A Randomized Controlled Trial and Cost Analysis Assessing the Value of Physical Therapy for Treatment of Adhesive Capsulitis

Abstract

Objectives:

Adhesive capsulitis of the shoulder when diagnosed from history, physical exam, and radiographic evidence is generally treated with a six-month course of conservative therapy. Currently, no study has compared the use of corticosteroid injections and physical therapy (PT) to corticosteroid injections and supervised neglect (SN). Due to the average duration of symptoms associated with this syndrome, the cost of physical therapy to the patient and the medical system can be considerable. The purpose of this study is to compare physical therapy and supervised neglect while assessing the costs associated with therapy to perform a value analysis.

Methods:

After IRB approval was obtained, patients were prospectively randomized into either PT or SN study arms. Patients in both groups were offered corticosteroid injections at the time of first visit and at follow up intervals of 3, 6, and 12 months following enrollment. Patient reported outcome measures (PROMs) including the Disabilities of the Arm, Shoulder, and Hand (DASH) Score and the American Shoulder and Elbow Surgeons (ASES) Shoulder Score were collected at enrollment and at 6 weeks, 3 months, 6 months, and 12 months follow-up. Costs to patients were evaluated by interview at the time of final follow-up and included copays, cost of transportation, and opportunity cost of time missed from work. Value measures were generated per patient by dividing longitudinal improvements in PROMs by the costs associated with adhesive capsulitis management (PT vs SN). Analyses for improvement within treatment groups as well as comparisons of the two groups were performed. Power analysis indicated minimum sample size of 10 (r >.8) to reach non-inferiority significance. All analyses were calculated with STATA 14.2 (STATA SE 2015) and Microsoft Excel (Microsoft Corp).

Results:

A total of 26 patients were enrolled, with 12 patients randomized to PT and 14 patients randomized to SN. Six (50%) of the 12 patients enrolled in PT and 7 (50%) of the 14 enrolled in SN have achieved 12 month follow-up. The mean follow-up time of both groups was 10 months. No statistically significant differences were observed in follow-up time between the PT and SN groups. Statistically significant improvements from enrollment to follow-up were observed in both the PT and SN groups for the DASH (Tables 1.1 and 1.2). A comparison of the PROM improvements between the PT and SN groups yielded no statistically significant difference (Table 1.3). Average patient costs due to adhesive capsulitis treated in the PT group were $6,150.46 (range, $191.52-30,000.00), whereas patient costs sustained in the SN group were $45.00 (range, $30.00-75.00). Statistically significant differences were observed when comparing the cost and value of PT to SN.

Conclusion:

Both PT and SN treatment modalities showed significant improvements in the DASH Score at a mean follow-up of 10 months, indicating potential non-inferiority of SN compared to PT. Financial analysis of the two groups showed a significant difference in value with respect to outcome measures. The combination of primary costs (copays) and secondary costs (travel, copays and opportunity costs) diminish the clinical value of physical therapy as an ideal treatment to the patient. This study suggests the consideration of supervised neglect as a viable treatment approach for adhesive capsulitis, without unnecessary financial burden to the patient and medical system.

Table 1.1

Comparison of Enrollment vs. Follow-Up Means for ASES and DASH in Physical Therapy Arm (ARM 1)

	Enrollment	Follow-Up	p-value
ASES	44.167 (29-57.667)	88.333 (47.5-98)	0.0630
DASH	.404 (.392-.6)	.067 (.025-.345)	0.0277*

Table 1.2

Comparison of Enrollment vs. Follow-Up Means for ASES and DASH in Watchful Waiting Arm (ARM 2)

	Enrollment	Follow-Up	p-value
ASES	58.583 (43.917-65.583)	89.167 (85.833-99)	0.0117*
DASH	.375 (.267-.538)	.033 (.009-.1)	0.0173*

Table 1.3

Comparison of % Change in ASES and DASH between Arm 1 and Arm 2

	Arm 1	Arm 2	p-value
Difference in ASES	44.167 (-9.167-69)	37.833 (23.5-44.5)	1.0000
Difference in DASH	-.346 (-.575- -.25)	-.312 (-.526- -.221)	0.6002

All data provided in tables are structured as median (interquartile-range) unless stated otherwise.