It is 9:40 on a clinic morning. The front desk hands a new patient a clipboard with a PHQ-9, the patient marks nine checkboxes with a pen, and a medical assistant scores it in the hallway between rooms: zero, plus two, plus one, plus three, carry on down the column. The total gets typed into the chart, the form goes into a folder, and the next patient is already waiting. Multiply that by every intake, every follow-up, and every annual screening visit, and you have a scoring pipeline that runs entirely on mental arithmetic performed under time pressure.
The paper part is not the problem. Validated screeners like the PHQ-9 and GAD-7 work well on a printed page, and for a waiting room full of patients with no portal logins and no clinic tablets, paper is often the only administration mode that reaches everyone. The fragile part is everything that happens after the pen: adding up item scores by hand and retyping the total somewhere useful. A patient whose true score is 10 but gets recorded as 9 lands in a different severity band, and possibly a different care pathway, because of a slip in hallway addition.
PaperSurvey.io keeps the administration exactly as it is, a printed form and a pen, and replaces the scoring. You assign each response option its 0 to 3 weight once, scan the completed forms, and every screener arrives as a scored record: per-item values, an automatic total, and flags you can filter, ready to check against the scanned form before anyone acts on it.
Why validated screeners stay on paper
Plenty of clinics have tried moving screening to tablets or patient portals and quietly moved back. The reasons are practical rather than nostalgic.
- The waiting room is a paper environment: a clipboard needs no login, no charge, no Wi-Fi, and no cleaning protocol between patients. It works for every patient who walks in, including those who would never complete a portal questionnaire before the visit.
- Instrument fidelity is non-negotiable: the PHQ-9 and GAD-7 are validated with specific item wording and the same four response anchors, from "Not at all" to "Nearly every day." A printed form preserves that layout exactly, every time, on any office printer.
- Paper does not compromise the measurement: a meta-analytic review of patient-reported outcome measures found paper and electronic administration give equivalent scores, with an average weighted correlation of 0.90 across 207 coefficients (Gwaltney, Shields & Shiffman, 2008). The mode you can actually administer in a waiting room measures just as well as a screen.
The trade-off clinics accepted in exchange was hand scoring. That is the part that no longer has to be manual.
What a PHQ-9 or GAD-7 score is made of
Both instruments are weighted sums, which is exactly what makes them so automatable. The PHQ-9 asks nine questions about the last two weeks, each answered on the same scale: Not at all scores 0, Several days scores 1, More than half the days scores 2, and Nearly every day scores 3. The total runs from 0 to 27, and the conventional severity bands are minimal (0 to 4), mild (5 to 9), moderate (10 to 14), moderately severe (15 to 19), and severe (20 to 27). A tenth question about how difficult these problems have made daily life is recorded but not scored.
The GAD-7 works the same way with seven anxiety items on the identical 0 to 3 scale, for a total of 0 to 21, with cut points at 5, 10, and 15 for mild, moderate, and severe anxiety. In both cases the score is a screening aid for a clinician, not a diagnosis, but it routes real decisions: whether to assess further, whether to follow up sooner, whether treatment is working between visits.
One item deserves special mention. Item 9 of the PHQ-9 asks about thoughts of self-harm, and most protocols require attention to any non-zero answer regardless of the total score. A workflow that only records the sum can lose that signal entirely, which is an argument for capturing every item, not just the total.
Where hand scoring and retyping go wrong
Scoring a single form is easy. Scoring the forty-third form of the day, between rooming patients, is where errors creep in.
- Arithmetic under pressure: adding nine small numbers sounds trivial until it happens hundreds of times a week. Columns get mixed up, a 2 is read as a 3, a skipped item is silently treated as zero by one staff member and left ambiguous by another.
- A second manual step: the total then has to be typed into an EHR field, a spreadsheet, or a study database. Every retyped number is another chance for a transposition, and the per-item answers usually never make it into the record at all.
- Cut points make small errors expensive: a one-point slip in the middle of the scale is invisible. The same slip at 9 versus 10, or 14 versus 15, changes the severity category and the recommended action.
- The item 9 flag disappears: when only totals are recorded, a patient scoring low overall but non-zero on the self-harm item looks unremarkable in the data.
Research and population screening programs feel this even harder than clinics do. A school district screening day or a longitudinal study can produce thousands of forms in a week, and hand scoring at that volume means either a large temp-staffing bill or a long delay before anyone sees results. Our breakdown of the cost of manual survey data entry puts numbers on that trade.
Scoring PHQ-9 and GAD-7 automatically in PaperSurvey
In PaperSurvey, a screener is built as a single-choice grid: nine rows, four columns, one mark per row. Each response option carries an answer weight, so "Not at all" is worth 0 and "Nearly every day" is worth 3, exactly as the instrument specifies. You set those weights once when you set up the survey. From then on, every scanned form is scored the same way, every time, with the per-item value and the automatic total stored together on each response.
The scanning side is ordinary office equipment. Generate the screener in PaperSurvey, print it on plain paper, collect the completed forms, and scan the stack with a multifunction scanner or a phone's document scanner. Optical mark recognition reads the checkboxes, and validation flags problems such as two answers marked in one row. Check each answer used for clinical scoring against the original form before relying on the total; an accepted reading is not proof that the mark was read correctly. PaperSurvey cannot process completed screeners printed from another platform.
The PHQ-9 Depression Questionnaire template below is ready to print, or open it in the editor and make it yours:
View the full PHQ-9 Depression Questionnaire template. A matching GAD-7 Anxiety Questionnaire template is in the gallery too, built the same way, so an anxiety and depression screening pair takes minutes to set up rather than an afternoon.
From the clipboard to a scored dataset
Once forms are scanned, the data behaves like data instead of like a filing cabinet.
- Every item, not just the total: exports include the 0 to 3 value for each question alongside the computed score, so item-level analysis, item 9 review, and re-verification against the original scan are all possible.
- Flags you can filter: filter responses where the self-harm item is above zero, or where the total crosses a cut point, and review those first. The original scanned page sits behind each response for context.
- Exports for clinic and research alike: send results to Excel or CSV for the practice manager, or straight to SPSS, R, Stata, or SAS for a research team, with weights already applied.
- Paper and web in one dataset: the same screener runs as an identical web survey for telehealth intakes or remote follow-ups, and web and paper responses land in a single combined dataset with the same automatic scoring, so nobody merges two spreadsheets at the end of the month.
The practical effect is that a stack of morning screeners becomes a scored, filterable dataset before lunch, and a research batch of two thousand forms becomes analyzable in days instead of months.
Keeping screening data confidential
Depression and anxiety screening data is among the most sensitive information a survey platform can hold, and it should be handled accordingly. PaperSurvey stores data in the EU under GDPR and runs on hosting infrastructure that is ISO 27001 certified and SOC 2 Type II audited. A data processing agreement is available on request, and SAML SSO on Enterprise Plus ties access to your organization's identity provider, which matters when screening data must only be visible to authorized clinical or research staff.
The Bottom Line
Paper is the right administration mode for a great deal of mental health screening: it reaches every patient in the waiting room, it preserves the validated instrument exactly, and the evidence says it measures as well as a screen. What paper never deserved was the hand scoring that came bundled with it. Setting the 0 to 3 answer weights once and scanning the stack gives you the same printed PHQ-9 and GAD-7 your clinic already trusts, with totals, severity bands, and item-level flags computed automatically and exported wherever your workflow lives. Start your free trial and score this week's screeners without touching a calculator.