• bronchoscopy
• pneumocystis pneumonia

We read with interest the case report by Newton et al¹ describing platypnoea and orthodeoxia in a patient with pneumocystis pneumonia (PCP). We would like to add our own experience and question whether the observation has an important clinical implication.

A 29 year old patient who had been born and raised in Zambia presented 10 months after arrival in the UK with a history of cough and progressive dyspnoea. A chest radiograph showed bilateral diffuse lung shadowing, maximal in the perihilar regions and lower zones, but with relative sparing of the lung apices. Pao₂ breathing air was 9 kPa. PCP was suspected and fibreoptic bronchoscopy and BAL performed. Sedation was with 2 mg midazolam. The bronchoscopy, which was performed with the patient sitting, was complicated by a fall in Sao₂ to 85% which was only partially corrected by supplemental oxygen. Pneumocystis was identified in the BAL fluid sample and treatment commenced with high dose co-trimoxazole. The day after the bronchoscopic examination, while supine and breathing air, the patient had an Sao₂ of 98%. Rising to the sitting position led to a steady fall in Sao₂ to 94%, which rapidly corrected to 99% when the patient returned to the supine position.

Our patient showed orthodeoxia similar to that described by Newton et al. This became clinically important when bronchoscopy was performed in the sitting position. Predominant mid and lower zone shadowing is a common radiographic feature in PCP² and we wonder whether the phenomenon of platypnoea and orthodeoxia may be widespread in such patients. A study to identify the frequency of this finding would be worthwhile and easy to do. We suggest that supine and sitting Sao₂ measurement should be routine before bronchoscopic investigations in such patients and that bronchoscopy in the supine position might be the approach of choice for patients with suspected PCP.