PUENTE

Platform for Unified Antibiotic Stewardship and Microbiology-enabled Evaluation

A mobile, cloud-based clinical platform supporting antibiotic stewardship and microbiology integration at the bedside. To be further developed between the Bernhard Nocht Institute for Tropical Medicine (BNITM) in Hamburg and the Instituto de Medicina Tropical (IMT) at Universidad Central de Venezuela, with a clinical pilot at Hospital Universitario de Caracas (HUC).

Status: Proof-of-concept pilot

---

The problem

In hospitals with limited diagnostic infrastructure, antibiotics are started empirically — before culture results are available. In most settings, there is no systematic process to revisit those decisions once results arrive. Broad-spectrum and Reserve-category antibiotics continue by default, not by deliberate choice. Laboratory results are issued on paper and rarely reach the prescriber at the right moment. Stewardship reviews occur sporadically and lack structured documentation. The result is accumulating antimicrobial resistance, avoidable drug toxicity, and no usable data to measure improvement.

PUENTE addresses this gap at the level of the individual patient encounter: every antibiotic, every ward round, every result, every decision — documented, time-stamped, and exportable.

---

How it works

The platform follows patients from admission to discharge through four structured steps:

1. Admission — the patient is registered and receives a unique QR code sticker
2. QR identification — any clinical team member scans the sticker to open the patient record instantly
3. AMS review — at every ward round, each antibiotic is reviewed using a structured 6-option decision set, linked to culture results and biomarker trends
4. Discharge with data — outcome, antibiotic duration, and stewardship decisions are recorded for audit and research

---

Who uses PUENTE

PUENTE is a single shared platform used by four clinical roles. Each role interacts with the same patient record but focuses on different parts of it.

Role	Primary use	Key screens
Attending physician / resident	Prescribing, reviewing antibiotic decisions at ward round	Dashboard, patient record, prescribing form
AMS team	Overseeing all antibiotics on the ward, triggering reviews, monitoring outcomes	Dashboard (flagged patients), AMS action panel
Clinical pharmacist	Checking prescribing appropriateness, AWaRe classification, dose and route	Prescribing form, AMS action record
Microbiologist / lab technician	Entering staged culture results linked to patient QR code	Culture results pipeline

---

Feature 1 — Ward dashboard and QR patient identification

The ward dashboard is the entry point for all clinical roles. It shows a live count of four key indicators — active patients, AMS reviews due, cultures pending, and results awaiting action — updated each time the screen loads.

Patients are listed with colour-coded status dots:

• 🟢 Green — stable, AMS review on track
• 🟡 Yellow — AMS review due or overdue
• 🟣 Purple — culture pending, no result yet
• 🔴 Red — critical alert (susceptibility result or Reserve antibiotic without documented review)

Each patient row shows name, age, sex, ward, bed, WHO AWaRe category of the current antibiotic, and treatment day. A ward filter bar restricts the view to a single ward or the ICU.

Each admitted patient receives a unique QR sticker (PNT-YYYY-NNNNN) placed on the wristband, bed card, and paper chart. Scanning it instantly opens the patient record — no manual search, no risk of selecting the wrong patient.

Used by: All roles at the start of each ward round. The AMS team filters for "AMS review due" to prioritise daily work. The pharmacist scans for Reserve-category antibiotics without documented justification.

---

Feature 2 — Patient record and antibiotic stewardship decision-making

The patient record opens immediately after QR scan. It displays a colour-coded identity chip, an alert banner when a culture result is actionable, and each antibiotic as a card showing drug name, dose, route, frequency, treatment day, clinical syndrome, and WHO AWaRe classification.

Below each antibiotic card are six structured AMS action buttons, colour-coded by clinical intent:

Action	Clinical meaning
Continue (sage)	Current antibiotic is appropriate — no change
De-escalate (lavender)	Switch to a narrower-spectrum agent
Escalate (gold)	Broaden coverage in response to clinical deterioration
Stop (red)	Discontinue — infection resolved or diagnosis changed
IV → Oral (coral)	Switch to oral route — patient stable, oral absorption adequate
Other (white)	Documented reason required

Selecting any action opens a confirmation sheet with a structured reason picker and a free-text note field. Every decision is recorded with a timestamp and the user's identity.

The prescribing form requires three fields before the antibiotic can be selected: working diagnosis (structured list of infectious syndromes), culture status (Yes / No / Not applicable), and antibiotic selection grouped by WHO AWaRe category. Selecting a Watch or Reserve drug immediately shows an explanatory banner.

Used by:

• Doctor / resident — selects AMS action at ward round, records clinical reasoning
• AMS team — reviews decisions made, audits undocumented Reserve antibiotic courses
• Pharmacist — validates dose, route, and AWaRe category

---

Feature 3 — Microbiology integration

Culture collection is initiated from within the patient record. The form enforces four quality checks before submission:

1. Sample collected from peripheral vein — not from an existing IV line
2. Skin disinfection applied twice, with 60-second drying time, no touching after
3. Correct volume collected (8–10 mL per bottle, aerobic bottle filled first)
4. QR sticker applied to the sample tube — this links the physical sample to the digital patient record; without it, results cannot be automatically attributed

Results are reported in four stages, each entered by the laboratory as it becomes available:

Stage	Laboratory event	Colour
1 — Received	Sample logged, accession number assigned	Green
2 — Gram stain	Gram reaction, morphology, and growth density reported	Green
3 — Isolate identified	Organism named, resistance phenotype noted	Red if MDR, Green if susceptible
4 — Susceptibility	Full minimum inhibitory concentrations (MICs) reported	Red if resistant

When stage 4 is filed, an alert fires on the prescribing clinician's dashboard with a direct link to the antibiotic review panel.

Used by:

• Microbiologist — enters each stage as results become available; turnaround time tracked automatically
• AMS team — monitors culture positivity, MDR rates, and time-to-result across all active patients
• Doctor — alerted when results arrive; can act directly from the notification

---

Feature 4 — Biomarker monitoring and stewardship analytics

CRP and procalcitonin (PCT) results are logged per patient in quantitative or semiquantitative format, each labelled with a clinical interpretation — Improving, Stable, Worsening, or New result — colour-coded on a vertical timeline. A mini sparkline visualises the trend across the admission.

The analytics screen aggregates all admissions into programme-level key performance indicators:

Indicator	Description
AWaRe prescribing ratio	Proportion of antibiotic courses by category — WHO target ≥60% Access
AMS reviews completed	Percentage of antibiotics reviewed within the target window
IV-to-oral conversions	Number of patients switched from intravenous to oral therapy per month
Median antibiotic duration	Median treatment days compared to pre-intervention baseline
Pre-antibiotic culture coverage	Percentage of prescriptions where a culture was ordered before the first dose
Reserve antibiotic justification	Percentage of Reserve-category courses with a documented clinical indication

All records export to CSV for analysis in standard statistical software.

⚠️ The analytics module is part of the research design of the pilot phase. It is not a commercially licensed reporting tool.

Used by:

• AMS programme coordinator — monthly review of stewardship targets
• Research team (BNITM / IMT) — extraction of pilot data for analysis

---

Feature 5 — WHO GLASS-ready data export

Generating data for national and global AMR surveillance is a recognised bottleneck for hospitals in low- and middle-income countries. It typically requires a dedicated data manager, manual data extraction, and familiarity with WHO GLASS submission formats — a process that takes days and is prone to transcription error. PUENTE eliminates this burden. Because every culture result, susceptibility interpretation, and antibiotic course is structured and coded at the point of care, the data required for WHO GLASS submission is already complete at the end of each reporting period. A single export screen produces a validated, submission-ready dataset covering both surveillance modules:

• GLASS-AMR (antimicrobial resistance): organism, specimen type, S/I/R interpretation, collection date, ward, patient age group and sex — all captured as a byproduct of routine microbiology result entry.

• GLASS-AMU (antimicrobial use): antibiotic name, ATC code, defined daily doses (DDD), route of administration, AWaRe category, and duration — captured at prescription and updated on discharge.

The export screen validates completeness before submission, flags missing fields (such as MIC values where only S/I/R interpretations were entered), and produces a CSV file formatted for direct import into WHONET — the WHO-recommended software for GLASS data preparation and submission.

This positions HUC as a potential sentinel surveillance site contributing to Venezuela's national AMR data and to the WHO GLASS global dataset — a research output that extends beyond the pilot's clinical objectives.

Used by: AMS programme coordinator · research team (BNITM / IMT) · national AMR focal point

---

Pilot site

The EKFS pilot will run across two inpatient wards (adult medical and paediatrics) and the ICU at Hospital Universitario de Caracas (HUC). Microbiology specimens are processed at the IMT laboratory at Universidad Central de Venezuela.

---

Team

Name	Institution	Role
Ricardo Strauss	BNITM Hamburg	Principal investigator, platform development
Elizabeth Hernandez	HUC / Universidad Central de Venezuela	Clinical lead
Belkisyolé Alarcón de Noya	IMT / Universidad Central de Venezuela	Microbiology lead
Heiman Wertheim	Radboud UMC / SPICE Program	External scientific advisor
Patricia Valenzuela	Sociedad Venezolana de Infectología	National infectious disease network liaison

---

Ongoing Project BNITM & IMT-UCV

PUENTE builds on SABER-SDG3 (Strengthening the Americas Basis to Reach the SDG3), a collaboration funded by the German Academic Exchange Service (DAAD) from 2025 to 2028 between BNITM and IMT. SABER-SDG3 provides proof of operational stability and institutional trust between the two sites, independent of PUENTE's funding trajectory.

---

Contact

Ricardo Strauss
Bernhard Nocht Institute for Tropical Medicine
Hamburg, Germany
ricardo.strauss@bnitm.de

Data privacy and security

In the production deployment, PUENTE runs on a server physically located within the HUC hospital network. Access is restricted by IP address: the application responds only to devices connecting from within HUC's internal network range. Access from outside the hospital — including personal devices on external networks — is technically blocked at the server level. Staff requiring remote access connect through the hospital VPN, which assigns them a hospital IP address before the application becomes reachable. No patient data is transmitted to external servers. All records remain within the hospital's own infrastructure. Every action in the app — prescribing decisions, AMS reviews, culture requests — is time-stamped and attributed to the authenticated user, generating a complete and tamper-evident audit trail. The current proof-of-concept demo contains exclusively fictional patient data and no personal information of any kind. Data handling for the research component complies with GDPR (Regulation EU 2016/679) on the BNITM side and with Venezuela's constitutional right to personal data protection (Article 60, Constitution of the Bolivarian Republic of Venezuela, 1999) on the HUC/IMT side. A formal data management plan and ethics approval will be obtained prior to pilot initiation.