Background: Usual practice in lower respiratory tract infections (LRTIs) includes administering complementary oxygen. The effectiveness of oxygen therapy and different methods of delivery is unknown. This review contributes to the rational use of oxygen in the treatment of LRTIs.
Objectives: To determine in the treatment of LRTIs: the effectiveness of oxygen therapy and oxygen delivery methods; the safety of these methods; and indications for oxygen therapy.
Search strategy: We searched the Cochrane Central Register of Controlled Trials (CENTRAL) (The Cochrane Library, 2008, issue 2); MEDLINE (January 1966 to March 2008); EMBASE (1990 to December 2007); and LILACS (January 1982 to March 2008).
Selection criteria: Randomised controlled trials (RCTs) comparing oxygen versus no oxygen therapy or methods of oxygen delivery for hypoxaemic LRTIs in children (3 months to 15 years of age). To determine indications for oxygen therapy, observational studies were included.
Data collection and analysis: We assessed 551 titles. No studies comparing oxygen versus no oxygen were found. Four RCTs comparing delivery methods and 12 observational studies assessing the accuracy of clinical signs indicating hypoxaemia were eligible. A meta-analysis of the RCTs comparing oxygen delivery methods was performed.
Main results: Three studies assessed the effectiveness of nasal prongs (NP) versus nasopharyngeal catheters (NPC). The pooled estimate effect showed no differences (OR 0.96; 95% CI 0.48 to 1.93) in treatment failure (number of children failing to achieve adequate SaO2). One study compared the effectiveness of NP versus nasal catheter (NC). No differences were found in treatment failure (the mean number of episodes of desaturation/child: NC group 2.75, SD +/- 2.18 episodes/child; NP group 3, SD +/- 2.5 episodes/child, p = 0.64). Another study compared face mask (FM) and head box (HB) versus NPC. Use of FM showed lower risk of treatment failure (failure to achieve PaO2 > 60 mmHg) than the NPC (OR 0.20; 95% CI 0.55 to 0.88). As did the use of HB compared with NPC (OR 0.40; 95% CI 0.13 to 1.12).Studies assessing the accuracy of signs and/or symptoms indicating hypoxaemia showed that cyanosis, grunting, difficulty in feeding and mental alertness have better specificity in predicting hypoxaemia and its results were consistent among studies.
Authors' conclusions: NP and NPC seem to be similar in effectiveness and safety when used in patients with LRTI. There is no single clinical sign or symptom that accurately identifies hypoxaemia. Studies identifying the most effective and safe oxygen delivery method are needed.