ORIGINAL ARTICLE
Evaluation of B-type natriuretic peptide in patients
with community acquired pneumonia
Eman O. Arram
a,*
, Mohammed Saleh
b, Enas Hammad
ca
Chest Medicine Department, Faculty of Medicine, Mansoura University, Egypt
b
Department of Microbiology & Immunology, Faculty of Medicine, Al-Azhar University, New Damietta, Egypt
c
Department of Microbiology & Immunology, Faculty of Medicine, Mansoura University, Egypt Received 27 April 2013; accepted 12 May 2013
Available online 14 June 2013
KEYWORDS
B-type natriuretic peptide Community acquired pneu-monia
C-reactive protein Erythrocyte sedimentation rate
The area under the curve
Abstract Background: Community acquired pneumonia (CAP) is a widespread disease, brain natriuretic peptide (BNP) shown to be elevated in patients with congestive heart failure, acute myo-cardial infarction and massive pulmonary embolism, BNP levels have been found to be significantly elevated in CAP with a positive correlation to CRP and also low levels of BNP predict treatment success.
Objective: The aim of this study was to evaluate the utility of cardiac stress marker B-type natriuretic peptide in patients with CAP without sepsis or shock.
Patients and methods: 26 patients with CAP were eligible for study in addition to 10 healthy con-trol cases, cardiac stress marker BNP and inflammation marker (CRP, white blood cell count, ESR) were measured in patients with CAP.
Results: The mean serum level of initial BNP level was significantly higher (86.75 ± 25.91 pg/ ml) in patients with CAP than the control group (30.70 ± 6.77 pg/ml) with p-value <0.001 and also
* Corresponding author. Address: Chest Medicine Department, Fac-ulty of Medicine, Ali basha Ebrahim Street, Mansoura 35116, Egypt. Tel.: +20 1094195910.
E-mail address:arram1_eman@hotmail.com(E.O. Arram).
Peer review under responsibility of The Egyptian Society of Chest Diseases and Tuberculosis.
Production and hosting by Elsevier
The Egyptian Society of Chest Diseases and Tuberculosis
Egyptian Journal of Chest Diseases and Tuberculosis
www.elsevier.com/locate/ejcdt www.sciencedirect.com
0422-7638ª 2013 The Egyptian Society of Chest Diseases and Tuberculosis. Production and hosting by Elsevier B.V.
http://dx.doi.org/10.1016/j.ejcdt.2013.05.003 Open access under CC BY-NC-ND license.
significantly lower after treatment (40.15 ± 9.60) with p-value <0.001, BNP significantly positively correlated with (CRP) in patients before treatment (r = 0.709, p < 0.001) and also positively cor-related with CRP after treatment (r = 0.490, p = 0.011), there was significant negative correlation between initial BNP level and partial arterial oxygen pressure (r = 0.409, p = 0.038). The optimal cut-off for BNP was 47.0 pg/ml at admission vs 39.5 pg/ml after treatment.
Conclusion: In this study BNP levels show a transient rise in patients with community acquired pneumonia and correlate well with CRP, low levels of BNP adequately identify patients with an uncomplicated disease course.
ª 2013 The Egyptian Society of Chest Diseases and Tuberculosis. Production and hosting by Elsevier B.V.
Introduction
Community-acquired pneumonia (CAP) is a common and lead-ing infectious cause of death throughout the world. Although the mortality rate in CAP patients has decreased with wide-spread use of antibiotic therapy, the mortality rate for hospita-lised CAP patients ranges between 10% and 15%[1,2].
Brain natriuretic peptide (BNP) is a neurohormone secreted by the heart in response to myocardial stretch caused by vol-ume overload. It has been demonstrated to be a neurohumoral marker for left ventricular dysfunction and a potent prognostic marker in congestive heart failure patients[3].
BNP regulates a wide array of physiological effects includ-ing natriuresis, diuresis and vasodilatation, the main stimulus for the secretion of BNP is cardiac stress reflected by myocar-dial stretch and pressure or volume overload[4].
Recently pro-inflammatory cytokines and the activation of the sympathetic nervous system have been identified as addi-tional triggers inducing BNP secretion[5].
They have been studied for the diagnosis and prognosis of patients with congestive heart failure, myocardial infarction, pulmonary embolism and sepsis[6,7].
Consequently, BNP levels may reflect the severity of pneu-monia, displayed by pulmonary hypertension, right ventricular pressure overload and the inflammatory cytokine response, as well as the presence of disease-relevant co-morbidities, namely heart failure and renal dysfunction. In a pilot study we found promising results for the ability of BNP to predict death in se-lected patients with CAP[8].
In patients affected by an infectious disease but without sepsis or shock, study results are still controversial[9,10].
It has been shown that increased BNP levels in patients with acute phase pf Kawasaki disease significantly decrease after treatment[11].
The aim of this study was to evaluate the utility of BNP in CAP without sepsis or shock.
Patients and methods
The case control study was conducted on 26 cases with the diagnosis of community acquired pneumonia (CAP) and healthy control cases, both patients and control were cross matched for age and sex.
CAP was defined by the presence of one or several of the following recently acquired respiratory signs or symptoms, cough, sputum production, dyspnoea, core body tempera-ture > 38.0C, auscultatory findings of abnormal breath sounds and rales, white blood cell count >10 or <4· 109cells L 1and an infiltrate on chest radiograph[12].
Exclusion criteria
Exclusion criteria were history of heart failure, clinical, radio-logical, or echocardiographic evidence of heart failure, history of renal failure, chronic obstructive pulmonary disease, cirrhosis of liver, hypertensive heart disease, pregnancy, pul-monary hypertension, pulpul-monary embolism, cystic fibrosis, active pulmonary tuberculosis, hospital-acquired pneumonia, immunodepression.
All patients underwent laboratory tests at admission and at 15th day of the treatment which include white blood cell count, erythrocyte sedimentation rate, C-reactive protein, arterial blood gas (ABG), and BNP level.
CRP was analysed with a Ray Bio-Human CRP ELISA (Enzyme-linked immunosorbent Assay) KIT (Ray Biotech, Inc, USA), BNP assay was performed by the use of micro-plate-based immunoassay (Abbot laboratories, Abott park, IL).
Statistical analysis
Data were analysed using SPSS (Statistical Package for Social Sciences) version 15. Qualitative data are presented as number and percent. Comparison between groups was done by Chi-Square test. Data were presented as mean ± SD. Paired t-test was used for comparison within groups. Student t-test was used to compare between two groups. Pearson’s correlation coefficient was used to test correlation between variables. P< 0.05 was considered to be statistically significant. Results
This study was conducted on 26 cases with the diagnosis of community acquired pneumonia (CAP) and 10 healthy control cases, both patients and control were cross matched for age and sex.
Table 1 Demographic and BNP (initial) data of the case control study population.
Cases (n = 26) Control (n = 10) PValue Age (years) 48.54 ± 5.46 45.40 ± 4.90 0.122
Sex 0.763
Male 17 (65.4%) 6 (60%) Female 9 (34.6%) 4 (40%)
BNP (pg/ml) 86.73 ± 25.91 30.70 ± 6.77 <0.001 Data are represented as mean ± SD.
Table 1presents the demographic data of the case control study population as regards age and sex, initial BNP level mea-surement in the cases and control.
The mean age among CAP patients was 48.54 ± 5.46 while in the control group was 45.40 ± 4.90. The mean serum level
of initial B-type natriuretic peptide (BNP) was significantly higher (86.73 ± 25.91 pg/ml) in patients group than the con-trol group 30.70 ± 6.77 (P < 0.001) (Fig. 1).
Statistically significant decrease was observed between ini-tial and after treatment levels of BNP (86.73 ± 25.91 pg/ml
0 10 20 30 40 50 60 70 80 90 100 l o r t n o C s e s a C pg/ml
Figure 1 Initial BNP level in cases and control group.
Table 2 Comparison between initial and after treatment levels of BNP, CRP, ESR and leucocytic count in patient group .
Before treatment After treatment PValue BNP (pg/ml) 86.73 ± 25.91 41.15 ± 9.60 <0.001 CRP (mg/l) 127.19 ± 57.84 40.38 ± 11.95 <0.001 ESR (mm/h) 59.31 ± 12.70 28.77 ± 8.86 <0.001 Leucocytic count (106/ml) 16.53 ± 3.55 9.97 ± 1.99 <0.001 Significant when P < 0.05. BNP 140.00 120.00 100.00 80.00 60.00 40.00 20.00 CRP 250.00 200.00 150.00 100.00 50.00 0.00
vs. 41.15 ± 9.60 pg/ml, P < 0.001), CRP (127.19 ± 57.84 mg/ l vs 40.38 ± 11.95 mg/l, P < 0.001), ESR (59.31 ± 12.70 mm/ h vs 28.77 ± 8.86 mm/h, P < 0.001) and leucocytic count (16.53 ± 3.55· 106/ml vs 9.97 ± 1.99· 106/ml, P < 0.001)
as shown inTable 2.
Figs. 2 and 3show significant positive correlation between BNP and CRP before and after treatment (r = 0.709, P< 0.001 and r = 0.490, P = 0.011 respectively).
Fig. 4shows significant negative correlation between initial BNP and PaO2 (r = 0.409, P = 0.038). BNP_after 60.00 50.00 40.00 30.00 20.00 CRP_after 60.00 50.00 40.00 30.00 20.00
Figure 3 Correlation between BNP and CRP levels after treatment (at 15th day) in patient group.
BNP 140.00 120.00 100.00 80.00 60.00 40.00 20.00 PaO2 85.00 80.00 75.00 70.00 65.00 60.00
To evaluate the utility of BNP to predict the disease and treatment success in cases with CAP without sepsis or shock, a receiver operating characteristic (ROC) analysis was per-formed, The area under the curve in cases with CAP at admis-sion (AUC 0.98; SD 0.014; 95% CI 0.962–1.015), optimal BNP cut-off was 47.0 pg/ml, sensitivity and specificity were 96%, 100% respectively. The area under the curve in cases with CAP after treatment (AUC 0.82; SD 0.068; 95% CI 0.690– 0.956), optimal BNP cut-off was 39.5 pg/ml, sensitivity and specificity were 65%, 100% respectively as shown inTable 3, Figs. 5 and 6.
Discussion
Community acquired pneumonia (CAP) is the leading infec-tious cause of death in developed countries and remains the sixth most common overall cause of death[13,14].
Community-acquired pneumonia is prevalent worldwide. Community-acquired pneumonia in adults is commonly caused by viruses, bacteria, Mycoplasma pneumoniae and
Chy-lamidia pneumoniae. Antibiotic treatment for CAP is usually empirical because the causative pathogen is rarely identified. The majority of bacterial infections are caused by Streptococ-cus pneumoniae, Haemophilus influenzae and Mycoplasma pneumoniae and beta lactams, macrolids and flourocinolons are used frequently for treatment of CAP[15,16].
B-type natriuretic peptide (BNP) was first isolated from the porcine brain and subsequently from the heart of porcine and rats[17]. BNP is a 32 amino acid peptide and plays a role in vascular tonus and natriuresis. The mammalian heart synthes-ises and secretes BNP, which has potent diuretic, natriuretic and vascular smooth muscle-relaxing effects as well as complex interactions with the hormonal and nervous systems[18].
Recently, BNP has been investigated as a predictive bio-marker in CAP, in addition, several serum biobio-markers such as white blood cell (WBC) counts, C-reactive protein (CRP), d-dimer and procalcitonin have also been investigated as po-tential predictors of mortality of CAP[19–23].
Several studies demonstrated how BNP levels rise signifi-cantly in patients affected by sepsis[24,25].
Table 3 Prediction of pneumonia and treatment success in all patients with community acquired pneumonia. Area under the curve (AUC) of receiver operating curve (ROC) characteristic plot analysis, BNP threshold to predict pneumonia and treatment success, Sensitivity and specificity at both cut-off levels.
AUC SD 95% CI PValue Cut off BNP Sensitivity Specificity BNP at admission 0.98 0.014 0.962–1.015 0.001 47.0 pg mL 96 100 BNP after treatment 0.82 0.068 0.690–0.956 0.003 39.5 pg mL 65 100 SD, standard deviation; CI, confidence interval; BNP, B-type natriuretic peptide.
1 - Specificity 1.0 0.8 0.6 0.4 0.2 0.0 Sensitiv ity 1.0 0.8 0.6 0.4 0.2 0.0
ROC Curve
But in patients affected by an infectious disease, without sepsis or shock study results are still controversial, indeed, if some authors found elevated BNP levels in microbial infec-tions without sepsis, others did not find any difference between patients having early sepsis and control[9,10].
This study showed that mean BNP values are significantly increased transiently in CAP patients compared to control healthy cases, similar results have been found by Yetkin et al.[26]. C-reactive protein (CRP) is an acute-phase protein, which is synthesized by the liver in response to infection or tis-sue inflammation. A range of cytokines, including interleukin-6, interleukin-1b and tumour necrosis factor, stimulate its pro-duction[27]. CRP was first identified in 1930 from the obser-vation of patients with pneumonia. Its name was derived from the fact that it reacted with the pneumococcal C-polysac-charide in patients’ plasma during the acute phase of pneumo-coccal pneumonia. Subsequently, it has been recognized as a marker of inflammation, a so-called ‘acute-phase protein’. An acute-phase protein has been described as a protein whose plasma concentration increases or decreases by at least 25% during an inflammatory process[28].
Circulating levels of proinflammatory cytokines, interleu-kin-6 (IL-6), interleukin-1 beta (IL-1b) and tumour necrosis factor alpha (TNF-a) are usually elevated in pneumonia [29,30].
According to this knowledge proinflammatory cytokines might increase plasma BNP from ventricles and CRP from liver in pneumonia. It has been reported that BNP levels might be affected by increased pulmonary artery pressure and proinflammatory cytokines released from lung tissue [8,31].
Recent studies have shown significant correlation between CRP and BNP in patients with chronic renal failure and heart failure[32,33].
Investigations comparing CRP level to white blood cell count in patients with CAP, CRP levels were consistently found to superiorly diagnose pneumonia[34]. Another recent paper[35]confirmed a significant association between progno-sis and CRP and WBC. In this study, CRP and white blood cell count levels are high before treatment and significantly de-crease on 15th day of the treatment, also CRP levels signifi-cantly correlated with BNP before and after treatment.
As a significant correlation has been shown between BNP and IL-1b in patients with myocardial dysfunction, it can be speculated that IL-1b released from the alveolar macrophages might play a role in secretion of CRP and BNP. It is reason-able to expect that the correlation between CRP and BNP ex-ists in patients with pneumonia[26].
Burn et al.[36]concluded that C-reactive protein measure-ments are useful in the first week in follow up antibiotic treat-ment for severe community acquired pneumonia similar results were seen in our study.
The secretion of BNP has also been suggested to be trig-gered by hypoxia, leading to pulmonary vasoconstriction, pul-monary hypertension and right heart overload[37,38].
Patients with CAP usually suffer from partial hypoxia in the pulmonary vascular system, in our study, BNP levels neg-atively correlate with partial oxygen pressure.
Previous studies demonstrated that BNP level negatively correlated with hypoxaemia in patients with COPD[39].
Moine et al. [40]reported that decreased level of partial oxygen pressure correlated with severity of pneumonia while
1 - Specificity 1.0 0.8 0.6 0.4 0.2 0.0 Sensitivity 1.0 0.8 0.6 0.4 0.2 0.0 ROC Curve
Diagonal segments are produced by ties.
Mueller et al. [8] detected higher levels in sever pneumonia, and the highest levels of BNP were obtained in subjects who died of pneumonia.
Also initial high level of BNP that decreased after antibiotic treatment may be explained by proinflammatory cytokines [26].
In this study, patients with CAP without sepsis or shock and after exclusion of any diseases can affect BNP level, the cut-off level to predict improvement was 39.5 pg mL, sensitiv-ity and specificsensitiv-ity were 65%, 100% respectively.
While Christ et al.[10]concluded that, BNP cut-off level to predict improvement in 302 CAP patients with sepsis or shock was 95.5 pg mL while sensitivity and specificity were 47.8%, 85.8% respectively, this difference may be attributed to small number of this study.
In conclusion, BNP levels show a transient rise in patients with community acquired pneumonia, and correlates well with CRP, so we can use BNP as an acute phase reactant in pneu-monia, and can be used as predictor of treatment success, the limitation of this study is the small number of patients caused by limited availability of appropriate subjects free of major dis-eases which can be accused as a cause of increase in the level of BNP.
Reference
[1]G.L. Colice, M.A. Morley, C. Asche, et al, Treatment costs of community-acquired pneumonia in an employed population, Chest 125 (2004) 2140e45.
[2]M.J. Fine, M.A. Smith, C.A. Carson, et al, Prognosis and outcomes of patients with community-acquired pneumonia: a meta-analysis, JAMA 275 (1996) 134e41.
[3]A. Palazzuoli, G. Antonelli, I. Quatrini, R. Nuti, Natriuretic peptides in heart failure: where we are, where we are going, Intern. Emerg. Med. 6 (2011) 63–68.
[4]T. Breidthardt, K. Laule, A.H. Strohmeyer, et al, Medical and economic long-term effects of B-type natriuretic peptide testing in patients with acute dyspnea, Clin. Chem. 53 (2007) 1415– 1422.
[5]M.S. Weinfeld, G.M. Chertow, L.W. Stevenson, Aggravated renal dysfunction during intensive therapy for advanced chronic heart failure, Am. Heart J. 138 (1999) 285–290.
[6]M. Brueckmann, G. Huhle, S. Lang, et al, Prognostic value of a plasma N-terminal pro-brain natriuretic peptide in patients with severe sepsis, Circulation 112 (2005) 527e34.
[7]J. Mair, A. Hammerer-Lercher, B. Puschendorf, The impact of cardiac natriuretic peptide determination on the diagnosis and management of heart failure, Clin. Chem. Lab. Med. 38 (2001) 571e88.
[8]C. Mueller, K. Laule-Kilian, A. Scholer, A.P. Perruchoud, B-type natriuretic peptide for risk stratification in community-acquired pneumonia, J. Intern. Med. 258 (2005) 391–393. [9]E. Kandil, J. Burack, A. Sawas, et al, B-type natriuretic peptide:
a biomarker for the diagnosis and risk stratification of patients with septic shock, Arch. Surg. 143 (2008) 242–246.
[10]M. Christ-Crain, T. Breidthardt, D. Stolz, et al, Use of B-type natriuretic peptide in the risk stratification of community-acquired pneumonia, Intern. Med. 264 (2008) 166–176. [11]T. Kawamura, M. Wago, H. Kawaguchi, M. Tahara, M. Yuge,
Plasma brain natriuretic peptide concentrations in patients with Kawasaki disease, Pediatr. Int. 42 (2000) 241–248.
[12]R.R. Watkins, T.L. Lemonovich, Diagnosis and management of community- acquired pneumonia in adults, Am. Family Phs. 83 (2011) 1299–1306.
[13]M.S. Niederman, Guidelines for the management of community acquired pneumonia. Current recommendations and antibiotic selection issues, Med. Clin. North Am. 85 (2001) 1493–1509. [14]E.M. Mortensen, C.M. Coley, D.E. Singer, et al, Causes of
death for patients with community-acquired pneumonia: results from the pneumonia patient outcomes research team cohort study, Arch. Intern. Med. 162 (2002) 1059–1064.
[15]T.J. Marrie, Therapeutic implications of macrolide resistance in pneumococcal community-acquired lower respiratory tract infections, Int. J. Clin. Pract. 58 (2004) 769–776.
[16]V. Dudas, A. Hopefl, R. Jacobs, et al, Antimicrobial selection for hospitalized patients with presumed community-acquired pneumonia: a survey of nonteaching US community hospitals, Ann. Pharmacother. 34 (2000) 446–452.
[17]T. Sudoh, T. Kangava, N. Minamino, et al, A new natriuretic peptide in porcine brain, Nature 332 (1988) 78–81.
[18]Y. Saito, Y. Nakao, H. Itoh, Brain natriuretic peptide is a novel cardiac hormone, Biochem. Biophys. Res. Commun. 158 (1989) 360–368.
[19]R.P. Smith, B.J. Lipworth, I.A. Cree, et al, C-reactive protein A clinical marker in community-acquired pneumonia, Chest 108 (1995) 1288e91.
[20]R.A. McIvor, Plasma d-dimer for outcome assessment in patients with CAP: not a replacement for PSI, Chest 126 (2004) 1015e16.
[21]J.M. Querol-Ribelles, J.M. Tenias, E. Grau, et al, Plasma d-dimer levels correlate with outcomes in patients with community-acquired pneumonia, Chest 126 (2004) 1087e92. [22]D.T. Huang, L.A. Weissfeld, J.A. Kellum, et al, Risk prediction
with procalcitonin and clinical rules in community-acquired pneumonia, Ann. Emerg. Med. 52 (2008) 48–58.
[23]M. Masia, F. Gutierrez, C. Shum, et al, Usefulness of procalcitonin levels in community-acquired pneumonia according to the patients outcome research team pneumonia severity index, Chest 128 (2005) 2223–2229.
[24]F. Post, L.S. Weilemann, C.M. Messow, C. Sinning, T. Mu¨nzel, B-type natriuretic peptide as a marker for sepsis-induced myocardial depression in intensive care patients, Crit. Care Med. 36 (2008) 3030–3037.
[25]M. Falcone, S. Corrao, M. Venditti, P. Serra, G. Licata, Performance of PSI, CURB-65, and SCAP scores in predicting the outcome of patients with community acquired and healthcare-associated pneumonia, Intern. Emerg. Med. 65 (2011) 431–436.
[26]O. Yetkin, S.S. Hacievliyagil, H. Gunen, et al, Assessment of B-type natriuretic peptide in patients with pneumonia, Int. J. Clin. Pract. 62 (2008) 488–491.
[27]B. Clyne, J.S. Olshaker, The c-reactive protein, J. Emerg. Med. 17 (1999) 1019–1025.
[28]J.J. Morley, I. Kushner, Serum c-reactive pritein levels in disease, Ann. N. Y. Acad. Sci. 389 (1982) 406–418.
[29]A.J. Puren, C. Feldman, N. Savage, et al, Patterns of cytokine expression in community acquired pneumonia, Chest 107 (1995) 1342–1349.
[30]J.J. Bouwman, F.L. Visseren, P.K. Bouter, R.J. Diepersloot, Azithromycin inhibits interleukin-6 but not fibrinogen production in hepatocytes infected with cytomegalovirus and chlamydia pneumoniae, J. Lab. Clin. Med. 144 (2004) 18–26. [31]Q. He, M.C. LaPointe, Interleukin-1b regulation of the human
brain natriuretic peptide promoter involves Ras-, Rac-, and p38 kinase dependent pathways in cardia myocytes, Hypertension 33 (1999) 283–289.
[32]O. Ortega, P. Gallar, M. Munoz, et al, Association between C-reactive protein levels and N-terminal pro-B-type natriuretic peptide in pre-dialysis patients, Nephron. Clin. Pract. 97 (2004) 125–130.
[33]M.S. Sabatine, D.A. Morrow, J.A. de Lemos, et al, Multimarker approach to risk stratification in non-ST elevation acute coronary syndromes: simultaneous assessment of troponin I, C-reactive protein, and B-type natriuretic peptide, Circulation 105 (2002) 1760–1763.
[34]R.P. Smith, B.J. Lipworth, I.A. Cree, E.M. Spires, J.H. Winter, C-reactive protein. A clinical marker in CAP, Chest 108 (1995) 1288–1291.
[35]S. Kruger, S. Ewig, R. Marre, et al, Procalcitonin predicts patients at low risk of death from community-acquired pneumonia across all CRB-65 classes, Eur. Respir. J. 31 (2008) 349–355.
[36] A.H.W. Burns, J.J. Oosterheert, E. Hak, A.I.M. Hoepelman, Usefulness of consecutive c-reactive protein measurements in follow up sever community-acquired pneumonia, Eur. Respir. J. (2008) 26–732.
[37]W.E. Hopkins, Z. Chen, N.K. Fukagawa, C. Hall, H.J. Knot, Le- Winter MM. Increased atrial and brain natriuretic peptides in adults with cyanotic congenital heart disease: enhanced understanding of the relationship between hypoxia and natriuretic peptide secretion, Circulation 109 (2004) 2872–2877. [38]L.B. Yap, D. Mukerjee, P.M. Timms, H. Ashrafian, J.G. Coghlan, Natriuretic peptides, respiratory disease, and the right heart, Chest 126 (2004) 1330–1336.
[39]T. Ando, K. Ogawa, K. Yamaki, et al, Plasma concentrations of atrial, brain, and C-type natriuretic peptides and endothelin-1 in patients with chronic respiratory diseases, Chest 110 (1996) 462–468.
[40]P. Moine, J.B. Vercken, S. Chevret, P. Gajdos, Severe community acquired pneumococcal pneumonia. The French study group of community-acquired pneumonia in ICU, Scand. J. Infect. Dis. 27 (1995) 201–206.
