Adults diagnosed with Severe and Persistent Mental Illness (SPMI)

ACORN adult questionnaires have been employed in a variety of settings treatment patients with Schizophrenia, Bipolar and other disorders with symptoms of psychosis and thought disorders who are likely to need long term care. Much, but not all, of the data on patients with SPMI are from programming specializing in these disorders including residential and intensive outpatient programs. The following describes the outcomes for these patients, using outcomes with patients with non-SPMI diagnoses as a comparative sample. Differences in patterns of scores on both the global distress and alliance scales are evident across the diagnostic grouping.

To view the animated graph of scores over time, as well summary data tables used to create the graphs in this report, download:

• TrajectoryofChange-LongTerm-25plus-SPMI.xls

Description of samples

For purposes of these analyses, the sample was restricted to patients with more than 25 assessments within an episode of care. The purpose of this restriction was to provide an estimate of outcomes for patients receiving relatively long term care. While this selection criteria narrowed the sample, the sample selected was found to be comparable to cases with fewer sessions with regard to intake scores and patterns of change during early phases in treatment. The samples selected for this analysis averaged greater change over all than comparable patients with fewer than 25 sessions. This is strong evidence that the patients selected for this samples were benefiting from the longer periods of care.

• Psychotic disorders
  • n=229 (cases with > 10 assessments)
  • n=74 (cases with > 25 assessments)
• Bipolar disorder
  • n=272 (cases with > 10 assessments)
  • n=58 (cases with > 25 assessments)
• Comparison sample (all other cases)
  • n=9244 (all cases with > 10 assessments)
  • n=2695 (cases with > 25 assessments)

Change over repeated measures

Averages of global distress and alliance scores were calculated at assessments numbers 1 , 5, 10, 15, 20, 25 and at the final assessment in the episode, whenever that occurred. The score for the first session was obtained from the form for the second assessment, which asked respondents to look back on the prior session. Both scales were scored in such a way that higher scores represented greater global distress or concerns regarding the alliance.

The data from obtained used several different versions of the Adult Outcome Questionnaire, including versions developed for use in SPMI populations.

In all three samples, both global distress and alliance scores show higher elevation at the first session, with scores falling rapidly during the first 5 to 10 sessions, but continuing to show improvement throughout the episode of care. The diagnostic groups differed in severity of distress and alliance concerns at intake. Bipolar reported the highest level of global distress, with Psychotic disorders reported significantly less distress at intake than the comparison samples. In contract, psychotic disorders reported the highest levels of concern on alliance scale, while Bipolar disorders reported the lowest levels of concern on the same scale.

Many of the Psychotic Disorders completed questionnaires that included symptoms of psychosis. The symptoms tended to be endorsed at a lower level than other symptom items, even this population of psychotic disorders. However, after adjusting for differences in intake scores, all versions of the questionnaires produce comparable patters of change over time.

The following graphs displays mean scores for the Global Distress and Alliance Scores across the repeated assessments.

Score stability

All ACORN measures are designed to be sensitive to change over repeated assessments. In this sense, they can be described as reactive to short term changes in the patients subject level of distress. However, if treatment if effective, one would also expect that the scores would not only fall over time but also become more stable as the patient experienced fewer extreme variations in mood or functioning. In order to estimate stability of scores, the standard deviations for the scores were calculated for five session "waves" of assessments. If treatment were effective in helping patients lead more stable lives with lower levels of distress, one would expect the standard deviations of scores to decrease over time as the average score fell.

This was found to be the case for both Global Distress and Alliance scores. However, clear differences in patterns of score stability was observed across the diagnostic groups. Bipolar disorders displayed a higher level of both global distress and instability of global distress scores over time. Psychotic disorders exhibited a rapid stabilization of global distress scores within the first 10 assessments. However, the alliance scores exhibited significantly greater instability for Psychotic scores compared to the other samples, with s slower downward trend on the standard deviations of scores.

The following display the standard deviations of the Global Distress and Alliance scales for waves of five assessments each.

Alliance and outcome

The relationship between Alliance scores and outcome was explored by correlating the Alliances scores at different points in the treatment episode with the benchmark score for each case. This was done to control for any differences in intake scores. Pearson r correlations with the benchmark score were calculated for the initial alliance score, the mean of alliance scores for the first 25 sessions, the mean of alliance scores for sessions 21-25, and the alliance score at the last assessment. Correlations are calculated so that the association between low alliance scores (absence of alliance problems) and low Global Distress Scores is a presented as a positive correlation.

The patterns of correlated demonstrated that the alliance scores near the end of the treatment episode were much more predictive of outcome than scores at the start of the episode or even average alliance scores across the first 25 sessions. Strikingly, the Alliance scores near the end of treatment are significantly more predictive for the cases with Psychotic or Bipolar disorders that for the comparison sample of other disorders. This is strong evidence for the importance of routine monitoring of alliance when working with the SPMI population.

Change in Global Distress as a function of intake scores

Sample may differ significantly with regards to mean scores at different assessment points, while being very similar in patterns of change over time after controlling for intake scores using a general linear model (GLM) regression analysis where intake score predicts a score a subsequent measure point. Higher scores are expected to show more positive change over time, while very low scores can be expected to remain stable or show a slight increase over time.

The ACORN Animated Change Graph permits to user to visual the scores over time as a function of the intake score. A macro contained within the Excel file alters the intake score. Formulas containing the slope and intercept values from the GLM at each assessment point permit the file to display expected scores are subsequent measurement points. The following image displays to the results for patients with an intake score of 2.

As is apparent, the pattern of change over time is similar across all diagnostic groups, with Psychotic disorders exhibiting a faster rate of change during the first 5 sessions than the comparison samples. Psychotic disorders also reported the lowest levels of distress at the final assessment. This pattern was evident over the range of intake scores. The user is invited to download the attached Excel file and open the tab for "Intake Scores Predict Change". The user may need to lower the security setting in Excel to permit the user to give permission for the macros to run.

• TrajectoryofChange-LongTerm-25plus-SPMI.xls

-- JebBrown - 14 Jul 2011