This study evaluated the quality and content of transmitted digital images in a prehospital telemedicine system under routine clinical conditions. Pictures were taken and transmitted in 42.8% of the analysed missions. Most of them were deemed of “good quality” (76.4%) and mainly captured previously existing documents, like medical reports, medication lists, or 12-lead ECGs. In only 6.1% of the pictures, an identifiable face of the patient was displayed, but most of the documents contained confidential personal data. Between the groups, patient characteristics differed significantly in terms of age and the frequency of acute coronary syndromes. Therefore, no conclusions from a comparison of these groups should be drawn regarding the influence of image transmission on prehospital time intervals.
Digital image transmission was used frequently and resulted in the transmission of photographs with predominantly satisfactory picture quality. Camera shaking, which was a common reason for impaired quality, is nearly impossible to prevent in an emergency setting. Using the flash while taking pictures of documents frequently caused pictures to be overexposed, thereby resulting in pictures without an identifiable content. Therefore, we recommend adjusting the default settings of the flash used. Quality and readability improved significantly when the uncompressed transmission mode was used, but the file sizes were about 50 times higher. Compressed jpeg-files seem to be reasonable in this setting, but studies to evaluate whether lesser file compression improves the quality in the same way as the uncompressed mode are needed. Especially in rural areas, the amount of data is crucial in terms of determining acceptable transmission times because mobile networks with lower uplink capacities are mostly available.
Most of the photos contained personal data, but an identifiable face was photographed in only a few missions. In most instances, no new data was generated (e.g., picture of a face); instead, previously existing data (e.g., medical documents) were converted into a digital format and transmitted. To transmit video files from the emergency site could have been an alternative, but technical and practical arguments like described above led to the use of still picture transmission. In the future, when mobile networks enable faster and more stable upload capacities, mobile video transmission might be easier to realise and could be more meaningful in certain situations (e.g., remote neurological assessment). However, the focus of the described telemedicine system was to achieve sufficient data for remote teleconsultation within a narrow time period. Most of the time documents were photographed and such information can be extracted easily from single pictures. Thus, the tele-EMS physician was able to gain detailed information without committing to an overly time consuming audio communication with the EMS team. To get sufficient data for necessary medical decisions within the shortest possible time period is crucial for emergency teleconsultation. Unfortunately, the extent to which measurable benefits to patient care were reached remains unclear. If electronic patient records become accessible via the Internet in the future or if localized electronic health information (e.g., electronic health card) becomes routinely used, the need for photographs of medical documents will be reduced noticeably. The described system enabled image transmission of previously printed 12-lead ECGs (e.g., by the general practitioner) in addition to or instead of the 12-lead ECG transmission from the vital data monitor. Consequently, this data transmission enabled the pre-notification of the cardiologist on-call when needed. Patients with acute coronary syndromes that were treated with additional teleconsultation received significantly more often urgent percutaneous coronary intervention compared to patients that were treated by regular EMS [8]. Previous research has clearly demonstrated that transmission of the 12-lead ECG enhances treatment processes and improves outcome [9–13]. In a study using a cellular video-phone for remote interpretation of prehospital 12-lead ECGs, the interpretation quality was comparable to the evaluation of a printed ECG [14]. Ohtsuka et al. demonstrated that the transmission of 12-lead ECGs was feasible and rapid using an older type camera phone [15]. All pictures of ECGs (n = 28) in our study were completely readable (Figure 3). In situations where a previously recorded ECG differs from the current ECG, picture transmission seems to offer a meaningful contribution to the standard ECG transmission modes. However, the main intention of picture transmission was to enable rapid medical assistance by a remote experienced physician who was the receiver of all transmitted data. Medical decisions can be based on transmitted pictures if they contain critical information (e.g., medical report, 12-lead-ECG) and this approach can reduce the amount of audio communication needed during emergency teleconsultation.
To evaluate the influence of picture transmission on prehospital time requirements, we first analysed the comparability of both groups regarding patient characteristics. Although no significant differences in the NACA score and gender distribution were detected, patients of group 1 were significantly older and significantly more patients were diagnosed with acute coronary syndrome compared to group 2. Therefore, no conclusions from a significantly prolonged on-scene time should be drawn due to the different patient characteristics that may have caused the prolongation of the on-scene time. Overall, the meaningfulness of time interval comparisons between both groups is questionable. However, the ambulance arrival to hospital arrival intervals did not differ significantly, but data about the exact driving route or the use of emergency lights and sirens on route to the hospital were not available.
The operation of the described system represents a considerable effort with associated costs. In contrast, the use of smartphones could be a comparatively inexpensive alternative. Indeed, pictures taken with a smartphone can be transmitted to any e-mail address or to another smartphone. Unfortunately, this transmission occurs mostly without proper encryption, and the reliability is unknown. Furthermore, local storage of confidential data seems to be problematic. Smartphones have already been evaluated for similar purposes in different disciplines. For example, in the plastic surgery context, image transmission led to shorter treatment intervals with a comparable diagnostic accuracy in the assessment of free flaps when compared to the classic in-house assessment [16]. Even with an older 1.1 megapixel camera phone, satisfactory assessments of the replantation potential of completely amputated fingers were possible [17]. Pirris et al. demonstrated that a patient’s cell phone camera can be used for the remote evaluation of infected wounds [18]. In such an ambulatory setting, ultra-short transmission times are not required, and in situations with non-urgent communication between a patient and the surgeon, compromises in the reliability of transmission are acceptable. However, for prehospital teleconsultation a single smartphone does not enable multiple telemedical applications. Indeed, secure availability, high reliability, and encoded transmission are required for teleconsultation in EMS. Our developed pilot system was designed to meet these demands, but in 4.5% of all missions (n = 289), a complete drop-out occurred. Prior to a routine use the reliability must be improved in order to achieve the advantage offered by the combined use of several networks. Successful functioning is crucial for the introduction of potentially helpful telemedical applications. Prior to our project, different systems with similar applications were developed but not evaluated in clinical routine [19]. In a previous observational study with a precursor of our system, a tablet-computer with a built-in camera was used. The frequency of picture transmission was comparable; however the content of the photos was not evaluated, and the picture quality was only rated cumulatively [20]. Poor reliability and stability of this tablet-PC led to the described changes.
In addition to the operation of digital cameras within a telemedicine system, they can be useful for documentation and teaching in EMS. Photos taken on-scene enable an in-hospital trauma team to get realistic impressions of accident kinematics. A large display of images is desirable, but even the built-in screen of the camera allows this information to be distributed. Prehospital images should be saved in picture archiving and communication systems (PACS) so that they become available for the whole treatment process.
Limitations
The frequency of picture transmission probably depended not only on the medical necessity but also on the team’s attitude towards the system. Pictures were taken during day-time from spring to autumn due to restricted funding capacities. Pictures taken during the night or under different weather conditions (e.g., snow) would probably result in varying picture quality. Unfortunately, no technical guidelines for camera use were provided to the EMS. If such recommendations would have been implemented, different results may have been detected. As mentioned above, the non comparable groups led to very limited assessable results regarding the comparison of time intervals. Moreover, incorrect assessments by the investigators cannot be ruled out definitely, but we minimised this risk by using the described assessment procedure. This study was designed to evaluate feasibility, quality, and content of transmitted images and influences on time requirements. Influences on the patient outcomes were not measured and not purpose of this study. If the described approach would be implemented into routine care to support paramedic staffed ambulances this study has to be repeated in this different setting. The indications for still picture transmission might be different when no physician in on-scene.
