{"resourceId":"imagetrend-st-johns-fire-rescue-ai-assist-case-study-202609","versions":[{"version":"external-2211ee8029a3c39eefd8b0a7aa687d7ac7a7c34ea7cfd323e17c3a2f08ee3ae9","resource":{"id":"imagetrend-st-johns-fire-rescue-ai-assist-case-study-202609","title":"ImageTrend case study: St. Johns County Fire Rescue uses AI Assist on 85% of fire and EMS reports, 422 rejections in 20,000+ completions","organization":"St. Johns County Fire Rescue (Florida) / ImageTrend","sector":"County fire rescue and EMS / patient care documentation","geography":"United States (Florida)","publishedAt":"September 2026","publicationDate":null,"eventDate":null,"sourceName":"ImageTrend","sourceLabel":"Vendor customer case study PDF","sourceUrl":"https://www.imagetrend.com/wp-content/uploads/2026/09/2026-St_Johns_Fire-Case-Study-1.pdf","evidenceClass":"vendor-claim","outcomeClass":"emerging","topics":["knowledge-work","accessibility-workforce","data-security"],"finding":"ImageTrend's case study (marked published September 2026) says St. Johns County Fire Rescue, with nearly 500 field personnel and about 67,000 incidents a year, uses ImageTrend Elite AI Assist (voice-to-text, image scanning of medication lists and insurance cards, auto-population) on 85% of fire and EMS reports, about 2.9 uses per incident. AI Assist populated 472,948 fields (44.7 per incident); more than 20,000 completions were applied with 422 rejected. Shorter completion time, less QI review and faster billing are stated without figures. A CQI Check feature is next.","sledRelevance":"One of the few fire/EMS AI documentation deployments reporting usage and acceptance counts at agency scale, relevant to any department weighing AI-assisted ePCR.","evidence":"Vendor case study with agency quotes. Counts (fields populated, completions, rejections) are system telemetry; time savings, accuracy and billing effects are asserted without baselines or numbers. Measurement period not stated.","architectureImplications":"AI features embedded in the existing ePCR platform: speech-to-text narratives, OCR of IDs, insurance cards and medication lists, and automatic field population.","governanceImplications":"A low rejection count measures crew acceptance, not correctness; agencies should audit a sample of accepted AI-filled fields against source documents.","securityPrivacyImplications":"Image scanning captures IDs, insurance details and medication lists at the bedside; PHI handling and model processing location are not described.","caveats":"Vendor-authored; month-level date only; no stated timeframe for the totals; no independent accuracy audit; 'rejected completions' undefined.","streamIds":["emergency-services"],"roles":{"sales":"Frame St. Johns as scale evidence from inside an existing ePCR: AI Assist on 85% of reports and roughly 473,000 fields populated across about 67,000 incidents a year. Use it to open a conversation about documentation burden and QI rework, but keep the value claim bounded: time savings and accuracy gains are asserted without numbers, so any engagement should measure them locally.","engineering":"The architecture is voice, OCR and text analysis inside the ePCR, auto-filling fields that crews accept or reject. Engineers should ask how rejections are defined, whether edits after acceptance are logged, where audio and images are processed, and how AI-populated fields are flagged for QA. Field-level audit against source documents is the key validation the case study does not provide.","delivery":"Embed AI Assist in go-live training as St. Johns did, but add what is missing: a baseline of report completion time and QI rework, a monthly audit of accepted AI fields, and a rule for PHI captured by image. Use the rejection rate as an adoption signal only, and judge success on accuracy and time measured before and after. Plan for CQI Check as a separate change with its own review."},"roleDetails":{"sales":{"customerProblem":"EMS crews split attention between patients and screens, and manual entry errors create billing and QI rework.","stakeholders":["EMS deputy chief","QI/QA coordinator","Billing office","Field crews","Privacy officer"],"discoveryQuestions":["How long do reports take today?","How much QI rework comes from data entry errors?","Do you already have AI features in your ePCR?"],"valueHypothesis":"Embedded AI capture may cut manual entry and rework if accuracy holds on audit.","potentialEngagement":"Baseline and field-level accuracy audit for AI-assisted ePCR.","unsupportedClaims":["Specific time savings","Accuracy improvement percentages","Billing revenue gains"]},"engineering":{"fit":"AI-assisted ePCR data capture for fire-based EMS.","architectureIntegration":"Voice, OCR and text analysis inside ImageTrend Elite auto-populating report fields.","prerequisites":["ePCR with AI features enabled","Device cameras and microphones","QA workflow"],"constraints":["Vendor telemetry only","No timeframe stated","Rejection metric undefined"],"security":"Confirm where scanned IDs, insurance cards and audio are processed and retained.","proofOfValue":["Field-level accuracy on audited sample","Report completion time vs baseline","QI rework rate"]},"delivery":{"work":"Train at go-live, baseline metrics, audit accepted fields, review PHI capture.","dependencies":["ImageTrend","Billing","QI staff"],"ownership":"Fire rescue EMS operations with QI.","skillsAdoption":"Crews learn scanning and dictation workflows and to check auto-filled fields.","governanceCheckpoints":["Go-live training sign-off","Monthly accuracy audit","Annual PHI review"],"acceptanceCriteria":["Audited AI-field accuracy reported monthly","Completion time measured against baseline"],"risks":["Accepted but wrong fields","PHI over-capture","Adoption counted as success"]}},"retrievedAt":"2026-10-05T22:06:27-05:00","enrichedAt":"2026-10-05T22:06:27-05:00","enrichmentBasis":"retrieved source","accessibilityWorkforceImplications":"Deputy chief says crews spend less time on screens and more on patients; providers adapted the tool to their own workflows.","procurementImplications":"Comes as a feature of an incumbent ePCR platform, so adoption may bypass a separate AI procurement review; contract terms on AI processing should be checked.","operatingModelImplications":"Documentation shifts toward real time, including before leaving the hospital; QA moves toward pre-submission checks.","sourceVerification":{"responseId":"curl-imagetrend-stjohns-2026-10-05T22:06:27-05:00","openedUrl":"https://www.imagetrend.com/wp-content/uploads/2026/09/2026-St_Johns_Fire-Case-Study-1.pdf","referenceExcerpt":"More than 20,000 completions applied with only 422 rejected completions.","promptVersion":"sled-research-v4.0","model":null,"basis":"agent-reported inspection"}}}]}