigation are important to ensure an unbiased comparison. In four studies, 18.20.21.25 changes in the physical characteristics of patients with chronic
conditions after participating in a work hardening program were reported. Mayer et aI18 noted an increase in back
extensor strength and trunk range of motion and showed a positive correlation to an increase in rate of return
to work. Hazard et a120 noted that "physical capacities" in the treatment group had increased by the
termination of the program. Sachs et al2l noted an increase in functional capacity and spinal range of motion at
the conclusion of their program. Edwards et a125 noted that those subjects who returned to work had a greater decrease
in pain and a greater increase in upper-extremity strength than those who did not return to work. These reports
of physical changes in strength, range of motion, functional capacity, and pain indicate a trend that could justify
interventions that address these variables in industrial rehabilitation. More detailed and carefully controlled
studies are needed, however, to confirm this trend. A more thorough examination of the effects of work hardening
and work conditioning programs on a worker's physical impairments and their correlation to the functional outcomes
will be an important aspect of future research. Outcome measures used in future studies should have documented
reliability and validity. Traditionally, work conditioning and work hardening have not been initiated until the patient
was felt to have reached maximum medical improvement. The APTA guidelines 12 for work conditioning and work hardening
state that these services should occur "at the point of resolution of the initial or principal injury."
There is an increasing trend toward early intervention with industrial injuries, because time off work is inversely
correlated to the rate of return to work.27 The longer an individual is off work, the less likely that individual
is to return to work.27 In some programs, conditioning exercises and work simulation are encouraged at low intensity
levels during the medical treatment phase of recovery. The rationale behind early intervention is that individuals
are less likely to assume a "sick role" and more likely to return to work if they are kept as active
as possible throughout their rehabilitation. While incorporating this early intervention approach, we must be careful
to allow sufficient time for healing of injured structures. Furthermore, if treatment is begun at an early stage,
when should the functional capacity evaluation (FCE) be administered? If administered before the initiation of
this early treatment, the FCE may be too physically demanding for the patient with a recent injury. If treatment
is initiated before the FCE is performed, how will the treating practitioner know what to treat? Do we need to
develop a multilevel approach with evaluations and treatments specifically designed for patients with more acute
conditions? Further research is needed to determine the optimal point in the rehabilitation process for work conditioning
and work hardening interventions. In summary, a great deal of variability was found in the methodologies of the studies reviewed and in the extent to which the authors described their subjects, programs, and outcomes. This variability created difficulty when attempting to identify types of programs that are most likely to result in the highest percentage and most efficient return to work of injured workers. Future studies should 1. Have both a treatment group and a control or comparison group that are compared at the beginning as well as the end of the study. 2. Describe the recruitment process and group assignment process in detail. 3. Describe the patients in detail, classifying them according to type and severity of symptoms and time since injury. 4. Report details of the program, including evaluation and treatment procedures; disciplines involved; and duration, frequency, and cost of treatment. 5. Report details of any treatments or interventions applied to the comparison/control group. 6. Report outcome measures in objective, well-defined terms (eg, rate of return to work, physical
demand level of the work to which the patients returned, necessary job modifications, time requried to return to
work or to close the case, cost-benefit analysis. rate of reinjury, need for further treatment, impact of unresolved
litigation and disability payments, changes in physical impairments and their correlation to functional outcomes).
Conclusions There is evidence to suggest that work hardening and work conditioning programs accomplish their
stated and implied goals. Lindstrom et all', produced evidence in a randomized study that work conditioning programs
produced a higher percentage of return to work and earlier return to work in a group of patients who had been off
work for an average of 2 months. Hazard et a120 demonstrated similar results using a work hardening program in
a group of patients who had been off work for greater than 4 months. Other studies suggested positive results,
but more carefully documented, randomized, and controlled studies are needed to confirm their findings. Further
development of a body of well-designed studies documenting the effectiveness of work hardening and work conditioning
programs and determining the best and most cost-effective methods of treatment is needed. References 1 Peters P. Successful return to work following a musculoskeletal injury. American Association of Occupational Health Nurses journal. 1990:38:264-270. 2 Lepping V. Work hardening: a valuable resource for the occupational health nurse. American Association of Occupational Health Nurses journal. 1990:38:313-317. 3 Yeater D. The occupational health nurse as a disability manager: a vital health care management strategy. American Association of Occupational Heallb Nurses journal. 198735: 116-118. 4 Snook SH. The cost of back pain in industry. Occup. Med. 1988:3(l):1-5. 5 Basmajian JV. Therapeutic F-Exercise. Baltimore. Md: Williams & Williams; 1980:178-187. 6 Spengler DM. Bigos SJ. Martin NA. et al. Back injuries in industry: a retrospective study-overview and cost analysis. Spine. 1986: 11:241-245. 7 Quebec Task Force on Spinal Disorders. Scientific Approach to the Assessment and Management of Activity-Related Spinal Disorders. Spine. 1987:12(supp. 7):SI-S59. 8 Helm-Williams P. Industrial rehabilitation: developing guidelines. PT~Magazine of Physical Therapy. 1993:1(3):65-68. 9 Matheson LN. Ogden I.D. Violette K. Schultz K. Work hardening: occupational therapy in industrial rehabilitation. Am J Occup Ther. 1985:39:314-321 10 Commission on Practice. American Occupational Therapy Association: Work hardening guidelines. Ariz J Occup Ther. 1986:40:841--843. 11 National Advison- Committee Recommendations for Work Hardening Program. Tucson. Ariz: Commission on Accreditation of Rehabilitation Facilities: 1988. 12 APTA guidelines for programs in industrial rehabilitation. PT-Magazine of Physical Therapy. 1993;1(3):69-72. 13 Wyrick JM. Niemeyer LO. Ellexson 114, et al. Occupational therapy work-hardening programs: a demographic study. Am J Occup Ther. 1991;45:109-112. 14 Lindstrom L Ohlund C. Eek C. et al. The effect of graded activity on patients with subacute low back pain: a randomized prospective clinical study with an operant-conditioning behavioral approach. Phys Ther. 1992:72:279293 15 Aberg J. Evaluation of an advanced back pain rehabilitation program. Spine. 1984;9:317318. 16 Haig AJ, Union P. McIntosh M. et al. Aggressive early medical management by a specialist in physical medicine and rehabilitation: effect on lost time due to injuries in hospital employees. J Occup. Med. 1990:32:241-244. 17 Mitchell RI, Carmen GM. Results of a multi center trial using an intensive Active exercise program for the treatment of acute soft tissue and back injuries. Spine. 1990;15:514-521. 18 Mayer TG, Gatchel RJ, Kishino N, et al. Objective assessment of spine function following industrial injury: a prospective study with comparison group and one-year follow-up. Spine. 1985;10:483-493. 19 Mayer TG, Gratchel RJ, Mayer H. et al. A prospective two-year study of functional restoration in industrial low back injury. JAMA. 1987;258:1763-1767. 20 Hazard RG. Fenwick JW, Kalisch SM, et al. Functional restoration with behavioral support: a one-year Prospective study of patients with chronic low-back pain. Spine. 1989;14:157-161. 21 Sachs BL, David ' IF, Olimpio D, et al. Spinal rehabilitation by work tolerance based on objective physical capacity assessment of dysfunction: a prospective study with control subjects and twelve-month review. Spine. 1990;15:1325 1332. 22 Oland G. Tveiten G. A trial of modern rehabilitation for chronic low-back pain and disability: vocational outcome and effect of pain modulation. Spine. 1991;16:457-459. 23 Catchlove R. Cohen K. Effects of a directive return to work approach in the treatment of workman's compensation patients with chronic pain. Pain. 1982;14:181-191. 24 Caruso LA, Chan DE, Chan A. The management of work-related back pain. Am J Occup Ther. 1987;41:112-117. 25 Edwards BC, Zusman M, Hardcastle P, et al. A physical approach to the rehabilitation of patients disabled by chronic low back pain. ,Med J Aust. 1992;156:167-172. 26 Horal J. The clinical appearance of low back disorders in the city of Gothenburg, Sweden: comparisons of incapacitated problems with matched controls. Acta Orthop Scand Suppl. 1969:118:15-37 27 McGill CM. Industrial back problems: a control program. J Occup Med. 1969; 10: 174178. 28 Turner JA, Ersek M, Herron L, etal. Patient outcomes after lumbar spinal fusions.JAMA. 1992;268:907-911. 29 Bergquist-Ullman M, Larsson U. Acute low back pain in industry: a controlled prospective study with special reference to therapy and confounding factors. Acta Orthop Scand Suppl. 1977; 170:9-11. 30 Troup JD. Martin JW, Lloyd DC. Back pain industry: a prospective survey. Spine. 1981A 61-69. 31 Dillane JB, Fry J, Kalton G. Acute back syndrome: a study from general practice. Br3led J 1966:9:82-84. 32 Frank A. Low back pain. Br Med. 1993~ 306:901-909.
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Send mail to
barry@oakes.org with
questions or comments about this web site.
|
