Pattern of aerobic sputum bacteriology in respiratory tract infections and pneumonias in adultsSR Kamat, GH Tilve, VR Hoskote, Parvin S Anklesaria, UK Sheth
Department of Chest Medicine, K.E.M. Hospital and Seth G.S. Medical College, Parel, Bombay 400012, India
A 4 year study on 93 pneumonias, 70 respiratory tract infections and 20 normals for sputum bacteriology is reported. In controls, mainly Streptococcus pyogenes Staphylococcus and Branhamella catarrhalis were seen and frequently duplicated. In patients, there was a predominance of gram negative bacteria, mainly Enterobacter, Pseudomonas, Klebsiella species. There was some change in the individual frequency over these years, probably attributed to changes in the patterns of antibiotic usage in the general community.
Since the introduction of antibiotics, the pattern of organisms causing infections has changed. This is well documented by Finland. , from data derived from monitoring of cases with bacteraemia in Boston, U.S.A. His main findings are that the number of admissions with bacteraemia over 30 years from 1935 has increased in adult by 3˝ times. Case fatality rates dropped from 60% in 1935 to 30% in 1947 and rose to 40% in 1961. The mortality was the lowest between 10-29 years of age and then increased steadily. The striking feature in his report was a persistent change in pathogenic bacteria which now are gram negative bacilli mainly enteric organisms such as E. coli, Proteus, Pseudomonas and Enterobacter species.
The change in the pattern in respiratory infections was confirmed by Tillotson and Lerner,  and Sullivan et al.  Comparable data has not come from India but Sainani and Fulambarkar  have reported Pneumococci in 44%, Streptococcus haemolyticus in 10%, Staphylococcus in 31% and Klebsiella, H. influenzae, E. coli, Pseudomonas in 1 to 3 per cent of pneumonia cases.
In this institute, a prospective survey of cases of lobar pneumonia in adults was done from 1972 to 1974 in 108 cases. For standardising technical methods in 1975, we did duplicate cultures in normals, in cases with acute 'cold' as outpatients and in patients of chronic lung disease admitted for acute exacerbations. The bacteriological aspects will be discussed in this report.
From patients admitted to the medical wards with history of fever, respiratory symptoms and radiographic evidence of segmental or lobar consolidation, 108 patients were selected over 3 years. Cases (a) with a suspicion of malignancy (b) with obvious tuberculosis (one case from first series, three & four cases respectively from the second & third series were omitted as they were found to have Pulmonary Tuberculosis at a later date). (c) who had received chemotherapy in previous 3 days, (d) who were pregnant, (e) with age below 12 years and (f) who were moribund were excluded. Two patients died from the first series in 24 hours. One and four cases respectively from series first and second were uncooperative and were not included in the series. In each case the sputum sample was collected on admission by a doctor after thorough clearing of the nose and mouth. Then the patient was made to gargle with sterile saline and an expectorate from the chest was collected. Except in the first 29 patients, another sample was collected and processed after 48-72 hours.
The control group consisted of 20 normal persons. In addition, 20 persons with acute `cold' and 50 patients of chronic lung disease with acute lower respiratory exacerbations were also studied [Table 1]. The normals and "acute cold" patients consisted basically of hospital employees. The latter had basic diagnosis as chronic bronchitis (16), bronchiectasis (17), bronchial asthma (5), lung abscess (6), and miscellaneous (6).
The sputum sample was plated directly on superimposed blood agar and chocolate agar plates and incubated at 37°C for 24 hours. A smear was prepared and stained by Gram stain for all samples except in category 3 of [Table 1]. These were classified as Gram positive, negative or mixed organisms. The isolated colonies were inoculated (in nutrient broth and glucose broth for Streptococci). The broth culture was inoculated and further identification was done according to the method of Cowan & Steel.  Sensitivity test was done by the method of Bauer et al  on a 6 hour broth culture.
[Table 1] lists the number of cultures studied and the organisms isolated. In controls, the frequency of bacteria per culture seemed higher probably because of Branhamella, catarrhalis.
Bacteriology in control groups
In normals, the predominant organisms were Streptococci, Staphylococci and Branhamella catarrhalis [Table 2].
There were only two isolates of other Gram negative bacteria. In acute `cold' group, while organisms were similar, there were 10 isolates of other Gram negative bacteria (4 Klebsiella).
Thus it appeared that with the presence of infection, a greater frequency of Gram negative bacteria is seen. For control group, our techniques in bacterial isolation were reliable; in 31 of 40 subjects (78%) the same bacteria were isolated [Table 3]. In the majority of the control smears showing Gram positive bacteria, we isolated mixed organisms on culture [Table 4]. Of 14 cases from the normal group with mixed culture status, 12 cultures were due to Branhamella catarrhalis. In acute `cold' group, of 13 with mixed culture status, 8 were due to Branhamella catarrhalis.
Results of bacterial drug sesitivity are shown in [Table 5]. In controls, while majority of strains of Steptococcus pyo genes were sensitive to penicillin, all were sensitive to streptomycin and chloramphenicol. A third of Streptococcal strains seemed resistant to new antibiotics (ampicillin, cephalexin and gentamycin).
Similar trends were seen for Staphylococci but a large proportion were sensitive to gentamycin.
The trends for other bacteria were similar but numbers were too small for any reliable conclusion.
Bacteriology in Respiratory Infections
Over 4 years at this hospital, Streptococcus pyogenes is becoming increasingly frequent [Table 2]. Thus, successive proportional percentages are 16, 2, 17 and 46. Streptococcus pneumoniae seemed uncommon, Staphylococci were isolated in a small proportion, being most frequent in 1974. Branhamella catarrhalis, was uncommon in cases with pneumonia but was seen often in acute infective exacerbations, with chronic lung disease.
Over the period, infection with Klebsiella has become commoner; so also Enterobacter, especially in pneumonia group. Pseudomonas was commoner in chronic lung cases but was not seen in pneumonia cases. In the latter group coliforms were seen oftener.
Duplicate cultures were done in 7 cases of the first series and in all cases of the second and third pneumonia series. Thus out of 64 patients with pneumonia, in 20 (31%) duplicate cultures showed the same organisms.
Of 42 patients with acute infective exacerbations, 17 (40%) showed same organisms on second culture.
In 1973, Staphylococci, Klebsiella and Enterobacter were predominant while in 104, they were Streptococci and Enterobacter: in 1975, these were Streptococci, Branhamella catarrhalis and Pseudomonas.
[Table 4] reveals the smear culture correlation in 93 pneumonia cases.
[Table 5] shows results of drug sensitivities in pneumonia cases.
Streptococcus pyogenes: A large proportion of Streptococcus pyogenes were resistant to penicillin, sulphadiazine and epicillin; a majority were sensitive to ampicillin, gentamycin and cephalexin while almost all were sensitive to streptomycin, tetracycline and chloramphenicol.
Staphylococci: Almost all strains were resistant to penicillin, less often to epicillin, sulphadiazine and ampicillin. Most strains were sensitive to streptomycin tetracyline, chloramphenicol and gentamycin. Most strains of Branhamella catarrbalis, were sensitive to sulphadiazine, epicillin, streptomycin. tetracycline and chloramphenicol.
Klebsiella: The strains seemed resistant to penicillin, sulphadiazine, ampicillin and epicillin. The best drug seemed to be streptomycin or chloramphenicol.
Enterobacter aerogenes: The strains formed the largest group which did not seem sensitive to pencillin, sulphadiazine, streptomycin and ampicillin. A few were sensitive to epicillin and cephalexin, while most were suseptible to chloramphenicol. The miscellaneous Gram negative group showed sensitivity to streptomycin then to tetracycline and chloramphenicol. The strains were frequently resistant to penicillin, ampicillin, epicillin, sulphadiazine and many times to gentamycin.
From India we are unaware to such serial data over 4 years on patterns in bacterial isolations from sputum in pneumonia patients, control group and miscellaneous respiratory infections.
These results reveal that bacteria causing respiratory infections in urban community of Bombay have changed. As we have excluded cases which had received chemotherapy in preceding 3 days, this may not be the immediate cause of these changes in flora. But the trends seen over 4 years may have a bearing upon the prevailing practice of antibiotic usage in the community. The work from abroad quoted earlier ,, supports this interpretation. As hospital inpatients are a selected sample, our data can not represent a true status in the community.
It is interesting to note that Streptococcus pneumoniae was uncommon in our study. We checked our bacteriologic techniques in a control group and found that normals (Presumably with oropharyngeal secretions) showed different organisms which were consistently isolated. In pneumonia cases, consistent double isolation was seen in 35 per cent. Over 3 years, there was a changing pattern with a predominance of gram negative bacteria.
Drug sensitivity patterns revealed that majority of strains were surprisingly resistant to penicillin, sulphadiazine and ampicillin. As epicillin is related to the latter, resistance to it may be cross linked. It is possible that this pattern may be related to over usage of these agents over a prolonged period in urban community.
The newer antibiotics, viz. cephalexin also showed resistance for several strains but emergence of similar phenomenon to streptomycin, tetracycline and chloramphenicol is not evident. This data may have relevance to a rational approach to drug therapy in our situation. Similar changes in the pattern has been shown by Wysocki and Drunner  from Germany. They showed in cultures from wound and urine, that Staphylococci, E. coli, Pseudomonas aeruginosa, Aerobacter and Proteus group formed the major organisms.
We thank Dr. C. K. Deshpande, Dean for allowing us to carry out this work. Kind help given by Prof. K. G. Nair and various residents is appreciated. Dr. H. C. Barbhaiya helped in obtaining drugs and other help.
[Table 1], [Table 2], [Table 3], [Table 4], [Table 5]