Skip to main content

Table 3 Association of full-time equivalent (FTE) capacity measures with nonpharmacologic therapy utilization in the purchased care system: Results from ten multivariate regression models (n = 1300 observations)

From: The relationship between capacity and utilization of nonpharmacologic therapies in the US Military Health System

Independent variable:b Regression Coefficient Estimate (Standard Error)a
Regression Model Dependent Variable: Users per 1000 sample soldiers
Model 1
TE
Model 4
Massage
Model 5
TENS
Model 6
Oth PT
Model 7
Oth PsychE
Physical Therapist FTE 0.0047 (.0230) 0.0114 (.0170) 0.0302 (.0177) −0.0055 (.0103) −0.0018 (.0029)
Chiropractor FTE 0.1239 (.0889) 0.0842 (.0640) 0.1573 (0.0806)* 0.1003 (.0483)* 0.0273 (.0124)*
Behavioral Health Provider FTE −0.0077 (.0071) −0.0063 (.0050) − 0.0038 (.0051) −0.0056 (.0029)* − 0.0002 (.0010)
Independent variable:b Regression Model Dependent Variable: Encounters per therapy userc
Model 1
TE
Model 4
Massage
Model 5
TENS
Model 6
Oth PT
Model
7 Oth PsychE
Physical Therapist FTE 0.0226 (.0170) 0.0061 (.0060) 0.0116 (.0060)* 0.0014 (.0046) 0.0030 (.0077)
Chiropractor FTE 0.2251 (.0812)** 0.1052 (.0588) 0.0517 (.0248)* 0.0847 (.0284)** 0.0403 (.0490)
Behavioral Health Provider FTE −0.0097 (.0046)* −0.0023 (.0060) −0.0016 (.0016) − 0.0016 (.0013) 0.0018 (.0019)
  1. Models 2 (CH/OM) and 3 (ACU) had insufficient observations to support analysis and are excluded from this table
  2. Abbreviations: ACU Acupuncture, CH/OM Chiropractic and Osteopathic Manipulation, Oth PsychE Psychoeducation (see Additional file 1 for detailed list), OthPTs Other Physical Therapy (see Additional file 1 for detailed list), TE Therapeutic Exercise, TENS Transcutaneous Electrical Nerve Stimulation
  3. *p ≤ 0.05; ** p ≤ 0.01; ***p ≤ 0.001
  4. aEach regression model also included facility type (clinic vs hospital), facility branch (Army vs other) and 6 regions
  5. bAnnual measure computed as FTE per 100,000 patients at the MTF
  6. cFor facilities with no utilization the number of visits per user was imputed as zero