Executive Summary
GI procedures create substantial clinical and financial value only when patients arrive on time, prepared correctly, and medically ready to proceed. Inadequate bowel cleansing, missed fasting windows, medication errors, transportation barriers, and late cancellations can leave endoscopy capacity unused and expose patients to repeat procedures, delayed diagnoses, and avoidable safety events.
Florence addresses these failure points through automated, two-way SMS engagement beginning up to two to three weeks before a procedure. Rather than functioning as a simple reminder service, Florence delivers procedure-specific education, confirms critical preparation steps, interprets patient replies, and escalates the patients who need human assistance while there is still time to protect the case.
In an initial hospital-system pilot, Florence-supported GI preparation was associated with a 46% reduction in the combined occurrence of no-shows and cancellations, without requiring additional staff. This preliminary operational result is consistent with a growing body of evidence showing that reinforced digital education and appropriately timed messaging can improve bowel-preparation quality, attendance, adherence, and patient satisfaction. [1–8]
For an integrated GI group, with approximately 5,000 annual outpatient colonoscopies, every one-percentage-point improvement in successful completion represents approximately 50 protected procedures – potentially over $100,000 in clinical and procedural income. A focused pilot can rapidly determine the effect on procedure completion, preparation quality, staff workload, and return on investment—and create a repeatable model for broader evaluation across a broader network.
The Opportunity: Turn Scheduled Volume into Completed Procedures
A colonoscopy or advanced endoscopy appointment remains operationally at risk until the patient completes every required step. Patients must obtain the preparation, follow diet restrictions, alter selected medications only as directed by the clinical team, arrange transportation, observe fasting windows, and take the preparation at the correct time.
The updated U.S. Multi-Society Task Force recommendations emphasize split-dose regimens, patient-centered selection of bowel-preparation regimens, and systematic monitoring of adequacy rates, with an endoscopy-unit target of at least 90% adequate preparation.[9] The challenge is that traditional written instructions and broad reminder calls provide limited visibility into whether each patient is actually progressing through the required steps.
This creates a preparation-visibility gap. A patient who cannot obtain the laxative, misunderstands the second-dose timing, continues solid food, or lacks transportation may not become visible until the procedure slot can no longer be saved. At the same time, staff may spend hours calling low-risk patients while higher-risk patients remain unreachable.
Why generic reminders are not enough
The evidence is nuanced. One large, randomized trial found that a series of automated texts did not improve attendance or preparation quality when layered onto an existing workflow that already included written instructions and a nurse call.[1] This finding argues against treating all messaging solutions as equivalent. Digital outreach is most useful when it is timed to specific actions, reinforces understanding, supports patient questions, and produces actionable escalation—not when it merely repeats an appointment date.
What the Medical Literature Shows Since 2021
| Evidence theme | Selected findings | Implication for Florence |
|---|---|---|
| Reinforced SMS education | A 2022 meta-analysis of seven randomized trials (5,889 patients) found higher adequate-prep rates with SMS (81.7% vs 75.7%) and lower nonattendance (RR 0.74).[2] | Supports timed, structured reinforcement rather than one-time instructions. |
| Interactive mobile messaging | Randomized trials using WhatsApp or WeChat reported improved preparation quality, diet and medication adherence, satisfaction, and anxiety.[3,4] | Two-way communication can address comprehension and behavioral barriers. |
| Timed SMS instruction | A 2024 blinded randomized trial reported adequate preparation in 84.3% of SMS patients versus 71.3% of controls, with higher compliance and satisfaction.[5] | Action-specific cues near the procedure window can improve execution. |
| Multichannel automation | An 833-patient multicenter trial combining automated calls, SMS, and web content improved adequate preparation from 54.5% to 60.7%.[6] | Automation can scale preparation support across large procedure populations. |
| AI-supported preparation | A 2025 randomized pilot found higher mean BBPS scores with an AI smartphone tool (6.78 vs 5.35); AI reminders were independently associated with adequate preparation.[7] | AI can personalize guidance, but SMS avoids app download and login barriers. |
| Operational efficiency | Digital preprocedure instructions have reduced no-shows and same-day cancellations and achieved high patient satisfaction in real-world implementation.[8] | Patient engagement can protect procedure capacity as well as clinical quality. |
AI voice versus SMS
Published GI-preparation evidence remains substantially deeper for SMS, mobile messaging, telephone reinforcement, and multichannel reminders than for conversational AI voice. Voice may help selected patients; however, calls can be screened as spam, arrive at inconvenient times, or require a patient to remain available for a synchronous interaction.
Florence uses the patient’s existing text inbox, permits asynchronous replies, preserves a written record of instructions, and requires no app, portal, password, or smartphone. The strategic advantage is not the channel alone: it is the combination of low-friction access, two-way engagement, procedure-specific logic, and timely escalation.
The Florence GI Procedure Preparation Model
Florence supports colonoscopy, EGD, capsule endoscopy, EGD Bravo, ERCP, and upper EUS through a single protocol engine. Engagement can begin two to three weeks before the procedure to identify medication, transportation, pharmacy, language, and social barriers, then intensify during the final week and final 24 hours.
| What Florence manages | What Florence surfaces |
|---|---|
| • Split-dose and same-day prep timing | • Failure to obtain or tolerate preparation |
| • Clear-liquid, low-residue, and NPO confirmation | • Incorrect dose timing or continued solid food |
| • Practice-approved medication checkpoints | • Medication uncertainty requiring clinical review |
| • Transportation and caregiver readiness | • Repeated nonresponse and likely no-show risk |
| • Bilingual, fifth-grade-level education | • Transportation or caregiver barriers |
| • Teach-back and plain-language patient replies | • Possible post-procedure complications |
| • Post-procedure symptom checks when appropriate | • Social barriers requiring staff intervention |
A care-team multiplier—not a replacement for clinical judgment
The Gastroenterologists and Endoscopy unit retains control of all clinical content, medication instructions, escalation thresholds, and routing. Florence operationalizes the practice’s approved workflow and directs staff attention toward the patients most likely to lose a procedure. The goal is to reduce repetitive outbound work while giving schedulers and nurses earlier, more actionable visibility.
What patients say about Florence’s GI prep support
I forgot to bring my printed instructions with me when I went to the pharmacy to pick up my prep. Fortunately, Flo had sent the instructions to my phone, so I could pull them up right there.
Flo reminded me when to start my prep and when it was time to take the second dose. That really helped because I had several things going on that evening and could easily have lost track of the time.
I forgot to tell the nurse that I take Ozempic. Flo reminded me to contact the clinic about it before my colonoscopy. The nurses helped me get everything straightened out, so I didn’t have to cancel. My colonoscopy went fine. Thank you, Flo!
Why Your GI Group and Endoscopy Unit Should Pilot Florence
The proposal is designed for speed and low implementation burden. A pilot can begin with one endoscopy center, one procedure type, or a defined physician cohort. Scheduled-patient data can be supplied through a routine file or agreed integration; your approved instructions are configured into the protocol; and Florence operates alongside the existing scheduling and clinical workflow.
The economic logic is straightforward. At 5,000 annual colonoscopies, a one-percentage-point improvement protects approximately 50 procedures and potentially $100,000 in revenue. Additional value may come from fewer repeat examinations, earlier cancellation notice, improved room utilization, reduced staff calling, and better patient experience. The pilot should use the endoscopy unit and gastroenterologists own contribution-margin and staffing assumptions to calculate a credible local ROI.
The larger strategic value is reproducibility. A successful pilot can generate a standardized implementation playbook, outcome definitions, and economic model suitable for evaluation by other practices and ambulatory endoscopy centers.
Recommended Rapid-Pilot Design
| Pilot element | Recommended approach |
|---|---|
| Scope | One endoscopy center or defined physician cohort; begin with outpatient colonoscopy and expand after proof of value. |
| Duration | Approximately 12–16 weeks, including baseline confirmation and active enrollment. |
| Comparison | Historical baseline, phased implementation, or concurrent control based on operational feasibility. |
| Primary endpoint | Combined no-show, late-cancellation, and inadequate-preparation rate, with each component also reported separately. |
| Quality measures | Adequate preparation (BBPS ≥6 and all segments ≥2), first-attempt completion, repeat procedure, cecal intubation, and ADR where sample size permits. |
| Operational measures | Cancellation lead time, staff outbound calls, escalations, protected procedures, rescheduling time, and estimated contribution margin preserved. |
| Engagement measures | Enrollment, response, prep-acquisition confirmation, split-dose adherence, medication-query escalation, transportation confirmation, and patient satisfaction. |
Conclusion
The literature supports digitally reinforced preparation, but it also shows that simple reminders alone are insufficient. High-performing programs reach patients at the right time, reinforce specific actions, identify barriers, and alert staff before the opportunity to intervene has passed. Florence brings these functions together through low-friction, two-way SMS.
Generated Health’s preliminary 46% reduction in no-shows, cancellations, and poor preparations suggests meaningful potential to improve procedure completion without adding staff. A focused pilot offers a fast, measurable way to validate that result in a high-volume community gastroenterology setting and determine whether Florence can become a scalable procedure-preparation infrastructure for you.
Every scheduled procedure is an opportunity to diagnose, treat, and generate value. Florence helps ensure that opportunity is not lost because the patient was left to prepare alone.
References
- 1. Mahmud N, et al. Effect of text messaging on bowel preparation and appointment attendance for outpatient colonoscopy: a randomized clinical trial. JAMA Network Open. 2021;4(1):e2034553. PMID: 33492374. Link
- 2. Reinforced education by short message service improves the quality of bowel preparation for colonoscopy: a systematic review and meta-analysis. 2022. PMID: 35192000. Link
- 3. Lam TYY, et al. Nurse-led reinforced education by mobile messenger improves the quality of bowel preparation of colonoscopy in a population-based colorectal cancer screening program: a randomized controlled trial. International Journal of Nursing Studies. 2022;133:104301. PMID: 35764027. Link
- 4. Wang C, et al. Effect of WeChat messaging on improving the quality of bowel preparation for outpatient colonoscopy: a randomized controlled study. Gastroenterology Nursing. 2022;45(2):120-126. PMID: 35283439. Link
- 5. Kılınç T, et al. The effect of informing patients who will undergo a colonoscopy via short messaging service on the procedure quality and satisfaction: an endoscopist-blinded, randomized controlled trial. Journal of PeriAnesthesia Nursing. 2024;39(3). PMID: 37999686. Link
- 6. Preoperative automatic reminder systems and impact on quality and compliance with colonoscopy preparation: a multicenter randomized controlled trial. 2023. PMID: 37610349. Link
- 7. Zhong H, et al. A clinical pilot trial of an artificial intelligence-driven smartphone application of bowel preparation for colonoscopy: a randomized clinical trial. Scandinavian Journal of Gastroenterology. 2025;60(1):116-121. PMID: 39709551. Link
- 8. Richter JM, et al. A digital preprocedure instruction program for outpatient colonoscopy. Telemedicine and e-Health. 2020;26(4):468-476. PMID: 31298628. Included as a foundational operational implementation study. Link
- 9. U.S. Multi-Society Task Force on Colorectal Cancer. Optimizing bowel preparation quality for colonoscopy: consensus recommendations. American College of Gastroenterology / American Gastroenterological Association / American Society for Gastrointestinal Endoscopy. 2025. Link
- 10. Comparative effectiveness of enhanced patient instructions for bowel preparation before colonoscopy: network meta-analysis of 23 randomized controlled trials. Journal of Medical Internet Research. 2021. PMID: 34694227. Link
- 11. Effectiveness of bowel preparation innovative technology instructions on clinical outcomes among patients undergoing colonoscopy: a systematic review and meta-analysis. 2023. PMID: 37402823. Link
- 12. Mobile health technologies supporting colonoscopy preparation: a systematic review and meta-analysis of randomized controlled trials. 2021. PMID: 33735320. Link
Notes and limitations
** The 46% reduction is a preliminary Generated Health pilot result supplied for this white paper. Based on reduction in no shows and cancellations from 2% to 1.2%. In this a cohort of 5,166 patients randomized to usual outreach or Flo assisted outreach (3,151), 2,307 patients engaged (73%) and had 2,206 completed endoscopy procedures. It has not yet been independently peer-reviewed and should be validated using prospectively defined outcomes in the proposed pilot.
The estimated annual colonoscopy volume and organizational description used for pilot planning were supplied by Generated Health. Final proposal materials should confirm current procedure volumes, locations, staffing, and baseline cancellation and inadequate-preparation rates directly with the gastroenterology group and endoscopy unit.
Florence supports practice-approved education, confirmation, and escalation workflows. Medication changes and clinical decisions remain the responsibility of the treating clinicians.