Professional Documents
Culture Documents
Formulae used in manuscript IBW (male) = 50 kg + 0.9 kg for each cm > 150 cm in
height
Measured creatinine clearance (mL/min)
IBW (female) = 45 kg + 0.9 kg for each cm > 150 cm in
urine creatinine concentration × volume
= height
plasma creatinine con
ncentration
Creatinine clearance (mL /min) = 186 ¥ (Scr/88.4)-1.154 ¥ age-0.203 ¥ 0.742 if female ¥ 1.21
[140 − age (years)] × weight (kg) if African American
= (×0.85 if female)
0.814 × plasma creatinine (µmol / L)
‘Unadjusted’ eGFR (mL/min)
Cockroft Gault adjusted for ideal body weight
(IBW)45,46 = eGFR ¥ BSA/1.73
Results
Funding: None. One hundred and twenty-three patients were identified
Conflict of interest: None. with a diagnosis of NF or Fournier’s gangrene. Eighty-
Table 1 Patient characteristics at initial assessment Twenty-two (27%) patients reported non-steroidal
Patients with necrotizing anti-inflammatory drug (NSAID) use, most commonly
fasciitis (n = 82) diclofenac (12 patients, 15%). Forty-one per cent of the
n (%) group known to be on NSAIDs at admission were febrile
Male 44 (54)
compared with 50% of the patients that were not known
Ethnicity to be on NSAIDs, but this was not statistically significant
Pacific Islander 33 (40) (P = 0.47). Twenty-five (30%) patients had a previous
European 32 (39) diagnosis of gout. Twenty-nine (35%) had known dia-
Maori 15 (18) betes mellitus, and 18 (62%) of those were on medical
Clinical features treatment, including nine (31%) patients who were on
Pain 73 (89)
regular insulin treatment. No intravenous drug users
Swelling 71 (87)
Blue or purple skin discolouration 36 (44)
were identified in this cohort.
Pyrexia (>37.5°C) 36 (44) The mean white blood cell count at presentation was
Ulceration 31 (38) 20.5 ¥ 109/L (SD 11.6) and 58 (71%) had a glucose of
Blistering 18 (22) >6.0 mmol/L at admission. Renal impairment at presen-
Rigours 10 (12) tation was common with 51 (62%) patients having an
Comorbidities: elevated creatinine (>0.12 mmol/L).
Diabetes mellitus 29 (35)
A number of patients spoke only limited English and
Gout 25 (30)
Chronic respiratory disease 18 (22)
14 patients (17%) required an interpreter. Ten (12%)
Congestive heart failure 17 (21) patients did not have a recorded telephone number and
Obesity 16 (20) six (7%) patients lived alone.
Smoker 14 (17)
Chronic renal impairment 11 (13)
Ischaemic heart disease 10 (12) Microbiology
Sixty-one (74%) patients had blood cultures completed
within 72 h of admission and only 15 (18%) of these
two patients fulfilled inclusion criteria and were evalu- grew a pathogen (Table 2). None of these patients had
ated. Of the 41 patients excluded: 28 were transferred received antibiotics prior to blood cultures being
from another institution (Middlemore Hospital is a ter- obtained. Streptococcus pyogenes was isolated from tissue or
tiary plastics referral centre) and 13 did not fulfil our blood cultures in 33 (40%) patients, and S. aureus
definition of NF and likely had only superficial infection. (including MRSA) was isolated in seven (9%) patients;
The number of patients presenting with NF each year 26 (32%) patients had polymicrobial infection, although
ranged from five to 21 patients. these were not reflected in the blood cultures. Five (6%)
patients had MRSA isolated; one from blood cultures and
Initial assessment four patients from tissue specimens. Four of the MRSA
isolates were only resistant to b-lactam antibiotics con-
The mean age at presentation was 54.9 years (SD 18.5). sistent with community-acquired MRSA. Group G strep-
Most patients presented with symptoms of pain and tococcus was isolated in four patients, including two
swelling. Patient demographics, presenting symptoms patients with positive blood cultures. The bacterial cause
and comorbidities are outlined in Table 1. Forty-four was unknown in eight (10%) patients, five of whom died
(54%) patients were initially assessed by a general prac-
titioner prior to admission, but only two (5%) referral
letters included NF as a possible diagnosis. Fifteen (18%)
patients received antibiotics prior to admission and one of Table 2 Isolates from blood cultures completed within 72 h of admission
these patients was on community intravenous antibiotic Blood culture isolate Number of isolates (%)
treatment. Twenty-six (32%) patients had an initial (n = 15)
Australasian triage score of 4 or 5 at presentation.16 Streptococcus pyogenes 7 (47)
The site of infection was the lower limb in 46 (56%) Group G streptococcus 2 (13)
patients, upper limb in 12 (15%) patients, perineum in Staphylococcus aureus 2 (13)
13 (16%), abdomen in 6 (7%), thorax in 4 (5%) and face Serratia marcescens 1 (7)
in 1 patient. Thirty-four (41%) patients reported either Clostridium spp. 1 (7)
a surgical procedure or trauma immediately prior to Fusobacterium spp. 1 (7)
Vibrio vulnificus 1 (7)
presentation.
shortly after admission. Of these five only one had patients had an initial surgical procedure more than 24 h
surgery. after admission. Six of 14 (43%) patients who had sur-
gical intervention between 24 and 48 h after admission
died. Triage scores were recorded in 18/23 (78%) of the
Treatment and follow up
patients who had surgery delayed by 24 h or more, and
Forty-nine (60%) patients had a surgical procedure all patients had an admission Australasian triage score of
within 24 h of admission. The mean length of acute hos- 3 or 4 except for one patient whose score was 2.
pital stay was 24.5 days (SD 24.6, range 1–109) and 43 Age, chronic renal impairment, ischaemic heart
(52%) patients were admitted to intensive care. Only disease, congestive cardiac failure, gout and late surgical
seven (9%) patients required amputation of a limb, but intervention were all associated with mortality (Table 3).
most patients required extensive debridement. The Independent predictors of mortality on stepwise multi-
median number of surgical procedures per patient was variable logistic regression were congestive heart failure
two (range 0–15). (odds ratio (OR) 4.84; 95% confidence interval (CI) 1.13,
A wide range of antibiotic combinations was used as 20.63; P = 0.033) and gout (OR 4.08; 95% CI 1.09, 15.33;
initial therapy. Nineteen (23%) patients received clinda- P = 0.037). None of the patients with MRSA isolated from
mycin and ciprofloxacin ⫾ a beta-lactam as first-line cultures died.
therapy. Only three (4%) patients received a carbapenem
as part of their initial medical treatment. Fifty-six (68%)
Discussion
patients were treated with clindamycin in combination
with other antimicrobials. NF remains an important cause of infection-related mor-
Twenty-five (30%) patients died. The median time bidity and mortality in our community. Pacific Islanders
from admission to death was 2 days (range 0–63). Seven- are disproportionately represented, making up 40% of
teen (64%) patients died within 72 h of admission. Ten our cohort compared with only 22% of all admissions to
(12%) deaths occurred in patients who did not have Middlemore Hospital for the duration of the study. The
surgical debridement. Seven (9%) of these patients had 2001 New Zealand census data identified that 19% of the
significant comorbidities precluding surgical intervention, residents of the Middlemore Hospital area were Pacific
including two (2%) patients with severe chronic obstruc- Islanders.
tive pulmonary disease, two (2%) with severe congestive Pacific Islanders do not have an overall increased inci-
cardiac failure, two (2%) with advanced dementia and dence of admission for cellulitis to Middlemore Hospital
one (1%) with advanced myelofibrosis. Two of the (M Nisbet and D Holland, unpublished data, 2006), but
patients who died had severe irreversible septic shock and we found that they had an increased incidence of severe
died shortly after admission to hospital. One patient infection, although this did not translate into an increased
declined surgery and died 4 days after admission. risk of death. There may be a number of confounding
Eight of 49 (16%) patients who had a surgical proce- factors, such as delayed presentation, impaired commu-
dure within 24 h of admission died. Twenty-three (28%) nication (including need for interpreters and no ready
access to a telephone) and comorbidities, such as diabetes NF complicating gout has been previously observed,
mellitus and renal impairment. Diabetes mellitus (both and more recently Yu et al. described 15 patients with
previously diagnosed and undiagnosed disease) is likely to gout and NF.19 In that series three patients (20%) died, all
be an important contributing risk factor in this popula- from septic shock.19 Gout is a common and difficult
tion. The New Zealand Health Survey (2002/2003) found problem in South Auckland, but its link to NF in our
that 10.1% of Pacific Islanders in New Zealand had dia- community has not been studied previously.20 Although
betes compared with only 2.9% of the European popula- gout was an independent predictor of mortality there
tion.17 Compared with Europeans aged >40 years, the may be a number of confounding factors, including undi-
prevalence of undiagnosed diabetes is more than fourfold agnosed diabetes, use of NSAIDs, which may have been
among Pacific Islanders.18 underreported and delayed presentation because of
Many patients presented with classical symptoms of incorrect attribution of symptoms to gout. The impor-
pain and swelling, but only 60% of patients had a surgical tance of gout as a marker for mortality in this cohort
procedure within 24 h of admission. Twelve per cent warrants further awareness and investigation.
either declined surgery or were not considered as surgical
candidates. This suggests that the diagnosis of NF was
uncertain at admission for a significant number of Conclusion
patients resulting in delayed intervention and increased
NF remains a difficult condition to diagnose and has a
mortality. Fewer than 50% were pyrexial at admission,
high mortality in our community. Early diagnosis of NF
which may have been influenced by prior treatment with
can be difficult and requires a high index of suspicion in
NSAIDs. A lower percentage of patients on NSAIDs at
all patients presenting with cellulitis or unexplained
admission were afebrile when compared with patients
sepsis. Congestive heart failure and gout were indepen-
that were not on NSAIDs. NSAID use was identified in
dent predictors of mortality in our cohort, and patients
27% of our cohort, perhaps an underestimate in this
with these conditions and sepsis especially need early
retrospective review. Over 30% of patients had a triage
assessment, aggressive triaging and rapid intervention.
score of 4 or 5, suggesting haemodynamic stability at
presentation. These findings emphasize the importance
of considering NF even though clinical signs may be
Acknowledgements
minimal, and also emphasize that overt shock and insta-
bility are often late and frequently a preterminal state. The authors gratefully acknowledge the assistance of
MRSA was isolated from five patients and all of these Katrina Poppe (Biostatistician, UniServices, University of
received early treatment with clindamycin to which their Auckland), Paula Wallis, Karen Gunson and Rebecca
strains were susceptible. The possibility of MRSA must be Findlay (Outpatient Intravenous Antibiotic Service,
considered when deciding appropriate antibiotic therapy Middlemore Hospital), Stephen Park and Andrea Rigby
prior to culture availability, especially in communities (Clinical Coding Department, Middlemore Hospital) and
where MRSA is relatively common. the Department of Surgery (Middlemore Hospital).
11 Laucks S. Fournier’s gangrene. Surg Clin APACHE II score, dissemination, and sur- Health Intelligence. Charting Our Health.
North Am 1994; 74: 1339–52. vival. World J Surg 2007; 31: 1858–62. Wellington: Ministry of Health; 2004.
12 O’Louglin R, Roberson A, Cieslak P, 15 Tiu A, Martin R, Vanniasingham P, 18 Simmons D, Thompson C. Prevalence of
Lynfield R, Gershman K, Graig A et al. MacCormick A, Hill A. Necrotizing the metabolic syndrome among adult
The epidemiology of invasive group A fasciitis: analysis of 48 cases in South New Zealanders of Polynesian and
streptococcal infection and potential Auckland, New Zealand. ANZ J Surg European descent. Diabetes Care 2004;
vaccine implications: United States, 2005; 75: 32–4. 27: 3002–4.
2000–2004. Clin Infect Dis 2007; 45: 16 ACEM. The Australian National Triage 19 Yu K, Ho H, Chen J, Luo S. Gout
853–62. Scale: A User Manual. Melbourne: complicated with necrotizing fasciitis –
13 Tillou A, Hill C, Brown C, Velmahos G. Australian Government Publishing report of 15 cases. Rheumatology 2004;
Necrotizing soft tissue infections: Service: Commonwealth Department of 43: 518–21.
improved outcomes with modern care. Health and Family Services (DoH & FS); 20 Stamp L, Gow P, Sharples K, Raill B.
Amer Surg 2004; 70: 841–4. 1997. The optimal use of allopurinol: an audit
14 Yilmazlar T, Ozturk E, Alsoy A, Ozguc H. 17 Ministry of Health. 2002/2003 New of allopurinol use in South Auckland.
Necrotizing soft tissue infections: Zealand Health Survey Datacubes – Public Aust N Z Med J 2000; 30: 567–72.