Culture of all sputum samples irrespective of quality adds value to the diagnosis of pneumococcal community-acquired pneumonia in the elderly

Culture of expectorated sputum in the microbiological diagnosis of community-acquired pneumonia (CAP) is considered valid only if high-quality (HQ) samples are obtained, but evidence regarding pneumococcal etiology specifically is lacking. We studied 323 radiologically confirmed CAP cases in patients aged ≥ 65 years. Sputum samples were assessed for quality microscopically and cultured. Two quality criteria sets were applied to delineate HQ from low-quality (LQ) sputa: leukocytes/epithelial cells ratio > 5 and ≤ 2.5 epithelial cells/400× magnification field (HQ1), or leukocytes/epithelial cells ratio > 1 (HQ2). A sputum sample was obtained and the quality assessed in 224 cases; 47% were HQ1 and 76% HQ2. Encapsulated pneumococci (EPnc) were cultured in 25 (24%), 14 (12%), 35 (21%), and 4 (7%) of the HQ1-, LQ1-, HQ2-, and LQ2-samples, respectively. If another pneumococcal test (blood culture, urine antigen, or ≥ twofold increase in CbpA or PsaA antibodies) was positive, EPnc were cultured at similar proportions in HQ1- and LQ1-sputa; if the other test was negative, EPnc were cultured less often in LQ1- than HQ1-sputa. EPnc were found less often in LQ2- than in HQ2-sputa. Our results suggest similar specificity in LQ- and HQ-sputum cultures. All sputum samples add value to the pneumococcal CAP-diagnosis in the elderly. ELECTRONIC SUPPLEMENTARY MATERIAL: The online version of this article (10.1007/s10096-019-03536-9) contains supplementary material, which is available to authorized users.