As clinicians, we routinely make critical decisions for our patients with depression. Because of the uncertainty of factors that affect diagnosis and treatment, clinicians may find an objective, quick measurement tool helpful. Measurement-based care (MBC) provides specific and objective information on which to base clinical decisions and should therefore enhance quality of care and treatment outcomes.^1-3

MBC rests on these assumptions.

• Compared with general questions that are typically asked during a patient evaluation, specific measurements (administered by clinicians or self-reported by patients) provide more accurate information on which to establish a diagnosis, assess treatment outcomes, and modify treatments.

• Patients who complete these mea-surement tests will better understand their disorder and treatment effects, which will enable them to better manage their depression.

• Medical records that include the results of specific measurements will assist subsequent clinicians in understanding the results of prior treatments.

• The routine use of the same mea-surements in practice and clinical research studies will help clinicians translate research findings into their own practices.

• For most outpatients with depression, self-report methods are available that are free and that take little time and effort.

Diagnostic measurements

Researchers have used criterion-based diagnostic methods for years. After DSM-III was introduced in 1980, the Structured Clinical Interviews for DSM-III (SCID) (and later for DSM-IV) were developed to obtain lifetime diagnoses.^4,5 Briefer structured interviews were then developed, including the Mini-International Neuropsychiatric Interview (MINI), which assesses only current diagnoses, and the MINI-Plus, which elicits information about current and past diagnoses.^6-8 The MINI takes 30 to 40 minutes to administer, while the MINI-Plus may take up to 60 minutes.

Studies have shown that structured or semistructured interviews provide more accurate diagnoses than typical practice. For example, clinically rendered diagnoses were compared with those made based on SCID results.⁹ Major diagnostic differences were found in 40% of outpatients with clinical diagnoses of schizophrenia or bipolar or major depressive disorders. In addition, when clinicians were provided with a diagnosis that was determined using SCID, they changed the chart diagnosis in a substantial proportion of cases and prescribed fewer medications.¹⁰

Symptom measurements

Once a diagnosis has been made and therapy has been initiated, the regimen must often be modified because of intolerance, adverse effects, or other less-than-desirable symptomatic outcomes. Medication and somatic therapies are typically aimed at treating symptoms, but psychotherapy and disease self-management may also address other aspects of treatment (eg, medication adherence, social/occu-pational function, self-esteem). The Texas Medication Algorithm Project (TMAP) and the Sequenced Treatment Alternatives to Relieve Depression (STAR*D) studies showed that diligent assessment of symptoms and adverse effects enhances outcomes.^11-14

The goal of therapy for depression is symptom remission and, ultimately, sustained remission and functional recovery.¹⁵ Most patients require more than 1 treatment revision (eg, altered dosage, treatment, or delivery). When implementing guideline-driven or evidence-based care, initial treatment is continued until remission or maximal symptom improvement is obtained or until the patient cannot tolerate the regimen. Thereafter, the dosage or type of treatment may be changed.

Specific implementation is associated with a significant number of errors that do not reflect clinical need (such as too rapid or too slow treatment change). Some clinicians try an antidepressant for too short a duration (eg, only 2 to 4 weeks) and then change medication if the patient globally sees “no benefit.” Because patients with depression have negative cognitive biases, they are less likely to recognize modest improvements when asked global questions about their symptom response. Thus, a medication that has some modest effect after 2 to 4 weeks may be mistaken as “ineffective” by the patient and then stopped, when its continued use could have produced remission. Yet, when itemized symptom measurements are used, clinicians can often detect some meaningful degree of improvement, leading to the continuation of treatment—perhaps with a dose escalation rather than a treatment change. This scenario highlights the drawbacks of depending on global judgment rather than a more careful, detailed assessment to make appropriate treatment decisions.¹⁶

Conversely, a clinician may continue a treatment that affords some improvement but not remission—even after optimal therapeutic effect could be expected. Perhaps the clinician and his or her patient with depression are willing to settle for “better, but not well” (ie, a response) as a good enough outcome. Or perhaps the clinician is not aware that the patient is still experiencing substantial symptoms. Again, a sensitive symptom measurement provides better and more accurate information about “residual” symptoms that could be diminished by adjusting the dosage or type of treatment.

The impetus for developing symptom measurements came from research trial needs. The gold standard has been the clinician-rated 17-item Hamilton Rating Scale for Depression (HAM-D-17), which is also available in longer formats.^17,18 The Montgomery-sberg Depression Rating Scale (MADRS) was developed from a long list of items, 10 of which are optimal for differentiating drug efficacy from placebo in trials.¹⁹ MADRS items are sensitive to symptom changes with treatment. Neither the HAM-D-17 nor the MADRS, however, includes all 9 DSM-IV diagnostic symptom domain criteria that define a major depressive episode.