Effect of Total End Range Time on Improving Passive Range of Motion

Once informed consent was obtained, 20 digits from 15 volunteer subjects were studied from June 1991, through April 1993. There were 11 volar plate injuries that had been managed conservatively, 1 volar plate injury that had been treated surgically, 5 flexor tendon lacerations that had been repaired (4 zone V, 1 zone II), and 3 distal radius fractures with controlled RSD. There were 11 little fingers, 5 middle fingers, 3 ring fingers, and 1 index finger. There were nine men and six women. The ages of the subjects ranged from 18 to 84 years (mean = 38 years).

Once a candidate for PIP serial casting had been identified and had met the inclusionary criteria, he or she was randomly assigned to one of two groups, A or B. Group A patients wore the initial casts for 6 days and the subsequent casts for 3 days. Group B patients wore the initial casts for 3 days and the subsequent casts for 6 days. All patients (groups A and B) wore the preliminary resting casts for 2 days prior to beginning the actual testing period.

An initial resting cast was applied with the PIP extended as much as was comfortable according to the patient’s subjective response. The cast allowed the distal interphalangeal joint (DIP) to be free. Each patient returned 2 days later to have the resting case removed. The dry cast was cut open with scissors using a Z line running distal to proximal on the dorsal surface (Fig. 1). Once the preliminary cast had been removed, a 5-oz weight was hung on the patient’s finger for 15 minutes to achieve maximum PIP extension. At this point, the joint was considered to have been preconditioned. During this preconditioning procedure, the patient’s hand was rested, palm up, on the edge of a table with the metacarpals and proximal phalanges as flat on the table as possible. The weight was looped over the DIP crease with a nylon cord which was padded with a single layer of moleskin at the contact area (Fig. 2). Immediately after the 15-minute preconditioning period, a torque passive range of motion text (TPROM) was performed. A passive extension force of 500 g was applied for 20 seconds through a nylon cord attached to a Haldex AB tensiometer (Fred Sammons, Inc., Brookville, IL). The cord was padded with moleskin and looped under the DIP crease. The line of the cord was run perpendicular to the shaft of the middle phalanx. A flat, metal digital goniometer was placed on the dorsum of the digit, the arms of the goniometer resting on the proximal and the middle phalanges and the axis of rotation of the goniometer aligned with the axis of rotation of the PIP (Fig. 3). The tester read aloud the goniometer baseline extension reading to the nearest degree and a member of the recording staff recorded it.

At this time, the first treatment cast was applied. As soon as the plaster had been applied, an 800-g extension force was applied via a nylon line at the DIP crease in the same manner as described to obtain the TPROM reading. The force was applied for 60 seconds with the therapist supporting the patient’s hand by holding the proximal phalanx throughout the 60-second period.

The patient was instructed not to remove the cast unless there was marked pain, or unless the tip turned white or purple or became numb. No patient throughout the study removed his or her own cast. The patient was instructed to actively flex the SIP to its full available end range (50 repetitions [reps], for times a day [QID]) to keep the DIP supple and the flexor digitorum profundus (FDP) gliding. The patient was to fully flex and extend the metacarpophalangeal joint (MP) (50 reps, QID) to promote tendon gliding and MP motion. All “uninvolved digits” were to be fully flexed and extended (50 reps, QID). Activities of daily living (ADL) with the involved hand were discouraged by the therapist.

Another staff member who had access to the grouping assignments (A or B) gave the patient a return to clinic appointment for 6 days later (group A patients) or for 3 days later (group B patients).

The cast was cut off when the patient returned to the clinic. A TPROM reading was obtained using the 500-g pull for 20 seconds technique. A member of the recording staff recorded the reading. Immediately thereafter, another case was applied in the manner previously described, along with an 800-g pull after the plaster was applied. At this point, the patients in the group A were given an appointment for 3 days later, while those in group B were given an appointment for 6 days later.

The cast was removed and the final TPROM reading was obtained when the patient return to the clinic. At this point, each patient had worn three cast: an initial resting case for 2 days followed by two treatment cases (group A patients wore the first treatment case for 6 days and the second treatment case for 3 days, while group B patients wore the first treatment cast for 3 days and the second treatment case for 6 days). Each patient had also been measured three times. The first time was just after the resting case had been removed and the PUP had been preconditioned for 15 minutes. The second time was after the first treatment cast had been removed. The final time was after the second treatment cast had been removed.

Once the testing periods had been completed and the data had been recorded, each patient was evaluated on a case-by-case basis as to the need for additional therapy.

The casting technique was standardized throughout the course of the study. The cast was constructed of three strips of fast-setting plaster, each 1 × 6-8 inches, depending on the length of the involved digit. The first and third strips were applied proximal to distal, beginning at the web space and ending at the DIP crease. The second strip was applied distal to proximal from the SIP crease to the web space. Between applications of the strips, the case was massaged to make a contoured fit and to milk out any excess water. When the last strip had been applied, the distal, volar edge of the cast was smoothed off to allow comfortable flexion of the SIP. The final step was the 800 g of extension force for 60 seconds, as previously described. Casts were applied and goniometric tests were performed by a single tester (P.C.L.).

Because we were anticipating relatively small increments of change in ROM and because the entire hypothesis rested upon these outcomes, we felt it necessary to test the accuracy of our specific system of goniometry. Therefore, a preliminary reliability study was designed to test our TPROM technique. Breger-Lee et al. have reported the reliability of a similar technique. A “fused joint model” was chosen. Seven volunteer subjects with a total of 20 fused joints (5 index, 5 long, 5 ring, and 5 little fingers) were tested. Sixteen of the fusions had been accomplished by miniature Herbert’s screw fixation. The remaining four had spontaneously fused as a result of their primary disorder (e.g., rheumatoid arthritis).

Because the joints had been fused, no preconditioning procedure was performed on the joints. Using the previously described TPROM technique with 500 g applied at the SIP crease for 20 seconds, an initial reading was taken and recorded. Throughout the entire study, all measures were read to the nearest degree; no fractional estimate was determined. A second measure was obtained with the same technique 2 to 8 weeks after the initial reading. All measures were obtained by one treating therapist (P.C.L.). All records were kept by another staff member. The results of the reliability study are shown in Figure 7.

In designing the study we were concerned for two reasons about the accuracy of our technique for measuring outcome, that is, our goniometry. One reason was that we were anticipating range changes on the order of less than 10°. The other reason was the much-quoted claim that goniometry is only accurate to ±5°. This 5° figure appears to have originated from a 1978 article by Boone et al. However, in a study that looked at the larger joints of the upper extremity, not the small joints of the hand. In addition, we incorporated the torque principle advocated by Brand for which limited reliability data are available to increase our accuracy. In our reliability study we were measuring changes on the order of 0° to 3°. The intraclass correlation coefficient of r = 0.98 gave us the needed level of confidence to proceed with the TERT investigation (Fig. 7).

Before interpreting the findings, we want to caution against a potential pitfall. The reader may be tempted to focus on the concept of serial casting for the management of PIP flexion contractures. This would be an error. In fact, this article is not about casting at all, nor is it about a specific method of treating PIP flexion contractures. Rather, the central theme of this study is the relationship between the amount of time a stiff joint is positioned at its available end rand and the consequent improvement in PROM.

Of course, in the Methods and Materials section there was a detailed description of the casting procedure, which was modified from earlier descriptions by Brand and Bell and Kolumban. However, the reader should not focus on the casting method, except to understand the research design.

The casting of the PIPs was chosen only because it was a convenient model with which to study the central question of the TERT–PROM relationship. Our intention is to draw a broad conclusion about this relationship that goes far beyond any single end range technique (e.g., PIP flexion contractures).

In a similar vein, the reader needs to recognize that the casting procedure, as described, was developed as a research design to satisfy certain scientific considerations relating to the eventual data analysis. The description of the method is not intended to be a recommendation of a clinical technique. For example, the 800-g force applied as the plaster dried is not a recommended clinical procedure. Many authorities have suggested that forces of such magnitude on the small joints of the hand may be injurious. Therefore, we are not recommending it as a clinical approach. However, we observed that no subject in this study demonstrated evidence of inflammation or of an adverse reaction to any of the procedures throughout the study. In fact, typically the subjective position followed by enforced end range positioning through an unyielding device such as a cast or a splint. The concept of “cranking on a joint and holding it there” is mentioned only to condemn it. Our research design described a gentle preconditioning session prior to casting.

Casting was selected as the model for end range positioning because it had two distinct advantages over the other forms of splinting. The primary advantage was that it ensured continuous end range positioning compared with a splint that may have been removed by the subject unbeknownst to us. The issue of compliance was critical in obtaining accurate TERT calculations. No subject removed his or her cast at any time throughout the study. The secondary advantage to the cast was that it did not permit any direct exercise of the joint, which would have introduced an additional variable to contend with when interpreting the results.

There are many clinical devices available, other than casts, that are capable of delivering large TERT values. In interpreting the data and the results, we suspect that the conclusions apply to any such devices/splints. Numerous additional factors will influence the selection of the most appropriate end range tool. This decision must be made on a case-by-case basis. Rigid protocols are fraught with difficulties and, again, we mention them only to condemn them.

The concept of a measurable dosage of stress for treating joint stiffness was addressed in this study, not the effectiveness of any specific stress delivery system, such as the cast model used. The data clearly showed that the 6 days of TERT positioning produced a significantly greater increase in PROM than did 3 days of TERT positioning. This was supported by the results of the t test (p < 0.005). But for our hypothesis to be fully satisfied, we needed to show not only a significant difference for 6 days versus 3 days, but also an amount of improvement for the 6-day period that was twice as much as that for the 3-day period. If we could demonstrate this 2:1 relationship between TERT and ROM gains, we could conclude that the improvement in joint range was directly proportional to the time the joint was held at its end range. For all subjects we found a relationship between the 6-day and the 3-day periods that closely approximated the hypothetical ideal for 2.0 or 2:1. Our ratios were 1.76 using the method shown in Figure 2, Figure 5.446 using the alternative method described in the Results section. Though these two methods seemed to produce different results, the difference was only a matter of which mathematical method of analysis was selected. Both gave results close to 2.0.

Our results serve as data, gathered in a controlled study, to substantiate Brand’s suggestion that if a joint is held at its moderately lengthened position (i.e., end range) for a significant time (i.e., high TERT values), it will grow (longer). Such growth of the dense connective tissues about the joint allows for increased motion.

This finding is consistent with a model of would remodeling that involves continuous dynamic shifting of the elements within the connective tissue to produce the final length and strength of tissue. How the total stress message is monitored and acted upon by the fibroblast to arrange the fibers, cross-links, and glycosaminoglycans is not fully understood. But, regardless of the biochemical and/or electrical mechanisms involved, it appears that total time of stress delivery greatly affects the process.

However, before concluding that ROM increases were proportional to TERT, we needed to look at the possible effect of order on the outcome. We found that whether the 6-day cast of the 3-day cast was used first did not make a significant difference; group A (1.6) and group B (2.0) had similar outcomes (Fig. 6). This absence of effect of order was borne out with the t test.

But is TERT the only factor that influences stress? What about the question of force? Is there a relationship between force and ROM changes? One could argue that if total stress were the active agent in promoting tissue change, then force would likewise alter the stress message. And, indeed, it may. However, again, there is no substantiation in the literature. Furthermore, there is a very narrow range of clinically safe amounts of force, compared with the very broad rand of clinical TERTs. For example, on the small joints of the hand, safe forces may range from 0 to perhaps 800 g, a very narrow range. But TERTs can range from a few minutes per day to 24 hours per day, a very broad range. Therefore, increases in TERT can multiply the total stress message manyfold, whereas increases in force may not produce dramatically more total stress. Nevertheless, a similar study on force variance may provide enlightening. If a relationship were established between force and ROM increase, a stress prescription for a stiff joint would perhaps read something line “splint for 2 hours at 500 g, QID,” where 2 hours would be the wearing time of an end range device and the TERT for the day would be 8 hours. In altering the stress dosage to meet the joint’s needs, one could alter either the force of the TERT, or both. Until force is studied, the stress prescription may read “splint for 2 hours, QID.” In practice we routinely adjust the stress dose from visit to visit by adjusting the TERT. Such an approach makes stress delivery more objective and controllable than do more empirical clinical approaches.

Based on the results of this study and on our clinical experience with many different joints, this principle should apply to most synovial joints, not just the PIP. After all, the biologic processes involved in this phenomenon are universal and are not specific to the PIP. It would be interesting to conduct a follow-up investigation using a similar approach at a different synovial joint, such as the knee, to verify this supposition. In fact, Light et al. found that stiff knees treated with LLPS for 2 hours per day improved about four times as much as knees treated for one-half hour per day. Though the study by Light et al. was not as controlled as this study, it also suggested a direct relationship between TERT and ROM gain.

The initial 2-day resting cast applied prior to the testing period was included to eliminate the commonly observed clinical phenomenon that simply resting an irritated joint for 48 hours, allowing it to calm down, can produce a quick improvement in ROM. We wanted range changes to be a result not of a changing inflammatory status but, rather, of the joint stiffness itself.

The study could have been improved by including more than two TERT values, such as 1 day or 9 or 12 days, to determine whether the same linear relationship held true as TERT varied. However, the level of significance (p < 0.005) suggests that the relationship is very strong as it stands, without further substantiation. Keeping the investigator from seeing the goniometer would have eliminated all possibility that the tester was influenced by seeing or by knowing the scores; this was not practical in the clinical setting. However, the investigator was unaware of the grouping assignments within this study.

In conclusion, the increase in PROM of a stiff joint is directly proportional to the length of time the joint is held at its end range, or TERT.