Depression Screening Scores Rise While Referral Slots Stay Empty
Primary care clinicians are detecting depression at higher rates than a decade ago, yet the proportion who reach a therapist or psychiatrist has not moved in step. The gap between a positive PHQ-9 and a booked appointment is the subject here, and the evidence on what actually closes it.
The Screening Paradox in Primary Care
Walk into most primary care clinics in the US, UK, or Germany and you will find the PHQ-9 embedded in the intake form, the annual review, or the chronic disease template. The nine-item scale takes under three minutes to complete, and it has become close to universal. Scores in the moderate range, 10 to 14, are now flagged routinely in patients who came in for something else entirely.
What happens next is less uniform. In many systems, a score above a threshold triggers a referral request rather than a treatment decision. That request enters a queue. In parts of the UK, talking therapy waiting times have stretched into months for some patients; in US metropolitan areas, child and adolescent psychiatry openings can be scarce enough that families drive hours. The detection has improved. The capacity has not.
The result is a familiar clinical scene: a patient leaves with a leaflet, a phone number, and a score documented in the record. The clinician has done what the template asks. The patient has not been treated. This site has argued before that depression guidelines assume a clinician who has time, and that assumption is doing a lot of work in the referral gap.
What the Evidence Actually Supports
The PHQ-9 was introduced in 2001 as a severity measure and diagnostic aid for adults in primary care. It performs well at that job. It was not designed to decide who needs a psychiatrist rather than a GP, and it does not reliably make that distinction. Using it as a triage tool is a common drift, not a validated use.
Screening programs in isolation have a weak record. Systematic reviews of routine depression screening in primary care have generally found that detection rises but outcomes do not improve unless the screening is paired with a structured treatment pathway. The score is a signal. It is not an intervention.
Integrated and collaborative care models, where a care manager and a consulting psychiatrist support the GP, do show modest but consistent benefits for depression outcomes. The effect sizes are not dramatic. They are real, and they depend on the treatment arm being staffed. A collaborative model with no therapist attached is a meeting schedule, not a service.
Referral capacity sits underneath all of this. No amount of better screening fixes a system where the receiving end is full.
Why Clinicians Keep Screening Anyway
Quality metrics reward completion. In many health systems, the proportion of patients with a documented PHQ-9 is a reported indicator, and the proportion referred is sometimes another. The metric measures the act of screening, not the outcome of care. Clinicians respond to what is measured, and the measured thing is the form.
Medico-legal pressure reinforces it. A documented score and a documented referral attempt are defensible. An undocumented conversation about why watchful waiting might be reasonable for a patient with mild symptoms is harder to defend if something goes wrong later. Documentation drives behaviour, and the safest documentation is a referral.
For mild presentations, there is often no attractive alternative in the room. A GP with a fifteen-minute slot and a patient scoring 11 has few tools that fit the time. The questionnaire is the default because nothing better has been placed within reach.
The Empty Slot Problem
Therapy openings are scarce across high-income systems, and the scarcity is not evenly distributed. Psychiatrist shortages are documented in the US, the UK, and much of the EU, with rural and lower-income areas hit hardest. A patient in a city may wait weeks. A patient two hours from the city may wait longer or travel.
Digital tools fill part of the gap. Guided self-help platforms, app-based CBT, and telephone coaching can reach patients who would otherwise sit on a list. They work for some presentations and not others, and they do not suit patients without reliable internet, a private space, or the literacy to engage with a text-heavy program.
Patients get lost between the score and the session. They are told to call a number. They call, leave a message, and do not call back. The record shows a referral was made. The patient shows nothing. A related piece on how COPD admissions track the cold, not the smoking rate makes a similar point about upstream factors that sit outside the clinical encounter.
Rethinking the Referral Pathway
Stepped care matches severity to setting. Mild symptoms get brief GP-led intervention or guided self-help. Moderate symptoms get structured psychological therapy. Severe symptoms, or those with risk features, get specialist input. The model is well described. Its weak point is the step between mild and moderate, where waiting lists are longest.
Collaborative models embed a therapist or care manager in the clinic, which shortens the distance between detection and treatment. When the therapist is in the building, a warm handoff becomes possible, and the patient does not have to navigate a phone tree alone. The cost is staffing, and staffing is the constraint everywhere.
GP-led brief interventions have a place for mild depression. Behavioural activation, structured follow-up at two to four weeks, and sleep and activity planning can be delivered in primary care without a specialist. They are not a substitute for therapy in moderate or severe illness. They are a way to avoid leaving mild cases entirely untreated while they wait.
Measuring function, not just symptom scores, changes the picture. A patient whose PHQ-9 falls from 14 to 11 but who has returned to work and is sleeping again has improved in ways the score does not capture. Function measures are harder to collect and less tidy in a dashboard.
Practical Steps for Busy Clinicians
Ask about access before referring. A patient who cannot take weekday appointments, cannot afford a co-pay, or has no transport will not attend a service that assumes all three. Knowing this changes the referral, not just the conversation.
Use warm handoffs where the service allows it. A direct introduction, a shared note, or a same-day call converts more referrals than a printed number, and it costs a few minutes.
Document the referral attempt and the outcome. If the patient is still waiting at four weeks, that is clinical information, and it belongs in the record alongside the original score.
Consider watchful waiting for mild scores with a clear follow-up date. A PHQ-9 of 8 in a patient who is sleeping, working, and not expressing hopelessness can be reviewed in two to four weeks rather than referred immediately. The follow-up date is the intervention.
Advocate for local integrated services using the data already in the system. The number of patients referred, the number seen, and the median wait are three figures that make the case better than a position paper. This site has argued that stewardship targets built for hospitals miss nursing homes, and the same pattern applies when mental health targets are built for specialist settings and applied to primary care.
A Case from Clinic: The Score That Didn't Tell the Story
Last month, a 58-year-old man came in for a routine diabetes review. His PHQ-9 was 12, flagged by the electronic template. He denied feeling depressed but mentioned he had stopped playing golf and wasn't sleeping well. We talked about his sleep, his loss of interest, and his recent retirement. He agreed to try a structured sleep and activity plan with a follow-up in three weeks. At that visit, his score had dropped to 7. He had resumed his morning walks and was sleeping better. The score alone would have sent him to a waiting list; the conversation revealed a treatable adjustment reaction. This is not an argument against screening, but a reminder that the number is a starting point, not a diagnosis.
This article is informational and does not constitute personalised medical advice. Anyone concerned about their mental health should speak with a qualified clinician.
