I know the vibration was not normal.

— Jack Lemmon

In the classic 1979 American disaster thriller film, China Syndrome, a Shift Supervisor at a fictional nuclear power plant Jack Godell (played by actor Jack Lemmon) notices an unusual vibration in his cup of coffee. Without giving away the plot, the plant executives choose to dismiss this ominous warning sign, and the result was a nuclear meltdown, ostensibly melting reactor into the underlying earth, “all the way to China”.

In the nonfictional world, vibration analysis also known as acoustic signature identification, and vibration signature analysis has been used for over 6 decades to monitor the condition of rotating machinery, from nuclear power plants1 to computer hard drives.2 Popularized in the early-mid 1980s by such groups as Bruel and Kjaer (B&K),3 the technology is based on the fundamental principle that all rotating machines exhibit periodic disturbances due to minor imperfections or contamination.

Consequently, if a thrombus is adherent to the impeller of a rotodynamic (rotary) blood pump—whether ingested or deposited in situ—it will vibrate. The resulting vibration is typically measured by accelerometer; and because vibration may create an audible sound (aka “noise”), a microphone can be used to detect the vibration less invasively. In turn, the acceleration (or sound) can be analyzed in a number of ways, the simplest of which is to assess its overall magnitude (or loudness, for example with a stethoscope). Frequency domain analysis can add a wealth of information.

Over the past 2 decades, several investigators, including the authors of this article, have used acoustical signal gainfully to detect thrombosis in a variety of clinically used ventricular assist devices (VAD). Schalit and colleagues were one of the first to use an accelerometer to detect thrombosis in the HVAD, an excellent subject for such a study. Their initial study published in 20154 focused on the third harmonic (3× the rotational speed) and demonstrated excellent sensitivity and specificity, with AUC greater than 0.85—significantly better than monitoring pump power (watts). Their current study reported in this edition5 provided yet better performance by focusing on nonharmonic frequencies—a rather clever hypothesis.

On account of their remarkable results, these investigators conclude that vibration analysis will improve clinical outcomes by detecting thromboembolic complications in the HVAD.

The logical next step, therefore, would be to hasten the translation of this technique to clinical practice. Lest this study be relegated as academic exercise, and the enormous effort and ingenuity applied by these investigators be in vain.

Clinical translation could ostensibly be achieved by clinicians, or patients themselves. This was actually done in the bygone era of the HeartMate I, circa 1997, when bearing wear was a critical mode of failure. Choudri et al. reported the use of both time and frequency analysis of biweekly transabdominal acoustic readings.6 Today, it is not uncommon for VAD personnel to auscultate implanted continuous flow pumps, while rounding on inpatients. But this is arguably an act of desperation, necessitated by the inaction of the device manufacturer.

In their prior 2014 report, Schalit et al. issued the recommendation that “An accelerometer integrated into the system will enable real time detection of LVAD thrombosis.”4 They are not alone in reaching this conclusion. A cursory online search uncovered over a dozen scientific reports demonstrating the efficacy of vibration analysis of rotary pumps—spanning the past decade, including eight that are cited in the present article.

This begs the question of why the technology was not incorporated into the HVAD, long ago? The problem articulated by Schalit et al., “thromboembolic events during LVAD treatment can lead to stroke or peripheral organ dysfunction … [or] pump malfunction” is undisputed, and the technology has been known to the MCS community since, at least, 1997.

Is anybody listening?

Federal guidelines require that a manufacturer of life-sustaining medical devices perform rigorous hazard analysis before releasing their device into circulation. Official guidances lay out the best practices to identify and assess potential the risks in terms of their severity and their likelihood of occurrence. In this case, there is no debate about the severity of a neurologic adverse event. And, similarly, the likelihood of thromboembolic events with the HVAD need not be speculated; it is a documented fact, and unacceptable. For several years.7,8

In such cases, ISO14971 (Section 6.2) states9:

“The manufacturer shall identify risk control measure(s) that are appropriate for reducing the risk(s) to an acceptable level… [using] one or more of the following, in the priority order listed:

a) inherent safety by design; b) protective measures in the medical device itself or in the manufacturing process; c) information [and/or training] for safety.”

When this process fails it then becomes the obligation of the FDA to recall the product as the risks outweigh the benefits. Which is what in fact did occur on June 3, 2021, with FDA distribution of a class 1 recall of the HeartWare devices across the nation because “There is an increased risk of neurologic adverse events and mortality associated with the internal pump.”10

Let us step back in time, and consider the three risk mitigation steps above. The first option, to redesign the HVAD to render it safe is eliminated by the inherent design of the blood-lubricated bearing of the device. The third option has been the topic of considerable clinical and biomedical research, leading to improvement in practices with respect to blood pressure management,11 anticoagulation,12 impeller speed regulation,13 and so on. Therefore option (b) would be the last line of defense—namely to introduce an alarm system to alert patient and clinician of imminent danger. The cost of a $2 accelerometer and software upgrades to incorporate the algorithm of Schalit et al. would be inconsequential as compared to the benefits. Although there would be an intangible cost of admission of failure, or worse, guilt.

Looking not too far forward in time, as safer (and more effective) devices become commonplace, the necessity for vibration analysis of Schalit et al. will hopefully become unnecessary and obsolete.

But for the for the greater than 3,000 HVADs currently in use, Medtronic is more than capable of providing an extracorporeal device based on this technology to patients presently supported—and possibly avert the VAD-equivalent of a China Syndrome.

1. Parry DL: Nondestructive flaw detection in nuclear power installations Elper EP, Roux DP (eds), in: Incipient Failure Diagnosis for Assuring Safety and Availability of Nuclear Power Plants. Gaitlinsburg, TN: United States Atomic Energy Commission, 1968, pp. 107–126. 2. Wang Y, Miao Q, Pecht M: Health monitoring of hard disk drive based on Mahalanobis distance. 2011 Prognostics and System Health Management Conference (PHM-2011 Shenzhen). 2011. 3. Bruel Kjaer: Condition Monitoring of Industrial Machinery Using Mechanical Vibration as a Machine Health Indicator, 1980. 4. Schalit I, Espinoza A, Sørensen G, et al.: LVAD thrombosis detection using third harmonic frequency measured with 3D accelerometer. J Hear Lung Transplant 34: S214–S215, 2015. 5. Schalit I, Espinoza A, Pettersen FJ, et al.: Improved detection of thromboembolic complications in left ventricular assist device by novel accelerometer-based analysis. ASAIO J 68: 1117–1125, 2022. 6. Choudhri A, Salehizadeh B, Levin H, Oz M: Acoustic frequency spectrum shifts in failing artificial hearts. J Acoust Soc Am 101: 3187, 1998. 7. Rogers JG, Pagani FD, Tatooles AJ, et al.: Intrapericardial Left Ventricular Assist Device for Advanced Heart Failure. N Engl J Med 376: 451–460, 2017. 8. Chiang YP, Cox D, Schroder JN, et al.: Stroke risk following implantation of current generation centrifugal flow left ventricular assist devices. J Card Surg 35: 383–389, 2020. 9. ISO 14971: 2019 Medical devices—Application of risk management to medical devices ANSI/AAMI/2019. 10. FDA: Stop New Implants of the Medtronic HVAD System – Letter to Health Care Providers. 2021. Available at: https://www.fda.gov/medical-devices/letters-health-care-providers/stop-new-implants-medtronic-hvad-system-letter-health-care-providers. Accessed July 30, 2022. 11. Teuteberg JJ, Slaughter MS, Rogers JG, et al.: The HVAD left ventricular assist device: Risk factors for neurological events and risk mitigation strategies JACC Hear Fail 3: 818–828, 2015. 12. Jennings DL, Gellatly RM, Szandzik EG, Leet A, Lanfear DE: Anticoagulation for the HeartWare HVAD: An International Comparison of Strategies and Outcomes. J Hear Lung Transplant 33: S313–S314, 2014. 13. Granegger M, Thamsen B, Schlöglhofer T, et al.: Blood trauma potential of the HeartWare ventricular assist device in pediatric patients. J Thorac Cardiovasc Surg 159: 1519–1527.e1, 2020.