Systemic review: the accuracy of lactic dehydrogenase in the diagnosis of pneumocystis pneumonia / 中华危重病急救医学

ABSTRACT

Objective To study the accuracy of lactic dehydrogenase (LDH) in the diagnosis of pneumocystis pneumonia (PCP). Methods The data of this systemic review was retrieved from the PubMed, China Biology Medicine disc, Wanfang, Weipu and China National Knowledge Infrastructure (CNKI) databases from establishment till to October 31st, 2017. Case-control studies about the diagnosis of PCP were enrolled. Enrolled studies were required that patients in case group ware PCP and patients in control group were lung diseases other than PCP. The QUADAS tool was used to evaluate the quality of studies. The RevMan 5.3 software was used to draw a forest plot. The StataMP 14 software was used to make subgroup analyses by drawing receiver operator characteristic (SROC) curves for the whole group, the acquired immune deficiency syndrome (AIDS) group, and the not all-AIDS group, and calculating their diagnostic odds ratio (DOR) and 95% confidential interval (95%CI). Results Thirteen studies, all in English, were included. There were 825 patients in the case group, in which 650 patients were AIDS. There were 1 341 patients in control group, in which 888 patients were AIDS and most of them were Pulmonary Kaposi Sarcoma, bacterial pneumonia, pulmonary tuberculosis etc. Although there were different positive values of LDH in different studies, from 200 U/L to 598 U/L, sensitivities were good, especially in AIDS patients all values were above 80% (80%-100%). The specificities had big fluctuations, from 6% to 85%, which made them poor. The DOR (95%CI) of LDH in PCP diagnosis of all patients, AIDS patients and not-all AIDS patients were 6.73 (3.19-14.21), 9.17 (3.79-22.18) and 5.07 (1.30-19.80) respectively. Conclusions The sensitivity of LDH in the diagnosis of PCP is high, especially in AIDS group. In practice if LDH is negative, there should be more evidences to support the treatment of PCP.