There has been a recent explosion in the use of brief screening tools as a prevention method for treating depression. Yet research is not keeping up with practice, and concern is rising that the most popular depression screen used today – the PHQ-9 – is flawed and its use can sometimes do more harm than good. Here is a summary of the those concerns, and what can be done to overcome them:

"Depression" is not a clinical diagnosis

When a health professional identifies a case of Major Depressive Disorder, it’s usually the result of an in-depth interview that explores 9 key problem areas, each of which has a very distinct set of symptoms that have persisted for several weeks, months, or even years. The interview process is quite lengthy, and serves to rule out relatively mild examples that are in fact normal, and to ensure that there is not a different, more appropriate diagnosis. 

When a screening test like the PHQ-9 identifies a case of depression, it is the result of a quick checklist identifying how frequently 9 very common symptoms have occurred in the past couple of weeks. The difference between the two methods - and their outcomes - is very significant. A diagnosis of Major Depressive Disorder reflects a serious problem that requires immediate treatment. A positive screen for “Depression” suggests only that something might be wrong, and should cue the clinician to ask many more questions in order to determine the exact nature of the problem, so that an appropriate diagnosis can be made if it in fact exists. 

Most people who are mildly or moderately depressed according to the PHQ-9 do not meet criteria for an episode of Major Depressive Disorder.

The available evidence suggests that screens like the PHQ-9 are highly sensitive. In other words, they catch almost every true case of major depressive disorder. However, there is also something called specificity. A test with high specificity only catches the target condition, and does not mix this up with other diagnoses that share similar symptoms. Unfortunately, highly sensitive tests usually have to sacrifice their specificity. As they diagnose more people, they make more mistakes by including people who aren’t depressed. Because the PHQ-9 sacrifices specificity so that no true cases are missed, many more people are mis-diagnosed. This means that a lot of people who have a positive screen will be at risk of getting treatment they don’t even need.

Medication does not work for mild cases of Major Depressive Disorder, yet it is the most commonly prescribed treatment

With most screening tools, the lower your score is, the less likely you are to have the problem. This means that as scores get lower the likelihood of an incorrect diagnosis increases dramatically. Tools like the PHQ-9 try to get around this fact by specifying the “severity” of depression. To do this, they create cut-off scores to isolate cases with “mild”, “moderate”, or “severe” depression. But what these categories really should be called is “probably isn’t experiencing major depressive disorder”, “might be experiencing major depressive disorder”, or “probably is experiencing major depressive disorder”. The difference is important, since the evidence suggests that someone who is identified with a “mild” case of depression using the PHQ-9 will probably be prescribed medication, whereas someone who “probably doesn’t have major depressive disorder” shouldn’t be prescribed anything. What’s more, even if someone is experiencing major depressive disorder, if their symptoms are mild, research suggests that medication probably won’t have much impact on their symptoms.

Many people who are severely depressed according to the PHQ-9 actually have another problem, or more than one psychiatric disorder

Recent research suggests that the cut-off for the PHQ-9 needs to be 14 or 15 to catch a genuine episode of major depressive disorder. However, even then it tends to misdiagnose people who have other diagnoses, such as Bipolar Disorder, Panic Disorder, or a Personality Disorder. Also, many people with more serious symptoms of a depressive disorder usually have more than one diagnosis. But when a more appropriate (i.e., higher) cutoff is used, the specificity of the PHQ-9 plummets. This means that it performs reasonably well as a screening tool that can spot whether a more general mental health problem exists, but is not very good at picking out cases of major depressive disorder.

Recommendations to use the PHQ-9 do not reflect available evidence

Don’t think things could get worse? Consider the fact that in spite of the PHQ-9 being available for close to two decades there is no empirical evidence that it provides any benefit. That’s right, when it comes to the gold standard for research studies - randomized, controlled trials - there have been none that look at whether patients who get a positive screen actually benefit from this information. While people have spent a lot of time trying to figure out what cut-offs are best, and literally thousands of studies have used the PHQ-9 to measure the severity of symptoms, almost no one has looked at what happens to people when it is used routinely to identify cases of major depressive disorder and start antidepressant treatment. 

Routine use of screens like this to diagnose a major depressive episode may actually increase health care costs

You would think that if screens are supposed to help people, someone might have looked at how much money is saved. After all, we know that the fallout that results from an episode of major depressive disorder can be severe in terms of the cost to society at large. It often involves losses at work, with family, and in overall productivity. So if we can accurately diagnose and treat this problem, the savings to society should be huge. On the other hand, antidepressant treatment is relatively expensive, and if the medications are prescribed inappropriately, there is more to worry about than the money wasted filling a prescription. These drugs can have a lot of unwanted side effects - particularly for people who take them but don’t need them. So it is perhaps not surprising that a recent Quebec study showed that antidepressant use is actually associated with dramatically higher health care costs. What’s more, it found that many of the people who had prescriptions had no recent history of depression or anxiety.

Is there a solution?

While it may sound like the PHQ-9 and other screening tools should never be used, there are actually some pretty straightforward fixes that allow them to be used responsibly and effectively. A diagnosis of Major Depressive Disorder, or any psychiatric diagnosis for that matter, is a serious thing that requires attention, so coming up with ways to ensure that a problem is not missed is important. Here are some of the most helpful guidelines to make sure that screening tools help more than they hurt:

1. Mental health screens are only appropriate in settings where a qualified mental health professional is available to follow up on any positive results. 
2. Ideally, screening tools should only be used with people who are presenting with some initial complaints or concerns about how they are feeling. They should not be given randomly to people who don’t even have symptoms.
3. The cut-off on screening tools should be high enough to ensure that mild symptoms of normal distress are not confused with a psychiatric diagnosis. In the case of the PHQ-9, the evidence suggests that a score of 15 is the lowest cut-off that should be used. 
4. Follow up needs to rule out other diagnoses that may share similar symptoms. Additional screens may be used to get this started, but the process will still require a thorough interview.
5. There is more than one “gold standard treatment” for major depressive disorder. Medication is just one of these options. Be sure that you are making an informed choice and have all the information on effective treatment.
6. Screens like the PHQ-9 actually work better when used to track symptoms once a person has received an appropriate diagnosis. If you are just starting or already getting treatment, be sure to use some sort of symptom monitor so that you can spot whether things are improving, and know when you’ve achieved your goals. 
7. If you are using a screen for monitoring purposes, be sure that it includes the full range of symptoms you are experiencing. Plus, be sure to include a measure of positive symptoms. For example, Myndplan routinely uses a Life Satisfaction monitor to accomplish this. After all, feeling better involves more than just getting rid of negative symptoms. 

My own experience suggests that screens like the PHQ-9 work best when used by a clinician who understands their limits, and how to appropriately follow up when results indicate that their might be a problem. On the other hand, they perform worst in busy clinics where providers are pressed for time and rely on their prescription pads as the sole method of treatment.

References

Inoue, T., Tanaka, T., Nakagawa, S., Nakato, Y., Kameyama, R., Boku, S., Toda, H., Kurita, T., & Koyama, T. (2012) Utility and limitations of PHQ-9 in a clinic specializing in psychiatric care. BMC Psychiatry. 12(73). https://bmcpsychiatry.biomedcentral.com/articles/10.1186/1471-244X-12-73

Thombs, B., Ziegelstein, R.C., Roseman-Kloda, L.A., & Loannidis, J.P.A. (2014). There are no randominized controlled trials that support United States Preventive Services Task Force guidelines screening for depression in primary care: a systematic review. BMC Med. 12(13). DOI: 10.1186/1741-7015-12-13. 

Vasiliadis, H.M., Latimer, E., Dionne, P.A., & Preville, M. (2013). The costs associated with antidepressant use in depression and anxiety in community-living older adults. Can J Psychiatry. 58(4), 201-9