Resuscitation following OHCA
Presenting heart rhythms
When the heart rhythm is first monitored after OHCA a patient may present with a shockable rhythm (ventricular fibrillation or ventricular tachycardia - treated by delivering an electric shock using a defibrillator) or non-shockable rhythm (asystole or pulseless ventricular activity and bradycardia) on the electrocardiogram ( ECG)  . The treatment and prognosis depend on presenting heart rhythm with better survival after OHCA with shockable rhythm , .
Information about presenting heart rhythm was available for 91.1% of our cohort. The proportion of OHCA cases with a shockable presenting heart rhythm was 25.1% (Figure 5). Other OHCA registries and academic studies have reported a similar proportion of shockable rhythms  . Some report lower proportions of shockable rhythms and also a trend towards diminishing proportions of patients with shockable rhythms over recent years  . Part of the reason for our relatively high proportion of shockable presenting rhythms may be the fact that survivors are likely to be over-represented in our linked dataset.
Figure 5: Distribution shockable and non-shockable heart rhythms by gender
Figure 5 also shows that a larger proportion of males present with a shockable heart rhythm during their OHCA. This is in agreement with earlier published figures  . In our dataset both males and females between 45 and 65 years of age were most likely to present with a shockable rhythm.
Bystander Cardiopulmonary Resuscitation
The positive effect on survival of the early start of cardiopulmonary resuscitation ( CPR) is well established [32-35] . In the ideal situation, a bystander starts performing chest compressions on the OHCA victim as soon as possible.
Figure 6 shows that in just over 40% of OHCA cases where the Scottish Ambulance Service attempted resuscitation it is recorded that a bystander started CPR before arrival of the ambulance service. It is an idiosyncrasy of the way CPR data is collected on the electronic patient report form used by the Scottish Ambulance Service that it is unclear whether in the remaining 60% of cases there was no bystander CPR observed or the data is simply missing or unknown.
Scotland's 40% bystander CPR rate is low compared with the numbers published for other centres, including some parts of England  . For example it is substantially lower than the 60% bystander CPR reported by the London Ambulance Service in their 2015/2016 annual report  . The relatively low overall survival rate in Scotland during the period our data was collected, lends plausibility to the suggestion that our bystander CPR rate is substantially lower than other parts of Europe with better survival outcomes. It is also possible that because our 2011-15 dataset bystander CPR information was often missing or ambiguous, the actual percentage of bystander CPR may be higher, and our findings simply reflect significant under-reporting.
A fundamental tenet of Scotland's OHCA Strategy is that increasing bystander CPR rate is crucial to increasing survival. It is therefore a priority to improve completeness of information about bystander CPR. Work is underway to capture baseline bystander CPR data for the period 2011-2015 by extracting additional data relating to telephone guided CPR initiated by emergency call handlers from the Scottish Ambulance Service control centres ( see section about future work for more details).
Figure 6: Total proportions of known and unknown/missing bystander CPR given
More detailed analysis show that the likelihood of receiving bystander CPR differs between subgroups of people in the dataset. The upper left panel of Figure 7 shows that the likelihood of receiving bystander CPR is higher among males. This data conflicts with some of the current literature reporting higher bystander CPR rates in witnessed OHCA among females  , but are in line with other published findings  . The current literature describing bystander CPR rates across age groups is inconsistent. Some authors report higher proportions receiving bystander CPR among younger patients while others report higher proportions among older patients , . Our data show the highest proportions with a record of bystander CPR among the youngest age groups (Figure 7, left upper panel). Figure 7, bottom left panel, shows that patients from deprived areas ( SIMD1) are less likely to receive bystander CPR. This is in line with a recently published study from Taiwan showing that patients who suffer from an OHCA in a low socio-economic status area are less likely to receive bystander CPR compared with more affluent areas  . The figure in the right bottom panel shows a trend towards increasing proportions of records of bystander CPR over time. This trend towards more bystander CPR in Scotland in the years before the launch of the Scottish OHCA strategy might be partly explained by an international increase in emphasis on the importance of bystander CPR. It is also likely that the recording of bystander CPR has improved in recent years. Our overall increase in bystander CPR over time is concordant with observations from other centres , .
Figure 7: Percentages of OHCA with a record of receiving bystander CPR by age, sex, SIMD quintiles and year of OHCA