evidoria

← Back to browse

Good practice Imported

CLEO — Costa Rica's AI Assistant for Purging Surgical Waiting Lists

Costa Rica · San José · See the Costa Rica profile · See the San José profile

Evidence: Observational / pre–post Top 22% 73/100 · Ask Evidence Copilot about this practice

Costa Rica's public health service (CCSS) uses an AI assistant, CLEO, to contact surgical-waitlist patients via WhatsApp, phone and email, checking records against health and civil registries — resolving 367,403 cases since 2023, with mandatory human sign-off on removals.

367,403
Total waiting-list cases resolved since 2023 (2023 to Q1 2026)
136,774
Cases purged specifically as no longer needing surgery (2023 to Q1 2026)
31.2 %
National list-depuration rate, 2023 (2023)
18.2 %
National list-depuration rate, Q1 2026 (Q1 2026)
1,500 records/day
Maximum records processed per day
₡7,122 million (~US$14 million)
Budget allocated for the May–October 2026 targeted push (May–Oct 2026)
100,125
Target additional resolutions in the May–October 2026 push (May–Oct 2026)
15
Official resolution categories used by CLEO
CLEO — Costa Rica's AI Assistant for Purging Surgical Waiting Lists

Details

Maturity
Scaling
Promoter
Caja Costarricense de Seguro Social (CCSS)
Period
2023–2026 (ongoing; targeted expansion May–Oct 2026)
Keywords
healthcare, public administration, service delivery, generative AI

Context

Costa Rica's Caja Costarricense de Seguro Social (CCSS), the country's single public health insurer, has deployed CLEO, a virtual-assistant tool that cross-references the EDUS unified digital health record, the national Civil Registry, and provider networks to identify patients on surgical, procedure, and consultation waiting lists who may no longer need the service.

Objectives

CLEO aims to reduce backlogs on CCSS's waiting lists by identifying patients who no longer need a scheduled procedure and confirming this with a defined, auditable set of resolution categories rather than removing anyone unilaterally.

Activities

CLEO contacts flagged patients directly by WhatsApp, phone call, and email, applying one of 15 official resolution categories — such as 'no longer needs surgery', 'deceased', or 'already treated elsewhere' — to each case. Every AI-flagged removal requires human validation by CCSS staff before it is executed, so the system recommends but does not unilaterally decide. Deployed from 2023 and scaled through the first quarter of 2026, the system can process up to 1,500 records per day.

Results

CLEO has resolved 367,403 patients off surgical waiting lists, including 136,774 cases specifically purged as no longer needing surgery. Over that period, CCSS's national list-depuration rate fell from 31.2% in 2023 to 18.2% in Q1 2026. CCSS allocated ₡7,122 million (roughly US$14 million) for a further targeted push between May and October 2026, aiming at 100,125 additional resolutions.

Conclusions

The figures are self-reported by CCSS but corroborated consistently across four independent Costa Rican outlets over more than a year of coverage, which strengthens confidence in the topline numbers even though no independent technical audit of CLEO's accuracy or categorisation logic has been published.

Implementation

Indicative cost
Medium (€50k–€500k) — CCSS allocated ₡7,122 million (roughly US$14 million) for a targeted expansion push between May and October 2026 alone; no full platform build cost is published.
Time to results
Medium (1–3 years) — Deployed from 2023 and scaled through Q1 2026, with a further targeted expansion planned for May–October 2026.
Staffing & skills
CCSS staff who validate every AI-flagged removal before it is executed, CLEO virtual-assistant tool integrated with EDUS, the national Civil Registry, and provider networks

Conditions for success

  • integration with the EDUS unified digital health record and the national Civil Registry for identity and eligibility cross-referencing
  • 15 standardised, defined resolution categories to structure human decisions
  • mandatory human sign-off before any removal is executed
  • dedicated phased budget for expansion (₡7,122 million for the May–October 2026 push).

Common failure modes

  • figures are self-reported by CCSS, and no independent technical audit of CLEO's accuracy or categorisation logic has been published.

Where it fits

Governance type
national single-payer public health insurer
Scale
national
Income level
upper-middle-income

Commonly funded by

National / regional programmes

Indicative funding routes for practices of this type — always check each programme's current calls and eligibility rules.

Do you run this practice? Claim it — verified implementers get a public contact pathway and can propose corrections.

Data sources

Where this practice's information was retrieved from, and when.

Attachments

Similar practices you may find useful