Executive Summary
Bangladeshi hospitals are scaling rapidly but the clinical workforce is not. Bed counts at leading tertiary facilities have doubled over the last decade while nurse-to-patient ratios have remained stubbornly above 1:12 in general wards. SGT Systems Limited's IoMT — Connected Patient Monitoring service is engineered to amplify the reach of the clinical team: real-time vitals at every bed, intelligent alerts that reach the right clinician in seconds, asset tracking that ends the daily hunt for infusion pumps, and a single central station that turns ward-walking into ward-managing.
The service is targeted at hospital administrators, medical directors, ICU heads and IT leaders responsible for facilities of 50 beds and upwards, as well as at diagnostic chains, dialysis networks and home-care providers who need to monitor patients across multiple sites. The deliverable is a working, validated platform that integrates with the hospital's existing HIS / EMR via HL7 FHIR, complies with the Bangladesh Digital Security Act 2018 and the Health Services Division's draft eHealth Data Protection Guidelines, and is fully supported under SLA.
This paper sets out the methodology, the technology stack, the engagement model and the kind of clinical and operational outcomes a mid-sized hospital can expect.
Industry Context
The global Internet of Medical Things market was valued at approximately USD 60 billion in 2024 and is projected to reach USD 285 billion by 2032. The drivers are well known: ageing populations in OECD markets, post-COVID emphasis on remote monitoring, and the cost pressure on hospital systems that makes continuous monitoring of stable patients in lower-acuity beds attractive.
The Bangladesh picture is one of rapid private-sector expansion. The country now has more than 5,000 registered hospitals and clinics, of which roughly 600 operate ICUs. Premium tertiary facilities in Dhaka, Chittagong, Sylhet and Khulna are competing on patient experience and clinical outcomes, and increasingly on the digital sophistication of the bedside. Government tertiary hospitals are also moving: the Health Services Division's Digital Health Strategy 2023–2027 calls for HL7 FHIR adoption, unique patient identification and interoperable EMRs across the public sector.
Regulatory expectations are tightening. The Bangladesh Medical & Dental Council, the Directorate General of Drug Administration and the Health Services Division have collectively signalled that medical-device data must be traceable, tamper-resistant and stored within Bangladesh except where bilateral agreements allow. The draft eHealth Data Protection Guidelines align broadly with HIPAA and GDPR principles — consent, minimisation, breach notification and right of access.
Clinically, the case for continuous monitoring outside the ICU is increasingly compelling. Multiple international studies have shown that continuous vitals on general wards reduce code-blue events by 30–50 percent, cut unplanned ICU transfers by 20–25 percent and shorten length of stay by 0.6–1.1 days. For a 200-bed private hospital in Bangladesh, those numbers translate to material revenue and reputational benefit.
Challenges We Solve
- Spot-check blindness. Vitals taken every four hours miss the early-warning trajectory that precedes most adverse events.
- Alarm fatigue. Bedside monitors that alarm on every threshold breach desensitise nursing staff; smart, context-aware alerting cuts non-actionable alarms by 70–80 percent.
- Lost or misplaced assets. Infusion pumps, ECG machines and portable ventilators wander between wards; staff spend hours per shift looking for them.
- Manual charting. Nurses spend 25–35 percent of their shift on documentation that could be auto-populated from the bedside.
- Siloed data. Vitals, lab results, medication administration and notes sit in separate systems, making clinical review fragmented.
- Compliance reporting. Hand-hygiene compliance, bed-turn-around times and infection-control rounds are hard to audit without instrumentation.
- Remote and home-care scaling. Post-discharge monitoring of cardiac, dialysis and chronic-disease patients is operationally heavy without an IoMT platform.
Our Approach
Discovery & Requirements Gathering
We begin with a clinical workflow study: shadowing nurses, residents and consultants in target wards, mapping handover points, alarm flows and documentation steps. We deliver a written clinical-and-technical assessment, including a regulatory gap analysis against the eHealth Data Protection Guidelines and a clinical-risk register reviewed with the medical director.
Solution Architecture
We design a layered architecture that respects the hospital's existing investments. Bedside medical devices (multi-parameter monitors, infusion pumps, ventilators) connect via their native HL7 or proprietary protocols to a clinical-integration engine. Wearable patches and BLE peripherals connect through ceiling-mounted gateways to the same backbone. All data flows through an HL7 FHIR API into the central platform and is pushed back to the HIS / EMR.
Hardware Selection & Procurement
We work with established medical-device vendors (Philips, Mindray, Masimo, Welch Allyn, Nihon Kohden) and CE / FDA-cleared wearable patches for ambulatory monitoring. RFID and BLE asset tags are sourced from Stanley Healthcare, Zebra and equivalents. We handle DGDA registration, customs and bio-medical engineering coordination.
Implementation & Integration
Implementation is staged ward by ward, starting with a pilot of two beds in a non-critical ward to validate end-to-end data flow, alarm logic and EMR integration. Every device is bio-medical-engineering checked, MAC-address registered and added to the clinical asset registry before commissioning.
Deployment & Commissioning
Commissioning is performed under joint sign-off by the clinical lead, the IT lead and our project manager. We run a two-week shadow period where the existing manual workflow continues in parallel; only after the platform has been validated against manual observation is it adopted as the source of truth.
Operations & Optimisation
Post-go-live we operate a 24x7 monitoring service for the platform itself (not the patients — that remains the clinical team's responsibility). The optional AMC includes biomedical engineering coordination, alarm-threshold tuning, quarterly clinical-governance reviews and integration of new device types as they are introduced.
Technology Stack
| Layer | Technologies |
|---|---|
| Edge / Devices | Philips IntelliVue, Mindray BeneVision and Masimo Root multi-parameter monitors; Welch Allyn Connex spot-check; Masimo Radius PPG and VitalConnect VitalPatch wearable patches; BLE-enabled SpO2, BP, glucose and thermometer peripherals; medical-grade infusion pumps with HL7 output |
| Connectivity | Ceiling-mounted BLE-to-MQTT gateways with hospital-grade Wi-Fi 6 fallback; segmented medical-device VLAN; private 4G backup for critical alarm path |
| Integration | HL7 v2 and HL7 FHIR R4 connectors; integration engines (Mirth Connect, Rhapsody) for legacy HIS; DICOM gateway for imaging modalities |
| Backend | EMQX MQTT broker tuned for low-latency alarm path; TimescaleDB for waveform and vitals history; PostgreSQL for clinical records; Redis for live state; Kubernetes deployment on-premises within the hospital data centre |
| Visualisation | Central nurses' station dashboard (web), tablet-based bedside view for clinicians, native mobile app with priority push for on-call consultants, large-format ward-overview displays |
| Asset Tracking | BLE / RFID tags from Stanley Healthcare, Zebra; floor-plan visualisation with last-seen and zone-dwell analytics; hand-hygiene compliance modules |
| Security & Compliance | TLS 1.3, mTLS between devices and gateways, AES-256 at rest, role-based access via Active Directory / Azure AD, full audit logging, configuration aligned with HIPAA, the Bangladesh DSA 2018 and draft eHealth Data Protection Guidelines |
Engagement Model
| Phase | Duration | Deliverables | Payment Trigger |
|---|---|---|---|
| 1. Clinical Workflow Study | 3–4 weeks | Workflow maps, regulatory gap analysis, clinical-risk register, opportunity quantification | Fixed fee on report acceptance |
| 2. Architecture & Compliance Design | 4–6 weeks | Reference architecture, HL7 FHIR integration specification, network & security design, DPIA | 30% of phase on sign-off |
| 3. Pilot Ward Deployment | 8–12 weeks | Two beds live, EMR integration validated, alarm logic tuned, shadow period passed | 30% on pilot acceptance |
| 4. Hospital-wide Roll-out | 4–9 months | Remaining wards instrumented, central station live, asset-tracking live, clinical training delivered | Milestone-based per ward accepted |
| 5. Clinical Hyper-care | 90 days | 24x7 platform monitoring, on-call clinical-informatics support, weekly stability reports | Included in roll-out |
| 6. AMC (optional) | Annual | Platform SLA, alarm tuning, new-device onboarding, quarterly governance reviews | Quarterly in advance |
Case Study Example
A 280-bed multi-specialty private hospital in Dhaka engaged SGT Systems to address two pain points: a steady rise in code-blue events in the post-surgical ward and chronic friction in locating shared infusion pumps. Over five months we instrumented 96 general-ward beds with continuous wearable vitals patches and bedside integration to the existing Mindray monitors, tagged 240 mobile assets, and integrated the platform with the hospital's Oracle Cerner-based EMR via HL7 FHIR.
Six months after full go-live the measured outcomes were: code-blue events in monitored wards down 41 percent, unplanned ICU transfers down 23 percent, average nursing documentation time reduced by 19 minutes per shift, and time spent searching for infusion pumps reduced from a measured 38 minutes per shift to under 4. The clinical director reported a meaningful improvement in junior-doctor handover quality because the trend graphs are now consulted at the bedside rather than reconstructed from memory.
Why SGT Systems
Healthcare IT is unforgiving. Clinical adoption is the single biggest determinant of programme success, and clinical adoption is impossible without engineering discipline that respects the realities of the bedside. We have built our IoMT practice around three principles.
First, clinician-led design. Every workflow and alarm decision is reviewed with the actual clinicians who will use it, and our delivery teams include nurse-informaticists who have worked in Bangladeshi wards. We do not ship dashboards that look beautiful in a sales deck but fail in the heat of a 2 a.m. emergency.
Second, integration depth. HL7 v2 and FHIR R4 are not just badges on a brochure; we have shipped working bidirectional integrations against multiple commercial HIS / EMR platforms and several locally built systems, including write-back of vitals into the patient chart with the correct audit trail. We carry our own integration engineers rather than depending on the EMR vendor for every change.
Third, regulatory and security maturity. Our delivery follows ISO 27001-aligned procedures, includes a Data Protection Impact Assessment as a standard deliverable, and is reviewed against the Bangladesh Digital Security Act 2018, the draft eHealth Data Protection Guidelines and HIPAA principles. The DGDA notification, biomedical-engineering coordination and clinical-governance committee engagement are not optional extras — they are part of the standard delivery.
Pricing & Procurement
IoMT engagements are quoted per bed, per monitored asset and per integration. Indicative ranges: BDT 95,000–220,000 per general-ward bed for continuous vitals (including wearable patches, gateway share and platform license), BDT 4,500–9,000 per RFID-tracked asset, and BDT 12–28 lakh for HL7 FHIR integration with an existing HIS / EMR. ICU and step-down beds are quoted separately depending on the existing monitor estate.
Quotes are issued in BDT and USD with milestone-based payment per ward accepted. We support both capital-purchase and managed-service models; under the managed-service model the platform, devices and AMC are bundled into a per-bed-per-month fee. AMC for capital-purchase clients is typically 15–20 percent of capital cost per year and includes 24x7 platform support, biomedical engineering coordination and clinical-informatics consultation.
Frequently Asked Questions
Where is the patient data stored?
By default, all clinical data is stored on-premises inside the hospital's data centre, on infrastructure the hospital owns. Hybrid models with anonymised analytics in a regional cloud are available where the hospital has explicit consent. We do not export identifiable patient data outside Bangladesh.
How does this integrate with our existing EMR?
We use HL7 FHIR R4 wherever the EMR supports it, and HL7 v2 with an integration engine (Mirth Connect or Rhapsody) where it does not. We have working integrations with Oracle Cerner, Epic, OpenMRS and several locally developed HIS platforms.
What happens when the network goes down?
Bedside devices continue to alarm locally as they always did. Our gateway tier buffers data and replays it when the network returns. The alarm path can be configured to fail over to private 4G for critical wards.
How do you handle alarm fatigue?
Our alarm engine applies context (patient acuity, trend direction, time-of-day, recent intervention) to suppress non-actionable alarms and escalate genuinely deteriorating patients. Threshold tuning is a continuous activity during the hyper-care period and the AMC.
Are the wearable patches DGDA-registered?
The patches we deploy carry CE Class IIa or FDA 510(k) clearance and we complete DGDA notification before deployment. We maintain a full device registry with batch numbers, calibration dates and decommission records.
Can we extend this to home monitoring after discharge?
Yes. The same platform supports a discharge-care module that lets a patient continue to wear a vitals patch at home for a configured period. Data flows over the patient's home Wi-Fi or a cellular hub we provide, and is visible to the consultant and care coordinator.
How is the clinical-governance committee involved?
The hospital's clinical-governance committee owns alarm escalation policy, threshold protocols and any algorithmic decision support that influences clinical action. Our role is to implement what the committee approves; we never deploy a clinical-action-influencing change without sign-off, and the platform records the committee version that was in effect for every alert generated.
What is the expected lifetime of the wearable patches?
The reusable patches we typically deploy are rated for 1,000+ wears with single-use adhesive electrodes that are changed per patient. The patch electronics carry a 3-year warranty and a documented refurbishment programme so that capital cost amortises predictably.
Next Steps
The recommended first step is a 90-minute discovery call with your medical director and IT lead, followed by a half-day on-site clinical workflow review at no charge. We will leave you with a written opportunity estimate and a sandbox login to an anonymised demo dashboard. Reach out via the contact page and a member of our clinical-informatics team will respond within one business day.