Preoperative assessment of patients for thoracic surgery is a multidisciplinary process designed to offer appropriate surgical treatment with acceptable risk. UK guidelines for pulmonary resection associated with malignant disease involved review of available evidence concerning operative risk. Patients displaying cardiopulmonary physiological parameters above previously recommended threshold values remain classified as acceptable risk. However, less certainty exists about the utility of predicted postoperative pulmonary function values and preoperative performance status to confer unacceptable risk. These guidelines suggest a tri-partite risk assessment combining risks of operative mortality, perioperative adverse cardiac events and postoperative dyspnoea, to be discussed by the multidisciplinary team and with the patient.

The driver for the development of cardiopulmonary bypass (CPB) was the desire to operate on a still heart in a blood-free surgical field. The concept was first applied in the animal laboratory by Gibbon in 1935. Research continued until the first clinical application of CPB by Dennis in 1951, in which the patient died owing to surgical complications. In 1953, Gibbon successfully closed an atrial septal defect, using his design of heart–lung machine, and triggered the evolution of present-day cardiac surgery.

Purpose of reviewPulse oximetry is ubiquitous but detailed understanding of the technology is poor. This is illustrated by publications addressing knowledge of pulse oximetry and those warning against the use of transmission pulse oximeter sensors in a reflectance manner, unintended by the manufacturers, owing to the potential for iatrogenic problems. Reflectance oximetry sensors are distinct and their application rather specific. Users must adhere to the manufacturer's guidelines to be assured of approximating the claimed accuracy and other specifications. Moreover, a thorough understanding of the device's shortcomings will optimize performance and avoid misuse. Cautious skepticism is appropriate with use of any technology but particularly with indirect measures of vital signs.Recent findingsThe studies of reflectance sensors described here illustrate a diversity of successful applications and opportunities for further research. The genesis of applications for some sensors, for example fetal sensors, has proven helpful in other clinical settings where low mean arterial pressure and need for accurate monitoring of a SpO(2) of less than 80% is poorly provided by transmittance sensors. Reflectance sensors are more prone to placement over contaminating sources (for example arteries and pigmentation), but their more sophisticated design can provide greater versatility than transillumination methods.SummaryThis invited review highlights recent developments and applications of reflectance oximetry with an emphasis on the potential clinical and research advantages.

Patients presenting for thoracic surgery often have concomitant cardiac disease, and because many are smokers, they may have related complications. The authors explain how to predict and assess cardiac risk factors and how to assess the respiratory system in these patients. The most valid single test for predicting postoperative lung function is the predicted postoperative forced expiratory volume in one second.

Many patients with respiratory disease present for anaesthesia. Clinical examination should elicit signs of respiratory pathology, especially airway obstruction and impaired gas exchange. A chest radiograph is mandatory in most patients. The authors give detailed advice on how to deal with smokers and patients who have chronic obstructive airway disease, asthma, upper respiratory tract infection, obstructive sleep apnoea, pulmonary hypertension or cystic fibrosis.

Respiratory support in the adult respiratory distress syndrome (ARDS) remains a difficult area. In severe ARDS, survival rates rarely exceed 50%. Death is most commonly due to sepsis and multiorgan failure (MOF). Unsupportable respiratory failure as the primary cause of death is unusual, a 16% incidence reported in 1983 was revealed unchanged when repeated at the same institution in the early 1990s.1,2Such statistics allow no complacency, however, in our attitude to respiratory support which, as a purely supportive endeavour, must be achieved without contributing to disease progression or the onset of sepsis or MOF. The potential for mechanical ventilation to cause or exacerbate existing lung damage has stimulated the search for improved respiratory support technology. In addition, reassessment of goals in respiratory support and more aggressive manipulation of cardiopulmonary interactions have contributed to considerable recent changes in our approach to respiratory support in ARDS.