aLiverpool School of Tropical Medicine, Liverpool, UK; bThe Royal Liverpool and Broadgreen University Hospital, Liverpool, UK; cArrowe Park Hospital, Wirral, UK; and dKenyatta Hospital, Nairobi, Kenya.
Received: 28 March 2002; revised: 9 May 2002; accepted: 20 May 2002.
Pneumonia is a common and serious cause of ill health in sub-Saharan Africa, particularly in HIV-infected adults. Streptococcus pneumoniae is the most frequently identified aetiological agent responsible for up to 80% of cases [1,2], with HIV co-infection present in between 30 and 90% of cases [3,4]. The diagnosis of pneumococcal pneumonia is presumptive, based on clinical assessment and lobar/segmental consolidation present on the chest X-ray. Microbiological confirmation is usually unavailable. Using clinical and radiological findings, benzyl penicillin, an effective, cheap and readily available antibiotic, can be commenced with confidence of a good clinical response. However, several case reports have described atypical presentations of pneumococcal pulmonary infections [5,6]. This has cast doubt on the reliability of clinical and radiological findings to identify pneumococcal pneumonia correctly, as has occurred with tuberculosis in regions of high HIV prevalence.
We compared by HIV status, the chest radiographic findings in 122 consecutive admissions to the Kenyatta National Hospital Nairobi with pneumococcal pneumonia. Pneumonia was defined by an acute (less than 28 days’ duration of symptoms) febrile (axillary temperature > 37.4°C) respiratory illness, with lower respiratory tract signs and pulmonary parenchymal abnormality on chest X-ray. Pneumonia was said to be pneumococcal in aetiology if S. pneumoniae was isolated from the blood or from a lung aspirate. HIV testing was performed after informed consent. The study had local institutional ethical clearance.
Chest radiographs were reviewed by two independent radiologists in the UK who were blind to HIV status. In the case of discrepancies between reports, the two radiologists met and reached a consensus view.
All but seven of the 122 cases were blood culture positive (95%), the remaining seven were diagnosed on the results of a lung aspirate. HIV co-infection was present in 62 individuals. HIV-infected individuals were more likely to die or have a dual infection with non-typhi Salmonella bacteraemia or tuberculosis (Table 1).
Pleural effusions and multilobar involvement were a frequent finding. Effusions were usually unilateral (n = 21, 78%) and occupied less than one third of a hemithorax (n = 23, 85%). Consolidation affected multiple lobes in over half of the cases (53%). No significant differences were found in the X-ray appearances by HIV status (Table 1). Mediastinal lymphadenopathy was present in seven of the cases with no association with HIV but a strong association with tuberculosis, five out of seven, P < 0.01. Interstitial or alveolar patterns of consolidation and cavitation were not seen in any of the X-rays. There was no association between multilobar involvement or effusion and inpatient outcome.
These findings suggest that the radiological features of bacteraemic pneumococcal pneumonia are not significantly different in HIV-infected and uninfected adults in Kenya and the appearances are in keeping with conventional descriptions of pneumococcal pneumonia. Cavitation, pneumothoraces and interstitial patterns of consolidation, reported infrequently as a feature of pneumococcal infection, were not seen in either group.
The individuals in this study were all hospital inpatients and in most cases were bacteraemic. An aetiological diagnosis based solely on sputum isolation of pneumococcus was excluded to avoid the poor specificity associated with this technique. They thus represent severe cases of pneumococcal pneumonia, and it would be unwise to generalize our findings to less severe cases.
HIV staging data (CD4 cell counts or clinical parameters) were not available for the HIV-infected group, and it may be argued that the failure to document a difference in radiographic appearances is a consequence of the HIV-infected group being relatively immunocompetent. It is only with advanced HIV disease that the radiographic appearances of tuberculosis become more atypical. This explanation would seem unlikely. Pneumococcal disease rates increase with advancing immunosuppression. Previous reports from Nairobi would predict that more than half of the HIV-infected cases in this study would have CD4 T cell counts less than 300 . The frequency of non-typhi Salmonella bacteraemia would support this prediction.
S. pneumoniae–Mycobacterium tuberculosis co-infection occurred in seven cases (5.7%) with a non-significant excess in the HIV-infected, rates similar to previous reports from Africa [2,8]. The presence of mediastinal lymphadenopathy, independent of HIV status, was strongly associated with this co-infection, and this radiological finding in the setting of acute pneumonia in Africa should precipitate a thorough search for M. tuberculosis irrespective of HIV status.
In summary, we found no significant difference between the radiographic appearances of pneumococcal pneumonia in HIV-1-infected and uninfected Kenyan adults. In clinical practice lobar and segmental consolidatory changes on chest X-ray with or without a pleural effusion can be used to support a diagnosis of pneumococcal pneumonia irrespective of HIV status. Variations from this pattern, particularly the presence of mediastinal lymphadenopathy, would suggest dual or alternative pulmonary pathology other than S. pneumoniae, of which tuberculosis is the most important.
The authors would like to thank the patients and staff of the Kenyatta National Hospital for their help with this study.
Sponsorship: This study was supported by The Wellcome Trust, UK.
1.Aderaye G. Community acquired pneumonia in adults in Addis Ababa: etiologic agents and the impact of HIV infection. Tuberc Lung Dis 1994, 75:308–312.
2.Scott JA, Hall AJ, Muyodi C, et al. Aetiology, outcome, and risk factors for mortality among adults with acute pneumonia in Kenya. Lancet 2000, 355:1225–1230.
3.Gilks CF, Brindle RJ, Otieno LS, et al. Life-threatening bacteraemia in HIV-1 seropositive adults admitted to hospital in Nairobi, Kenya. Lancet 1990, 336:545–549.
4.Feldman C, Glatthaar M, Morar R, et al. Bacteremic pneumococcal pneumonia in HIV-seropositive and HIV-seronegative adults. Chest 1999, 116:107–114.
5.Rodriguez Barradas MC, Musher DM, Hamill RJ, Dowell M, Bagwell JT, Sanders CV. Unusual manifestations of pneumococcal infection in human immunodeficiency virus-infected individuals: the past revisited. Clin Infect Dis 1992, 14:192–199.
6.Collazos J, Fernandez A, Mayo J, Martinez E. Pulmonary gangrene secondary to pneumococcal pneumonia in a patient with AIDS. Clin Infect Dis 1997, 24:268–269.
7.Gilks CF, Ojoo SA, Ojoo JC, et al. Invasive pneumococcal disease in a cohort of predominantly HIV-1 infected female sex-workers in Nairobi, Kenya. Lancet 1996, 347:718–723.
8.Allen SC. Lobar pneumonia in Northern Zambia: clinical study of 502 adult patients. Thorax 1984, 39:612–616.
© 2002 Lippincott Williams & Wilkins, Inc.