Score the PHQ-9 and GAD-7 separately. The PHQ-9 total runs from 0 to 27; the GAD-7 total runs from 0 to 21. Both cover the past two weeks, but adding the totals creates a number with no supported meaning. Neither questionnaire provides a diagnosis or treatment plan.
The two scoring systems
The original PHQ-9 adult validation study supports its five bands.
PHQ-9: nine answers, each worth 0 to 3 points.
| Score | Severity |
|---|---|
| 0-4 | None or minimal |
| 5-9 | Mild |
| 10-14 | Moderate |
| 15-19 | Moderately severe |
| 20-27 | Severe |
The original GAD-7 adult primary-care study supports its four bands.
GAD-7: seven answers, each worth 0 to 3 points.
| Score | Severity |
|---|---|
| 0-4 | Minimal |
| 5-9 | Mild |
| 10-14 | Moderate |
| 15-21 | Severe |
For each questionnaire, add the item values. There is no weighting or reverse scoring in the native Survey Doctor definitions. Every scored item is required.
Simple arithmetic.
Never a combined score.
The bands describe reported symptom frequency. They are not care instructions, and a low band does not mean that no care is needed. A full assessment can identify concerns that either questionnaire misses.
Context still matters.
For a closer explanation of either scale, use the PHQ-9 score guide or GAD-7 score guide.
Read the pair without inventing a combined result
The two scores answer different questions. Look at the pattern. The PHQ-9 covers depression symptoms, including mood, interest, sleep, energy, appetite, concentration, movement, and thoughts of death or self-harm. The GAD-7 focuses on worry, tension, restlessness, irritability, and fear.
When both scores are high, the pair shows that both sets of symptoms were reported often. It does not establish two diagnoses or prove that one condition caused the other.
When one score is higher, inspect the individual answers. The gap may help organize a conversation, but it does not determine treatment emphasis. Functioning, history, medical factors, other symptoms, and the person's priorities still matter.
Do not stop there.
When both scores are low, do not dismiss a concern that brought someone to care. Panic, trauma reactions, grief, mania, substance effects, physical illness, and specific fears may not fit these short questionnaires. Ask anyway.
Item 9 of the PHQ-9 stands alone
Only the PHQ-9 asks about thoughts of death or self-harm. Any answer above zero on item 9 needs direct follow-up, regardless of either total. The PHQ-9 is not a complete suicide risk assessment, and the GAD-7 has no safety item that can replace one.
Do not average it away.
If you need help now. In the US, Call 988 or text 988. Call 911 if you are in immediate danger. Outside the US, contact your local emergency number or find support in your country.
Functional impact stays separate from the scores
Published PHQ-9 materials can include a separate functional-difficulty question. Some GAD-7 materials also append a difficulty question. Neither answer contributes to the scored total.
Survey Doctor shows each optional companion when any scored item on that questionnaire is above zero. It stores the selected response separately and never adds it to the PHQ-9 or GAD-7 total.
An all-zero result does not show the question. Historical, manual, and imported rows without companion data mean it was not collected. A combined review should still ask directly about work, home responsibilities, relationships, and other daily activities.
Know what a score-only record omits
Manual entry accepts one whole-number total in the native range for each questionnaire. It cannot establish the item answers, prove that all items were completed, recover PHQ-9 item 9 status, or show whether the functional question appeared.
That boundary matters.
Some imported historical summaries also lack enough evidence to confirm their form, completion, or scoring convention. Survey Doctor keeps incompatible summaries outside current bands, trends, and clinical comparisons. A number that happens to fall within 0 to 27 or 0 to 21 is not enough to prove that it is a current native result.
Comparing the two scores over time
Repeated native results can show whether the reported symptom pattern is changing. There is no universal point change, percentage change, remission threshold, or retest interval that applies to both questionnaires in every setting.
Compare like with like.
Keep each history separate. Compare complete results from the same form under similar conditions. Review item-level changes and functional impact instead of relying only on a band crossing.
A one-point move can cross a boundary. That does not prove a sudden clinical change. Likewise, an unchanged total may hide important movement among individual symptoms.
The bottom line
Add each questionnaire's answers, keep the totals separate, and treat the bands as descriptions. Then add the missing context: functioning, safety, history, and the person's own experience.
For more about their different purposes, see PHQ-9 vs GAD-7.