Scottish Medical Journal

Three years experience of adults admitted to hospital with E.coli o157 in North-East Scotland

SMJ 2002; 47(5): 110-112

A.M. Cadwgan*, R.B.S. Laing, L. Dargie, M. Beadsworth, A.R. Mackenzie, J.G. Douglas.

Infection Unit, Aberdeen Royal Infirmary, Aberdeen.

Key Words:   E. coli O157, Haemolytic Uraemic Syndrome, Gastroenteritis.

*Author For Correspondence

The Infection Unit

Aberdeen Royal Infirmary

Foresterhill

Aberdeen

AB25 2ZN

tony.cadwgan@arh.grampian.scot.nhs.uk


ABSTRACT

Aims: To describe the epidemiology, clinical features, treatment and outcomes of adults with E. coli O157 infection presenting to Aberdeen Royal Infirmary over a three year period.

 

Methods: A retrospective casenote review.

 

Results: Thirty-two confirmed cases of E. coli O157 infection were admitted between 1997 and 2000. The median age was 58 years (range 16-93). Ten patients (31%) were from the city of Aberdeen and 22 (69%) from surrounding rural areas. Twenty-seven patients (85%) presented between May and October. The source of infection was unknown or unconfirmed in all cases. Bloody diarrhoea was present in 30 (94%). Leucocytosis was present in 18 (63%) but only four patients (13%) had a fever. Six of the 32 patients (19%) developed Haemolytic-Uraemic Syndrome (HUS) of whom 2 died. Ten patients received antibiotics of whom two developed HUS. Twenty-seven of the 32 (85%) had made a full recovery by time of discharge, three (9%) had impaired renal function and two (6%) died in hospital.

 

Conclusion: E. coli O157 infection tends to occur sporadically in rural areas in North East Scotland. It is not usually associated with fever. Infection occurs more commonly in the summer and autumn. HUS complicates infection in almost one fifth of patients.

 

INTRODUCTION

First recognised as a human pathogen in 19821-4, human infection with E. coli O157 has now been reported in over 30 countries - most commonly the United Kingdom, Canada and the United States5. Within the U.K., Scotland has consistently had the highest incidence since 19876. In a recent report from Grampian 16% of patients admitted to the Regional Infection Unit with proven bacterial gastroenteritis had E. coli O157 isolated7.

 

As well as consumption of contaminated food, transmission of E. coli O157 is known to be associated with contaminated water supplies and direct person-to-person contact5. Contact with animals or farms has been shown to be a highly significant risk factor for infection8.

 

In light of the significant number of cases of E. coli O157 being admitted to the Regional Infection Unit and other adult wards in Aberdeen Royal Infirmary, we decided to review of our experience of this condition.

 

METHODS

Information was collected from a retrospective case note review of all adult patients (16 years and over) admitted to all wards in Aberdeen Royal Infirmary between 1997 and 2000 who were microbiologically confirmed as having E. coli O157 infection. Fever was defined as a temperature >37.5C on at least one occasion. Haemolytic Uraemic Syndrome was said to have developed if the patient had evidence of red cell haemolysis, thrombocytopenia and acute renal impairment.

 

RESULTS

Thirty-two confirmed cases of E. coli O157 infection were admitted to hospital over the three year study period from January 1997 to December  2000. Ten patients (31%) were from Aberdeen City and 22 (69%) from surrounding rural areas. Nineteen were admitted to the Regional Infection unit, seven to the Gastroenterology Unit, two to General Medical and three to General Surgical wards.

 

The median age of the cases was 48 years and the range was 16 to 93. Sex ratio was 17 male and 15 female cases. The duration of admission ranged from 2 to 44 days with a median of 5 days. A seasonal variation was evident with 27 patients (85%) presenting between May and October.

 

The source of infection was unknown or unconfirmed in all cases. Twenty-six of the 32 had no contacts with E. coli O157 infection. Three patients had family contacts with E. coli O157 infection and one each had a contact in school, at work or with a friend. Four out of the 32 (13%) were farm workers.

 

The clinical and laboratory features on admission are shown in Table I and Table II. Six of the 32 patients (19%) developed HUS, of whom two died. The age of those who developed HUS ranged from 16 to 70 years with a median age of 41. Of the six who developed HUS four (66%) had a co-morbid illness: two hypertension, one cerebrovascular disease, one epilepsy. All six patients had a raised white cell count on admission. Thirteen of the 26 patients (50%) who did not develop HUS had a co-morbid illness. Comparison between those with and without HUS is made in Table 3.

 

The time from admission to development of HUS ranged from 0 to 5 days with a median of 0.5 days. Time from onset of symptoms to HUS ranged from 4 to 8 days, with a median of 8 days. Three patients received haemodialysis; three patients received therapeutic plasma exchange (TPE); two patients received both dialysis and TPE, both of whom died.

 

Ten patients were treated with antimicrobials - oral or intravenous ciprofloxacin in six patients and intravenous cefotaxime plus metronidazole in four. With one exception all these anti-microbials were started in other units for presumed intra abdominal sepsis prior to confirmation of E.coli O157 infection. Two of the patients who received antimicrobials (20%) developed HUS.

 

Four of those with HUS developed neurological features, one transient limbweakness, one seizures and two intracerebral haemorrhage both of whom died. Twenty-seven patients had made a full recovery by time of discharge, three had impaired renal function on discharge and two died in hospital. 29 out of the remaining 30 patients had made a complete recovery at six months leaving one impaired renal function.

 

DISCUSSION

Awareness of the pathogenic potential of E. coli O157 in humans was increased by the Lanarkshire outbreak of 19969. In 1998, the incidence of infection was estimated at 4.2 cases per 100,00 in Scotland and 1.7 per 100,000 in England and Wales8. In this three year study, only adults who were sufficiently unwell to warrant hospital admission were included. None of the cases occurred as part of an outbreak.

 

 Seasonal variation was identified with most patients presenting in the warmer months, between May and October. It might be speculated that this is due to more people taking part in outdoor/rural pursuits during these months and there is also evidence that faecal shedding of E. coli O157 by cattle increases in the summer10. A Scottish study of patients under 15 years of age found the rate of infection was highest between June and September11 but seasonal variation in adults has not been previously reported in the region.

 

Most cases of  E. coli O157 infection (69%) arose in patients from outwith Aberdeen in rural areas. In such areas the farming of cattle and sheep is a main industry and many houses receive a private water supply. In an outbreak in the Grampian village of Tarves, the local drinking water supply was implicated12. In this study, 4 out of 32 patients (13%) worked in the farming industry. By comparison, less than 4% of the Grampian population works in agriculture13.

 

Attempts to look for a food source of infection were not undertaken in any of the cases reported here. This is usually performed in the outbreak setting.

 

The clinical features described in this study are comparable to those previously published14,15. Most patients had diarrhoea initially with stools becoming bloody one or two days later. Abdominal pain and tenderness were prominent features. Leucocytosis was present in the majority of admission bloods. Fever was not commonly reported, this contrasts with the majority of our patients with bacterial gastroenteritis in whom fever is present on admission in 65%7.

 

In a large American study, clinical signs and symptoms independently associated with E. coli O157 compared with campylobacter, salmonella or shigella included reported bloody diarrhoea, visibly bloody stool samples, a peripheral leucocyte count greater than 10 and abdominal tenderness. At least three of these features were present in 65.4% of patients with E. coli O157 infection in contrast they were present in 18.5 % of those who had infection with campylobacter, salmonella or shigella15. Our experience would suggest that the absence of fever should be added to those aforementioned criteria in deciding the likelihood of E. coli O157 infection.

 

Ten out of the 32 patients were treated with antibiotics on initial presentation. A body of literature has, for several years, supported an association between use of antibiotics and progression to HUS16. A prospective cohort study has confirmed that giving antibiotics to children with E.coli O157 infection increases their risk of developing HUS17. No similar study exists in adults. In this review of 32 cases we found no association between antibiotic use and HUS but the small numbers may be relevant to this negative finding.

 

Three patients in this study were treated with Therapeutic Plasma Exchange. Although this form of therapy has not been proven to be effective in randomised trials, a general consensus supports it's use in adults with idiopathic HUS18. Experience from the 1996 Lanarkshire outbreak suggested that treatment was promising in patients with HUS secondary to E. coli O157 infection9.

 

Age less than four years or more than 65 years is perceived as a risk factor for the development of HUS subsequent to infection with E. coli O157 infection19. The median age of those who developed HUS in this study was 41 years, all were adults  and only three were aged over 65 years. Although the numbers are small, these patients affected by HUS tended to be younger than described in previous studies and may suggest that serious the clinical sequelae associated with this infection are not confined to the extremes of age.

 

REFERENCES

1. Riley L.W., Remis R.S., Helgerson S.D., et al

    Haemorrhagic colitis associated with a rare Escherichia coli serotype

     New England Journal of Medicine 1983; 308:681-85

 

2. Karmali M.A., Steele B.T., Petric M., et al

   Sporadic cases of haemolytic uraemic syndrome associated with faecal

    cytoxin-producing Escherichia coli in stools

    Lancet 1983;I:619-20

 

3. Karmali M.A., Petric M., Lim C., et al

    The association between idiopathic haemolytic uraemic syndrome and infection

      by verotoxin producing Escherichia coli

     Journal of Infectious Disease 1985;151:775-82

 

4. Bitzam M., Ludwig K., Klemt M., et al

    The role of Escherichia coli O157:H7 infections in the classical (enteropathic)

     haemolytic uremic syndrome : results of a central european, multicentre trial.

     Epidemiology and Infection 1993;110:183-96

 

5. Griffin P.M., Mead P.S.

    Escherichia coli O157

     Lancet 1998;352:1207-1212

 

6. Advisory Committee on the Microbiological Safety of Food : Report on

    verocytotoxin-producing Escherichia coli.

    Her Majesty's Stationery Office Publications, London (1995) pp.19-20

 

7. Cadwgan A.M., Watson W.A., Laing R.B.S., et.al.

      Presenting clinical features and C-reactive protein in the prediction of a positive

      Stool culture in patients with diarrhoea.

      Journal of Infection 2000:41;159-161

 

8.      Locking M.

A case-control study of sporadic cases of Escherischia coli O157 infection in Scotland.

SCIEH Weekly Report 2000;34:9-10

 

9     Dundas S., Murphy J., Souter R.L., et al

       Effectiveness of therapeutic plasma exchange in the 1996 Lanarkshire

        Escherichia coli O157:H7 outbreak

        Lancet 1999;354: 1327-1330

 

10. Ostroff S.M., Kobayashi J.M., Lewis J.H., et al

      Infections with Escherichia coli O157:H7 in Washington State: the first year of

       statewide disease surveillance

       JAMA 1989;262:355-9

 

11. Douglas A.S., Kurien A.

     Seasonality and other epidemiological features of haemolytic uraemic syndrome

     in Scotland"

      Scottish Medical Journal 1997; 42:166-171

 

12.  Dev V.J., Main M., Gould I.

       Waterborne outbreak of Escherichia coli O157

       Lancet 1991;337:1412

 

13. MacDonald I.A.R., Gould I.M., Curnow J.

     Epidemiology of infection due to Escherichia coli O157: a 3-year prospective

      study

     Epidemiology and Infection 1996;116:279-284

 

14. Borczyk A.A., Carlson A.O., Carlson A.K., et al

      A severe outbreak of Escherichia coli O157:H7 - associated colitis in a Nursing

      Home

      New England Journal of Medicine 1987;317:1496-500

 

15. Sltusker L., Ries A.A., Greene K.D., et al

      Escherichia coli O157:H7 Diarrhoea in the United States: clinical and

       epidemiologic features

       Annals of Internal Medicine 1997;136:505-513

 

16. Boyce T.G., Swerdlow D.C., Griffin P.M.

    Escherichia coli O157:H7 and the Haemolytic Uremic Syndrome

     New England Journal of Medicine 1995;336 :364-368

 

17. Wong C.S., Jelacic S., Habeeb R.L., et al

      The risk of hemolytic-uremic syndrome after antibiotic treatment of Escherichia coli O157 : H7 infections.

       New England Journal of Medicine 2000;342:1930-6

 

18. Ruggenenti P, Lutz J, Remuzzi G.

      Pathogenesis and treatment of thrombotic microangiopathy.

      Kidney Int 1997;51:S97-101

 

19. Dundas S, Todd WT, Stewart AI, Murdoch PS, Chaudhuri AK, Hutchinson SJ.

      The central Scotland E.coli O157:H7 outbreak: risk factors for the development of haemolytic uremic syndrome and death among hospitalized patients.

            Clinical Infectious Diseases 2001;33:923-31

 

 

Back to October Index