The Danish EMS is a two-tiered ambulance system consisting of ambulances as first-line response and an advanced response from PCCAs in rapid-response vehicles or helicopters [14,15,16].

Healthcare in Denmark is a tax-supported service. Each of the five regions has its own EMS and a varying number of regional general hospitals capable of treating most common medical and surgical conditions. The Central Denmark Region has 69 ambulances staffed by ambulance clinicians (emergency medical technician (EMT) students, ambulance assistants, EMTs, or paramedics) and ten rapid-response vehicles staffed by a PCCA and an EMT or a paramedic.

We identified all EMS personnel in Central Denmark Region from mailing lists provided by the Prehospital EMS, Central Denmark Region and Falck Denmark A/S. The survey was distributed to 810 ambulance clinicians and 152 PCCAs.

The present guideline for prehospital fluid administration in the Central Denmark Region recommends 1000–2000 ml NaCl 0.9% as a fast infusion for septic shock; however, no recommendations for suspected infection or sepsis without hypotension are given.

To examine the fluid treatment administered by ambulance clinicians and PCCAs in patients with suspected infection, an online survey was developed.

The survey instrument included both 5-point Likert-scale questions rated from ‘strongly disagree’ to ‘strongly agree’ and multiple-choice options for different parameters. More attitude-based questions included a “Don’t know” option, to ensure, that respondents would complete the survery in stead of abandoning the survey or answering randomly. The survey consisted of three sections: (1) baseline information, (2) perception of, education and knowledge about fluid administration and (3) scenarios about fluid administration (two hypothetical scenarios of patients brought in by EMS). The survey was in Danish. A translated version is provided in the Additional file.

The online survey was developed and refined through a thorough, stepwise process. First, the survey was developed with a group of paramedics and physicians with special knowledge of and interest in fluid therapy, inspired by a similar in-hospital survey which was modified to reflect prehospital challenges: i.e. about fluid administration to patients with suspected infection as patients are in the prehospital setting in stead of a firm diagnosis of sepsis and to cover only the period of transportation to the hospital. Then, the survey was pretested by six EMS employees who volunteered to do so, followed by a 30–60-min semi-structured interview. The interview first covered the volunteers’ perception of the entire field of prehospital fluid administration in general, followed by going though all questions ensuring, they perceived the questions as intended. The interviews resulted in a more exact and improved survey in terms of understanding.

Baseline information covered age, educational level, years of experience and gender. In the section ‘Perception of, education and knowledge about fluid administration’, the respondents were asked how much they agreed to statements about fluid therapy, as well as whether they were aware of and used the current guideline. Afterwards, they had to indicate if they faced challenges and if yes, then which kinds of challenges they faced from a predefined list, with the option to add further challenges as comments. Finally, they were asked to select the five most commonly used triggers, from a list of 22 predefined triggers, for fluid initiation. We created two almost identical clinical scenarios: A 55-year-old, previously healthy woman (70 kg), complaining of fever and dyspnoea was assessed by the ambulance clinicians. For the last 14 days she had been coughing and spitting. She was slightly confused (Glasgow Coma Scale (GCS) 15), blood pressure 120/75, pulse rate 120, respiration frequency 28, temperature: 39.1 Celsius and saturation 92% (3 Litre nasal oxygen per minute). The expected transportation time was 30 min.

In the second, she was hypotensive with a blood pressure of 88/60 mmHg. For each scenario, respondents were asked (1) how much fluid they would administer to the patient during 30 min of transportation, (2) how fast they would infuse and (3) how they made their decision.

The survey was distributed through emails with an embedded survey link. Emails were sent on June 9th, 2021, with up to six reminders. The survey was closed on July 19th, 2021. Responses were anonymous. Participation was voluntary, and informed consent was obtained at the beginning of the survey. No incentives were offered. Partly completed surveys were excluded from the analysis. All responses were collected through REDCap, version 10.6.16, hosted at Aarhus University [17].

All data were presented as numbers and proportions (%) for dichotomous and categorical variables. Data were analysed using Stata Version 16 (StataCorp LP, College Station, TX, USA).