The PHQ-9 and the question you cannot skip
The PHQ-9 is the most widely used depression measure in the world: nine items, two minutes, free, validated across an enormous range of settings. It is also the instrument most likely to put a suicidality disclosure in front of a clinician who was not expecting one, which is why every practice using it needs a written protocol for item 9.
The PHQ-9 maps the nine DSM criteria for a major depressive episode onto nine self-report items. Respondents rate how often each has bothered them over the last two weeks on a four-point scale, and the item scores sum to a total between 0 and 27.
| Response | Score |
|---|---|
| Not at all | 0 |
| Several days | 1 |
| More than half the days | 2 |
| Nearly every day | 3 |
Severity bands
| Total | Commonly described as | Typical implication |
|---|---|---|
| 0–4 | Minimal | Monitoring generally sufficient |
| 5–9 | Mild | Watchful waiting, re-administer |
| 10–14 | Moderate | Treatment plan warranted |
| 15–19 | Moderately severe | Active treatment indicated |
| 20–27 | Severe | Active treatment, consider medication referral |
These bands are conventional rather than absolute. A score of 10 is frequently used as a screening threshold, but the appropriate cutoff varies by population and by whether you are optimising for sensitivity or specificity. In a setting where missing a case is costly, a lower threshold is defensible; in a setting where false positives generate expensive follow-up, a higher one may be.
Item 9, and why it needs a protocol
The ninth item asks about "thoughts that you would be better off dead, or of hurting yourself in some way." It is scored like the others and contributes to the total, but operationally it is not like the others at all.
Any endorsement above zero requires clinical follow-up. Not because the item is diagnostic of risk — it is a coarse screener that conflates passive ideation with self-harm intent, and its positive predictive value for subsequent attempts is limited — but because a client has just told you something in writing, and a record showing a positive endorsement with no documented response is the worst possible artefact to hold.
The scenario every practice should have planned for. A client completes the PHQ-9 from home at 11pm the night before their session and endorses item 9 at 2. Nobody sees it until the following afternoon. What is your process? If the honest answer is "there isn't one," that is worth fixing before it happens rather than after.
A workable protocol has four parts: who monitors incoming results and within what window; what threshold triggers contact; what that contact consists of; and how the response is documented. It should also say explicitly what the instrument is not — remote screening is not crisis monitoring, and clients should be told, in writing, what to do in an emergency rather than assuming a form reaches someone quickly.
What to do with a positive item 9
Follow-up is a clinical assessment, not a form. The endorsement is the prompt; the assessment covers ideation and its character, intent, plan, means and access, protective factors, history of attempts, and the reasoning that connects those findings to a disposition. Where risk is present, the plan follows — safety planning, means restriction, increased contact frequency, referral — recorded with what was agreed and with whom.
Note that a client can score 3 on item 9 and be at low acute risk, and can score 0 and be at high risk. The instrument is a door-opener. The assessment is the thing.
The tenth question
The PHQ-9 includes a final functional-impairment item — how difficult the problems have made work, home life and relationships — which is not included in the 0–27 total. It is frequently ignored, which is a shame, because functional impairment is often what actually justifies treatment intensity to a payer. A client with a moderate score and severe functional impairment is a different clinical picture from one with the same score and intact functioning.
PHQ-2
The first two items — depressed mood and anhedonia — form the PHQ-2, a two-item screener scored 0–6. It is designed for high-volume settings where administering nine items to everyone is impractical, with a positive screen triggering the full instrument. In outpatient behavioral health there is rarely a reason to prefer it over the full nine, which take about a minute longer.
Limits worth stating
- It is self-report. It measures what someone is willing and able to report, which is affected by insight, stigma, and what they think you will do with the answer.
- Somatic items confound with physical illness. Fatigue, sleep disturbance and appetite change are depression criteria and also consequences of chronic illness, chemotherapy, pregnancy and sleep disorders. Elevated scores in medically complex clients need interpreting, not just recording.
- Two weeks is a short window. It captures current state, not course. A single score says little about trajectory, which is the entire argument for repeat administration.
- It was validated predominantly in adults. Adolescent use has its own literature and a modified version exists; do not assume adult bands transfer.
Documentation
Record the total, the date, the administration method, and — where relevant — the item 9 response and what followed. A score in a chart with no date is close to useless, and a score recorded without the response it triggered is worse than useless.
Then use it. A note that says "PHQ-9 16, down from 21; continuing weekly with behavioural activation focus, will reassess at 12 or below for step-down" does three jobs at once: it records the measure, it demonstrates clinical reasoning, and it establishes the criterion for a future change in intensity. See the golden thread.
Licensing
The PHQ-9 is free to use, reproduce and translate without permission or fee. It was developed by Robert Spitzer, Kurt Kroenke and Janet Williams with an educational grant from Pfizer, and its open availability is a substantial part of why it is used so widely.
Read the trend, not the number
A single PHQ-9 is a snapshot of a two-week window, taken on a day that may or may not have been representative. Three scores across twelve weeks tell you something a single score cannot, and the clinical conversation that matters is almost always about the shape of the line rather than its current position.
This is also the most underused therapeutic application of the instrument. Clients frequently do not notice improvement, particularly in depression, where the condition itself distorts recall toward the negative. Showing someone that their score has moved from 21 to 16 across two months, during a period they would describe as uniformly awful, is an intervention rather than an administrative step — and it is free.
Adolescents
The PHQ-9 was developed and validated primarily in adults. A modified adolescent version exists with adjusted wording and additional items, and adult severity bands should not be assumed to transfer directly. Where you are working with under-18s, use an age-appropriate version and be explicit in the record about which instrument was administered — a chart that simply says "PHQ-9 14" for a fourteen-year-old leaves a later reader unable to interpret it.
Language and translation
The PHQ-9 exists in a very large number of validated translations, which is one of its practical advantages over instruments available only in English. Use a validated translation rather than an ad hoc one — a family member interpreting items live introduces variation that makes serial scores incomparable, and it places a relative in the position of relaying a suicidality question.
Record which language version was administered. A client whose scores were collected in two different languages across an episode of care has a chart that needs that fact to be interpretable.
Verified 29 July 2026. Instrument specifications are as published by their developers; interpretation thresholds vary by population and setting and are presented as commonly cited starting points rather than diagnostic rules. Screening instruments do not establish diagnoses. Primary references: VA National Center for PTSD; APA Services. This page is clinical reference, not clinical or legal advice.