Sir,

Vitreous haemorrhage is known to occur by a wide variety of mechanisms. We present the case of a 27-year-old gentleman who developed a vitreous haemorrhage following cardiopulmonary resuscitation (CPR). We postulate that this occurred as a result of the chest compressions performed, through a mechanism similar to valsalva retinopathy. To our knowledge this is the first reported case of vitreous haemorrhage arising in this way.

Case report

A 27-year-old man presented to the accident and emergency department with sudden onset of severe retrosternal chest pain. There was no relevant past medical or ocular history and no previous history of chest pain. He was a cigarette smoker. Initial electrocardiography (ECG) showed ST segment elevation in the anterior chest leads. Shortly afterwards the patient collapsed on a trolley having had a ventricular fibrillation cardiac arrest. He was shocked into asystole and after 1 mg of adrenaline intravenously and CPR, he regained a perfusing rhythm. He was transferred for cardiac angiography, which showed 100% occlusion of his left anterior descending (LAD) coronary artery. The decision was taken not for thrombolysis but angioplasty instead. A proximal occlusion of the LAD was readily canalised and two stents were inserted. The patient was then transferred to the intensive care unit where he received intravenous dobutamine and noradrenaline for resuscitation of presumed stunned myocardium.

Following extubation 8 days later he complained of blurred vision in his right eye. He was subsequently reviewed in the ophthalmology department. Snellen visual acuities were 3/60 right eye and 6/4 left eye. He was found to have a dense vitreous haemorrhage and no fundal view in the right eye. Ultrasound revealed a flat retina. The haemorrhage failed to resolve and 3 months following myocardial infarction he underwent right vitrectomy. At surgery, dilatation of peripapillary vessels with adherent vitreous haemorrhage was noted, but no other retinal abnormalities. He made an uneventful recovery and postoperative visual acuity in his right eye was 6/6. Dilated fundoscopy since the surgery has not revealed any associated pathology.

Comment

This case describes a presumed iatrogenic vitreous haemorrhage. In the absence of contributing ocular pathology, the only causative factor we can identify is the cardiac compressions the patient received during his resuscitation.

Chest compressions during CPR are known to cause several complications, for example rib fractures, pulmonary haemorrhages and liver lacerations. However, ocular complications are more unusual. Chest compression by safety belt in automobile accidents is known to cause traumatic retinal angiopathy (Purtscher's disease) – even from a minor compression injury from a lap-shoulder belt1 – and retinal detachment.2 We also note the recent report by Chandra et al3 of a case of Purtscher's retinopathy in one eye, and Valsalva retinopathy in the other, following a compressive chest injury.

The mechanism and efficacy of chest compressions as part of CPR have become better understood in recent years. There has been deliberation as to whether sternal compression generates a direct increase in arterial forward flow by cardiac compression, or whether it is the relaxation phase that encourages venous return through negative intrathoracic pressure. However, it is now generally thought that both mechanisms are acting to provide some cerebral perfusion, though the flow generated is insufficient to force vessel aneurysm or rupture. Babbs4 used a mathematical model to show that even the most efficient compression technique would only generate a maximum of 3.1 l/min and 58 mmHg of systolic perfusion pressure. Pinming et al5 confirmed this by performing transoesophageal echocardiography during closed-chest CPR, in cardiac arrest. Peak forward aortic flow was measured at 58.8±11.6 cm/s; less than half of normal peak aortic flow.

Valsalva retinopathy is a well-described phenomenon. A preretinal haemorrhage in the macula area is the usual finding at presentation. The cause is thought to be a rupture of a retinal vessel, following physical exertion – usually a sudden and rapid rise in intrathoracic pressure (such as when coughing, heavy lifting or straining at stool). This is thought to cause a rise in the intravenous pressure, then a rise in retinal vessel intraluminal pressure, with consequent rupture of a retinal venule or capillary, perhaps at the site of a pre-existing vessel wall weakness.6 Extension of the haemorrhage into the vitreous is rare, but is thought to occur if the force of the intrathoracic pressure rise is sufficiently high7 or in the presence of a blood disorder such as sickle cell disease.8

In our case, we can rule out other causes of vitreous haemorrhage. There was no previous ocular pathology, no retinal pathology was identified during vitrectomy surgery and the patient had not been thrombolysed, or given any anticoagulant therapy. We therefore hypothesise that this patient's vitreous haemorrhage was due to the chest compressions received during CPR, via a mechanism similar to valsalva retinopathy.