Interpretive Database

The added benefit of this procedure is that the clinician can quickly determine whether the SOS-10 is valid (scores of 0 or 60 suggest a biased response set and are considered invalid), and get a total score before a session starts to have a sense of their patient’s current level of distress.

Since its publication, the SOS-101 has been used as an outcome measure in a broad range of clinical programs and research studies. Many of these treatment programs and research projects have contributed data to aid in the development of an interpretive database for the SOS-10. SOS-10 data have been collected for non-patients, outpatients, and inpatients. Presently, this database contains over 9000 subjects with 2336 non-patients, 1598 outpatients and 5119 inpatients. The sample is 62 percent female and has a mean age of 30.45 (SD = 14.2). As can be seen in Table 1, the SOS-10 provides strong separation between both the total patient and nonpatient groups, as well as within each patient group. Using the average coefficient alpha (.93) of the published studies (described later) that have used the SOS-10, the standard error of measurement was calculated (SEM = 4) in order to provide a confidence interval around SOS-10 scores. In other words a change of 4 or fewer points between any two measurement periods is most likely due to chance and/or measurement error and should not be interpreted as clinical improvement.

Using the large patient and nonpatient samples in the database, a cutoff score of 41 was calculated using the formula suggested by Jacobson and Truax5 to be the threshold that separates the functional (non-patient) and dysfunctional (patient) distribution of SOS-10 scores. In other words, patients who score at or above a 41 are thought to have a good sense of well-being and a minimal level of distress, while patients that score below a 41 are considered to be in the clinical range. As it was hoped this measure would be used to track treatments over time, a reliable change score (+/- 8.5) has also been calculated using the recommended formula from Jacobson and Truax5 to help reduce misinterpretation. Therefore an increase or decrease of at least 8.5 points on the SOS-10 between any two measurement periods of the same patient or group is considered unlikely due to chance alone, thus suggesting the impact of treatment. Combining the clinical cutoff score of 41 with the reliable change score of 8.5 points allows us to operationalize a rigorous definition of improvement that includes a meaningful increase in well-being (score change of 8.5 or greater), as well as a total score that falls into the functional (non-patient; > 41) range of scores.