Documente Academic
Documente Profesional
Documente Cultură
D. J. M U S C A T E L L O "*, K.-A. O G R A D Y #, K. N E V I L L E $ J. M A N U L T Y %
" Communicable Diseases Sureillance and Control Unit, and NSW Public Health Training Program,
NSW Health Department, Locked Bag 961, North Sydney NSW 2059, Australia
# Centre for Population Health, Macquarie Area Health Serice, 62 Windsor Parade, East Dubbo,
NSW 2830
$ Sydney Childrens Hospital, High Street, Randwick, NSW 2031
% Communicable Diseases Sureillance and Control Unit, NSW Health Department, Locked Bag 961,
North Sydney, NSW 2059
SUMMARY
Acute poststreptococcal glomerulonephritis (APSGN) is an inflammatory kidney condition that
can complicate Group A streptococcal infections. Two clusters of APSGN occurred recently in
New South Wales (NSW), Australia ; one in a rural town in December 1999 and the other in a
Sydney suburb in January 2000. We interviewed carers of the affected children but found no
common exposures except three of the Sydney cases were cousins in frequent contact. To assess
the probability of these clusters occurring, we analysed hospital admissions for acute
glomerulonephritis, as a proxy for APSGN in younger patients. The incidence of acute
glomerulonephritis in NSW during 1989\901997\8 in residents aged under 20 years was
2n2\100 000\year (95 % CI 2n02n5). Incidence was highest in children aged 59 years, boys and
Aboriginal children. We found no evidence for other clusters during that period. The recent
clusters highlight the continued potential for unexpected future outbreaks of APSGN.
blood and proteins to permeate into the urine. The Table 1. Case definition for acute poststreptococcal
process by which the glomeruli become injured during glomerulonephritis
streptococcal infection is uncertain, but is believed to
Case definition
be related to the bodys immune response to the
All four of the following criteria are required for a
invading organism. Presenting symptoms are typical clinical case. For a sub-clinical case, only criteria 2, 3 and
of any form of acute nephritic syndrome : haematuria, 4 are required.
proteinuria, oliguria, oedema, hypertension and acute 1. Clinically compatible illness with one or more of :
renal failure. APSGN can be distinguished from other $ Oedema (swelling of the face or limbs).
$ Macroscopic haematuria (visibly dark urine).
forms of glomerulonephritis by evidence of recent
$ High blood pressure (diastolic 80 mmHg if 13
Group A streptococcal infection through serological
years of age or younger, or 90 mmHg if older
markers of streptococcal infection, and low serum C3 than 13 years).
complement. Prognosis is generally good, with most 2. Microscopic glomerular haematuria
patients recovering completely with only supportive $ If urine sent to a laboratory for microscopy : red
therapy [6,7]. blood cells (RBC) 10\ml.
The epidemiology of APSGN in a modern, de- $ If urine tested using dipstick urinalysis :
(584) were excluded. The ISC has been fully enum- none had evidence of skin sores. None of these cases
erated since July 1993, and partially enumerated for were related, or had any apparent common exposure,
some hospitals prior to July 1993. A sample weighting except that the two younger children attended the
factor on the data base [14] was used to estimate total same school and had been swimming at the same local
hospitalizations in those hospitals that contributed pool during the week preceding their throat
sampled data prior to this date. However, 99 % of the symptoms. They had no contact with each other at
acute glomerulonephritis admissions were to fully school. One child was admitted to hospital and the
enumerated hospitals, and therefore the influence of others were treated in the outpatient department
sampling would be small. Interstate admissions of (Table 2).
NSW residents were included. In Sydney, the cluster involved four boys aged
To better estimate the true population incidence of under 10 years living in the same suburb, three of
acute glomerulonephritis leading to hospitalization, whom were cousins and of Polynesian background.
we removed obvious multiple admissions of the same The first boy hospitalized was unrelated to the three
person by deleting records with the same age in years cousins, attended a different school and had no
(to two decimal places), sex and Statistical Local Area apparent common exposure. The three cousins each
[15] of residence occurring within 60 days of the initial lived in separate households but played together
hospitalization. The departure status of each ad- regularly. Only one of the Sydney cases had evidence
mission episode aided in verifying multiple admissions of a throat infection (peritonsillar abscess). The first
arising from patient transfers. boy hospitalized had a preceding purulent lesion on
Probabilities for counts of hospital admissions for his arm. The three cousins had evidence of possible
acute glomerulonephritis were calculated assuming a skin infection related to insect bites or superficial
Poisson distribution of the admissions over time ; for injury (Table 2). One case was subclinical.
all NSW the frequency distribution of admissions per
month was not significantly different from a theor-
Hospital admissions
etical Poisson distribution with the same average
frequency (goodness of fit : # l 0n16, P l 1n0). The There were 347 admissions for acute glomerulo-
confidence intervals of crude admission rates were nephritis in patients aged under 20 years between July
calculated using the exact method for the Poisson 1989 and June 1998. Males comprised two-thirds of
distribution. Confidence intervals for proportions the admissions in the period. Admissions were
were calculated using the exact method for the concentrated in patients aged 59 years (n l 149,
binomial distribution [16]. Analysis was performed 43 %) (Table 3), and peaked at age 7 years (n l 39,
using SAS statistical software [17]. 11 %). The number of admissions declined from 47 in
Knoxs method was applied to the hospital ad- 1989\90 to 35 in 1990\1, increased to 53 in 1992\3
mission data to assess the tendency of acute glomerulo- and then declined steadily to 20 admissions in 1997\8
nephritis in NSW to cluster in time and space (Table 3). Eight percent (95 % CI 511 %) of
(geographically) [16]. The method involves a pair-wise admissions were of people identified as being of
comparison of each event with every other event Aboriginal or Torres Strait Islander origin (Table 3).
to determine whether each pair is adjacent in time This compares with only 3 % of the total NSW
and\or space. In our analysis, adjacent in time was population aged under 20 years identifying as being
defined as two admissions occurring 31 days or less of Aboriginal or Torres Strait Islander origin in the
apart, and adjacent in space was defined as two ad- 1996 Australian census [18]. The principal diagnosis
missions occurring in residents of the same Statistical was proliferative glomerulonephritis (which includes
Local Area. APSGN) in two-thirds and unspecified glomerulo-
nephritis in almost one-third of the admissions in
this age group (Table 3).
RESULTS Admissions for acute glomerulonephritis occurred
at an average annual rate of 2n2 (95 % CI 2n02n5) per
Cluster investigation
100 000 residents aged under 20 years across NSW
The rural cluster consisted of two girls and a younger over the 9-year period to June 1998. The rates in the
boy aged 519 years and living in the same rural town. regions in which the clusters occurred were un-
All three reported preceding throat symptoms, and remarkable (the rural region ; Macquarie Health Area :
368
D. J. Muscatello and others
Table 2. Description of cases of acute poststreptococcal glomerulonephritis according to the case definition applied
Clinically compatible illness Microscopic Recent streptococcal infection ?
haematuria
Area Health Age Date of Visibly Diastolic BP Red blood cells ASO serology Anti-DNase B Serum C3
Service group onset of dark on admission (no.i10'\ml) Characteristics (IU\ml) serology complement (g\l)
and case no. (years) Sex APSGN Oedema urine (normally 80) (normally 10) and date of onset Skin swab Throat swab (normally 200*) (normally 170) (normally 0n831n70)
Macquarie
1 1014 F 5 Nov 1999 No Yes 80 10100 Throat infection n.d. n.d. 426 (24\11\99) ; 200 n.s.a.
31 Oct 1999 397 (05\12\99)
2 1519 F 10 Dec 1999 Yes Yes 85 10100 Throat infection n.d. n.d. 858 n.d. n.s.a.
23 Nov 1999
3 59 M 13 Dec 1999 Yes Yes 80 100 (14\12\99) ; Throat infection n.d. n.d. 4449 (14\12\99) ; 1600 n.s.a.
100 (23\12\99) 24 Nov 1999 4864 (20\12\99)
Central Sydney
1 59 M 8 Jan 2000 Yes Yes 87 100 Purulent arm lesion n.d. kve 300 680 (12\1\00) ; 0n21
25 Jan 1999 320 (13\1\00)
2 59 M 16 Jan 2000 Yes Yes 90 100 Dry knee wound\ n.d. kve 200 680 0n16
insect bites
1 Jan 2000
3 59 M 26 Jan 2000 Yes Yes 80 100 Peritonsillar abscess n.d. kve 100 (26\1\00) ; 680 0n56 (25\1\00) ;
17 Jan 200. 100 (24\1\00) 0n45 (24\1\00)
4 04 M Asymptomatic No No 71 11100 (29\1\00) ; Insect bites 1 Jan n.d. kve 100 1360 0n39
100 (31\1\00) 2000. Became wet
itchy sores
BP, blood pressure ; kve, no pathogens found ; n.d., not done ; n.s.a., no serum available; ASO, antistreptolysin O antibody to Group A streptococci titre ; anti-DNase B,
anti-deoxyribonuclease B antibody to Group A streptococci titre.
*ASO titres may be 200 or more in healthy, school-age children.
Streptococcal glomerulonephritis in NSW 369
Age (years)
04 63 18n2 (14n222n6)
59 149 42n9 (37n748n3)
1014 78 22n5 (18n227n2)
1519 57 16n4 (12n720n8)
Sex
Male 225 64n8 (59n669n9)
Female 122 35n2 (30n140n4)
Aboriginality
Non-Aboriginal 298 85n9 (81n889n4)
Aboriginal\Torres Strait Islander 28 8n1 (5n411n5)
Not stated\missing 22 6n3 (4n09n4)
Year of admission
1989\90 47 13n5 (10n117n6)
1990\1 35 10n1 (7n113n7)
1991\2 42 12n1 (8n916n0)
1992\3 53 15n3 (11n719n5)
1993\4 48 13n8 (10n417n9)
1994\5 41 11n8 (8n615n7)
1995\6 32 9n2 (6n412n8)
1996\7 29 8n4 (5n711n8)
1997\8 20 5n8 (3n68n8)
Type of acute glomerulonephritis
Proliferative (including APSGN) 227 65n4 (60n270n4)
Rapidly progressive 3 0n9 (0n22n5)
Other specified 14 4n0 (2n26n7)
Unspecified 103 29n7 (24n934n8)
Total 347 100n0
5
December (Fig. 1).
4
The clustering of cases in Sydney in January was
3 highly unusual ; the mean number of admissions in
2 January of residents of the inner Sydney region
1 (Central Sydney Health Area) aged under 20 years
0 over the 9-year period was 0n22 (95 % CI 0n00n8).
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Using the mean and its confidence limits as the
Month
parameters of Poisson distributions, the probability of
Fig. 1. Average, minimum and maximum number of three or more admissions in January in inner Sydney
admissions for acute poststreptococcal glomerulonephritis
was 0n002, with a range of 0n0000n047. With only one
in each month from July 1989 to June 1998 in persons aged
under 20 years in NSW. #, minimum ; i, average ; , of the three cases from the rural cluster being
maximum. admitted, we were unable to determine from hospital
370 D. J. Muscatello and others
admission data whether the observed cluster in that In the absence of literature describing the epi-
area was unusual, although we are unaware of other demiology of APSGN in modern developed settings,
such clusters of non-hospitalized cases occurring in the hospital admissions data provided a readily
NSW. For residents aged under 20 years of the central available means by which to describe quickly its
western region of NSW (Macquarie Health Area), the epidemiology in NSW. The data have some
mean number of admissions in December over the limitations, however. First, we were unable to de-
period was 0n37 (95 % CI 0n091n04). The probability termine the epidemiology of non-admitted cases.
of one or more admissions in the central west in Second, hospital medical records are often incomplete
December was then 0n309 (0n0860n647). or unclear [20], possibly explaining why the specific
The hospital admission data did not reveal a type of acute glomerulonephritis was not coded for
tendency to cluster in time and space. Over the 9 almost one-third of the admissions. We therefore had
years, there were 18 pairs of patients living in the same to analyse all admissions coded as acute glomerulo-
Statistical Local Area and admitted within 31 days of nephritis as a proxy for APSGN. However, pro-
each other, compared with 19n4 expected pairs using liferative glomerulonephritis, which includes APSGN,
Knoxs method. The probability of 18 or more pairs accounted for two-thirds of the admissions. Finally,
occurring was 0n66. the indigenous status of admitted patients is almost
certainly under-reported in NSW [21].
Attack rates of glomerulonephritis following Group
A streptococcal infection vary by the ability of a
DISCUSSION
particular strain of the bacteria to cause glomerulo-
We identified two clusters of acute glomerulonephritis nephritis (nephritogenic). Attack rates from nephrito-
occurring around the same time but in geographically genic strains of 543 % have been reported, depending
distinct parts of NSW. The first, in late 1999, occurred on strain, age and site of infection [22]. A population-
in a small rural town and involved three unrelated based study reported an attack rate of 0n7 % following
children of Caucasian origin with no apparent streptococcal pyoderma in African-American children
common exposure. The second, in early 2000, oc- aged 26 years living in rural United States of America
curred in an inner Sydney suburb, and involved three [23]. In the first half of the twentieth century, when
related children of Polynesian origin who played APSGN was more common in developed countries,
together regularly and who therefore were likely to outbreaks of the disease demonstrated a variety of
have a high chance of a common exposure. A fourth, epidemiological manifestations : sporadic instances,
Caucasian, child in the same suburb was also affected, geographically localized clusters, clusters within
but we could find no common exposure. families and widespread epidemics [24]. The pro-
Analysis of hospital admission data for NSW for nounced variability in attack rates and in the
the 9 years to June 1998 revealed that admissions for prominence of particular strains of Group A
acute glomerulonephritis occurred more commonly in streptococci over time [25] highlight the unpredictable
boys, in Aboriginal children, and around the age of nature of these infectious agents and the potential for
7 years. There was a declining trend in admissions in further clusters of APSGN to occur in future in NSW.
each year since a peak in 1992\3 and no strong At present, prophylactic antibiotic use is the only
seasonal pattern was evident, contrary to the New available means of controlling Group A streptococcal
Zealand experience where a distinct seasonal peak in infection. Whether antibiotic treatment is protective
admissions occurs between April and June [19]. Based against progression to glomerulonephritis following
on expected counts of admissions in the areas affected, streptococcal infection is unclear, with contradictory
we were able to determine that the recent cluster of results reported in the literature [26, 27]. However, a
four patients with APSGN from one suburb in Sydney recent systematic review and meta-analysis of trials of
was a highly unusual event. With only one of the antibiotic treatment following sore throat showed a
recent rural cases being admitted, we could not trend towards being protective against glomerulo-
determine from the hospital admissions data whether nephritis [28].
that cluster was unusual. However, analysis of state- The Northern Territory of Australia has evidence-
wide admissions data did not reveal any tendency for based [29] recommendations for community-wide
acute glomerulonephritis to occur in clusters during prophylactic antibiotic use when clusters of APSGN
198998. occur in isolated Aboriginal communities [8], a setting
Streptococcal glomerulonephritis in NSW 371
prone to epidemics probably due to high background Aboriginal children. J Paediatr Child Health 1995 ; 31 :
rates of streptococcal skin infections. Should prophy- 2458.
lactic antibiotics be recommended to prevent the 5. Infectious Diseases and Immunization Committee,
Canadian Paediatric Society. Group A streptococcus : a
spread of APSGN when clusters occur in modern
re-emergent pathogen. CMAJ 1993 ; 148 : 190911.
developed urban and rural settings ? Unlike the far 6. Couser WG. Glomerulonephritis. Lancet 1999 ; 353 :
north of Australia, clusters of more than two cases of 150915.
APSGN are rare in both urban and rural NSW. 7. Hricik DE, Chung-Park M, Sedor JR. Medical prog-
Furthermore, there is no evidence that further ress : glomerulonephritis. N Engl J Med 1998 ; 339 :
intervention in relation to the clusters we described 88899.
8. Territory Health Services (Australia). Guidelines for
would have affected further spread of APSGN. We
the control of acute poststreptococcal glomerulo-
therefore make the following recommendations for nephritis. Casuarina : Territory Health Services, 1997.
the public health response to clusters of APSGN in 9. Epidemiology and Surveillance Branch. NSW Inpatient
modern developed settings such as NSW : household Statistics Collection (HOIST) [computer file]. Sydney :
contacts with overt throat or skin infection should be NSW Health Department, 2000.
treated with oral antibiotics (generally penicillin) ; 10. National Coding Centre. The Australian version of The
International Classification of Disease, 9th Revision,
families of cases should be educated to seek medical
Clinical Modification (ICD-9-CM). Sydney : National
advice if signs or symptoms of throat or skin infection Coding Centre, 1996.
or acute glomerulonephritis occur ; clinicians should 11. Zhang Y, Shen Y, Feld LG, et al. Changing patterns of
report clusters to Public Health Units who can glomerular disease at Beijing Childrens Hospital. Clin
investigate for common exposures to exclude possible Pediatr 1994 ; 33 : 5427.
point sources of infection. Where ongoing risk is 12. Yap H-K, Chia K-S, Murugasu B, et al. Acute
identified, additional control measures may be glomerulonephritis changing patterns in Singapore
children. Pediatr Nephrol 1990 ; 4 : 4824.
required.
13. Rosenberg HG, Vial SU, Pomeroy J, et al. Acute
glomerulonephritis in children : an evolutive morpho-
logic and immunologic study of the glomerular inflam-
A C K N O W L E D G E M E N TS mation. Path Res Pract 1985 ; 180 : 63343.
14. Epidemiology and Surveillance Branch. HOIST Docu-
The authors would like to thank the members of the mentation. Sydney : NSW Health Department, 2000.
NSW Infectious Disease Advisory Committee for 15. Australian Bureau of Statistics. 1996 census dictionary.
ABS catalogue no. 2901.0. Canberra : Australian Bu-
their assistance in developing the recommendations.
reau of Statistics, 1996.
We would also like to thank staff of the Institute for 16. Armitage P, Berry G. Statistical methods in medical
Clinical Pathology and Microbiological Research, the research, 3rd ed. Oxford : Blackwell Scientific Publi-
Royal Prince Alfred Hospital, Concord Repatriation cations, 1994.
Hospital, Sydney Childrens Hospital and the New 17. SAS [computer program]. Version 6.12. Cary (NC) :
Childrens Hospital pathology laboratories for their SAS Institute Inc., 1997.
assistance. Dr Fay Johnston of Territory Health 18. Australian Bureau of Statistics. Estimated resident
populations (HOIST) [computer file]. Sydney : NSW
Services provided valuable advice.
Health Department, 2000.
19. Meekin GE, Martin DR. Autumn the season for
poststreptococcal acute glomerulonephritis in New
REFERENCES Zealand. NZ Med J 1984 ; 97 : 2269.
20. Wilson RM, Runciman WB, Gibberd RW, et al. The
1. Markowitz M. Changing epidemiology of Group A quality in Australian health care study. Med J Aust
streptococcal infections. Pediatr Infect Dis J 1994 ; 13 : 1995 ; 163 : 45871.
55760. 21. Public Health Division. The health of the people of
2. Gray BM. Streptococcal infections. In : Evans AS, New South Wales report of the Chief Health Officer.
Brachman PS, eds. Bacterial infections of humans : Sydney : NSW Health Department, 1997.
epidemiology and control. New York : Plenum Medical, 22. Nissenson AR, Baraff LR, Fine RN, et al. Post-
1998. streptococcal acute glomerulonephritis : fact and
3. Carapetis JR, Wolff DR, Currie BJ. Acute rheumatic controversy. Ann Intern Med 1979 ; 91 : 7686.
fever and rheumatic heart disease in the Top End of 23. Nelson KE, Bisno AL, Waytz P, et al. The epidemiology
Australias Northern Territory. MJA 1996 ; 164 : 1469. and natural history of streptococcal pyoderma : an
4. Streeton CL, Hanna JN, Messer RD, et al. An epidemic endemic disease of the rural southern United States.
of acute poststreptococcal glomerulonephritis among Am J Epidemiol 1976 ; 103 : 27083.
372 D. J. Muscatello and others
24. Rammelkamp CH, Weaver RS. Acute glomerulo- 27. Lasch EE, Frankel V, Vardy PA, et al. Epidemic
nephritis : the significance of the variations in the glomerulonephritis in Israel. J Infect Dis 1971 ; 124 :
incidence of the disease. J Clin Invest 1953 ; 32 : 34558. 1417.
25. Colman G, Tanna A, Efstratiou A, et al. The serotypes 28. Del Mar CB, Glasziou PP, Spinks AB. Antibiotics for
of Streptococcus pyogenes present in Britain during sore throat (Cochrane Review). In : The Cochrane
19801990 and their association with disease. J Med Library [computer file], Issue 1, 2000. Oxford : Update
Microbiol 1993 ; 39 : 16578. Software, 2000.
26. Stetson CA, Rammelkamp CH, Krause RM, et al. 29. Johnston F, Carapetis J, Patel MS, et al. Evaluating the
Epidemic acute nephritis : studies on etiology, natural use of penicillin to control outbreaks of acute post-
history and prevention. Medicine (Baltimore) 1995 ; 34 : streptococcal glomerulonephritis. Pediatr Infect Dis J
43150. 1999 ; 18 : 32732.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.