Background This audit looked at the use of chest-x-rays (CXRs) in newly diagnosed HIV patients at an inner-city GUM clinic. The Departmental guidelines recommended that all newly diagnosed HIV patients had a baseline CXR. The 2011 British HIV Association guidelines (which were under consultation at the time of the audit) proposed that CXRs should only be done in patients with current or previous chest disease, high-risk for TB or in intravenous drug users.

Aims The aims of this audit were twofold, first to confirm whether all newly diagnosed patients had a CXR as per departmental guidelines, and second to review the results of the CXR in order to see whether, if any, factors predicted abnormal results.

Methods The audit considered all newly diagnosed HIV patients from 1 April 2009–31 March 2011. Data including demographic details, past medical history, health on diagnosis and details of the CXR (if performed) were collected from electronic HIV summaries and radiology records. Statistical analysis was performed using SPSS.

Results A total of 196 patients were identified, 69% of whom had had a CXR. In those who had a CXR and in whom results were available (n=132), 92% had a normal CXR and 8% abnormal. Significant predictors of abnormal CXRs included chest symptoms at diagnosis (p<0.001) and a CD4 count <200 (p=0.001). There was no significant link with the patients' country of origin but there was a pattern of association which was clinically relevant.

Conclusions In this audit all CXRs in asymptomatic patients newly diagnosed with HIV were normal. No latent chest disease was identified as a result of routine asymptomatic screening with CXRs and the practice is not justified. Following this audit clinic guidelines have been appropriately amended in keeping with current British HIV Association guidelines with the additional criteria for performing CXRs in patients presenting with CD4 <200.