Test-Retest Reliability of the Dizziness Symptom Profile

Objectives: This investigation was conducted to measure the test-retest reliability of the Dizziness Symptom Profile (DSP). The DSP was developed to assist primary care providers, general otolaryngologists, and other health care providers in the development of a differential diagnosis for patients who present with dizziness, vertigo, or unsteadiness. The DSP yields a score ranging from 0 to 100% for each of 7 subscales. Each subscale represents a different diagnosis including benign paroxysmal positional vertigo, Meniere's disease, persistent postural-perceptual dizziness (PPPD), superior semi-circular canal dehiscence, vestibular migraine, vestibular neuritis, and general unsteadiness. Design: Subjects were 150 adult patients (mean age 56.79 years, SD 15.69 years) referred to the Balance Disorders Clinic at Vanderbilt University Medical Center. Subjects completed two administrations of the DSP. The mean interval between test administrations was 1.58 days (SD 1.78 days). The response modes for the DSP were both a 0 to 100 mm visual analog scale (scored 0 mm = "strongly disagree" to 100 mm = "strongly agree") and, by extrapolation, the original 5-point Likert scale where the anchors were "strongly disagree" (scored 0 points) and "strongly agree" (scored 4 points). Results: Pearson correlation coefficients were calculated to assess test-retest reliability for individual DSP items, and ranged from r = 0.67 to 0.91 (mean 0.80; p < 0.001). Cronbach's alpha coefficients were calculated to assess internal consistency reliability of items comprising the seven subscales. Each subscale had an acceptable level of internal consistency (Cronbach's alpha coefficients > 0.7) with the exception of PPPD which approached 0.7. Intraclass correlation coefficient estimates and their 95% confidence intervals were also calculated to assess the relative reliability of the subscales. All 7 subscales showed moderate to strong test-retest reliability, with intraclass correlation coefficients ranging from 0.85 to 0.94. Minimal detectable change (MDC) scores were calculated to assess absolute variability/measurement error for the seven subscale scores (which range from 0 to 100%). MDC values ranged from 16% (PPPD) to 25% (unsteadiness). Conclusions: (1) The test-retest reliability of the DSP is moderate to strong. (2) MDC values for each subscale were determined. (3) The DSP coupled with the Dizziness Handicap Inventory enables the clinician to evaluate the constructs of dizziness impairment, and disability/handicap. (4) The DSP may help provide a window to the natural history of dizziness disease(s). (5) The DSP provides a less biased assessment of the symptoms reported by the patient.