Clinical Evidence

DGDA · GHTF/IMDRF principles · CER · Class B/C/D

Overview

DGDA clinical evidence requirements follow GHTF/IMDRF principles. The required level of evidence scales with device risk class. Clinical evidence submitted in a dossier must demonstrate that the device achieves its intended clinical purpose with an acceptable benefit-risk profile.

Requirements by Class

Class	Clinical Evidence Required
A	None — notification only; no clinical evidence required
B	Clinical evaluation summary — literature review demonstrating safety and performance is sufficient for most devices
C	Detailed Clinical Evaluation Report (CER); clinical investigation data may be required for novel devices without recognised equivalent
D	Comprehensive clinical evidence — full CER; clinical investigation data typically required for novel devices; strong overseas regulatory approval (FDA PMA, EU MDR, TGA) can support the dossier

Clinical Evaluation Report (CER)

The CER is submitted in the registration dossier for Class B, C, and D devices. A compliant CER must:

CER Section	Description
Device description and intended purpose	Clinical context — what the device does and for whom
Literature search	Database, search terms, date range, inclusion/exclusion criteria
Literature appraisal	Critical assessment of relevant published studies
Equivalence justification (if used)	Basis for using another device's clinical data
Clinical investigation data	Study design, endpoints, results — if available/required
Benefit-risk conclusion	Clinical benefits vs. residual risks
Ongoing PMS plan	How clinical performance will be monitored post-registration

Using Overseas Clinical Data

DGDA accepts clinical data generated outside Bangladesh, provided:

• The study population and clinical conditions are relevant to Bangladesh patients
• The device design and intended use are the same as the product to be registered
• The data is from well-designed studies (RCTs, prospective studies, or appropriate evidence for the device type)
• Any ethnic, environmental, or healthcare-system differences between the study population and Bangladesh are addressed

Regulatory approvals as supporting evidence: Strong regulatory approval from FDA (PMA or 510(k) clearance), CE marking under EU MDR, or TGA approval provides supporting evidence of clinical acceptability. For Class D devices, prior approval from a major regulator can significantly strengthen the dossier.

Equivalence

For Class B and C devices, clinical data from an equivalent predicate device can support the CER without new clinical investigations. Equivalence must be demonstrated across:

• Clinical equivalence: Same intended purpose, patient population, body site, conditions of use
• Technical equivalence: Same design principles, materials (where in contact), specifications
• Biological equivalence: Same materials in contact, same type and duration of body contact

If full equivalence cannot be demonstrated, clinical data from the subject device itself is required.

Clinical Investigations in Bangladesh

Clinical investigations in Bangladesh are governed by the National Ethical Guidelines for Health Research and require:

• Ethics Committee approval from an BMRC (Bangladesh Medical Research Council) recognised IEC
• Regulatory review by DGDA for investigational devices not yet registered
• Compliance with ICH GCP and Declaration of Helsinki
• Sponsorship by or with written agreement from the manufacturer or AR

Clinical investigation requirements apply when a Class C or D device without adequate overseas data needs Bangladesh-specific clinical evidence.

IVD Performance Evaluation

IVDs do not use a CER — they require performance evaluation data demonstrating:

• Analytical performance: Accuracy, precision, linearity, detection limits, interference effects
• Diagnostic performance: Sensitivity, specificity, positive/negative predictive value (for diagnostic tests)
• Stability: Claimed shelf life and open-vial stability supported by data
• Comparison study: Performance demonstrated against a reference method or predicate

IVD performance data must reflect the intended use setting (professional laboratory vs. point-of-care vs. self-test).