Recently in The Journal of Family Practice, Nease and Malouin (“Depression screening: a practical strategy,” J Fam Pract 2003; 52:118–124) did an excellent job of identifying a rationale and strategy for screening for depression in primary care. They underscore 2 critical issues for practitioners:
- There are a large number of patients with depression, and they are more likely to be treated in primary care than in the specialty mental health system1
- Primary care is insufficiently equipped to provide that care, often does not detect its need, and often insufficiently treats the problem even if identified.2
The response they propose is sound, but a number of points might be emphasized to further clarify the evidence. The first concerns the implicit assumption that screening will lead to better treatment. The second concerns the populations to be screened. The third concerns the high comorbidity of depression with other psychiatric diagnoses. The last concerns the relationship of screening data to best patient care for depression.
There is good reason to be concerned that depression-screening data do not necessarily affect patient care or physician behavior.3 If we are going to improve identification of depression, access to the best pharmacologic and psychologic treatments must be equally improved. It has been suggested that access to best practice in those areas is inaccessible to most primary care patients.
While the article notes that when and whom to screen are important to consider, the selection has consequences for the cost and effectiveness. Coyne and colleagues4 have noted that screening instruments are positive for depression symptoms 18%–30% in primary care populations, with an estimated cost of $60 for the follow-up assessment to every positive screen. However, 70% or more of these positive screens do not result in a diagnosis of depression, which generates significant expense that needs to be carefully evaluated. Coyne et al further observe that generalized screening has generated lower rates of depression treatment than expected.
There has been ample demonstration of high comorbidity levels between depression anxiety disorders and substance abuse. Further, anxiety disorders and substance abuse, considered individually, are seen almost as frequently in primary care populations as depression, have as great comorbidity with acute and chronic medical conditions, and use astoundingly high amounts of health care resources.5 One of the important features of the Patient Health Questionnaire (PHQ) is that it has modules for anxiety and substance abuse; scoring time needed to include these 2 modules is negligible while producing important additional data.
In summary, the entire point of depression screening is to improve the access to and outcomes of depression care. To do so, primary care must make considerable educational, behavioral, and organizational changes. Depression screening is a key dimension of those needed changes, and the Nease and Malouin article is certain to assist that.
However, screening alone may also be seen as insufficient to produce the requisite changes needed to accommodate their suggestions. We must pursue the development and implementation of programs in primary care that change physician behavior, reorganize practice capability, implement appropriate screening, and use the best practices for depression treatment. If depression screening is to escape the criticism of being too narrow an intervention, too costly, and limited in its ability to improve patient care, then we must broaden our target.
Rodger Kessler, PhD,
Berlin Family Health, Central Vermont Medical Center,
Berlin, Vt
Drs Nease and Malouin respond:
We appreciate Dr Kessler’s comments on our paper and the opportunity to respond. We agree wholeheartedly with his first 2 points. Depression screening must be linked to effective systems and strategies of care in order to realize success in improving patient outcomes. This is an important point of the recent United States Preventive Services Task Force report.6 Because the PHQ-9 can facilitate the monitoring needed in any system of depression care,7 we emphasized this point in our paper.
We also agree that it is important to strategically consider who and when to screen. This issue is closely linked with the first. Where practices successfully utilize effective and efficient mental health care strategies, a clinical “cost-benefit” equation may suggest that frequent screening is justified. Practices must carefully consider their individual situations as they develop their own depression identification and treatment strategies.
Finally, we’d like to offer an alternative approach to the issue of comorbidity. Depression and anxiety disorders show a great deal of overlap in epidemiology and pharmacology. Our understanding of the significance of this overlap is even limited by criteria used to define these disorders.8 Similarly, substance abuse disorders overlap greatly with depression. While the full PHQ9 permits screening for these additional disorders at potentially negligible additional patient and scoring burden, it increases the interpretation and treatment burdens identified in our paper. We believe that if a busy primary care practice decides to implement screening, they should focus on depression primarily, recognizing that depression is often a sentinel marker for co-occurring anxiety and substance abuse disorders.