Rapid aetiological diagnosis of pneumonia based on routine laboratory features

The values of some basic laboratory features on admission to hospital were recorded and compared in 418 adult patients with community-acquired pneumonia, namely erythrocyte sedimentation rate, C-reactive protein, white blood cell (WBC) count, serum lactate dehydrogenase (S-LD), serum alanine-aminotransferase, and serum sodium. Discriminant analysis was performed to obtain an aetiological diagnosis. WBC value of greater than 15 x 10(9)/l strongly indicated a bacterial and, especially a pneumococcal aetiology, whereas increased S-LD could imply a mycoplasmal infection. For patients less than 50 years of age the equation C2 = -1.788 + 0.204 x WBC-0.0909 X S-LD was constructed, in which C2 greater than 0 indicated a pneumococcal aetiology. This function correctly classified 31/33 (93.9%) patients with a mycoplasmal and 20/31 (64.5%) patients with a pneumococcal infection. Patients with viral, Haemophilus influenzae or chlamydial infection could not be discriminated from each other. The age of the patient, WBC and possibly S-LD on admission are easily accessible parameters and these results could therefore be of value in daily clinical practice in hospitals.

Pneumococcal antigens in sputa: ELISA for the detection of pneumococcal C-polysaccharide in sputa from pneumonia patients

An improved ELISA, the LKB UltroBact Pneumococcus Kit detecting pneumococcal C-polysaccharide, has been tested. Sputum samples from 72 patients with community acquired pneumonia were included in the study. The sensitivity obtained was 96.1% and the specificity 92.6%. This ELISA might offer a useful diagnostic method in major clinical microbiologic laboratories for demonstrating Streptococcus pneumonia in sputa from patients with pneumonia.

Etiologic diagnosis of pneumonia by antigen detection: crossreactions between pneumococcal C-polysaccharide and oral microorganisms

Crossreactions between bacteria occurring more or less frequently in the respiratory tract were investigated using an enzyme-linked immunosorbent assay (ELISA) developed for the detection of pneumococcal C-polysaccharide. A collection of 218 strains was investigated: 30 Streptococcus pneumoniae, 120 alpha-streptococci, and 68 strains representing other species. Strong crossreactions were observed with 36% of the alpha-streptococci and with two of 11 Staphylococcus aureus strains. The collection of alpha-streptococci consisted of 90 fresh clinical isolates and 30 stock strains. Almost all crossreactions of alpha-streptococci were found among the clinical isolates. Among the stock strains only one of four Streptococcus mitis strains was positive. Pneumococcal C-polysaccharide and phosphorylcholine inhibited the reactions in ELISA with monoclonal antibodies against pneumococcal C-polysaccharide, as well as with a polyclonal antiserum against pneumococcal C-polysaccharide. We suggest that the cross reactions between alpha-streptococci and pneumococci depend on the presence of phosphorylcholine as a common antigenic determinant. The crossreaction in the ELISA with some Staphylococcus aureus strains may be explained by the presence of protein A binding to the Fc portion of the antibodies. When the 10 alpha-streptococci that showed the strongest crossreactions and ten pneumococci representing different types were tested in different concentrations the absorbance values were lower for most alpha-streptococci compared with the pneumococci. This explains that false positive results with alpha-streptococci do not seem to constitute a practical problem in this ELISA developed for detection of pneumococcal C-polysaccharide in samples from patients with pneumonia.

A new coagglutination test for detecting pneumococcal C-polysaccharide

A new coagglutination test (PnC-CoA) for detecting pneumococcal C-polysaccharide (PnC) was compared with a commercial kit for detecting capsular polysaccharide using sputum samples from 105 patients with pneumonia. The sensitivity obtained with PnC-CoA was 95.8% and with the commercial kit 83.3%; the specificity was 96.5% and 91.2%, respectively. The PnC-CoA is simple to perform and it is a rapid, sensitive and specific test for detecting Streptococcus pneumoniae in sputa from adult patients with pneumonia.

Antibiotic therapy in pneumonia: a comparative study of parenteral and oral administration of penicillin

An open, randomized study of treatment of radiologically verified community-acquired pneumonia is described. 33 patients were treated with phenoxymethylpenicillin orally in an average dose of 2 g every 8 h, and 36 patients were treated intravenously with benzylpenicillin 3 g every 8 h. When temperature was normalized the antibiotic dose was reduced in both groups to oral phenoxymethylpenicillin in an average dose of 1 g every 8 h. 24 and 26 patients in respective groups completed 10 days of therapy. No statistically significant differences between the two groups were found when compared for duration of fever, hospital stay, CRP, ESR, leukocyte counts and X-ray normalization. In spite of the low number of patients included in this study it gives a clear indication that treatment of pneumonia with penicillin by the oral route results in the same outcome as parenteral treatment when patients suffering from vomiting, diarrhoea and severe illness with respiratory distress are excluded. In addition, oral treatment is cheaper than parenteral and more simple to manage.

Aetiology of community-acquired pneumonia in hospital treated patients

From May 1982 a prospective 1-year study of adult patients with community-acquired, radiologically verified, hospital treated pneumonia was performed at the Department of Infectious Diseases, Orebro Medical Center Hospital, Orebro, Sweden. The study included 147 patients with a median age of 71 years. Special efforts to diagnose a pneumococcal aetiology were accomplished by antigen detection of the pneumococcal C-polysaccharide (PnC) in sputum and saliva samples and by serological methods for determination of antibody titres against PnC. A pneumococcal aetiology was established in 46.9% of the patients, including 8.1% with double infections. Altogether Haemophilus influenzae A virus were noted in 9.5%, respectively, Mycoplasma pneumoniae in 5.4%, legionnaires' disease in 2.7% and Branhamella catarrhalis in 2.0%, whereas enteric gram-negative bacilli as aetiological organisms were not found in any patient. These findings imply that penicillin should still be the first drug of choice in hospitalized adult patients with community-acquired pneumonia in Sweden.

Diagnosis of pneumococcal pneumonia by detection of antigen in saliva

A sandwich enzyme-linked immunosorbent assay was used for the detection of pneumococcal C polysaccharide in saliva samples from patients with radiologically verified pneumonia. In 16 of 29 patients (55%) with pneumococcal pneumonia, as verified by conventional culture methods, the antigen was present in saliva specimens. The test was negative in 35 of 36 patients (97%) with non-pneumococcal pneumonia and in all saliva samples from a control group, consisting of 25 patients with no signs of respiratory disease. It is concluded that the detection of pneumococcal C polysaccharide in saliva offers a valuable complement to conventional diagnostic methods for pneumococcal pneumonia.

Comparison of enzyme-linked immunosorbent assay with coagglutination and latex agglutination for rapid diagnosis of pneumococcal pneumonia by detecting antigen in sputa

A sandwich enzyme-linked immunosorbent assay detecting the species-specific pneumococcal C polysaccharide was compared to latex agglutination and a coagglutination test which detected capsular pneumococcal antigens in sputum specimens with regard to specificity and sensitivity. Specimens from 52 patients with clinical and radiological evidence for pneumonia were tested. Twenty-one patients with Streptococcus pneumoniae isolated in sputum and 31 patients with a non-pneumococcal etiology were included. The predictive values for a positive test by enzyme-linked immunosorbent assay was 0.91 and for a negative test 0.97, by latex agglutination 0.90 and 0.91, and by coagglutination 0.84 and 0.85 respectively; these values did not show a statistically significant difference. Whereas agglutination tests are technically more simple and can be performed more rapidly, the enzyme-linked immunosorbent assay has the advantage of detecting pneumococcal C polysaccharide, an antigen common to all pneumococci. Thus it provides an interesting alternative to tests based on serum containing antibodies to all 83 different capsular polysaccharides.

Determination of antibodies to pneumococcal C polysaccharide in patients with community-acquired pneumonia

The pneumococcal C polysaccharide (PnC) is species specific and believed to be a cell wall component of all capsular types. Antibodies against PnC in human sera have been demonstrated previously, but the question of whether a rise in these antibodies occurs during pneumococcal infections has not been investigated. We used an indirect enzyme-linked immunosorbent assay (ELISA) for the estimation of PnC antibodies in 124 hospital-treated patients with pneumonia. In 3 of 6 patients with pneumococcal bacteremia and in 17 of 44 patients with S. pneumoniae isolated in the blood, sputum, or nasopharynx, a significant rise in antibody levels was recorded, accounting for a sensitivity of 38.6%. Of 35 patients with pneumonia of other known or suspected etiology, 1 gave a positive result, corresponding to a specificity of 97.1%. In addition, 3 of 8 patients with PnC antigen in the sputum as the only etiological finding and 5 of 37 patients with unknown etiology gave positive results. The PnC antibodies did not seem to have any protective capacity against pneumonia caused by pneumococci. The ELISA, in which only one antigen preparation was used, was more simple than other tests in which traditional capsular antigen preparations are used. It might therefore be used as a supplemental method in the diagnosis of pneumococcal pneumonia. The problems involved in expressing serum titers obtained with the ELISA are discussed.

Detection of C polysaccharide in Streptococcus pneumoniae in the sputa of pneumonia patients by an enzyme-linked immunosorbent assay

The pneumococcal C polysaccharide (PnC) is species specific and believed to be a cell wall component of all pneumococcal types. A sandwich enzyme-linked immunosorbent assay (ELISA) for detection of PnC in sputa has been developed by using a monoclonal antiphosphorylcholine antibody and a polyclonal rabbit anti-PnC antiserum in the test system. A 1-year study of adult hospitalized patients with community-acquired pneumonia was performed. A total of 147 patients with clinical and radiological evidence for pneumonia were accepted for the study. Of these, 105 patients provided a sputum sample upon admission to the ward. The sputa were cultured semiquantitatively as well as tested for the presence of antigen. Of the sputum samples from patients with Streptococcus pneumoniae, 27 of 33 (accounting for a sensitivity of 82%) were positive in the ELISA test. Of the sputum samples from patients with pneumonia of some other known or suspected etiology, 32 of 34 (accounting for a specificity of 94%) were negative. In addition, 7 sputum samples from 31 patients with pneumonia of unknown etiology were positive. The ELISA test described here is in our opinion a sensitive and specific test for detecting PnC from S. pneumoniae in sputa from patients with untreated pneumonia.

Cross-reactions between alpha-streptococci and Omniserum, a polyvalent pneumococcal serum, demonstrated by direct immunofluorescence, immunoelectroosmophoresis, and latex agglutination

In recent years several groups have used serological methods to demonstrate pneumococcal capsular antigens in sputum. In the present study 123 strains of alpha-hemolytic streptococci (including 97 strains from sputum or pharyngeal specimens) were tested for cross-reactions with a polyvalent antipneumococcal serum (Omniserum). Representatives of the following species were included: Streptococcus bovis, S. equinus, S. intermedius, S. lactis, S. milleri, S. mitis, S. mutans, S. sobrinus, S. salivarius, S. sanguis, S. suis, and Aerococcus viridans. Serological reactions were detected by direct immunofluorescence, immunoelectroosmophoresis, and latex agglutination. Fifteen (12%) of the strains gave positive reactions by all three methods. Positive reactions were also observed with another 32 strains (26%) with two of the methods, whereas 37 strains (30%) gave positive reactions by just one technique. Altogether 84 (68%) strains gave positive reactions with one or more of the methods. Latex agglutination gave positive reactions with 26 (21%) strains compared with 57 (46%) in immunofluorescence and 63 (51%) in immunoelectroosmophoresis. Absorption of the antiserum with one alpha-hemolytic strain reduced but did not entirely eliminate the cross-reactions with five tested strains. These findings indicate a potential risk of cross-reactions with polyvalent antipneumococcal serum in tests carried out on sputa or other specimens which may be contaminated with alpha-hemolytic streptococci.

Studies on the determination of plasma prolactin

A double antibody solid-phase (DASP) and a plyethylene glycol (PEG) radioimmunoassay method for plasma prolactin determination were developed and evaluated as to their reliability. The iodination of prolactin was carried out using a lactoperoxidase sorbent with iodination conditions being optimized by studying the effect of reaction buffer pH in the range 5.0--7.0 at two ionic strengths. With the DASP-technique, intra-assay coefficient of variation was 6% (n = 30) and the inter-assay coefficient of variation 12% (n = 12), and with the PEG-technique, intra-assay coefficient of variation was 10% (n = 10). The sensitivity, with the DASP system, was 0.098 ng/tube and the detection limit for plasma, 0.8 microgram/l. The use of the PEG-technique was abandoned because of poor precision and nonparallelism of plasma dilution curves with the standard curve. The whole mean basal plasma value was 34.0 +/- 19.5 microgram/l (mean +/- 2SD, n = 35). The response after administration of 200 mg sulphiride was an increase of between 119--220% of basal values and after administration of 100 mg of chlorpromazine, an increase of 85--140%. After administration of TRH (thyrotropine releasing hormone) an increase of between 60 and 330% was obtained, the mean increments after 20 min in outpatients being 87.0 microgram/l and in hospitalized subjects, 51.2 microgram/l.

A study of factors promoting and inhibiting lactation

The present series of studies, made between 1972 and 1975, has shown that existing routines in the maternity ward, such as weighing the baby before and after breast feeding, can inhibit the establishment of lactation and increase the number of early failures. Giving information to the father in the maternity ward may be an important factor in promoting breast-feeding, as judged from the findings on duration of breast feeding and from the mothers' comments. The greatest effect observed was that of skin-to-skin and suckling contact during the first hour after delivery, which increased the median duration of breast feeding by 2 1/2 months. These findings illustrate that existing routines during the neonatal period should be re-examined as to their influence in promoting or inhibiting breast feeding.