4  Discussion
When a wide range of sociodemographic, clinical, social and proximal factors are examined in the same sample population, the results of this study indicate that different factors are associated with the two different components of pre-hospital delay. This may be important in helping to develop interventions which target specific components of delay. Our findings support previous research showing that decision time accounts for almost two thirds of pre-hospital delay [19,20]. Shorter decision times were associated with being married, attributing symptoms to a heart attack, and symptom onset in the presence of a bystander. Home-to-hospital delays were shorter in patients who had other symptoms as well as chest pain, particularly non-pain symptoms. Two factors predicted both short decision times and home-to-hospital delays: the type of ACS and symptom onset away from home. Previous studies showed that both attribution of symptoms to a heart attack and diagnosis of an STEMI predicted shorter decision times and total pre-hospital delays [4,12,21,22] and this combination of factors may help to reduce decision time. Patients who experienced both STEMI and a greater number of non-pain symptoms (nausea, breathlessness etc) which may be more salient than pain, were more likely to have short home-to-hospital delays, a finding that has not been reported before. Home-to-hospital delays were not associated with attribution of symptoms to a heart attack, but to the number and variety of symptoms. Patients who are found to have a STEMI possibly experience symptoms that are perceived as serious, and this may increase their motivation to reach hospital quickly. The type of ACS may therefore be associated with shorter delays in both components but for different reasons.
Moser et al. [1] have argued that social and proximal factors such as being married, having a bystander present and location at symptom onset play an important role in the cognitive aspects of symptom appraisal. Our findings support this, showing that both being married and the presence of a bystander predicted short pre-hospital delay, and are associated specifically with the decision time component. Perhaps not surprisingly, a bystander is more likely to be present if patients are married. Previous studies have shown conflicting findings concerning the role and relationship of the bystander in pre-hospital delay, since although patients tend to have a short decision time if a bystander is present, this is not the case if the bystander is a relative because family members, particularly spouses, often recommend strategies that increase delay [6,8]. Our results support previous findings showing that patients who are at home when symptoms begin are likely to have longer pre-hospital delays [8,23], indicating that the relationship between the context in which cardiac symptoms occur and pre-hospital delay is complicated and warrants detailed investigation. Educational and counseling interventions aimed at reducing decision delay have been found to increase patients knowledge of ACS [24] but it may be useful if future interventions were to incorporate ways of optimizing bystander assistance, particularly among relatives of patients previously identified as at risk of ACS.
Our study highlights the critical role of making initial contact with the EMS in promoting short pre-hospital delays, since the results show that patients who contacted the EMS were more likely to have short decision times. Patients experiencing acute cardiac symptoms are recommended to contact the EMS directly, not only as the quickest mode of transport to hospital but also because the EMS are trained and equipped to treat life threatening cardiac arrhythmias which often accompany acute cardiac symptoms. Patients who use the EMS receive reperfusion therapy more promptly than others [1]. Previous research has shown that patients are often reluctant to use the EMS because they do not believe that their symptoms are serious [5,6,13,25]. In our study, attribution of symptoms to a heart attack was not a predictor of contacting EMS, however, patients who reported greater intensity of pain were more likely to contact the EMS. The experience of intense pain may override doubts about the cause or how serious the symptoms are and patients' reluctance to seek emergency care. We also observed that contact with the EMS was more likely when symptoms began at the weekend, perhaps because patients are unwilling to contact their physicians at the weekend thus limiting their choices. The decision to contact the EMS may therefore reflect broader cognitive and social aspects of symptom appraisal and coping involved in the decision making process.
This study has some limitations. The sample size was relatively small in comparison with some other studies. The average age of patients was younger than in many contemporary clinical cohorts, with a higher proportion of men and patients admitted with STEMI compared with NSTEMI/UA [26,27]. The exclusion of patients with co-morbidities that can affect symptom presentation, mood or troponin measurements is likely to have excluded older patients. Only patients with chest pain were included due to the requirements of the larger study [17], leading to the possibility that a proportion of patients who presented without chest pain were excluded. However, despite these factors, the study cohort remains comparable to similar recent studies of pre-hospital delay [11,12,28]. It was necessary to collect data retrospectively and to restrict analyses to patients who could recall the time of symptom onset. Although every effort was made to interview patients soon after admission, data may have been affected by recall bias. Only survivors could be interviewed thus factors that influenced delay in individuals who did not survive ACS may not have been captured. We carried out our study in the UK where there is no charge for health care at the point of contact and in a densely urban environment, and different factors may operate in rural areas.

4.1  Future recommendations
Pre-hospital delay remains a significant obstacle to improving treatment for ACS. Further work investigating how patients and their relatives appraise symptoms in the two component phases of delay may help in the development of more effectively targeted interventions. Interventions aimed at helping patients and their families appraise symptoms more accurately and seek help more quickly are likely to reduce delay times here. The pattern of symptoms, including both pain and non-pain symptoms played an important part in reducing delay in the home-to-hospital component, suggesting that wider educational interventions aimed at the general public as well as clinical staff in recognizing the atypical symptoms of ACS may help to improve the rapidity of hospital admission. Interventions should focus on challenging unhelpful illness beliefs, and providing information and education concerning atypical patterns of symptoms in relation to ACS and be made more widely available to include patients' families, health care professionals and the general public.