Human Factors and Postmarket Surveillance at FDA
Thomas P. Gross, MD, MPH
Division of Postmarket Surveillance
Office of Surveillance and Biometrics
Center for Devices and Radiological Health
Food and Drug Administration
Goals of Postmarket Surveillance
Human factors principles and practices have an important role to play in postmarket surveillance (PMS) at the Food and Drug Administration (FDA). To appreciate their role, one must understand the goals of and the regulatory authorities governing PMS. Briefly, the principal goal of PMS is to discover additional safety information on marketed medical and radiation-emitting devices. This is achieved through use of specific regulatory tools that aid in detecting adverse events (both real and potential), estimating adverse event frequencies (both absolute and relative), and identifying groups at risk. Among the desired outcomes in the process are improvements in device design and use, achieved through feedback of safety information to product manufacturers and evaluators.
FDA Postmarket Authorities
With regard to regulatory authorities governing PMS, two principal authorities, the Medical Device Reporting (MDR) Regulation and the PMS Statute, are noteworthy in terms of human factors principles and practices. The MDR Regulation governs reporting of adverse events related to the use of medical and radiation-emitting devices. It is an essential element in the broad range of FDA postmarket activities that protect the public health by capturing additional safety information that would not be gathered
otherwise. This can also be said of the PMS Statute, which governs the required study of designated newly marketed critical devices as well as the discretionary study of any device-related health issue deemed necessary to protect the public health. Although both surveillance entities can provide safety information about a device used in the general population under actual conditions, the MDR Regulation provides for passive surveillance, whereas the PMS Statute provides for active surveillance through formal study. Both of these authorities will now be explored in more depth, with references to the MDR Regulation based on the proposed new regulation.
Adverse Event Reporting
Means for reporting adverse events is currently provided through two principal programs at FDA. The voluntary program is known as MedWatch, and the mandatory program is referred to as MDR. MedWatch, the FDA's voluntary reporting program for the health care and consumer community, has developed a partnership with major health care organizations in the United States to assist FDA in promoting reporting, particularly those events that are of a serious nature. Mandatory reporting of adverse events, on the other hand, is required of user facilities, distributors, and manufacturers.
User facilities include hospitals, nursing homes, ambulatory surgical facilities, and outpatient diagnostic and treatment facilities, but do not include physician or other health care provider offices located outside of these facilities. These facilities are required to report device-related deaths and serious injuries. Serious injuries are defined as life-threatening events, events that result in permanent impairment of a body function or permanent damage to a body structure, and events that require medical or surgical intervention to preclude permanent impairment or damage. User facilities report individual deaths to FDA, with a copy to the manufacturer, and report serious injuries to the manufacturer. The reports must be submitted within 10 days of knowledge of the event.
Reportable events for both the distributor and manufacturer are the same. Both must report device-related deaths, serious injuries, and device malfunctions that are likely to cause or contribute to death or serious injury if they were to recur. Distributors are required to report all of these events to FDA and send a copy of these reports to the manufacturer. All reports must be submitted within 10 days of knowledge of the event. Manufacturers must submit reports on these incidents to FDA within 30 days of knowledge of the event as well as reports of remedial action within 5 days. Remedial actions are those occurrences, other than routine device maintenance or servicing, where action is necessary to prevent recurrence of a reportable event and include actions such as recalls, replacements, relabeling, notifications, and inspections.
The term "device-related," as it pertains to adverse events, means that the event was or may have been attributable to a medical device, or that a device was or may have been a factor in an event, including those occurring as a result of malfunction, poor manufacture, inadequate labeling, or improper design. Regarding the latter two, human factors principles are of course an important aspect because flaws in labeling and/or design potentially can induce human error.
Aside from device-related deaths, serious injuries, and malfunctions, some events can be classified as "near misses." These can be considered unintended actions or lack of actions that require corrective means to prevent injury. These events may be indicative of device-user interface problems, and thus are important from a human factors perspective. FDA encourages the health care community and consumers to report near misses voluntarily through the MedWatch program.
Upon the effective date of the Medical Device Reporting regulation, specific forms will be designated for reporting adverse events. The mandatory reporting form, also known as FDA form 3500A, will be required for the reporting of each event to FDA and the manufacturer. The form for voluntary reporting is nearly identical to the first page of this two-page mandatory form. Sections of the form will require mandatory reporters to use designated terms (and their corresponding numeric codes) from our coding manual, and not text, to characterize the event. For instance, user facilities and distributors will be required to select, from our coding manual of approximately 2,000 terms and codes, up to three event problem terms (codes) that describe what happened to the patient and up to three that describe what happened to the device. For example, terms (codes) that describe what happened to the patient might be "hypothermia" (code 1915), "stroke" (code 2086), and "headache" (code 1880). Terms that describe what happened to the device encompass human factors-related problems as well as those less likely to be human factors-related, such as electromechanical or software problems. Examples of the former might include terms directly related to design problems such as "difficult to deflate" or "difficult to position," terms indirectly related such as "misreading of display" or "miscalibration," terms directly related to labeling such as "mislabeled" or "incorrect instructions," and terms related to consequences where cause is unspecified such as "overinfusion."
Similarly, manufacturers will be required to select evaluation terms (codes) from the coding manual. Up to four terms (codes) addressing methods of evaluation, up to four addressing results, and up to four addressing conclusions will be allowed. Again, human factors-related terms can be chosen. For example, method terms might be "electrical tests performed" or "user-interface test performed;" results terms might be "design/user interface problem" or "operating steps confusing;" and conclusion terms might be "user interface contributed to event," "operational context caused event," or "labeling-related."
Once adverse events are reported to FDA, they are immediately evaluated by a clinical staff of 16 professionals, the majority of whom are nurses, for problem identification to determine if the events pose a real or potential risk to the public health. The evaluation process varies according to type of device, patient or device outcome, information in the current report, and information from previous reports and other data sources (such as the literature). As our clinical staff learn more about human factors principles, these are also taken into account in report evaluation. An initial evaluation may result in:
- The report being read with no further action planned;
- A request for additional information from the manufacturer or reporter;
- A request for an inspection of the reporting site and/or manufacturer;
- The initiation of internal FDA meetings for information sharing and/or consideration of educational activities, risk communication, and compliance options.
Examples of human factors problems identified through adverse event report evaluation include problems related to:
- Infusion devices and poor display depth resulting in injuries from over-infusion due to viewing angle difficulties;
- Ventilators and alarm failure resulting in injuries from shorting of alarms due to battery misinstallation;
- Oxygen delivery devices and inadequate flow control resulting in injuries from hypoxia due to misunderstanding of control settings.
Required and Discretionary PMS Studies
Knowledge of adverse event reports is important for a variety of reasons, not the least of which is to assist in defining outcomes in mandated PMS studies. As stated earlier, the PMS Statute, which governs these studies, is another tool to assess device safety and does so by active study of the device in its early postmarket life. The statute contains two separate, but related, authorities: required postmarket surveillance (RPS) studies apply to certain types of newly marketed devices, and discretionary postmarket surveillance (DPS) studies represent a broader authority that can be used for any device if FDA believes that there are postmarket questions about the product. It is important to understand the motivation behind the passing of this new legislation. First, Congress believed that FDA needed more and better postmarket surveillance tools, in addition to the MDR Regulation, to detect device problems as soon as possible. Secondly, Congress recognized that there was a limit to premarket safety data. Congress intended that this statute would allow for the clinical monitoring of the earliest experiences with the device once it is distributed in the general population under actual conditions of use.
The statutory requirements for RPS studies state that manufacturers must submit study protocols within 30 days of introducing the product into U.S. interstate commerce. In the case of DPS studies, a manufacturer has 30 days from the date of FDA's notification to submit a study protocol. FDA has 60 days to review both RPS and DPS study protocols. We are required by law to use several key review criteria. First, we must examine the qualifications and experience of the principal investigator to determine if that individual is qualified to conduct the study. We must also determine whether the study conducted according to the protocol will result in the collection of useful data to protect the public health and provide additional safety and/or effectiveness information for the device. For example, useful data might include estimation of rates of specified outcomes, re-operation, and/or explant. The statute also specifies that FDA cannot approve a postmarket surveillance protocol until it has been critiqued by an appropriately qualified scientific and technical review committee. We meet this requirement by putting together a team of experts from within FDA to evaluate each of these study protocols.
RPS applies only to newly marketed devices, that is those first marketed after 1 January 1991. The law also defined three categories of products that would be subject to RPS:
- Permanent implants, the failure of which may cause serious adverse health events or death;
- Devices that are intended for a use in supporting or sustaining human life;
- Devices that potentially present a serious risk to human health.
DPS broadens the scope of RPS by allowing FDA to use its discretion to call for studies on any device, newly marketed or not, where further safety and/or effectiveness data are desired or where knowledge of the given device-related issue is deemed necessary to protect the public health.
From the standpoint of human factors, the PMS Statute is likely more applicable to devices that are life-sustaining or supporting as opposed to permanently implantable devices, because user interface is more prominent with the former. Thus, if human factors problems figure prominently in the postmarket safety assessment of life-sustaining or supporting devices, then these factors and related outcomes might be formally studied under RPS or DPS. To date, the PMS Statute has not been implemented for this category of devices due to priority given to permanently implantable devices.
Human factors principles and practices have an important role to play in PMS at the FDA. The role of human factors is better appreciated after one understands the regulatory authorities/tools that may be brought to bear in their study. These postmarket tools exist and should be increasingly used to address human factors in medical and radiation-emitting devices. Furthermore, human factors concepts need to be incorporated more fully in the use of these tools. Feedback loops to product manufacturers and product evaluators, based on information garnered from adverse event reports or postmarket study, should be strengthened. Finally, the end user should be encouraged to be a partner in the adverse event discovery process.