The situation
Every Author Health appointment opened the same way: a Medical Assistant spending real time with the patient before the clinician ever joined the call, walking them through required clinical questionnaires — depression and anxiety screeners, and for many patients, additional condition-specific assessments. For a new-patient Intake visit, that added up to roughly 30 minutes of MA time before a provider even said hello.
None of this work was optional — these assessments feed directly into treatment decisions and quality reporting. But there was nothing about them that required a staff member to be the one asking the questions. The bottleneck wasn't the assessments themselves; it was where and how they were happening.
Designing a self-service path — without losing rigor
I led this initiative end to end: defining the approach, writing the requirements, and building the metrics framework we'd use to know if it actually worked. The plan had two connected parts, built in phases rather than as one large release:
-
Rework the cadence
Standard questionnaires had been coming due every 30 days regardless of clinical need. Stretching that to 90 days — while preserving a provider's ability to trigger one anytime — cut how often patients needed to engage with these at all, before touching how they were delivered.
-
Move it into the portal
Assessments became something a patient could complete asynchronously before their appointment, gated behind identity verification and consent-to-treat, with real safeguards built in — including a specific handling path for any answer indicating suicidality on the depression screener.
-
Add a faster cognitive screener
For patients 55 and older, we introduced a new self-administered cognitive assessment as a patient-facing alternative to the traditional in-visit cognitive test — a test that, by its nature, can only ever be done by the patient themselves.
-
Keep staff informed, not blocked
None of this could work if providers walked into a visit unsure what was already done. We added visibility directly into the clinical workflow, so staff could see at a glance what a patient had already completed before the appointment even started.
What changed
Since launch, the shift has been real and sustained — not a one-time spike, but a consistent pattern month over month:
Looking at it from the other direction confirms the same story: of patients who completed their questionnaires independently in a recent month, only about 3% still needed a staff member's help to get through it — meaning the self-service path isn't just technically available, it's genuinely usable without assistance for the vast majority of patients using it.
That's real MA time given back on nearly half of all standard assessments, and for the cognitive screener, a workflow that literally could not have existed in its old form — the self-administered version doesn't just save staff time, it opens up a faster, more sensitive screening option for patients who might otherwise have gone without one in a time-constrained visit.
Figures reflect internal product metrics current as of mid-2026, generalized for public sharing.