Every patient, every visit: Routine tests yield clinically useful data
Mining your database can reveal response patterns, improve patient outcomes.
General psychiatry practitioners such as myself traditionally have relied on writing case reports to describe our clinical experience. One obstacle to getting cases published is that many research journals require submitted articles to include large samples and rating scales as measures of change in the conditions of patients being studied.
I have published articles about my clinical experiences using patient data collected with the Clinical Global Impressions (CGI) scale and other standardized tests. Research instruments such as the CGI can gather empiric data and are easy to use in clinical practice.1
This article describes how routine standardized testing provides useful data for research and improves diagnostic accuracy—and patient outcomes—even before I meet my patients for the first time.
Why use standardized tests?
Benefits. All my new patients undergo screening before their first face-to-face meeting with a psychiatrist. This registration visit takes about 2 hours, after which they are scheduled for an appointment based on clinical urgency. We charge no fee for the screening visit; the benefits of gathering a comprehensive database before the clinical evaluation outweigh the cost of the tests, software, and staff time.
Along with completing insurance and biographical paperwork, patients perform self-administered psychosocial and medical histories and a battery of standardized tests. This information allows me to focus on interpersonal issues—rather than fact-finding—during the first interview. It also ensures a comprehensive patient history.
Bipolar disorder is difficult to diagnose in patients presenting with depressive symptoms. In a 5-year chart review,2 we used data from Structured Clinical Interview for DSM-IV (Mini-SCID) screening tests to assess this tool’s usefulness in diagnosing depressed patients. Data also included each patient’s demographic information, initial clinical diagnosis, current clinical diagnosis, and Symptom Checklist-90 (SCL-90) results.
Among 796 patients who took the Mini-SCID at their initial visit, 256 had a current clinical diagnosis of bipolar disorder and 540 had nonbipolar diagnoses. The Mini-SCID had a sensitivity of 0.58 and specificity of 0.63 in predicting a current diagnosis of bipolarity. This compared with a sensitivity of 0.35 and specificity of 0.98 for the clinician’s initial diagnosis. Among patients with bipolar II disorder, the MiniSCID’s sensitivity was 0.55, compared with 0.20 for the clinician’s initial diagnosis.
Patients who endorsed mania/hypomania on the Mini-SCID yet had a diagnosis of nonbipolar illness had SCL-90 profiles more like those of bipolar than unipolar patients. Therefore, using the Mini-SCID with the SCL-90 might improve in-office recognition of bipolar illness.
Limitations. One limitation to using rating scales to publish experiences in clinical practice is that clinical need, rather than a research protocol, determines the frequency of visits. Another is that we ask patients to rate symptoms they experience in the week before office visits. Thus, the data do not capture changes that occurred in other weeks.
Standardized tests we use
Except for the Quick Inventory of Depressive Symptomatology (QIDS), I selected the tests I use in the late 1980s because of:
- their ease of use and affordability
- my familiarity with them from my academic work
- their suitability for a mood disorder clinical practice such as mine.
Psychosocial history. Patients use an office computer to complete a questionnaire about family and developmental history, financial and employment history, education, health, alcohol and drug history, current stressors, and the presenting problem. Software from Multi-Health Systems (See Related Resources) allows me to add or remove questions as needed.
To ensure privacy when the next patient uses the computer, each patient’s report is deleted after it is printed. I receive the printed report, which details all responses and flags those that may require clarification.
Medical history. A standardized form asks patients about whether they have had most common medical conditions, their present symptoms, and family members’ health. An additional form inquires into psychiatric treatment, family history of psychiatric illnesses, and present medications.
Mini-SCID. The Mini-SCID has several advantages over the Structured Clinical Interview for DSM (SCID):
- Patients self-administer the test on a computer at the office.
- For research purposes, Mini-SCID results are protected from clinician biases because patients are interviewed using uniform questions and circumstances.
