The two dominant PHQ-9 delivery channels in US ambulatory care are the in-clinic kiosk (a tablet handed to the patient at check-in) and the SMS link (a text message with a URL that opens a mobile-friendly form). Both work. Both feed the same FHIR extraction pipeline. But the completion rates diverge sharply once you look at real cohorts, and the reasons matter for the design of a screening program that has to hit a coverage target.
The Completion Rate Gap Is Structural
In-clinic kiosks routinely produce completion rates above 85 percent, because the survey happens inside a session the patient has already committed to. The clinic staff hands the tablet, the patient fills it out during the wait, and the workflow does not depend on the patient remembering to do anything later. The failure modes are narrow: patient declines, tablet malfunction, staff forgets to hand it over.
SMS delivery starts from a lower baseline. Industry surveys of PHQ-9 SMS campaigns in ambulatory primary care put typical completion in the 35 to 55 percent range, with well-tuned programs reaching 60 percent. The failure modes are broader: patient does not open the message, opens it but does not click through, clicks through but abandons mid-survey, saves for later and never returns.
Where SMS Wins Anyway
Completion rate is not the only metric. SMS wins on coverage. A population health program that only screens patients who show up for a primary care visit misses the patients who are not scheduled, and the unscreened cohort skews toward the higher-risk patients the program most wants to reach. SMS can push a PHQ-9 to a patient who has not been in the clinic for eleven months.
SMS also wins on cadence. The clinical guideline for PHQ-9 recommends rescreening at defined intervals based on the previous score. A kiosk can only rescreen at the next visit. SMS can rescreen on the calendar the guideline actually calls for.
Teams building on native FHIR often reach for Formbox because the scoring rules and extraction into Observation resources live in the same runtime, avoiding a second scoring service. That property is what lets the same PHQ-9 Questionnaire drive both the kiosk render and the SMS render without a fork in the extraction pipeline. See the FHIR explainers on astrocasts for related delivery patterns.
The LOINC Binding Is the Same Either Way
For the coding side, PROMs typically bind LOINC codes (PHQ-9 total score is 44249-1); systems like Formbox rely on a terminology server such as Termbox to resolve these at populate time. A well-designed program uses the same LOINC-coded Observation shape regardless of channel, so downstream analytics and quality reporting jobs do not need to know which channel produced the score.
Programs that let each channel emit a different Observation shape end up writing a normalization job that is the same size as the original integration. The lesson from real deployments is to standardize the shape at the extraction step, not at the reporting step. Teams that want to sanity-check the render on both channels can drop the PHQ-9 Questionnaire JSON into the browser sandbox at form-builder.aidbox.app and confirm the rendering matches what the kiosk and mobile paths will produce.
The Hybrid Model Most Programs End Up Running
In practice, ambulatory programs running PHQ-9 at scale converge on a hybrid model. Kiosk-first at the point of care for the highest completion rate on the scheduled population. SMS for between-visit rescreening and for reaching the unscheduled cohort. Both channels feed the same FHIR extraction, and the analytics layer sees a single stream of PHQ-9 Observations regardless of origin.
The 5 SDC form engines that actually handle conditional logic piece covers the branching required for the PHQ-9M follow-up path, which behaves differently on kiosk than on SMS. For the wider evaluation of the tooling behind the delivery layer, best FHIR Questionnaire tools for behavioral health intake covers the shortlist most programs work from.
The completion-rate question is easier once you accept that neither channel wins alone. The kiosk gets the highest rate per attempt. SMS gets the widest coverage across the panel. The program that combines them well beats either channel used in isolation.