A large cross-sectional analysis in JAMA Health Forum is drawing attention to a persistent weakness in modern medicine: many high-risk therapeutic devices reach patients with limited clinical evidence, yet their most serious safety problems may only become visible after widespread use. The study reports that more than one-third of the high-risk devices examined were later associated with serious recalls, raising new questions about how regulators, manufacturers, clinicians, and health systems monitor technologies once they enter routine care.
Therapeutic medical devices include products designed to diagnose, treat, support, or replace bodily functions. Unlike drugs, which are absorbed or metabolized by the body, these devices can interact mechanically, electrically, biologically, or digitally with patients. Examples include implantable pumps, cardiac devices, surgical equipment, neurostimulation systems, and other technologies whose failure can cause severe injury or death. Their risks may depend not only on the device itself, but also on surgical technique, software, maintenance, patient anatomy, and how the product is used in real-world settings.
The analysis focused on high-risk therapeutic medical devices and examined whether serious recalls were associated with characteristics visible before or around the time of regulatory authorization. These characteristics included features of the device, the amount and type of premarket clinical evidence, and regulatory factors. Serious recalls generally involve products that may cause significant harm, require urgent corrective action, or pose a substantial risk to patients if they remain in use.
The central result was striking: more than one-third of the devices in the study were subject to serious recalls. Yet the investigators did not identify consistent relationships between recall risk and the device characteristics, premarket clinical evidence, or regulatory characteristics they evaluated. In practical terms, the signals available before approval did not reliably distinguish which technologies would later face serious safety problems.
That finding does not mean that clinical evidence or regulatory review is unimportant. Premarket studies are essential for identifying whether a device can work and for detecting common or immediate complications. However, clinical trials often involve relatively small numbers of carefully selected patients and may last for limited periods. Rare failures, delayed complications, manufacturing defects, software errors, and problems that emerge only after thousands of procedures can remain undetected until a device is widely adopted.
Medical devices also evolve in ways that can complicate safety assessment. A product may be modified after authorization, combined with new software, manufactured at a different facility, or used for patients who were not well represented in the original studies. Even when each change appears minor, the cumulative effect can alter performance. In addition, clinicians may adapt devices for new procedures or patient populations, creating real-world conditions that were not fully captured by premarket testing.
Recalls are therefore an important component of the device safety system, but they are not a complete substitute for continuous surveillance. A recall may occur only after adverse-event reports, clinician complaints, hospital investigations, manufacturer testing, or regulatory analysis reveal a pattern. Underreporting, inconsistent terminology, and delays in recognizing signals can make it difficult to determine how frequently a device is failing or which patients are most vulnerable.
The study’s implications extend beyond individual products. Hospitals and health systems increasingly rely on complex networks of implants, monitoring equipment, connected software, and automated treatment systems. A failure in one component can affect an entire clinical workflow. Stronger postmarket surveillance could help identify problems earlier by linking device identifiers with electronic health records, procedural data, maintenance histories, imaging findings, and patient outcomes. Such systems could also make it easier to determine whether a safety signal is concentrated in a particular model, manufacturing lot, clinical setting, or patient group.
Experts have increasingly advocated for surveillance methods that use large-scale data analysis, active monitoring, and rapid communication rather than depending primarily on voluntary reports. Machine-learning tools may help detect unusual patterns, but algorithms require accurate and standardized data and cannot replace clinical judgment. Patients and clinicians also need clear information about recalled products, the severity of potential harm, recommended corrective actions, and whether an alternative treatment is available.
The authors’ conclusion is especially relevant as medical innovation accelerates. A device can be technically sophisticated, approved through an established regulatory pathway, and supported by clinical evidence, yet still present risks that become apparent only through broad and prolonged use. The findings suggest that safety should be treated as an ongoing process rather than a single decision made before market entry. For high-risk therapeutic technologies, robust postmarket surveillance may be essential to ensure that innovation improves care without allowing hidden failures to spread across the health system.
Subject of Research: Safety and serious recalls of high-risk therapeutic medical devices
News Publication Date: Information not provided
Web References: https://doi.org/10.1001/jamahealthforum.2026.2626
References: Mooghali M, et al. Study published in JAMA Health Forum. DOI: 10.1001/jamahealthforum.2026.2626
Keywords: Medical treatments, medical technology, medical equipment, risk factors, regulatory policy, therapeutic medical devices, serious recalls, postmarket surveillance, patient safety
