Our study demonstrates that it is feasible to obtain physiological data from several data sources during acute care trajectories until definitive care in the ICU. In this study, physiological variables from the pre- and in-hospital phase for patients with ICH/SAH, OHCA and severe TBI were merged, offering a possible continuously sampling of data in the initial phase of critical illness or injury. However, the high number of 5-min intervals without measurements of important physiologic variables does not justify the use of the term continuous physiological data in this study. Frequent deviations from physiological treatment goals were observed, but the study was too small to assess any possible effects on patient outcome.
It may be challenging to establish and maintain monitoring, especially when patients are handled between different health care providers. Inadequate documentation in this phase has been demonstrated to be associated with increased mortality, although this may be confounded by the severity of illness or injury . The pre-hospital operational setting with bad weather, temperature changes, and movement may contribute to the lack of monitoring. In addition, prioritization of immediate treatment over monitoring and the use of several different monitors along the treatment trajectory may affect the degree of monitoring. Clinical and operational decisions are continuously made, balancing the need for rapid transfer to hospital versus establishing advanced therapy before and during transport. For the patient groups included in this study, an optimization of oxygenation and circulation on site and during transport may have a better impact on patient outcome than rapid transfer to hospital. Although most patients had arrived at hospital within 2 h after dispatch, lack of monitoring and deviations from treatment goals continued to be an issue.
For some patients, monitoring was not initiated by either P-EMS or ground-EMS before a considerable amount of time had passed and was often incomplete. Moreover, it occurred that patients were detached from the monitor leaving possible physiological deviations undetected at several occasions. This may be related to that Ground-EMS and P-EMS used separate monitors. However, it is unlikely that detachment and re-attachment of equipment should take more than a couple of minutes in a normal operational setting, suggesting that other interventions may have been prioritized . Occasionally, physiological variables in the ED were not registered. To obtain accurate documentation in the ED and during radiologic examinations may be challenging but favorable [22, 23]. Our observations agrees with a study by Chen et al. on in-hospital documentation of vital signs, which illustrated the difficulty of maintaining continuous monitoring of critically ill hospitalized patients . As we have demonstrated in Fig. 2b, the proportion of patients residing in the ED (i.e. the relative height of the grey field) was highest between 1 and 2 h after dispatch (20–50% of patients). During this period, the overall proportion of deviations reached nearly 40% (i.e. the relative height of the red field in Fig. 2a), and this was mainly due to hypoxia and hypotension (Supplementary Fig. 1). The low number of patients precludes an in-depth analysis of these issues.
The degree of physiological deviations experienced by the patients may be due to insufficient treatment provided or difficulties providing this given their critical condition. SpO2 was the variable that deviated most frequently in the pre-hospital phase. This is unfavorable, as hypoxia in the acute phase is known to affect the outcome negatively . SpO2 was also the most frequently monitored variable in the pre-hospital phase in our study. For some of the patients, monitoring of SpO2 was discontinued when patients were hypoxic, possibly leaving the patients hypoxic for an unknown amount of time. However, some missing SpO2 measurements might be related to the technology itself, as reduced blood perfusion might impede SpO2 measurements.
Systolic blood pressure had the highest proportion of deviations in the ED, and hypotension is associated with increased mortality after brain injury . Insufficient monitoring of systolic blood pressure is also documented in other studies [11, 21]. However, in the majority of patients, only non-invasive blood pressure monitoring was performed. The accuracy of this method is uncertain in hypotensive patients, and clinical signs may be considered as more useful by the clinicians—a possible cause for lack of measurements . After admittance to the ICU, wide fluctuations in systolic blood pressure were observed, which might occur due to patients being in an unstable circulatory status before and during initial resuscitation.
Exposure to hypercapnia after brain injury is associated with a poor clinical outcome . In four patients, potential deviations in end tidal CO2 were not registered due to lack of monitoring. EtCO2 had the lowest proportion of deviations in the ED, but also the highest proportion of missing data. However, in one patient (patient 2) where EtCO2 was monitored continuously throughout the pre-hospital phase, 11 deviations in EtCO2 were observed prior to hospital arrival. This agrees with the general self-fulfilling observation that the variables that deviated most frequently were the variables most often registered.
This study has several limitations. First, few patients were included in the study and they suffered from a wide range of medical conditions. This makes it difficult to draw specific conclusions regarding exposure to adverse physiological events and long-term outcome. Second, this was a one-center pilot study giving preliminary information needed before commencing further studies. Third, non-invasive systolic blood pressure was documented in the pre-hospital phase and invasive systolic blood pressure in the early in-hospital phase, which may cause variations in the results. Finally, with respect to EtCO2 measurements, we defined hypercapnia (> 6 kPa) as a deviation from optimal physiology. Hypocapnia (< 4.5 kPa) may also be harmful in these groups of patients. We considered possible gas leakage when using supraglottic devices, differences in end-tidal and arterial CO2 and poor circulation in patients with return of spontaneous circulation (ROSC) as potential sources of bias in this regard. For these reasons, we did not report the extent of hypocapnia in ROSC patients, and only two out of seven non-ROSC patients in this sample had measurements of EtCO2 limiting further exploration of this issue.