Mental health clinical decision support
Clinical decision support tools often focus on treatment choices for providers after a diagnosis is made. This is because for most specialties, there are objective data sets for making diagnostic decisions, like radiologic imaging, labs, vital signs, or other direct physical exam findings. In mental health care, most of these modalities are not easily applicable to our field, thus leaving clinicians to make decisions mainly on the reported verbal elicitation of symptoms and by using acute observational findings from interactions with the patient. This process leaves much variability in the diagnostic and symptom tracking process as there are many patient and provider interaction variables that impact the statistical sensitivity of the data collected in diagnostic and follow-up assessments. Over the past couple of decades, researchers in mental health have developed many rating scales to bring some standardization to the diagnostic evaluation and treatment tracking process.
Many instruments have been created to include diagnostic-specific and general distress scales. Some of these are available in the public domain while others are proprietary. Some are clinician-rated scales based on observations with the patient and others are patient self-reported scales. Some studies have explored whether clinician-rated scales are similar to patient self-report scales. One of the largest studies that collected both types of data is the STAR*D (Sequenced Treatment Alternatives to Relieve Depression) trial that looked at standardized treatment of depression. This study found that the response and remission rates for depression were found to be nearly equal when comparing clinician-rated to patient-rated scales (1).
Common patient self-reported scales include:
Depression: Beck Depression Inventory (BDI), Patient Health Questionnaire 9 (PHQ-9), Quick Inventory of Depressive Symptoms (QIDS)
Bipolar Mania: Altman Mania Scale
Anxiety: Generalized Anxiety Disorder (GAD-7), PHQAnxiety, Hospital Anxiety and Depression Scale (HADS)
Alcohol Use Disorders: Alcohol Use Disorders Identification Test (AUDIT)
Post Traumatic Stress Disorder: PTSD Checklist (PCL)
Benefits for Using Standard Scales:
Using standard scales allows for improved diagnostic accuracy and therefore better treatment choices based on a more accurate and complete diagnosis. For example, there was improved inter-rater agreement when standardized data was used in emergency psychiatry cases and in outpatient settings (2,3). One can also start measuring degree of change in the symptom burden of a mental illness by using these scales. Many of these scales have shown to detect disease well and to also measure clinically significant change. By using these standard scales, we can start to track clinical response and remission more accurately and consistently than by the current common practice of asking the patient how he/she has been doing recently. Clinicians often see that patients struggle to accurately assess how they are doing due to cognitive and emotional errors in recall of past and current events that are common to the illnesses being treated. By iteratively utilizing standardized methods of asking how a patient is doing in the moment at a single cross-section of time may allow for the most accurate assessment possible of current functioning. Doing this at each visit then allows for the potential of improving psychological awareness of patients about their own functioning. Building these cross-sections into a timeline record then allows a clinician and a patient to view progress of these target symptoms more accurately.
Current Computer Decision Support Tools for Mental Health:
There are several initiatives over the past decade that have tried to bring self-report tools mentioned above into a coherent software package for patient data entry and for clinician viewing. Some of the most prominent efforts are the OQ Analyst (4), Carepaths (5), and Q-Local(6). The OQ Analyst utilizes the Outcomes Questionnaires (OQ) in electronic form to track general distress among patients. The Carepaths system uses the OQ Analyst functions along with other disease-specific scales and integrates them into an online mental health electronic medical record offering. The Q-logic system uses the Brief Symptom Inventory (BSI-18) for its main outcome measure. These systems attempt to collect data directly from a patient via a self-report portal and then automatically score the scales and report findings to a clinician in real-time. The clinician can then use this data and share it with a patient in a treatment session.
Evidence of Effectiveness:
The OQ-Analyst has the most robust research evidence showing that use of standard self-report scales in a computer-assisted design actually affects patient outcomes. Over about 10 years, several studies and meta-analyses have shown that routine use of a general distress measure assists in tracking symptom distress and helps clinicians gauge when treatment is not improving. This allows the clinician to engage in a re-evaluation to determine if treatment efforts need to shift and to engage the patient to encourage continuation of treatment. When both the provider and the patient can view this outcome data, the outcomes compared to treatment as usual without such tracking shows impressive outcomes. Effect sizes for patients identified at risk for dropping out of care range from -0.53 for provider-only viewing of the data to -.70 for both provider and patient viewing of the data with embedded clinical support tools (22). Drs. Duncan and Miller have also shown that utilization of the very simple Outcome Rating Scale for clinicians and patients to track progress can vastly improve the effectiveness of treatment by an effect size of 0.79 after 1-2 years of implementation (8).
The use of standardized rating scales for diagnostic assessment and tracking treatment progress is sorely needed in the mental health field. Once the field starts to use these tools in regular practice, we will be able to start to define and describe sub-populations more consistently and to potentially view differential responses to treatment efforts. This will then lead to possibilities in driving true quality improvement efforts as we become more targeted in our treatment efforts of heterogeneous mental health conditions.
1. Rush AJ et al. An Evaluation of the Quick Inventory of Depressive Symptomatology and the Hamilton Rating Scale for Depression: A Sequenced Treatment Alternatives to Relieve Depression Trial Report. Biol Psychiatry 2006;59(6):493–501
2. Miller, P. Inpatient diagnostic assessments: 2. Inter-rater reliability and outcomes of structured vs. unstructured interviews. Psychiatry Research 2001. 105: 265-271
3. Basco, M R; Bostic, J Q; Davies, D. Methods to Improve Diagnostic Accuracy in a Community Mental Health Setting Am J Psychiatry 2000; 157:1599–1605
4. www.oqmeasures.com 
6. psychcorp.pearsonassessments.com/HAIWEB/Cultures/en-us/Productdetail.htm?Pid=P02010 
7. Shimokawa, K; Lambert, M and Smart, D. Enhancing Treatment Outcome of Patients at Risk of Treatment Failure: Meta-Analytic and Mega-Analytic Review of a Psychotherapy Quality Assurance System. Journal of Consulting and Clinical Psychology. 2010, Vol. 78, No. 3, 298–311
8. Miller, S; Duncan, B; Sorrell, R; Brown, G. The partners for change outcome management system. Journal of Clinical Psychology 2005 61(2): 199–208
Submitted by Millard Brown, MD