Interpreting physiological data

Using the aforementioned case above as an example, how does the pretest probability of arrhythmia in this individual known to have PAF influence the interpretation of the tachycardia and the subsequent treatment of a presumed arrhythmia? As far as we are aware, this technology is not validated to assess cardiac arrhythmias in any population, least of all the military. In a fit, healthy military population, who typically have no prior history of cardiac arrhythmias, it is vital to understand the consequences of producing data that may influence perceptions of occupational suitability. Should an episode of tachycardia in an otherwise healthy individual mean that they are medically downgraded?

Many questions are raised, which extend to other types of physiological monitoring (box). As with any source of clinical information, the data made available to the clinicians sitting in judgement must be assessed for its quality and reliability. International standards exist not only for the methods by which physiological strain responses may be evaluated in healthy individuals, as used with personnel working under high thermal stress,2 but also for the devices and software that are employed to monitor medical patients. There is the potential for misleading information to be provided and unnecessary interventions undertaken.3 Thus, issues of central importance relate to the point at which an individual may be said to have fallen off the ‘healthy’ stress–strain curve and should henceforth be considered a casualty. In addition, it must be determined whether the system used to identify this departure from the healthy state is adequate for monitoring aberrant physiology in disease.

Accessing and using the information

Military personnel operate under a duty of care, in conditions and with constraints that may differ markedly from civilian populations. In the case described, the subject elected to continue running and covered the intended distance without further reported issue. In a military context, a number of constraints exist. These include the requirements to define who owns and uses the data; whether data is reviewed contemporaneously or in retrospect; how the individual is alerted to new information, as with an alarm or voice-activated advice; whether they could choose to disregard the information and how the RMO contributes to decision-making, be it through real-time monitoring or ‘after-action’ review of events. Precedents from civilian life include both incidental and deliberate withholding of performance data from professional athletes by their medical and management teams, which may be considered necessary for optimising precompetition training or winning a race. In addition, there is the requirement for hospitalised patients and their relatives to apply to medical authorities, such as NHS Trusts in the UK, for the release of information collected during their inpatient admission. The additional ethical issues raised by capturing data from healthy military personnel extend beyond the more easily grasped beneficence and non-maleficence axis to challenge issues of personal autonomy. These include the freedom to achieve peak physical performance, which may on occasion fall a hair’s breadth from physiological failure,4 as well as justice, whereby limited resources may need to be targeted at those perceived to be at greatest risk.

Monitoring in context

Other challenging issues that may be of relevance in military settings include the scope for false reassurance. An example of this is a peripheral body measurement reading lower than true core temperature, the parameter of clinical interest in heat stroke.5 In a similar vein, misdiagnosis of exertional heat stroke as supraventricular tachycardia has been described, whereby focussing on one measured variable and neglecting to address other important parameters resulted in incorrect management.6 On the other hand, undue alarm from artefactual or methodological overestimation of a measured or derived variable could result in pre-emptive withdrawal of personnel from important activities, which they might otherwise have successfully completed.

In relation to heat stress and exercise, an issue of perennial military importance, it has been stated that ‘Man is not a pulse rate, a rectal temperature, but a complex array of phenomena.’7 In trauma research, HR monitoring in isolation has shown no value in differentiating injury from exercise-induced tachycardia. However, a non-invasively monitored combination of HR, blood pressure, stroke volume and pulse oximetry was able to reliably predict blood volume in a model of acute haemorrhage.8 The lesson is that, as in clinical practice, more than one variable may be required to support diagnosis and decision-making. Tools to analyse variability and changes in the complexity of recurring signals, such as ‘rolling’ HR over time, may help to characterise physiological strain, impending decompensation9 or overt illness10 and could mature to have practical utility in the future. On the battlefield, it may become possible for learning algorithms to help to identify and triage wounded military personnel according to the degree of deviation from ‘healthy’ physiology highlighted by monitoring, thereby producing an automated form of personalised medicine.11 12 However, any such approach must first be validated in populations representative of the military, before attempting to translate or extrapolate their use into occupational and clinical practice.

Implications for defence

Wearable technology represents an appealing and potentially hugely beneficial enhancement to the equipment currently used by UK military personnel, which could improve healthcare across a number of domains. It may prove possible to close a number of declared capability gaps in diverse areas. This includes managing heat illness in the jungle and other prehospital environments,13 predicting post-traumatic stress disorder in advance of operational deployment14 and optimising rehabilitation following musculoskeletal injury.15 The breadth, depth and accuracy of physiological data capture may be further increased by moving beyond the body surface to use indwelling devices. This may include implantable loop recorders to detect bradyarrhythmias and tachyarrhythmias at high altitude16 17 or continuous interstitial sampling to monitor clinical chemistry, such as lactate concentration during high-intensity exercise.18

Just because technology exists, however, does not mean we are mandated to use it. Rather, we must define the questions we wish to answer. Hypothesis-driven research must determine if the wearable technology in question can answer our specific questions and whether it benefits the wearer or, ultimately, the mission. The data obtained must then be validated and proven to be reliable and reproducible. Examples of ongoing work towards such outcomes include instantaneous evaluation of physiological responses to prolonged endurance exercise in female military personnel crossing Antarctica, investigation of the effects of high altitude on energy and metabolism using subcutaneous continuous glucose monitoring and characterisation of heat acclimatisation status by analysis of HR variability. Importantly, these projects have benefited from direction, guidance and coordination from the Academic Departments of the Royal Centre for Defence Medicine. Additional collaboration has occurred with subject matter experts at established homes of physiological research within Defence, including the Institute of Naval Medicine, the Army Recruiting and Training Division and Royal Air Force Centre for Aviation Medicine.