Case 12 Difficulty with diarrhoea

Mrs Brooks is an 84-year-old woman who is referred by her GP to the general medical take with weakness, nausea, malaise, and shortness of breath (SOB). She has been unable to cope at home because of the weakness and multiple co-morbidities which include rheumatoid arthritis, chronic renal impairment, gastro-oesophageal reflux, chronic heart failure and recurrent urinary sepsis. The current problems started three weeks earlier with urinary frequency and pyrexia for which she received a seven-day course of oral co-amoxiclav from her GP.

Her current medication is methotrexate 10 mg once a week, omeprazole 40 mg daily, furosemide 40 mg daily, lisinopril 7.5 mg daily, as well as folinic acid, co-codamol, senna and lactulose.

Dr Papadakis, the medical CT1 who first examines Mrs Brooks, finds her to be weak and frail. She is apyrexial and hypotensive with a systolic blood pressure of 95 mm Hg. She has a heart rate of 125 in atrial fibrillation. There are widespread crackles in both lung fields. The chest X-ray shows cardiomegaly and a small left pleural effusion with patchy consolidation at the right lung base. The initial blood results show she has a normochromic normocytic anaemia with a haemoglobin of 9.6 g/dl and a white blood count of 11.9. She has renal failure with a creatinine of 221 mmol/l (previously 155) and an albumin of 29 g/l. The CRP is 39. A diagnosis of heart failure and possible chest infection is made and she is commenced on intravenous cefuroxime and given an intravenous infusion of 1L of normal saline over 6 hours. She is given a loading dose of oral digoxin with a plan to continue with 62.5 micrograms daily thereafter.

Would you have managed the situation in the same way?

On the post take ward round 2 hours later she is somewhat improved and the consultant agrees with the management plan. She is reviewed the next day (a Friday) when she is still apyrexial, her systolic blood pressure is now 120 mmHg and her apical rate has slowed to 80. She complains of some abdominal discomfort and her dose of laxative is increased.

At handover that evening the FY1 doctor identifies Mrs Brooks as a patient who is unwell and who requires review over the weekend but as she is improving no change in her management plan is communicated to the cover team.

On the Sunday evening the nurse looking after Mrs Brooks is concerned that she is less communicative and is complaining of intermittent abdominal pain. Her temperature has risen to 38.5 °C and she is once again tachycardic. The CRP taken earlier that day has risen to 140 and the albumin has fallen further to 22. The cover FY1 comes to see Mrs Brooks and is worried by the temperature, believing she has pneumonia which is not responding to antibiotics. She adds in a macrolide antibiotic to the cephalosporin and arranges a plain abdominal film because Mrs Brooks by now has a very tender abdomen and has just had two episodes of diarrhoea.

The plain abdominal film shows some colonic dilatation but no further action is taken at this stage. The next morning Mrs Brooks is very unwell and has had persistent diarrhoea overnight. Her abdomen is distended and very tender. The nurse in charge reports that the diarrhoea is very suggestive of Clostridium difficile infection and a stool sample is sent which confirms C difficile toxin. The result is available the next day and oral vancomycin and intravenous steroids are started but Mrs Brooks rapidly deteriorates and dies on the Tuesday night.

At the team’s Mortality and Morbidity review meeting, the notes and clinical history are reassessed and the conclusion is reached that Mrs Brooks died of overwhelming Clostridium difficile colitis which had not been detected soon enough to allow adequate treatment. The cause of death was noted as 1a C difficile colitis, automatically triggering a Serious Incident Requiring Investigation (SIRI) process.

Expert opinion

Mrs Brooks has multiple risk factors for healthcare acquired infection and for C difficile in particular. She is elderly, immunosuppressed with methotrexate and has rheumatoid arthritis, she is receiving a proton pump inhibitor and she has renal impairment and heart failure. She has recently been treated with a seven-day course of co-amoxiclav and on admission to hospital was given a broad spectrum intravenous cephalosporin and later a macrolide antibiotic as well. The use of laxatives may also have been relevant.

The evidence that she had an infective problem on admission was not convincing and her symptoms were compatible with a combination of dehydration and renal impairment in the context of heart failure, anaemia, and atrial fibrillation. The use of cefuroxime was inadvisable at the outset and should have been questioned on the post take ward round. The abdominal pain, the pyrexia, the falling albumin, and the rise in the CRP are all features compatible with the onset of colitis.

C difficile infection is strongly associated with antibiotic usage and in this case the prescription of cefuroxime for an unproven pneumonia was very ill advised. Indeed, a seven-day course of co-amoxiclav for an uncomplicated UTI is excessive. This was further compounded by the failure to review the prescription after 24 hours and for it to continue over the weekend. Many hospitals have introduced automatic stop or review dates for intravenous antibiotics and even the simple act of writing the proposed duration of therapy on the drug chart might have empowered the cover team to have modified the prescription. The handover was inadequate – no patient should be left receiving intravenous or broad spectrum antibiotics over a weekend without being reviewed and careful plans put in place and communicated.

If C difficile is suspected, appropriate treatment with either vancomycin or metronidazole should commence before waiting for the results of stool sampling. Delays in starting treatment may adversely affect outcome.

A SIRI was undertaken which concluded that death was the result of C difficile colitis in a lady with significant risk factors. This case was one of several which occurred over a six-month period and which triggered a major review of antibiotic stewardship within the hospital, resulting in a change in antibiotic policy and a programme of education for all medical staff. Ward pharmacists were instructed to challenge prescriptions for intravenous antibiotics over 48h duration and all prescriptions of antibiotics were required to include the indication for, and duration of, antibiotic therapy. This has since been continuously audited and reported to the hospital board. The incidence of C difficile infection has fallen very significantly since these changes were instituted.

The Department of Health continues to closely manage rates of C difficile infection in the NHS in England. A heavy focus on reducing the number of cases identified may be counterproductive with C difficile as the detection of C toxin does not necessarily indicate disease in a patient with loose stool (rather incidental asymptomatic carriage) and these ‘targets’ may lead to a more conservative approach to testing for C difficile in general, potentially resulting in failure to diagnose in a timely fashion.

Legal comment

The patient’s death should also be notified to the coroner, as death was ‘unnatural’ (due to hospital acquired infection) and possibly neglectful if the C difficile colitis had not been detected soon enough to allow adequate treatment. Some of the terms in the reporting requirements of the Coroner’s Act 1988, supplemented by the Coroner’s Rules 1984, are imprecise, for example, ‘unnatural’ in the context of an elderly patient such as Mrs Brooks. When a coroner is informed of such a death, he/she has the discretion to dispense with an Inquest, with or without a post-mortem. Informal discussion with the coroner’s office will clarify whether or not a patient’s death should be reported.

When clinicians discuss a case ‘with the coroner’; a coroner’s officer usually serves as intermediary. The police previously employed coroner’s officers but they are now employees of the Local Authority. They are administrators and do not have legal responsibility. Coroner’s officers are usually the first point of contact in the coroner’s office for advice about whether a patient’s death should be formally notified to the coroner. The coroner’s officer will discuss with the coroner, if required on more complex matters, and where necessary will facilitate direct communication between the coroner and the referring doctor.

If the coroner does decide to open an inquest, the SIRI report and action plan should be shared with the coroner as evidence of organizational learning and changes that have been made to antibiotic stewardship in the Trust. This may provide sufficient assurance to coroner to avoid a Rule 43 letter of steps taken to improve patient safety. It is also good practice for the patient’s family to be given a copy of the SIRI report and invited to meet with appropriate clinicians/managers to discuss their concerns.

The SIRI report and action plan is also notified to the PCT and SHA. If this case is one of several in a cluster, then it is likely that this will have triggered closer monitoring under the Trust’s performance contract with the PCT of the Trust’s management of C difficile until audit results demonstrated an improvement.

All NHS Trusts also have to self-register with the Care Quality Commission to confirm the status of their compliance with infection control standards and are open to being inspected by the CQC as to the accuracy of the data provided. All NHS Trusts now require staff to undertake mandatory training in infection control to raise awareness of the potential for healthcare acquired infection.

Stay updated, free articles. Join our Telegram channel

Join