--- title: PHQ-9 and GAD-7 scoring and trend documentation | Nextvisit description: When to give PHQ-9 and GAD-7, how to score and trend them, and how to keep scores structured for MBC reporting. url: "http://localhost:3000/guides/phq9-gad7-best-practices" type: static generatedAt: "2026-09-04T06:57:01.747Z" --- Guide # PHQ-9 and GAD-7 scoring and trend documentation When to give PHQ-9 and GAD-7, how to score and trend them, and how to keep scores structured for MBC reporting. Guide / For Therapists, NPs and PAs, Psychiatrists / 3 min / Published February 3, 2026 / Updated March 22, 2026 PHQ-9 and GAD-7 are the screeners that show up most often in payer-mandated MBC reporting. They are short, free, and well-validated. They are also captured poorly: scored inconsistently, stuck in the HPI as free text, or never trended. ## When to administer Give PHQ-9 and GAD-7: - At every new-patient intake. - At every follow-up during active treatment of depression or anxiety. - On a defined cadence (every visit or every other visit) for patients in maintenance. - Before any major treatment change (medication initiation, dose change, regimen switch, course change). - When the patient or family raises concern about symptom trajectory. Most practices land on every visit during active treatment, every other visit during maintenance. That is a reasonable default. ## How to administer Two patterns work: Pre-visit, on a tablet or via the patient portal. The patient completes the screener before the visit. Scores arrive in the chart and you discuss them in the room. Most efficient, if check-in can support it. In the visit, by interview. You ask the questions during the visit. Slower, more accurate for patients with literacy issues, language barriers, or cognitive concerns. You also hear how they interpret the question. Avoid the third pattern: the patient fills it out and the clinician glances without engaging. Then the screener is a checkbox. ## How to score Validated cutoffs: PHQ-9: - 0-4: minimal depression - 5-9: mild - 10-14: moderate - 15-19: moderately severe - 20-27: severe GAD-7: - 0-4: minimal anxiety - 5-9: mild - 10-14: moderate - 15-21: severe PHQ-9 item 9 is a suicide-risk question. Any positive response is a clinical event. Document the C-SSRS or your standard risk assessment, regardless of the total score. The total is one input. The item pattern is another. Two patients with a PHQ-9 of 12 can look different: high scores on items 1-4 (mood, anhedonia, sleep, energy) vs high scores on items 7-9 (concentration, psychomotor, suicidality). Look at the items. ## How to chart Keep the score as a structured field, not a sentence in the HPI. Trending. A structured score can be charted over time in the record and in practice-level reporting. Free-text “PHQ-9 of 12” in the HPI cannot. Reporting. Payer-mandated MBC reports query structured scores. They do not parse free text reliably. In Nextvisit, PHQ-9 and GAD-7 (plus AUDIT, CRAFFT, PCL-5, and the practice-configurable list) are structured fields. The score lands when responses are captured. The trend graph shows in the chart header. Aggregate reporting rolls up across providers. If your system does not support this, use a dedicated EHR-template section with score and date, separate from HPI prose. Keep the format consistent across visits. ## How to trend A trend is more useful than any single score. Watch for: - Sustained improvement. Two consecutive scores down 5 or more points typically reflects clinically significant improvement, often the “responder” threshold. - Sustained worsening. Two consecutive scores up 5 or more points warrants a treatment-plan revisit. - Plateau at a partial response. Stuck at a moderate score (10-14) for 6 to 8 weeks despite an adequate trial. Usually the signal to consider augmentation or rotation. - Disconnects. The score does not match the clinical impression. The patient’s function can improve while reported symptoms remain. Reported symptoms can improve while function remains impaired. Discuss the disconnect during the visit. ## What MBC reporting looks like Aggregate reports of score trends can include: - Number of patients screened in the quarter. - Distribution of baseline scores. - Percentage of patients with at least one follow-up score within 90 days of baseline. - Distribution of change from baseline among patients with follow-up scores. The reports want the data, not narrative. Structured scores come out of the warehouse. Free-text HPIs become a manual chart review. ## A note on patient experience Some patients find the screeners mechanical. The mitigation is to talk about the score in the room, briefly. “Your PHQ-9 is 14 today, up from 9 last visit. That tracks with what you described about sleep and energy. Let’s talk about whether we adjust the dose.” The screener becomes part of the conversation. Adherence and accuracy both improve. See it on your workflow ## Twenty minutes, one mock visit. You leave with a note in your template. We run a mock session live, draft the note, and walk through what the downstream claim would look like. No slides. No sales deck. [Book a demo](https://nextvisit.cal.com/ryan/demo) [Talk to sales](mailto:hello@nextvisit.ai) Live in 2 weeks or less BAA signed by default