Are they really refusing to travel? A qualitative study of prehospital records

Ethical approval for this study was obtained from Lincolnshire Local Research Ethics Committee (reference no. 04/Q2405/46).

There were 76635 emergency requests resulting in activation of an ambulance response from April 2003 to March 2004. This resulted in 63650 transportations, 9068 instances of refusal to travel (RTT), 967 deaths and 2950 calls with other reasons for non-transportation (see below). The rate of RTT as a proportion of the total requests was therefore 11.8% and this was stable throughout the year without significant seasonal variation. A single month (October 2004) was therefore selected as typical for both the pilot and the main study. Consecutive cases of RTT were selected for analysis until saturation was reached. Cases where treatment only had been declined but transport accepted were excluded.

An initial pilot study (n = 50) was undertaken to develop the data collection tool and template for the full study. 397 (i.e. a further 347) records were required to achieve theoretical saturation, i.e. no new data emerging to contribute to theoretical development of new or existing categories. Free text information from records was transcribed using data collection headings agreed by all the authors (Table 1). Recorded data were entered into QSRN6 for analysis [13] Data were further organised into categories and grouped into themes by two independent researchers (JM, DS). These were described and named after discussion by the whole research team (which included paramedic, nursing, medical, allied health professional and lay members) reflecting a multidisciplinary perspective. Descriptive statistics were used to assess frequencies within themes.

A count of clinical observations recorded on the patient report was included in the analysis because this indicated that the crew had made a clinical assessment of the patient. This was important because some instances of RTT, for example when the patient was intoxicated, could involve refusal of clinical assessment as well as transport.

Presenting conditions were recorded and grouped into six main problem themes (Table 2). In some instances there was more than one presenting condition recorded for a single patient. Sometimes information was recorded as a medical diagnosis such as chronic obstructive pulmonary disease (COPD) whilst at other times using lay terminology, such as breathing difficulties. Whatever the preferred terminology failure to transport patients to hospital was usually explained in medical terms. Most patients who 'refused to travel' were recorded as having a medical condition or problems with mobility. Themes were further subdivided into specific or more general conditions.

Clinical observations made by ambulance staff were routinely recorded in 229 of 397 (57.7%) cases. Electrocardiography (ECG) was included as a form of immediate management because it may have indicated that the patient's condition was potentially serious.

Over two thirds (69.8%) of patients in the sample who 'refused to travel' received immediate management at the scene (Table 3). The commonest category of immediate management where patients subsequently refused to travel were those cases where the patient was assisted off the floor, out of the bath or elsewhere, either to bed or to a suitable chair.

Immediate management was sometimes administered by others at the scene rather than the ambulance crew, for example one patient received oxygen from the fire service. Giving oral glucose to patients suffering hypoglycaemia resulted in clinical improvement obviating the need for transport. The minor theme of 'transport' reflected cases where transport was provided other than to hospital, for example, to a patient's home when there had been a fall in a public place. A small number of patients, particularly cases associated with alcohol consumption or assault, refused treatment as well as transport to hospital. Clinicians had no choice but to make a decision about the patients' competence to refuse treatment in this situation.

There was a clear statement of why a patient 'refused to travel' in 80.6% (320/397) of cases; over half of these 68.8% (220/320) reported the clinician viewpoint, 31.3% (100/320) of the statements reported reasons that the patients themselves had given. The remaining forms gave no explanation other than the patient refused. Where the reason was recorded, there were seven themes under reasons given by patients (Table 4) and four themes for reasons given by the clinicians (Table 5).

In just over half of the cases where there was a clear statement of why a patient 'refused to travel' (54.5%; see Table 5), there was either minor injury or no apparent injury. In a further quarter (24.5%) of cases the patient responded well to treatment. In 5.5% of cases the clinician recorded that there was no medical emergency and labelled this as a social problem or inappropriate call-out.

However, in only 32 of 397 (8.0%) cases did the ambulance crew explicitly record that the patient refused to travel against advice.

This study has shown that the term 'refusal to travel' was misleading as only 8% of our sample were recorded as not travelling to hospital against professional advice. The findings show that recorded reasons for completing a 'refusal to travel' form were complex and diverse. Sometimes these consisted of patient generated reasons and at other times clinically derived reasons. These were expressed variably as medical diagnoses or using lay terminology and in physical terms more often than psychological or social descriptions. In most (over 90%) cases, although a RTT form was completed, it appeared that refusal to travel was a negotiated shared decision. Although RTT was estimated to cost £1.45 million this was not an entirely wasted resource because ambulance crews did provide an on-site assessment. However, it is clear from the data that alternatives to an ambulance response based on prior telephone assessment may have been a more appropriate use of some of this resource.

One third of patients who subsequently 'refused to travel' had suffered trips, slips and falls without sustaining injury, a finding consistent with other studies [4] Management for these patients, often the frail elderly, was to lift them from the floor or bath. Telephone triage before sending an ambulance may have determined an alternative response, reducing the likelihood of transfer to hospital [11] Although meeting NHS needs, little is known of the acceptability of this approach to patient or carer and whether other psychological or environmental needs are being met.

Third-party requests against the patient's wishes resulting in RTT may have been prevented by a question incorporated into the triage process, for example "does the patient know about the call-out being made on their behalf?" If the answer is "no", an attempt could be made, for example using a triage nurse, to speak with and gather information directly from the patient, thereby potentially facilitating a more appropriate response. Obviously where the patient is unconscious or involved in a road traffic accident this is not a consideration.

Calls were labelled by clinicians as inappropriate for very few patients. Where calls were for trivial complaints or by frequent callers, mental health issues may have been an underlying factor. Calls requesting medication supplies could have been addressed by patient education.

One limitation of the study included the data source, the written clinical records, from which qualitative information which was derived. The data gleaned from these records included reasons given by patients but interpreted and documented by crews. These were sometimes brief, consisting of a few words and therefore not enough to gain any in-depth understanding of why certain phenomena were occurring. It cannot always be assumed that the interpretations of patient reasons were accurate.

The focus on physical rather than psychological or social terms by clinicians could have been due to the way the form was constructed or because of expectations of clinicians. For example, only one social problem was explicitly recorded in our sample.

Another anomaly in the recording system occurred when patients had recovered significantly since the call, which obviated the need for transport. The system at the time of the study prevented ambulance crews from doing anything other than take patients to hospital, even if their condition did not warrant it, unless a 'refusal to travel' form was signed. Not transporting after treatment of hypoglycaemia, for example, has been shown to be safe [14] with adequate follow up[15] but was not an option for crews at the time of the study. A 'treat and leave' protocol would have removed the perception that these successfully treated cases were inappropriate calls instead of appropriate requests with satisfactory outcome for patient and crew.

The strength of the study is in its careful analysis of qualitative data from a multidisciplinary including patient perspective.

New systems for responding to emergency calls, whether in the UK or in comparable health systems abroad, including the use of telephone triage, on-scene assessment and referral could lead to changes in the rate of 'refusal to travel'.

The data suggested that the processes leading to refusal to travel and non-transportation to hospital were complex. Many interacting factors were often involved, including medical, emotional and social. In-depth interviews with ambulance crews and patients would enable a fuller understanding of these factors.

There are implications for future ambulance service education and training, not just in recording information, but also shared decision making, informed consent and greater integration with other health professionals in order to develop more multi-disciplinary solutions for those cases not leading to transport to hospital.

Further research is recommended into the impact of Emergency Care Practitioners and Community Paramedics on non-transportation. These recommendations should be considered within the context of the new Government plans for widening the role of ambulance services outlined in the Bradley report [16].